07-30-2003, 12:06 PM
|
Free Member
|
|
Join Date: Apr 2003
Location: Fallston, MD. USA
Posts: 419
Thanks: 0
Thanked 0 Times in 0 Posts
|
|
Hi kwag:
Last night I encoded a 72 minutes movie, and this morning, when I checked the file it was over 1.1 G. Then the rosary of errors started to take place:
1.- Thinking that everything was OK, I closed CQmatic, without saving the log.
2.- Then, I said to myself: You have not changed any of the settings at all, therefor you can go back and rerun CQmatic in prediction only mode, and I did. BUT I did not change the output file name, and the new prediction erased my 1.1 Gig file, so now, I do not have the encoded file. Live and learn. But I have the npr files and the new log.
The CQmatic .tpr file is telling me that it encoded the file at a CQ of 80.92
Quote:
Job.OutputConfig_Graphic.JPEGCompressionQuality = 80
|
While the log run this morning, is telling me:
Quote:
Last CQ = 80.924927
Current CQ = 81.090996
CQ difference = 0.166069
CQMatic complete!
Total minutes of process: 69.05
Process ended at 12:14:52
On 07/30/2003
|
Which is saying 81.09.
The second problem that I had, is that the Mpeg file was horizontally stretched. The film that I was backing up had a ratio of 4:3, but when I used such ratio I had the stretched output, this was resolved by setting Tmpenc to 16:9 ratio. (I had experimented with and without gripfit, without grip fit I had script errors, perhaps because I did not know where to place the border line).
Thanks for your help Kwag
P.S>- I have not tried the new version, I will do that after lunch
|
Someday, 12:01 PM
|
|
Site Staff / Ad Manager
|
|
Join Date: Dec 2002
Posts: 42
Thanks: ∞
Thanked 42 Times in 42 Posts
|
|
|
07-30-2003, 12:15 PM
|
Free Member
|
|
Join Date: Apr 2002
Location: Puerto Rico, USA
Posts: 13,537
Thanks: 0
Thanked 0 Times in 0 Posts
|
|
Quote:
Originally Posted by totonho03
The CQmatic .tpr file is telling me that it encoded the file at a CQ of 80.92
Quote:
Job.OutputConfig_Graphic.JPEGCompressionQuality = 80
|
|
That's JPEG compression quality setting. Not MPEG Quote:
While the log run this morning, is telling me:
Quote:
Last CQ = 80.924927
Current CQ = 81.090996
CQ difference = 0.166069
CQMatic complete!
Total minutes of process: 69.05
Process ended at 12:14:52
On 07/30/2003
|
Which is saying 81.09.
|
On the last run, the CQ to be used will actually be Last CQ = 80.924927 because last is actually the "Current", and the last "Current" never did execute
-kwag
|
07-30-2003, 04:22 PM
|
Free Member
|
|
Join Date: Apr 2003
Location: Fallston, MD. USA
Posts: 419
Thanks: 0
Thanked 0 Times in 0 Posts
|
|
Hi kwag:
Thank you for the explanation regarding current and current CQs
I followed your advise, and downloaded tha latest version, and the results were completely different to the one I received in my previous prediction.
Using v 1.0-RC 5, I had the following:
Quote:
Last CQ = 80.924927
Current CQ = 81.090996
CQ difference = 0.166069
CQMatic complete!
Total minutes of process: 69.05
Process ended at 12:14:52
On 07/30/2003
|
but now, using the same parameters, as far as avs and tmpgsettings, the log is given me this:
Quote:
http://www.kvcd.net
CQMatic Version 1.01 Release.
Copyright Softronex Corporation, 2003.
All rights reserved.
Time: 15:58:41 Date: 07/30/2003
Ready!
Project: F:\DVDProjects\CQMATIC\tmpgenc\Output\pinochiootto .tpr
Creating: CQMatic.tpr
F:\DVDProjects\CQMATIC\tmpgenc\Output\pinociootto. m1v
Execute.
Movie Time: 72
Average Bitrate: 1397
Prediction Only mode
Executing Prediction Phase...
Process started at 15:59:13
On 07/30/2003
Entering CF phase.
Using CQ of 50.00
FAST source analysis...
Process time: 1.38 minutes.
Encoder end.
Correction Factor = 1.4174
Exit CF phase.
Using CQ of 60.00
......................
....................
Last CQ = 79.611168
Current CQ = 81.011963
Using CQ of 81.01
Encoder started...
Process time: 10.55 minutes.
Encoder end.
File size difference = 1.116484
CQ ABOVE watermark. CQ set to 90.0
Using CQ of 90.00
CQMatic complete!
Total minutes of process: 54.13
Process ended at 16:53:21
On 07/30/2003
|
Now, I am afraid to encode because if at CQ8o+, the file was over 1.1Giga, then with a CQ of 90 (Am I reading this log correct?), the file size would be bigger)
Thanks
Totonho03
|
07-30-2003, 04:45 PM
|
Free Member
|
|
Join Date: Apr 2002
Location: Puerto Rico, USA
Posts: 13,537
Thanks: 0
Thanked 0 Times in 0 Posts
|
|
Hi totonho03,
What is your source
What I can't understand is how it's possible that even at CQ=80, a 72 minute movie can make a file over 1GB in size
This doesn't make sense at all
If I take a 528x480 2 hour movie, and I set CQ to 80, then "maybe" I get a 1GB size. But not even at 704x480 CQ=80 you get a 72 minute movie to that size
There's something wrong in your encoder settings or elsewhere
What is your MIN bitrate and your MAX bitrate set to
-kwag
|
07-30-2003, 06:14 PM
|
Free Member
|
|
Join Date: Apr 2003
Location: Fallston, MD. USA
Posts: 419
Thanks: 0
Thanked 0 Times in 0 Posts
|
|
Hello kwag:
Thank you for your response. Here are the answers to your questions:
Quote:
The source is a decrypted dvd.
Quote:
There's something wrong in your encoder settings or elsewhere
|
You may very well be correct in this ascertion. I used dvd2avi and the d2v file was used in moviestacker, which is where I found the lenght and the average bit rate of the film.
Quote:
What is your MIN bitrate and your MAX bitrate set to
|
The encoder settings are as follow:
KVCD-CQ-352x480-_NTSC 120 min .Film_-PLUS
Stream type: Mpeg-1
Size= 352X40; Aspect ratio: 16:9. 525 lines (Output is not stretched with this setting); Frame rate= 29.97 fps
Constant quality= (CQ) - Using Motion estimate search;
Max bit rate= 2000; MIN bit rate 796
Source range is unchecked, and I have start frame at 0 and end at -1
Have saved the .tpr as text, and have not touched any other TMPGenc settings
I will do another test, actually will do 2 more tests, starting all over again from dvd2avi.......... One with Pinochio, and the other with another DVD.
Thanks
Totonho
P.S.- Please let me know if my settings are correct. Thanks again
|
07-30-2003, 06:50 PM
|
Free Member
|
|
Join Date: Apr 2002
Location: Puerto Rico, USA
Posts: 13,537
Thanks: 0
Thanked 0 Times in 0 Posts
|
|
Hi totonho03,
It might (still!) be a problem dealing with 29.97fps interlaced sources
I'm having problems predicting my "Styx" concert, which is VIDEO 29.97fps Interlaced.
I'm working specifically on this issue.
-kwag
|
07-30-2003, 09:28 PM
|
Free Member
|
|
Join Date: Apr 2003
Location: Fallston, MD. USA
Posts: 419
Thanks: 0
Thanked 0 Times in 0 Posts
|
|
Thanks kwag:
I will wait a bit before ending another one, but I will still try the on prediction mode only.
I will let you know how this turns out.
Regards
Totonho
|
07-31-2003, 07:44 AM
|
Free Member
|
|
Join Date: Apr 2003
Location: Fallston, MD. USA
Posts: 419
Thanks: 0
Thanked 0 Times in 0 Posts
|
|
Hello again:
Last night I started the prediction, using the latest program. The only thing that I did not re-do, is to convert the files from vob to d2v, but went from d2v to QCmatic, via Moviestacker and Tmpgenc.
Here is my log:
Quote:
http://www.kvcd.net
CQMatic Version 1.03 Release.
Copyright Softronex Corporation, 2003.
All rights reserved.
Time: 23:55:26 Date: 07/30/2003
Ready!
Project: F:\DVDProjects\CQMATIC\tmpgenc\Output\pinochio2.tp r
Creating: CQMatic.tpr
F:\DVDProjects\CQMATIC\tmpgenc\Output\pinochio2.m1 v
Execute.
Movie Time: 72
Average Bitrate: 1397
Prediction Only mode
Executing Prediction Phase...
Process started at 00:00:06
On 07/31/2003
Entering CF phase.
Using CQ of 50.00
FAST source analysis...
Process time: 1.33 minutes.
Encoder end.
Correction Factor = 1.4138
Exit CF phase.
Using CQ of 60.00
Encoder started...
Process time: 11.55 minutes.
Encoder end.
File size difference = 1.218494
Last CQ = 60.00
Current CQ = 66.55
Using CQ of 66.55
Encoder started...
Process time: 11.50 minutes.
Encoder end.
File size difference = 1.153124
Last CQ = 66.55
Current CQ = 69.83
Using CQ of 69.83
Encoder started...
Process time: 11.53 minutes.
Encoder end.
File size difference = 1.150079
Last CQ = 69.83
Current CQ = 71.47
Using CQ of 71.47
Encoder started...
Process time: 11.48 minutes.
Encoder end.
File size difference = 1.121916
Last CQ = 71.47
Current CQ = 72.29
Using CQ of 72.29
Encoder started...
Process time: 11.43 minutes.
Encoder end.
File size difference = 1.114702
Last CQ = 72.29
Current CQ = 72.70
Using CQ of 72.70
Encoder started...
Process time: 11.40 minutes.
Encoder end.
File size difference = 1.112297
Last CQ = 72.70
Current CQ = 72.90
Low diff. CQ bumped!
Using CQ of 90.00
Encoder started...
Process time: 11.40 minutes.
Encoder end.
File size difference = 1.084036
Last CQ = 90.00
Current CQ = 90.00
Using CQ of 90.00
Encoder started...
Process time: 11.50 minutes.
Encoder end.
File size difference = 1.084033
CQ ABOVE watermark. CQ set to 90.0
Using CQ of 90.00
CQMatic complete!
Total minutes of process: 93.15
Process ended at 01:33:15
On 07/31/2003
|
Please note that I found one setting that was incorrect, that is I was using interlace in Tmpg, in reality the setting should had been progressive. This was changed prior to running the prediction.
I have noticed something different with this new version, that is, the difference between QCs never goes below 1.11xxx, and the QC jumped from 72.90 to 90.
Regards
Totonho03
|
07-31-2003, 10:52 AM
|
Free Member
|
|
Join Date: Apr 2002
Location: Puerto Rico, USA
Posts: 13,537
Thanks: 0
Thanked 0 Times in 0 Posts
|
|
Quote:
Originally Posted by totonho03
File size difference = 1.112297
Last CQ = 72.70
Current CQ = 72.90
Low diff. CQ bumped!
|
This happens because the last sample file size difference to the current sample was just so low, that I force a CQ change out of scale, so we don't stay in a long cycle trying to increase CQ in very small steps. This is a problem with TMPEG's internal CQ scale. It's not linear, and has some "null" segments where you can make a small change in the CQ value, but there's no file size difference. Quote:
Using CQ of 90.00
Encoder started...
Process time: 11.40 minutes.
Encoder end.
File size difference = 1.084036
Last CQ = 90.00
Current CQ = 90.00
Using CQ of 90.00
Encoder started...
Process time: 11.50 minutes.
Encoder end.
File size difference = 1.084033
CQ ABOVE watermark. CQ set to 90.0
Using CQ of 90.00
CQMatic complete!
Total minutes of process: 93.15
Process ended at 01:33:15
On 07/31/2003
|
And of course, your movie is 352x240 and 72 minutes, so even at CQ=90, CQMatic couldn't get to the wanted file size, so it sets CQ=90 as the maximum limit. Quote:
Please note that I found one setting that was incorrect, that is I was using interlace in Tmpg, in reality the setting should had been progressive.
|
That was the problem right there
Encode your movie at CQ=90, and your file will fit on a CD, because it will never reach the wanted file size. Your final file size will be ~ ( Your Wanted / 1.084033 ) as seen from the last File size difference = 1.084033
-kwag
|
07-31-2003, 11:00 AM
|
Free Member
|
|
Join Date: Apr 2003
Location: Fallston, MD. USA
Posts: 419
Thanks: 0
Thanked 0 Times in 0 Posts
|
|
Thanks kwag:
So now all I need to do is to go to TMPgenc, and start the encoding with a CQ of 90. There is no need to go back to CQmatic and begin the encoding proces to encode Correct? Thanks again. Will do that later on today.
Totonho03
|
07-31-2003, 11:14 AM
|
Free Member
|
|
Join Date: Apr 2002
Location: Puerto Rico, USA
Posts: 13,537
Thanks: 0
Thanked 0 Times in 0 Posts
|
|
Quote:
Originally Posted by totonho03
There is no need to go back to CQmatic and begin the encoding proces to encode Correct?
|
That's right
Just encode at CQ=90.
-kwag
|
08-01-2003, 08:18 AM
|
Free Member
|
|
Join Date: Apr 2003
Location: Fallston, MD. USA
Posts: 419
Thanks: 0
Thanked 0 Times in 0 Posts
|
|
Hello Kwag:
Last night I encoded the dvd at CQ 90, and ended up with a file of 1.255,032 Giga. I was using version 1.3. Will try it w/ version 1.4
Regards
Totonho
Edit.- I am not so sure of this, but I think that I found the problem:
On TMPGenc, under the Advanced tab, I was setting the video source type as progressive (Non interlaced); however, I was not checking the settings under the Video tab, which are grayed, and when I was redoing everything, I noticed that the encode mode, under the Video tab, was still still showing interlace mode. I changed it and is now reading non-interlaced. I will let you know the results of this test in an hour or so........
|
08-01-2003, 10:51 AM
|
Free Member
|
|
Join Date: Apr 2003
Location: Fallston, MD. USA
Posts: 419
Thanks: 0
Thanked 0 Times in 0 Posts
|
|
Hello Kwag:
I believe that I can finally say that CQmatic is working in my machine, I have not encoded it yet, but the prediction run with no problems, (well a small one, I was viewing Tmpgenc setting while it was encoding, and it did not like that at all, after I closed the settings window the prediction run smoothly).
After 6 iterations, the final CQ reads as follow:
Quote:
Using CQ of 72.85
Encoder started...
Process time: 11.47 minutes.
Encoder end.
File size difference = 1.114229
Last CQ = 72.85
Current CQ = 73.06
CQ difference = 0.207291
CQMatic complete!
Total minutes of process: 92.87
Process ended at 11:24:54
On 08/01/2003
|
And just to confirm it, I should encode the film at the last CQ number, which in this case is CQ 72.85, correct?
One last question here, I remember reading that if I were to use 23.xx, instead of the the 29.xx fps, I would not have a problem, Is this also correct?
Thank you for your help kwag, you have Job's patience...............
Totonho03
|
All times are GMT -5. The time now is 11:26 PM — vBulletin © Jelsoft Enterprises Ltd
|