12-22-2003, 06:50 PM
|
Free Member
|
|
Join Date: May 2003
Posts: 10,463
Thanks: 0
Thanked 0 Times in 0 Posts
|
|
Read carefully all what is indicated at the top of this topic and do EXACTLY what is said. Everything is well indicated.
Remember also to use only SHORT names and path (for instance, put all you files in a directory called "c:\kvcd").
|
Someday, 12:01 PM
|
|
Site Staff / Ad Manager
|
|
Join Date: Dec 2002
Posts: 42
Thanks: ∞
Thanked 42 Times in 42 Posts
|
|
|
12-23-2003, 04:16 PM
|
Free Member
|
|
Join Date: Dec 2002
Posts: 18
Thanks: 0
Thanked 0 Times in 0 Posts
|
|
I have read all 37 pages and did not find this problem. I have found problem where TMPGEnc goes over the movie quickly and gives CQ of 90.
My problem is that CQmatic goes through 4 prediction cycles and takes approximately 10min for each cycle and gives me 89.77 value.
When you say to keep path short for which file. I have CQmatic under program files and TMPGenc under C drive.
|
12-23-2003, 05:01 PM
|
Free Member
|
|
Join Date: May 2003
Posts: 10,463
Thanks: 0
Thanked 0 Times in 0 Posts
|
|
Quote:
Originally Posted by ss
When you say to keep path short for which file. I have CQmatic under program files and TMPGenc under C drive.
|
Not for the tools, for ALL FILES INVOLVED IN THE PROJECT : the source (d2v or avi file), the avs script, the tmpgenc project file (tpr)...
(that's not mandatory for all, but do not complicate the things : put EVERYTHING there and you will be sure )
|
12-23-2003, 07:06 PM
|
Free Member
|
|
Join Date: Dec 2002
Location: Buenos Aires, Argentina
Posts: 131
Thanks: 0
Thanked 0 Times in 0 Posts
|
|
I may also suggest to use paths without spaces on them.
Hope it helps, as it helped me.
Gaudi
|
12-26-2003, 11:06 AM
|
Free Member
|
|
Join Date: Jan 2003
Posts: 85
Thanks: 0
Thanked 0 Times in 0 Posts
|
|
Quote:
Originally Posted by cweb
Quote:
Originally Posted by Abond
Quote:
When using CalcuMatic to get the Avg bitrate, use the 'custom' CD setting and subtract the overhead from the CD size
|
That means you know how much is the overhead. Do you really know?
|
I don't know exactly - perhaps a few megabytes. I thought there might be some way to calculate it more precisely. I'll keep experimenting.
|
I'm currently down to 775 megabytes per 80min CD... as 780 was a bit too much and I had to reduce the audio of one segment to 128 from 160...
CqMatic is great, I must say again. Happy Christmas to you Kwag and all those who contribute here.
|
12-26-2003, 01:48 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 cweb
CqMatic is great, I must say again. Happy Christmas to you Kwag and all those who contribute here.
|
Thanks cweb, and Merry Christmas to you too
-kwag
|
12-30-2003, 05:24 PM
|
Free Member
|
|
Join Date: May 2003
Location: UK
Posts: 201
Thanks: 0
Thanked 0 Times in 0 Posts
|
|
After prediction is complete, i'm getting an error from TMPGE; "Illegal Stream Error" , then i press OK, and TMPGE continues encoding, but then i notice the encoded film looks bad quality and colours are very blurry and dark?
Script;
AviSource("C:\**.avi")
BlindPP(cpu=4)
Blockbuster(method="noise",detail_min=1,detail_max =3,variance=0.1,seed=1)
Convolution3D(1, 6, 12, 6, 8, 2.8, 0)
GripCrop(480, 480, overscan=1, source_anamorphic=false)
GripSize(resizer="LanczosResize")
Undot()
TemporalSoften(2,7,7,3,2)
DctFilter(1, 1, 1, 1, 1, 1, 0.5, 0)
GripBorders()
any help plz?
__________________
TheDJ®
|
01-15-2004, 04:22 AM
|
Free Member
|
|
Join Date: Jul 2003
Location: Blighty, PAL Land
Posts: 133
Thanks: 0
Thanked 0 Times in 0 Posts
|
|
I've noticed that Kwag isn't replying much lately.
Lets hope that he's busy updating CQMatic
While we're on the subject, any chance of CQMatic runiing in batch mode ?
Cheers Kwag
__________________
MisterMickster
|
01-27-2004, 08:23 AM
|
Free Member
|
|
Join Date: Jan 2003
Posts: 85
Thanks: 0
Thanked 0 Times in 0 Posts
|
|
I'm recording episodes of the show My Hero from satellite, and this one (some episodes at least) causes CQMatic to give me a large incorrect CQ (over 89), and the file difference indicated in CQMatic is a very large negative number... The correct CQ would be around 60 (I'm putting 3 episodes in KVCDx3 per 80 minute CD with about 6M for the SVCD wrapper overhead).
Just thought it would be useful to know - perhaps this could be fixed somehow.
|
01-27-2004, 08:25 AM
|
Free Member
|
|
Join Date: May 2003
Location: UK
Posts: 201
Thanks: 0
Thanked 0 Times in 0 Posts
|
|
i've stopped using CQMatic to be honest, i prefer dvd2svcd.
__________________
TheDJ®
|
01-27-2004, 08:28 AM
|
Free Member
|
|
Join Date: Jan 2003
Posts: 85
Thanks: 0
Thanked 0 Times in 0 Posts
|
|
I prefer CQMatic because I don't like loss of control.. I prefer to set up everything myself - avisynth, tmpgenc, filters, script etc. CQMatic just helps with file prediction size.
In the above case I mentioned, I am using X1 prediction.
|
01-27-2004, 09:27 AM
|
Free Member
|
|
Join Date: May 2003
Posts: 10,463
Thanks: 0
Thanked 0 Times in 0 Posts
|
|
Quote:
Originally Posted by TheDJ®
i've stopped using CQMatic to be honest, i prefer dvd2svcd.
|
It seems you are lukier than I am because on the last 10 attempts with this software, 7 got way under the target (100 MB under ) and only one was correct
BTW, are you aware that DVD2SVCD use a 20% B-spoilage and that is definitely not recommanded ?
|
01-27-2004, 09:42 AM
|
Free Member
|
|
Join Date: May 2003
Location: UK
Posts: 201
Thanks: 0
Thanked 0 Times in 0 Posts
|
|
Quote:
Originally Posted by Dialhot
Quote:
Originally Posted by TheDJ®
i've stopped using CQMatic to be honest, i prefer dvd2svcd.
|
It seems you are lukier than I am because on the last 10 attempts with this software, 7 got way under the target (100 MB under ) and only one was correct
|
I'm sorry to hear that, seems the program isnt ready yet? I'm sure Kwag is dealing with it.
Quote:
Originally Posted by Dialhot
BTW, are you aware that DVD2SVCD use a 20% B-spoilage and that is definitely not recommanded ?
|
I'm not quite sure wot u mean, but its doing a better job for me than CQMatic is. I had the freeze problem every split second if u remember, and converted same file in dvd2svcd, and that did it perfectly, so stick to wot works is my motto. Quality is still good with divx/xvid, but there is still this smudgy look to overall picture (even with CQmatic) and skin colour on face especially. Other than that, its great! I have burned many many films, still have alot to watch
But i still have that problem with shorter films/documentary's, the prediction (trial and error), doesnt encode the whole documentary, would always miss maybe 2-3 minutes off the end. I have yet to try short films/dcoumentarys (under 1hour) with CQMatic, as i had problems doing normal movies.
__________________
TheDJ®
|
01-27-2004, 09:54 AM
|
Free Member
|
|
Join Date: May 2003
Posts: 10,463
Thanks: 0
Thanked 0 Times in 0 Posts
|
|
Quote:
Originally Posted by TheDJ®
I'm sorry to hear that, seems the program isnt ready yet? I'm sure Kwag is dealing with it.
|
I was atlking about DVD2SVCD ! And there is no way to deal with that as the problem is internal to the tool (I talk about that a lot in the DVD2SVCD forum).
Quote:
I'm not quite sure wot u mean, but its doing a better job for me than CQMatic is. I had the freeze problem every split second if u remember, and converted same file in dvd2svcd, and that did it perfectly, so stick to wot works is my motto.
|
CQMatic just calculate the CQ, and does nothind else. So your probmem is not with CQMatic. It's possible that indeed the B-spoilage used by DVD2SVCD fix you freezing problem
Quote:
Quality is still good with divx/xvid, but there is still this smudgy look to overall picture (even with CQmatic) and skin colour on face especially.
|
Thats is a divx/xvid issue
Quote:
Other than that, its great! I have burned many many films, still have alot to watch
|
What are your settings in the "Encoder" tab, under the section "CQ Test Values" ?
Quote:
doesnt encode the whole documentary, would always miss maybe 2-3 minutes off the end.
|
Did you ever read what I told to you at this subject ???
Look at the avs script generated by D2S : there is no way the movie is not encoded completly ! For that you need a "trim" instruction at the end of the script. Look into it : there isn't any line like this
Stop asking to DVD2SVCD to remove the temporary files and stop begging it to make CD image and you will see that you have ALL the movie there but split in two files.
I told you that so much...
|
01-27-2004, 10:50 AM
|
Free Member
|
|
Join Date: May 2003
Location: UK
Posts: 201
Thanks: 0
Thanked 0 Times in 0 Posts
|
|
Quote:
Originally Posted by Dialhot
What are your settings in the "Encoder" tab, under the section "CQ Test Values" ?
|
Offset CQ Value: 80, CQ Value Factor: 1.000, % of movie to test: 2
Quote:
Originally Posted by Dialhot
Stop asking to DVD2SVCD to remove the temporary files and stop begging it to make CD image and you will see that you have ALL the movie there but split in two files.
I told you that so much...
|
I never delete temporary files, only manually, and never make CD images coz i like to check the movie before burning.
__________________
TheDJ®
|
01-27-2004, 10:56 AM
|
Free Member
|
|
Join Date: May 2003
Posts: 10,463
Thanks: 0
Thanked 0 Times in 0 Posts
|
|
Quote:
Originally Posted by TheDJ®
Offset CQ Value: 80, CQ Value Factor: 1.000, % of movie to test: 2
|
And how much test does it take to d2s generally to find the CQ ?
Quote:
I never delete temporary files, only manually, and never make CD images coz i like to check the movie before burning.
|
At least I have an answer on this point !
So, as you do that, did you check the mvp file (the video stream) that is generated. Does it contain all the video or not ?
Don't you have also 2 files generated by bbmpeg (called xxx00.mpg and xxx01.mpg) ?
|
01-27-2004, 11:38 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 TheDJ®
I'm sorry to hear that, seems the program isnt ready yet? I'm sure Kwag is dealing with it.
|
Nope. No more
It's IMPOSSIBLE to achieve a high accuracy by sampling only parts of a movie. There is NO WAY for me, or for ANYONE, to devise a highly accurate prediction method, unless the complete footage can be pre-analyzed. The best I did was the X3 mode in CQMatic, and this can fail, if the sampled parts of the movie vary from the non-sampled parts, if they are too different in action content. If the movie is very linear, meaning the action is well distributet throughout the movie, then the result will be excelent and below 2% in final file size. But this is not the case for all movies. So I've done my best, and that's it.
BTW, for KDVDs, I'm done with prediction as it's not needed, at least for KDVDs Look at the thread "ffvfw" in this forum
-kwag
|
01-27-2004, 02:00 PM
|
Free Member
|
|
Join Date: May 2003
Location: UK
Posts: 201
Thanks: 0
Thanked 0 Times in 0 Posts
|
|
Quote:
Originally Posted by Dialhot
And how much test does it take to d2s generally to find the CQ ?
|
On average i would say between 4-8 tests.
Quote:
So, as you do that, did you check the mvp file (the video stream) that is generated. Does it contain all the video or not ?
|
I did not check, but will check on the next conversion.
Quote:
Don't you have also 2 files generated by bbmpeg (called xxx00.mpg and xxx01.mpg) ?
|
Nope, never got 2 files.
__________________
TheDJ®
|
01-27-2004, 02:04 PM
|
Free Member
|
|
Join Date: May 2003
Location: UK
Posts: 201
Thanks: 0
Thanked 0 Times in 0 Posts
|
|
Quote:
Originally Posted by kwag
It's IMPOSSIBLE to achieve a high accuracy by sampling only parts of a movie. There is NO WAY for me, or for ANYONE, to devise a highly accurate prediction method, unless the complete footage can be pre-analyzed.
-kwag
|
Hi Kwag, thnx for the reply.
One question, when doing a prediction, how does it find different parts of the movie to test? I was wondering if it finds highest peaks in the film, and if so how? Or is this the part which is impossible?
__________________
TheDJ®
|
01-27-2004, 02:21 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 TheDJ®
One question, when doing a prediction, how does it find different parts of the movie to test?
|
It divides the movie in "slices" (2 minutes of footage), and then encodes those portions. Quote:
I was wondering if it finds highest peaks in the film, and if so how? Or is this the part which is impossible?
|
In order to do that, a scan would have to be done on the complete movie.
Even though this is possible, there's no way to determine the compressibility that the encoder will appply to the particular movie, so it's useless.
This would have to be done internally to an encoder, where THE encoder knows how much it will compress the material.
-kwag
|
All times are GMT -5. The time now is 07:55 PM — vBulletin © Jelsoft Enterprises Ltd
|