[DIY]Shutterglasses test bench

[Petrus]

I made a simple test bench to compare the performances of my different shutterglasses.
It is basically a piece of cardboard with two white leds and two photodiodes in front on each leds :

To get an acurate measure of the light passing through the glasses, I use a simple circuit with an operationnal amplifier :


I adjust the resistance to get 10V at the output without the glasses between the led and the photodiode. Then I can easyly determine the brightness loss caused by the glasses.

With the output connected to an oscilloscope I can also see the shutter speed.
Example on the Elsa wired @120Hz:

A: DDC signal
B: measured brightness

As you can see, the shutter doesnt go totally clear, thus I measure Toff from 100% (shutter closed) to 10% (shutter almost fully open)

So here are the results :

Elsa revelator wired :

color fidelity:

transparency (power off):29,5%
transparency (on@120Hz) :10%
Ton :380µs
Toff (100-10%) :4,56ms


Simuleyes (powered with 5V):

color fidelity:

transparency (power off):13%
transparency (on@120Hz) :6%
Ton :750µs
Toff (100-10%) :4,4ms


Asus glasses (on the simuleyes controller) :

color fidelity:

transparency (power off):27%
transparency (on@120Hz) :10%
Ton :600µs
Toff (100-10%) :4,26ms

I have ordered Nvidia's 3D vision glasses and I will soon make the same tests with it.

[edit]
Nvidia 3D vision :

color fidelity:

transparency (power off):31,9%
transparency (on@120Hz) :16,5%
Ton :700µs
Toff (100-10%) :2,3ms

[edit]
Xpand X101 :
pi state (ON state with both shutters opened):


Off state :


color fidelity:

transparency (pi state):27%
transparency (on@120Hz):15,8%
Ton :600µs
Toff (100-10%) :1,54ms

[edit]
Crystaleyes CE2 :

I've repaired these glasses, that explain the screws and the white foam for nose piece

Transparency (power off):29,6%
Transparency (on@120Hz) :11,6%
Ton :450µs
Toff (100-10%) :4,3ms

[cybereality]

Interesting. It would be cool to hear your conclusions because I am not exactly sure what the numbers mean. It will also be interesting to see how the Nvidia glasses stack up.

[iondrive]

hurray, more tech notes.

OK, so lower times are better and ...

they all turn off (go from dark to clear) in about 4.4 ms? (+/-0.2)

and the revelators turn on (go from clear to dark) about twice as fast as the others? (Ballpark generalization)


Yes?

[Petrus]

hurray, more tech notes.
OK, so lower times are better and ...
they all turn off (go from dark to clear) in about 4.4 ms? (+/-0.2)
and the revelators turn on (go from clear to dark) about twice as fast as the others? (Ballpark generalization)
Yes?

Yes exactly.
I've edited the topic with the mesurements on the Nvidia 3Dvision.
The results are far better than all my other (and older) glasses concerning the Turn off time and Transparency, but the color fidelity is a bit disappointing, not worse than the others, but I expected better results.

[Petrus]

I've edited the post with the Xpand X101
It's the fastest glasses thanks to its Pi-cell technology but the Nvidia 3D vision is still the brightest.
As for the color fidelity, it's neither worse nor better than Nvidia 3D vision, but different :

[Petrus]

I made the schematic of the test bench :

[Petrus]

I've edited the initial post with the measurements on the Crystaleyes CE2 glasses

[Nanite]

Is the transparency same with the 'glasses active and shutter open' compared 'glasses totally powered off'?

I am thinking of calibrating my coming 3DTV set 3D mode by putting the glasses (or one of it lenses) between the tv and the calibrating probe. I am wondering if I have to do similar kind of circuit to keep the lense active and open.

Great post and setup btw!

[Petrus]

Except on the Xpand, on all other glasses the OFF state (0v across the LCD) is the same as the clear state. So the color will be the same too. maybe there is a slight change caused by transitions but I think it is negligible. All glasses I tested have the same dominant color when working or not.
If you have Xpand glasses, they stay in clear state a couple of minutes after they lost the IR signal.

[Zerofool]

I just saw this thread, very nice idea .
It would be cool if you also post the oscillograms for each pair of glasses, because I remember a strange observation made by xbit labs regarding 3D Vision glasses' timings, they stay open for very short time. Look at the image here:



Did you get the same result?
Anyway, thanks for the good work.

[tritosine5G]

Thats because of LCD's, they mask screen defects like that. Its worse than I tought.
When I tried 2233rz (for 30 min, then gave up), I was very suprised theres no public stuff goin on because of this, it was immidietly apparent.
Then again , with 3d vision & DLP projectors its sure not like that, but 50% duty cycle.

[Zerofool]

They are using different timings for each supported device? Now it makes sense to me why each device has to be certified to work with 3D Vision...
Thanks for the info.

[quakelore]

Petrus,

After all of the testing and specifications which of these glasses ended up being your go to in the long run? It is always fun playing around with glasses and trying to see what we can do to make them better. I am still on the hunt for what fits my niche the most. I feel like the fact I am so hard to please might be backfiring on me when it comes to 3D. I know that IT staffing companies have been emphasizing the education in 3D and this makes me realize why.

[Petrus]

I use Nvidia 3Dvision glasses.
Xpand glasses does not have enough contrast and the others are too dark.