Thanks for all the information. I’m glad to see that after so many years this section of the forum still has so much activity from all the old proffesionals.
But about this importon and IP thing. One problem of mine was never resolved in the last thread, which still makes me very hesitant to invest any more time in it until it’s further developed (was gonna wait till 2010 to come back to importons)
The issue was basically that, it does not accurately bounce light around corners. In interior scenes this was negligible, as it didn’t have to take into account light from every angle to produce adequate results, but I’m sure it still has issues.
I’m still suspecting that any sort of improvment into contrast or ‘vibrance’ of an image with importons/IP is really due to the fact it’s not taking into account every light bounce and is artifically darkening areas. I mean, final gathering has the ability to be every bit as accurate and detailed as IP/importons, that is if you turn up the settings and wait a bit more time. So it’s not like there is some inherit more ‘vibrant’ component to the ip/importon algorithm, changes in color and contrast is due to it bouncing the light differently. And as I understand, final gathering with very high settings is the current benchmark for accuracy in mental ray, so if importons are coming out different or darker in more areas, more contrasted, it’s not because of more accuracy or improvement over final gathering, it’s because of error, it’s missing bounces.
Or am I completely wrong there?
Admittedly, I have not done new tests since the lasts tests in that old thread. But I’ve seen no mention of any changes to the algorithm in the recent service packs, so I’ve not taken the time to retry it. So I am just curious, has anyone run into anything about this issue? Or looked into it a little more?
I just transferred my entire system and applications in one step flawlessly to a RAID array with all applications working with custom libraries intact (and no third party software or boot disks), I’m previewing 32-bit HDR and RAW files in the OS X Finder (Vista can’t even preview .CR2 files), running a Unix command line and have never seen this Mac Pro crash. I’m using keywords and Spotlight to work faster with textures:
I could do the same with fonts and get a full preview of the typeface right there.
So I’m wondering how it is you claim that Vista is a better experience. Vista is a better experience for people who have no idea how to use the Mac (Exposé, menu bar proxies, drag and drop into open/save dialogs, etc). There may be more 3D apps for Windows but that’s about it. I think you better tell the editor of CGSociety Paul Hellard why he’s wrong when he said this to me in an email (when I pitched him an article about free 3D workflow tweaks for the Mac):
Gee, you’re talking to the right man here. I’m an OS X user to the
point of wondering why anyone would waste time on any other platform.
But I'll take the word of a guy who trolls all Mac-related threads over his any day. Or maybe I should take the word of Microsoft, who has so much confidence in the abilities of Vista that they are scratching all references to it and giving away Windows 7 for free.
edit: it doesn't bother you that the Mac side of your MacBook Pro's 3D hardware is supported by Autodesk but the Vista side isn't? So what part of this experience is better again?
Its just that some of us spend very little time using the OS for anything other than hosting our primary application. I for one, spend a very small portion of my day doing anything other than clicking in Maya. With a couple easy tweaks Vista has excellent network performance, disk performance, graphics performance, and 64bit support. Thats 98% of what I care about.
I come from the IRIX and Linux world originally and I hate Microsoft. That doesn’t change the fact that Vista runs Maya very well.
sorry guys but rygoody posted a really good question regarding the initial thread topic of Importons and irradiance particles, something i would also like to know the answer too, and cgbeige you have just replied with non related crap. Go write it somewhere else. (cool set up tho. Thought i would lighten the mood lol) If you wanna reply to this then PM me. Don’t fill this thread with non related comments.
@Dagon@withego Is there any news on whether mental images are improving the importon/ IP workflow/algorithms?
lol - third party software is needed to see the files at all (I was talking about how you can preview RAW and HDR files right in the Finder without any additional software):
that also works in open dialogs so you can see what HDR file you are importing for an IBL probe.
Okay, so how about HDR files previewed in Explorer by default? anyway, I won’t derail this thread anymore - I was just responding to a retarded blanket statement that needed to be refuted.
There it is. RAW in Vista. .dll files aren’t “third party software”, if drivers are considered “third party” then every driver in your computer is “third party”, since Apple doesn’t manufacture any of that software or hardware. Do you believe your hard drive was made by Apple with drivers written by Apple? Canon also made the RAW driver files for the Mac, in case you weren’t aware.
Also, regarding HDR… Do you just have a rough time using Google? Came right up. Your Vista install didn’t have an updated Direct X, evidently, CGBeige. Please update your computers before making claims about things not working & software elitism in general.
Re: Is there a way to view HDR’s as thumbnails in Windows Explorer? The new version of the Direct X SDK seems to include HDR support for Windows explorer. I’m downloading it at the moment and will report back once I’ve tested it.
Followed by:
Re: Is there a way to view HDR’s as thumbnails in Windows Explorer? Okay, tried it and it works a treat. A little slower but considering it’s processing 32-bit images on-the-fly what do you expect. If the slight speed decrease bothers you then go with the option suggested previously and save a jpeg version of each HDR in the same folder.
Seems pretty obvious your experience with Vista is very biased and perhaps it’s just too complex. No more from me here, though. Re-railing thread.
If Importons/IP don’t accurately bounce light, are they simply workarounds for using GI and/or FG? I’ve never been able to get either (Importons & IP, that is) to work properly, but in my quick-deadline workflow I have little time to experiment; high turnarounds demand medium-quality renders at best, and the only reason to switch to Importons/IP for me would be if they sped up my workflow itself, as render times are already decent with GI and sometimes GI+FG. Are these features just not very well implemented in MR for Maya yet? What about physical accuracy?
My initial understanding was that irridance particles were to compete with Vray irridance particles. As in being, the one click ultimate solution to all GI, a replacement for FG, but better.
But given there current state of the function, I suspect there still being developed. Autodesk just put them in there for people who want to expiriment right now.
and is there anyway to turn on the image caching framebuffer in maya 2009 without ctrl_ghost?
Apple doesn’t manufacture any of that software or hardware.
Google “Aperture review.” Mine is at the top - I know how RAW works in OS X and actually, yes, it is Apple that writes the RAW decoder for Canon files, much like any other 3rd party RAW decoder.
I’ll stop derailing the thread when people stop writing misinformation that needs correcting.
This thread is fast becoming diluted once again unfortunately. I still want to know the pros and cons about IP and Importons. Appears to me that GI and FG are still kings perhaps not as fast, but in any case I wont have to learn how to use the newer technology…for now
@Dagon@withego Is there any news on whether mental images are improving the importon/ IP workflow/algorithms?
But given there current state of the function, I suspect there still being developed. Autodesk just put them in there for people who want to expiriment right now.
and is there anyway to turn on the image caching framebuffer in maya 2009 without ctrl_ghost?
For the first question since Dagon is betatesting MR, I guess he got more news than me…!
Concerning the second question about “image caching framebuffer in Maya 2009” after searching a few threads:
I guess that with Maya 2009 you can trigger the “image caching framebuffer” by adding a string option in the [size=2]miDefaultOptions, such as:[/size]
[size=2][/size]
[size=2]Name= fb_virtual cached[/size]
[size=2]Value= 2[/size]
[size=2]Type= Integer[/size]
[size=2][/size]
[size=2]I’m just guessing, so I will appreciate if Dagon, Max (Ctrl.Studio), Bart (lamrug.org) or Archoury (Rachid) could explain how to enable the “image caching framebuffer” function with Maya 2009, since it was already present in maya 2008 SP1 through the Ctrl Ghost shader…!!![/size]
IP is a mental image algorithm, i think it’s patent pending right now, there is not irradiance particles in vray, vray use irradiance cache for primary diffuse rays, this is similar to FG
for secondary (or even primary) rays vray can use the lightcache algorithm that is similar (a kind of reverse photon mapping), but also different from IP
mmm i dont know, in max you can’t use the string options, you may try the max version of the ctrl.ghost, i remember the fb option was hidden, but never tried it (probably you need to use backburner)
So that means that with Maya 2009, we can just add a new string [28] in the miDefaultOptions, in my case [28], since string [27] is used to set the IMPORTONS emission.
And feed the values as follows:
And go on to batch render from the rendering menu…?
The reason I’m asking this, is that on the Lamrug forum they discussed how to enable
the fb cache via the strings you mention, but only with batch rendering from the command line, which I don’t use often…!
Ciao Alex,
yes you’re right, the string option doesn’t work, i dont know exactly why, maybe if Max is reading this he can give us an hint, because ctrl.ghost is working very well
Hi Dagon,
After testing on my “extreme” (309 mb) Maya & Mr exterior scene, whit a lot of ZBrush 32 bit displacement maps, Miss Fast Skin materials, massive use use of Ivy generator & Pfx, Car paint shader etc. etc, it really seems to work also with the string option [28] enabled in the mi_default Options…!
I’ve batch rendered the scene without the string option and after 4 hrs. at 100% rendering done, MR start to flush and fails to write the FB exiting with status 1.
With the string option, MR at 100% rendering done start to flush but amazingly write out the FB and no exit code is displayed in the render log file…!
End of render log without string:
JOB 0.3 progr: 100.0% rendered on ego-c20e829e2d5.3
RC 0.2 info : rendering statistics
RC 0.2 info : type number per eye ray
RC 0.2 info : eye rays 6911455 1.00
RC 0.2 info : transparent rays 2042044 0.30
RC 0.2 info : reflection rays 7976137 1.15
RC 0.2 info : refraction rays 3262416 0.47
RC 0.2 info : shadow rays 565439349 81.81
RC 0.2 info : environment rays 3094640 0.45
RC 0.2 info : finalgather rays 9693193 1.40
RC 0.2 info : probe rays 266395675 38.54
RC 0.2 info : fg points interpolated 80131276 11.59
RC 0.2 info : on average 88.10 finalgather points used per interpolation
RCI 0.2 info : main bsp tree statistics:
RCI 0.2 info : max depth : 40
RCI 0.2 info : max leaf size : 520
RCI 0.2 info : average depth : 35
RCI 0.2 info : average leaf size : 9
RCI 0.2 info : leafnodes : 4919670
RCI 0.2 info : bsp size (Kb) : 257261
PHEN 0.2 progr: calling output shaders
PHEN 0.2 progr: maya_shaderglow(): Computing glow…
PHEN 0.2 info : maya_shaderglow(): Glow: Filter Width … 63
PHEN 0.2 info : maya_shaderglow(): Resolution … 0.638926
PHEN 0.2 info : maya_shaderglow(): Normalization … 7.938000
PHEN 0.2 info : maya_shaderglow(): Halo: Filter Width … 157
PHEN 0.2 info : maya_shaderglow(): Resolution … 0.260841
PHEN 0.2 info : maya_shaderglow(): Normalization … 49.298000
PHEN 0.2 progr: maya_shaderglow(): Done!
PHEN 0.2 progr: writing image file C:/Documents and Settings/ALEX/Desktop/SPONZA2009/SPONZA16.exr (frame 1)
MEM 0.2 info : allocation of 48000000 bytes failed: flushing
MEM 0.2 info : try ‘-memory 1150’ for future runs
MEM 0.2 progr: scene cache flushed 280 MB in 0.13s, now: 951 MB
MEM 0.1 progr: scene cache flushed asynchronously 103 MB for module JOB, now: 877 MB
MEM 0.2 info : allocation of 48000000 bytes failed: flushing
MEM 0.2 progr: scene cache flushed 6 MB in 0.00s, now: 870 MB
MEM 0.1 progr: scene cache flushed asynchronously 234 MB for module JOB, now: 712 MB
MEM 0.2 info : allocation of 48000000 bytes failed: flushing
MEM 0.1 progr: scene cache flushed asynchronously 178 MB for module JOB, now: 626 MB
MEM 0.2 info : allocation of 48000000 bytes failed: flushing
MEM 0.1 progr: scene cache flushed asynchronously 156 MB for module JOB, now: 560 MB
MEM 0.2 info : ================ memory error post-mortem ================
MEM 0.2 info : -------- memory summary --------
MEM 0.2 info : module max MB curr MB %bytes
MEM 0.2 info : MEM 506 51 10.39
MEM 0.2 info : DB 888 228 45.86
MEM 0.2 info : PHEN 405 176 35.42
MEM 0.2 info : RC 205 0 0.02
MEM 0.2 info : GAPMI 37 0 0.01
MEM 0.2 info : IMG 6087 1 0.25
MEM 0.2 info : JOB 1 0 0.19
MEM 0.2 info : GEOMI 1495 0 0.07
MEM 0.2 info : TRANS 13 0 0.00
MEM 0.2 info : LIB 160 33 6.81
MEM 0.2 info : API 7 0 0.00
MEM 0.2 info : RCI 1117 0 0.02
MEM 0.2 info : RCGI 26 4 0.94
MEM 0.2 info : GAPDI 722 0 0.00
MEM 0.2 info : GAPPO 250 0 0.00
MEM 0.2 info : RCFG 104 0 0.00
MEM 0.2 info : DBN 169 0 0.01
MEM 0.2 info : other 5 0.01
MEM 0.2 info :
MEM 0.2 info : total 498 100.00
MEM 0.2 info :
MEM 0.2 info : max heap memory approx. 1563 MB, now approx. 560 MB
MEM 0.2 info : -------- database summary --------
DB 0.2 info : database elements by module:
DB 0.2 info : jobs MB #jobs MB #nonjobs module
DB 0.2 info : 43 2734 3 915 GAPMI
DB 0.2 info : 23 276 56 12 IMG
DB 0.2 info : 0 0 4 3 LIB
DB 0.2 info : 0 911 200 8325 API
DB 0.2 info : 3 4 0 10 RCGI
DB 0.2 info :
DB 0.2 info : database elements by type:
DB 0.2 info : jobs MB #jobs MB #nonjobs type
DB 0.2 info : 23 276 56 6 13 image
DB 0.2 info : 42 1823 199 908 40 nbox
DB 0.2 info : 3 1 0 8 63 invalid
DB 0.2 info : 0 0 3 915 95 region_tri_list
DB 0.2 info : 0 0 4 1 108 kdtree_pos
DB 0.2 info : 0 1826 2 10196 other
DB 0.2 info :
DB 0.2 info : 70 3926 265 12034 total
MEM 0.2 info : run with ‘-message mem debug --’ for more info.
MEM 0.2 info : please try lower -memory settings.
mental ray: out of memory
MEM 0.2 info : allocation of 48000000 bytes failed: flushing
MEM 0.2 progr: scene cache flushed 3 MB in 0.00s, now: 556 MB
MEM 0.1 progr: scene cache flushed asynchronously 136 MB for module JOB, now: 463 MB
mental ray: out of memory
MEM 0.2 info : allocation of 48000000 bytes failed: flushing
MEM 0.2 progr: scene cache flushed 10 MB in 0.00s, now: 452 MB
MEM 0.1 progr: scene cache flushed asynchronously 104 MB for module JOB, now: 347 MB
mental ray: out of memory
MEM 0.2 info : allocation of 48000000 bytes failed: flushing
MEM 0.2 progr: scene cache flushed 6 MB in 0.00s, now: 340 MB
MEM 0.1 progr: scene cache flushed asynchronously 23 MB for module JOB, now: 317 MB
mental ray: out of memory
MEM 0.2 info : allocation of 48000000 bytes failed: flushing
MEM 0.2 progr: scene cache flushed 12 MB in 0.00s, now: 304 MB
mental ray: out of memory
MEM 0.2 info : allocation of 48000000 bytes failed: flushing
mental ray: out of memory
MEM 0.2 fatal 031008: can’t allocate 48000000 bytes.
MEM 0.2 fatal 031008: can’t allocate 48000000 bytes.
MEM 0.2 fatal 031008: can’t allocate 48000000 bytes.
MEM 0.2 fatal 031008: can’t allocate 48000000 bytes.
MEM 0.2 info : cleaning up memory mapped frame buffers
MEM 0.2 error 031075: timeout cleaning up frame buffers
// Maya exited with status 1
End of render log with string:
JOB 0.4 progr: 100.0% rendered on ego-c20e829e2d5.4
RC 0.2 info : rendering statistics
RC 0.2 info : type number per eye ray
RC 0.2 info : eye rays 6911452 1.00
RC 0.2 info : transparent rays 2041931 0.30
RC 0.2 info : reflection rays 7976878 1.15
RC 0.2 info : refraction rays 3261300 0.47
RC 0.2 info : shadow rays 565430215 81.81
RC 0.2 info : environment rays 3094796 0.45
RC 0.2 info : finalgather rays 9692404 1.40
RC 0.2 info : probe rays 266390927 38.54
RC 0.2 info : fg points interpolated 80127942 11.59
RC 0.2 info : on average 88.10 finalgather points used per interpolation
RCI 0.2 info : main bsp tree statistics:
RCI 0.2 info : max depth : 40
RCI 0.2 info : max leaf size : 520
RCI 0.2 info : average depth : 35
RCI 0.2 info : average leaf size : 9
RCI 0.2 info : leafnodes : 4919670
RCI 0.2 info : bsp size (Kb) : 257261
PHEN 0.2 progr: calling output shaders
PHEN 0.2 progr: maya_shaderglow(): Computing glow…
PHEN 0.2 info : maya_shaderglow(): Glow: Filter Width … 63
PHEN 0.2 info : maya_shaderglow(): Resolution … 0.638926
PHEN 0.2 info : maya_shaderglow(): Normalization … 7.938000
PHEN 0.2 info : maya_shaderglow(): Halo: Filter Width … 157
PHEN 0.2 info : maya_shaderglow(): Resolution … 0.260841
PHEN 0.2 info : maya_shaderglow(): Normalization … 49.298000
PHEN 0.2 progr: maya_shaderglow(): Done!
PHEN 0.2 progr: writing image file C:/Documents and Settings/ALEX/Desktop/SPONZA2009/SPONZA14.exr (frame 1)
MEM 0.2 info : allocation of 48000000 bytes failed: flushing
MEM 0.2 info : try ‘-memory 1150’ for future runs
MEM 0.2 progr: scene cache flushed 278 MB in 0.07s, now: 950 MB
MEM 0.1 progr: scene cache flushed asynchronously 103 MB for module JOB, now: 876 MB
RC 0.2 progr: rendering finished
RC 0.2 info : wallclock 4:10:55.30 for rendering
RC 0.2 info : allocated 875 MB, max resident 1562 MB
GAPM 0.2 info : triangle count (including retessellation) : 11127699
PHEN 0.2 info : Reflection rays skipped by threshold: 21676859
PHEN 0.2 info : Refraction rays skipped by threshold: 19658492
Scene C:/Documents and Settings/ALEX/Desktop/SPONZA2009/SPONZADISPLACEMENTSSSBASTA5SPEROIVY__2948.mb completed.
Now I’m just not sure about the naming convention for the first value of the string which is the name of the string, because looking at the other strings no other values displays underscores like in this one: Name= fb_virtual
Anyway this string just saved my ass, with this rendering so it could be real lifesaver with large scene like mine…!
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.
