@dvd2svcd: CCE 1passVBR question/suggestion

[Bach]

Hi,

firstly, I would like to say that it is awesome the way that you had handled the CQ prediction while using tmpgenc.

I would like to know if it could be possible for you to use the same procedure to predict the Qfactor value while using CCE 1passVBR.

I know that nowaday you are using the calculated bitrate as the maximum bitrate while doing 1passVBR encoding. However there are 2 problems with this aproach:
a) the size prediction doesn't work very well if the Q_factor is bigger than 5;
b) the video quality is not so good if the calculated bitrate value is smaller than 2100, so that one should use at least 1 CD for each 50 minutes of movie time length.

In the other hand, using, in example, min=0, max=2400, Q_factor=50, I can put a whole 2 hours movie with 2 audio tracks (160 kbps each) into 2 CD, with a quality as good as (or even better than) that one achieved with tmpeg using CQ=90, and done much more faster.

btw: (this is for anyone who is reading this post): min=0 means that you will let CCE to choose the minimum bitrate which it will use.

thank you.

Cheers.

[MITSUI_1]

I agree that there can be some tweaks to one-pass VBR enclding with CCE.

A lot of us have found that one-pass VBR can lead to higher quality files at lower bitrates.

But three things would make the process OH so much nicer:

1. The calculated average, etc., simply should not be used for one pass vbr. Give the user the option of deciding exactly what he or she wants for min, max and Q.

2. Either so some sort of size prediction analysis like with TMPGenc, or give the user the option to have CCE take several "trys" to reach the appropriate file size. On the later option, you could have DVD2SVCD encode using the Q factor and bitrate settings selcted by the user and: (a) lower the Quantization number (i.e., increase quality) and reencode if the resultant MPEG2 file is far smaller than the size that would fill 2 cds; or (b) increase the quantization factor (perhaps up to a certain ceiling) if the MPEG2 file exceeds the target size, and again, keep reencoding until it gets it right. Perhaps cap the number of tries that DVD2SVCD will perform before it settles on an encoded file to split and mux.

[IGuaNAdoN]

And I'm still holding out for someone to figure out how to use the .vaf info and automatically set the bitrate and q factor. :) :) this has been said before... but I'm a dummy and don't have any idea how to do such things. It’s just a wish of mine and I'm sure others. :)

I figure if I bring this up every few months it might happen... or not! :D


Peace,

IgGy :D

[dvd2svcd]

Bach: Good idea, will be implemented some time.