Page 25 of 70 FirstFirst ... 15232425262735 ... LastLast
Results 361 to 375 of 1041

Thread: MUX'ing, VSplit, and MPG2 files.

  1. #361
    Join Date
    Jan 2002
    Posts
    4,809
    Yeah. Hilarious is not quite the word I would use. Especially now that it has me freaked out. What a scary thing. If the venom has gone just a touch further, <1mm, there would have been permanent nerve damage and possibly amputation. Now that is frightening...

    --jdiner

  2. #362
    Join Date
    Jan 2002
    Posts
    4,809
    Welp folks. I am tempted to quit right now and just give up on this tivo stuff...

    I went downstairs to test the latest version on the player I have hooked up to my tv to find that my machine down there, where I have been storing all of my transcoded shows for the last year, and my unprocessed TyStreams from my Tivos had a hard drive fail.

    My raid array is now a worthless pile of junk. I can't tell you how frustrating that is... I could just scream. All of last season of Enterprise, Smallville, etc... Just gone. Maybe IBM can pull the data off of the drive if I warantee it. But come on... the drive is about... 10 months old.

    --jdiner

  3. #363
    Join Date
    Jan 2002
    Posts
    4,809
    Hummm. That is wierd. In re-working the stream interleaving code I discovered something truly wierd.

    In doing the new SCR calcuation. I need to know exactly how many bytes went by for the audio and video. So I started counting out 30 frames, which is what the data reports as the frame rate. Only to find that things worked for a short time and then totally fritzed...

    So I started digging into the PTS values in the data. It quickly became obvious that it is actually 25 frames/sec. (Keep in mind the doubling that takes place...) But that is wierd. That is the PAL frame rate but that is what the data shows...

    Anyone heard anything about that kind of thing? It is just a bit to bizarre... But before you all think I am going crazy. I changed the mux'er from 30 frames/sec to 25 frames/sec and it works perfectly... Go figure...

    Any comments? Thoughts? Rumours you have heard?

    --jdiner

  4. #364
    Join Date
    Aug 2002
    Posts
    52
    Ooof, sorry about the drive - that sux!

    Q

  5. #365
    Join Date
    Jan 2002
    Posts
    4,809

    Vsplit #13c Mux1d...

    Alright folks here you go. It is not perfect. But it is testable...

    This version includes audio as well as video output in the mux'ed stream. However, this version WILL NOT play an SA or UK Tivo stream. I had to tighten things up a bit while trying to figure out some problems. The next version will once again support those platforms.

    Stick with 5 minute files as I have never had anything actually run that long with this one. I get between 1:30 and 2:35 minutes worth of good sync'ed playback before things get... whacked out. And when it goes it is pretty much immediate.

    I know what is causing the current problem it is just too late and I have had enough for today... The bad news is I have still have yet to fully fix it. The good news is that I have determined the proper calculation steps for all of this. You would be amazed at what all had to be brute forced to get even this far.

    Things to look for:

    1- That it starts out sync'ed correctly.

    2- That stay sync'ed while it plays normally.

    3- That the video and audio play smoothly.

    4- That the program does not crash, etc...

    The output has been toned down but it is still pretty massive. Redirect it and then for the most part ignore it.

    Let me know what kind of results you get.

    --jdiner

  6. #366
    Join Date
    Jan 2002
    Posts
    4,809
    Speaking of kitchens I did some work for a company in CA recently. I found on one of the trips that the CFO of this up and coming young tech company, had decided one night after working long and hard to run some chemistry experiments in the kitchen at the office.

    The gathering of Hydrogen was going well. The process working smoothly. The only real problem appears to have been the measuring of exactly how much had been gathered. Although perhaps the biggest single mistake was the burning split test. The concussion should have killed him, the flying glass and shattered equipment cut through entire cases of soda... And he, being the luckiest man alive, got not even a major scratch...

    Wild...

    --jdiner

  7. #367
    Join Date
    Jan 2002
    Posts
    4,809
    BTW, I was right about what is going wrong with the mux'ing engine at 2:35 in my Smallville test clip. The SCR value resets.

    The SCR is what is used for seeking around inside of a playing stream, and for most of the rest of the syncing in playing in general.

    My code correctly generates that value to a certain point. But after 2:35 in the obscura clip it resets to 0:0:0.0 which is a no-no. Shouldn't be that hard to track down though. I expect that it is a matter of the math. I multiply by some pretty large numbers at times. I will have to switch to a long double or find another way to do it. Since wrapping at 2:35 is pretty useless over all...

    I think it is prececision because in 7 minutes of source it wraps 3 times at 2:35...

    In my DeadZone clip it wraps at 2:35 as well.

    Technically it is going by so fast I am not sure, so let me check for sure. Actually it wraps at

    !!!: SCR cal = 00:02:39.072

    Same place everytime give or take .001...

    Now that is what I like to see. A nice repeatable error... Makes them easy to find at least.

    --jdiner

  8. #368
    Join Date
    Jan 2002
    Posts
    4,809
    You know I seem to be the only one here. Here I work late into the evening and no one is even around to be impressed or take it for a spin...

    Fine, then I will go to bed too... I doubt I will be up early so give is a whirl and let me know what happened...

    --jdiner

  9. #369
    Join Date
    Jun 2002
    Posts
    30

    Smile

    Originally posted by jdiner
    You know I seem to be the only one here. Here I work late into the evening and no one is even around to be impressed or take it for a spin...
    Plenty of people up over here in the UK... but the latest version doesn't work on UK tivos! :-)

  10. #370
    Join Date
    Feb 2002
    Posts
    116
    You're not the only one here... about to give it a shot.

  11. #371
    Join Date
    Feb 2002
    Posts
    116
    Tried it on a Showtime clip, which the previous (non-audio) version liked. PowerDVD blasts through the 5 minutes of clip in about 10 seconds, like it's on super fast forward, and I get little clicks and chirps for audio. WinDVD starts to do the same thing then crashes after a couple seconds. This clip has Dolby Digital sound, which is probably what's breaking.

    On a clip from a local channel (no DD sound) it's in perfect sync for the first minute in PowerDVD. I can even seek! At exactly 1 minute in, the video gets choppy all of a sudden, and gets worse over time. The same happens in WinDVD, but 10 seconds in.

    Will try on some more clips tomorrow.

  12. #372
    Join Date
    May 2002
    Posts
    234
    jdiner,

    is it possible that the show you were examining looking at the frame rate was attempting to do 24 fps (i.e. film movie frame rates) instead of PAL or NTSC

    SA tivos would use 30 fps, but Dtivos would probably use whatever was most appropriate for the program being encoded (just like the scan direction varies show to show)

  13. #373
    Join Date
    Sep 2002
    Posts
    110
    Originally posted by jdiner
    All of last season of Enterprise, Smallville, etc... Just gone.
    --jdiner
    I have about half of the last season of Enterprise saved, there yours if you want whem.

  14. #374
    Join Date
    Feb 2002
    Posts
    54
    I've been super busy lately, but I did find the time to check out several clips.

    Observations so far:

    The few clips I've tried so far run great initially, then massive stuttering commences. The duration in which they run well isn't fixed, and varies from 27 to 73 seconds.

    Oddly enough, to clear the stuttering effect, I have to close and re-open the file. If I just rewind back to the beginning and press play, it keeps doing it (PowerDVD XP decoder).

    One bonus so far however though:

    One of my test clips when muxed with tmpgenc, simply locked up every player I've ever tested it in if I allowed it to run past the 19th second from the beginning. The mpg muxed with this new vsplit-mux13d or whatever it's called (new as of 10/10/02) plays fine past this point.

  15. #375
    Join Date
    Jul 2001
    Posts
    657
    I'm sorry I've been up to my neck at work, missed out on alot of the testing, but trying to keep up with the thread. I hope to be "back to normal" by the weekend.

    but I have to say, jdiner, I have all of the last season(and all of this season so far) of enterprise on disk, thanks mostly to you and would be glad to help.

    just let me know.

    Thanks!

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •