ettore
04-02-2003, 08:31 PM
where i can find some tutorial for c4d that show me how i can simulate the light from the windows with radiosity on?
View Full Version : window light
|
|
LucentDreams
04-02-2003, 08:50 PM
Step one get an image with a window in it
Step two put image in luminosity channel
Step three apply material to sky
Step four render
One note and HDRI image with a window rendered from 8.1 will look better since there is a more dynamic range between the light outside the window and the light inside the window within the image itself.
Step two put image in luminosity channel
Step three apply material to sky
Step four render
One note and HDRI image with a window rendered from 8.1 will look better since there is a more dynamic range between the light outside the window and the light inside the window within the image itself.
FusionDG
04-02-2003, 10:12 PM
Hi ettore,
Are you asking how to light an interior scene just from sunlight coming in through a window?
Are you asking how to light an interior scene just from sunlight coming in through a window?
FusionDG
04-03-2003, 04:33 PM
ettore wrote on 04-03-2003 06:08 AM:
Yes,it's what i want to do,a room only illuminated with light from the sun.
Can you send me a tutorial for this?
Hi ettore,
It's funny you ask this because I am trying to do the very same thing. Last night i was experimenting with a small 20' x 20' enclosed room w/a ceiling, floor, 4 walls, and one window.
I then setup only one light, a direct spot light setup to be distant w/hard shadows. I then started testing my scene to see what it took to get the room lit up just by the sun alone.
It seems harder than I imagined unfortunately. I ended up getting the light up to 1000% brightness and the radiosity strength to about 350%. That was getting ok results, but then things were getting real blotchy. For me it seems whenever I put the radiosity strength above 150%, things starting getting blotchy, like the colors run hot and have extreme results for me.
I did notice that the higher I set the Stochastic samples at, the better and brighter lit the scene was which makes sense. I think I ended up getting the Stoch setting up to 800 or so. And my D Depth was at 10.
So this is trying for me also. What I would like to have is this:
I do architectural rendering, so for me it seems that I need the radiosity strength at around 100 –150%, and no higher. This is because when I start throwing other lamps and can lights in there, the scene’s lighting gives me results that are hard to predict when the strength gets above 200 –300%.
What I think would be helpful is we could have a thread with a simple model setup as I mentioned above, then apply textures to the walls & ceiling to make them white, and add a floor texture like a wood floor. Then let everyone do his best simulation for a natural daylight setting. I and others would have alot to gain from such an experiment.
~ paul
Yes,it's what i want to do,a room only illuminated with light from the sun.
Can you send me a tutorial for this?
Hi ettore,
It's funny you ask this because I am trying to do the very same thing. Last night i was experimenting with a small 20' x 20' enclosed room w/a ceiling, floor, 4 walls, and one window.
I then setup only one light, a direct spot light setup to be distant w/hard shadows. I then started testing my scene to see what it took to get the room lit up just by the sun alone.
It seems harder than I imagined unfortunately. I ended up getting the light up to 1000% brightness and the radiosity strength to about 350%. That was getting ok results, but then things were getting real blotchy. For me it seems whenever I put the radiosity strength above 150%, things starting getting blotchy, like the colors run hot and have extreme results for me.
I did notice that the higher I set the Stochastic samples at, the better and brighter lit the scene was which makes sense. I think I ended up getting the Stoch setting up to 800 or so. And my D Depth was at 10.
So this is trying for me also. What I would like to have is this:
I do architectural rendering, so for me it seems that I need the radiosity strength at around 100 –150%, and no higher. This is because when I start throwing other lamps and can lights in there, the scene’s lighting gives me results that are hard to predict when the strength gets above 200 –300%.
What I think would be helpful is we could have a thread with a simple model setup as I mentioned above, then apply textures to the walls & ceiling to make them white, and add a floor texture like a wood floor. Then let everyone do his best simulation for a natural daylight setting. I and others would have alot to gain from such an experiment.
~ paul
AdamT
04-03-2003, 05:56 PM
Good idea, Paul. I'd be up for that experiment too.
handige_harrie
04-03-2003, 06:46 PM
I'm willing to join in that experiment. I did some tests for my own interior-scene. And it's pretty tough with all the variables:
-light
-mesh (how big are the windows)
-materials
-radiosity
Still not perfect, and test-rendering takes a lot of time unfortunately.
I am currently rendering an example of my interior with my 'best' settings so far. Expect an uploaded picture in about 2-3 hours :rolleyes:.
-light
-mesh (how big are the windows)
-materials
-radiosity
Still not perfect, and test-rendering takes a lot of time unfortunately.
I am currently rendering an example of my interior with my 'best' settings so far. Expect an uploaded picture in about 2-3 hours :rolleyes:.
FusionDG
04-03-2003, 07:11 PM
That would be great of we could get a few people to run tests on the same model! :bounce:
I don't have too much trouble w/exterior lighting, but interior is so much tougher for me for some reason.
If I could just nail down lighting for an interior scene that was lit solely by direct sun, that would be a huge step FWD for me.
Any other takers?
Originally posted by handige_harrie
Expect an uploaded picture in about 2-3 hours :rolleyes:.
Looking FWD to it HH
I don't have too much trouble w/exterior lighting, but interior is so much tougher for me for some reason.
If I could just nail down lighting for an interior scene that was lit solely by direct sun, that would be a huge step FWD for me.
Any other takers?
Originally posted by handige_harrie
Expect an uploaded picture in about 2-3 hours :rolleyes:.
Looking FWD to it HH
Cris-Palomino
04-04-2003, 08:36 AM
I'll try my hand at it, too.
Cris
Cris
Phasmatis
04-04-2003, 09:23 AM
Here's my attempt. I used one distant light 80% brightness with a slight yellow colour and soft shadows (shadow map res = 750x750).
Pink Room (http://www.planetdeusex.com/tnm/phasimages/pink.jpg):
Settings:
Strength = 140%
Accuracy = 64%
PrePass = 1/1
Diffue Depth = 7
Stoch Samples = 200
Min Res = 16
Max Res = 18
Recompute = Always
Render Time 1.27mins
Brown Room (http://www.planetdeusex.com/tnm/phasimages/brownroom.jpg):
Same settings as above but with a darker texture.
Pink Room2 (http://www.planetdeusex.com/tnm/phasimages/pink2.jpg):
Settings:
Strength = 140%
Accuracy = 64%
PrePass = 1/1
Diffue Depth = 5
Stoch Samples = 500
Min Res = 32
Max Res = 34
Recompute = Always
Render Time about 5mins
Pink Room (http://www.planetdeusex.com/tnm/phasimages/pink.jpg):
Settings:
Strength = 140%
Accuracy = 64%
PrePass = 1/1
Diffue Depth = 7
Stoch Samples = 200
Min Res = 16
Max Res = 18
Recompute = Always
Render Time 1.27mins
Brown Room (http://www.planetdeusex.com/tnm/phasimages/brownroom.jpg):
Same settings as above but with a darker texture.
Pink Room2 (http://www.planetdeusex.com/tnm/phasimages/pink2.jpg):
Settings:
Strength = 140%
Accuracy = 64%
PrePass = 1/1
Diffue Depth = 5
Stoch Samples = 500
Min Res = 32
Max Res = 34
Recompute = Always
Render Time about 5mins
AdamT
04-04-2003, 02:17 PM
I think the one with higher samples will look better. Could you post the model somewhere so we're all working with the same scene? If not could you post the diminesions of the room? The size of the object will have a significant impact on radiosity settings, so direct comparison won't be possible if we're all working with different-size models.
Phasmatis
04-04-2003, 03:07 PM
Yeah pink room 2 does look a lot better, less blotches.
Room file. (http://www.planetdeusex.com/tnm/phasimages/daylight.zip)
It has the radiosity settings of Pink Room2.
It's an R8 by the way.
Room file. (http://www.planetdeusex.com/tnm/phasimages/daylight.zip)
It has the radiosity settings of Pink Room2.
It's an R8 by the way.
handige_harrie
04-04-2003, 04:05 PM
Rendering took a bit longer :shrug: so I stopped about halfway last night (I had to get some sleep ;)) and I started again about 4 hours ago (with an alpha map, so I didn't need to rerender the whole image). Still busy at the moment...
Settings are too high to call it a simple test, but the result is already looking very good, except for some flaws that are easily fixed.
I will upload the result when it's finished :).
Reaction to Phasmatis:
1) I can't see where the light is coming from. Is it coming thorugh the window that is visible in the image? If so, why are there two recantuglar illuminated spots on the sides?
2)you took a pink and brown room for an example. If we are trying to simulate real-life lighting, we should use a white room. In that case you can see shadow and light much better. Another point is the excessive color-blending you get with radiosity strength set above 100%. If you would have a white room with a wooden (beige) floor and set radiosity strength to 140% probably the whole room will be beige. This is not realisitic.
Settings are too high to call it a simple test, but the result is already looking very good, except for some flaws that are easily fixed.
I will upload the result when it's finished :).
Reaction to Phasmatis:
1) I can't see where the light is coming from. Is it coming thorugh the window that is visible in the image? If so, why are there two recantuglar illuminated spots on the sides?
2)you took a pink and brown room for an example. If we are trying to simulate real-life lighting, we should use a white room. In that case you can see shadow and light much better. Another point is the excessive color-blending you get with radiosity strength set above 100%. If you would have a white room with a wooden (beige) floor and set radiosity strength to 140% probably the whole room will be beige. This is not realisitic.
AdamT
04-04-2003, 04:18 PM
Originally posted by Phasmatis
Yeah pink room 2 does look a lot better, less blotches.
Room file. (http://www.planetdeusex.com/tnm/phasimages/daylight.zip)
It has the radiosity settings of Pink Room2.
It's an R8 by the way.
Cool, I'll have a crack at it today or sometime this weekend.
Yeah pink room 2 does look a lot better, less blotches.
Room file. (http://www.planetdeusex.com/tnm/phasimages/daylight.zip)
It has the radiosity settings of Pink Room2.
It's an R8 by the way.
Cool, I'll have a crack at it today or sometime this weekend.
FusionDG
04-04-2003, 05:33 PM
Hey Phasmatis, thx for posting the file. :applause:
If everyone works off of this same file, then we can really compare things apples to apples like Adam said.
I had the same Q's/thoughts as HH had:
Originally posted by handige_harrie
1) I can't see where the light is coming from. Is it coming thorugh the window that is visible in the image? If so, why are there two recantuglar illuminated spots on the sides?
2)a]you took a pink and brown room for an example. If we are trying to simulate real-life lighting, we should use a white room. In that case you can see shadow and light much better.
b]Another point is the excessive color-blending you get with radiosity strength set above 100%. If you would have a white room with a wooden (beige) floor and set radiosity strength to 140% probably the whole room will be beige. This is not realisitic.
IMHOI would think it best to use white materials on the walls/ceiling at first. No specular/diffuse, just to get pure results. And keep the rad strength as close to 100% for the reason that HH mentions above.
Once we get the sunlight lit room looking good in it's purest form, then maybe we can throw a red wall in there or some wood floors to see how the color bounces around.
Thx again for the file. I'll definitely be messing around with it this weekend! :thumbsup:
If everyone works off of this same file, then we can really compare things apples to apples like Adam said.
I had the same Q's/thoughts as HH had:
Originally posted by handige_harrie
1) I can't see where the light is coming from. Is it coming thorugh the window that is visible in the image? If so, why are there two recantuglar illuminated spots on the sides?
2)a]you took a pink and brown room for an example. If we are trying to simulate real-life lighting, we should use a white room. In that case you can see shadow and light much better.
b]Another point is the excessive color-blending you get with radiosity strength set above 100%. If you would have a white room with a wooden (beige) floor and set radiosity strength to 140% probably the whole room will be beige. This is not realisitic.
IMHOI would think it best to use white materials on the walls/ceiling at first. No specular/diffuse, just to get pure results. And keep the rad strength as close to 100% for the reason that HH mentions above.
Once we get the sunlight lit room looking good in it's purest form, then maybe we can throw a red wall in there or some wood floors to see how the color bounces around.
Thx again for the file. I'll definitely be messing around with it this weekend! :thumbsup:
Phasmatis
04-04-2003, 05:41 PM
I have no idea why there are two bright squares at the sides of the shadows since there is only one light coming through the window, if anyone can explain it that would be great.
The strength and be turned down quite a bit especially if you turn up the diffuse depth, which you can see in the first pink room picture... It's way too bright.
I wanted to use different colours to see if the radiosity settings translated well with different textures but perhaps it is better to just use a white room.
The strength and be turned down quite a bit especially if you turn up the diffuse depth, which you can see in the first pink room picture... It's way too bright.
I wanted to use different colours to see if the radiosity settings translated well with different textures but perhaps it is better to just use a white room.
handige_harrie
04-04-2003, 07:07 PM
At last...after about 8 hours in total :D but that's mostly due to high poly count and heavy procedurals.
http://members.home.nl/hcgl.hermans/designkeuken04.jpg
Flaw: the ceiling is too bright were the kitchen touches the ceiling (because of the many bounces of light). This is easily fixed by turning down the GI generate percentage (40-50%) of the kitchen materials (the top of the kitchen and the edge that touches the ceiling)
Easy trick to prevent excesive color-bleeding is turning down the GI saturation of the materials (30-40%)
Example using Phasmatics scene coming up in a few minutes :)
http://members.home.nl/hcgl.hermans/designkeuken04.jpg
Flaw: the ceiling is too bright were the kitchen touches the ceiling (because of the many bounces of light). This is easily fixed by turning down the GI generate percentage (40-50%) of the kitchen materials (the top of the kitchen and the edge that touches the ceiling)
Easy trick to prevent excesive color-bleeding is turning down the GI saturation of the materials (30-40%)
Example using Phasmatics scene coming up in a few minutes :)
handige_harrie
04-04-2003, 08:24 PM
http://members.home.nl/hcgl.hermans/daylight01.jpg
Rendering took 2min25sec on P4 2.0@2.66 w/ 768MB pc2700.
download daylight_h_h_version.c4d (http://members.home.nl/hcgl.hermans/daylight_h_h_version.c4d)
Rendering took 2min25sec on P4 2.0@2.66 w/ 768MB pc2700.
download daylight_h_h_version.c4d (http://members.home.nl/hcgl.hermans/daylight_h_h_version.c4d)
AdamT
04-04-2003, 08:44 PM
Both of those renders look way too flat to me. Are you sure your light is set to cast shadows?
handige_harrie
04-04-2003, 08:49 PM
Yes, getting nice contrast and not getting too dark spaces is the problem.
Of course the lights (both single light scenes) cast (hard) shadows, you can see that ;).
I'd say, download the scene and start testing :)
Of course the lights (both single light scenes) cast (hard) shadows, you can see that ;).
I'd say, download the scene and start testing :)
FusionDG
04-04-2003, 09:35 PM
00:01:18 on my Dual AMD 2000+ MP 512 MB RAM
I didn't have the sky object BMP so it's white [i kinda prefer this anyways for now :D]
Everything else is straight from HH's file as he left the file.
http://fusiondesigngroup.com/pics/sun0001.jpg
At the risk of sounding biased, IMHO I think my image shows the shadows and depth better even though I didn't change one lighting/GI parameter.
Why is that? :shrug:
Anyways, I'm rather happy w/the way this looks so far :D
I didn't have the sky object BMP so it's white [i kinda prefer this anyways for now :D]
Everything else is straight from HH's file as he left the file.
http://fusiondesigngroup.com/pics/sun0001.jpg
At the risk of sounding biased, IMHO I think my image shows the shadows and depth better even though I didn't change one lighting/GI parameter.
Why is that? :shrug:
Anyways, I'm rather happy w/the way this looks so far :D
FusionDG
04-04-2003, 09:49 PM
Well, ok, I guess the material on the sky object is what the variable was.
Here’s the result I got when I took off the material from the Sky object because I was getting tired of the warning everytime.
00:01:11 Dual AMD 2000+ MP 512 MB RAM
640 x 480 [ I like the 4:3 ratio]
http://fusiondesigngroup.com/pics/sun001.jpg
and then when I put the material back on the Sky object. I didn’t take into account the luminance properties of the material on the Sky object even though the BMP wasn’t present :D
http://fusiondesigngroup.com/pics/sun002.jpg
Here’s the result I got when I took off the material from the Sky object because I was getting tired of the warning everytime.
00:01:11 Dual AMD 2000+ MP 512 MB RAM
640 x 480 [ I like the 4:3 ratio]
http://fusiondesigngroup.com/pics/sun001.jpg
and then when I put the material back on the Sky object. I didn’t take into account the luminance properties of the material on the Sky object even though the BMP wasn’t present :D
http://fusiondesigngroup.com/pics/sun002.jpg
handige_harrie
04-04-2003, 10:21 PM
Argh stupid mistake :wip: that luminating sky was effectively hdri lighting (image based lighting), and that's why everyting is lit so evenly. This is what causes the 'flatness'.
And also dumb that I didn't include the sky texture :hmm:.
Looking much better already FusionDG :)
I will do some more testing tomorrow.
And also dumb that I didn't include the sky texture :hmm:.
Looking much better already FusionDG :)
I will do some more testing tomorrow.
FusionDG
04-04-2003, 10:28 PM
Originally posted by handige_harrie
Argh stupid mistake
Looking much better already FusionDG
Glad to know i'm not the only one that does them ;)
here's another with higher max res
back to 512 x 512 ‘cause can see room better
*altered the Max res to 60*
http://fusiondesigngroup.com/pics/sun05.jpg
I think i'm getting some light leaks cause those white spots in the corners aren't JPG comression. :shrug:
Argh stupid mistake
Looking much better already FusionDG
Glad to know i'm not the only one that does them ;)
here's another with higher max res
back to 512 x 512 ‘cause can see room better
*altered the Max res to 60*
http://fusiondesigngroup.com/pics/sun05.jpg
I think i'm getting some light leaks cause those white spots in the corners aren't JPG comression. :shrug:
flingster
04-04-2003, 11:09 PM
what are light leaks? i can see what they are of course but what causes them?
also the min and max settings...the documentation i'm confused about...they seem to be the reverse of what i'd expect them to do...can anybody give me a better understanding/interpretation of what they can do...what high or low settings would do...and what a high low setting region would be?
thanks.
also the min and max settings...the documentation i'm confused about...they seem to be the reverse of what i'd expect them to do...can anybody give me a better understanding/interpretation of what they can do...what high or low settings would do...and what a high low setting region would be?
thanks.
flingster
04-04-2003, 11:12 PM
oh yeah and FusionDG & handige_harrie the corner part where there is leak is to dusty looking. on the lighter coloured floor boards you get the light change on the wall...but i feel its just to bright. other than that...pretty damn good....that render time handige_harrie is friggin fast...gonna go look at your file in a minute
:thumbsup:
:thumbsup:
FusionDG
04-04-2003, 11:22 PM
Originally posted by flingster
what are light leaks? i can see what they are of course but what causes them?
also the min and max settings...the documentation i'm confused about...
this is the best explanation of min & max res that i have found, maybe you have seen it already though :shrug:, if not:
MV's radiosity toots (http://www.mvpny.com/RadTutMV/RadiosityTut1MV.html)
Light leaks i think can happen in C4D when you are connecting two planes w/out any depth.
The room that HH supplied in this file is a nice simple room [i'm not rippin here FYI, really apreciate the file :D] which is good in our case. But sometimes if the walls have depth [i.e., another plane to simulate depth such as for 2 x 4 stud thickness or such] , that can cut down the likelihood of light leaking in through the joints.
what are light leaks? i can see what they are of course but what causes them?
also the min and max settings...the documentation i'm confused about...
this is the best explanation of min & max res that i have found, maybe you have seen it already though :shrug:, if not:
MV's radiosity toots (http://www.mvpny.com/RadTutMV/RadiosityTut1MV.html)
Light leaks i think can happen in C4D when you are connecting two planes w/out any depth.
The room that HH supplied in this file is a nice simple room [i'm not rippin here FYI, really apreciate the file :D] which is good in our case. But sometimes if the walls have depth [i.e., another plane to simulate depth such as for 2 x 4 stud thickness or such] , that can cut down the likelihood of light leaking in through the joints.
FusionDG
04-04-2003, 11:24 PM
Originally posted by flingster
the corner part where there is leak is to dusty looking.
That's just simulating the real world to keep the CG image from looking too clean. I'll add some dust bunnies and cobwebs in the corners as well in my next rendering ;)
the corner part where there is leak is to dusty looking.
That's just simulating the real world to keep the CG image from looking too clean. I'll add some dust bunnies and cobwebs in the corners as well in my next rendering ;)
AdamT
04-05-2003, 02:22 PM
Here's my effort using a single area light in place of the distant light:
http://bellsouthpwp.net/A/d/AdamTrachtenberg/daylight1.jpg
And here are the settings I used:
http://bellsouthpwp.net/A/d/AdamTrachtenberg/Rad_Stngs.jpg
This scene obviously uses the sky texture, as I rendered it before reading the that last 4-5 posts. Render time was about 1 1/2 minutes on my dual 2.4Ghz Xeon.
http://bellsouthpwp.net/A/d/AdamTrachtenberg/daylight1.jpg
And here are the settings I used:
http://bellsouthpwp.net/A/d/AdamTrachtenberg/Rad_Stngs.jpg
This scene obviously uses the sky texture, as I rendered it before reading the that last 4-5 posts. Render time was about 1 1/2 minutes on my dual 2.4Ghz Xeon.
ThirdEye
04-05-2003, 02:38 PM
here's my attempt
http://www.cgtalk.com/attachment.php?s=&postid=425045
strenght 100%
accuracy 90%
prepass size 1/1
dd 10
stochastic samples 120
min res 3
max res 20
http://www.cgtalk.com/attachment.php?s=&postid=425045
strenght 100%
accuracy 90%
prepass size 1/1
dd 10
stochastic samples 120
min res 3
max res 20
AdamT
04-05-2003, 03:06 PM
Looks good Thirdeye, but hard to compare with the different textures and view angle.
ThirdEye
04-05-2003, 03:09 PM
The major difference between mine and the others is the area shadow that gives more realism to the pic and the higher DD (i usually use 10-20 dd for interior scenes)
AdamT
04-05-2003, 03:42 PM
Yep, the area shadows look nice. I'm afraid to ask what the render time was on that puppy.
ThirdEye
04-05-2003, 04:42 PM
Originally posted by AdamT
Yep, the area shadows look nice. I'm afraid to ask what the render time was on that puppy.
10 mins on my dual PIII 1ghz
Yep, the area shadows look nice. I'm afraid to ask what the render time was on that puppy.
10 mins on my dual PIII 1ghz
AdamT
04-05-2003, 04:49 PM
Originally posted by ThirdEye_01
10 mins on my dual PIII 1ghz
That's not bad at all. Definitely worth it.
10 mins on my dual PIII 1ghz
That's not bad at all. Definitely worth it.
flingster
04-05-2003, 09:20 PM
guy i went a bit overkill here. (took original room posted).
white room 89% brightness generate gi 85%
parallel spot round light 770% brightness, volumetric, noise illumination etc..(helped kill me render times) also light gel -fusion -blured sema +red-orange.
offset sky slightly - generate gi 79%
took friggin ages...i like the look...can't get shadow right...soft looks sh**, hard is...well too hard, area is a rubbish compromise really...again affects render time...when you see light through window you get a mix of hard and soft areas...also would be affected by say window glass...that said...it was a compromise test.
rad settings....
strength 120% (tried the 100% as people posted, but with noise on light, high brightness etc couldn't get room to light up very easily)
accuracy 80% (but worked fine down to about 69%)
prepass 1/1
diffuse depth 7 (had to lift from 5 originally...as i said with light type, noise, gel, brightness etc only was i could think to light level in room...i may stand corrected...but i tweaked and this setting worked.
sto samples 500 (had it around 420 up fine)
minres 29
maxres 71
let me know what i did wrong...can't give you an accurate render time really cos i was busy doing other stuff on pc at time..so ran in background at around 1hr30!!! jeez...2gig athlon..with gig ram.
some of my pretests ran at about 20mins...i tried so much stuff...can't give definite this killed it...that killed it...cos i got loads of settings on that would kill any render.
white room 89% brightness generate gi 85%
parallel spot round light 770% brightness, volumetric, noise illumination etc..(helped kill me render times) also light gel -fusion -blured sema +red-orange.
offset sky slightly - generate gi 79%
took friggin ages...i like the look...can't get shadow right...soft looks sh**, hard is...well too hard, area is a rubbish compromise really...again affects render time...when you see light through window you get a mix of hard and soft areas...also would be affected by say window glass...that said...it was a compromise test.
rad settings....
strength 120% (tried the 100% as people posted, but with noise on light, high brightness etc couldn't get room to light up very easily)
accuracy 80% (but worked fine down to about 69%)
prepass 1/1
diffuse depth 7 (had to lift from 5 originally...as i said with light type, noise, gel, brightness etc only was i could think to light level in room...i may stand corrected...but i tweaked and this setting worked.
sto samples 500 (had it around 420 up fine)
minres 29
maxres 71
let me know what i did wrong...can't give you an accurate render time really cos i was busy doing other stuff on pc at time..so ran in background at around 1hr30!!! jeez...2gig athlon..with gig ram.
some of my pretests ran at about 20mins...i tried so much stuff...can't give definite this killed it...that killed it...cos i got loads of settings on that would kill any render.
flingster
04-05-2003, 09:21 PM
oooh and it did look ok...before the 10k limit!!!! honest...:thumbsdow
AdamT
04-06-2003, 07:33 AM
http://bellsouthpwp.net/A/d/AdamTrachtenberg/beige_room2.jpg
http://bellsouthpwp.net/A/d/AdamTrachtenberg/Rad_Stngs2.jpg
http://bellsouthpwp.net/A/d/AdamTrachtenberg/Rad_Stngs2.jpg
LucentDreams
04-06-2003, 08:46 AM
Wow Adam nice render, Is there caustics in there?
Phasmatis
04-06-2003, 08:54 AM
They're all looking pretty darn cool. :)
I really like flingster's, is there any way you can show an uncompressed pic?
I think I'll have a play around with your settings
I really like flingster's, is there any way you can show an uncompressed pic?
I think I'll have a play around with your settings
flingster
04-06-2003, 12:34 PM
Adamt : love that...works really well...the only bit needs a little work is imho around the picture and that right hand corner...not sure how you woul lift it without affecting rest of look/quality though...its still so damn cool though...what did you do with the light?
Phasmatis:
90%compression
http://www.flingster.com/cgtalk/radio.jpg
and the file if you wanna check out what i'm talking about..warning though guys expect high render time...worth checking out light settings..if you wanna tweak stuff...let us know how ya get on...i recon you can get that dusk look easier than i got it though...but as i said got carried away.
http://www.flingster.com/cgtalk/daylight.c4d
have fun...:thumbsup:
Phasmatis:
90%compression
http://www.flingster.com/cgtalk/radio.jpg
and the file if you wanna check out what i'm talking about..warning though guys expect high render time...worth checking out light settings..if you wanna tweak stuff...let us know how ya get on...i recon you can get that dusk look easier than i got it though...but as i said got carried away.
http://www.flingster.com/cgtalk/daylight.c4d
have fun...:thumbsup:
Phasmatis
04-06-2003, 01:18 PM
That pic is great, I can't see any blotches or artifacts and thanks for posting the file. :thumbsup:
Here's what I got after playing around with it for a little while.
http://www.planetdeusex.com/tnm/phasimages/flingsterchange.jpg
The quality isn't as good as flingster's but it only took 11mins and 29secs to render.
Here's (http://www.planetdeusex.com/tnm/phasimages/flingsterchange.c4d) the file.
Here's what I got after playing around with it for a little while.
http://www.planetdeusex.com/tnm/phasimages/flingsterchange.jpg
The quality isn't as good as flingster's but it only took 11mins and 29secs to render.
Here's (http://www.planetdeusex.com/tnm/phasimages/flingsterchange.c4d) the file.
flingster
04-06-2003, 02:34 PM
Phasmatis : you clever guy...thanks for this...you got time down a hell of a lot...you can also now see the noise on the light alot better. i was thinking about trying 2 lights...basically on top of each other...to con into looking like one light...but being able to change shadow on each one to hard on one light, and soft on another...then offset one ever so slightly to give the hard/soft shadow i talked about earlier. Also did anybody play about with highlights very very low...didn't mean over the top...but to get the reality con we all want....or is this just out of order on my part?
AdamT
04-06-2003, 03:35 PM
Thanks Kai. No caustics in there, but I may add some to see what comes of it (other than really long render times :)).
Thanks too Flingster. The lighting is a single area light casting soft shadows. I had to bump the light's brightness to something like 500%. I probably need to raise the samples another 50-100 units to clear those artifacts on the frame, but render times are already pretty high at just over 2 hrs. Amazing how much longer the render times are when you add objects to this scene than they are with the basic room and light. Without the objects the render time is like 2-3 minutes!
BTW, in the above render the table and lamp are from the "interior 3D for cinema 4D" CD sold by Maxon, and the chair and photo/frame are by yours truly. The CD is really high quality, although there's not much diversity in the style of the objects.
Thanks too Flingster. The lighting is a single area light casting soft shadows. I had to bump the light's brightness to something like 500%. I probably need to raise the samples another 50-100 units to clear those artifacts on the frame, but render times are already pretty high at just over 2 hrs. Amazing how much longer the render times are when you add objects to this scene than they are with the basic room and light. Without the objects the render time is like 2-3 minutes!
BTW, in the above render the table and lamp are from the "interior 3D for cinema 4D" CD sold by Maxon, and the chair and photo/frame are by yours truly. The CD is really high quality, although there's not much diversity in the style of the objects.
handige_harrie
04-06-2003, 04:36 PM
I combined a lot of things seen here and came up with a result that I myself am very pleased with (for my interior scene):
http://members.home.nl/hcgl.hermans/daylight04.jpg
http://members.home.nl/hcgl.hermans/daylight_final.c4d
without the environment sphere tex.
http://members.home.nl/hcgl.hermans/daylight04.jpg
http://members.home.nl/hcgl.hermans/daylight_final.c4d
without the environment sphere tex.
handige_harrie
04-07-2003, 11:58 AM
With some furniture:
http://members.home.nl/hcgl.hermans/daylight05.jpg
PS use for brightness/contrast and hue.
http://members.home.nl/hcgl.hermans/daylight05.jpg
PS use for brightness/contrast and hue.
Coen
04-07-2003, 04:10 PM
Nice One Harry,
one small thingy though, the shadows on the two chairs seem a bit odd probably because they suggest that the light that casts them is positioned so low.
Nice volumetrics too,
later
Coen
one small thingy though, the shadows on the two chairs seem a bit odd probably because they suggest that the light that casts them is positioned so low.
Nice volumetrics too,
later
Coen
handige_harrie
04-07-2003, 05:33 PM
Originally posted by Coen
one small thingy though, the shadows on the two chairs seem a bit odd probably because they suggest that the light that casts them is positioned so low.
The light that casts the shadows is actualy the bright part on the floor. This is pure indirect lighting because of radiosity. I must say I'm surprised by the 'shadow'quality with these settings.
one small thingy though, the shadows on the two chairs seem a bit odd probably because they suggest that the light that casts them is positioned so low.
The light that casts the shadows is actualy the bright part on the floor. This is pure indirect lighting because of radiosity. I must say I'm surprised by the 'shadow'quality with these settings.
Coen
04-07-2003, 05:47 PM
Thought of that after I posted.
me stands correced :D
me stands correced :D
FusionDG
04-07-2003, 07:33 PM
The tests everyone is posting looks great. Some great ideas here!
I'll dable later on when I get some projects out of the way.
paul
I'll dable later on when I get some projects out of the way.
paul
flingster
04-08-2003, 02:18 PM
ok guys i've been playing about. a couple of things i've noticed on the light in my living room...as it streams through the window...near the window sill...the shadows are hard...but on the carpet a further distance you get more area shadow mixed with hard areas...if this makes sense...
so with this in mind i mentioned earlier a two light test...one with hard shadow...and the second with area...if you look at my earlier posting the shadow was area...which looks nice..but i'm not so sure its correct...and the eye picks up these subtle differences.
well i had a go at the two light theory...but can't really get what i want...(not spent to much time fiddling about though)...so some of you guys may want to have a go at it or may have an alternative idea to simulate the effect in my living room.
i have enclosed a couple of thinks to try an show what the hell i'm talking about...in the second link you have the large diagonal shadow along the wall...hard shadow....but when it hits the right hand wall...its more area.
http://www.photo.net/photodb/photo?photo_id=1408737
http://www.photo.net/photodb/photo?photo_id=1411668
also for those of ya who haven't seen this tut..its worth a read...not so much to this but still interesting.
http://www.warpedspace.org/tutorials.htm
so with this in mind i mentioned earlier a two light test...one with hard shadow...and the second with area...if you look at my earlier posting the shadow was area...which looks nice..but i'm not so sure its correct...and the eye picks up these subtle differences.
well i had a go at the two light theory...but can't really get what i want...(not spent to much time fiddling about though)...so some of you guys may want to have a go at it or may have an alternative idea to simulate the effect in my living room.
i have enclosed a couple of thinks to try an show what the hell i'm talking about...in the second link you have the large diagonal shadow along the wall...hard shadow....but when it hits the right hand wall...its more area.
http://www.photo.net/photodb/photo?photo_id=1408737
http://www.photo.net/photodb/photo?photo_id=1411668
also for those of ya who haven't seen this tut..its worth a read...not so much to this but still interesting.
http://www.warpedspace.org/tutorials.htm
AdamT
04-08-2003, 03:05 PM
Hi Flingster,
I'm not sure if there's really a difference in the hard/soft shadow in the second linked photo. I think it just looks softer closer to the camera. Sunlight outdoors shouldn't cast an area shadow in the CG sense, because an area shadow represents the spreading of the shadow due to light attenuation. I think an area shadow can look okay when sunlight is coming through a window because the contribution of indoor lighting and bounced light will soften the shadow in some spots. As I think about it, I guess the same can occur to a lesser extent outdoors.
I'm not sure if there's really a difference in the hard/soft shadow in the second linked photo. I think it just looks softer closer to the camera. Sunlight outdoors shouldn't cast an area shadow in the CG sense, because an area shadow represents the spreading of the shadow due to light attenuation. I think an area shadow can look okay when sunlight is coming through a window because the contribution of indoor lighting and bounced light will soften the shadow in some spots. As I think about it, I guess the same can occur to a lesser extent outdoors.
flingster
04-08-2003, 03:34 PM
yeah i know what you mean...but take a look at light coming in through your window...the shadows are hard...at the window sill area very defined....but when they reach the floor...they tend to be hard edged still...but with the area or softer look/feel inside the hard area....kinda fall off.
i'm not convinced that its the fact that they are closer to the camera....then again i'm probably wrong....its all about getting a realistic look....so how the do we go about it in cinema....the single room is turning out to be an excellent example...not just for radiosity testing but also shadows. yeah i know the area shadow is the wrong thing in a cg sense..but it was a fudge that looks ok...
all the images i posted are not ideal for what i was talking about but the second one comes closest...you also find because this is just a plain white room there is a difference between this and say handige_harrie room with shiney wood floor....as would be the results if you had a thick pile carpet...with say a blanket scrunched up on it and possibly a glass window...areas of shadow would look different and the part about a mix of shadow quality would come more into play...with hard areas...and then tapering/blending into soft/area look on parts of the image. its really hard to describe what i'm talking about without being able to post an image and say that it!! sorry for this...and hopefully you can get what i mean.
i just think the area and the hard shadow...just doesn't cut it for realism imho...but i don't know how to solve it in cinema...:wip:
i'm not convinced that its the fact that they are closer to the camera....then again i'm probably wrong....its all about getting a realistic look....so how the do we go about it in cinema....the single room is turning out to be an excellent example...not just for radiosity testing but also shadows. yeah i know the area shadow is the wrong thing in a cg sense..but it was a fudge that looks ok...
all the images i posted are not ideal for what i was talking about but the second one comes closest...you also find because this is just a plain white room there is a difference between this and say handige_harrie room with shiney wood floor....as would be the results if you had a thick pile carpet...with say a blanket scrunched up on it and possibly a glass window...areas of shadow would look different and the part about a mix of shadow quality would come more into play...with hard areas...and then tapering/blending into soft/area look on parts of the image. its really hard to describe what i'm talking about without being able to post an image and say that it!! sorry for this...and hopefully you can get what i mean.
i just think the area and the hard shadow...just doesn't cut it for realism imho...but i don't know how to solve it in cinema...:wip:
Phasmatis
04-08-2003, 03:51 PM
I've just formatted so I haven't got cinema installed so I can't try this myself at the moment...
What about using soft shadows with a 2000x2000 shadow map? Or even an area shadow at 2kx2k? It may make it hard enough to look realistic but you'll have the slight falloff from the soft or area shadows, if you know what I mean. :hmm:
What about using soft shadows with a 2000x2000 shadow map? Or even an area shadow at 2kx2k? It may make it hard enough to look realistic but you'll have the slight falloff from the soft or area shadows, if you know what I mean. :hmm:
flingster
04-08-2003, 05:12 PM
yeah ive been searching the net for good examples...
i think if you get fall off looking right and you may have cracked it.
i think if you get fall off looking right and you may have cracked it.
flingster
04-08-2003, 11:05 PM
tried the soft shadow map...not impressed with my results let me know how others got on.
what i did get better results on was simple just taking Phasmatis edit of flingster file...then just drop shadow density...maybe add a little colour...but mainly drop the area shadow width....this gave me a better hard edge line to the shadow...which you would expect also flattened it slightly in appearance...but i just missed that somewhere along the line...got carried away with the radiosity part really..:shrug:
i might just see now how i get on with maybe adding to the plain room...floor...maybe blinds...window glass...ornaments etc as you guys played around and then see what sort of settings i need...
this thread has turned out really useful to me...learned a lot...let me know how ya got on...always interested in just plain white room as a start off point though.:applause:
what i did get better results on was simple just taking Phasmatis edit of flingster file...then just drop shadow density...maybe add a little colour...but mainly drop the area shadow width....this gave me a better hard edge line to the shadow...which you would expect also flattened it slightly in appearance...but i just missed that somewhere along the line...got carried away with the radiosity part really..:shrug:
i might just see now how i get on with maybe adding to the plain room...floor...maybe blinds...window glass...ornaments etc as you guys played around and then see what sort of settings i need...
this thread has turned out really useful to me...learned a lot...let me know how ya got on...always interested in just plain white room as a start off point though.:applause:
flingster
05-27-2003, 09:43 PM
Ok guys been a long time since this thread was alive...
but saw something in another thread that thought might be useful in this case.
archrendr recon'd he could put a large coloured sphere at the side of an image and get a warm effect on his image...
so i thought how about if i changed the colour of the backwall for this image and see its effects on the image...
so the result is as follows hope this helps guys...sorry if you no longer care or forgot all about this thread...but thought it might interest some of ya...colour is a little orange etc...but it is fairly warm glow to it...kinda mexican feel to it light wise.
lemme know your thoughts etc.
cheers
but saw something in another thread that thought might be useful in this case.
archrendr recon'd he could put a large coloured sphere at the side of an image and get a warm effect on his image...
so i thought how about if i changed the colour of the backwall for this image and see its effects on the image...
so the result is as follows hope this helps guys...sorry if you no longer care or forgot all about this thread...but thought it might interest some of ya...colour is a little orange etc...but it is fairly warm glow to it...kinda mexican feel to it light wise.
lemme know your thoughts etc.
cheers
CGTalk Moderation
01-14-2006, 07:00 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.