Womble Mpeg Video Wizard cutlist
Womble Mpeg Video Wizard cutlist
Hi. I have been testing comskip for a couple of months now and like it very much. I would love to see support for Womble "Mpeg Video Wizard" which I use for all cutting and merging of mpeg files. I think the output file is named <trim list file>.tls. I know comskip has support for Womble output, but that files can't be used with Mpeg Video Wizard. Must be for some old or another Womble application? What do you think?
I sure could use this also.
Below is example of .wme file format in mpeg video wizard:
//
// Bookmark List (Thursday, April 27, 2006)
//
// Usage:
// 1. A bookmark list block is formatted by the following consecutive lines
// (1) bookmark list identification string
// (2) complete MPEG movie file path name
// (3) video stream id, 0 for default
// (4) bookmark value format id = frame number
// (5) total number of bookmarks that follow
// (6) bookmark frame offset and check box status
// 2. Text line started with two consecutive slashes are comment lines,
// and they will be ignore by the reader.
//
[BookmarkList]
PathName= C:\VidTst\Will - Grace - Secrets - Lays.mpg
VideoStreamID= 224
Format= frame
Count= 19
0 1
3940 0
4844 0
5296 0
7545 0
9345 0
10695 1
28511 0
29382 0
29831 0
30282 0
32081 0
32530 0
33429 0
35858 0
36315 1
51372 0
52281 0
53197 0
Below is example of .wme file format in mpeg video wizard:
//
// Bookmark List (Thursday, April 27, 2006)
//
// Usage:
// 1. A bookmark list block is formatted by the following consecutive lines
// (1) bookmark list identification string
// (2) complete MPEG movie file path name
// (3) video stream id, 0 for default
// (4) bookmark value format id = frame number
// (5) total number of bookmarks that follow
// (6) bookmark frame offset and check box status
// 2. Text line started with two consecutive slashes are comment lines,
// and they will be ignore by the reader.
//
[BookmarkList]
PathName= C:\VidTst\Will - Grace - Secrets - Lays.mpg
VideoStreamID= 224
Format= frame
Count= 19
0 1
3940 0
4844 0
5296 0
7545 0
9345 0
10695 1
28511 0
29382 0
29831 0
30282 0
32081 0
32530 0
33429 0
35858 0
36315 1
51372 0
52281 0
53197 0
I can see the pathname as a potential problem;--however for me it would not be a problem.
I believe the program may "complain" if the path is different, but I think it will still load the file. I will try to check this out.
The 1 behind a frame number does refer to a show segment. I am not sure which is best, i.e. to show the show segments or not as the file is loaded. Mpeg Video Wizard can make its own decisions as to which to include by the user allowing it, but its decisions may not be as good as Comskip.
Perhaps it might be best to have Comskip set its selections in the output file, and then they can be over ridden in Mpeg Video Wizard if desired.
Thank You!
I believe the program may "complain" if the path is different, but I think it will still load the file. I will try to check this out.
The 1 behind a frame number does refer to a show segment. I am not sure which is best, i.e. to show the show segments or not as the file is loaded. Mpeg Video Wizard can make its own decisions as to which to include by the user allowing it, but its decisions may not be as good as Comskip.
Perhaps it might be best to have Comskip set its selections in the output file, and then they can be over ridden in Mpeg Video Wizard if desired.
Thank You!
I tried the 79_39 version for the mls output. I didn't get any output with that selection for .mls . Tried it with file names with spaces and with short file names. I didn't see it in any other directories either. Program would produce other file output formats however. Also tried it with 0 and 1 set for .mls in .ini file.
I thought it was strange that the new comskip.exe file was the same size as the prior version, but a contents compare said they were different, although I didn't attempt to see what exactly was different.
I tried the latest with the correct .exe file. It did produce a .mls file output. One problem with it noticed as opened in MPEG Video Wizard. It complains about file format. The problem is the Count= line in the .mls output. For the mpg file I tested, it has a count of 9 listed, but there are only 7 clip points. Looking at the .wme output file, there is a total of 7 Cliplist points. So the count appers to be wrong. It did mount in Mpeg Video Wizard after "complaining" about the file format.
I will try it for awhile and see how it works and report back.
One comment about Comskip results with testing I have done so far. It seems to produce good results except on some break points which are "fade to white" at commercial stop or start. This is without any changes in tuning. It would seem that fade to white would be treated about the same as fade to black in the detection process as they appear to be essentially inverse processes. Not sure how comskip addresses this. This is USA cable channels.
I tried the latest with the correct .exe file. It did produce a .mls file output. One problem with it noticed as opened in MPEG Video Wizard. It complains about file format. The problem is the Count= line in the .mls output. For the mpg file I tested, it has a count of 9 listed, but there are only 7 clip points. Looking at the .wme output file, there is a total of 7 Cliplist points. So the count appers to be wrong. It did mount in Mpeg Video Wizard after "complaining" about the file format.
I will try it for awhile and see how it works and report back.
One comment about Comskip results with testing I have done so far. It seems to produce good results except on some break points which are "fade to white" at commercial stop or start. This is without any changes in tuning. It would seem that fade to white would be treated about the same as fade to black in the detection process as they appear to be essentially inverse processes. Not sure how comskip addresses this. This is USA cable channels.
It is possible to recognize the fade-to-white by tuning the "non uniformity" setting.
Did you review the uniformity of a white frame in the debugwindow to check you setting? Scan frame by frame through the transition and record the lowest value of the non-uniformity (U=) Double that value for the ini file.
But also do have a look at the log file. Just above the black frame list.
It is possible that the uniformity clues are disabled because of a too strict validation rule, this is not yet tunable.
Did you review the uniformity of a white frame in the debugwindow to check you setting? Scan frame by frame through the transition and record the lowest value of the non-uniformity (U=) Double that value for the ini file.
But also do have a look at the log file. Just above the black frame list.
It is possible that the uniformity clues are disabled because of a too strict validation rule, this is not yet tunable.
I have tried the .mls output many times and it appears to work OK with the exception of the line "Count=<number>" The number is not correct for the number of cut points. If this is changed to the number of entries in the list, it works fine. The number of entries in the list does match the .wme and .VPrj files.
I have not had much luck with the "fade to white" consideration I wrote about in a previous post.
Per your comment about too strict a limitation that may be noted in the log file just above the black frame list, the following lines appear:
Confidence of SceneChange cutting: 0 out of 0 are strict
Confidence of Uniform cutting: 15 out of 32 are strict
Confidence of Volume cutting: 0 out of 0 are strict
Black Frame List
I am going to try Comskip on some other program material to see how it works with other channels. (A few I have tried, it seems to do well) I need to see if the one I have had trouble with is just a special case.
Regards
I have not had much luck with the "fade to white" consideration I wrote about in a previous post.
Per your comment about too strict a limitation that may be noted in the log file just above the black frame list, the following lines appear:
Confidence of SceneChange cutting: 0 out of 0 are strict
Confidence of Uniform cutting: 15 out of 32 are strict
Confidence of Volume cutting: 0 out of 0 are strict
Black Frame List
I am going to try Comskip on some other program material to see how it works with other channels. (A few I have tried, it seems to do well) I need to see if the one I have had trouble with is just a special case.
Regards