CQ vs VBR 2pass

[bruceleroy]

This might be a repeat topic, but I thought i'd ask your opinions.
I am not real happy with 2pass VBR's SVCD product for 2 CDR's.
(In my opinion, it does not compare to CCE's 4pass product, but I dont
have CCE so...)In longer movies the action seems a little choppy as the bitrate suffers with size. Can anyone comment on CQ vs. 2 pass VBR for SVCD production? I have not used CQ for SVCD's so any advice would help.
thanks

[SiCN]

Hell yeah, it is a repeat topic : Click

[banjazzer]

Can anyone comment on CQ vs. 2 pass VBR for SVCD production? I have not used CQ for SVCD's so any advice would help.

I use CQ most of the time, not particularly because it is better than 2-pass VBR(which it may or may not be!) but because it takes half the time. You just need to allow for a few test encodes and a bit of simple maths to estimate final filesize. I use a min bitrate of 350, *enable padding* ticked, max bitrate of 2520 and a CQ setting between 75 and 85. Good results every time. 8)

[biggantdogg]

I think some DVD standalone players are picky about CQ. Mine is. Panasonic SC-HT75. I tried CQ with the same bitrate as when I use 2 pass VBR, and it played back stuttery after about 5min.

2passVBR, no problem.

[adam]

biggantdogg, all vbr encoding modes are imperfect and cannot be completely controlled. It is quite common to see the bitrate peak well above the max bitrate you set in the encoder, and this can sometimes cause playback problems on standalone dvd players. I think this may be what's happening in your case because from my experience CQ respects the max bitrate much less than 2-pass vbr.

[sterno]

I think CQ is very dependent on the input quality, more so than with 2-pass or CBR. People say they get great results with it on high-quality sources (DVD rip), but I have had consistently bad results with noisy sources (VHS capture). CQ (300-2500, 80%, no padding) gave me a file with an avg video bitrate of somewhere around 2200kbps, 2100k 2-pass VBR and CBR both gave me better quality. The CBR encode was much faster than 2-pass, CQ was somewhere in between (closer to 2-pass than CBR, though).

After I ran the input through VirtualDub with some pretty heavy noise filtering the output was acceptable with CQ, though medium-speed panning scenes still looked terrible - different parts of the background appeared to be moving at different speeds. I found that dropping from 80% to 75% gave me a smaller file with no noticable change in quality. Sometime I'm going to try with some DVD samples and see how it looks.

I tried CQ with the same bitrate as when I use 2 pass VBR, and it played back stuttery after about 5min.

Did you try both with and without padding? It seems that some DVD players require it, some don't care, and I know that on my player (GE1101P) enabling padding gives me jerky playback and sync problems with both 2-pass VBR and CQ.

[biggantdogg]

Adam,
What you say makes sense, but if that were the case, the problem would be with 2vbr and not CQ. (if it were a bitrate issue) (especially a over max bitrate issue)

Sterno,
Good points, I'll do some experimenting

[TheSeven7hSon]

I like CQ. For me it's faster with good quality. My only gripe is that when I add black video to the end, Nero cuts it off to the point where there's motion. Can't figure out why. What I end up doing is do the last 10-15 seconds or so as CBR. Silly, I know.

Seven

[banjazzer]

Adam,
What you say makes sense, but if that were the case, the problem would be with 2vbr and not CQ. (if it were a bitrate issue) (especially a over max bitrate issue)

I think Adam is right about this. I remember reading elewhere that TMPG's CQ setting can exceed the maximum bitrate chosen (which seems a bit odd). Quite likely this is more likely with CQ than 2-pass VBR because the former is not working to any filesize constraint. I seem to recall the figure of 1850 being mentioned as a suitable max. setting, which still yielded actual max figures of around 2500! I never experimented further because my player seems to be able to handle a max setting of 2520.

[adam]

Adam,
What you say makes sense, but if that were the case, the problem would be with 2vbr and not CQ. (if it were a bitrate issue) (especially a over max bitrate issue)

No that's exactly the opposite of what would happen, assuming that is the problem. Like I said, CQ doesnt appear to respect max bitrates as well as 2-pass vbr so thats why you will often get much higher bitrate peaks despite what you set your max bitrate to.

2-pass vbr is more controlled than CQ, not the other way around.