Using CBR or VBR on SVCD

[summit15]

I'm somewhat new and can't seem to get a clear answer to this question:

If file size is not an issue, will I get better, the same, or worse results by using the 2-pass VBR over CBR?

I am capturing a DV stream via firewire. I am then using TMPG to encode the AVI as an MPEG-2. When encoding, if I use CBR at 2520 bitrate and motion search to highest quality, will the quality be any better or worse than if I used the 2-pass VBR with the average bitrate set at the same 2520?

My general understanding is that the higher the "average" bitrate or the bitrate used in CBR, then the better the quality of picture. By lowering the bitrate in either case, then I will reduce the file size but the quality a little as well. Is this understanding correct?

This logic leads me to believe that CBR at 2520 and VBR with an average of 2520 will give me the same result. Is this correct?

Thanks for the help.

[kinneera]

Since 2520 is the maximum bitrate for SVCD in most players, CBR and 2-pass VBR at that average (the maximum will be equal to the average) will be the same. Single pass VBR may be your best bet in this circumstance, as you will then be able to fit more video at essentially the same quality onto a disc.

For lower bitrates, 2-pass VBR will be superior, but takes longer.

[summit15]

So, if file size is not an issue, and I encode at CBR at 2520, this is as good of quality as I am going to get?

I'm tired of playing around with all the settings... I just want to find one way to encode and know that I am not sacrificing any quality that I don't have to!

[kinneera]

Yes, but as I said, a single-pass VBR with the maximum set to 2520 will generally allow you to fit more video into the same space at the same quality. (This includes the CQ modes in TMPGEnc).

[Boooooommm]

What VBR does is allocate the kbit ammount depending on quality without passing the average kbit specified. This means that if a frame has little detail, the encoder will use less kbits on that frame, but if a frame has a lot of detail the encoder will use enough kbits to make it looks nice. This, will result in a good quality and well sized movie file equal to, or smaller to the same file made with the same average kbit ammount in CBR. CBR uses the same kbit ammount on every frame, which means that no matter what the level of detail on the frame, it will use the same kbit ammount. The more kbit, the more detail is used on the frame.

[kinneera]

That is correct for 2-pass VBR, and only when the average is not equal to the usable maximum. If the target is not SVCD, then 2-pass with an average of 2520 will be superior to CBR.