Testing 75Hz
-
- Certif-Eyable!
- Posts: 1139
- Joined: Tue Sep 18, 2012 10:32 pm
Testing 75Hz
I'm currently testing out running the Rift at 75Hz. So far it seems to be working fine on the desktop. I also underclocked my laptop screen to 75 Hz (normally it is 120Hz).
Now I just need to see if I can get demos to run at a smooth 75Hz. Does anyone know if any demos will support a 75Hz Rift?
BTW, I tried running it at 76Hz, and it gave me an error saying it was unsupported, then after that error the nvidia control panel went a bit buggy temporarily. So don't try 76Hz yet. But 75Hz seems to work OK without issues.
Now I just need to see if I can get demos to run at a smooth 75Hz. Does anyone know if any demos will support a 75Hz Rift?
BTW, I tried running it at 76Hz, and it gave me an error saying it was unsupported, then after that error the nvidia control panel went a bit buggy temporarily. So don't try 76Hz yet. But 75Hz seems to work OK without issues.
-
- Binocular Vision CONFIRMED!
- Posts: 203
- Joined: Fri Aug 31, 2012 11:21 pm
Re: Testing 75Hz
I got my DIY Rift clone's 5.6" LCD running at 75Hz. I took it all apart but mirrored the display and played Half life 2 on it with a controller. Everything was smoother(but there was some ghosting) than when running at 60Hz. I couldn't tell if it was dropping frames to make it 60fps and just using more updated frames. Try running the UDK demo with the RIft and update the ini so it's not 60fps limited. Mirror the display and have a fps counter overlay and see on the laptops panel what the demo is running at. compare the Rift at 60 and 75 to see if 75 is smoother or not.
-
- Certif-Eyed!
- Posts: 529
- Joined: Tue Jan 19, 2010 6:38 pm
Re: Testing 75Hz
Still using a CRT as my main display, have not yet found anything that wont run at any refresh rate I want it to. (Less than 75Hz is headache inducing on CRT) A few of the recent big budget games insist that I want 60Hz, but the AMD drivers have settings to correct that. Nvidia? I don't know.
EDIT- Try using FRAPS. I know it is a bit dodgy, and can't tell you for certain if Rift is receiving 75Hz, but if my understanding is correct it will tell you for certain if it is not. Of course if the rift is displaying the 75Hz it receives is much more difficult to check..
EDIT- Try using FRAPS. I know it is a bit dodgy, and can't tell you for certain if Rift is receiving 75Hz, but if my understanding is correct it will tell you for certain if it is not. Of course if the rift is displaying the 75Hz it receives is much more difficult to check..
"If you have a diabolical mind, the first thing that probably came to mind is that it will make an excellent trap: how do you get off a functional omni-directional treadmill?"
- Fredz
- Petrif-Eyed
- Posts: 2255
- Joined: Sat Jan 09, 2010 2:06 pm
- Location: Perpignan, France
- Contact:
Re: Testing 75Hz
I don't know how to measure the display frequency without a high speed camera, but I guess something in a similar vein to this could lead to some solution with a standard camera maybe :
http://www.brekel.com/?p=749
http://www.youtube.com/watch?v=oSiVii2qaJY
http://www.brekel.com/?p=749
http://www.youtube.com/watch?v=oSiVii2qaJY
-
- Certif-Eyable!
- Posts: 1139
- Joined: Tue Sep 18, 2012 10:32 pm
Re: Testing 75Hz
I used Custom Resolution Utility to get fullscreen games to recognise new resolutions. Custom resolutions in Nvidia Control Panel only seem to affect the desktop (and games that run in a window).
It turns out that 75Hz isn't actually possible on the Rift DK1 without some colour distortion on the left half of the right eye
. The maximum frame rate without any colour distortion is 71Hz. The maximum frame rate with a lot of colour distortion is 83Hz. Between those two may be points where the colour distortion is acceptable.
But it's possible to minimize the the amount of distortion by using the optimal timings that I calculated. These timings are optimal (I think) for 71Hz or above:
Active: 1280, 800
Front porch: 4, 1
Sync width: 8, 1
Back porch: 60, 4
Blanking: 72, 6
Total: 1352, 806
Sync polarity: +, - (doesn't seem to make a difference)
Refresh Rate: between 71 and 83 Hz depending on how much latency or colour distortion you want
These timings are the minimums that it will accept. If your Rift DK1 doesn't accept them, you may need to increase them slightly. The vertical timings are the important ones to keep as low as possible.
Below 71Hz you may want to increase the back porch (or is it front porch or sync width?) to force it to write out the pixels to the screen as fast as possible without getting the colour distortion. I don't have a latency tester yet, and don't completely understand how LCD screens work, but that's my guess.
83Hz instead of 60Hz makes a huge difference on the Rift!
Even with the huge colour distortion, 83Hz feels much more real.
The only times you should ever run your Rift at less than 71Hz are when the game won't support other frame rates, or when your game can't keep up.
If your game runs fast enough to render 62 frames per second then in 1280x800 60Hz mode with vsync, you will get 60 FPS. But if you run the same game in 1280x800 71Hz mode with vsync, you will get 35.5 FPS. That's because if every frame isn't rendered in time, they all have to wait for the next frame, so you get half the speed. But I recommend reducing the visual quality to try to get a constant 71 FPS (or 72-83 depending on how much colour distortion you can tolerate). On the other hand, if your game can render at a constant 58 FPS then 71Hz will give you 35.5 FPS instead of 30 FPS with vsync.
You may be wondering about other resolutions. If there is any latency with the downscaling, it is < 3ms and too small for Valve to detect. Other resolutions have about the same amount of colour distortion for the same frame rate. I was surprised by that, because the colour distortion gets worse the higher the total number of lines. But it seems to be the higher the total number of lines for that resolution. So there doesn't seem to be any advantage or disadvantage to running in a higher resolution.
Active: 1440, 900
Front porch: 4, 1
Sync width: 8, 1
Back porch: 60, 5
Blanking: 72, 7
Total: 1512, 1058
Sync polarity: +, - (doesn't seem to make a difference)
Refresh Rate: between 71 and 83 Hz depending on how much latency or colour distortion you want
Active: 1680, 1050
Front porch: 4, 1
Sync width: 8, 1
Back porch: 60, 6
Blanking: 72, 8
Total: 1992, 1058
Sync polarity: +, - (doesn't seem to make a difference)
Refresh Rate: between 71 and 83 Hz depending on how much latency or colour distortion you want
Active: 1920, 1200
Front porch: 4, 1
Sync width: 8, 1
Back porch: 60, 6
Blanking: 72, 8
Total: 1992, 1208
Sync polarity: +, - (doesn't seem to make a difference)
Refresh Rate: between 71 and 82 Hz depending on how much latency or colour distortion you want
Unfortunately, my Rift screen filled up with dust while I was experimenting with this
(damn you Nvidia Control Panel that only shows on the screen I am adjusting
). Does anyone know how to clean dust off the Rift screen?
It turns out that 75Hz isn't actually possible on the Rift DK1 without some colour distortion on the left half of the right eye
![Anaglyph :anaglyph](./images/smilies/icon_Red%20Blue%20Anaglyph%20Small.png)
But it's possible to minimize the the amount of distortion by using the optimal timings that I calculated. These timings are optimal (I think) for 71Hz or above:
Active: 1280, 800
Front porch: 4, 1
Sync width: 8, 1
Back porch: 60, 4
Blanking: 72, 6
Total: 1352, 806
Sync polarity: +, - (doesn't seem to make a difference)
Refresh Rate: between 71 and 83 Hz depending on how much latency or colour distortion you want
These timings are the minimums that it will accept. If your Rift DK1 doesn't accept them, you may need to increase them slightly. The vertical timings are the important ones to keep as low as possible.
Below 71Hz you may want to increase the back porch (or is it front porch or sync width?) to force it to write out the pixels to the screen as fast as possible without getting the colour distortion. I don't have a latency tester yet, and don't completely understand how LCD screens work, but that's my guess.
83Hz instead of 60Hz makes a huge difference on the Rift!
![Woot! :woot](./images/smilies/woot.gif)
![Shocked :shock:](./images/smilies/icon_eek.gif)
The only times you should ever run your Rift at less than 71Hz are when the game won't support other frame rates, or when your game can't keep up.
If your game runs fast enough to render 62 frames per second then in 1280x800 60Hz mode with vsync, you will get 60 FPS. But if you run the same game in 1280x800 71Hz mode with vsync, you will get 35.5 FPS. That's because if every frame isn't rendered in time, they all have to wait for the next frame, so you get half the speed. But I recommend reducing the visual quality to try to get a constant 71 FPS (or 72-83 depending on how much colour distortion you can tolerate). On the other hand, if your game can render at a constant 58 FPS then 71Hz will give you 35.5 FPS instead of 30 FPS with vsync.
You may be wondering about other resolutions. If there is any latency with the downscaling, it is < 3ms and too small for Valve to detect. Other resolutions have about the same amount of colour distortion for the same frame rate. I was surprised by that, because the colour distortion gets worse the higher the total number of lines. But it seems to be the higher the total number of lines for that resolution. So there doesn't seem to be any advantage or disadvantage to running in a higher resolution.
Active: 1440, 900
Front porch: 4, 1
Sync width: 8, 1
Back porch: 60, 5
Blanking: 72, 7
Total: 1512, 1058
Sync polarity: +, - (doesn't seem to make a difference)
Refresh Rate: between 71 and 83 Hz depending on how much latency or colour distortion you want
Active: 1680, 1050
Front porch: 4, 1
Sync width: 8, 1
Back porch: 60, 6
Blanking: 72, 8
Total: 1992, 1058
Sync polarity: +, - (doesn't seem to make a difference)
Refresh Rate: between 71 and 83 Hz depending on how much latency or colour distortion you want
Active: 1920, 1200
Front porch: 4, 1
Sync width: 8, 1
Back porch: 60, 6
Blanking: 72, 8
Total: 1992, 1208
Sync polarity: +, - (doesn't seem to make a difference)
Refresh Rate: between 71 and 82 Hz depending on how much latency or colour distortion you want
Unfortunately, my Rift screen filled up with dust while I was experimenting with this
![Sad :(](./images/smilies/icon_e_sad.gif)
![Evil or Very Mad :evil:](./images/smilies/icon_evil.gif)
-
- Petrif-Eyed
- Posts: 2708
- Joined: Sat Sep 01, 2012 10:47 pm
Re: Testing 75Hz
A possible explanation of color distortion from running too fast is that the Rift uses temporal dithering to obtain a wider color gamut.
This can be observed as a slight "dot crawl" similar to what can be seen on brightly colored edges in NTSC composite video. I could see this when I had the C-cup lenses in my Rift (for larger sharply focused pixels) and I viewed the screensaver image from "The Gallery, Six Elements" you can download from the kickstarter page. The high contrast edge near the hallway corner shows slight green moving diamond shapes, as a visible artifact of this temporal dithering.
It is very possible that some screen refresh rates have temporal aliasing with that color dithering frequency. Perhaps a slight frequency tweak could alter or eliminate the "too fast" color problem.
Or at least that is what I suspect, not having performed such experiments myself yet. But I HAVE observed temporal dithering (dot crawl) artifacts in my Rift as described above, so I know this could potentially be a cause of the color artifact symptoms as described when running faster than 71 Hz.
Note: That lagom.nl site linked above has some GREAT LCD testing and analysis tools, for measuring and adjusting many things including latency, gamma, and much more.
This can be observed as a slight "dot crawl" similar to what can be seen on brightly colored edges in NTSC composite video. I could see this when I had the C-cup lenses in my Rift (for larger sharply focused pixels) and I viewed the screensaver image from "The Gallery, Six Elements" you can download from the kickstarter page. The high contrast edge near the hallway corner shows slight green moving diamond shapes, as a visible artifact of this temporal dithering.
It is very possible that some screen refresh rates have temporal aliasing with that color dithering frequency. Perhaps a slight frequency tweak could alter or eliminate the "too fast" color problem.
http://www.lagom.nl/lcd-test/black.php wrote:... On many LCD monitors, the darker shades show up as dynamic or static dithered patterns. ... Temporal dithering, frame-rate control (FRC), or dynamic dithering look like quickly moving patterns ... Many monitors cannot make small brightness steps, especially in the darker shades. Instead, they rapidly alternate between darker and brighter shades for the individual pixels, such that at least on average the brightness is correct. ... In older and/or low-end monitors, the dithering pattern may not look like noise, but rather like regular patterns ... Older LCD monitors and many laptop screens do not do temporal dithering at all, but rather static dithering, i.e. they use a fixed pattern of darker and lighter pixels ... How much you notice dithering is a combination of factors such the response time of the monitor, how intelligent the dithering algorithm is, the voltage response curve of the LCD pixels, and finally the number of bits. ... In any case, if you are interested: if a monitor has an 8-bit driver, it does not necessarily mean that there is no dithering. Such a driver produces a voltage that can vary in 256 steps, but unfortunately the luminance (amount of light emitted) from the pixels is a highly nonlinear function of the voltage ... In order to approximate the ideal luminance curve, dithering is still necessary, which is most noticeable in the darker shades where the natural response of the liquid crystals in the display deviates most strongly from the ideal luminance curve. To make it more complicated, an advanced dithering algorithm can also produce 256 shades with the correct luminance curve on a 6-bit display. The difference is that the dithering ("noise") will be more pronounced ...
Notice how "30 Hz flicker" is mentioned regarding such temporal dithering. This could imply that it works best at 60 Hz, with potential color artifacts at other frequencies. Such temporal color artifacts could change significantly with minor frequency variations, perhaps barely noticeable at some particular fractional Hz framerate variations.http://en.wikipedia.org/wiki/Frame_Rate_Control wrote:Frame rate control (FRC) is a method for achieving higher color quality in low color resolution display ... FRC is a form of temporal dithering which cycles between different color shades with each new frame to simulate an intermediate shade. This can create a potentially noticeable 30 Hz flicker. FRC tends to be most noticeable in darker tones, while dithering appears to make the individual pixels of the LCD visible.
Or at least that is what I suspect, not having performed such experiments myself yet. But I HAVE observed temporal dithering (dot crawl) artifacts in my Rift as described above, so I know this could potentially be a cause of the color artifact symptoms as described when running faster than 71 Hz.
Note: That lagom.nl site linked above has some GREAT LCD testing and analysis tools, for measuring and adjusting many things including latency, gamma, and much more.
This work is licensed under a Creative Commons Attribution-ShareAlike 3.0 Unported License. ![Image](http://i.creativecommons.org/l/by-sa/3.0/80x15.png)
![Image](http://i.creativecommons.org/l/by-sa/3.0/80x15.png)
-
- Cross Eyed!
- Posts: 108
- Joined: Thu May 30, 2013 5:46 am
Re: Testing 75Hz
nice...so you can have it run at 71 all the time!!
I'll try this tonight.
At Geekmaster...i couldn't figure out why a digital stream could start fudging up colors...nice to see you on top of things.
I'll try this tonight.
At Geekmaster...i couldn't figure out why a digital stream could start fudging up colors...nice to see you on top of things.
- KBK
- Terrif-eying the Ladies!
- Posts: 910
- Joined: Tue Jan 29, 2013 2:05 am
Re: Testing 75Hz
Now we are stepping into my fun area. I'll definitely play with this.
A much needed bit of information.
Thanks for the efforts.
A much needed bit of information.
Thanks for the efforts.
Intelligence... is not inherent - it is a point in understanding. Q: When does a fire become self sustaining?
-
- Certif-Eyable!
- Posts: 1139
- Joined: Tue Sep 18, 2012 10:32 pm
Re: Testing 75Hz
Yes, I was thinking about you when i was doing it.KBK wrote:Now we are stepping into my fun area. I'll definitely play with this.
A much needed bit of information.
Thanks for the efforts.
- KBK
- Terrif-eying the Ladies!
- Posts: 910
- Joined: Tue Jan 29, 2013 2:05 am
Re: Testing 75Hz
We are digitally controlling an analog monitor. the pixels may be considered to be 'digtal' but the signals are still analog in fashion and the given signals still occur in a simple analog fashion.
One can push a given input system that sets it's channel timing and widths via a analog signal trigger and controllable timing windows.
Just like we did in the old days, with VGA and even old computer systems.
hot-rodding digital still requires a full understanding of the analog electrical system that it is built on.
I will admit that my rift is already heavily modded out and probably has more contrast and color range than any other. Subjectively, about 15-20% higher.
I would also hazard it has more stable framing and I can 'see' the behavior of the IMU far better, with much less 'noise' getting in the way.
One can push a given input system that sets it's channel timing and widths via a analog signal trigger and controllable timing windows.
Just like we did in the old days, with VGA and even old computer systems.
hot-rodding digital still requires a full understanding of the analog electrical system that it is built on.
I will admit that my rift is already heavily modded out and probably has more contrast and color range than any other. Subjectively, about 15-20% higher.
I would also hazard it has more stable framing and I can 'see' the behavior of the IMU far better, with much less 'noise' getting in the way.
Intelligence... is not inherent - it is a point in understanding. Q: When does a fire become self sustaining?
-
- Sharp Eyed Eagle!
- Posts: 427
- Joined: Fri Jan 18, 2013 7:13 am
- Location: Gothenburg, Sweden
Re: Testing 75Hz
Very interesting! Think i have to try this : )