CCE q.factor. How much of difference matters?

[markrb]

I tried searching this like crazy, but didn't come up with anything. If you find a thread that helps please link it.

I am doing an encode and if I keep the original menu's I can get an average q.factor in CCE of 5.03, but if dump the original menu's and create my own I have enough room to drop the average to 4.48. My question is how much real difference does this roughly .5 of q.factor make in the end?

At what point can you really start noticing the difference?
In the guides it only says to stay below a q.factor of 9. All the parts that matter are in the green area or what I think is the green area as the background looks black to me. I can see the definate grey areas, so I am just assuming if it's not grey then it's ok. I never fully understood the green/grey thing and I have tried hard to. I can't tell a thing from the picture in the CCE guide, but that is not my question after all.

Thanks,
Mark

[waldok]

Mark,

I know you're going to tell me that you need scientific ane mathematical evidence there, but I'll still pop in with my ever "my eyes told me" stuff.

I couldn't notice the slightest difference. To be honest, I couldn't even see any difference between Q=5 and Q=10 on most movies (is it me ? any eye doctor around ?).

From my own (modest) experience with the grey/green mess, I would say this "2nd pass" optimization is a real waste of time, considering I never could "green" everything and I got a very hard time trying to lower Q factor to the recommended values (not speaking about the hours spent adjusting the grey parts). Plus, the results I compared with the tweaked 1pass Roba method didn't show me there was a clear added value to the initial 2pass method.
All this together mean I hardly think a 0.5 Q difference would ever bring anything to the subjective quality of the movie (add to this the quality of your TVSet, the quality of your standalone decoder and video outputs and you'll lose your 0.5 on the way. I'd rather keep the menus if I were you (but I'm not ;))

By the way, in you eternal quest for the best possible quality, do you ever have time to sit down and watch a movie ? (just joking) :D

SOrry I can't help with "facts" but I like discussing these "eyes" vs "mathematics" stuff, and you always bring good subjects here :-)

Waldok:cool:

[markrb]

Maybe this would help more. If I dump the menu's I can use an AVG bitrate that is 500 kbit higher. I am not going through and doing the section by section bitrate increase. I am only modifying the global bitrate while watching it in advanced to get an idea of what the video quality is across the board.

Mark

[waldok]

Maybe you could first use the "keep menus" version, open the advanced section, find out some areas where you have "bitrate problems" (i.e Grey or worse zones), print them out (or screen capture them). THen do the same with your "no menus after all" +500kbit/s bitrate and see the difference in those areas in terms of grey parts. I doubt lots of them will become green.

In my experience, playing around with the global bitrate will not help much, since it will be waste bits on already OK areas if you increase it a lot and not enough correction on problematic areas if you increase it too little. I had to increase it to very high values to get green only values (and like I said I never had them all) and this was a real waste of bits and produced too big files in the end, and this led to a nearly unnoticeable gain in quality.

Myabe I'm wrong here, so please someone correct this if needed.

Waldok:cool:

[Taokane]

Markrb
I've actually done some basic experiments with this. Answer is easy = you will not be able to tell. What I did is encoded a movie (moulin rouge, lots of colors, motion and dark scenes) with several different settings (Q=60 and Q=40) and at different quality levels (q.factor = 3.5 and q.factor = 4.5). Then I went to Fry's, found two side by side 54" digital HDTV's and ran my results on the same scenes for a side by side comparison. I put it to the highly scientific "eyebal" test. To me, I could not tell the difference. A couple of random passerby's couldn't see a better one either. I'll tell you what though, on those big-ass tv's there is a definate quality loss from the original. It's not much, but you can tell my copy just isn't quite as crisp as the original. Particularly in some of the darker scenes, there appeared to be noise in the backgrounds.
Something I do to save space is dump the menu, always. Just movie, or make your own. Additionally, I like to keep director's commentary, but I always encode it to the lowest possible satisfactory level. I usually stick to 64, sounds good without being staticky. If your thing isn't sound, do the main track as well, or avoid DTS.

[markrb]

Taokane thanks for the info, but now I am more undecided then ever.
However not with the bitrate issue. If I keep it this way I will just leave the menu and the lower bitrate.
Now I am thinking of splitting the movie onto two discs and not re-encoding at all. It's not an issue anyone, but me can decide.

The main thing is that this video is hard to re-encode. I am trying to backup Band of Brothers. I would really rather not loose a $90 set like I did with Earth to the Moon when I lost 1 disc and was forced to buy the entire set again to get that one disc back.

Thanks again,
Mark

[waldok]

didn't I just tell you ?

:D

Waldok:cool: