Hi, guys.

After spent several months making VCDs from gigantic AVI files, waiting hours to encode them, I decided to move up to MPEG-2 realtime capture. My objectives: capture TV shows (analog video), edit them and burn SVCDs of course without reencoding the files.

So I decided to buy a VideoBlaster Digital VCR. My choice was 70% based in the price. I read good reports and also bad ones about it, and it seems to me this is kind of love-or-hate-it board.

It took me a long path to make good SVCDs, and this wasn't done without lots of persistence. I'm sharing my experience with others owners of this board, hoping for some better ideas than mine, or at least, giving some hope to people that is frustrated with this card.

I started capturing with a custom record setting of 480x480, 1.8kb/s, 32kHz, but the exported mpeg file was totally unusable, with a peak (?) bitrate of 4.8kb/s The only recording setting that worked for me was "Good (EP)" (480x480, 2.2kb/s, 32kHz), which result in an almost-compliant SVCD mpeg file. The only problem was the 32kHz audio.

My next natural move was demux the mpeg file, upsample it to 44kHz, reeconde it back to mp2 and remux to finally edit it. I got a complete out-of-sync movie, impossible to be resynchronized, video became shorter than audio. I also changed the sample rate to make its time equal to the video. Sometimes they are in sync, sometimes not. Besides, all the process tooks several minutes and I felt it would be unpractical.

So I tried to first edit the original video, and instead of cut off the commercials (to get one continuous file), I tried to split an 1-hour show in small parts, and make the long process of demux-upsample-remux for each part. Bingo! Ops, at least for part #1 With part 2, 3, 4 & 5 I got something like 50, 100, 150, 200 ms audio delay. Well, at least a conclusion!

Today I'm making my SVCDs this way. I'm using Womble MPEG2VCR to edit (split) the show and the audio process I described. By chance I installed Ulead Video Studio and realized that Audio Editor can extract directly the audio from a mpeg file and save a reencoded mp2 back to 44kHz in one pass . I also realized that the progressive audio delays in the last parts of the show could be easily located in the wave window and deleted.

The job is finished by TMPGEnc and Nero, using a SVCD template and dragging all mpegs together in order (of course!), putting 0 seg delay between tracks insted of 5 seg set by Nero. This is a little boring process, but is the only one that worked to me, and I'm taking about 1 hour in the entire process. Not so bad.

My final comments: the results are fantastic! I'm making 45min-SVCDs with regular 700MB CDRs and 52min with ProDisc 800MB (and some overburning). Of course I'm waiting for a new (or even a patched) future release of the software with 44kHz support.

I have some points to discuss, and I'd appreciate some help:
.Does anybody tried M2-Edit (forgot the company's name) ?
.Does anybody have a comparision between M2-Edit and Womble?
.Why Womble not always suceeded in make exact edition in these mpeg-2 files?

Sorry for long message and English mistakes...
Thank you!