VideoHelp Forum




+ Reply to Thread
Results 1 to 8 of 8
  1. Is there a significant gain for using 2-pass mpeg encoding when the source is home video (DV cam quality) ?

    I can't see an obvious difference, but where would the difference be obvious between 1-pass & 2-pass encoding?

    (using TMPGENC for example...)
    Quote Quote  
  2. Always Watching guns1inger's Avatar
    Join Date
    Apr 2004
    Location
    Miskatonic U
    Search Comp PM
    If the footage is mostly taken with a static camera and of people sitting around not moving much, then you may not get much out of 2-pass encoding, other than slightly more efficient packing of the data.

    If, on the other hand, you have fast pans or zooms (shudder), or a lot of on screen action, then you will probably see much better results from a 2-pass method.

    In the end though, it is down to your preference and your eye. If you cant see the difference and are happy with 1-pass, go with it. Or, so a sample edit of 30 seconds of mixed footage from you tape, encode one copy with 1-pass, and second copy with 2-pass, and a third copy with a high bitrate CBR, put all three on a disk, and get different people to pick which one they think looks best.

    I know I have a much more critical eye than most of my friends - just ask the one who hates me for pointing out how crappy the picture on his $4000 plasma TV is - so I use a 2-pass VBR on 90% of my footage.
    Quote Quote  
  3. Cheers for that.

    With lower range DV cams (mine 1M pixel 3CCDs) that the video would be noisier than broadcast / less definition & that extra bits might have been allocated for that rather than wanted artifacts.

    I guess it all does boil down to if one can actuall tell the difference - (the reason why I drink wine out of a box)

    Rog.
    Quote Quote  
  4. Always Watching guns1inger's Avatar
    Join Date
    Apr 2004
    Location
    Miskatonic U
    Search Comp PM
    I have a single CCD DV camera, but unless I am shooting in low light, I get very little noise most of the time. Depending on the feel I want, I may even add grain in post. Even with this, I still prefer 2-pass VBR in most instances.

    I should also add that I edit in Vegas and have the mainconcept Mpeg encoder, which is much faster than TMPGEnc. While adding a second pass effectively doubles the encode time, a 2-pass encode is probably only 15 - 20 % longer than a single pass in TMPGEnc.
    Quote Quote  
  5. Member
    Join Date
    May 2004
    Location
    New Jersey USA
    Search Comp PM
    Cinemacraft Encoder actually improves DV footage in my opinion. I use 3 or more passes in VBR. It cant hurt to have more passes unless your in a hurry. I usually start the encoding on a 6o min DV AVI before I leave for work or before I go to sleep.
    "There's a sucker born every minute"
    Quote Quote  
  6. Originally Posted by Muppet Meat
    Cinemacraft Encoder actually improves DV footage in my opinion. I use 3 or more passes in VBR. It cant hurt to have more passes unless your in a hurry. I usually start the encoding on a 6o min DV AVI before I leave for work or before I go to sleep.
    What does more than 2 passes acheive?

    Doesn't the first pass calculate the bits required per frame and then the second do the encoding, scaled by the metric calculated in the first?

    What do the extra passes actually do?
    Quote Quote  
  7. Always Watching guns1inger's Avatar
    Join Date
    Apr 2004
    Location
    Miskatonic U
    Search Comp PM
    Mainconcept only allows up to 2 passes under Vegas. I use CCE with Rebuilder for backing up, and haven't seen any great difference between say 2 and 4 passes, except for the longer encode times.

    I suspect that after 2 passes the law of dimishing returns takes effect, and the differences you see are exponentially smaller on each subsequent pass. 3 passes might make a difference on a particularly difficult section (eg. a fade to black or very high action scene), but after that there isn't much more it can do, so 4 passes will give you very little, and 5 or more is just wasting CPU cycles for ego's sake.

    I am happy to be proven wrong if someone has some proof that 6 is demonstrably better than 2
    Quote Quote  
  8. Member
    Join Date
    May 2004
    Location
    New Jersey USA
    Search Comp PM
    Good question, Im not sure how much of a difference it makes but I have came across warez SVCDs with the ripper's methods stated in text files which sometimes use 5 pass CCE SP encodes.
    Would a $2000 encoder have the ablity to do more then 2 passes for no reason at all? Im not sure but I guess its there for a reason.
    I have read in cce tutorials that after 3 passes there is less and less improvement. But it doesnt hurt to have more passes if your are sleeping for 6 hours or coming home from work after 10 hours.
    The first pass is to create that video file so you can check it for greyed areas in need of higher bitrates. I rarly see areas in trouble if I use a minimum bitrate of 2000 instead of 1. If you want ultra efficiency use 1 minimum and tweak it all before the the next pass or passes. Ultra efficiency is another reason why cce sp is so expensive.
    "There's a sucker born every minute"
    Quote Quote  



Similar Threads

Visit our sponsor! Try DVDFab and backup Blu-rays!