Spacelord
10-01-2009, 09:31 AM
http://www.mentalimages.com/products/iray
Iray Looks quite promising.
Iray Looks quite promising.
View Full Version : Mental Images I-Ray
|
|
AthorNZ
10-01-2009, 10:27 AM
Hope this is included in mentalray for maya
wurp
10-01-2009, 10:35 AM
it does actually look promising, but can't tell untill it's out ofcourse..
it states it's part of mr 3.8 and not a separate product, so that's good at least:
http://www.mentalimages.com/fileadmin/user_upload/PDF/iray_FAQ-090929.pdf
they render an almost noise free version of an interior in 120 seconds in the pdf, not bad, it's quite lowrez but still, not bad at all.
still doesn't have motionblur or programmable shaders, you can live without those things at least for arch viz, perhaps in the future they will at least add mblur.
it states it's part of mr 3.8 and not a separate product, so that's good at least:
http://www.mentalimages.com/fileadmin/user_upload/PDF/iray_FAQ-090929.pdf
they render an almost noise free version of an interior in 120 seconds in the pdf, not bad, it's quite lowrez but still, not bad at all.
still doesn't have motionblur or programmable shaders, you can live without those things at least for arch viz, perhaps in the future they will at least add mblur.
Kzin
10-01-2009, 11:40 AM
this is a thing i dont understand, they support dof but not mb, anyone knows why? whats the problem here?
Spacelord
10-01-2009, 11:45 AM
I wonder if Iray was used for these test renders on Kims blog.
If so maybe 3dsmax might have it in the next release :)
http://area.autodesk.com/blogs/ken/autodesk_on_vacation_did_our_research_team_earn_it
If so maybe 3dsmax might have it in the next release :)
http://area.autodesk.com/blogs/ken/autodesk_on_vacation_did_our_research_team_earn_it
CHRiTTeR
10-01-2009, 11:53 AM
I wonder if Iray was used for these test renders on Kims blog.
If so maybe 3dsmax might have it in the next release :)
http://area.autodesk.com/blogs/ken/autodesk_on_vacation_did_our_research_team_earn_it
hmmm the results arent identicall
If so maybe 3dsmax might have it in the next release :)
http://area.autodesk.com/blogs/ken/autodesk_on_vacation_did_our_research_team_earn_it
hmmm the results arent identicall
Glacierise
10-01-2009, 11:55 AM
For anybody who doesn't want to go through the full document, and for discussion purposes, pros and cons:
Up to you:
-unbiased
Pros:
-speed from gpu utilization, scalability
-quality
-is a mode of mental ray instead a new product
-iterative, but 'not interactive'
-support for well known shaders.lights like the mia_material/A&D, pro materials, spot/omni/directionals, photometrics, HDRIs
-can also run on CPU
-supports DOF
Cons:
-no support for non-NVIDIA GPUs
-probably inefficient on gaming cards, will require more memory, hence quadros
-it's not a preview for mental ray, i.e. - not as full featured as mr
-no motion blur, hence no animation
-no MetaSL support
Looks supercool, but will probably use it for shading/lighting design only, because of the lack of motion blur. Still, integration in mental ray is great.
Up to you:
-unbiased
Pros:
-speed from gpu utilization, scalability
-quality
-is a mode of mental ray instead a new product
-iterative, but 'not interactive'
-support for well known shaders.lights like the mia_material/A&D, pro materials, spot/omni/directionals, photometrics, HDRIs
-can also run on CPU
-supports DOF
Cons:
-no support for non-NVIDIA GPUs
-probably inefficient on gaming cards, will require more memory, hence quadros
-it's not a preview for mental ray, i.e. - not as full featured as mr
-no motion blur, hence no animation
-no MetaSL support
Looks supercool, but will probably use it for shading/lighting design only, because of the lack of motion blur. Still, integration in mental ray is great.
wurp
10-01-2009, 12:11 PM
motionblur requires one to store several frames of the same scene in memory etc, its way more complex than dof.
by the way you can see iray in action:
http://www.nvidia.com/object/gpu_technology_conference.html
click on the "view webcast" link down where it says "Opening Keynote with Jen-Hsun Huang "
One thing I noticed is that they run this on a pretty large cluster of boxes, not sure if the 120 second example mentioned in the pdf is running with a large cluster like that, if so it's not that impressive...
this is a thing i dont understand, they support dof but not mb, anyone knows why? whats the problem here?
by the way you can see iray in action:
http://www.nvidia.com/object/gpu_technology_conference.html
click on the "view webcast" link down where it says "Opening Keynote with Jen-Hsun Huang "
One thing I noticed is that they run this on a pretty large cluster of boxes, not sure if the 120 second example mentioned in the pdf is running with a large cluster like that, if so it's not that impressive...
this is a thing i dont understand, they support dof but not mb, anyone knows why? whats the problem here?
Kzin
10-01-2009, 12:15 PM
motionblur requires one to store several frames of the same scene in memory etc, its way more complex than dof.
by the way you can see iray in action:
http://www.nvidia.com/object/gpu_technology_conference.html
click on the "view webcast" link down where it says "Opening Keynote with Jen-Hsun Huang "
One thing I noticed is that they run this on a pretty large cluster of boxes, not sure if the 120 second example mentioned in the pdf is running with a large cluster like that, if so it's not that impressive...
mmh, ok, that makes things clearer. any informations about framebuffer rendering, like vector motion, such things?
in this video they talk about 15 gpu's (if i understood it correcly), the renderperformence is really good.
http://www.youtube.com/watch?v=TQsXualxLVs
by the way you can see iray in action:
http://www.nvidia.com/object/gpu_technology_conference.html
click on the "view webcast" link down where it says "Opening Keynote with Jen-Hsun Huang "
One thing I noticed is that they run this on a pretty large cluster of boxes, not sure if the 120 second example mentioned in the pdf is running with a large cluster like that, if so it's not that impressive...
mmh, ok, that makes things clearer. any informations about framebuffer rendering, like vector motion, such things?
in this video they talk about 15 gpu's (if i understood it correcly), the renderperformence is really good.
http://www.youtube.com/watch?v=TQsXualxLVs
CHRiTTeR
10-01-2009, 12:27 PM
considdering that demo is running on 15 gpu's im not really impressed.
I expected more, compared to how vrayrt (gpu edition) runs on 1 gaming gpu.
I expected more, compared to how vrayrt (gpu edition) runs on 1 gaming gpu.
davius
10-01-2009, 01:55 PM
Cons:
-no support for non-NVIDIA GPUs
-probably inefficient on gaming cards, will require more memory, hence quadros
HUGE letdown. Why is ATI/AMD being outcast from this party?
-no support for non-NVIDIA GPUs
-probably inefficient on gaming cards, will require more memory, hence quadros
HUGE letdown. Why is ATI/AMD being outcast from this party?
Cheesestraws
10-01-2009, 01:56 PM
Because mental images is owned by Nvidia, and of course CUDA is ready now. :wise:
davius
10-01-2009, 02:00 PM
Yeah! I forgot that...
jeremybirn
10-01-2009, 02:44 PM
Looks supercool, but will probably use it for shading/lighting design only, because of the lack of motion blur. Still, integration in mental ray is great.
I agree it's clearly targeted at arch/vis work, for now. But just to play devil's advocate: Film productions that render final elements in MR have tended to fake motion blur in post anyway. Other companies have baked GI solutions using MR, then rendered final frames with other software.
I don't know how good this is, or how it'll compare with the competition as the products mature, but if this (or one of its competitors) were really good, I could imagine people working them into film pipelines. So far I haven't seen proof that buying a lot of GPUs for rendering is smarter than just adding CPUs to a render farm, but I'm sure people will get whatever seems cost-effective.
-jeremy
I agree it's clearly targeted at arch/vis work, for now. But just to play devil's advocate: Film productions that render final elements in MR have tended to fake motion blur in post anyway. Other companies have baked GI solutions using MR, then rendered final frames with other software.
I don't know how good this is, or how it'll compare with the competition as the products mature, but if this (or one of its competitors) were really good, I could imagine people working them into film pipelines. So far I haven't seen proof that buying a lot of GPUs for rendering is smarter than just adding CPUs to a render farm, but I'm sure people will get whatever seems cost-effective.
-jeremy
Array
10-01-2009, 04:22 PM
Can't wait to test iRay on one of these!
http://techreport.com/articles.x/17670
:thumbsup:
http://techreport.com/articles.x/17670
:thumbsup:
bjoern
10-01-2009, 04:38 PM
looks likes crap against Vray RT (on a GPU)
Kzin
10-01-2009, 04:49 PM
looks likes crap against Vray RT (on a GPU)
you should compare the scene complexity. the mr demo is a complete different thing, glossy shader, many area lights, indoor gi, all stuff that i would like to see from vrayrt. so we should wait until we can test the two renderer. ;)
you should compare the scene complexity. the mr demo is a complete different thing, glossy shader, many area lights, indoor gi, all stuff that i would like to see from vrayrt. so we should wait until we can test the two renderer. ;)
DrBalthar
10-01-2009, 07:03 PM
Disappointing this looks even worse performance-wise than NVidia's OptiX demo. Not even close to what Caustic's latest demo showed off or even VRayRT(GPU demo) at Siggraph. I think Mental Images completely have lost it they seem to be so far behind the curve.
Fireantz
10-01-2009, 08:47 PM
Disappointing this looks even worse performance-wise than NVidia's OptiX demo. Not even close to what Caustic's latest demo showed off or even VRayRT(GPU demo) at Siggraph. I think Mental Images completely have lost it they seem to be so far behind the curve.
Isn't a bad idea that mental images release new engine. There is also nothing disappointing to keep comparing both of the renderers. Just think about it, we (artists) should appreciated them, because in the end...they will always make our life getting more easier from their effort ! :buttrock:
Isn't a bad idea that mental images release new engine. There is also nothing disappointing to keep comparing both of the renderers. Just think about it, we (artists) should appreciated them, because in the end...they will always make our life getting more easier from their effort ! :buttrock:
inkblack
10-01-2009, 09:16 PM
Isn't a bad idea that mental images release new engine. There is also nothing disappointing to keep comparing both of the renderers. Just think about it, we (artists) should appreciated them, because in the end...they will always make our life getting more easier from their effort ! :buttrock:
yeah i think you are right....i have watched the whole presentation of NIVDIA and i must say that i like it as much as the vray thing and the caustics demo....they are just completely different scenes with completely different light setups.
You really can only compare speed with the same scene and settings....
from what i saw just now, it seems that they are all equal, but i could be wrong of course since i only have seen 1 movie of each company/product.
But after all, its all good for us...as mentioned earlier...
man, what do i have to do at work when the waiting time is gone?? :)
cheers
Madthijs
yeah i think you are right....i have watched the whole presentation of NIVDIA and i must say that i like it as much as the vray thing and the caustics demo....they are just completely different scenes with completely different light setups.
You really can only compare speed with the same scene and settings....
from what i saw just now, it seems that they are all equal, but i could be wrong of course since i only have seen 1 movie of each company/product.
But after all, its all good for us...as mentioned earlier...
man, what do i have to do at work when the waiting time is gone?? :)
cheers
Madthijs
PiotrekM
10-01-2009, 09:19 PM
If its final quality renderer then for me its big thing...preview quality as vray RT is okay for lighting& material testing here
jasonio
10-01-2009, 09:41 PM
I'll pass judgement when I see the results from both renderers on a single GPU, quad core PC. Vray RT should be quicker, considering you have to fork out for it. iRay still looks like a great step forward for people using a renderer that comes included with their software.
Array
10-01-2009, 11:03 PM
I'll pass judgement when I see the results from both renderers on a single GPU, quad core PC. Vray RT should be quicker, considering you have to fork out for it. iRay still looks like a great step forward for people using a renderer that comes included with their software.
I don't think that the monetary value which you pay for a renderer is necessarily a good benchmark for how good said renderer is. MentalRay isn't free, its cost is plugged into the price of whichever 3d software that it comes with.
I don't think that the monetary value which you pay for a renderer is necessarily a good benchmark for how good said renderer is. MentalRay isn't free, its cost is plugged into the price of whichever 3d software that it comes with.
eem
10-02-2009, 12:32 AM
iRay vs VRayRT
As cool as Vray is. I'm a bigger fan of Nvidia.
Nvidia backing iRay would lead me to believe that iRay will come out better.
Nvidia has been directing it's GPU development towards the purpose of full film quality animation for years. I do not doubt that the Nvidia acquisition of Mental Images was right along the same lines of this intention. And iRay is further along the same lines. Which means iRay is gonna be a pretty big forefront for the benefits of CUDA and if thats the case, it's gonna be a star CUDA application with alot of support, alot of development and alot of Nvidia money behind it.
If mental images does it. It could be just another tack on. But if Nvidia gets behind it full force, I don't think VrayRT stands a chance. An internally developed CUDA renderer will be done right.
As cool as Vray is. I'm a bigger fan of Nvidia.
Nvidia backing iRay would lead me to believe that iRay will come out better.
Nvidia has been directing it's GPU development towards the purpose of full film quality animation for years. I do not doubt that the Nvidia acquisition of Mental Images was right along the same lines of this intention. And iRay is further along the same lines. Which means iRay is gonna be a pretty big forefront for the benefits of CUDA and if thats the case, it's gonna be a star CUDA application with alot of support, alot of development and alot of Nvidia money behind it.
If mental images does it. It could be just another tack on. But if Nvidia gets behind it full force, I don't think VrayRT stands a chance. An internally developed CUDA renderer will be done right.
Spacelord
10-02-2009, 12:42 AM
I think the great thing about Iray is that is a final frame renderer.
You look at the latest CausticRT video (which is impressive) but the highest resolution
they mention is 800x600px
@Glacierise I'm not too sure that it will only run on quadro/tesla cards.
they do say platforms of choice in most situations, hopefully any nvidia card that can run cuda will work.
Jeff Mottles Iray vid Rocks !!
Fingers crossed to see Iray in 3dsmax 2011
You look at the latest CausticRT video (which is impressive) but the highest resolution
they mention is 800x600px
@Glacierise I'm not too sure that it will only run on quadro/tesla cards.
they do say platforms of choice in most situations, hopefully any nvidia card that can run cuda will work.
Jeff Mottles Iray vid Rocks !!
Fingers crossed to see Iray in 3dsmax 2011
Kabab
10-02-2009, 01:59 AM
Lets not forget a key point here...
Its a hosted solution so you need to upload your entire scene to their render farm the upload time alone could diminish any rendering speed improvements....
If I've misinterpreted this please correct me.
These new Fermi GPU will be the game changer since well you can run C++ on the things so porting renderers across will be much easier...
Its a hosted solution so you need to upload your entire scene to their render farm the upload time alone could diminish any rendering speed improvements....
If I've misinterpreted this please correct me.
These new Fermi GPU will be the game changer since well you can run C++ on the things so porting renderers across will be much easier...
Spacelord
10-02-2009, 02:09 AM
Lets not forget a key point here...
Its a hosted solution so you need to upload your entire scene to their render farm the upload time alone could diminish any rendering speed improvements....
If I've misinterpreted this please correct me.
These new Fermi GPU will be the game changer since well you can run C++ on the things so porting renderers across will be much easier...
Nope read the pdf, it uses Nvidia cuda cards, Quadro, tesla, etc...
All cards that you can use in a workstation or plugin into a workstation.
Its not Mental Images Reality Server, probably similar technology.
cheers
Its a hosted solution so you need to upload your entire scene to their render farm the upload time alone could diminish any rendering speed improvements....
If I've misinterpreted this please correct me.
These new Fermi GPU will be the game changer since well you can run C++ on the things so porting renderers across will be much easier...
Nope read the pdf, it uses Nvidia cuda cards, Quadro, tesla, etc...
All cards that you can use in a workstation or plugin into a workstation.
Its not Mental Images Reality Server, probably similar technology.
cheers
dagon1978
10-02-2009, 11:36 AM
@Glacierise I'm not too sure that it will only run on quadro/tesla cards.
they do say platforms of choice in most situations, hopefully any nvidia card that can run cuda will work.
it needs a cuda capable card, but obviously the quadro are faster
they do say platforms of choice in most situations, hopefully any nvidia card that can run cuda will work.
it needs a cuda capable card, but obviously the quadro are faster
dagon1978
10-02-2009, 11:40 AM
Lets not forget a key point here...
Its a hosted solution so you need to upload your entire scene to their render farm the upload time alone could diminish any rendering speed improvements....
If I've misinterpreted this please correct me.
These new Fermi GPU will be the game changer since well you can run C++ on the things so porting renderers across will be much easier...
in this video they're showing iray inside reality server (with a server doing the gpu rendering), but iray can work in your workstation, using your gpu/acceleration card with cuda
Its a hosted solution so you need to upload your entire scene to their render farm the upload time alone could diminish any rendering speed improvements....
If I've misinterpreted this please correct me.
These new Fermi GPU will be the game changer since well you can run C++ on the things so porting renderers across will be much easier...
in this video they're showing iray inside reality server (with a server doing the gpu rendering), but iray can work in your workstation, using your gpu/acceleration card with cuda
wurp
10-02-2009, 12:13 PM
and how fast is it on a single highend box? how long does it take to render for example an interior like seen in the example? they mention 120 seconds in the pdf, but is that in a 15 cpu cluster or what?
in this video they're showing iray inside reality server (with a server doing the gpu rendering), but iray can work in your workstation, using your gpu/acceleration card with cuda
in this video they're showing iray inside reality server (with a server doing the gpu rendering), but iray can work in your workstation, using your gpu/acceleration card with cuda
dagon1978
10-02-2009, 01:15 PM
and how fast is it on a single highend box? how long does it take to render for example an interior like seen in the example? they mention 120 seconds in the pdf, but is that in a 15 cpu cluster or what?
i think it's in a single gpu, if you look at the video you can see the same scene rendered in few seconds and noiseless
i think it's in a single gpu, if you look at the video you can see the same scene rendered in few seconds and noiseless
CHRiTTeR
10-02-2009, 01:37 PM
i think it's in a single gpu, if you look at the video you can see the same scene rendered in few seconds and noiseless
cant see anything with those compression artifacts
cant see anything with those compression artifacts
dagon1978
10-02-2009, 01:58 PM
cant see anything with those compression artifacts
is not that hard (for me) to distinguish the compression artifacts from the brute force noise, if you can't, just stop looking at these videos and wait to try the full softwares ;)
is not that hard (for me) to distinguish the compression artifacts from the brute force noise, if you can't, just stop looking at these videos and wait to try the full softwares ;)
^Lele^
10-02-2009, 03:17 PM
i think it's in a single gpu, if you look at the video you can see the same scene rendered in few seconds and noiseless
They changed the original PDF faqs, so no more samples on the new one, but i remained convinced the samples shown on the first pdf were all done on a reality server (the 15gpu solution).
After 120 seconds, however, the image (not very compressed at all on the pdf)showed enough noise to produce flickering in animation, particularly in the mid to low light areas, usually problematic with bruteforce approaches.
The image i am speaking of is visible on the iray page (first link), the one with the blown out brights.
*IF* it is this good, at that size, and needs 15 gpus, there's something wrong.
Let's hope infos are a bit more forthcoming, i'd be glad to have one more such a tool in my arsenal.
They changed the original PDF faqs, so no more samples on the new one, but i remained convinced the samples shown on the first pdf were all done on a reality server (the 15gpu solution).
After 120 seconds, however, the image (not very compressed at all on the pdf)showed enough noise to produce flickering in animation, particularly in the mid to low light areas, usually problematic with bruteforce approaches.
The image i am speaking of is visible on the iray page (first link), the one with the blown out brights.
*IF* it is this good, at that size, and needs 15 gpus, there's something wrong.
Let's hope infos are a bit more forthcoming, i'd be glad to have one more such a tool in my arsenal.
dagon1978
10-02-2009, 03:27 PM
They changed the original PDF faqs, so no more samples on the new one, but i remained convinced the samples shown on the first pdf were all done on a reality server (the 15gpu solution).
conviced by what?
After 120 seconds, however, the image (not very compressed at all on the pdf)showed enough noise to produce flickering in animation, particularly in the mid to low light areas, usually problematic with bruteforce approaches.
The image i am speaking of is visible on the iray page (first link), the one with the blown out brights.
*IF* it is this good, at that size, and needs 15 gpus, there's something wrong.
Let's hope infos are a bit more forthcoming, i'd be glad to have one more such a tool in my arsenal.
this is a snapshot of an alpha software, i can't understand how you can say there is something wrong on this, i saw many "wrong" things in the caustics/chaos demos too, but, as long as they're showing a w.i.p. software i dont mind about it
conviced by what?
After 120 seconds, however, the image (not very compressed at all on the pdf)showed enough noise to produce flickering in animation, particularly in the mid to low light areas, usually problematic with bruteforce approaches.
The image i am speaking of is visible on the iray page (first link), the one with the blown out brights.
*IF* it is this good, at that size, and needs 15 gpus, there's something wrong.
Let's hope infos are a bit more forthcoming, i'd be glad to have one more such a tool in my arsenal.
this is a snapshot of an alpha software, i can't understand how you can say there is something wrong on this, i saw many "wrong" things in the caustics/chaos demos too, but, as long as they're showing a w.i.p. software i dont mind about it
CHRiTTeR
10-02-2009, 03:51 PM
is not that hard (for me) to distinguish the compression artifacts from the brute force noise, if you can't, just stop looking at these videos and wait to try the full softwares ;)
i can distinguish it also up to the point it gets blurred, so its impossible to be sure if the noise is really gone or not. you can 'guess' but thats it.
why should i stop watching their videos because of that?
i can distinguish it also up to the point it gets blurred, so its impossible to be sure if the noise is really gone or not. you can 'guess' but thats it.
why should i stop watching their videos because of that?
dagon1978
10-02-2009, 03:55 PM
i can distinguish it also up to the point it gets blurred, so its impossible to be sure if the noise is really gone or not. you can 'guess' but thats it.
why should i stop watching their videos because of that?
if you
cant see anything with those compression artifacts
you need to stop watching that videos, because you can't see anything :rolleyes: it's just... logic?
but if you
i can distinguish it also up to the point it gets blurred, so its impossible to be sure if the noise is really gone or not
it's something different, so you can "guess" as i'm doing now, right?
why should i stop watching their videos because of that?
if you
cant see anything with those compression artifacts
you need to stop watching that videos, because you can't see anything :rolleyes: it's just... logic?
but if you
i can distinguish it also up to the point it gets blurred, so its impossible to be sure if the noise is really gone or not
it's something different, so you can "guess" as i'm doing now, right?
CHRiTTeR
10-02-2009, 04:01 PM
i assumed if one saw the video he would understand what i meant. i thought it was quite obvious
dagon1978
10-02-2009, 04:06 PM
i assumed if one saw the video he would understand what i meant. i thought it was quite obvious
yes, and for me it's pretty obvious in this video that it doesn't take 120 seconds to make it clean, it's a matter of 5-10 seconds and if you use a logical equation 120/16 you can get my point
yes, and for me it's pretty obvious in this video that it doesn't take 120 seconds to make it clean, it's a matter of 5-10 seconds and if you use a logical equation 120/16 you can get my point
mental
10-02-2009, 08:20 PM
Quit acting like a jerk Dagon.
^Lele^
10-02-2009, 10:27 PM
conviced by what?
By what was written under the images?
this is a snapshot of an alpha software, i can't understand how you can say there is something wrong on this, i saw many "wrong" things in the caustics/chaos demos too, but, as long as they're showing a w.i.p. software i dont mind about it
Alpha?
It's due to be released in a month from now.
And they benchmarked it publicly.
As i wrote, i SINCERELY HOPE i misread the pdf, and that more news are forthcoming in the days ramping up to release, as i'd be HAPPY of having it as an added tool to my arsenal.
By what was written under the images?
this is a snapshot of an alpha software, i can't understand how you can say there is something wrong on this, i saw many "wrong" things in the caustics/chaos demos too, but, as long as they're showing a w.i.p. software i dont mind about it
Alpha?
It's due to be released in a month from now.
And they benchmarked it publicly.
As i wrote, i SINCERELY HOPE i misread the pdf, and that more news are forthcoming in the days ramping up to release, as i'd be HAPPY of having it as an added tool to my arsenal.
dagon1978
10-02-2009, 11:44 PM
By what was written under the images?
under the images there was written something about the system used? :curious:
i don't think so
Alpha?
It's due to be released in a month from now.
And they benchmarked it publicly.
As i wrote, i SINCERELY HOPE i misread the pdf, and that more news are forthcoming in the days ramping up to release, as i'd be HAPPY of having it as an added tool to my arsenal.
for alpha i meant just an early version (alpa, beta, call it as you like), and i'm pretty sure you misread the pdf
under the images there was written something about the system used? :curious:
i don't think so
Alpha?
It's due to be released in a month from now.
And they benchmarked it publicly.
As i wrote, i SINCERELY HOPE i misread the pdf, and that more news are forthcoming in the days ramping up to release, as i'd be HAPPY of having it as an added tool to my arsenal.
for alpha i meant just an early version (alpa, beta, call it as you like), and i'm pretty sure you misread the pdf
TAVO
10-03-2009, 12:01 AM
alpa, beta, call it as you like
can i call it Susan ??, that would make me happy.......
The software needs to past the Beta test at least for a public release, maybe some info is wrong somewhere
can i call it Susan ??, that would make me happy.......
The software needs to past the Beta test at least for a public release, maybe some info is wrong somewhere
Spacelord
10-03-2009, 01:05 AM
under the images there was written something about the system used? :curious:
i don't think so
for alpha i meant just an early version (alpa, beta, call it as you like), and i'm pretty sure you misread the pdf
For those images in the pdf I have no idea of what system was used.
At the Nvidia conference the video they said it was using 15 gpus and it looks like the same scene. It would be good to know what resolutions they test were and what resolutions its cappable of. I wonder if the resolution might be based of how much ram the gpu card has.
i don't think so
for alpha i meant just an early version (alpa, beta, call it as you like), and i'm pretty sure you misread the pdf
For those images in the pdf I have no idea of what system was used.
At the Nvidia conference the video they said it was using 15 gpus and it looks like the same scene. It would be good to know what resolutions they test were and what resolutions its cappable of. I wonder if the resolution might be based of how much ram the gpu card has.
dagon1978
10-03-2009, 01:32 AM
For those images in the pdf I have no idea of what system was used.
At the Nvidia conference the video they said it was using 15 gpus and it looks like the same scene. It would be good to know what resolutions they test were and what resolutions its cappable of. I wonder if the resolution might be based of how much ram the gpu card has.
the same scene doesn't mean the same system ;)
the pdf was older than the presentation (and is not available anymore), it was showing the capabilities of iray, so why rendertimes based on a 16GPUs system? why not 32 or 64 or 1024?
in the presentation they're using this kind of system:
1- to show what you can do with reality server
2- to show a production scene rendered in a couple of seconds interactively
i think it's pretty obvious, but i dont want to repeat myself again and again ;)
just waint the release if you dont believe me
can i call it Susan ??, that would make me happy.......
The software needs to past the Beta test at least for a public release, maybe some info is wrong somewhere
and where you can see any "public release" announcement? ;)
this is the point, mental ray is not a "public software", it's integrated by the OEMs
At the Nvidia conference the video they said it was using 15 gpus and it looks like the same scene. It would be good to know what resolutions they test were and what resolutions its cappable of. I wonder if the resolution might be based of how much ram the gpu card has.
the same scene doesn't mean the same system ;)
the pdf was older than the presentation (and is not available anymore), it was showing the capabilities of iray, so why rendertimes based on a 16GPUs system? why not 32 or 64 or 1024?
in the presentation they're using this kind of system:
1- to show what you can do with reality server
2- to show a production scene rendered in a couple of seconds interactively
i think it's pretty obvious, but i dont want to repeat myself again and again ;)
just waint the release if you dont believe me
can i call it Susan ??, that would make me happy.......
The software needs to past the Beta test at least for a public release, maybe some info is wrong somewhere
and where you can see any "public release" announcement? ;)
this is the point, mental ray is not a "public software", it's integrated by the OEMs
Spacelord
10-03-2009, 01:44 AM
the same scene doesn't mean the same system ;)
the pdf was older than the presentation (and is not available anymore), it was showing the capabilities of iray, so why rendertimes based on a 16GPUs system? why not 32 or 64 or 1024?
in the presentation they're using this kind of system:
1- to show what you can do with reality server
2- to show a production scene rendered in a couple of seconds interactively
i think it's pretty obvious, but i dont want to repeat myself again and again ;)
just waint the release if you dont believe me
Hey I don't know why your getting defensive I was just pointing what the system specs they used at Nvidia conference.
I thought Reality server is different to Iray ? I thought it was more of an Internet based application.
Once its in our hands I'm sure there will be lots to speed tests.
the pdf was older than the presentation (and is not available anymore), it was showing the capabilities of iray, so why rendertimes based on a 16GPUs system? why not 32 or 64 or 1024?
in the presentation they're using this kind of system:
1- to show what you can do with reality server
2- to show a production scene rendered in a couple of seconds interactively
i think it's pretty obvious, but i dont want to repeat myself again and again ;)
just waint the release if you dont believe me
Hey I don't know why your getting defensive I was just pointing what the system specs they used at Nvidia conference.
I thought Reality server is different to Iray ? I thought it was more of an Internet based application.
Once its in our hands I'm sure there will be lots to speed tests.
dagon1978
10-03-2009, 01:47 AM
p.s.
yes, for higher resolution you need more memory, this depends also on the complexity of the scene
yes, for higher resolution you need more memory, this depends also on the complexity of the scene
dagon1978
10-03-2009, 01:53 AM
Hey I don't know why your getting defensive I was just pointing what the system specs they used at Nvidia conference.
I thought Reality server is different to Iray ? I thought it was more of an Internet based application.
Once its in our hands I'm sure there will be lots to speed tests.
sorry Duncan, but my english is pretty poor, so many times i'm not able to use the right nuance on my messages ;)
iray is part of mental ray 3.8 and reality server 2.4, so yes and not they're not differents, you can use iray in RS or in ray, in the nvidia presentation they choosed to use RS
I thought Reality server is different to Iray ? I thought it was more of an Internet based application.
Once its in our hands I'm sure there will be lots to speed tests.
sorry Duncan, but my english is pretty poor, so many times i'm not able to use the right nuance on my messages ;)
iray is part of mental ray 3.8 and reality server 2.4, so yes and not they're not differents, you can use iray in RS or in ray, in the nvidia presentation they choosed to use RS
dnashj33
10-03-2009, 02:42 AM
considdering that demo is running on 15 gpu's im not really impressed.
I expected more, compared to how vrayrt (gpu edition) runs on 1 gaming gpu.That's true, but if it works across multiple GPU's then that's a boon for the industry as well. No CG application has been able to leverage multi-GPU's yet, as far as I know. Gamers have been able to for a few years, but I can't think of a single CG program that benefits from it.
I expected more, compared to how vrayrt (gpu edition) runs on 1 gaming gpu.That's true, but if it works across multiple GPU's then that's a boon for the industry as well. No CG application has been able to leverage multi-GPU's yet, as far as I know. Gamers have been able to for a few years, but I can't think of a single CG program that benefits from it.
^Lele^
10-03-2009, 07:30 AM
under the images there was written something about the system used? :curious:
i don't think so
for alpha i meant just an early version (alpa, beta, call it as you like), and i'm pretty sure you misread the pdf
Ah well, thanks for all the wisdom, now i see the light. I was wrong, as you say so.
The true spirit of constructive discussion.
English nuances or not, mate.
i don't think so
for alpha i meant just an early version (alpa, beta, call it as you like), and i'm pretty sure you misread the pdf
Ah well, thanks for all the wisdom, now i see the light. I was wrong, as you say so.
The true spirit of constructive discussion.
English nuances or not, mate.
dagon1978
10-03-2009, 09:56 AM
Ah well, thanks for all the wisdom, now i see the light. I was wrong, as you say so.
The true spirit of constructive discussion.
English nuances or not, mate.
eheh Lele, i can't understand how your posts can be "constructive" as i asked you to clarify your belief and your answer is "by what was written under the images"
what i meant was:
- where in the video you can see a single frame that needs 120 seconds to clean the scene?
- why do you think that the pdf was releated to this presentation (as it was older and probably related to SIGGRAPH presentation = 1GPU)?
- which is the logic to report in a pdf rendertimes about a 16GPU system? without writing system specs? don't you think is much more logical to write rendertimes about 1 single high-end GPU?
- if there was any relation between the pdf and this presentation (scene apart) why they removed it with a newer pdf?
The true spirit of constructive discussion.
English nuances or not, mate.
eheh Lele, i can't understand how your posts can be "constructive" as i asked you to clarify your belief and your answer is "by what was written under the images"
what i meant was:
- where in the video you can see a single frame that needs 120 seconds to clean the scene?
- why do you think that the pdf was releated to this presentation (as it was older and probably related to SIGGRAPH presentation = 1GPU)?
- which is the logic to report in a pdf rendertimes about a 16GPU system? without writing system specs? don't you think is much more logical to write rendertimes about 1 single high-end GPU?
- if there was any relation between the pdf and this presentation (scene apart) why they removed it with a newer pdf?
vlad
10-05-2009, 02:18 PM
Hey I don't know why your getting defensive I was just pointing what the system specs they used at Nvidia conference...
Look at the pink text link at the bottom of his posts ;)
Look at the pink text link at the bottom of his posts ;)
morimitsu
10-05-2009, 03:04 PM
Cons:
-it's not a preview for mental ray, i.e. - not as full featured as mr
-no motion blur, hence no animation
Looks supercool, but will probably use it for shading/lighting design only, because of the lack of motion blur. Still, integration in mental ray is great.
Are they planning to make it a full featured mental ray render capable, with motion blur, all shader types, etc?
In other words, eventually we may just use mr in maya as usual (without any lack of features) and just select the option to render with gpu?
If so, well, it's outstanding.
-it's not a preview for mental ray, i.e. - not as full featured as mr
-no motion blur, hence no animation
Looks supercool, but will probably use it for shading/lighting design only, because of the lack of motion blur. Still, integration in mental ray is great.
Are they planning to make it a full featured mental ray render capable, with motion blur, all shader types, etc?
In other words, eventually we may just use mr in maya as usual (without any lack of features) and just select the option to render with gpu?
If so, well, it's outstanding.
jupiterjazz
10-05-2009, 05:08 PM
Well well,
I can’t really let pass a new mental announcement without some constructive deconstruction... so:
~ Ahhh... yet another half-baked mental “solution”! ~
Proof that it’s half baked: none
History that backs me up: MetaSL, mental mill, reality server, mental queue, mental etc. (mmm maybe someone should patent this last idea :) ).
But let’s move to some further thoughts:
* iray in reality server
==============
- At NV’s gpu conference iray is shown embedded within reality server. Why?
Simply because mental is still trying to find a way to market reality server, which is their biggest product failure as it is basically unsold as of today so they are putting inside anything they have to make it appealable. Truth is that one can simply build rs just by using opnesource softwares (see old threads). For who is still confused about rs (it’s not your fault, just mental’s) it is basically a pipeline to do sever side rendering with remote users being able to control and tweak certain parameters. All packed in a ultra awkward, complicated - hey let’s use the right word for one time! - “mental” pipeline. It includes several software and hardware renderers, iray being one of the hw ones.
- iray in sceniX: We will see if how it gets in.
Also mental mill is integrated in FXComposer, and it wasn’t a revolution for the industry...
* iray
====
First of all: sincerely congrats to mental for having restrained from putting in the faq that iray is for production usage. Then:
- developed by mental, NOT by nvidia (hence the sadness)
- does not use optix api (optix if you don’t know is not a renderer but an API, it is the official name of the old codename “nvirt”: nvidia raytracing api). Hint: further proof of the how few a close-minded company as mental collaborates with an open one as nvidia (this is because of mental management, not nvidia). The orange bugatti shown at the start of gpu conference talk is optix:
http://www.youtube.com/watch?v=vPeZf8WJDM4&feature=related
and this is also optix:
http://www.youtube.com/watch?v=BAZQlQ86IB4&feature=channel
- closed: Optix API will allow developers to write their own “iray”, in a more open and controllable fashion (when Optix API will be released you will see for yourself the API and code examples) again
- hardcoded shader (based on archmat) and no programmable shaders (yes, not even MetaSL! hint: even mental does not use its own technology...)
- no motion blur
- no curves (so no hair nor fur)
- no particles
- no volumes
- no SDS
- no NURBS
- no displacement
- no sss
- performance: see for yourself results on 15 GPUs
hint: mental removed images from PDF...
- hint: the scene used for demo features a very large horizontal window, on a more tricky setup, with smaller light entrance noise should appear and converge-time increase
- no bidirectional correlation to mental ray (so you basically need a double pipeline)
- not useful for relighting: if you think you can use iray to relight your mental ray setup and then switch back to mr for final rendering, well, this is not the case.
- misleading faq: mental ray 3.8 integration: iray is not “integrated” with mray 3.8, it will be “shipped along”.
- integration on autodesk softwares: I would be very surprised if they do it, it has no value for 3dsmax/maya/softimage. Hint: no words from 3dsmax prod managers in this thread.
- No non-nvidia gpu, no OpenCL support
- mental is again late. VRay RT has at least 1.5 years of advantage.
* MetaSL
======
- still no VM compiler! Hello mental! 5 years have passed. And now it’s too late: RSL2.0 and OSL.
- metasl shaders support in mental ray: slow, unstable, not full feature
- showstopper performance issues for compilation of real production shaders
- no adoption in film industry
- oems: afaics only autodesk and dassault are using metasl and only for HW previews (as I said there are serious performance issues for on-the-fly compilation of real shaders)
- metasl has probably a very limited future because of RSL2.0 and the upcoming OSL (which will provide all infrastructure, including compiler, opensource).
These are just my opinions anyway, I guess now is time to just sit back and enjoy the usual nice comments ;)
edit: I also find ridiculous that specially with mental products the symbol of TM or (c) or (r) is omnipresent, even on thread title.... It definitely shows up some mental attitudeTM.
P.
I can’t really let pass a new mental announcement without some constructive deconstruction... so:
~ Ahhh... yet another half-baked mental “solution”! ~
Proof that it’s half baked: none
History that backs me up: MetaSL, mental mill, reality server, mental queue, mental etc. (mmm maybe someone should patent this last idea :) ).
But let’s move to some further thoughts:
* iray in reality server
==============
- At NV’s gpu conference iray is shown embedded within reality server. Why?
Simply because mental is still trying to find a way to market reality server, which is their biggest product failure as it is basically unsold as of today so they are putting inside anything they have to make it appealable. Truth is that one can simply build rs just by using opnesource softwares (see old threads). For who is still confused about rs (it’s not your fault, just mental’s) it is basically a pipeline to do sever side rendering with remote users being able to control and tweak certain parameters. All packed in a ultra awkward, complicated - hey let’s use the right word for one time! - “mental” pipeline. It includes several software and hardware renderers, iray being one of the hw ones.
- iray in sceniX: We will see if how it gets in.
Also mental mill is integrated in FXComposer, and it wasn’t a revolution for the industry...
* iray
====
First of all: sincerely congrats to mental for having restrained from putting in the faq that iray is for production usage. Then:
- developed by mental, NOT by nvidia (hence the sadness)
- does not use optix api (optix if you don’t know is not a renderer but an API, it is the official name of the old codename “nvirt”: nvidia raytracing api). Hint: further proof of the how few a close-minded company as mental collaborates with an open one as nvidia (this is because of mental management, not nvidia). The orange bugatti shown at the start of gpu conference talk is optix:
http://www.youtube.com/watch?v=vPeZf8WJDM4&feature=related
and this is also optix:
http://www.youtube.com/watch?v=BAZQlQ86IB4&feature=channel
- closed: Optix API will allow developers to write their own “iray”, in a more open and controllable fashion (when Optix API will be released you will see for yourself the API and code examples) again
- hardcoded shader (based on archmat) and no programmable shaders (yes, not even MetaSL! hint: even mental does not use its own technology...)
- no motion blur
- no curves (so no hair nor fur)
- no particles
- no volumes
- no SDS
- no NURBS
- no displacement
- no sss
- performance: see for yourself results on 15 GPUs
hint: mental removed images from PDF...
- hint: the scene used for demo features a very large horizontal window, on a more tricky setup, with smaller light entrance noise should appear and converge-time increase
- no bidirectional correlation to mental ray (so you basically need a double pipeline)
- not useful for relighting: if you think you can use iray to relight your mental ray setup and then switch back to mr for final rendering, well, this is not the case.
- misleading faq: mental ray 3.8 integration: iray is not “integrated” with mray 3.8, it will be “shipped along”.
- integration on autodesk softwares: I would be very surprised if they do it, it has no value for 3dsmax/maya/softimage. Hint: no words from 3dsmax prod managers in this thread.
- No non-nvidia gpu, no OpenCL support
- mental is again late. VRay RT has at least 1.5 years of advantage.
* MetaSL
======
- still no VM compiler! Hello mental! 5 years have passed. And now it’s too late: RSL2.0 and OSL.
- metasl shaders support in mental ray: slow, unstable, not full feature
- showstopper performance issues for compilation of real production shaders
- no adoption in film industry
- oems: afaics only autodesk and dassault are using metasl and only for HW previews (as I said there are serious performance issues for on-the-fly compilation of real shaders)
- metasl has probably a very limited future because of RSL2.0 and the upcoming OSL (which will provide all infrastructure, including compiler, opensource).
These are just my opinions anyway, I guess now is time to just sit back and enjoy the usual nice comments ;)
edit: I also find ridiculous that specially with mental products the symbol of TM or (c) or (r) is omnipresent, even on thread title.... It definitely shows up some mental attitudeTM.
P.
Buexe
10-05-2009, 05:27 PM
Does anybody really expect a good product that makes production efficient from mi?
Kzin
10-05-2009, 05:38 PM
...
i think now it should be clear that MI does not look on tv/broadcast or filmmarket. it looks like that they concentrate their product for archvis/cad only. so lets hope that some new stuff for mr will follow. ;)
i think now it should be clear that MI does not look on tv/broadcast or filmmarket. it looks like that they concentrate their product for archvis/cad only. so lets hope that some new stuff for mr will follow. ;)
rBrady
10-05-2009, 05:43 PM
Its possible, Personally I think its way too early to tell if this is good or not. All the tech demos I have seen can be very misleading. In a few weeks (or months) I am sure we will see some vrayRT vs iRay benchmarks and production experiences that will shed some light. Either way, the GPU has massive potential.
I don't know how close we are to using the GPU for production rendering but were moving in that direction faster at least! Even if its not ready, it can still be cool.
I don't know how close we are to using the GPU for production rendering but were moving in that direction faster at least! Even if its not ready, it can still be cool.
DrBalthar
10-05-2009, 06:27 PM
These are just my opinions anyway, I guess now is time to just sit back and enjoy the usual nice comments ;)
Great summing up on the situation. May I add one thing. Unbiased rendering engine, at least three years late (Maxwell,FryRender Indigo, LuxRender) already there since years.
As I said before I think mental images have completely lost it, they used to be leaders in the market now they just tyring catching up (very late).
Great summing up on the situation. May I add one thing. Unbiased rendering engine, at least three years late (Maxwell,FryRender Indigo, LuxRender) already there since years.
As I said before I think mental images have completely lost it, they used to be leaders in the market now they just tyring catching up (very late).
Glacierise
10-05-2009, 07:04 PM
All the current unbiased renderers are not GPU accelerated and in fact very slow, having one that is fast enough for 'interactive' use will be a new thing.
Spacelord
10-06-2009, 01:17 AM
Hi Glacierise, I agree I think it will be a great asset for the Arch Vis artist.
It might not be for the Movie industry offering every single whizz bang feature.
But for Arch Vis artists, if its fast its going to be great.
It will be interesting to see how it handles caustics since most Unbiased renderers struggle clearing the noise.
It might not be for the Movie industry offering every single whizz bang feature.
But for Arch Vis artists, if its fast its going to be great.
It will be interesting to see how it handles caustics since most Unbiased renderers struggle clearing the noise.
mister3d
10-06-2009, 03:41 AM
It will be interesting to see how it handles caustics since most Unbiased renderers struggle clearing the noise.
I'm wondering why? :rolleyes: Vray RT GPU shwed that it does it fast, but it's still noisy!
I'm wondering why? :rolleyes: Vray RT GPU shwed that it does it fast, but it's still noisy!
Spacelord
10-06-2009, 04:16 AM
I'm wondering why? :rolleyes: Vray RT GPU shwed that it does it fast, but it's still noisy!
VrayRT isn't a final frame renderer, where as Iray is so i wouldn't want noise to be in the final render.
VrayRT isn't a final frame renderer, where as Iray is so i wouldn't want noise to be in the final render.
mister3d
10-06-2009, 04:22 AM
VrayRT isn't a final frame renderer, where as Iray is so i wouldn't want noise to be in the final render.
Why it's not? It is actually, considering it supports almost all features and is identical to the cpu rendering quality. The lack of mb, sssis a drawback, but not for everything. Vray cpu will give you the exact noisy caustics.
Why it's not? It is actually, considering it supports almost all features and is identical to the cpu rendering quality. The lack of mb, sssis a drawback, but not for everything. Vray cpu will give you the exact noisy caustics.
Spacelord
10-06-2009, 04:32 AM
Why it's not? It is actually, considering it supports almost all features and is identical to the cpu rendering quality. The lack of mb, sssis a drawback, but not for everything. Vray cpu will give you the exact noisy caustics.
So do you know for certian Iray is going have the same troubles as a unbiased renderer has with Caustics ?
I thought VrayRT was for only tweaking materials, lights Et cetera
and when you wanted to render a final 10k Frame you had to go through the normal Vray.
So VrayRt can render high res stills ? is this what your saying ?
So do you know for certian Iray is going have the same troubles as a unbiased renderer has with Caustics ?
I thought VrayRT was for only tweaking materials, lights Et cetera
and when you wanted to render a final 10k Frame you had to go through the normal Vray.
So VrayRt can render high res stills ? is this what your saying ?
mister3d
10-06-2009, 05:22 AM
So do you know for certian Iray is going have the same troubles as a unbiased renderer has with Caustics ?
No, I don't know much. And what troubles Caustics has?
I thought VrayRT was for only tweaking materials, lights Et cetera
Vray RT yes, but Vray RT GPU has the identical image quality as the standard Vray.
and when you wanted to render a final 10k Frame you had to go through the normal Vray.
So VrayRt can render high res stills ? is this what your saying ?
I don't know unfortunatrly.. but considering Caustics showed you can increase the resolution, I'm wondering why Vray RT GPU won't?
No, I don't know much. And what troubles Caustics has?
I thought VrayRT was for only tweaking materials, lights Et cetera
Vray RT yes, but Vray RT GPU has the identical image quality as the standard Vray.
and when you wanted to render a final 10k Frame you had to go through the normal Vray.
So VrayRt can render high res stills ? is this what your saying ?
I don't know unfortunatrly.. but considering Caustics showed you can increase the resolution, I'm wondering why Vray RT GPU won't?
Spacelord
10-06-2009, 05:32 AM
There might be some confusion here, I'm talking about Caustics through glass, etc..
Not CausticRT the hardware, the renderer.
Not CausticRT the hardware, the renderer.
jupiterjazz
10-06-2009, 07:56 AM
All the current unbiased renderers are not GPU accelerated and in fact very slow, having one that is fast enough for 'interactive' use will be a new thing.
iray aswell is *not* unbiased, it's as "consistent".
Essentially "consistent" means that your image *can* get brighter or darker with more iterations but converges to the theoretical correct result with time, while for an unbiased renderer, the average brightness of every iteration is correct (at least the expected value). The funny thing is that you can actually understand this from the keynote, just look at how the image refines and gets brighter with each iteration: that's because it picks up more light, thing that wouldn't happen with an unbiased renderer.
p
iray aswell is *not* unbiased, it's as "consistent".
Essentially "consistent" means that your image *can* get brighter or darker with more iterations but converges to the theoretical correct result with time, while for an unbiased renderer, the average brightness of every iteration is correct (at least the expected value). The funny thing is that you can actually understand this from the keynote, just look at how the image refines and gets brighter with each iteration: that's because it picks up more light, thing that wouldn't happen with an unbiased renderer.
p
Spacelord
10-06-2009, 09:38 AM
iray aswell is *not* unbiased, it's, using mental terminology "consistent" (these guys are always amazing with names...)
Essentially "consistent" means that your image *can* get brighter or darker with more iterations but converges to the theoretical correct result with time, while for an unbiased renderer, the average brightness of every iteration is correct (at least the expected value). The funny thing is that you can actually understand this from the keynote, just look at how the image refines and gets brighter with each iteration: that's because it picks up more light, thing that wouldn't happen with an unbiased renderer.
p
Not sure about that, Rendering in unbiased renderer, it starts of quite dark and as it samples more, the more light shows up.
cheers
Essentially "consistent" means that your image *can* get brighter or darker with more iterations but converges to the theoretical correct result with time, while for an unbiased renderer, the average brightness of every iteration is correct (at least the expected value). The funny thing is that you can actually understand this from the keynote, just look at how the image refines and gets brighter with each iteration: that's because it picks up more light, thing that wouldn't happen with an unbiased renderer.
p
Not sure about that, Rendering in unbiased renderer, it starts of quite dark and as it samples more, the more light shows up.
cheers
Glacierise
10-06-2009, 10:04 AM
If it's not unbiased, what have they accelerated then? Photon mapping? That should be good for mr too?
jupiterjazz
10-06-2009, 10:52 AM
Not sure about that, Rendering in unbiased renderer, it starts of quite dark and as it samples more, the more light shows up.
cheers
...We are not speaking the same language here. The avg brightness of samples changes in iray at each pass, aka a sample changes its brightness, while in an unbiased renderer not, samples cannot get brighter. And again, at each pass they change again so they -theoretically- converge to a correct solution.
If it's not unbiased, what have they accelerated then? Photon mapping? That should be good for mr too?
Argh... unbiased rendering is not an acceleration technique!
Iray is *not* mental ray and does not use photon mapping.
It's a path tracer, there is no need to accelerate anything if you have 15 GPUs plugged, and ,besides, considering all that power I think it's not so fast...
The point is that feature-wise does not support anything. It's barely useful besides for marketing announcements.
It seems there is a lot of confusion here, I am not surprised since mental is known to be unable to communicate with clients and the general audience (it's done on purpose, it is a strategy to trap users in cumbersome api: one example to rule them all mental ray).
So, forget the consistent Vs unbiased discussion, with so many limitations (see my other pot on page 4 of this thread) iray proves to be just at the very least at experi-mental stage (regardless what they say in their press release where it seems a finished product), and knowing the infamous mental usability saying that the pipeline from any application to iray will be troublesome and plenty of caveats is highly probable.
For mental ray fans: truth is that mental is unable to progress with mental ray, especially with hw assistance, see this amazing sentences which I find even damaging the image of nvidia:
http://events.animationblogspot.com/2009/08/07/desowitz-at-siggraph-2009-day-5/
mental is losing marketshare because of competition in arch/product viz world (vray / vrayrt / modo / maxwell) and film/animation (3delight, PRMan, Arnold).
Once ADSK will get tired of promises and you will see a change. It looks to me this time ADSK is not buying mental words, so it seems there is hope.
P.
cheers
...We are not speaking the same language here. The avg brightness of samples changes in iray at each pass, aka a sample changes its brightness, while in an unbiased renderer not, samples cannot get brighter. And again, at each pass they change again so they -theoretically- converge to a correct solution.
If it's not unbiased, what have they accelerated then? Photon mapping? That should be good for mr too?
Argh... unbiased rendering is not an acceleration technique!
Iray is *not* mental ray and does not use photon mapping.
It's a path tracer, there is no need to accelerate anything if you have 15 GPUs plugged, and ,besides, considering all that power I think it's not so fast...
The point is that feature-wise does not support anything. It's barely useful besides for marketing announcements.
It seems there is a lot of confusion here, I am not surprised since mental is known to be unable to communicate with clients and the general audience (it's done on purpose, it is a strategy to trap users in cumbersome api: one example to rule them all mental ray).
So, forget the consistent Vs unbiased discussion, with so many limitations (see my other pot on page 4 of this thread) iray proves to be just at the very least at experi-mental stage (regardless what they say in their press release where it seems a finished product), and knowing the infamous mental usability saying that the pipeline from any application to iray will be troublesome and plenty of caveats is highly probable.
For mental ray fans: truth is that mental is unable to progress with mental ray, especially with hw assistance, see this amazing sentences which I find even damaging the image of nvidia:
http://events.animationblogspot.com/2009/08/07/desowitz-at-siggraph-2009-day-5/
mental is losing marketshare because of competition in arch/product viz world (vray / vrayrt / modo / maxwell) and film/animation (3delight, PRMan, Arnold).
Once ADSK will get tired of promises and you will see a change. It looks to me this time ADSK is not buying mental words, so it seems there is hope.
P.
Kzin
10-06-2009, 11:10 AM
at the moment it looks like Mental Images dont know in which direction they should go, a bit headless to me. they developed irradience particals, a promising new method for gi, but nothings go further with it. now they introduced iray and all features that we need are missing (motion blur, hair, displacements). so the real question is what plans have mental images for future mental ray releases? or is the development concentrate on iray now?
and for the arnold thing, it is not a good sign that a mental ray customer start to write their own renderer and start to selling it. "i see dark clouds on the horizon" ;)
and for the arnold thing, it is not a good sign that a mental ray customer start to write their own renderer and start to selling it. "i see dark clouds on the horizon" ;)
inguatu
10-06-2009, 02:31 PM
...
For mental ray fans: truth is that mental is unable to progress with mental ray, especially with hw assistance, see this amazing sentences which I find even damaging the image of nvidia:
http://events.animationblogspot.com/2009/08/07/desowitz-at-siggraph-2009-day-5/ (http://events.animationblogspot.com/2009/08/07/desowitz-at-siggraph-2009-day-5/)
P.
So what did 3Delight showcase as new technology at Sig09? Didn't see any articles about them. I'm curious to see what they have up their sleeves.
For mental ray fans: truth is that mental is unable to progress with mental ray, especially with hw assistance, see this amazing sentences which I find even damaging the image of nvidia:
http://events.animationblogspot.com/2009/08/07/desowitz-at-siggraph-2009-day-5/ (http://events.animationblogspot.com/2009/08/07/desowitz-at-siggraph-2009-day-5/)
P.
So what did 3Delight showcase as new technology at Sig09? Didn't see any articles about them. I'm curious to see what they have up their sleeves.
inguatu
10-06-2009, 02:39 PM
and for the arnold thing, it is not a good sign that a mental ray customer start to write their own renderer and start to selling it. "i see dark clouds on the horizon" ;)
Arnold has been around for years. It was written when running MR on XSI was the only real option. I wouldn't spout off doom and gloom when comparing or referring to the Arnold renderer. What was used in Meatballs was an internal version Sony crafted for their own needs. I'm not sure how much original code is left. A Sony dude would have to chime in on that. Other than that, Arnold is supposidly still in Messiah Studio, although obviously not as well used as Mental Ray, Renderman, Vray, etc.
Arnold has been around for years. It was written when running MR on XSI was the only real option. I wouldn't spout off doom and gloom when comparing or referring to the Arnold renderer. What was used in Meatballs was an internal version Sony crafted for their own needs. I'm not sure how much original code is left. A Sony dude would have to chime in on that. Other than that, Arnold is supposidly still in Messiah Studio, although obviously not as well used as Mental Ray, Renderman, Vray, etc.
Kzin
10-06-2009, 02:46 PM
Arnold has been around for years. It was written when running MR on XSI was the only real option. I wouldn't spout off doom and gloom when comparing or referring to the Arnold renderer. What was used in Meatballs was an internal version Sony crafted for their own needs. I'm not sure how much original code is left. A Sony dude would have to chime in on that. Other than that, Arnold is supposidly still in Messiah Studio, although obviously not as well used as Mental Ray, Renderman, Vray, etc.
no, its not the old arnold one. the version which will be sold in the future is the sony renderer, their inhouse arnold verion. so i hope it will be good as expected, its production proven and really fast, the first images and infos goes in that direction.
edit: but i have to agree, we have to wait until we can test the renderer. ;)
no, its not the old arnold one. the version which will be sold in the future is the sony renderer, their inhouse arnold verion. so i hope it will be good as expected, its production proven and really fast, the first images and infos goes in that direction.
edit: but i have to agree, we have to wait until we can test the renderer. ;)
inguatu
10-06-2009, 02:48 PM
no, its not the old arnold one. the version which will be sold in the future is the sony renderer, their inhouse arnold verion. so i hope it will be good as expected, its production proven and really fast, the first images and infos goes in that direction.
edit: but i have to agree, we have to wait until we can test the renderer. ;)
that is a NICE surprising turn of events! Thanks for the correction.
edit: but i have to agree, we have to wait until we can test the renderer. ;)
that is a NICE surprising turn of events! Thanks for the correction.
jupiterjazz
10-06-2009, 03:18 PM
So what did 3Delight showcase as new technology at Sig09? Didn't see any articles about them. I'm curious to see what they have up their sleeves.
It has noting to do with the subject, anyway, you should ask D'nA, we simply use their tools and develop on top via their APIs and language for our clients.
But I can tell you that there are companies that "talk before do" and other that "do and then talk". I worked for the former -you know who- and I got bored beyond belief...
So now we just use the best tools for our needs, and for sure the ones from D'nA conform to the second category.
P.
It has noting to do with the subject, anyway, you should ask D'nA, we simply use their tools and develop on top via their APIs and language for our clients.
But I can tell you that there are companies that "talk before do" and other that "do and then talk". I worked for the former -you know who- and I got bored beyond belief...
So now we just use the best tools for our needs, and for sure the ones from D'nA conform to the second category.
P.
TwinSnakes
10-06-2009, 05:26 PM
Have you even seen the VrayRT(GPU) demo?
VrayRT(GPU) could've done the entire building on one GPU, that took 15 GPU's for iRay to do one floor.
you should compare the scene complexity. the mr demo is a complete different thing, glossy shader, many area lights, indoor gi, all stuff that i would like to see from vrayrt. so we should wait until we can test the two renderer. ;)
VrayRT(GPU) could've done the entire building on one GPU, that took 15 GPU's for iRay to do one floor.
you should compare the scene complexity. the mr demo is a complete different thing, glossy shader, many area lights, indoor gi, all stuff that i would like to see from vrayrt. so we should wait until we can test the two renderer. ;)
Kzin
10-06-2009, 05:43 PM
of course i 've seen the demo, live. and again, the vray demo has not the visual complexity like the iray ones. look at the shader and lights. ;)
Array
10-06-2009, 05:45 PM
Great summing up on the situation. May I add one thing. Unbiased rendering engine, at least three years late (Maxwell,FryRender Indigo, LuxRender) already there since years.
As I said before I think mental images have completely lost it, they used to be leaders in the market now they just tyring catching up (very late).
You are comparing apples and oranges here. MentalRay was developed as a general purpose, programmable, renderer for use in a variety of situations, including animation production. Maxwell was built from the ground up as a physical light simulator for use in product and architectural visualization. Just because MentalRay does not yet offer an unbiased rendering mode does not mean that Mental Images (now Nvidia) have "lost it". Think of all the ways that Maxwell would have to change to fill the same niche that MentalRay currently occupies. It's not even worth discussing. By your metric, Pixar's PRMan renderer is a dinosaur.
As I said before I think mental images have completely lost it, they used to be leaders in the market now they just tyring catching up (very late).
You are comparing apples and oranges here. MentalRay was developed as a general purpose, programmable, renderer for use in a variety of situations, including animation production. Maxwell was built from the ground up as a physical light simulator for use in product and architectural visualization. Just because MentalRay does not yet offer an unbiased rendering mode does not mean that Mental Images (now Nvidia) have "lost it". Think of all the ways that Maxwell would have to change to fill the same niche that MentalRay currently occupies. It's not even worth discussing. By your metric, Pixar's PRMan renderer is a dinosaur.
DrBalthar
10-06-2009, 06:51 PM
You are comparing apples and oranges here. MentalRay was developed as a general purpose, programmable, renderer for use in a variety of situations, including
No I'm not have a look at the mental ray API it has enough functionality in there that they could simply expose an unbiased front-end to it. They just did not do it. Hence if you have enough time you could write your own unbiased render on top of mental ray.
No I'm not have a look at the mental ray API it has enough functionality in there that they could simply expose an unbiased front-end to it. They just did not do it. Hence if you have enough time you could write your own unbiased render on top of mental ray.
inguatu
10-06-2009, 10:55 PM
It has noting to do with the subject, anyway, you should ask D'nA, we simply use their tools and develop on top via their APIs and language for our clients.
But I can tell you that there are companies that "talk before do" and other that "do and then talk". I worked for the former -you know who- and I got bored beyond belief...
So now we just use the best tools for our needs, and for sure the ones from D'nA conform to the second category.
P.
sorry, I figured as many times as you go off on tangets (off subject) when anything regarding the word "mental" is used, that you'd put forth some information about 3Delight.
Yes you are correct, DNA never puts out any information.
It's all good.
But I can tell you that there are companies that "talk before do" and other that "do and then talk". I worked for the former -you know who- and I got bored beyond belief...
So now we just use the best tools for our needs, and for sure the ones from D'nA conform to the second category.
P.
sorry, I figured as many times as you go off on tangets (off subject) when anything regarding the word "mental" is used, that you'd put forth some information about 3Delight.
Yes you are correct, DNA never puts out any information.
It's all good.
jupiterjazz
10-07-2009, 12:16 AM
sorry,
No need to be sorry, "that’s just like, uh, your opinion, man" :)
P.
No need to be sorry, "that’s just like, uh, your opinion, man" :)
P.
robinmitra
10-11-2009, 09:36 PM
Impressive! I have been waiting for Mental Ray to create something like this for a long time!
dagon1978
10-26-2009, 09:54 PM
Not sure about that, Rendering in unbiased renderer, it starts of quite dark and as it samples more, the more light shows up.
cheers
another time, he like to talk about what he dont know, you can see it very well from his words
maxwell it's working like this (starting darker and calculating the bounces on the way), vray ppt/RT it's working like this, fry render it's working like this
for paolo they're all biased :rolleyes: LOL
the funny thing is that i'm pretty sure he NEVER tried maxwell, he never tried vray ppt/RT (or he can't see how they're working, i dont know which one is worst), but he like to talk about iray taht he NEVER tried :rolleyes: ahaha funny, funny man
p.s.
iray is un unbiased progressive renderer just like maxwell, vray RT, fry and so...
cheers
another time, he like to talk about what he dont know, you can see it very well from his words
maxwell it's working like this (starting darker and calculating the bounces on the way), vray ppt/RT it's working like this, fry render it's working like this
for paolo they're all biased :rolleyes: LOL
the funny thing is that i'm pretty sure he NEVER tried maxwell, he never tried vray ppt/RT (or he can't see how they're working, i dont know which one is worst), but he like to talk about iray taht he NEVER tried :rolleyes: ahaha funny, funny man
p.s.
iray is un unbiased progressive renderer just like maxwell, vray RT, fry and so...
Kzin
10-26-2009, 11:37 PM
p.s.
iray is un unbiased progressive renderer just like maxwell ...
is iray also doing spectral rendering like maxwell or just the "simple" path tracing approach?
iray is un unbiased progressive renderer just like maxwell ...
is iray also doing spectral rendering like maxwell or just the "simple" path tracing approach?
jupiterjazz
10-27-2009, 12:31 AM
another time, he like to talk about what he dont know, you can see it very well from his words
maxwell it's working like this (starting darker and calculating the bounces on the way), vray ppt/RT it's working like this, fry render it's working like this
for paolo they're all biased :rolleyes: LOL
the funny thing is that i'm pretty sure he NEVER tried maxwell, he never tried vray ppt/RT (or he can't see how they're working, i dont know which one is worst), but he like to talk about iray taht he NEVER tried :rolleyes: ahaha funny, funny man
blah blah blah :)
p.s.
iray is un unbiased progressive renderer just like maxwell, vray RT, fry and so...
[/quote]
Wrong.
Sorry to disappoint you, but iray is not "unbiased", the right technical term is "consistent".
Just dig a bit and you should grasp the concept.
P.
maxwell it's working like this (starting darker and calculating the bounces on the way), vray ppt/RT it's working like this, fry render it's working like this
for paolo they're all biased :rolleyes: LOL
the funny thing is that i'm pretty sure he NEVER tried maxwell, he never tried vray ppt/RT (or he can't see how they're working, i dont know which one is worst), but he like to talk about iray taht he NEVER tried :rolleyes: ahaha funny, funny man
blah blah blah :)
p.s.
iray is un unbiased progressive renderer just like maxwell, vray RT, fry and so...
[/quote]
Wrong.
Sorry to disappoint you, but iray is not "unbiased", the right technical term is "consistent".
Just dig a bit and you should grasp the concept.
P.
dagon1978
10-27-2009, 12:48 AM
Wrong.
Sorry to disappoint you, but iray is not "unbiased", the right technical term is "consistent".
Just dig a bit and you should grasp the concept.
P.
so maxwell, vray ppt/RT, fry are unbiased or "consistent"? tell me paolo
and also, tell me how you can talk about something you never tried, because you never tried iray right? :rolleyes:
i think i gonna use this as my next firm
"just look at how the image refines and gets brighter with each iteration: that's because it picks up more light, thing that wouldn't happen with an unbiased renderer." paolo berto
too funny paolo, too funny
Sorry to disappoint you, but iray is not "unbiased", the right technical term is "consistent".
Just dig a bit and you should grasp the concept.
P.
so maxwell, vray ppt/RT, fry are unbiased or "consistent"? tell me paolo
and also, tell me how you can talk about something you never tried, because you never tried iray right? :rolleyes:
i think i gonna use this as my next firm
"just look at how the image refines and gets brighter with each iteration: that's because it picks up more light, thing that wouldn't happen with an unbiased renderer." paolo berto
too funny paolo, too funny
dagon1978
10-27-2009, 12:49 AM
is iray also doing spectral rendering like maxwell or just the "simple" path tracing approach?
it's an unbiased progressive path tracer
it's an unbiased progressive path tracer
wurp
10-27-2009, 01:05 AM
it's an unbiased progressive path tracer
from the iray faq:
"iray is an intuitive to operate, interactive, consistent, high-performance
global illumination rendering technology that generates photorealistic
imagery by simulating the physical behavior of light."
http://www.mentalimages.com/products/iray
from the iray faq:
"iray is an intuitive to operate, interactive, consistent, high-performance
global illumination rendering technology that generates photorealistic
imagery by simulating the physical behavior of light."
http://www.mentalimages.com/products/iray
dagon1978
10-27-2009, 01:20 AM
from the iray faq:
"iray is an intuitive to operate, interactive, consistent, high-performance
global illumination rendering technology that generates photorealistic
imagery by simulating the physical behavior of light."
http://www.mentalimages.com/products/iray
and then? where you read it is biased (or is not unbiased)?
"iray is an intuitive to operate, interactive, consistent, high-performance
global illumination rendering technology that generates photorealistic
imagery by simulating the physical behavior of light."
http://www.mentalimages.com/products/iray
and then? where you read it is biased (or is not unbiased)?
jupiterjazz
10-27-2009, 01:29 AM
so maxwell, vray ppt/RT, fry are unbiased or "consistent"? tell me paolo
and also, tell me how you can talk about something you never tried, because you never tried iray right? :rolleyes:
i think i gonna use this as my next firm
"just look at how the image refines and gets brighter with each iteration: that's because it picks up more light, thing that wouldn't happen with an unbiased renderer." paolo berto
too funny paolo, too funny
The only funny thing here is that you don't know what you are talking about :)
p
and also, tell me how you can talk about something you never tried, because you never tried iray right? :rolleyes:
i think i gonna use this as my next firm
"just look at how the image refines and gets brighter with each iteration: that's because it picks up more light, thing that wouldn't happen with an unbiased renderer." paolo berto
too funny paolo, too funny
The only funny thing here is that you don't know what you are talking about :)
p
dagon1978
10-27-2009, 01:31 AM
The only funny thing here is that you don't know what you are talking about :)
p
really? lol
again you are not answering my question, are you using iray paolo? :)
have you ever used (or seen) maxwell, vray pp/RT, fry render? :rolleyes:
which one is unbiased and which is "consistent"?
p
really? lol
again you are not answering my question, are you using iray paolo? :)
have you ever used (or seen) maxwell, vray pp/RT, fry render? :rolleyes:
which one is unbiased and which is "consistent"?
jupiterjazz
10-27-2009, 02:03 AM
really? lol
again you are not answering my question, are you using iray paolo? :)
have you ever used (or seen) maxwell, vray pp/RT, fry render? :rolleyes:
which one is unbiased and which is "consistent"?
I just prefer to leave you in the mental cloud (TM) (http://cloud.mental.com/direct/) , it's more entertaining :)
P
again you are not answering my question, are you using iray paolo? :)
have you ever used (or seen) maxwell, vray pp/RT, fry render? :rolleyes:
which one is unbiased and which is "consistent"?
I just prefer to leave you in the mental cloud (TM) (http://cloud.mental.com/direct/) , it's more entertaining :)
P
dagon1978
10-27-2009, 02:24 AM
I just prefer to leave you in the doubt-cloud state, it's more entertaining :)
P
oh my god, i think i'm not gonna sleep tonight :beer:
:wavey:
P
oh my god, i think i'm not gonna sleep tonight :beer:
:wavey:
ThE_JacO
10-27-2009, 02:45 AM
so maxwell, vray ppt/RT, fry are unbiased or "consistent"? tell me paolo
and also, tell me how you can talk about something you never tried, because you never tried iray right? :rolleyes:
i think i gonna use this as my next firm
"just look at how the image refines and gets brighter with each iteration: that's because it picks up more light, thing that wouldn't happen with an unbiased renderer." paolo berto
too funny paolo, too funny
A) both you and Paolo should really drop these schoolyard bickering sessions. It doesn't lend anything to either member's credibility.
B) before you quote something that is correct as "funny" you might want to look a bit into progressive estimators and the various montecarlo and quasi-montecarlo papers, and see for yourself that a change in the averaged luminance of the result (which does happen in iray) points to a consistent, not unbiased estimator (if MI's own PDFs weren't enough for you to trust somebody on that call), which basically means the result of the estimator before the limit can be statistically off-mark compared to pure unbiased algorithms.
The fact that a number of software houses abuse and confuse the latest buzzword (unbiased) doesn't suddenly legitimate mis-use or invalidate a correct technical note. Much the same of when everybody and their dog were mis-using words like radiosity, global illumination and montecarlo all over the place and, usually, completely in the wrong context.
You might also want to look into the montecarlo lawsuit (Pixar's) that makes it so that MRay is one of the few rendering engines out there that is on its toes in terms of implementations and can't use the same algorithms used in some other engines (which are sometimes in violation of a patent but ignored for the time being).
Get some forum manners, get a clue, then come back to the discussion :)
Cheers,
Raff
and also, tell me how you can talk about something you never tried, because you never tried iray right? :rolleyes:
i think i gonna use this as my next firm
"just look at how the image refines and gets brighter with each iteration: that's because it picks up more light, thing that wouldn't happen with an unbiased renderer." paolo berto
too funny paolo, too funny
A) both you and Paolo should really drop these schoolyard bickering sessions. It doesn't lend anything to either member's credibility.
B) before you quote something that is correct as "funny" you might want to look a bit into progressive estimators and the various montecarlo and quasi-montecarlo papers, and see for yourself that a change in the averaged luminance of the result (which does happen in iray) points to a consistent, not unbiased estimator (if MI's own PDFs weren't enough for you to trust somebody on that call), which basically means the result of the estimator before the limit can be statistically off-mark compared to pure unbiased algorithms.
The fact that a number of software houses abuse and confuse the latest buzzword (unbiased) doesn't suddenly legitimate mis-use or invalidate a correct technical note. Much the same of when everybody and their dog were mis-using words like radiosity, global illumination and montecarlo all over the place and, usually, completely in the wrong context.
You might also want to look into the montecarlo lawsuit (Pixar's) that makes it so that MRay is one of the few rendering engines out there that is on its toes in terms of implementations and can't use the same algorithms used in some other engines (which are sometimes in violation of a patent but ignored for the time being).
Get some forum manners, get a clue, then come back to the discussion :)
Cheers,
Raff
dagon1978
10-27-2009, 02:49 AM
B) before you quote something that is correct as "funny" you might want to look a bit into progressive estimators and the various montecarlo and quasi-montecarlo papers, and see for yourself that a change in the averaged luminance of the result (which does happen in iray) points to a consistent, not unbiased estimator (if MI's own PDFs weren't enough for you to trust somebody on that call), which basically means the result of the estimator before the limit can be statistically off-mark compared to pure unbiased algorithms.
The fact that a number of software houses abuse and confuse the latest buzzword (unbiased) doesn't suddenly legitimate mis-use or invalidate a correct technical note. Much the same of when everybody and their dog were mis-using words like radiosity, global illumination and montecarlo all over the place and, usually, completely in the wrong context.
You might also want to look into the montecarlo lawsuit (Pixar's) that makes it so that MRay is one of the few rendering engines out there that is on its toes in terms of implementations and can't use the same algorithms used in some other engines (which are sometimes in violation of a patent but ignored for the time being).
Get some forum manners, get a clue, then come back to the discussion :)
Cheers,
Raff
so, maxwell, vray ppt/RT, fry render, are unbiased or consistent engine? :rolleyes:
The fact that a number of software houses abuse and confuse the latest buzzword (unbiased) doesn't suddenly legitimate mis-use or invalidate a correct technical note. Much the same of when everybody and their dog were mis-using words like radiosity, global illumination and montecarlo all over the place and, usually, completely in the wrong context.
You might also want to look into the montecarlo lawsuit (Pixar's) that makes it so that MRay is one of the few rendering engines out there that is on its toes in terms of implementations and can't use the same algorithms used in some other engines (which are sometimes in violation of a patent but ignored for the time being).
Get some forum manners, get a clue, then come back to the discussion :)
Cheers,
Raff
so, maxwell, vray ppt/RT, fry render, are unbiased or consistent engine? :rolleyes:
ThE_JacO
10-27-2009, 02:57 AM
Haven't used them so I wouldn't know, but the point remains that calling a chicken a duck and making fun of those calling it a chicken remains childish, confrontational, and ultimately wrong, no matter what marketing departments around the world think of it.
Answer me this one question, do you even know what Montecarlo means/is for, or what estimators are, or for that matter, what statistical effect of a bias and the biasing ruleset mean?
You're not discussing literature or the definition of art, you are discussing rather factual nomencaltures and algorithm implementations here. Arguing semantics doesn't get very far :)
Now, if for once in history somebody from MI could come in and shed some light and contribute to a forum, like every other company under the sun does on occasion, things would probably be easier :)
Answer me this one question, do you even know what Montecarlo means/is for, or what estimators are, or for that matter, what statistical effect of a bias and the biasing ruleset mean?
You're not discussing literature or the definition of art, you are discussing rather factual nomencaltures and algorithm implementations here. Arguing semantics doesn't get very far :)
Now, if for once in history somebody from MI could come in and shed some light and contribute to a forum, like every other company under the sun does on occasion, things would probably be easier :)
dagon1978
10-27-2009, 03:11 AM
Haven't used them so I wouldn't know, but the point remains that calling a chicken a duck and making fun of those calling it a chicken remains childish, confrontational, and ultimately wrong, no matter what marketing departments around the world think of it.
Answer me this one question, do you even know what Montecarlo means/is for, or what estimators are, or for that matter, what statistical effect of a bias and the biasing ruleset mean?
You're not discussing literature or the definition of art, you are discussing rather factual nomencaltures and algorithm implementations here. Arguing semantics doesn't get very far :)
Now, if for once in history somebody from MI could come in and shed some light and contribute to a forum, like every other company under the sun does on occasion, things would probably be easier :)
bingo! you haven't used them! like paolo
i can tell you, they're working like iray in this field, so they're all "consistent" engine, but they claim to be "unbiased"
so i was referring to "the common sense" of unbiased engine
has you can see here (http://forums.cgsociety.org/showpost.php?p=6139548&postcount=67) :
All the current unbiased renderers are not GPU accelerated and in fact very slow, having one that is fast enough for 'interactive' use will be a new thing.
iray aswell is *not* unbiased, it's as "consistent".
p
paolo was trying to say it is different from other "unbiased" renderers (or, if not, i can't understand his reply)
but it's not
so if you like to be so pernickety iray is a consistent path tracer, like maxwell
or, in the "common sense" it's an "unbiased" path tracer, like maxwell
you see, nothing change
p.s.
the fact that MI is calling it in the right way is a good or a bad thing? :rolleyes:
Answer me this one question, do you even know what Montecarlo means/is for, or what estimators are, or for that matter, what statistical effect of a bias and the biasing ruleset mean?
You're not discussing literature or the definition of art, you are discussing rather factual nomencaltures and algorithm implementations here. Arguing semantics doesn't get very far :)
Now, if for once in history somebody from MI could come in and shed some light and contribute to a forum, like every other company under the sun does on occasion, things would probably be easier :)
bingo! you haven't used them! like paolo
i can tell you, they're working like iray in this field, so they're all "consistent" engine, but they claim to be "unbiased"
so i was referring to "the common sense" of unbiased engine
has you can see here (http://forums.cgsociety.org/showpost.php?p=6139548&postcount=67) :
All the current unbiased renderers are not GPU accelerated and in fact very slow, having one that is fast enough for 'interactive' use will be a new thing.
iray aswell is *not* unbiased, it's as "consistent".
p
paolo was trying to say it is different from other "unbiased" renderers (or, if not, i can't understand his reply)
but it's not
so if you like to be so pernickety iray is a consistent path tracer, like maxwell
or, in the "common sense" it's an "unbiased" path tracer, like maxwell
you see, nothing change
p.s.
the fact that MI is calling it in the right way is a good or a bad thing? :rolleyes:
ThE_JacO
10-27-2009, 03:31 AM
bingo! you haven't used them! like paolo
i can tell you, they're working like iray in this field, so they're all "consistent" engine, but they claim to be "unbiased"
so i was referring to "the common sense" of unbiased engine
Since you don't know what consistent or unbiased mean, apparently, how do you know what they actually are?
Are you taking at heart the statements of people whose opinion you've been busy belittling for months to make that call? That would be ironic :)
you see, nothing change
Lots change.
Lemme quote you:
"just look at how the image refines and gets brighter with each iteration: that's because it picks up more light, thing that wouldn't happen with an unbiased renderer." paolo berto
too funny paolo, too funny
That implies that Paolo's assertion was wrong, and that he's funny because of it. It wasn't funny, and it was correct.
p.s.
the fact that MI is calling it in the right way is a good or a bad thing? :rolleyes:
Where did I bash MI and why would you think I care what anybody thinks of the company or their software?
I'm addressing your manners and the fact you accuse many people of incompetence while attempting a defense of a software you use, when you clearly don't have the competence or understanding of the matters at hand to put yourself on the moral high ground you try to occupy.
Does that make things any clearer?
Now, again, stop bickering with everybody and being a fanboy thinking MI needs you to shield them from the rest of the world. They don't need that, and I'm not talking of iRay compared to maxwell or whatever other engine.
I am not comparing iRay to anything, I'm not defending Paolo (who's been warned for his forum manners too afaik), and I'm not saying anything about maxwell or any other engine. I'm asking you to make an effort towards better forum manners and being more consistent in your technical contributions.
Is that enough context to my post now? :)
i can tell you, they're working like iray in this field, so they're all "consistent" engine, but they claim to be "unbiased"
so i was referring to "the common sense" of unbiased engine
Since you don't know what consistent or unbiased mean, apparently, how do you know what they actually are?
Are you taking at heart the statements of people whose opinion you've been busy belittling for months to make that call? That would be ironic :)
you see, nothing change
Lots change.
Lemme quote you:
"just look at how the image refines and gets brighter with each iteration: that's because it picks up more light, thing that wouldn't happen with an unbiased renderer." paolo berto
too funny paolo, too funny
That implies that Paolo's assertion was wrong, and that he's funny because of it. It wasn't funny, and it was correct.
p.s.
the fact that MI is calling it in the right way is a good or a bad thing? :rolleyes:
Where did I bash MI and why would you think I care what anybody thinks of the company or their software?
I'm addressing your manners and the fact you accuse many people of incompetence while attempting a defense of a software you use, when you clearly don't have the competence or understanding of the matters at hand to put yourself on the moral high ground you try to occupy.
Does that make things any clearer?
Now, again, stop bickering with everybody and being a fanboy thinking MI needs you to shield them from the rest of the world. They don't need that, and I'm not talking of iRay compared to maxwell or whatever other engine.
I am not comparing iRay to anything, I'm not defending Paolo (who's been warned for his forum manners too afaik), and I'm not saying anything about maxwell or any other engine. I'm asking you to make an effort towards better forum manners and being more consistent in your technical contributions.
Is that enough context to my post now? :)
dagon1978
10-27-2009, 03:47 AM
Since you don't know what consistent or unbiased mean, apparently, how do you know what they actually are?
Are you taking at heart the statements of people whose opinion you've been busy belittling for months to make that call? That would be ironic :)
Lots change.
Lemme quote you:
That implies that Paolo's assertion was wrong, and that he's funny because of it. It wasn't funny, and it was correct.
if it was correct, it implies that maxwell/etc are not unbiased renderers, but consistent renderers, so again, if you wanna call it consistent call it this way, unbiased it's much more common
it doesnt change the thing that iray is not different from the others
Where did I bash MI and why would you think I care what anybody thinks of the company or their software?
where did i say you bashed MI? :rolleyes:
I'm addressing your manners and the fact you accuse many people of incompetence
many? :curious:
while attempting a defense of a software you use...
:curious:
i'm writing in a public forum, if you dont like what i write dont read it, if what i write it's violating some policy then ban me, but i dont accept anything else from you
Are you taking at heart the statements of people whose opinion you've been busy belittling for months to make that call? That would be ironic :)
Lots change.
Lemme quote you:
That implies that Paolo's assertion was wrong, and that he's funny because of it. It wasn't funny, and it was correct.
if it was correct, it implies that maxwell/etc are not unbiased renderers, but consistent renderers, so again, if you wanna call it consistent call it this way, unbiased it's much more common
it doesnt change the thing that iray is not different from the others
Where did I bash MI and why would you think I care what anybody thinks of the company or their software?
where did i say you bashed MI? :rolleyes:
I'm addressing your manners and the fact you accuse many people of incompetence
many? :curious:
while attempting a defense of a software you use...
:curious:
i'm writing in a public forum, if you dont like what i write dont read it, if what i write it's violating some policy then ban me, but i dont accept anything else from you
ThE_JacO
10-27-2009, 04:29 AM
i'm writing in a public forum, if you dont like what i write dont read it, if what i write it's violating some policy then ban me, but i dont accept anything else from you
It is actually in violation of several of them. You could consider this a warning then if you wish.
Drop the personal attacks, ignore people who tick you off and let them do the same (or let us deal with them if they persist), and stop "calling people out", and everybody's life will be a lot easier.
You don't need to "accept" anything else from me other than a fair warning, which is what I did, including articulating what is against policies/forum etiquette insofar.
Have a lovely day :)
It is actually in violation of several of them. You could consider this a warning then if you wish.
Drop the personal attacks, ignore people who tick you off and let them do the same (or let us deal with them if they persist), and stop "calling people out", and everybody's life will be a lot easier.
You don't need to "accept" anything else from me other than a fair warning, which is what I did, including articulating what is against policies/forum etiquette insofar.
Have a lovely day :)
Buexe
10-27-2009, 04:32 AM
will iray work on my iphone and can I buy it in itunes?
ibuexe
ibuexe
wurp
10-27-2009, 10:52 AM
All these technical detail aside, if you look at the video presentation here: http://www.youtube.com/watch?v=TQsXualxLVs
you can see how iray refines the render, it does appear to go a bit brighter as it refines (due to the noise) but then so does maxwell and all other renderers that I've tried that are "unbiased" so where is the REAL difference here?
The quote i posted earlier from the iray FAQ earlier mentions consitent, I'm not sure anymore this means anything regarding the technology iray uses or if it's just a marketing word.
There was a discussion about this on a beta email list I was on, the conclusion was that an unbiased renderer produces the correct result starting off with a noisy image that becomes smoother and smoother the more time/samples you give it to render, consistent on the other hand means that during the rendering process the image can have various artefacts like photon blotches etc, but given enough time it also produces the correct image.
there is a quote in this paper here:
http://graphics.stanford.edu/papers/veach_thesis/thesis.pdf
"Intuitively, an unbiased estimator computes the correct answer, on average. A biased
estimator computes the wrong answer, on average. However, if a biased estimator is also
consistent, then the average error can be made arbitrarily small by increasing the sample
size."
According to this iray IS an unbiased renderer because it behaves exactly the way described in the paper, iray produces a noisy image at first than gets refined, and does not show interpolation artefacts or anything like that which a "consistent" renderer would.
you can see how iray refines the render, it does appear to go a bit brighter as it refines (due to the noise) but then so does maxwell and all other renderers that I've tried that are "unbiased" so where is the REAL difference here?
The quote i posted earlier from the iray FAQ earlier mentions consitent, I'm not sure anymore this means anything regarding the technology iray uses or if it's just a marketing word.
There was a discussion about this on a beta email list I was on, the conclusion was that an unbiased renderer produces the correct result starting off with a noisy image that becomes smoother and smoother the more time/samples you give it to render, consistent on the other hand means that during the rendering process the image can have various artefacts like photon blotches etc, but given enough time it also produces the correct image.
there is a quote in this paper here:
http://graphics.stanford.edu/papers/veach_thesis/thesis.pdf
"Intuitively, an unbiased estimator computes the correct answer, on average. A biased
estimator computes the wrong answer, on average. However, if a biased estimator is also
consistent, then the average error can be made arbitrarily small by increasing the sample
size."
According to this iray IS an unbiased renderer because it behaves exactly the way described in the paper, iray produces a noisy image at first than gets refined, and does not show interpolation artefacts or anything like that which a "consistent" renderer would.
wurp
10-27-2009, 12:25 PM
I think this is wrong, from what I read the lawsuit was ages ago and MI and everyone else are now free to implement montecarlo sampling instead of the quasi monte carlo type MI uses now, which does actually produce very ugly sampling patterns...
You might also want to look into the montecarlo lawsuit (Pixar's) that makes it so that MRay is one of the few rendering engines out there that is on its toes in terms of implementations and can't use the same algorithms used in some other engines (which are sometimes in violation of a patent but ignored for the time being).
Raff
You might also want to look into the montecarlo lawsuit (Pixar's) that makes it so that MRay is one of the few rendering engines out there that is on its toes in terms of implementations and can't use the same algorithms used in some other engines (which are sometimes in violation of a patent but ignored for the time being).
Raff
simonenastasi
10-27-2009, 01:01 PM
Sincerely, who the hell cares? Results will speak, not the "unbiased" label...
Usability? All purpose? Production ready? Ok for both big size stills and flicker free animations? Will the intrinsic limits (3d moblur, materials etc) be surpassed by the advantages? and so on...
Usability? All purpose? Production ready? Ok for both big size stills and flicker free animations? Will the intrinsic limits (3d moblur, materials etc) be surpassed by the advantages? and so on...
pix3lm0nk
10-27-2009, 01:29 PM
It is actually in violation of several of them. You could consider this a warning then if you wish.
Drop the personal attacks, ignore people who tick you off and let them do the same (or let us deal with them if they persist), and stop "calling people out", and everybody's life will be a lot easier.
You don't need to "accept" anything else from me other than a fair warning, which is what I did, including articulating what is against policies/forum etiquette insofar.
Have a lovely day :)
I hope that goes for a couple others in this thread who consistantly attack and attempt to derail any constructive threads regarding Mental Ray/Images, or any topic for that matter.
;)
Drop the personal attacks, ignore people who tick you off and let them do the same (or let us deal with them if they persist), and stop "calling people out", and everybody's life will be a lot easier.
You don't need to "accept" anything else from me other than a fair warning, which is what I did, including articulating what is against policies/forum etiquette insofar.
Have a lovely day :)
I hope that goes for a couple others in this thread who consistantly attack and attempt to derail any constructive threads regarding Mental Ray/Images, or any topic for that matter.
;)
ha-dou-ken
12-22-2009, 06:48 PM
Did a search for this with no results:
"
iray® is the world’s first interactive and physically correct, photorealistic rendering solution. With iray, users can quickly create life-like images of their creations by using intuitive, real world approaches and interactively exploring their results through the processing power of NVIDIA® graphics processing units (GPUs).
As a ready to integrate solution with ties to industry standards, iray enables software developers to rapidly add exceptionally fast and productive photorealism to their products. With iray, interactive realism is now a reality.
iray accelerates the creative process by accurately simulating the physical world, enabling designers, engineers and consumers to use materials, lighting and settings that truly resemble what they experience every day. iray makes this process interactive by progressively refining the image until the desired detail is achieved, becoming increasingly faster as more GPUs are employed, enabling the play of light, shadow and reflection to be studied.
As part of mental ray® 3.8 – the predominant ray trace renderer used on feature films and within the leading design and content creation tools of Autodesk, Dassault Systèmes and Parametric Technology – iray is poised to become an industry standard for visually simulating the physical world. Companies can join this upcoming convergence by adding iray directly to their products (directly or with mental ray) or by utilizing iray within their RealityServer® applications."
VIDEO DEMO:
http://www.mentalimages.com/index.php?id=634
It seems similar to Vray's equivalent.
"
iray® is the world’s first interactive and physically correct, photorealistic rendering solution. With iray, users can quickly create life-like images of their creations by using intuitive, real world approaches and interactively exploring their results through the processing power of NVIDIA® graphics processing units (GPUs).
As a ready to integrate solution with ties to industry standards, iray enables software developers to rapidly add exceptionally fast and productive photorealism to their products. With iray, interactive realism is now a reality.
iray accelerates the creative process by accurately simulating the physical world, enabling designers, engineers and consumers to use materials, lighting and settings that truly resemble what they experience every day. iray makes this process interactive by progressively refining the image until the desired detail is achieved, becoming increasingly faster as more GPUs are employed, enabling the play of light, shadow and reflection to be studied.
As part of mental ray® 3.8 – the predominant ray trace renderer used on feature films and within the leading design and content creation tools of Autodesk, Dassault Systèmes and Parametric Technology – iray is poised to become an industry standard for visually simulating the physical world. Companies can join this upcoming convergence by adding iray directly to their products (directly or with mental ray) or by utilizing iray within their RealityServer® applications."
VIDEO DEMO:
http://www.mentalimages.com/index.php?id=634
It seems similar to Vray's equivalent.
AJ
12-22-2009, 08:33 PM
http://forums.cgsociety.org/showthread.php?f=59&t=812125&highlight=iray
ha-dou-ken
12-22-2009, 08:35 PM
Thanks. Nuke this thread please.
CGTalk Moderation
12-22-2009, 08:35 PM
This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.
vBulletin v3.0.5, Copyright ©2000-2012, Jelsoft Enterprises Ltd.