What can I do better? (VCD)

[biggantdogg]

Executioner,
With my P4 1.7ghz, the first simple demux took about 15min for video and 3 min for audio on an 800mb file. (VCD 352X240)

[Executioner]

biggantdogg, thanks for the info.

[wulf109]

Executioner

On a 2GB file it takes about 12 minutes to do the demux and remux. You can do a straight remux,eliminate the demux,but I find the resulting mpg sometimes won't open.

[Truman]

Truman

You should read,I said there is no effect on file in CBR or VBR.

I did read:
>...Resolution does not effect file size in VBR or CBR. In CQ it will.

CQ is 1passVBR. A higher resolution do not effect a noticable larger filesize.

[adam]

If your using ntsc and your goal is the highest quality encode then you definitely should not skip the frameserving process. If your dvd is stored as ntscfilm, which almost all are, then you can use forced film in dvd2avi to preserve this framerate. This literally increases your quality by %20 since you now have %20 less pixels to encode. If your source is a progressive ntsc dvd then it is foolish to skip frameserving in dvd2avi.

The ONLY reason not to frameserve is to save time but with ntsc dvd's as a source, this comes with a huge decrease in quality.

[Executioner]

Thanks Adam, good points on frame serving. I tried it last night (not using DVD2AVI), but did not have very good success. I was able to rip the DvD as one large VOB file, but after a simple demux, the sound was way out of sync. Maybe I did not perform the steps correctly, but the mpeg created was not the entire movie either, and along with the sound not being in sync, I'll probably try it again this weekend to see where I went wrong. If it does not work again, then I'll go back to my original method of using DVD2AVI for frame serving.

[neomaine]

Well, tried the 'big file' approach to TMPGenc last night. Waste of time...and space. It took 30+ minutes to mux/demux and I've doubled the amount of space needed for the rip. Actually, its three times since you have the original, the demux AND the demux still exists while it remuxes. So, for a 4gb rip expect to need at least 12gb. And, it didn't even speed up the re-encode. Then I used DVD2AVI to save a project and only required 1gb extra (for the wav, my choice) and only took 15 mins to run.

So, the technique takes twice as long, uses at least twice as much space and provides 0% additional quality or functionality. Oh well, it was a fun excercise. I guess the only reason you would use this method is if you can't get DVD2AVI working. Well, I don't do DVD-/+R(W), so it may be usefull to do this without needing to re-encode for your final output.

[wulf109]

Truman

You saying that I am incorrect in stating the resolution does effect file size in CQ mode. You say that resolution does not effect file in CQ mode. Please explain the following.

45 min. at CQ70 2600bit,all encoding parameters anf template are the same only the resolution was changed.


720 x 480 = 621MB

352 x 240 = 245MB

The 352 encode resulted in a file less than half that of the 720 encode. I think that qualifies as a difference.

[adam]

Resolution itself has no effect on filesize regardless of what mode you are encoding in. Filesize is determined by your bitrate which is bits per second multiplied by the number of seconds in your movie.

However, in automatic vbr encoding modes like CQ the amount of bitrate used is determined by the encoder according to the complexity of your movie. With a higher resolution you have more total pixels per second so your movie needs more bitrate to achieve the same quantization level, so the encoder allocates more bitrate which in turn increases the filesize.

The resolution is not affecting the filesize, the bitrate allocated is. But even still, you really can't say that resolution is a factor in filesize even in CQ mode, because you could say that about literally every other parameter of your encode ex: framerate, noise reduction, gamma correction etc... Every little thing you can tweak about your source can affect how much bitrate TMPGenc allocates in CQ mode but the point is that your filesize is still being determined solely by your bitrate.

Try doing two identical encodes in CQ except use heavy noise reduction on one and not the other...you'll see a size difference there too.

[wulf109]

Biggantdogg

Referring to your original post about using 2-passVBR to create a VCD. I had never tried that in VCD. Your post prompted me to try it. I am very pleased with the results,thanks for turning me onto that option.

I would be happy to send a disc-to-disc copy of a file made by the method outlined in this thread to anyone interested. You can decide if works. You can email me at:

wulf109@constant.com