Making decent rip from PAL-BD - frame rate conversion

[Hgp3]

Hello,
I have problems making a decent "rip" of a PAL-Bluray I own.

The story so far:
Last weekend I decided to rip a TV series I bought on Bluray. Ending up with ~72GB of Data after riping with MakeMKV I wanted to have a smaller rip because it seems to me like unnecessary wasting of space. This is where my problems began. Because I don't know much about encoding I wanted to use a well-made encode from the web. At this point I noticed that my BD was in PAL, so it had 25 fps. (I believe it is because it was first released DVD-only in 2012 and got the Bluray-Release last year.) Digging a bit deeper in my source, I recognized it as a PAL speedup (wuh, what a surprise) with resampled, not pitch corrected audio.

I searched for the problems that I now have but couldn't find satisfying answers. (maybe using wrong search terms)

What I'm looking to do now is:
-obtain a well encoded release at original 23.976 fps and using video and original audio of it (because resampling of the lossy audio I have back to original seems pretty dumb for me)
-do a framerate conversion of the audio in my language, which leaves me with two not really good options if I got them right:
-resample without changing pitch (would results in correct playback of the great soundtrack, noises etc. but give me deeper syncronisation - meh)
-resample and adjust pitch (soundtrack, noises etc. would still sound wrong but syncronisation is played correctly; also audio would be even more fucked up because of the filtering done to it - more meh)
-get the PGS-Subs of the Bluray playing at the speed of the NTSC version (didn't find much about that topic)
I actually care more about the subtitles, but it would be a very-nice-to-have if I had the the syncronization in my rip too.

Why I dont want to stay with a PAL rip:
-I would have to encode it myself and end up with worse video quality using standard options and no filtering (even when filtering bluray isn't making that much difference afaik)
-the wonderful soundtrack would still sound like crap and the original dub is distorted
-and as a third little reason, I want to learn something by getting this done

What I want from you:
Please tell me if this is my way to go or if I'm doing it completely wrong. Help me deciding what to do with the audio and point me to software capable of doing what I'm wanting to. I'm familar with the main functionalities of MKVToolNix GUI and did some very light avisynth magic.

Thanks in advance!

[Sharc]

If you want to reduce the video file size you have to reencode. You cannot "rip" to a lower file size.

Is your 25fps video true interlaced or progressive (fake interlaced)?
Why do you want to change the framerate? Is it just for correcting the audio pitch?

I suggest that you upload a sample of your .mkv of few seconds duration.

You could try BD-Rebuilder for re-encoding of your .mkv. It probably comes close to what you are looking for.

[Hgp3]

Is your 25fps video true interlaced or progressive (fake interlaced)?

I don't really know what you mean with true or fake interlaced. I would guess it is progressive being animated and not filmed and not showing the well known interlacing artifacts. But please take a look in the MediaInfo: Video
ID : 1
Format : AVC
Format/Info : Advanced Video Codec
Format profile : High@L4.1
Format settings, CABAC : Yes
Format settings, ReFrames : 4 frames
Codec ID : V_MPEG4/ISO/AVC
Duration : 5s 840ms
Bit rate mode : Variable
Bit rate : 30.8 Mbps
Maximum bit rate : 38.0 Mbps
Width : 1 920 pixels
Height : 1 080 pixels
Display aspect ratio : 16:9
Frame rate mode : Constant
Frame rate : 25.000 fps
Color space : YUV
Chroma subsampling : 4:2:0
Bit depth : 8 bits
Scan type : MBAFF
Scan order : Top Field First
Bits/(Pixel*Frame) : 0.593
Stream size : 21.4 MiB (91%)
Language : English
Default : No
Forced : No
SOURCE_ID : 001011
_STATISTICS_WRITING_APP : mkvmerge v7.6.0 ('Garden of Dreams') 64bit built on Feb 8 2015 13:04:34
_STATISTICS_WRITING_DATE_UTC : 2016-09-02 22:00:32
_STATISTICS_TAGS : BPS DURATION NUMBER_OF_FRAMES NUMBER_OF_BYTES
DURATION : 00:00:05.760000000
NUMBER_OF_FRAMES : 144
NUMBER_OF_BYTES : 22962690

Why do you want to change the framerate? Is it just for correcting the audio pitch?

It seems I wasn't clear enough, sorry. It is not only about the audio but connected with it. I don't have any deeper knowledge about encoding and this is why I want to take an already well encoded version from someone else (video and original audio) and put it together with my edited audio and subtitles. The original (and every reencode made from it) has 23,976 fps, so I would get the whole thing with ideal conditions I think. (Good encode, original audio untouched, second audio matches after edit, subtitles match after edit)

If I only edit the pitch of my audio files, I would have to perform an unnecessary reencode of the original audio (lower quality) and I would have to encode the video myself (lower quality/bigger file). It wouldn't greatly effect my second audio that I definitly have to edit and I could leave the subtitles untouched.

I would definitly want to go with the first option, but you are free to convince me that the second one is the right for me.

I suggest that you upload a sample of your .mkv of few seconds duration.

Here it is: sample (https://my.mixtape.moe/jlskne.mkv)

[Asmodian]

The amount of labor required to get the audio from a PAL release to match the video from a NTSC release is NOT worth it, you will constantly be finding another episode where the starting point was a few frames different between the releases, even if you manage to get the durations to usually match by slowing down your audio. Re-timing the subtitles is even worse. :(

[Sharc]

Agree.
If at all, it is probably safer to start from the all-in-sync 25fps blu-ray version, reencode the progressive video to smaller file size, slowing it down to 23.976 (assumefps(23.976)) and slow the audio down (25 => 23.976 without applying pitch correction) in order to restore the "original" 23.976 pitch and match the video duration .....
Subtitles could possibly be adjusted/restamped using BDSup2Sub.

Probably not worth the pain. Good luck .....

[Q-the-STORM]

Why do you "have" to encode video if you stay with the PAL rip? you can just change framerate with mkvtoolnix and keep original video quality...
then sync and mux any audio/subs you want...

[Sharc]

Why do you "have" to encode video if you stay with the PAL rip? you can just change framerate with mkvtoolnix and keep original video quality...
then sync and mux any audio/subs you want...
I understood that the OP wants to reduce the file size also (first post) ........

Edit:
And when just changing the framerate from 25 to 23.976fps with mkvtoolnix the progressive video stream will still remain flagged as interlaced which MAY cause playback issues with certain players. At least it is not blu-ray compliant.

[huhn]

assumeFPS(24000, 1001, sync_audio=true)

maybe a program with better resampling should be used but that's it.

[Hgp3]

Hey Guys,
thanks to your help I made some progress. So far I got a "testing version" of the first episode as I wanted it to be.

The amount of labor required to get the audio from a PAL release to match the video from a NTSC release is NOT worth it, you will constantly be finding another episode where the starting point was a few frames different between the releases, even if you manage to get the durations to usually match by slowing down your audio.
Surprisingly as it seems now I am very lucky. The release i got is constantly 4 frames earlier (I made a dirty stackvertical avisynth script and skipped through some scenes of episode 01 to prove this.) and I don't have big problems matching my stuff to it.

Subtitles could possibly be adjusted/restamped using BDSup2Sub. Thank you for leading me to this, adjusting the subs was really easy with the conversion options. Only thing is that I experienced a bug with the delay option. Every time I "Ok" this, it lowers the delay by one frame (in my case from -166.833 to -125.122).


The audio was a bit complicated and I'm not satisfied with the procedure. I used audacity (and it was bitchy ...) and used the change speed filter. Then I cut the first ~167ms away and exported to flac (for testing, I don't want to stay with 330MB of audio). After telling audacity to not convert the 5.1 track into mono it came up with channel mapping. Using this (https://ubuntuforums.org/showthread.php?t=1573976&p=9841781#post9841781)information, I mapped the channels this way (http://i.imgur.com/7G8jvCa.png). Is that accurate?

For compressing the audio I want to use opus (Mapping 5.1 surround: front left, front center, front right, rear left, rear right, LFE source (https://tools.ietf.org/html/rfc7845#section-5.1.1)) for being as good as every other codec at high bitrates and being open source. Any thoughts of you or tips how to encode are welcome.

maybe a program with better resampling should be used but that's it.
Do you know a software that is better and/or easier to use than audacity?

tl;dr
made a test with first episode and got everything right, still unsatisfied about audio editing and looking for correct audio encoding

[huhn]

eac3to of cause. http://forum.doom9.org/showthread.php?t=125966

it is a command line program i personally have no problem with that.

it can do what you are looking for.

-23.976/... define source fps to be "23.976", "24.000", "25.000", ...
-changeTo24.000 change source fps to "23.976", "24.000", "25.000", ...

[Hgp3]

You are right, eac3to is great. But I skipped in my file and compared the audio and I found out that I mapped the channels wrong. The example I picked has a centred voice in the original DTS, my audacity flac has it on the right ear, the output from eac3to on the left. (I'm using stereo headphones)

command line: eac3to "input.dts" "output.flac" -25.000 -slowdown -167ms -2,0,1,4,5,3
audacity mapping see previous post
(unimportant fun fact: the output from eac3to is with 660MB twice the size of audacity output, both flac)

Do you recognize where I failed at mapping?

And, as eac3to doesn't support opus output, should I simply go with AAC after my tests are finished?

[huhn]

sorry i don't know.