Thanks for your comments, I agree, Bluray is a better choice. But for those here, who are still making DVD's for their customers, is there a way to tweak Main concept to get a better look?

In my test, I've done 2 pass VBR encoding, 6400 average bitrate, 8500 highest bitrate (video total length is 1h30)

Seems that all MPEG-2 encoders aren't made equal !!! I found one that gives very good results:Tmpgenc 6.

Take a look at the grid definition:

I haven't done variable-rate encoding in many years -- used to have quality problems and have done constant bitrate encoding ever since... I understand in theory variable can provide better results, but transitions were just so horrible I never went back. Ought I?

VBR is attempting to get the smallest file using an algorithm to figure out the best minimal bitrate for each frame of video. A problem somtimes arises when you are fading to black because the algorithm says it can use a low bitrate and will sometimes cause pixellation. I am a big fan of CBR but use VBR when I need to fit a long program on a DVD. If CBR works for you, there is no need to change. VBR will not necessarily give you better quality. It just tries to optimize bitrates.

With two-pass vbr, set the minimum bitrate at 2,000,000. You should have no problems. The default is absurdly low.

Two-pass VBR takes longer to encode, yes?

So for a DVD that's, say, 1.5 hours or less of content, even a constant bitrate will be appx. 6,000,000... you think VBR will really add much better quality at that point? I'm willing to experiment again since many of my clients still want DVD (vs. blu-ray, digital download, thumb drive, etc)

VBR puts the bits where they are needed, when they are needed. Even if the content fits at 6,000,000 cbr, high-complexity, high-motion scenes will benefit immensely from 9,500,000 bursts when they are called for.

2,000,000/6,000,000/9,500,000 is a sane default configuration for a 90 minute program.

You will get lots of discussion from hobbyists and purists here, but a commercial production shop would never deliver cbr for video that has motion. 8Mbs cbr will actually overheat many older players.

If it's a slideshow with straight cuts, however, it will render around 250Kbps no matter how you set the bitrate.

Interesting... OK.. my new computer is fast enough I'm going to be humored and test this out right now... just rendered a 1.5 hr dance show with a bunch of detail, 4 cameras and always moving people.. will see if I can visibly notice anything different.

Of course, my bigger problem will be sizing it now to fit on the actual disc.. with CBR it is very predictable.

I'm rendering now with Low at 2,000,000, avg at 5,888,888 and high at 9,500,000, variable 2-pass. Looks like render time will be way longer:
VBR 2-pass: 2:25 minutes
CBR: 1:10 minutes

I guess it's the two-pass thing, eh?

As long as your Average bitrate is also 6,000,000 there is no theoretical difference in fit. Treat it like cbr - - the math is all the same.

Do let us know if the extra time rendering two pass is "worth it" with your material.

Do let us know if the extra time rendering two pass is "worth it" with your material.

Will do ... Although right about now I see it as keeping me up a bit later tonight ;-)

Musicvid and I have always disagreed on this. I have never had a DVD player overheat from playing at a higher bitrate. When possible, I render CBR at 8,000,000. You will see that even at a CBR bitrate, a DVD is never at a constant bitrate. It will vary based on scenes. But I do agree, bump up the low bitrate on VBR. I push it higher than 2,000,000 when I use it.

@Former user OK. So then I ask you -- if CBR is not actually a constant bitrate, then how are we able to compute, quite precisely, what bitrate to set it to to achieve a certain size (in MB/Mb) based on length of content? It's really just math, and it's only possible because there is a known # of bits/sec (as I see it). It will *spin* at different speeds based on where it is on the disc, because it needs to so it can keep up with the data rate.

If I'm wrong about that, then explain how we're able to compute the size/bitrate based on length?

:-)

I am not a programmer or DVD expert so I don't know the exact mechanics of how bitrate is computed. I assume that at a CBR there is still an average bitrate achieved which gives you the math close enough to compute a file size. But if you look at a CBR disk on a bitrate graph, you will see variations in the bitrate. This is from Wiki

"When referring to codecs, constant bit rate encoding means that the rate at which a codec's output data should be consumed is constant. CBR is useful for streaming multimedia content on limited capacity channels since it is the maximum bit rate that matters, not the average, so CBR would be used to take advantage of all of the capacity. CBR would not be the optimal choice for storage as it would not allocate enough data for complex sections (resulting in degraded quality) while wasting data on simple sections.

The problem of not allocating enough data for complex sections could be solved by choosing a high bitrate to ensure that there will be enough bits for the entire encoding process, though the size of the file at the end would be proportionally larger.

Most coding schemes such as Huffman coding or run-length encoding produce variable-length codes, making perfect CBR difficult to achieve. This is partly solved by varying the quantization (quality), and fully solved by the use of padding. (However, CBR is implied in a simple scheme like reducing all 16-bit audio samples to 8 bits.)

In the case of streaming video as a CBR, the source could be under the CBR data rate target. So in order to complete the stream, it's necessary to add stuffing packets in the stream to reach the data rate wanted. These packets are totally neutral and don't affect the stream."

Here is a shot from Bitrate viewer for one minute of video encoded at 8,000,000 CBR. Notice even in that minute how much variation on actual bitrate.

well, except for the ramp-up at the beginning, it looks pretty darned constant to me!

that description from WIKI confirms that indeed CBR is a constant output/bitrate, and if the compressed source is too small for fill the bitrate the output is padded. It *does* add that the quality often varies (depending on CODEC) to more tightly compress as needed to maintain the bitrate (depending on source content)

Maybe not a good example, but there is a minimum and maximum bitrate listed as well as an average. I didn't say it varied by as much as VBR and as you read, it is sometimes padded to achieve the CBR. It was just a point I was trying to make, not argue about. I use CBR whenever possible and save VBR for longer shows that I need to assure will fit on a DVD. From the Telestream website:

"Even CBR bit rates can vary. If the encoder doesn't need the bits it's not going to use them otherwise you get "empty" bits. CBR is a tighter maximum so you don't overshoot the data rate too much for too long which can be a problem in certain server configurations. Usually the CDN will mention the requirement. It's also useful in multbitrate situations so each stream stays within it range limit. It also can help when you have limited overhead (bandwidth). "
 

OK.. so.. rendering finished -- VLC won't play them both at the same time (probably an option I have to set) so I'm watching one after the other in similar places, and both look the same to me.. of course, side-by side would be a better test so I'll drop both on my timeline and see that way (but then I'll be dependent on Vegas to decode both for me to view, which could influence results).

Also, VBR file ended up being physically smaller by a few hundred MB than the CBR

My subjective verdict on my content -- no obvious difference between VBR 2-pass render (6Mb/s) and CBR encoded at 6Mb/s. Content was a 1.4 hours dance recital with moving lights, various colors and a combination of wide, medium and tight shots.

- VBR was a slightly smaller file
- VBR took twice as long to encode (2-pass encoding) than CBR
- After both were rendered, I dropped them on two tracks of a timeline in a project setup as NTSC-Wide. Output was viewed at 100% (scaling option not enabled).

I spot-checked about 10 locations on the timeline. In each case I looked at the VBR then CBR and back again several times to see if I could see a difference in different parts of the image. Occassionally I would see a minor compression artifact difference, but I could not tell one image as being different than the other in terms of clarity/sharpness, brightness, color saturation or noise.

At least when there a target media that is around 1.5 hours, I will absolutely use CBR over VBR 2-pass. I'd be interested in seeing if VBR 1-pass would be any different.

When you have a chance, spot check some bitrates in real-time using VLC properties.

"Minor compression artifact difference" is all I would expect to see with moderate motion a la dance recital. The other things you mentioned are not a function of bitrate AFAIK.

Where this really matters is boats on water, action sports, game cap, and the like, so it's worth keeping in your bag of tricks. Capped bitrate is also the reason this same stuff looks so terrible on YouTube. Personally, I don't recall authoring in cbr in about a decade.

Anyway, we both learned something.