Originally Posted by Nillaz
Thank you for the clarity and the speed of your response. I also have an Integral laying around and after much wrestling with things I was able to get this working using your suggested method.
One question I have is how confident are you in the lag test results from the Diva? I am recording results that in some cases seem suspiciously low. It's possible that Samsung has tweaked their displays to achieve lower latency but there are several tests where I am repeatedly seeing results of less than 10ms. This is an astonishingly low figure for any tv even in game mode so I'm trying to determine if this is actually correct or due to user error.
There is much to say about input lag, could wrote a few pages about it as we already had extensive discussion with magazines and online reviewers on this topic.
bottom line is this:
- always compare several tv with same sources, same content, same signal, changing any parameters upfront can lead to huge difference at lag test results.
- make at least 10 or 20 measurements and the TV average lag is the measurement results average. (disconnect and reconnect HDMI cable between each test)
- wait at least 30 sec when you change a TV mode from one to an other, many TV start to activate or disable processing options gradually, for ex, C9 game mode is optimal after 20sec only.
- HDMI2.1 with QFT have lower input lag than any other.
- HDCP play a role on input lag, that's why DVI solution for testing input lag are irrelevant
- When testing 60Hz content, frame time is 16.66ms which means that depending of sync pulse, there is potentially a +/- 8ms possible difference on any 60Hz content (4ms on 120Hz, 2ms on 240Hz, etc..) so higher the refresh rate, more accurate is the lag measurement, that's why it is recommended to run multiple measurement on 60Hz and make an average from them and that represent the TV average input lag.
As conclusion i would say that it is misleading to give a TV input lag as a solid static number. such thing does not exist in reality.
input lag should be given as an average of what user can expect from the TV. (unless we have 1000Hz content and then input lag would be a solid not changing number always)
This being said, if you dig google a bit you will find much better technical explanation for all what i wrote above.