Two-pass encode MPEG2

[lemister]

What is the diference when creating an *.mpeg file (mpeg2) using variable bit rate and enabling "two-pass encode" box?. Besides the time subject. I haven't tryed yet but makes me think using variable bit rate in single pass mode gives same results.

For example, when encoding video with divx you can use multipass encoding; and the result is a variable bit rate video. A lot better than single pass; wich is constant bit rate.

[DLA]

With VS, the single-pass VBR will evaluate the bitrate needed on the fly. This gives a VBR file that is pretty much constant (almost).

With two-pass encode, the first pass will not produce a .mpg file but multiple flat text files where quality information for each frame is written. The second pass will use these files to better allocate the bitrate to the scenes that needs it the most.

End result is a better quality.

[Ken Berry]

A lot better than single pass; wich is constant bit rate.

Just to correct an apparent misconception on lemister's part -- just because it is single pass doesn't mean that it is automatically constant bitrate. You can also do a single pass with VBR. But as DLA has explained, you will usually get better quality with two pass encodes. However, it has to be noted that any difference in quality will often be fairly microscopic and may not be particularly noticeable to the naked eye. But if you want to know you are getting the best quality possible and are not worried about the extra time taken by the extra pass, then go for two pass.

Note also that if you choose constant bitrate, you can't use two pass encode because there is no need to. The bitrate is fixed so no calculations need to be made on how to improve quality using that chosen bitrate.

[erdna]

If VBR 18Mbs is selected as max bitrate or if, "do not re-encode compliant files, does two pass encoding still make sense?

[Ken Berry]

If, as I suspect you are, you are talking about AVCHD video, then I am not sure that two pass encoding will work with it. From my own experience and posts from others, it seems to cause VS to crash if selected. In any case, the quality is already very high and you are unlikely to be able to detect any significant improvement, at least with the naked eye.

If 'do no convert compliant mpeg files' is selected, and the files in the burning timeline are genuinely compliant, then they will not be re-encoded. And that means that two pass encode does not even enter into the picture. However, in the preceding steps in the Editor module, if you are rendering your project to a new Video File (Share > Create Video File and not Share > Create Disc), you will of course be able to choose whether you want to do a two pass encode or not at that stage. However, the proviso above about AVCHD applies here too.

[sjj1805]

I will try and explain the differences again.

Think of these three alternatives.

1. Constant Bit Rate.
The entire video is encoding at the same bit rate from start to finish.
This is perhaps the fastest way to create a newly rendered video but it also produces the largest file size because even areas that do not require a high bit rate will still have a high bit rate. More bits means bigger files which also means less space available on the DVD Disc.

2. Variable Bit rate - Single Pass.
The video uses more bit rates where there there is more action - such as a racing car driving past. It uses less bits where there is little action such as a building, or some cows grazing in a field. This produces a smaller file so you can get more onto that valuable DVD disc space.

3. Variable bit rate - Two Pass.
This time the encoder sort of does the same thing as variable bit rate - single pass. The difference is instead of encoding it straight away (which in effect is guess work) it reads through the file first making notes. It then goes through it all again a second time but having the benefit of the notes taken on its first pass it can get those required bit rates even better so producing not simply a smaller file but also one with better quality. Obviously this takes longer because of the extra work involved.

Hollywood film companies use Variable bit rates - multi-pass where their encoder will read through the video several times - perhaps even a dozen times or more.

[lemister]

It's now clear that using VBR along with 2-pass encode brings out better picture quality. I will take in mind next time I export to any DVD-video project.

[Trevor Andrew]

Hi

Variable bit rate was developed to allow more video, more footage to be burned to disc/DVD whilst maintaining quality.

A standard DVD can have 60 minutes (approx) of video at full bit rate.
Using Constant the video will fit to disc, encode faster, using full bit rate giving full quality.

If your video is greater than 60 minutes, then Variable bit rate may be considered.
At 90 minutes I still use Constant at 6000

We cannot directly compare Constant 6000 with Variable 6000
Constant is the Average, Variable is the Maximum.

Constant 8000 is the same as Variable 8000 (quality wise)

As a guess Constant 6000 may be equivalent to Variable 7000.

If you need more disc space use Variable.

If your DVD contains less than 60 minutes use Constant