3D vision on TV SONY kd-49xf8596

[BOLNICHKA39]

Nvidia Bringing Back Silently the Generic CRT display 3D Support?

Version 425.31 does not support 3D Vision, because NVIDIA Is Officially Ending Driver Support for 3D Vision. Only older drivers older than version 418.

You may wish to google "3D Vision Generic CRT mode" and study.

You will probably fail unless you:
(A) You have older version of NVIDIA drivers
(B) You have version of drivers that have the CRT mode

It won't work unless (A)=YES and (B)=YES

There will be some tricky adjustment to make it work, possibly an EDID to force the generic CRT hack to work.

You probably do not have the correct information, I have the 425.31 driver installed and I have stereo 3d support, and there is CRT monitor support. But as I said, the image becomes three-dimensional, but it is very twofold and I'm sure that the glasses are turned on and synchronized. I still see the frames in the glasses for the left and right eyes, although there is a peculiarity, if you look only with the right eye, then I see frames for the left and right eyes, but the frame for the right eye is clear, and for the left eye a little transparent, and also on the contrary, I see both frames with my left eye, but one frame is a bit transparent. Is it possible that the glasses are synchronized at the wrong frequency?

[Chief Blur Buster]

probably do not have the correct information

I did some research, and you're right, the 425.31 is the last driver that has 3D Vision support.

That was the link to NVIDIA's website, but they apparently delayed end-of-life to version 425.31

Is it possible that the glasses are synchronized at the wrong frequency?

If the transparency is perfectly constant, it is likely a phase problem and a shutter-open duty cycle problem, not a frequency problem. The timing of the shutter relative to the timing of the refresh cycles.

For the time budget of 1/60sec for two 120Hz 3D frames, on a sample-and-hold LCD without a strobe backlight:
Left eye open for 2ms
Both shutters closed for 6.33ms
Right eye open for 2.33ms
Both shutters closed for 6ms

3D Vision glasses in CRT mode, doesn't do this correctly for sample-and-hold displays, but it works fine for strobed displays.
This is caused by the problem of Not all pixels refreshes at the same time

For a strobe backlight (displays that have a backlight flashing mode), the cadence of the shutters is MUCH more forgiving. That's why NVIDIA originally invented LightBoost, a strobe backlight that improved 3D glasses operation, but also improved motion blur and reduced crosstalk.

Is the TV currently strobing? Adjustments on the TV-side may occaisonally improve things (e.g. turning on/off Game Mode, turning on/off "Motionflow IMPULSE" (don't use anything other than "Impulse"), turning on/off 3D Mode, etc)

Another technique is a custom emitter tuned with correct phase / correct shutter open duty cycles, properly optimized for the scanout behaviour of your particular 120Hz LCD (having a Samsung Galaxy 960fps high speed camera may help in programming a custom shutter duty cycle for 3rd party emitters or open sourcde implementations, other than NVIDIA 3D Vision emitters).

Unfortunately, there is very little information on customizing shutter cadences, but if 3D was still popular today, it would now be much easier to design shutter sequences today thanks to modern smartphones with 960fps-filming features that allows one to determine how long a clear refresh cycle is displayed for (that is not in mid-transition or mid-scanout), since not all pixels refresh at the same time.

Your language translator software probably will probably not correctly translate this post, because this veers into advanced 3D tweaking topics.

[BOLNICHKA39]

probably do not have the correct information

I did some research, and you're right, the 425.31 is the last driver that has 3D Vision support.

That was the link to NVIDIA's website, but they apparently delayed end-of-life to version 425.31

Is it possible that the glasses are synchronized at the wrong frequency?

If the transparency is perfectly constant, it is likely a phase problem and a shutter-open duty cycle problem, not a frequency problem. The timing of the shutter relative to the timing of the refresh cycles.

For the time budget of 1/60sec for two 120Hz 3D frames, on a sample-and-hold LCD without a strobe backlight:
Left eye open for 2ms
Both shutters closed for 6.33ms
Right eye open for 2.33ms
Both shutters closed for 6ms

3D Vision glasses in CRT mode, doesn't do this correctly for sample-and-hold displays, but it works fine for strobed displays.
This is caused by the problem of Not all pixels refreshes at the same time

For a strobe backlight (displays that have a backlight flashing mode), the cadence of the shutters is MUCH more forgiving. That's why NVIDIA originally invented LightBoost, a strobe backlight that improved 3D glasses operation, but also improved motion blur and reduced crosstalk.

Is the TV currently strobing? Adjustments on the TV-side may occaisonally improve things (e.g. turning on/off Game Mode, turning on/off "Motionflow IMPULSE" (don't use anything other than "Impulse"), turning on/off 3D Mode, etc)

Another technique is a custom emitter tuned with correct phase / correct shutter open duty cycles, properly optimized for the scanout behaviour of your particular 120Hz LCD (having a Samsung Galaxy 960fps high speed camera may help in programming a custom shutter duty cycle for 3rd party emitters or open sourcde implementations, other than NVIDIA 3D Vision emitters).

Unfortunately, there is very little information on customizing shutter cadences, but if 3D was still popular today, it would now be much easier to design shutter sequences today thanks to modern smartphones with 960fps-filming features that allows one to determine how long a clear refresh cycle is displayed for (that is not in mid-transition or mid-scanout), since not all pixels refresh at the same time.

Your language translator software probably will probably not correctly translate this post, because this veers into advanced 3D tweaking topics.

Yes, there are problems with the translation. I understand you this way, the 3d mode for CRT monitors is not suitable for LCD panels, because there is your own image display method. I have motionflow on my TV and game mode on. I also realized that for 3d vision you need a screen with a response of 2 milliseconds or less, because does the right lens flicker at a speed of 2.33 milliseconds, and when the right lens opens, the screen does not have time to change the frame for the other eye due to the large response, because of which it already falls into the left eye? Unfortunately I do not have a smartphone with a good camera to track the speed of the matrix. And I also did not find information about the response of the TV matrix, I found only the input signal delay of 13 milliseconds. Did you say something about emitter setup? Could you give more details about this? I also noticed one feature when I look at the TV through the glasses, and when I press the power button on the glasses, for 1 second the transparent frames become much more transparent and the image becomes almost perfect, but after that the image starts again very double. How to configure the emitter, I can try to configure the "blind"

[BOLNICHKA39]

UPD: I have disabled motionflow and the game mode is on, I can also try adjusting the brightness, contrast, and black level settings

[BOLNICHKA39]

Just now I noticed. The image is stronger than the twins at the top of the screen and the lower, the less ghosting. For example, in the 3d demo nvidia, the large nvidia logo is very double, and the letters of the word nvidia are much smaller, and the gender and small nvidia icon are almost not double. Does this confirm that the monitor matrix blinks out of sync with the glasses?

[Chief Blur Buster]

Does this confirm that the monitor matrix blinks out of sync with the glasses?

Let's see if the language translator software understood correctly:

The glasses are at the correct frequency, but wrong phase.

It is phase.

Like the moon phases. For example, full moon, half moon, new moon. Both TV and glasses have to be the same phase.

"Same phase"
Works: TV at half moon, and glasses at half moon
Works: TV at full moon, and glasses at full moon
Works: TV at new moon, and glasses at new moon

"Different phase"
Won't work: TV at full moon, and glasses at half moon
Won't work: TV at new moon, and glasses at full moon
Won't work: TV at half moon, and glasses at new moon

Numerically, phases are like:

Both TV and glasses at phase of:
T+0.5/120 second = left eye
T+1.5/120 second = right eye
T+2.5/120 second = left eye
T+4.5/120 second = right eye
etc.

Both TV and glasses at phase of:
T+0.8/120th second = left eye
T+1.8/120th second = right eye
T+2.8/120th second = left eye
T+3.8/120th second = right eye
etc.

Both TV and glasses at phase of:
T+0.3/120th second = left eye
T+1.3/120th second = right eye
T+2.3/120th second = left eye
T+3.3/120tn second = right eye
etc.

The TV and the glasses have to be the same phase.

This doesn't cover the shutter-open time, which will probably need custom emitter programming (e.g. home-made Arduino emitter).

Shutter open time needs to be very short for displays that sequentially scanout, like www.blurbusters.com/scanout ... because not all pixels change at the same time. The screen refreshes different pixels at different times. The 3D glasses should only open when all pixels are refreshed for the current refresh cycle, but close shutter before the next refresh cycle begins.

Because many LCDs have input lag, the phase will often be different.

Because of the phase of the scanout, there will be "bars of clarity" and "bars of double images", which is natural because of www.blurbusters.com/scanout .... Not all pixels of a screen's suface are refreshing all at the same time, and the shutter glasses have to correctly "time" the phase.

If that's what you meant by sync, then yes.

[BOLNICHKA39]

All right. But is it possible to programmatically change the frequency of the flicker of the glasses so that the phases of the flicker of the glasses are like on a TV. The second question is, if you use a monitor with a matrix response of 1 millisecond, then this should not cause problems with the display of 3d?

[BOLNICHKA39]

I found monitor acer niteo vg272xbmiipx [um.hv2ee.x01], it has a frequency of 240 hertz with a response of 1 millisecond, this monitor does not support 3d vision, however, if you replace the EDID, then it can work with 3d vision? Or acer nitro vg270upbmiipx [um.hv0ee.p01]?