Feature requests for TS cutting

Cela

Member
You can get close to binary searching using the thumbnail bar....
...Things can get very complex with joining and DVD processing when you have files with different audio codecs and multiple audio streams.
Yes, it must be frustrating, sometimes, that users only see their goals and forget the tediuos work needed to implement the details. The implementations are probably never as simple as the underlying ideas!

@phd, MrVideo, DanR
Thanks to your help and some additional e-mails from 'support@VideoReDo.com' (MrVideo in post #3 opened my eyes!) i had a phantastic quick start with VideoReDo. Since my post #15 I use it as first choice TS cutter for my current hobby projects. :)

Time code "HH:MM:SS.hh" (hundredth of seconds) works in my version, though I still would prefer "HH:MM:SS.mss"; not so much because of higher precision, I would be happy with an added digit "0", but because my "named chapters" notepad text file needs that format. I always forget some of the otherwise necessary manual corrections! I do not care what is under the hood.... The ".mss" format would help to eliminate sources of manual interaction errors.

I also prefer "copy&paste" (Ctrl+C & Ctrl+V) over manual copying of displayed numbers. Again I tend to get too many typos! I sent ideas to support, asking to make the current timecode display only field an edit field and to et inspired by VirtualDub's way of Ctrl+T edit field presentation. This would allow copy & paste and selecting full F# (what's internally used under the hood, I guess) and an external format ("HH:MM:SS.mss" as my choice).

What me also brought forward was making a personal copy of the Keyboard Shortcuts! Doing this I discovered that I can turn off (toggle) the Thumbnail windos bar" by F11. Toggle to OFF boosts slider searching performance! Do you agree?

DanR, having read your advice, I will give the thumbnails another chance. ;)

Regards,

Cela
 

Attachments

Cela

Member
You can get close to binary searching using the thumbnail bar. Set the thumbnail to something course, like 1 minute. Then when you find an area you are interested in, double click on that thumbnail. This frame will become the center frame and the interval will drop to the next lower value. For example, 1 minute will drop to 30 seconds. You can set the intervals on the Thumbnail parameter screen...
Comes close? That practically IS what I call 'binary search/refinement'! :)

Simple steps:
Repeat for each scene {

  • F11: Toggle OFF thumbnail bar
    Large slider: move quickly through very large sections (long strikes to get near point of interest)
    F11: Toggle ON thumbnail bar
    Ctrl+9: set 'large' section (60 seconds) for closing in
    Double click thumbnail: center focus and drop to the next lower value (60 -> 30 sec, etc.)
    ...
}

Just done! ~8GB ts in ~2 min, frame-accurate! :D

Edit:
Two more questions:
1. How to get quickly scenes from these chapters?
<chp_start><TAB>#<TAB><chp_text><TAB><chp_thumb_tc>
Code:
00:00:00.000 	1	 Eurovision Song Contest 2010| 29.05.2010 Oslo\Introduction	
00:02:13.260 	2	 2010 to Oslo/Norway	00:02:33.420	01:58:23.720	01:58:23.197 Poster
00:02:50.140 	3	 Share The Moment	
00:03:32.360	4	 Fairy Tale | Alexander Rybak	
00:05:33.380	5	 #01 Azerbaijan	
00:06:10.220 	6	 #01 Drip Drop | Safura 	00:07:11.480
00:09:26.140	7	 #02 Algo Pequeñito (Something Tiny)| Daniel Diges, Spain	
00:12:43.380	8	 #03 My Heart is Yours | Norway	
00:16:11.240	9	 #04 Run Away | Moldova
...
Legend:
<TAB>...horizontal TAB character \u0009
| ... in chp_text \n (new line)

2. How to Save each single scene of the resulting 30+ scenes as a separate mp4? (I already sucessfuly used the mp4 profile.)
 
Last edited:

Cela

Member
......prefer "copy&paste" (Ctrl+C & Ctrl+V) over manual copying of displayed numbers. Again I tend to get too many typos! I sent ideas to support, asking to make the current timecode display only field an edit field and to get inspired by VirtualDub's way of Ctrl+T edit field presentation. This would allow copy & paste and selecting full F# (what's internally used under the hood, I guess) and an external format ("HH:MM:SS.mss" as my choice). ...
I did not think possible implementation could be as simple as this! See attachment, item 4! :)

Hope it will not take long!

( MrVideo may know what I mean ;) )

P.S. Tags of first 13 frames - what is this? where does it come from?
P.P.S. View > Show Frame Number' needs to get a keybourd shortcut!
 

Attachments

Last edited:

phd

Super Moderator
Quick and dirty way to import scene markers if you have timecode.

Setup VRD to import chapter files.

Create a text file with the same name as your video.

30 sec.mpg
30 sec_Chapters.txt

Copy and paste the timecodes one per line
e.g. - 00:00:10:00

When you open the file the timecodes will appear as scene markers.
 

Cela

Member
@phd
Thanks a lot! This is what I need most for my next project step! :)

@all
But then, to complete the project, i would need to learn:
How can I ask VideoRedo not to append/join the scenes into one output file but to output each scene into its own output file?
 

Dan203

Senior Developer
Staff member
How can I ask VideoRedo not to append/join the scenes into one output file but to output each scene into its own output file?
Use the batch manager instead of the joiner. After you mark each scene use File->Queue to batch. When you're done marking all the scenes simply run the batch manager and it will produce a different file for each scene.

Dan
 

Cela

Member
@Dan203
Thanks, all clear now for actual project to output my scenes as different mp4 files for ipod. :)

@all
Scene List, again:
How to delete a scene from the Scene List?
How to reorder scenes in the Scene List?
(often needed for camcorder footage!)

Next, #21 & #23, please. ;)
 

phd

Super Moderator
select a scene on the scene list and click remove selected scenes.

Here is a procedure to rearrange segments of your video:

You may find this process easier in Scene Mode where you only select the segment you wish to output. Tools>Options>General Parameters and select "Scene Mode" for Editing Mode.

Mark the first segment you would like to add to the rearranged timeline. This segment will be green with the rest of the timeline red. Add to the Joiner. Joiner>Add Current Project to Joiner List

Clear the cutlist and select the next segment of the file you wish to add to the file. This new segment will be green with the rest of the timeline red. Add to the Joiner. Joiner>Add Current Project to Joiner List

Continue to add segments of the timeline. Clear the cut list each time and indicating the next sequential segment.

Output your file from the Joiner. Joiner>Create Video From Joiner List.
 

MACKerMD

New member
Match audio levels? Huh? There is nothing you can do to the audio levels, so I'm not sure what you are trying say here.

As for giving a second after and before, I try to be as tight as possible. In many network cases, there may only be a frame or two between the program and a commercial.

As for the audio helping you. My major complaint, which you can find in other threads, is the method by which they are displaying the audio. It is if there is an autogain control on it. You can be in an area that in the long black of a commercial break (which you won't get with an OTA capture), where there is enough low level noise that VRD slams the audio line against the top of the audio display. I won't rehash it here, other than to say that I've submitted samples of what it looks like and why it is extremely difficult at times to work with. I want a real waveform display, like what is displayed in all other professional NLE systems.

Because of the way the audio is displayed, I think I end up with more black before a commercial break than there should be.
I might be wrong, but my guess is, that C. would like to 'see' a waveform of the audio, so its visually editable as well. Then the waveform shows where the audio is 'soft/low' and the edit/cut-point can be made there.
 

Cela

Member
Unexpected problem

Quick and dirty way to import scene markers if you have timecode.

Setup VRD to import chapter files.

Create a text file with the same name as your video.

30 sec.mpg
30 sec_Chapters.txt

Copy and paste the timecodes one per line
e.g. - 00:00:10:00

When you open the file the timecodes will appear as scene markers.
Sorry, I could not make it?
I did the following, but it did not work for me! :(

>Setup VRD to import chapter files.
How?
I did: >Tools>Options>Chapter marks>hh:mm:ss.ff>Import chapter marks on file open
Then: Tools>Shift+Options>36. Display hundreths of seconds in timecode... :False -> display hh:mm:ss.ff
I renamed my ts to the simple name EVSC.ts
I transformed my hh:mm:ss.mss (milliseconds) time code to 50fps hh:mm:ss.ff timecode and made a file EVCS.txt starting with these entries:
Code:
00:00:00.00
00:04:26.13
00:05:40.07
00:07:04.18
00:11:06.19
00:12:20.11
When I open the file by >File>Open Video... EVSC.ts
the file EVSC.ts gets opened, but
the timecodes from file EVSC.txt do NOT appear as scene markers. NO scene markers!

What do I wrong?
 

phd

Super Moderator
EVSC.txt - wrong name

Try:
EVSC_Chapters.txt

Look at the example names I previously provided.
30 sec.mpg
30 sec_Chapters.txt

I just tried your data with the following file names and it worked fine.
h 1029.mpg
h 1029_Chapters.txt

Make sure the txt file has "_Chapters.txt"
 

Cela

Member
EVSC.txt - wrong name
...Make sure the txt file has "_Chapters.txt"...
Thanks! Details matter! :)

Now. with "_Chapters.txt", I get the "Chapters Umbrellas" in the Navigation Panel. How to get these Chapters/"Scenes" back into the Scene List (1-click, preferred) where they originally came from (by Save Chapter File...)?

I am used to more 'operation symmetry' (from other software which I enjoy to use since years): I/O ... Progress in cycles needs previous step's output for next step's input. Here its more a One Way road!

So you are gonna dump your other apps and make this your Top Model, right?...
You see, the Model still has a lot to learn to become Top M ;)
So do I!
 

Attachments

Last edited:

Cela

Member
I might be wrong, but my guess is, that C. would like to 'see' a waveform of the audio, so its visually editable as well. Then the waveform shows where the audio is 'soft/low' and the edit/cut-point can be made there.
Well, full audio editing (including X-fade) like in the big (expensive) programs would be cool, or simple audio features (i.e. FadeIn/FadeOut) like, since years, in small (and free) Mpeg2Schnitt (for PAL mpeg2 footage with mpa and ac3 audio) wouldn't be bad. Then you can freely choose the best 'looking' points. Now I try to choose the best 'sounding' points.

Even this limited scope needs a 'PreListen' (Play TO, Play From, Play Connection) which I would like to get here in a next update, soon.
Remember, what you see (and hear) in the original ts and what you hear in the outputted ts may or may not be the same! Therefore, some programs simulate the the real i/o by real i/o-processing of some seconds around the points of interest. So you can 'PreListen' - hear (and see) what you will get (after the real export)!

If you give me full audio editing - I'll take it! ;)
 
Last edited:

Cela

Member
Trying to follow your interesting advices with my 30+ scenes example, I got lost with another unexpected matter, which I call 'Connection flights (scenes) not possible' (see attached screenshot).

Please remember, the start point of these "Lost" sequels are the 'two simple questions' from Post #22:
1. How to get quickly scenes from these chapters?
2. How to 'Save as Video...' each single scene of the resulting 30+ scenes as a separate video file?

For this matter (and for my other open matters posted in this thread) my conclusion is:
Currently, there is no simple solution available!

Including missing chapters/scenes and timecode commodities, a lot of useful functions are still missing which are, since a long time, found in Mpeg2Schnitt for PAL mpeg2 footage. Once used to, who would like to miss them in a HD ts cutter?

In this thread (and in my e-mails) I provided a lot of input and ideas for improvements.

For the time being, and not to overstrain your patience here in the forum, I will give up and wait and see what the future will bring.

I will restrict myself to enjoying VRD TV Suite as basic frame-accurate h.264 TS cutter!
I am very pleased with its basic frame-accurate cutting feature.

For the rest, I am afraid, I will have to carry on with my old workflow tools and steps.

I am aware that, so far, I only explored "ts cutting territory" and there is far more exciting, useful unexplored "territory" in VideoReDo. But, for my pending h.264 footage, I have to start using what I am able to use instead of testing and contemplating about what I would expect to discover.

Thank you for your kind help and advice. :)

If you need more information, please feel free to ask! I will regularly check back here for news.

Regards,

Cela
 

Attachments

Well, full audio editing (including X-fade) like in the big (expensive) programs would be cool, or simple audio features (i.e. FadeIn/FadeOut) like, since years, in small (and free) Mpeg2Schnitt (for PAL mpeg2 footage with mpa and ac3 audio) wouldn't be bad. Then you can freely choose the best 'looking' points. Now I try to choose the best 'sounding' points.
Forget it, never gonna happen--especially with AC3 and probrably any HD audio. VRD is a video editing program, not an audio editing. If you really need to edit the audio then demux the audio, do your editing and then remux.

Even this limited scope needs a 'PreListen' (Play TO, Play From, Play Connection) which I would like to get here in a next update, soon.
Remember, what you see (and hear) in the original ts and what you hear in the outputted ts may or may not be the same! Therefore, some programs simulate the the real i/o by real i/o-processing of some seconds around the points of interest. So you can 'PreListen' - hear (and see) what you will get (after the real export)!

If you give me full audio editing - I'll take it! ;)
Really do not know what you are trying to accomplish, this prelisten thing. IMO, the current audio track is sufficient for VIDEO editing. If you want more accuracy then the VRD folks have a professional version they would like to sell you.

If you set the thumbnail display to 2 seconds you will ''see'' a nice audio display. But since most of us work at the frame level we see that display greatly zoomed-in or magnified--each frame becomes either part of a word or letter and/or whatever the music level is at that point in time and the display tends to look the same from frame to frame, again my opinion.

Mike
 
2. How to 'Save as Video...' each single scene of the resulting 30+ scenes as a separate video file?

For this matter (and for my other open matters posted in this thread) my conclusion is:
Currently, there is no simple solution available!

Including missing chapters/scenes and timecode commodities, a lot of useful functions are still missing which are, since a long time, found in Mpeg2Schnitt for PAL mpeg2 footage. Once used to, who would like to miss them in a HD ts cutter?

In this thread (and in my e-mails) I provided a lot of input and ideas for improvements.

For the time being, and not to overstrain your patience here in the forum, I will give up and wait and see what the future will bring.

I will restrict myself to enjoying VRD TV Suite as basic frame-accurate h.264 TS cutter!
I am very pleased with its basic frame-accurate cutting feature.
You are comparing apples to oranges. Your mpeg2schnitt is over 2 1/2 years old and no longer maintained. It may have been good then but now ... H264 editing is a totally new beast that VRD, I believe, is the only game in town. The VRD folks are and will be focusing on the core H264 functionality.

Mike
 

MrVideo

Active member
If you set the thumbnail display to 2 seconds you will ''see'' a nice audio display. But since most of us work at the frame level we see that display greatly zoomed-in or magnified--each frame becomes either part of a word or letter and/or whatever the music level is at that point in time and the display tends to look the same from frame to frame, again my opinion.
Frame level viewing is too close. I work at .1 seconds, which is 3 frames from thumbnail to thumbnail.

I do want waveform audio display. The current method, IMHO, is not good. I've tried to explain in various postings the issues that exist with the current audio display. Search the forum for those posted, complete with images.
 

Cela

Member
@Floobydust Posts #28, #36, #37
audio editing:
I did NOT request audio editing, though I'm used to FadeIn/FadeOUT from good old, according to you 'outdated' Mpeg2Schnitt!
MACKerMD, in Post #30 wrote "... my guess is, that C. would like to 'see' a waveform of the audio, so its visually editable as well".
I answered in Post #34: "If you give me full audio editing - I'll take it!" I did NOT request it!

Ergonomic and efficient human interface:
That is WHAT I ASK FOR!

If you "Really do not know what I am trying to accomplish...", feel free to do an example with mpeg2Schnitt, a 3 hour concert show, as I am doing right now with VDR. Extract some 30+ songs. Create individual clips. Create A DVD with chapter menu to access each clip in the long video. Transfer the long video and the individual clips to your iPhone (or equivalent pocket media player device I use COWON O2 using modified iPhone mp4 profile), etc. Then you will learn to know!

My recent market survey gave me the impression that VDR is not too far from what I am used to have with Mpeg2Schnitt (and other tools), and has a sound potential to become even better.
 

Cela

Member
@MrVideo
With your Post #38 you do not address me but Floobydust.
I agree with you. Wave form audio display would be a very userfriendly, useful feature.

As a quick first step, for next month's release, IMHO the current wave line should by made higher, e.g. twice its current hight.
Would this comply to your roadmap?

Again, if VDR will bring more, I will take more! ;)
 
Last edited:
Top Bottom