What's one pass and two pass?

[jammer20002]

I've been trying out one Sorenson Squeeze.
It has 3 compressions you can do.
Two of them are one pass and two pass.
What is 'pass'?
Is two pass better than one?
(I can't remember that other one.)

Thanks.


Jam

[Dr_Layne]

It's the number of times the video is examined and compressed. One pass means the video is compressed at the same time the computer is examing the video. Two pass means it's analyzed then compressed on the second pass. This way the computer can examine the video, decide what how much compression to use on a given group of frames and compress accordingly to minimize artifacts.

Steve

[edDV]

It's the number of times the video is examined and compressed. One pass means the video is compressed at the same time the computer is examing the video. Two pass means it's analyzed then compressed on the second pass. This way the computer can examine the video, decide what how much compression to use on a given group of frames and compress accordingly to minimize artifacts.

Steve

Further to that, the benefit is usually tighter compression, not significantly better quality.

The bad news, it takes approximately twice as long.

[junkmalle]

Single pass constant bitrate uses, as its name implies, the same bitrate throughout the video. A static black sequence will take as many bits as a high detail, high motion sequence even though it doesn't need it. You know how big the file will be when the process is done because the file size is the bitrate times the running time.

Single pass variable bitrate (aka constant quality or quantization mode with some codecs) varies the bitrate to get the quality level you request for each frame. Dark static scenes require fewer bits, bright high action scenes require more. The final file will come out to whatever size is required to get that level of quality.

Two pass variable bitrate is useful when you want a file of a specific size, say 700 MB to fit on a CD, or 4 GB to fit on a DVD. During the first pass the codec determines how much relative bitrate each frame will take. Then in the second pass it apportions the bits to create the file size you want.