Page 5 of 18 FirstFirst ... 3456715 ... LastLast
Results 61 to 75 of 263

Thread: TyTool 10r1 - Extraction/Frame Accurate Editing/DVD output/Closed Captioning...

  1. #61
    Join Date
    Dec 2004
    Posts
    831

    Poor Network Response

    I have been getting rather poor network response of late. I noticed this quite some time before installing 10r1, so I'm fairly confidant it isn't the new release. Prior to the last few weeks, I would consistently get around 2 MBps, sometimes up to 2.2MBps, but recently I've been getting around 0.6 MBps. I can't think what the issue would be. I can easily transfer more than 25MBps to and from the file server, and Double Socket mode is working just fine. FTP transfers to and from the TiVo run right near the 2MBps mark. The tivo is sitting around 78% idle, and tserver is the largest CPU user during a transfer. Does anyone have any ideas?
    Last edited by lrhorer; 12-17-2005 at 05:50 PM.
    Having trouble with TyTool? Try TyTool Documentation
    Need to hack an S3 / THD? Try S3 Hacking Script

  2. #62
    Join Date
    Dec 2002
    Posts
    72

    New LII audio bug ?

    I just tried to use tytoolr10 to download and multiplex an HD recording off an HDTIVO.
    What I get is continued
    New LII audio bug Detected... Correction in progress...

    messages, and the saved mpg file has video but no audio.

    Suggestions?

  3. #63
    Join Date
    Apr 2003
    Posts
    2,402
    Quote Originally Posted by lrhorer
    I have been getting rather poor network response of late. I noticed this quite some time before installing 10r1, so I'm fairly confidant it isn't the new release. Prior to the last few weeks, I would consistently get around 2 MBps, sometimes up to 2.2MBps, but recently I've been getting around 0.6 MBps. I can't think what the issue would be. I can easily transfer more than 25MBps to and from the file server, and Double Socket mode is working just fine. FTP transfers to and from the TiVo run right near the 2MBps mark. The tivo is sitting around 78% idle, and tserver is the largest CPU user during a transfer. Does anyone have any ideas?
    What happens when you try Single Socket mode?

    ew

  4. #64
    Join Date
    Dec 2004
    Posts
    831

    Huh?!?

    Quote Originally Posted by eastwind
    What happens when you try Single Socket mode?ew
    I hadn't tried it.

    Oh, now that's weird. The throughput doubles in Single Socket mode. It shouldn't do that. I'm getting about 1.3MBps in Single Socket mode.
    Having trouble with TyTool? Try TyTool Documentation
    Need to hack an S3 / THD? Try S3 Hacking Script

  5. #65
    Join Date
    Apr 2003
    Posts
    2,402
    Quote Originally Posted by lrhorer
    I hadn't tried it.

    Oh, now that's weird. The throughput doubles in Single Socket mode. It shouldn't do that. I'm getting about 1.3MBps in Single Socket mode.
    I just remembered seeing something about S2s transferring faster in Single Socket mode. Don't know if there was a reason given, but I think it was just an observation.

    ew

  6. #66
    Join Date
    Jul 2005
    Posts
    20
    Quote Originally Posted by kmt
    I just tried to use tytoolr10 to download and multiplex an HD recording off an HDTIVO.
    What I get is continued
    New LII audio bug Detected... Correction in progress...

    messages, and the saved mpg file has video but no audio.

    Suggestions?
    This actually seems to be a problem with other types of streams as well. I just pulled one off of Showtime from standard and it was all busted. May be a problem in the audio hole fix?

    Worked fine on a regular ESPN stream.

  7. #67
    Join Date
    Dec 2004
    Posts
    831

    Network Speed

    Quote Originally Posted by eastwind
    I just remembered seeing something about S2s transferring faster in Single Socket mode. Don't know if there was a reason given, but I think it was just an observation.
    It shouldn't (And reportedy doesn't) make much difference for an S2, but this is an S1.

    I think I've found the problem, however, and it's not good. Our company went to a new VPN client (Cisco) a few weeks ago, and it automatically sets the MTU to 1300. Without the MTU set to 1300, the client croaks. With the MTU set to 1300, double socket mode creeps along. I'm not sure why this is the case, at all, but when I set the MTU back to 1500, I get rates in excess of 2.2MBps.

    Jamie,

    Do you have any notion why an MTU of 1300 would cause tserver to slow down by 75%? I don't even see why it makes much of a difference, at all.
    Last edited by lrhorer; 12-17-2005 at 08:16 PM.
    Having trouble with TyTool? Try TyTool Documentation
    Need to hack an S3 / THD? Try S3 Hacking Script

  8. #68
    Join Date
    Apr 2003
    Posts
    2,402
    Quote Originally Posted by lrhorer
    It shouldn't (And reportedy doesn't) make much difference for an S2, but this is an S1.

    I think I've found the problem, however, and it's not good. Our company went to a new VPN client (Cisco) a few weeks ago, and it automatically sets the MTU to 1300. Without the MTU set to 1300, the client croaks. With the MTU set to 1300, double socket mode creeps along. I'm not sure why this is the case, at all, but when I set the MTU back to 1500, I get rates in excess of 2.2MBps.

    Jamie,

    Do you have any notion why an MTU of 1300 would cause tserver to slow down by 75%? I don't even see why it makes much of a difference, at all.
    Sorry if I'm butting in (since you were asking Jamie)....

    Could it be because the TiVo is trying to push an MTU of 1500? Would that cause enough overhead to slow it down that much?

    ew

  9. #69
    Join Date
    Aug 2004
    Posts
    4,075
    Quote Originally Posted by eastwind
    Could it be because the TiVo is trying to push an MTU of 1500? Would that cause enough overhead to slow it down that much?
    Yes, I expect if you set the tivo mtu to 1300 too your performance would improve (but it would be a little worse than with a 1500 mtu). I would have thought the linux kernel would do path mtu discovery to choose an appropriate end-to-end mtu, but maybe not. I suppose you can check /proc/sys/net/ipv4/ip_no_pmtu_disc. If 0, it _should_ be doing pmtu.

  10. #70
    Join Date
    Nov 2004
    Location
    Gurnee, IL
    Posts
    2,384
    Quote Originally Posted by eastwind
    Could it be because the TiVo is trying to push an MTU of 1500? Would that cause enough overhead to slow it down that much?
    Heck yeah...badbad fragmentation. As Jamie pointed out, linux ought to be doing mtu discovery, but if it's not...
    --
    Christopher D. Heer
    Quote Originally Posted by Oscar Wilde
    Perhaps, after all, America never has been discovered. I myself would say that it had merely been detected.

  11. #71
    Join Date
    Feb 2003
    Posts
    295

    quick network question

    I bought gig e adapters and switch that support jumbo frames. I have 2 HDTIVOs with the latest backported drivers and I have set them to jumbo frames as well. Do all the devices on the network that are running gig e have to have the exact same MTU settings or can they be set at different values. At one point I got over 5megs/sec using the new tytool but as I have been messing around with things it went down to 2.5 or so. Any guidance appreciated,

    Snoots

  12. #72
    Join Date
    Dec 2004
    Posts
    831

    Mtu

    Quote Originally Posted by Jamie
    I suppose you can check /proc/sys/net/ipv4/ip_no_pmtu_disc. If 0, it _should_ be doing pmtu.
    It's zero.
    Having trouble with TyTool? Try TyTool Documentation
    Need to hack an S3 / THD? Try S3 Hacking Script

  13. #73
    Join Date
    Apr 2002
    Location
    USA
    Posts
    116

    Auto-generation of .chp files?

    Is 10r1 missing a preference setting to disable auto-generation of .chp files, or did I miss it? I swear I had that disabled in 9r18 since I almost never go to DVD, but maybe I imagined it.

    Thanks,
    -Zak

  14. #74
    Join Date
    Jan 2002
    Posts
    4,809
    Quote Originally Posted by falcomadol
    JD: You previously indicated that music channel streams would be working in ?this version? (next public version):
    Yeah. It got eaten by the raid failure. I lost 4 or so months worth of development. I have the core of it in a previously forked dev tree and will be putting it back. It won't be that bad the second time around. I have it tracking the audio only but as it stands 10r1 is strill tying to lock onto a non-existant video stream for error detection/correction purposes.

    I'm one of those crazy people who has no C: drive btw

    Well, I have one intermittently actually. Relates to a system drive crash/rebuild issue.
    So how do you boot? Does it show up as a different letter or are you actually running windows with no hard drive at all. That would be... interesting.

    Oh yeah, if anyone is getting "bus error" when they execute tserver, the problem is that they ftp'd the file to the tivo using text mode. Reup in bin mode, and then chmod it.
    Still not sure if the one in the archive is good or bad. I have to run a hex diff and see. Will do it in the morning.

    --jdiner

  15. #75
    Join Date
    Jan 2002
    Posts
    4,809
    Quote Originally Posted by cheer
    That's 1.21 Gigawatts, not 3.2.

    Back on topic...THANK YOU for the resizeable GUI windows, JD! My 1400 x 900 laptop screen thanks you as well.
    Ah yes. Now that you have written it here that is what it was supposed to be... Seemed off to me and now I know as why.

    Yeah the resizable GUI has come in handy. The ability to shrink below the norm has made it useful on my HTPC when needed out on the TV. Should have done it a long time ago but it was minor compared to some of the other problems.

    --jdiner

Posting Permissions

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