Installed todays build of 13-5-2019 (20:26 dutch time).
My nvidea card isnt used (i think its all CPU render), it used to be suported in 2.7x but not anymore the 630M it was just as fast as the, the cpu (8 cores), so the differences to use it are negligible.
May 13 2019
Installed todays build of 13-5-2019 (20:26 dutch time).
Mar 24 2019
re-opened not fixed.
Transparant bones, to be able to set bone transparency (in pose mode , see overlays, the slider there should also have the option to make bones transparent ea too vanish them (which is much more useful (for compositional posing) as to currently setting all other objects to black), the slider can have a duo function. To get Black objects with bones (fow those who needs it), or to set the bones vaguely visible for fine-tuning compositions.
Mar 5 2019
Yes currently the bones are very visible, and that's problematic if you want to fine-tune poses for the best aesthetics.
Say you want to generate a beautiful pose, then one would like to see the environment, and figure mainly and be minimally distracted by the armature display.
Then just vaguely visible bones would be enough, as the main eye focus then is how does the model look like this (instead of how does my armature look like).
Mar 4 2019
Take a basic figure width a rig, go in pose mode. in solid display view.
In the main 3d view click enable bones in overlay then on the bottem menu go to shading Xray see how it behaves sliding from 0 to 1
While the anything else fades away except for bones. There has been an option (but i think it was in overlays menu then) where you could fade away the bones from 0 to 1
perhaps extend current shadow option with 1 to 2 to get that functionality back.
Mar 1 2019
Nov 16 2018
Jun 16 2018
Nov 2 2017
ah i see now, at first i responded quickly, i wasnt judging him, but the line was a general statement.
i should have written that poor sculpts, result in poor meshes, i was trying to explain that in general there (like a saying), but not to judge his specific level.
i am civil,
Let me give you some more numbers
Apply modifiers then save the file to a Blend file thats 667Mb of raw (but also compressed) blender data,
Without blend saving only applying the modifier, that raw data inside blender its about 1.7 Gig what you see on screen.
that's to be converted to obj since the formats are not the same it takes around 7Gig of data to be processed to become a obj file of more then 900Mb.
well it depends on the horsepower of your machine, over here its about 60 seconds.
And i'm kinda amazed that such a huge file is still processed that fast, for the amounts of data that it has to process.
Maybe add more Ram, or a faster processor (more cores) might help as well. (60 sec tested on Asus K95v a laptop 4 years old now).
Have you tried using bones ?
They are far better for finetuning relations between objects in animations.
Parent-child is normally not used for finetune tweaking positions between regular objects.
Another option is to use shapekeys, in edit mode but thats also not commonly used.
Shapekeys are there when bones would be to complex to use.
Its also not a problem on Windows, not in old builds not in todays builds,
If this is your regular job, then i stronlgy recommend you a SSD, this is not a bug in code; the waiting time you have is simply caused by the IO speed of your harddisk.
That suszane of 667MB took me about 2 to 3 secconds to load on SSD disk. So we realy cannot blame Blender here, but if it takes you 15 minutes upgrade your hardware, because your Disk IO speed is the bottle neck here.
Well it is solvable, also...
The art of sculpting isnt super small polygons... Its knowing where to put details and where to remove them.
So if your bad ad sculpting then that comes with a cost.
Its not Blender's fault that processing lots of faces take more time, its mostly the speed of your hardware, you could buy SSD's and a faster computer, or better learn to sculpt.
I couldnt even open the fbx file, i am not a developer but in some cases i give advice that can help you and them
Oct 23 2017
I'm not a developer
Its probably the extreme volume sampling, and light bounces for volumes, could your cards normally handle that ??
I rendered CPU and it did not crash, system stayed responsive, but had to kill blender cause such tasks take to much time, but windows was responsive.
I can imagine since windows wants drivers to be responsive this loaded to much work on your cards, and system gave up on the drivers, and your system came in a halt.
Halts are notorious to debug, but i would think the cause is obvious here.. remind that volume is one of the most time expensive shaders to use.
No, or if it should then its not working, neither for mkv movie or jpg image series.
Not for CPU and not for GPU (if only CPU worked would allready be fine to me).
Oct 22 2017
Apr 28 2017
Might have some weird boid uses as seen here https://www.youtube.com/watch?v=kyrhDTmPUAc
Its not that if you have not used some feature inside blender that you should remove it, instead learn it.
Things might not always have up to date documentation either, but still thats no reason to remove it.
Nov 20 2016
i'm using am orginal microsoft optical mouse, middle button (pusshing scrol wheel works).
Sep 9 2016
Mar 29 2016
There is a good reason for it.
As for now you like something pure random.
But maybe sometime, you or someone elese like a result and would like to repeat it
That wouldnt be possible if the random wasnt seed based.
Well all computer random number generates use a see, (results are put back to buildup newer pseudo numbers on.
So basicly each seed, a group of numbers.
Such group who look randomly but follow a pattern as set by a seed value are also used in games
Mar 12 2016
Mar 5 2016
It seams to be solved now, maybe the nodes got a bit slower, but its ok.
slower in the sense of fast clicking on a value to change it.
Code seams to wait now till rendering is finished before accepting another value.
I guess side effect of some threading result of gui-update
Feb 27 2016
the smoke sim doesnt handle scaling, all object scales should be applied.
not sure if it is a bug, its always been that way.
Think of smoke as an advanced cycles material.
if one scales the domain, then also the smoke doesnt change, as a material it only stretches.
Feb 24 2016
Well i dont see any error, it freezes and then Windows notifies that blender is not responding and blender crashes.
it took me 2 secconds to copy if from BA
It seams that the below node setup (for making clouds) seams to crash blender, i was wondering if i am doing something that isnt supposed to be done, or if others can confirm that this node setup tends to crash blender 2.77 (sometimes not directly). crash in viewport and on rendering. sometimes not directly sometimes after a few secconds.
Feb 22 2016
Jan 24 2016
if there are more threads active then there are cores then code execution might not be ideal, perhaps tuning on parallelism ea the tasker might solve it. (i've seen this in other parralel code that wasnt optimal, but i'm into the dark of how blender deals with it)
- merely an advice for parallel coding. *
Jan 17 2016
Well c++ isn't the problem here, i can code in about 12 languages..
But i usually don't work with code distribution systems, i only once used turtoise.
At work, we have a real small dev team each works dedicated on something we never had a need for it,
Jan 16 2016
@Zauber Paracelsus, that could be a neat idea, to have it only kicking only after x samples, and maybe ideally it should kick-in to recalculate not every x frames, but that ratio should be a curve, However i think this future is no longer being actively developed now.
Or is it still possible to compile it locally, and get this in cycles (that would be awesome) ?.
Dec 24 2015
that dropbox link didnt work here is a new link showing the errors
Dec 13 2015
Well it has no problem to do the BMW speed tests
And lots of other work i do too, but yes i know cannt update the gpu of a laptop and its not the fastest gpu
Depending on blender build, the 630M can outperform the laptop octocore i7 2.3Ghz
if the drawing of the edit viewport where to be handled as suggested it might work .
Dec 12 2015
Nov 28 2015
small note did you know for windows file path names you could use / as well, its some kind of compatibilty mode since the Dos age.
You might not need to do such conversions.
Nov 21 2015
Jun 13 2015
Ok its resolved here.
Jun 12 2015
i'm able to run 2.74
additional info, the python exe doesnt work either
Jun 11 2015
Apr 26 2015
Perhaps i'm wrong but isnt "is_last_step_empty" always false in above code ? (so line 699 no need to IF it), it was false since line 652
Mar 1 2015
I updated my intel HD4000 driver (after Sybren noted me about importance of openGL)
And me thinking a driver would likely update the openGL parts too.
And well so it did, intel driver 15.x.x.x.xxxx or so fixed this.
Now both methods work just as in 2.72
But strangley Blender 2.72 can use both modes for selection without crashing... (so this might not be driver related)
Ok just tried that, it was set at automatic.
I think thats default ?.
When manualy set to "openGL oclusion queries", it does not happen.
In the other mode it crashes blender. ..
It also happens with enough subdivides on a regular cube
But normally people dont do that, they are more likely get into this after sculpting.
Feb 20 2015
Hi, the experimental kernel need much more memory than the "supported".
I cant even render the default cube with my GTX 560Ti 1.28 GB, my GTX 760 4GB work fine.
There are some patches in review to reduce mem usage, also there is AMD team to work on split Cycles kernel in micro kernels what reduce it too.
You can only disable your 570 in UP atm..
If you have the 2.5 GB 570 it is may a different problem.
Jan 14 2015
But do we really need this feature? The description says it's useful when having big tiles, but why not make tiles smaller? You shouldn't have big tiles on CPU anyway..
Also on GPU i'm not sure how it'll work -- without atomics. Also not sure it's possible to access other tiles buffers from GPU.
Or maybe i'm missing something in the way how this feature works?
Nov 23 2014
Is there a difference with how this handles refraction materials when they have zero roughness or more than zero roughness? With one simple render of mine, I found that it was taking a very long time for a refraction material to render when it had zero roughness, but when the roughness was changed to any non-zero value, it was much faster. By changing the Refraction BSDF's roughness from 0 to 0.001, an image that previously took 20 minutes to render instead took only 41 seconds.
Nov 7 2014
I have some update on the CMJ part, as i did a lot of testing with render settings on the Wallace and Gromit render speed test
Wehre Razorblade achieved around 80% time reduction for simple movies (he's my collegue)
Nov 3 2014
I think a small reproduce-able bug was found on BA:
It doesnt put a PC into crash or halt.
Take some blend file
Oct 29 2014
Okay, now it applies again. There are no new features in this version, so there's no need for new builds.
@Peter Boos (PGTART) How does the patch disable CMJ? For me, it works with every adaptive option, even the progressive rendering. It produces way more noise than Sobol because CMJ is only low-discrepancy if the whole sequence is used, but it still works.
Could someone please continue review here? On BA, there is quite some evidence that it indeed gives a nice speedup in many scenes (of course not in every scene), the parameters are about as intuitive as it gets (two checkboxes, one quality slider and a setting that you can safely ignore if you don't know what it does) and many users seem interested in it. Of course, I'll fix things that need to be changed, but I need to know what these things are.
Oct 21 2014
I note that correlated multi jitter disables your sampler.
Its a different sampling method less advanced, but maybe its a tip... that if someone chooses your sampler to disable the other ones in the GUI.
Or better extend the 2 sample methods there with yours (so it becomes 3 choices, with yours as the default), >like> pattern : adaptive.
And then if users select another method, gray-out your sampler setting.
Then your settings fit in, i think in the right place of the GUI sampler settings, and not somewhere lost below it.
Oct 18 2014
Oct 17 2014
I've tested the patch with the agent shot here in the studio. It only gave about 8 seconds advantage in the total render time of 32minutes, which is less than a half of percent speedup. And in the same time i still don't see proof it doesn't lead to flickering in the animation (linking to the forum asking to read for feedback in there doesn't work as well because we don't have time to follow forums).
This makes me real skeptical about accepting this at this point, especially before better ways of sampling are investigated.
Comments in this patch seems to go to forum-ish like discussion, which is not really welcome here. It is a code review tool for developers, and stuff which is not directly related to the code review process should not be posted here.
Oct 16 2014
Well it think low values do make sense, say i would like to render a star-ship; mostly black background some stars and a detailed starship.
The "old" non adaptive renderer wouldn't make any difference between black space, or the detailed ship they took equal render time.
But now with low values one still could render fast "equal" all blocks of empty space, while the subject gets rendered normally, thats still an improvement.
The same setup would allow for rendering experiments, on transparent, to be later combined with foto's or film, where just a small part is CG.
Oct 10 2014
Now the intel drivers and the Nvidea display drivers (as i have two graphic adapters in the laptop) both got an update by windowsupdate, and i seam to be unable to reproduce the problem.
Oct 6 2014
Well in between video work (while i was eating) i did a render on the intel adapter, using the official release candidate of blender.
Oct 5 2014
I doubt that hardware is the issue here, since i wasn't rendering on GPU, and blender was just tiling the final render, and it has been doing fine for the start of the render job but didn't finish to the (final?) end. > i say final because that seamed to be the problem as i mentioned earlier in the blender artist thread.
Oct 4 2014
An additional comment since i had a Compositing setup to write the result to my c:\tmp folder
And i didnt found any result there, i think still the rendering itself didnt finish to the end, might be close related to the problem as i mentioned in the link although then i didnt got nvidia problems then, but like then the rendering process started fine, when i went sleeping 1/3 (one third) had been rendered fine then, the process was still going on.
Aug 25 2014
While looking at the render proces i got an idea that's maybe nice, as it might reduce render time, so i share the idea.
With adaptive rendering (i read from the threat this can be done on pixel scale as well), but that might not even be required for this idea.
Well i used camera with a focus and an empty as focus object / (zmap might work as well for this idea).
Jul 14 2014
While playing around with settings i noticed something strange using R19 Jemonn's build (thanks), on my i7 octocore (no GPU).
Jun 24 2014
@Adam Friesen (ace_dragon), as for reference what is your PC hardware (processor-speed&cores / gpu /memory ),
It seams a complex scene, how many faces its made of ?, how much lights ?
May 19 2014
May 15 2014
May i suggest "as an exception", this gets pushed to the Blender build bot ?.
Bypassing code reviews. This doesn't break the other cycles render methods or other stuff.
And we all know those builds are experimental anyway.
As a result we wouldn't disturb the programmer any more with build questions, or failing compilations.
Sure it might not have been peer viewed, but is it that so important ??..
Its experimental, and its the biggest blender change since well ehmm... cycles introduction.
Anyone loves a faster renderer, even if still under development.
Feb 22 2014
Today i tried to reproduce it again,
It was really obvious the last time, i was working on something small in another layer.
It took a long time to get a cycle preview of it.
For my main scene the bhv build was around 1.5 minute
I had the wait the same time for my small object.
It happened the whole evening while i worked on it, i wasn't sure if it was related to the latest build, but i didn't retried it with an older build.
Today worked on it again.. and so far it hasn't happen.
I'm not sure what is so much different from that day and today.
Feb 20 2014
Just loaded in Blender 2.69 official build 64 bit windows no problems with loading the file.
Maybe its your system ? (not saying linux is bad, but perhaps something got broken?), its not a general bug, but somehow operating system makes a difference