The 367.57 driver allowed to use the 3D under Debian linux on a DELL Precision T5810 with a BenQ 24" LED - XL2411Z monitor and a DELL DP to HDMI converter.
When switching to 375.10 there was a very obvious problem in the stereo display: The top ~3/4 of the screen gives a good stereo image. On the bottom then there is a zone without any stereo effect of about 1/8 of the height of the screen and below there is another zone where the stereo effect is actually inverted.
Closer inspection of the image for the 367.57 driver showed that it also had zones without correct 3D effect, one is a narrow band towards the top of the screen, and one at the very bottom of the display.
This problem occurs independently of the software used: PyMol, coot, chimera on all machines of the same kind. Is there any explanation for this peculiar problem?
Please provide nvidia bug report of you system as soon as issue hit? Also please share video recording showing this issue? What desktop env you are running gnome, kde, xfce or else? Is composite enabled/disabled? Are you sure no repro with 367.57 and repro with 375.10? Is there any change in system except nvidia driver?
I join some Screenshots taken with a digital camera looking through the left eye of the stereo glasses and also the debug files for different installations.
The problem is probably also related to the z-position of the object, the defect is more obvious for objects located more in the background.
We tested two versions of the driver, a screenshot and the debug file for driver version 367.57 are here (no additional installations):
https://filesender.renater.fr/?s=download&token=ba4b8ed1-0e58-bff3-140c-0b44e527eb6c
or prior to driver update on another machine:
https://filesender.renater.fr/?s=download&token=0d11f287-0ae5-812f-8917-3a339e98417c
and after update to driver 375.39 the files are here:
https://filesender.renater.fr/?s=download&token=84111f1e-c47c-b1ca-5357-8cd332647c99
We are using the xcfe desktop, I am unsure about composite enabled/disabled.
There are no additional changes in the system.
Best regards
Wim Burmeister
Hi wpbgrenoble,
Did you see this issue with glxgears -stereo application? If yes, Can you point the defect? Otherwise we need your application to try reproduction of this issue.
Hi,
the problem remains the same with glxgears -stereo. The stereo effect is correct on the top 3/4 of the screen and disappears towards the bottom. Due to the rapidly moving gears, the effect is a bit hard to judge and I cannot send any screenshot. It is possible that the problem starts almost in the middle of the screen, so that it extend further than with the applications coot, pymol or chimera, which I used previously.
Here some output from
glxinfo | grep OpenGL :
OpenGL vendor string: NVIDIA Corporation
OpenGL renderer string: Quadro M4000/PCIe/SSE2
OpenGL core profile version string: 4.4.0 NVIDIA 375.39
OpenGL core profile shading language version string: 4.40 NVIDIA via Cg compiler
OpenGL core profile context flags: (none)
OpenGL core profile profile mask: core profile
OpenGL core profile extensions:
OpenGL version string: 4.5.0 NVIDIA 375.39
OpenGL shading language version string: 4.50 NVIDIA
OpenGL context flags: (none)
OpenGL profile mask: (none)
OpenGL extensions:
OpenGL ES profile version string: OpenGL ES 3.2 NVIDIA 375.39
OpenGL ES profile shading language version string: OpenGL ES GLSL ES 3.20
OpenGL ES profile extensions:
And the output from glxgears -stereo :
Running synchronized to the vertical refresh. The framerate should be
approximately the same as the monitor refresh rate.
598 frames in 5.0 seconds = 119.432 FPS
600 frames in 5.0 seconds = 119.984 FPS
600 frames in 5.0 seconds = 119.984 FPS
566 frames in 5.0 seconds = 113.187 FPS
600 frames in 5.0 seconds = 119.982 FPS
600 frames in 5.0 seconds = 119.985 FPS
600 frames in 5.0 seconds = 119.983 FPS
600 frames in 5.0 seconds = 119.981 FPS
The framerate dropped while I moved the window from the top to the bottom.
I hope this information helps.
Greetings
Wim Burmeister
Is it possible for you share your stereo image/application with us? If not, Could you please take photos of glxgears and point the issue. I’m not able to access latest link FileSender .
Hello,
could you suggest a simple stereo demo program different from glxgears and running under debian for testing?
For glxgears, due to the speed of the movement, it is not possible to take any photograph through the stereo glasses showing the problem.
Something static would be better.
The software we are using here is quite accessible but specialized:
Index of /personal/pemsley/coot/binaries/release for binaries.
You have to work with pdb files (atomic structures containing lists of atom coordinates), for example
[url]http://www.rcsb.org/pdb/explore/explore.do?structureId=1ssp[/url] and than use the “Download Files” tab and chose “PDB format”.
For the expired files, could you please contact me directly by mail and I’ll send them?
wim.burmeister[at]ibs.fr?
Thanks
Wim
Also please share issue reproduction steps for the application you provided. We never used this application. Please share exact link which application should I download from Index of /personal/pemsley/coot/binaries/release ?
I hope you have NVIDIA 3D Vision 2 hardware like : http://www.nvidia.com/object/product-geforce-3d-vision2-wireless-glasses-us.html ?
Using a different download platform, I join again some Screenshots taken with a digital camera looking through the left eye of the stereo glasses and also the debug files for different installations.
We tested two versions of the driver, a screenshot and the debug file for driver version 367.57 are here (no additional installations):
and after update to driver 375.39 the files are here:
We have have NVIDIA 3D Vision 2 hardware.
I think the easiest software for a test is coot, but I would need some information about the platform and the distribution which you are using for a test in order to indicate the right executable. Still, there are no license issues.
Still, is there another program which would allow to test stereo than glxgears? glxgears is too fast and does not allow to take pictures using a digital camera.
Best
Wim
Is this issue still reproduce without DELL DP to HDMI converter? I mean just direct connect DP-DP cable or HDMI-HDMI cable and test?
The result is independent of the converter, either a genuine DELL converter or a generic one. The generic one shows always a message with “Incorrect cable”, probably from the NVidia Drivers, but works fine otherwise. Without a converter using only an adapter, there is no output to the screen at all.
Best
Wim
Hi wpbgrenoble, Does this monitor have also have inbuilt 3D emitter? I think you using Quadro M4000 which has DP ports. If you just connect DP-DP cable or DP-HDMI cable , Does the issue reproduced ?
>>Without a converter using only an adapter, there is no output to the screen at all.
Best
Can you explain more about it?
Hello,
sorry about adding some confusion.
In fact, I use a DP to dual link DVI converter, either the one from HP or a generic one.
When I use a passive adapter, there is no output at all (probably I would need to reduce the refresh rate to 60 Hz, what I did not do).
The Nvidia Quadro M4000 card has 4 DP connectors.
So please, when I wrote HDMI in my previous posts, please read dual-link DVI.
Connecting the DP of the Nvidia Quadro M4000 card via an adapter to the HDMI input of the BenQ 24" LED - XL2411Z again does not allow to use a refresh frequency of more than 60 Hz.
When I reduce the screen resolution to ~ 1024*750, 120 Hz and stereo become possible, but again with serious defects much worse than before.
Best
Wim
Hi wpbgrenoble,
I think you are using BenQ 24" LED - XL2411Z monitor, Is the same issue reproduce with any other model of monitor? Does BenQ 24" LED - XL2411Z monitor have also have inbuilt 3D emitter?
If you are using external NVIDIA 3D Vision 2 hardware, Please share the part number printed on eacj devices? Are you using 3 Pin Mini DIN Bracket Stereo Connector?
Part number of 3d Hub/IR Emitter?
Part number of USB cable used for connect 3d Hub to system?
Part number of 3d Glasses used?
Part number of 3 Pin Mini DIN Bracket Stereo Connector?
Part number of Sync cable? if any?
i also have the exact problem. It is quite annoying. It is with brand new dell debian 8 PC with asus 120hz monitor connected to Quadro card using dell hdmi to duallink DVI adapter and Nvidia 3d vision 2 glasses/IR emitter.
details:
os: Debian 8
nvidia driver: 375.26
graphics card: nvidia Quadro M4000
monitor: ASUS VG248QE
hdmi-dvi adapter: Dell KDP70
3d hub/emitter P/N: P854
emitter->PC USB cable: 030-0314-000-1816
emitter->graphics card sync cable: 030-0313-000-1816
3d glasses P/N: P1431
Here is an example of the serial numbers (we actually have 10 systems bought at the same time, all with the same problem. We use external IR emitters
Part number of 3d Hub/IR Emitter? 04213 160 18211
Part number of USB cable used for connect 3d Hub to system? 030-0314-000-4115
Part number of 3d Glasses used? 04212 160 11291
Part number of 3 Pin Mini DIN Bracket Stereo Connector? not accessible
Part number of Sync cable? if any? 030-0313-000-4115
I hope this helps
Wim
Hi 3dproblem,
nvidia Quadro M4000 is don’t have hdmi port. Do you mean with DP-DVI adapter? What the is mode of adapter you are using? Is this issue reproduce if you use DP-DVI cable? What desktop env you are running - GNOME, KDE, Unity or else? With which application you are testing? Can you share screenshots showing the issue? Please attach nvidia bug report.
the adapter/converter is: Dell KDP70 (converts from whatever to dvi dual-link)
Desktop environment is MATE (a fork of GNOME2). Testing with glxgears and COOT.
I cannot share screenshots, it is a very hard thing to photograph. Also, i dont have DP-DVI cable ( i dont think 120hz works with passive cable, hence the dell KDP70 converter/adapter)
thank you
please tell me how to send bugreport file, i dont see any method to attach it to this post or a private message.
thank you
Attach to your existing post.