2 pass encoding will create the best looking video at any specific data rate, as the encoder will best utilize the data bits available where they are needed.

i.e. For scenes where nothing much changes or with low detail, very few bits will be used. For scenes with rapid changes or very fine detail, the encoder will maximize the data bits available.

From personal experience, encoding at a constant fixed data rate may require as much as double the data rate to equal the performance of a VBR multi-pass encode.

Unless I have no time for the encode, I always use a 2-pass encoder to ensure the best possible quality for the data rate I am using.

Some encoders and DVD players handle VBR better than others. My experience has been that I always get the best results by using the max fixed bitrate for a video. A VBR will not do any better than the max bitrate. It only allows the bits to be allocated differently, mainly to fit more material in a smaller footprint.

Dave T2

Thank you for your replies!
I tried it for the first time. With my m2t files.
It resulted with heavy pixelation all over the video?
Where did I go wrong possibly?

Video encoding is an art. Two-pass encoding is used when you're trying to put 10 pounds of stuff into a 5 pound bag. Two-pass encoding is useful when you are using variable bit rate encoding (VBR) in order to fit more than about 72 minutes of video on a single-sided DVD and need to use a bitrate lower than a CBR of about 8Mbps.

Let's say you were trying to fit 2 hours on a DVD (using 192kbps .AC3 for the audio), that would require a video bitrate of about 4,896,000 bps in order to make it fit on the disc. You could encode the whole video at that bitrate, but the more effective method is to use VBR encoding where the bits available are used more flexibly to encode the video data more accurately, with fewer bits used in less demanding passages and more bits used in difficult-to-encode passages. The average bitrate over the length of the file would stay at 4,896,000 for example, but could go as high as whatever maximum you have set (usually 8,000,000 bps) for those passages which need it. Of course, other less complex passages would get fewer than 4,896,000 bps in order to maintain the average.

By doing a 2-pass encode, the encoder analyzes the file on the first pass looking for scenes which need the extra bits and those that don't, on the second pass it does the actual encode using the bitrate allocation strategy it calculated on the first pass. 2-pass encoding also ensures that the encoder more accurately hits the average bitrate target for which it was set. The only downside to 2-pass encoding is that it takes about twice as long to encode a file. I can easily see the difference.

This discussion was about DVDs, but it applies when you are trying to maximize the quality of an encode while making it fit within a fixed file size.

"It resulted with heavy pixelation all over the video?"

You didn't say which encoder, but if the MainConcept MPEG-2 encoder, the default minimum bitrate is way too low.

Using the example of a render for DVD, 9.5Mbs Max, 6Mbs Average, 2Mbs Minimum is a reasonable starting point.

You're using 2-pass with M2T files? I assume that this is HD stuff, what exactly are you trying to accomplish and what min/max/avg bitrates are you using?

Thank you all for your replies! I'm really appreciate them.
I was trying to make a video for vimeo.

http://vimeo.com/14241331

I used "constant bitrate" while rendering it.

(It's 2.00 a.m. now here, I'll make some tests using variable bitrate and I'll give more details about the results. Hope to learn the essence of it!
I have plenty of time and I prefere to render everything as "two-pass" naturally, of course. Thanks again!)

A fade, coming out of black, will always look bad if the minimum bitrate is too low.

You have some beautiful footage in there.

Aside from vision considerations, what on earth happened to your Audio? ... compare with this audio

"A fade, coming out of black, will always look bad if the minimum bitrate is too low. "

Yes, exactly. The default minimum bitrate in MainConcept should have been set at 1,500,000, not 192,000 !

Yeah, no kidding. 192K. WTF?

To really test the VBR vs. CBR, shoot a dark grey wall and encode using both methods. You will probably find less pixelation on the CBR than VBR, because the encoder sees minimal change in the pixels and will use a low bitrate. But the low bitrate makes the pixels vacillate in brightness and size.

I always use CBR unless I am trying to squeeze more than 1.5 hours on a disk. I don't like to go below 6.5mbps. If I am using AC3 audio, I will use 8mbps and if I am using PCM audio, I will use 7.5. I have had very good luck and no rejections with these numbers.

Dave T2