- User Since
- Sep 19 2016, 1:47 PM (160 w, 1 d)
Nov 30 2018
Nov 27 2018
In Unity an RGBA image could be imported with some extra settings; on of it that 'how to use the alpha'. When transparency needed, it acts like alpha, when it is an additional data channel, it acts as a greyscale channel.
Sorry for citing Unity; it is because of the flawless workflow in the pipeline.
If I could have a note: it would be great to make it work properly both for 'Color' and 'Non-Color' data as very specific usage could happen sometimes. Thanks.
Thanks you, it clearly seems to be a bug to me. As for real-time engines TGA is a very popular format, hopefully it will be fixed soon.
So it will keep the Cycles Principled-Unity-Unreal, etc. texture usage easy.
Nov 22 2018
Ok, so it works as it should, but it heavily affects the classic wireframe function (still can't see if will it cause a conflict or not).
I will check if I use the latest 2.8, but I answer it differently.
Now I understand, nut I think it is a very inefficient solution. While working when I hit 'Wireframe', I want to see a see-through wireframe, not hitting another button for 'X-Ray'. As it is a vital function, I suggest to make an additional pie where it happens together (also with hotkey) from UX aspect. I cannot find simple wireframe without xray useful at all for daily workflow; even in cases when it could be handy it is more rare than the case we need see-through wireframes (based on more than 20 years in archviz/game asset and as someone who works hard on putting Blender into focus with leaving AD products.
Hello Bastien, do you need info/data from the devs on this? Thx.
Nov 8 2018
OK, I tested, it is partially a bug, also UX.
Nov 5 2018
Thanks! I will test it, also I have an idea what have to be tested.
Thank you both.
I try to record it and upload as a GIF, I just played with the default cube, nothing has been changed, not scene-specific issue IMHO.
My biggest problem is that this wireframe mode - I risk to say it and I say it with no offense - not too useable. I work with very dense scenes where it is vital to see everything what happens in a scene in wireframe mode. As a user I'm not interested in X-Ray at all as it has no practical side to me expect of some very special occasions (for example having a double object in the same world space and I want to find these overlapping objects). But wireframes are everything, the most important foundation for modeling. If it does not work properly, the battle is lost. :(
It seems my former message lost somewhere, so:
Nov 2 2018
you guys are awesome.
Oct 3 2018
did anything happened to resolve this issue? I cannot see any changes, 2.79 still acts the same way, making me to stick with old Blender versions.
Sep 18 2018
OK, so I consulted with a Max plugin developer related to this issue.
He told me that the order of vertex IDs determines the triangulation of a mesh with ngons, which is probably related to the Max internal workflow where in Poly mode the invisible tris are there (could be switched on for the viewport) to see topology.
Sep 17 2018
Dear Bastien, if I remember correctly when this problem firstly appeared it was checked with Max, Modo, C4D and Blender (with the help with some friends). And if I remember correctly only Blender exports were unusable in other apps related to this issue (when issues came up, the user was notified, at least at export time, so it was easy to avoid them).
So they somehow managed to solve it (I will recheck it if I will have time).
I would be happy with at least a warning pop-up that something is not OK from topology aspect, because it is a 'flying blind' case now.
'It simply looks like tessellation of Max (and others) is more limited than the one we have in Blender'
'We just export polygons as they are in Blender, nothing else.'
As far as I know triangulation always happens, but it is not 'viewable'.
Thanks, here it is:
Sep 16 2018
Jun 6 2018
Jun 5 2018
"Quite honestly if Global Wireframe doesn't make it into Blender 2.8, I might even consider modelling in 2.79 and then moving everything into 2.8 to use it as a render engine."
Jun 1 2018
Completely agree with Mindinsomnia. Also a dedicated key for wireframe mode would be great. I don't wanna see surfaces.
Also colored wireframes are very useful (per object, per layer, per group, per collections, etc.), but I know Ton hates them, I just never was able understand why. I used them like 25 years ago and use them today when I'm not on Blender; pretty easy way of visually managing stuff.
Where to put the feedback on wireframes to be efficient?
Still cannot get how it works, outline disappears all the time when I touch the screen in Object mode.
Thanks, Brecht. I downloaded latest build, it is so buggy even Tab does not work for Edit mode, also forgets wireframe settings.
To honest it would be nice so have some infos related the 'classic' wireframe view from developers. I work with it like 80% in my time when modeling, 99% when putting scene together. Simple there is no faster way to put a dense scene together.
May 11 2018
Ok, maybe stupid question as I didn't have time to follow 2.8 development discussion, but where are the see-through wireframes? I work with that viewport in almost all apps I use (Blender, Max, C4D).
Apr 14 2018
Seems as a good idea using per object base.
Nov 24 2017
Do you have any idea when overall performance will be better? Will 2.8 bring change related to this (my former tests did not show improvements).
Nov 17 2017
Yep, Cinema and MODO both shocked me (different way).
Ok, some updates with test file:
Oct 27 2017
I made a new test, also here is the link for a testfile I made.
I shut it down again, still not finished. (clean setup)
Ok. I'm running the factory clean Blender 2.79 test now.
Hello Joshua, thanks.
Oct 25 2017
Oct 10 2017
Hello guys, since the thread I lost 5 times linked data with no warning, last time today. The linked node groups simple disappeared from the scene, no warning at all, no orphan data appeared in the Outliner, nothing. As talking about a few hundred materials with 5-6 textures per material + all the material settings it is significant time to remake, especially to do it 5 times.
Sep 19 2017
Thanks, it took a while, but now it works.
May I ask if there is any scientific/mathematical background behind the decision that it works this way for backfacing/single-sided materials or it was just 'let's do it this way'?
Hello Brecht, sorry for the delay, I made a small image to show what I would like to achieve.
Sep 7 2017
Addition: I made a test with 2.79 RCII with the Principled shader.
It results completely buggy render.
Aug 9 2017
Thanks for all, I tested it and now - I waited for this so much - my scene renders on AMD. Great job!
Thank you very much, I can't wait to test it.
Jul 29 2017
Here is the image I promised.
Hello Marius, could you offer me a link for that version? I would test it with the real scene. Thanks.
Jul 28 2017
'There can be missing image files and missing linked library .blend files, and those are quite different to Blender. Are you saying it's just the image texture that are missing, or are the materials and nodes missing (and so the image textures are also missing along with them)?'
I think there was some misunderstanding about the situation.
Jul 27 2017
Will it be supported? I just ran into this problem.
Jul 12 2017
Hello Guys, question: what are the current texture limits in Cycles?
I know last time I asked for increasing it CUDA had about 96 and if I remember correctly that became 1024, also on CPU. Thanks.
Jul 7 2017
'Having BW textures stored as single channel would also help to save 3/4 of the memory.' As I had no idea what causes the crash, I wanted to use as big textures as possible (no BW conversion) except of some simulating mixed usage.
Jul 6 2017
Here is the new file: https://ufile.io/ivcut
OK, quick test on real life scene:
Thanks, I'm on it.
Hello Mai, thanks for your effort!
When this commit will be included in a downloadable Blender version? I ran some tests; still not sure what makes the error, but it is clear that the same scene sometimes requires far more memory for OpenCL than CUDA.
Also I find it interesting that while final render time is comparable to CUDA, viewport render is far slower with it.
Jul 2 2017
Hello John; I'm on it; the last 2.79 I checked works on the example file, now I check it on the 'real' scenes. Some of them works (that is great), some results the error, but I have to examine what are differences, so I will need time for that. Thanks for the patience.
Jun 21 2017
Jun 18 2017
May 31 2017
Brecht, it is great to hear. I just made some short tests, but even reducing the 'darkness' of the effect would be a great step forward.
Conversation on Corona forum: https://corona-renderer.com/forum/index.php?topic=10392.0
In Mental ray it is solved, in Vray it is there, but less problematic (but I haven't tested it with the latest Vray).
May 29 2017
The question is then how a workaround could be made from algorithm side, am I right?
May 24 2017
BTW, have you tried with a former Blender version like 2.78a release (hash:e8299c8) on Win7 64 Pro and RX 480?
May 20 2017
May 19 2017
terminal = Win 7 command prompt?
Thank you, Bastien. I hope Sergey could help in this. I work on a project for 2 years now what could be a gigantic step forward in Blender's reputation for the CGI community, but I cannot recommend it as a honest/true alternative as these kind of bugs/limitations exist. I was pretty happy when developers eliminated the CUDA texture limits (I can't tell how much that annoyed me with big scenes), now it is time IMHO to push the limits to large scale usability (viewport performance issues, GLSL speed incrase, asset management, import/export fixes, handling terminator issues with shading, etc.).
Just for the test I tried it in Max + Vray RT, no artifacts on 12000 meter distance.
Dear Bastien, the artifact appears with only 400 unit away from the origin, too.
I many times worked with models far bigger than that in other packages without artifacts (think about archviz). How this could be solved in the future? Is there any plan for that?
I think it is a pretty important bug, please find a valid solution for this at least on the 2.8 releases. There should be a solution for that.
I will do my best to find an exact representation in my free time, but as it accidental, I have no promises. If I could find a solution to make a sample file what next?
I you read the details it is not a memory limit. If not, please do so.
Do you want to upload the file here?
It happens accidentally, that i why no file attached.
The problem exists with the latest official 2.78c. Thanks.
May 18 2017
May 5 2017
Sorry, I simply forgot to comment back, it still does not work. (Blender blender-2.78-3ab8895-win64).
Feb 3 2017
I made some tests with for the viewport.
'If your scene isn't rendering at all it means you've ran out of GPU memory, the only things to do in that case is to reduce scene complexity or use another device.'
I just made a quick test with the actual version of my 'big scene' on RX480 with the latest nightly build.
Hello Mai, thanks for the answer.
I have zero knowledge in python; would you please write me down what I exactly should do (and then how to revert back?)
Jan 17 2017
Hello Guys, is there any news on this? I downloaded the latest builds and still cannot use my RX480 for viewport rendering.
Nov 26 2016
Thanks, it does not work. Viewport rendering just reproduces the same error, it is not related to motion blur; it is related to the number of textures.
Nov 25 2016
I agree with Caetano, it is a real issue with low-poly characters what deserves to be solved.