Bitrates: Why does Everyone like CQ Instead of CBR?

[Shibblet]

Alright... I have been doing quite a bit of 480x480 movies with the 352x240x120min template. Bitrate is between 1800max and 300min. (My standalone won't go lower than 300). I am using CQ / Calcumatic and having some Awesome results with CQ Prediction. I am getting 90 minute movies on to one 80min CD, with the MPEG files turning up around 770megs - 780megs each.

Now CQ Matic takes a lot of prediction time. I would say for the high quality prediction, around about 20 minutes to encode each prediction.
(1.8 ghz Athlon XP, 1 Gig Ram, TMPEG Encoder 2.52 MPEG-1 Stream)

Now the templates for TMPEG Encoder provided on the KVCD.NET site are great ways to get going. The bitrates provided in each template works JUST FINE in almost any case.

But what I am having a hard time getting a grasp of is the way the CQ quality setting (0 - 100) adjusts the Bitrate if it's variable. If the quality setting is at 90-100, the bitrate stays near the maximum, almost like CBR. So wouldn't file prediction size be a lot easier with CBR? And if so, why all the trouble of working with CQ?

Couldn't you just apply the formula?

(Movie Minutes * 60) = Seconds
(790megs * 1024) = Approx CD Size (808960)
(Approx CD Size * 8) = Total CD Kbits (6471680)
(Total CD kbits) / (Seconds) = Constant Bitrate

So for a 90 minute movie...
(90 * 60) = 5400
(6471680 / 5400) = 1198.45
CBR = 1198
Which seems perfect for a 352x240 movie.

[rds_correia]

Hi Shibblet ,
Let me answer you this way.
Try encoding and burning to 2 CD-RWs the same movie in both possible was.
One with the classic CBR way and the other one with the CQ way.
Watch them in a good TVset, I mean something larger than 25 inches and tell me what you think.
I did that a LONG time ago and I found the CQ much better than the CBR, although it takes time doing the prediction.
Let us know your findings.
Cheers