Chapter Grabber extract chapters from DVD, BD, HDDVD and add chapter names

[Emp3r0r]

You need an API key to load chapters from the chapterdb.org website. If you want to do that then create one of those accounts.

If you have an oauth2 account provider, let me know and I can attempt to add it.

[kevbo]

I've been using ChapterGrabber for a while now on my NTSC DVDs, and I've found that the chapter marks it calculated seemed to be a bit off, with compounding inaccuracies the longer the program. (By "inaccurate", I mean that, upon creating a MKV using the times ChapterGrabber creates, and then moving through the chapter stops, they don't line up with the chapter stops in the original DVD, moreso the farther through the program you get.)

But other programs have been able to get accurate chapter extractions from the same .ifo files. Like, for example, MakeMKV. I don't want to have to use MakeMKV if I can help it, as I already have software that does almost everything I need. So I stared at this for a bit.

I made a discovery today. If I load an .ifo file into ChapterGrabber (5.4.0.0, hey, just noticed that the About box still says Copyright 2003-2011), it grabs chapters. The little tag in the status bar says "29.970fps". If I change that to "30.000fps" (or choose Edit/Current FPS/30.000), and then choose "Edit/Convert Chapter Times by FPS/to 29.970fps", I then get the correct times. They match what MakeMKV generated, and when I remux the file using the new output, I get exactly the chapter stops I wanted.

Which is interesting.

Basically, tell ChapterGrabber that the numbers it calculated were for 30.000fps and shift them back to 29.970fps.

This is great, because now I can finally get the chapter stops I wanted. But I'm curious how/why this is happening in the first place.

This can't be happening to everyone, or things like the ChaptersDb.org integration wouldn't work: every NTSC DVD would be off by a little bit.

Anyways, I'm just glad I figured a way out, and wanted to post with the hope that maybe this might help someone else.

Thanks.

[LeMoi]

I noticed the same problem with ChapterXtractor, but it didn't use to happen with chapters created by DVD Decrypter... Didn't know the solution would have been what you said !

[Emp3r0r]

Can someone else confirm?

Can you also confirm that most dvd chapters on chaptersdb.org is wrong as well?

Sounds like it'll be possible to mass fix if needed.

:thanks:

[osgZach]

ChapterGrabber uploads to database when you save a file (unless disabled in menu).

I'm looking for moderators to help clean up the database.

Then we can accept direct uploads to
http://chapterdb.org

Sorry, this turned into a rather long post. I just wanted to share my thoughts on the subject as this is a great resource I will be using now that I know about it.

I'm all for the general idea of useful Public Databases and such. But given the issues than can occur during importing and naming of chapters (timecode issues with FPS etc), or just general new-user tomfoolery as they learn the program; I think automatically uploading based on a file save is more trouble than it is worth. It would cost a lot of time and effort to scrub the site, for thousands of entries, especially if your user-base starts growing. Differences in naming standards and all that also create many duplicate entries, etc. I think it would be better to lock DB access, and take on approved users. Establish clear quality-control issues, etc. Naming standards, standard data fields that can be used to organize/sort what is displayed (both on the site and in GUI)

I don't know how you currently store data, but I'm assuming you use some sort of hash based on the Disc ID, or File properties of a given file? I've noticed when I open a Disc I get the correct FPS value automatically, but if I open say.. a Chapters.txt file generated by any number of utilities (they all appear the same) while I do sometimes get the correct title auto-filled into the search box, I do not get the correct FPS value. It seems to default to 30fps regardless, which is a hassle to remember/change if you are using a chapters.txt file extracted from a Blu Ray or other 23.976FPS source. Overall, I think having a given number of staff who are trustworthy, and the only ones who can update/modify the database (and requiring submissions to be generated from Discs) would be a far better and more reliable route to go.

You could still have a general submissions system, where users could choose to manually submit a chapters file, but instead of going to the main database they enter a moderation pool, where Moderators can review the entries manually for quality, etc. and correct them as needed. This may slow down the "Growth" of the main database, but you will reap benefits in quality and reputation. Separating the BD and DVD entries may also be a good idea, if its not done already. I'm assuming you at least have data fields for them based on what I saw on the web site, so giving the user the option to show only DVD, BD, or Both would be a nice option.

Similarly, having a rating system in place will help with quality as well. If there is an entry that is inaccurate, it is a lot easier to identify it with a low rating, and doubly so with a controlled database where the is only 1 entry for the film per media type (DVD/BD). If you insist on keeping an open database that is moderated after the fact, then I at least think you should develop a more intricate ID hashing system. You could take your submission and on the back end, convert it to all the file formats you support exporting to, then hash all of those so that you have a known quantity in the DB already and duplicates could be rejected that way. Then if I user opens up one of those files because its blank or whatever, it can easily identify what file it is and pull the info from the correct title.

(edit)
One final thing above all else. I think if you take away one suggestion from this post and implement it, that should be this... Add a "Verified" tag to the database. When flagged Verified it means that someone has physically put a disc into a player and checked the Chapter titles (and times if possible), or that those titles and times have come from another verifiable legitimate source. This should be a staffed position and privileges only given to individuals who have proven trustworthy. After an item is approved a general "confirm" accuracy option could be exposed, allows other users to endorse the original verification, stating "yep I checked too and this is accurate". That would go a long way to building a reliable and trustworthy DB that the public can start to adopt.

In closing:
Look, I am not ignorant of the fact a project like this probably takes a lot of work. I'm not trying to insult anyone or anything, just sharing my opinion on how I might do some things. If there is anything legitimately not thought of, then by all means take my advice if you want it. If you don't want it then that's fine too. I could be wrong about everything, after all.

[real.finder]

when I run it in my PC with windows 7 works normally

But in the server with Windows Server 2008 R2 get this directly

http://i.imgur.com/mjsuC3h.png


the details:
Description:
Stopped working

Files that help describe the problem:
C:\Users\RealFinder\AppData\Local\Temp\2\WEREB22.tmp.WERInternalMetadata.xml
C:\Users\RealFinder\AppData\Local\Temp\2\WER170.tmp.mdmp

Read our privacy statement online:
http://go.microsoft.com/fwlink/?linkid=104288&clcid=0x0409

If the online privacy statement is not available, please read our privacy statement offline:
C:\Windows\system32\en-US\erofflps.txt

edit: solved by doing this http://blogs.msdn.com/b/vijaysk/archive/2009/08/16/you-must-use-the-role-management-tool-to-install-or-configure-microsoft-net-framework-3-5.aspx

[Emp3r0r]

...

Thanks for your feedback. Here is the backlog:


Improve submission system to enforce quality data
Create staging area for submissions
Pre-generate hash values based on time codes (your-idea)
Improve voting, flagging, and moderation system
Improve default fps/duration values for chapter text files
Create and document chapterdb.org API
Document naming and quality standards

[isamu.dragon]

any plans to have this program to export the chapters as a .cvs file?

[Cheesegeezer]

Emp3r0r thanks for creating this amazing database.

I have signed up for an API key, however I can't find any API documentation on how to use the API or how to access it.

If you could point me in the right direction, it would be greatly appreciated.

Many thanks for your hard work

Cheese!!

[meRobs]

Can someone else confirm?

Can you also confirm that most dvd chapters on chaptersdb.org is wrong as well? ......

Yes, I can confirm that such time code errors occur.
I have been using ChapterGrabber to import/mux chapter points into MKV files.

I edit material in Adobe Premiere where I also insert chapter points. This is then saved to MP4 and also to Adobe Encore, an authoring App. I use the latter to create a DVD folder and thus IFO files with chapter values. For example, in one NTSC project with a duration of 32:10 min, the last three chapters, as seen in both Premiere and Encore were:
24:55:22 –– 27:32:12 –– 31:29:18 min.

With the Encore-created VTS_01_0.IFO file loaded into Chapter grabber, it gives the last three chapters at:
24:54.274 –– 27:30.741 –– 31:27.741 min.

I created a chapter TXT file from ChapterGrabber; and then imported it into MKVmerge with the above MP4 and muxed to an MKV.
When played in VLC, the chapters appear at
24:54 –– 27:31 –– 31:28 min,
which agree with ChapterGrabber but are progressively getting early (see correct values by Encore). This was first noted in this project because the chapters were loctaed in Premiere at scene changes.

I have not yet tried the above workaround, by kevbo, of tricking the App into thinking it had got the frame rate wrong, etc. Shall report back.

[meRobs]

I have just tried the workaround.
I loaded the VTS_01_0.IFO, which gave the 20th chapter for the 29.97 fps NTSC file as 31:27.741.
I clicked Edit > Current FPS & selected 30.00
Then, Edit > Convert Chapter Times by FPS & selected 29.970
This immediately changed the 20th chapter to 31:29:18.

This is correct since the value set in Premiere was 31:29:18 exactly. When muxed into the MP4 to give an MKV in MKVmerge, the result played perfectly in VLC!
The chapter jumps occurred exactly at the scene changes!

This is not surprising since 30.00 is exactly 0.1% greater that 29.97, which is equivalent to 3.6 sec in every hour.
Or ~2 sec in my 32 min clip -- a very noticeable error when a chapter jump arrives before a known scene change.

This relatively small error may easily go unnoticed in most cases unless the user is aware that the chapters should occur at particular spots such as at scene changes.

[meRobs]

I thought I had best check to see whether a similar bug occurs for PAL.

The short answer is no!
This is fortunate because there is no other framerate value close to 25 fps listed in ChapterGrabber.

Hence, the problem is limited to NTSC.

[szabi]

Hi

Last time I used this program it worked fine, click (http://forum.doom9.org/showthread.php?p=1561559#post1561559).
But now I got this result:
http://abload.de/img/ch_grabawbk7.gif
Not counting frames to times but writing it as name of chapters.
What happened?

bye
szabi

[Overdrive80]

Hi, if I save chapters (x264 QP file) as XXX.qpf, the output file hasnt structure as:

0 K -1
16364 K -1
29152 K -1
47625 K -1
65588 K -1
79564 K -1
89374 K -1
99771 K -1

if not:

<?xml version="1.0" encoding="utf-8"?>
<chapterInfo xml:lang="spa" version="3" extractor="ChapterGrabber 5.4" xmlns="http://jvance.com/2008/ChapterGrabber">
<title>VTS_01_0</title>
<source>
<name>E:\DB Movies\DSSD10378\VIDEO_TS\VTS_01_0.IFO</name>
<type>DVD</type>
<hash>828f48e16f83ae8f0fef17812a23cec8</hash>
<fps>23.976023976023978</fps>
<duration>01:19:31.8360000</duration>
</source>
<chapters>
<chapter time="00:00:00" name="Chapter 1" />
<chapter time="00:11:22.5340000" name="Chapter 2" />
<chapter time="00:20:15.8680000" name="Chapter 3" />
<chapter time="00:33:06.3680000" name="Chapter 4" />
<chapter time="00:45:35.5680000" name="Chapter 5" />
<chapter time="00:55:18.5020000" name="Chapter 6" />
<chapter time="01:02:07.6350000" name="Chapter 7" />
<chapter time="01:09:21.3020000" name="Chapter 8" />
</chapters>
</chapterInfo>

[lil vito]

Hi Jarrett,

I had an old version of ChapterGrabber, and now am happy to see that you have continued to update it, so I just downloaded and tried the new version. Thank you for the handy app.

A couple suggestions:

I saw a post on your blog, that import from web has been disabled, because Amazon no longer includes chapter names.

However, Barnes and Noble.com (www.bn.com) does include chapter names in its DVD descriptions, so I think you should re-enable search from web, and use B&N instead of Amazon.

Also--the program I find that I use most these days for re-encoding video to watch on portable devices is the freeware app Handbrake (www.handbrake.fr), which encodes with h.264, etc.

When one opens a DVD with Handbrake, the chapter markers are shown in the interface. Therefore, as HB detects the chapter markers itself, it doesn't need the chapter point detection aspect of ChapterGrabber.

However--HB just labels the chapters Chapter 1, Chapter 2, etc. To add actual chapter names, one has to do that manually, one at a time.

However, after manually setting the chapters, one can now save the chapters to a .csv file, and one can also open a previously saved .csv file.

You now have several options of saving the ChapterGrabber chapter file. I would suggest that you add one more Save option, to save to .csv. That would be very nice, if CG could save to a .csv file that could be opened in HB, with all the correct chapter titles.

Could you add "Save to .CSV"? (I would suggest in doing that, to test the resulting .csv with Handbrake, to make sure it works with how HB uses .csv.)

(Until you are able to include that in a new version, does anyone know of a workaround, a way to quickly convert one of the formats that CG saves to into a .csv readable by Handbrake?)

Thanks again!

Understand that HandBrake will not allow you to add chapters, if they are not already present in the source file.
The only use for the import button on the chapters dialog, is to allow for importing Chapter Names in the instance you have chapters there already.
That being said, I dont like having generic "Chapter 1, Chapter 2" style names, and so made a quick Powershell script to convert named chapters from ChapterGrabber's native "*.Chapters", into the csv format required by HandBrake for naming of chapters.

As the script is a powershell script, you may need to right-click and go to properties to "Unblock" it, as it will have obviously come from the internet if you download it here. (Optionally, just copy and paste the content into notepad and save as "Convert-ChaptersToHandbrakeCSV.ps1")

You will also need to make sure powershell execution policy is set to 'Unrestricted', or if you pasted content into notepad and saved, then 'RemoteSigned' will work fine.
Set this via:
Set-ExecutionPolicy Unrestricted
or:
Set-ExecutionPolicy RemoteSigned

Once this is done, you can either Run the script via Right-Click > Run with powershell
or
Run it from a powershell prompt.

Get-Help .\Convert-ChaptersToHandbrakeCSV.ps1 while in the folder where you saved the script for help.

Get-Help .\Convert-ChaptersToHandbrakeCSV.ps1 -Examples for examples.

Hope this helps someone...

[jmone]

Hi Emp3r0r, Thanks for your work on this!

As background, JRiver Media Center lets you create "particles" from BD (and DVD) that is basically a playback range. Myself and another JRiver Media Center user (MrC) have been working on how to create these particles for Blu-ray Music Videos and populate them with the relevant chapter information so it presents and plays like a CD rip.

At present, his script calculates the MD5 hash (or optionally uses a ChapterDB Disk ID) to query the ChapterDB in the creation of the required number of tracks/particles + adding the names and playback ranges in JRiver Media Center. It is all working really well, but a couple of Q's as we work to finalise the workflow.

ChapterDB has lots of duplicate entries of which some of which are not complete. Does anyone know if there is there a way of:
- prioritising the complete entries to be returned by an MD5 Hash search instead of incomplete ones
- editing/updating/combining existing entries to complete them (I tried updating one such entry by adding the missing chapter names in ChapterGrabber but it did not seem to make a change on the DB).

Thanks
Nathan

PS - Is this the correct forum/thread for ChatperDB?

[jmone]

Hi, is there any plans to add simple command line support?

Examples: If E: contains a bluray
chaptergrabber E:\ -3 D:\celltimes.txt -celltimes - chooses the third playlist of E:\ no switch after disk input would choose longest runtime playlist and outputs D:\celltimes.txt in celltimes format
chaptergrabber D:\playlist.mpls D:\chapters.xml -mkvxml - outputs D:\chapters.xml in mkv format
other switches could be: -mkv (chapter) -cg (chaptergrabber) -timecodes -tsmuxer -x264

This would allow frontends to make use of this tool.
.

+1, this would make our task easer to integrated as well.

[jmone]

FYI - For JRiver Media Center users, we have just posted a guide on managing BDs (mostly Music BD) by chapter using data pulled from ChapterDB - http://yabb.jriver.com/interact/index.php?topic=89627.0

[Aray]

Hi, Emp3r0r

I have never successfully used ChapterGrabber.

When I open a chapter file, the Title textbox will show a tip like "Database syncing...", Screenshot like this

https://i.postimg.cc/nrSvYjpn/Chapter_Grabber_Ver_5.4.0.0_Crash.jpg

And then the form will disappear without any message. How to fix it?


My environment:
Windows 10
ChapterGrabber ver 5.4.0.0

[hubblec4]

Hi Aray and welcome to Doom9 forum.

ChapterGrabber is not longer supported.
You could try my chapterEditor.

[Aray]

Hi Aray and welcome to Doom9 forum.

ChapterGrabber is not longer supported.
You could try my chapterEditor.

Thanks for your help and the Powerful ChapterEditor software. It's very easy to get started, Although the UI doesn‘t look so delicate.

[Flanker-B]

how to use this without the api key now that chapterdb is dead!?