DL3 fixture video speeds

kristaps.krkristaps.kr Registered User
edited May 2011 in DL.3/DL.2
Hello

What could cause DL3 video display speed variations when it is used without synchronisation.

We have 2xDL3, all the same firmware, they run on the similar hardware 1:1, one them after 5 minutes work plays video faster 2 seconds than other.

There is no question about syncing, question is about 2 independent DL3's which not playing at the same speed. Ok, I understand than in timeline after week it could be different, but in 5 minutes?

Thnx

Comments

  • schiefelsschiefels Registered User, HES Alumni
    edited May 2011
    In short, yes it is possible. There are many factors that could cause inconsistent playback between fixtures:

    1) Always the most important is Content and how it's encoded. Things like not having all i-frames, no End of Sequence headers, interlaced encoding instead of progressive, etc. can all cause slight variations in how the videos will playback. If you are seeing this with stock content, then none of those things should be factors since all of the stock content is encoded properly. However, if you are using User Content, be sure that it is encoded properly. I will be happy to take a look at a piece or two of the content and check it for encoding.

    2) How the movie playback is being started. We always recommend a 'set-up' cue that loads the movie and puts it into the "Pause and Rewind" play mode. Then, on the next cue, fade up the intensity/opacity and put the movie into "Loop forward" play mode. This process makes sure that the movies start at the same time. If you simply load the movies directly into loop forward and start them that way, there is a good chance the movies will not start at the same time due to differences in loading times between fixtures. You could see up to 1/4 second difference right off the bat unless the first process is followed.

    3) Computer hardware. You said they run "similar" hardware, but how similar? Small differences in the computer hardware can cause large differences in playback.

    4) Just the general way in which Windows XP and Windows applications work. Each computer has its own clock that the graphics engine application runs against, and the clocks are not sync'ed to anything. Also, we have tried to slim down the application and operating system as much as possible, but there are still background processes and other things running on the DL servers that can cause slight variations in playback performance.

    There's more to it, but these are just some of the reasons why network sync was created in the first place. When Collage Generator was first implemented (which was before network sync worked properly), we were happy when we could get two DL servers to stay in sync for a minute before there was noticeable playback drift. We have made a few playback improvements since then, but the big leap came with network sync. We always recommend using it for Collaging or in situations where two or more servers need to keep their movies in time.

    Hope this helps. I'll be happy to answer any questions.

    Thanks,
  • kristaps.krkristaps.kr Registered User
    edited May 2011
    Hello Matt

    1) It is original DL3 content. To be precise it is HES logo turning around itself. So this is not changed from factory.

    2)Playback is started from Wholehog III console. So at the first time it is accurate. But we will try then "Loop forward" play mode and let you know how it looks.

    3,4) Similar on hardware labels, technically ... who knows and yes, Windows is always right, no matter what. :)

    Problem with these speeds what we care about, happens in synchronized network.
    Problem comes the way that let's say "fixture 1" - F1 have constant speed X on which it works, then you have "fixture 2" - F2 with speed X+0,1. When they work together it comes that F2 finished video and have its video time shorter than F1 and it waits for F1 to end, BUT when F1 starts to move again, as it should not wait for anything to end and without interruption, then until F2 starts then it have some "speed up" to get together with F1.

    In short it is like pendulum with video playing and synchronisation together. Of course it is not coming like earthquake at the end, but it makes video lags in time, like "speedups" and "slowdowns" at video start and end.

    Whoaaaa, so complicated, hope that you understand me.
  • Marty PostmaMarty Postma Registered User
    edited May 2011
    Similar on hardware labels, technically ... who knows and yes, Windows is always right, no matter what. :)

    If you have CMA hooked up to your network you will be able to see if the DL.3 servers have the same hardware.....processor will make the biggest difference, if for example one of your units is a Core2Duo and the other is QuadCore.

    As Matt stated, proper setup/mark cues will help get past this (DO NOT use the built in Mark function on Hog-3 for this, you MUST do a manual mark with the Pause+Rewind playmode to get proper results).
    ...in synchronized network.
    Problem comes the way that let's say "fixture 1" - F1 have constant speed X on which it works, then you have "fixture 2" - F2 with speed X+0,1.

    Are you using different playback speed settings on the fixtures? Check to be sure that they are set the same.
    In short it is like pendulum with video playing and synchronisation together. Of course it is not coming like earthquake at the end, but it makes video lags in time, like "speedups" and "slowdowns" at video start and end.

    If you turn on "Statistics" from your Global Control Mode are you seeing a difference in playback speeds, dropped frames, etc?

    Hope this helps. :)
Sign In or Register to comment.