X-bit labs

Started by: MonkRX | Date 01/23/07

Comments: 51 | Last Comment:  07/08/08

[1-20 | 21-34]

1. Very In depth Stuff. Awesome. Good work.

2. Awesome, if not labriously-long-winded, article.

On second thought, a summary is in order.

3. Comprehensive article that is easy to understand even for nubs

4. I agree with others, this article is definitely very good and provides lots of useful information. Great job X-bit Labs!

5. Actually single link TMDS can be made to run 1920x1200@60Hz, by using reduced blanking. Just don't try that with low-end GeForce FX cards, which uses integrated 141MHz TMDS transmitter instead of 165MHz.

Also, single link DVI/HDMI is capable of 1920x1080p72 or higher. But achieving it depends on the sender+cable+receiver+DSP+display.

I have helped DTV forum acquaintances get the Pyrod 3711 37" LCD TV up to 1920x1080p72 (see http://www.dtvforum.info/index.php?showtopic=31037&st=44), and others have run an Acer AT3705-MGW 37" LCD TV at 1920x1080p74.4 over single link DVI (see http://www.dtvforum.info/index.php?showtopic=28451&st=1025).

6. Very detailed indeed. Would like to save to whole article in a single page or file for further reading and better understanding. Makes lot of sense, though a good knowlegde of physics is required to actually admire it.
I would like more points in the summary and how to make better use of LCD's. (like white paper for setting contrast is excellent)

Anyway excellent work and thanks for enlightening (at 60Hz :-)) us all.

"Would like to save to whole article in a single page or file for further reading and better understanding"

http://www.xbitlabs.com/articles/other/print/lcd-parameters.html
("Print article" link on the left sidebar)

7. "...you can only spot non-synchronization in movies when there is a difference of 200-300 milliseconds between video and audio."

This info is incorrect.

47ms can be detected. Plus lags can add up. For example if you watch video where audio is 40ms ahead of video, monitor adds another 47ms lag. This means that you are practically watching video with 87ms lag which is not only within range of detectability but almost out of range of acceptability.

"A study done in the 1940s by Bell Laboratories in the U.S. concluded that when audio led video by more than 35 milliseconds, or lagged video by more than 100 milliseconds, a/v out-of-sync will be detected."

http://www.tvtechnology.com/features/Tech-Corner/f_rf_technology_corner-05.14.03.shtml

A delay that is under a tenth of a second may be visible to someone who is searching for it but to a viewer unaware of any latency factors such a discrepancy will probably go unobserved. People also have a tendency to see problems that they have been told exist only after they hear of them as the author pointed out at one point in the article.

“You can even read some funny posts like, ‘I’ve been sitting at the discussed monitor for two weeks, but it’s only now, after I’ve read the forum, that I see the lag’.”

If a problem takes serious determination and effort to notice it is not an issue in my book.

8. Re: input lag

So... the author never played on-line FPSes? The ping of ~50ms is entirely playable, but if you add 50ms more to that it gets noticeable (esp. with long-range weapons), and if your /starting/ ping is ~100ms then a jump to 150ms is likely to have a significant effect on your performance.
Anything over 200ms will most likely make "twitch based" on-line FPSes just frustrating.

Ping and monitor latency are _very_ different things.

Sure they're different things, but the effect is the same: a further separation of the game state (on the server) from my brain state.

Regardless, it seems clear that monitors can have < 10ms lags without sacrificing other performance aspects, so serious gamers are not likely to consider monitors with 40ms+ lags. And there are many enthusiast gamers who are willing to pay substantial premiums for marginal hardware improvements.

No, effect is not the same.

Network lag 100 ms means that you have 10 frames per second. You videocard may produce 100 fps or more, but game information updates from server every 100 ms = 10 _effective_ fps. Game engine interpolate motion so it's still smooth, but it' s only interpolation.

Monitor lag 50 ms means that you have all your 60 fps (vertical refresh rate 60 Hz) but shifted in time by 50 ms.

So you can't say "100 ms network lag + 50 ms monitor lag = like 150 ms network like", they are different. Network lag is much more sensible than monitor lag.

9. Excellent article. It corrected a number of misconceptions I had about LCDs. Now I would like to see a comparision between the various panels; 6-bit TN, 8-bit TN, pva, mva and s-ips.
Hoping you come up with a detailed comparision soon.

"Now I would like to see a comparision between the various panels; 6-bit TN, 8-bit TN, pva, mva and s-ips"

Did you read this?
http://www.xbitlabs.com/articles/other/display/lcd-guide.html
http://www.xbitlabs.com/articles/other/display/response-compensation.html

10. Not sure I buy everything you said (especially the lag part), but I ganed a lot of insight into the color gamut and temperature as well as the response times.

11. Very interesting and engaging article!

12. That was very informative. Thank you.

13. By the way, I love the pig in the Article's Icon/Picture.

:)
http://olegart.ru/img/DCIM/pig-and-xbit.jpg

14. Very informative, very useful information.

15. Excellent article. You explain all the concepts that I read about in forums and reviews. And your opinions are supported by facts. A keeper

16. I' working with audio stuff, latency is a big subject, and for playing music the difference between 3ms or 23ms latency (to overdubs) is very sensible. The author spends pages trying to explain why input lag doesn't matter. No wonder why it's so long to explain, because it matters. Some people, like musicians, are used to sensing timings, and for those people 50ms late is very sensible: pc feels slugish.

Human brains are much more sensible to audio lag.

"Some people, like musicians, are used to sensing timings"

Audio timings, not video. Otherwise they could not watch movies -- there is up to 40 ms (1 frame @ 25 fps) lag between audio and video :)

I think your figure for the difference (about 300ms) between audio and video that isn't noticeable is quite a bit too high though. I would think such differences would be very noticeable with speech.

17. Great article, thanks a lot!

18. I've never seen anyone running their 17" LCD at native resolution. The pixel size is way too small!

I wish I could read this article before consenting to purchase my Samsung 730B. Back then (roughly a year ago), there was absolutely no discussion of this annoyance.

19. I agree with Cluster: input lag is real, it is noticeable (if just implicitly for most people), there are applications such as music and action, driving sim (and music) gaming where it's unacceptable for demanding users who will definitely use it as an important factor in purchase decision.

So why such, frankly a bit hysterical attempts to tell people that input lag doesn't matter? It clearly it does, and there are good and objective reasons why it does. It may not matter for the reviewer, but quite a few people do in fact disagree and don't quite buy the obsessive arguments against why they shouldn't care. If there's a screen with 50ms input lag and another similar one without, all other factors similar, then why even bother to attempt to convince oneself and other people that it doesn't matter?

FWIW, I have a PS2 hooked up to a PC capture card and a CRT monitor. There's about 2 frames of lag involved, which is quite negligible even for racing games, but there are a few, in fact RPGs like for Shadow Hearts and Magna Charta, that do require precise timing and are hindered by display lag.

"it is noticeable"

Where can I see results of double blind test?

I have already posted a link where you can read about results of scientific tests. Yes they are about AV synchronization but they provide the necessary info. They all conclude that 35-45ms lag is detectable and that acceptability thresholds (meaning passing that people will reject the video), are set at 90 ms.

Here&#8217;s another link: http://www.inventa.com.au/Audio-Video%20Out%20of%20Sync%20in%20PC%20Video%20Capture.htm
(&#8230;and many more similar articles can be found via google)

Now you provide us with some scientific tests that prove us wrong.

20. What an excellent article! Great work!

[1-20 | 21-34]