confusing audio syncing

[lucindrea]

ok i have a mpeg i got off my replay ( it's like a tivo for those who dont know ) .. i ran the mpeg through womble to remove commercials and tmpgenc in order to get it from 5G to 3.5 G .. used the dvd-ntsf template and double checked that my auido was encoded 48k16bit ... when i rin the finished mpeg through media player it skips now and agin but thats expected with media player .. when i run it through power-dvd i works great .. full ff and rweind , i can click anywhere on the movie and it instanly will skip to that part with no sound sync problems ...
now in order to get it into scenartist i need to demix the mpeg and convert the mp2 into a wav file ( i use winamp - 48k16bit ) because secnartist will not take an mp2 .....
i'm able to make the entire movie and even burn it onto a dvd-rw ... problem is the sound gets out of sync ... but it's a werid out of sync .. if i change chaptors ( i put chap points in scenartist ) .. the movie changes fine .. but the sound track is from the preivoius chap .. the more i change chaps the more out of sync it gets ( change chapt 3 times and the audio is 3 chapts behind ) ... .. opening the ifo created by scenartist in ifoedit just crashes ifoedit dvd viewer when i change chaptors.

the movie was recorded in pro-logic and it retains the chanels when converted to ac3 ( useing scenartist ) and when the thing runs first few chapts the audio is right in sync ( i got good ears and i cant see any delay when it's running fine ) ...

does anyone know of a way i can use either the raw mpeg , or the mpv/mp2 to create some sort of dvd image? i belive it's the conversion onto wav->scenartist/ac3->vob thats creating the problems

[TRILIGHT]

I don't know exactly how your streams are, lucindrea. I can tell you what I have to do with my TiVo streams though. It's a total pain in the ass since it's a manual process but it's the most reliable way I've found to do things so far.

When I extract my streams, I end up with a mpv and mpa file. I edit the mpv in m2edit to remove the parts I do not want. I then export this as multipart so I have a number of different video streams. Usually, removing commercials will give me about 4-6 parts depending on the commercial segments run.

Next, I convert the mpa file to WAV and load it up in Premiere. I cut the commercial segments out of the WAV file and then export the clips that I want from the timeline.

Next, I create a DVD2AVI project file and an AVS script so I can load the video up in Vdub. Once loaded up, I load up the WAV segment for that particular clip also and then play it. I've gotten so close before that no alteration was needed. However, usually you have to "zero in" on what sort of delay you will need. Change the "Interleaving" setting for the audio to set a positive or negative delay and then watch the clip again. Don't be surprised if you have to set it as far out as +/- 100-200ms! Most of my edits require setting a "delay" of anywhere from 50-150ms. It all depends on how close the audio edit got. I've had one be as far off as 300ms. Anyway, do the same for each segment and make a note of what delay setting got your audio in sync.

Once you have your delay settings, open the audio clip in Sound Forge (or something comparable) to make the required changes to your audio. The time settings (at least in Sound Forge) are listed as hr:min:sec.fractions. So, let's say for example, you determined that clip2 required a delay of -120ms. This means (since there is a negative delay) that your audio had to be started 120ms earlier in order to be in sync. Use your selection options to make a selection that is 00:00:00.120 from the beginning of the clip. Once selected, you will DELETE this selection. This makes it such that your audio is now starting 120ms earlier than it did before. By the same token, if for instance you determined you needed a delay of +80ms, you would use the insert silence option to insert silence at the beginning of the clip. Insert silence in the length of 00:00:00.080 and it will insert 80ms of silence. This effectively makes your audio start 80ms later than it did before, thus putting things in sync again.

In Maestro, simply drop your video clips in order onto the timeline. Then right-click the audio track and say that you want to make a sync audio track. Assign each of the audio clips you just made to their respective video clip. Maestro will make one large audio file for you to use. It does so, however in .aiff format. If you want to use this PCM audio, that's fine. However, if you are wanting to encode to AC3, you will need to convert this file to WAV (I use Sound Forge for this). Once it is WAV, you can convert to AC3 audio and use this file instead of the PCM so it takes up less space.

Whew! There! Pain in the ass, huh? ;) I'm all for just creating chapter points to skip over commercials with commands and I've done it before. However, there are times in which I simply have to have the extra space on the disk and I have to edit out the commercials to make things fit like I want. I know it's not a perfect solution but it is one that I've found to be quite reliable when it comes to the end results.