I agree with u on avoiding the dv25 compression and encoding straight from the timeline. However it's impossible since the projec t is made up with some jpeg files, and some avi files that have to be nothing but avi since it's mandatory for making the effects with Boris Red Open GL. Despite the fact u may disagree with me about rendering with Vegas, I can assure u, the TMPGENC results are much better. Now, could u please explain how to mux in dummy audio stream to make the 2 files generated by TMPGENC readable by VEgas or DVDA (they can be read by ULEAD DVD WORKSHOP2.
THANKS IN ADVANCE

Start TMPEnc, cancel project.
File>MPEG Tools>Simple Muliplex
Select the file you created previously as the video file and no audio file, Type as MPEG-2 Program (VBR).
That should do it!

Yes you cannot avoid the DV25 encoding on the AVIs however with your jpegs you can. Vegas will send frames from them to the encoder uncompressed so a significant improvement in quality is possible. The encoder will see 4:2:2 rather than 4:1:1 sampling.

Regarding your original problem, although TMPGEnc may do a better job of encoding than the MC encoder that comes with Vegas I was more concerned about your problems with levels being shifted.
Do you use the scopes in Vegas to monitor what is happening?
You may find as I did that the Histograms are the most useful for checking levels, on the kuminance channel you should have nothing below 16 and nothing above 235, one way to achieve this is with the Broadcast Colors filter in Vegas, I usually go for 'Conservative' preset.
This doesn't seem to be that necessary when going out to DVD HOWEVER if you;re having differences in levels between two DV25 codecs it'd be worth knowing just what you're starting with, the Vegas codec is way superior to the uSoft one so I'd put some effort in to trying to stay with it.

Simple answer: Click on either "System (Video only)" or "System (Video+Audio)" on the main TMPGEnc screen. This will create system MPEG files that can be read by Sony products.

Ok, done! But there's still a problem: this way the audio quality is very poor. Don't know why, have set all audio options as the best as possible with TMPEGENC. Help please!

Thanks farss (as always) for taking time to respond. I am just a newbie and all the histograms stuff seems not to be my beach (not yet... - sorry for the slang, not to be my field of knowledge).
The thing is: Mainconcept that comes with Vegas does an excellent job, but TMPGENC may even do a better job. I'm trying to keep in touch with Chienworks (Kelly) in order to upload on his homepage some files that will show better what I'm talking about.

DEar Farss
This way:
Start TMPEnc, cancel project.
File>MPEG Tools>Simple Muliplex
Select the file you created previously as the video file and no audio file, Type as MPEG-2 Program (VBR).
That should do it!
Doesn't work at all. It says "illegal MPEG video stream"
?????

Hmm,
I've only done it with a .m2v stream but you have a .mvp file, maybe there's something in that.
But doing it the way John suggested will give you the same result. I cannot understand why you say you're getting poor audio. The mpeg encoder doesn't touch the audio, it simply muxes the data into the stream. I assume you rendered the audio out from Vegas a separate .wav file. If so how did that sound?

Thanks for answering farss. I owe u some explanations. First, I'd like to appologize to have typed the following extension incorrectly, in fact, it's not .mpv but .m2v.
Now I have fixed everything (and learned a lot with u and the ones at www.dvdrhelp.com Let's see what I've done:
- I installed Debugmode plug-in to open a fake avi with TMPGENC. Works great and saves time ( as u've said).

for the audio problem...
- I installed tooLame, an external encoder since TMPGENC audio encoder is not one of the best available. I set the following options with TMPGENC also: Option>Guide Audio Engine>MPEG1 Audio Layer II encoder (external program, browse for tooLame); Same Audio Engine Guide>Do not use Audio layer III; Same Audio Engine Guide>Sample Frequency Covnertor> HIGH QUALITY (THAT'S WHAT I FORGOT TO SET).

Could also be done the following ways:
1.rendering the whole thing as elementary string m2v and later setting the file to be demuxed; rendering the ac-3 file in Vegas and burn the final result with DVDA;

2.rendering the whole thing as MPEG2 with audio. Openning it with DVDA and setting AC-3 option there;

About your question on using scopes...no I don't. But if u want me to do so in order to let u know what's happening around here, please let me know.

You may say that I am not an experienced user but I can reassure you that TMPGENC does a better job than Mainconcept. Looking at the final result one can tell the difference (the file encoded with TMPGENC is clearer toward the edges, clearer in general, I'd say. Someone would also say...it's a matter of eyemetrics.....lol but believe me, the difference is there.

I'd certainly not knock anyone for using TMPGEnc, I've used it a LOT myself. In some situations (mpeg-1 for VCD from poor VHS) it performs better than MC does.
On good quality video for mpeg-1 or 2 the difference isn't so noticable and as time is money I tend do go with what'll get the job done with acceptable rather than perfect results. TMPGEnc has a number of other advantages, being able to batch encode is great, it'll even shut the PC down when it's finished.
As for the scopes thing, they are one of the very powerful features built into Vegas. I gather from your posts you lean more to the creative than the technical side (my background is in engineering), in that case as I said before you'll probably fine the vectorscope and RGB parade or waveform a pretty daunting thing to look at. I always have the histogram display open in the bottom RH side of the screen. I don't yet have a decent studio monitor so I tend to use the scopes perhaps more than I should BUT once you understand what they're telling you they're a good way to check just what's going on.