What do you do to go from Vegas to TMPGenc?

render to AVI uncompressed or DV and then compress the resulting file using TMPGEnc

That's it exactly.

I'm pretty sure that 2-pass VBR makes no difference in quality at all in TMPGEnc. The purpose of VBR is to get more data into the same space by lowering the data rate during periods of very little motion. The purpose of the two-pass approach is to make a first pass in order to better predict how much the data rate can be lowered and still meet the average bitrate objective. If quality is your only concern, the best quality will be achieved by using a constant bitrate setting (CBR) at the maximum allowed by the SVCD "standard," namely 2520 kbs.

If you doubt this, pick a section of video (about 15 seconds) and encode it using 2-pass VBR. Then, encode the same video using CBR at 2520. Unless you set the VBR for a rate above 2520, the CBR should look better, and will of course take far less time.

Here are settings I use in order to get the best possible quality with SVCD in TMPGEnc:

Video tab
Rate Control Mode: CBR
Bitrate: 2520
VBV: Automatic
Profile: Main Profile & Main Level
DC Component Precision: 10 bits
Motion search precision: High Quality (slow) [I can't see any difference when using "Highest Quality"]

Advanced tab
No change from default settings (field order: bottom field first). I don't use any of the filters in the bottom half of the dialog box, although I've used them all at one time or another when converting from DVDs).

GOP structure
I Pictures: 1
P Pictures: 3
B Pictures: 2
Output Interval: 1
Max number: 0
Check "Detect scene change", but do not check "Output bitstream" or "Forced picture type setting."

Quantize Matrix
Quantize Matrix (the setting is below the matrix): MPEG standard
Output YUV Data: check
Use floating point DCT: check
No motion search for still picture: NO check
Soften block noise: check only if block noise is a real problem. Start with defaults of 35 for both Intra and Non-intra and if picture is too soft, try 15, 20, or 25 instead.

Option Menu
Environmental setting: Under the "General" tab, check "Interpolate YUV data from 4:1:1 to 4:4:4 if you are converting from AVI files saved using the DV format.

Hope this helps! Remember, SVCD is going to look a lot worse than DV no matter what you do because DV uses almost ten times the storage for the same amount of video. Since DV is already a compressed format, reducing the same video down to only 10% of the disk space means that something has to give.

Great!! Thanks for all the settings. I will make a template from them.

Johnmeyer - I could set most of the settings you recommended but some of them such as VBV cannot be set because they are grayed out on my copy of TMPGenc. Are they activated if you buy the "Pro" version?

If you don't get any error messages, you don't need to worry about VBV.

If you want to change all the settings, including those that are grayed out, you can load the Unlock template. This will retain all the settings you had prior to loading this template, but will -- as its name implies -- unlock all the locked settings. I believe this template is shipped with the regular version of TMPGEnc and is normally stored in the Templates\Extra subdirectory.

Thanks. I will try the unlock template when I get home. I ran a test encode last night and got no error messages. The product looked really good when completed. I burned an SVCD with Nero and it looks great! Thanks for your complete list of parameter settings. They really helped me out a lot.

I rendered a 1 hour 35 minute video to avi creating about a 20 Gig file. When going to TMPGenc, it only sees roughly the first 8700 frames (the first Gig?) What do I need to do to get TMPGenc to see the whole video? With Cinestream, I just export the file "with dependencies" which creates a very small file so TMPGenc can see the whole thing no matter how big the source files are. Can I do something similar with VV3.0a?