Here are the results from both CQmatic latest version and TOK for the same move and same avs script using KVCD3 528x480 template. CQMatic was way off.

================================================== ===========
ToK Log: C:\Phobic\Phobic.avs
================================================== ===========

Resolution (fps):528x480 (23.976 fps)
Total Frames: 126320
Total Time : 01:27:49
-------------------------------------------------------------

Audio Size: 73,762,012
Required Video Size: 745,437,988

Factor: 19.980
Desired Sample Size: 37,309,209

-------------------------------------------------------------
New Faster Prediction
-------------------------------------------------------------

Full Sample
Next CQ: 70.000. Sample Size: 22,463,687
Small Sample
Next CQ: 70.000. Sample Size: 1,994,036
Predicting...
NextCQ: 116.261 >=90, so check is needed
Next CQ: 90.000
Sample Size: 50,857,382

Next CQ: 66.024. Sample Size: 21,816,612
Next CQ: 78.414. Sample Size: 29,145,376
Next CQ: 83.488. Sample Size: 49,119,035
Next CQ: 80.719. Sample Size: 42,155,007
Next CQ: 79.713. Sample Size: 36,089,250
Next CQ: 80.066. Sample Size: 38,550,613
Next CQ: 79.889. Sample Size: 37,307,044

Exit Condition: 0.500 % reached ! yahoo !
Tries : 8


Final CQ: 79.889
Total Time For Predicition: 00:32:13


-------------------------------------------------------------
Encoding C:\Phobic\Phobic.avs
-------------------------------------------------------------

Encoding... CQ : 79.889
Final Encoded Size: 710,210,986
Muxing ...


Total Time (all operations): 03:22:47

Finished

With audio which is 112 Bitrate (mp2) this came out perfect at 800,004

http://www.kvcd.net
CQMatic Version 1.05.00
Copyright Softronex Corporation, 2003.
All rights reserved.
Time: 03:46:41 Date: 08/03/2003
Ready!
Project: C:\BBMPEG\Phobic.tpr

Creating: CQMatic.tpr

C:\MyEncoding\ToK_0.0.5.3\work\Phobic.m1v
Execute.
Movie Time: 87
Average Bitrate: 1140
Full Encode mode
Executing Prediction Phase...
Process started at 03:47:06
On 08/03/2003
Entering CF phase.
Using CQ of 50.00
FAST source analysis...
Process time: 0.48 minutes.
Encoder end.
Correction Factor = 1.6342
Exit CF phase.
Using CQ of 60.00
Encoder started...
Process time: 3.73 minutes.
Encoder end.
File size difference = 1.637004
Last CQ = 60.00
Current CQ = 90.00
CQ difference = 30.000000
Using CQ of 90.00
Encoder started...
Process time: 3.58 minutes.
Encoder end.
File size difference = 1.073913
CQ ABOVE watermark. CQ set to 90.0
Using CQ of 90.00
Encoding set to Full encode.
Full encode start...
CQMatic complete!
Total minutes of process: 7.80
Process ended at 03:54:54
On 08/03/2003


With out audio this came out to 870,463. TOK recommedned a CQ of 79.889 and CQ Matic at 90.

Bud

I just posted in another thread about CQ Matic requesting a CQ of 90 and my encode was way over. That was with version 1.05. I am now testing with RC5 and my CQ is on target back down to a CQ of around 79.

Note that I encoded the movie with ToK and it picked CQ of 81 and went over by 20MB so I know the CQ of 90 by CQMatic is way over.

I don't know what changed, but RC5 is more accurate for me then 1.05 or 1.02 which i both tried and got CQ90

I was thinking about what TOK was doing and what CQMatic was doing, so did a predictiion run only. I did not change the Max/Min in TMPGEnc, with TOK we don't change it, I left it at 2500/300 here is teh CQMatic prediction:

http://www.kvcd.net
CQMatic Version 1.05.00
Copyright Softronex Corporation, 2003.
All rights reserved.
Time: 06:41:01 Date: 08/03/2003
Ready!
Project: C:\BBMPEG\Phobicdefault.tpr

Creating: CQMatic.tpr

C:\MyEncoding\ToK_0.0.5.3\work\Phobic.m1v
Execute.
Movie Time: 87
Average Bitrate: 1079
Prediction Only mode
Executing Prediction Phase...
Process started at 06:41:19
On 08/03/2003
Entering CF phase.
Using CQ of 50.00
FAST source analysis...
Process time: 0.37 minutes.
Encoder end.
Correction Factor = 1.3328
Exit CF phase.
Using CQ of 60.00
Encoder started...
Process time: 3.72 minutes.
Encoder end.
File size difference = 1.488962
Last CQ = 60.00
Current CQ = 74.67
CQ difference = 14.668869
Using CQ of 74.67
Encoder started...
Process time: 3.60 minutes.
Encoder end.
File size difference = 1.050896
Last CQ = 74.67
Current CQ = 78.47
CQ difference = 3.800316
Using CQ of 78.47
Encoder started...
Process time: 3.58 minutes.
Encoder end.
File size difference = 1.047408
Last CQ = 78.47
Current CQ = 82.19
CQ difference = 3.720055
Using CQ of 82.19
Encoder started...
Process time: 3.62 minutes.
Encoder end.
File size difference = 0.820293
Last CQ = 82.19
Current CQ = 80.33
CQ difference = 1.860031
Using CQ of 80.33
Encoder started...
Process time: 3.58 minutes.
Encoder end.
File size difference = 0.897289
Last CQ = 80.33
Current CQ = 79.40
CQ difference = 0.930008
Using CQ of 79.40
Encoder started...
Process time: 3.60 minutes.
Encoder end.
File size difference = 0.969978
Last CQ = 79.40
Current CQ = 78.93
CQ difference = 0.465012
Using CQ of 78.93
Encoder started...
Process time: 3.58 minutes.
Encoder end.
File size difference = 1.024269
Last CQ = 78.93
Current CQ = 79.17
CQ difference = 0.232506
CQMatic complete!
Total minutes of process: 25.65
Process ended at 07:06:58
On 08/03/2003

As you can see it is very close to what TOK gave me in the tread above, 79.17. Again I used the same avs script file, I just did not change anything in TMPGEnc.

Bud

So you're going back to RC-5 too, eh? So am I at the moment. I just did an encode from dvd source and the m1v itself was 800mb. So you got more accurate predictions when you went back to 300/2500 for min/max bitrate? I'll try that next. Right now I'm still using 0.57*avgbitrate/2000, but instead seeing if using vcalc gives a more accurate avgbitrate than moviestacker. Just out of curiosity, which did you use? vcalc or MS?

I used vcal, but on to get the Max, I did not use the ave, and I only used vcal max to input to CGMatic not TMPGEnc, I left the defaults alone in TMPGE.

Bud