Converting TrueHD audio for importing into Cyberlink PowerDirector 10

[ShadowEdit]

I have a 2.98 GB big 6 channel TrueHD (.thd) file that I would like to import into Cyberlink PowerDirector 10. The result would probably be uploaded to YouTube. According to this the supported audio formats are WAV, MP3, WMA, M4A.
MP3 sucks, so that format is out of the picture right away. WAV is lossless, so that would be ideal. WMA could be lossless, if it's the Windows Media Audio Lossless codec, so that could work as well. I'm not sure if M4A or the lossy WMA codecs are better though.
I have managed to get the TrueHD file converted to M4A via eac3to and Nero AAC Codec. I have not managed to get it converted to WAV (xrecode II couldn't convert it and eac3to couldn't convert the whole file, if I remember correctly). I think I also had problems with getting TrueHD converted to WMA.

Sorry for this post being a bit confusing, what I'm wondering is:

1. Which format that TrueHD can be converted to and that can be imported into Cyberlink PowerDirector 10 would give the best results quality-wise and what software would I have to use to get it converted to that format?
2. Is it possible / worth it to convert TrueHD to another lossless format that can then be converted to WAV / Windows Media Audio Lossless? If so, what software should I use?
3. Will YouTube converting the audio to AAC-LC have any impact on things? Like, is it better to have the audio in a lossless format so that YouTube doesn't shittify things too much, or is it better to have it as M4A - which is basically AAC - and hope that YouTube will barely touch the audio due to it already being in the right codec?

I'm new to audio conversion, audio codecs and this forum, so sorry if something in this post seems stupid / doesn't make sense.

[Kerry56]

I have converted TrueHD to WAV in Audacity. This was a soundtrack from a Blu-ray movie, extracted using ClownBD and put into a .ts container. If you use Audacity, you need the ffmpeg library for Audacity installed also. You would also need to go to Edit-->Preferences-->Import/Export and put a mark in the circle to allow custom mix. This will let you export as a multi-channel file.

I do not know if Audacity will recognize a .thd file, even using the ffmpeg library because I haven't tried it.

If Youtube is going to re-encode, it is better to start with as good a copy as they will accept as input. Are there size limits on your upload?

[poisondeathray]

It's probably "cut off" in your eac3to or earlier conversion tests to WAV because of the 4GB WAV filesize limit. You would need to use Wave64, but it's probably not supported in Cyberlink

Another option is to wrap it into an AVI container, or some other container like a transport stream (audio in AVI , with no video), and it's likely Cyberlink will accept that

ffmpeg had a TrueHD decoder, and can do this

ffmpeg -i input.thd -vn -c:a pcm_s16le output.avi

If you need a ffmpeg GUI, try tencoder , or ffqueue. Set video codec to "none" , audio codec to "wav", container to "AVI"
https://www.videohelp.com/tools/TEncoder
https://www.videohelp.com/tools/FFQueue


Youtube won't pass thorugh the audio. You will take a quality hit either way when it re-encodes. But it's better to incur 1 lossy conversion vs. 2 lossy conversions. You should use 1080p or greater video when uploading to YT (Even if it's "blank" video), because YT will use higher audio bitrates when re-encoding

* although there are many recent changes at YT, including HTML5 , so using 1080p or higher video resolution when audio is the main focus might not be valid advice anymore. It used to be the higher the resolution video, the higher the audio bitrate allocated

[ShadowEdit]

Are there size limits on your upload?

I'm going to cut away almost everything until I'm left with a 37 second clip, so that shouldn't be a problem.

Another option is to wrap it into an AVI container, or some other container like a transport
stream (audio in AVI , with no video), and it's likely Cyberlink will accept that

I have the video track as AVC in an mkv file, by the way. Is it possible to have both a mkv track and an avi track in the same project? (I assume that avi doesn't support AVC since even mp4 doesn't support AVC.)

It's probably "cut off" in your eac3to or earlier conversion tests to WAV because of the 4GB WAV filesize limit. You would need to use Wave64, but it's probably not supported in Cyberlink

Yeah, I think it was cut off at 4 GB and if I remember correctly that 4 GB file was not importable into PowerDirector.

[poisondeathray]

Does Cyberlink support MKV container ? If it does, you can probably copy the video, convert the audio to WAV into a new MKV container


AVI container is problematic for AVC. Cyberlink probably won't support it . MP4 does support AVC, but uncompressed PCM WAV is poorly supported by open source muxers in MP4 container

The "best" container for your purposes is probably a transport stream , if the AVC video uses the "typical" 8bit encoding and "typical" settings. If it uses 10bit or "exotic" settings, Cyberlink will probably not support it. The 2nd "best" is probably MOV container

If you don't know what kind of AVC video you have, use mediainfo to identify
https://www.videohelp.com/tools/MediaInfo


This will copy the video, encode the audio to uncompressed pcm wav into a transport stream container. Most editing programs will accept it (if your AVC video uses "typical" settings)

ffmpeg -i input.mkv -c:v copy -c:a pcm_s16le output.m2ts

(or you can use one of the GUI's above, set video to "copy", audio to "wav" , container to transport stream)



EDIT:
http://www.cyberlink.com/support/product-faq-content.do?id=14951&prodId=4

It says MKV is supported, so you could use MKV container, if the video uses "typical" encoding

ffmpeg -i input.mkv -c:v copy -c:a pcm_s16le output.mkv

[ShadowEdit]

Neither MP4Box executed via cmd or YAMB could make an mp4 file out of the AVC video.

It's 8bit and Cyberlink can import the AVC track when it's inside an mkv file.

Could you rewrite that command to only apply to the audio track? I have already imported the video and audio tracks as separate tracks into PowerDirector and finished the editing one and a half month ago. However, it is possible to replace tracks in the project file and the editing will still be there. But since I have two separate tracks in the project file I am unsure if Cyberlink will be able to "apply" the editing to it the tracks again if it is imported as one track.

[poisondeathray]

Neither MP4Box executed via cmd or YAMB could make an mp4 file out of the AVC video.

Then something is wrong with your video, or commandline, or binary. "MP4" is a standardized container for AVC. Arguably it's was the first official container for AVC

It's 8bit and Cyberlink can import the AVC track when it's inside an mkv file.

Could you rewrite that command to only apply to the audio track? I have already imported the video and audio tracks as separate tracks into PowerDirector and finished the editing one and a half month ago. However, it is possible to replace tracks in the project file and the editing will still be there. But since I have two separate tracks in the project file I am unsure if Cyberlink will be able to "apply" the editing to it the tracks again if it is imported as one track.

What do you want the command to do? convert the audio so it's pcm wav in a MKV container only ? (ie. audio only mkv) ?

-vn will result in "no video" for ffmpeg

ffmpeg -i input.mkv -vn -c:a pcm_s16le output.mkv


If cyberlink doesn't "like" that mkv, pcm wav in AVI container is usually a safe bet for any windows program

ffmpeg -i input.mkv -vn -c:a pcm_s16le output.avi

[ShadowEdit]

I tried ffmpeg -i input.mkv -c:v copy -c:a pcm_s16le output.m2ts and this is what cmd told me:

[mpegts @ 00000000003a2b00] H.264 bitstream malformed, no startcode found, use the video bitstream filter 'h264_mp4toannexb' to fix it ('-bsf:v h264_mp4toannexb' option with ffmpeg)
av_interleaved_write_frame(): Invalid data found when processing input
frame= 1 fps=0.0 q=-1.0 Lsize= 1kB time=00:00:00.00 bitrate=11063.6kbit
s/s
video:1kB audio:0kB subtitle:0kB other streams:0kB global headers:0kB muxing overhead: unknown
Conversion failed!

So, if I wanted to just convert the extracted TrueHD track directly the command would be this?
ffmpeg -i input.thd -c:a pcm_s16le output.mkv

[poisondeathray]

Just add -bsf:v h264_mp4toannexb if you wanted the transport stream
ffmpeg -i input.mkv -c:v copy -bsf:v h264_mp4toannexb -c:a pcm_s16le output.m2ts

Since cyberlink supports mkv, you should use mkv. The overhead is about 5-7% less than a transport stream

Yes, the thd command is correct

[ShadowEdit]

Ok, I'll try the thd command and see if the mkv file will be importable into PowerDirector and see if the editing will still work.
Thanks for the help.

[ShadowEdit]

The mkv file was not importable into PowerDirector.

[poisondeathray]

Try the AVI as initially suggested. It will be the most compatible with any windows software

ffmpeg -i input.thd -vn -c:a pcm_s16le output.avi

[ShadowEdit]

The avi file was importable into PowerDirector, but there's no audio when I tried playing it in the built-in "preview player". The file didn't play at all in VLC, but it worked fine in MPC-HC. So, either the audio is not working at all in PowerDirector or it's too much for the preview player too handle. I'm on a pretty good HP ProBook laptop, so I think the preview player shouldn't have any problem playing the audio.

[poisondeathray]

The other audio only options are MOV and Transport stream.

Same song and dance:
ffmpeg -i input.thd -vn -c:a pcm_s16le output.mov

ffmpeg -i input.thd -vn -c:a pcm_s16le output.ts
ffmpeg -i input.thd -vn -c:a pcm_s16le output.m2ts
(there are slight differences between ts and m2ts, try them both)



It might be that powerdirector "doesn't like" audio only in certain containers . Can you not import video & audio in a container, but just disable the video in that program ?

For example, were you able to import AVC+PCM WAV in a MKV container ?

[ShadowEdit]

I'll try converting it to a ts file then.

I could import video + audio and then delete the video track, but then I would have to redo all the editing.

[poisondeathray]

I thought you could use a "replace" function ? Can't you just replace the "old" audio track with the "new" container, and disable or uncheckmark the video ?

Did the old project use "M4A" as the audio ? If so, then the audio in MOV would probably be the next most compatible . If that doesn't work, try renaming file.MOV to file.MP4


What kind of edits are you doing in PD ? Is it just simple cutting/trimming, or do you have other things going on like overlays, effects etc... ? How complex were your edits ?

[ShadowEdit]

What I did is I moved the audio file to another folder and then PowerDirector would ask me to locate the audio file and I would choose the replacement instead.

Not very complex, just cutting, but it took a while to get done.


I have now tried .ts and .mov. MOV wasn't importable and when I tried importing TS it got stuck at 60% and then crashed. I tried once more with the same result. So, either it's due to my PC not having good enough specs or PowerDirector just accepts audio formats for audio tracks. I'm guessing the latter.

[poisondeathray]

In some editing programs , there are options to replace the asset by right clicking on it, or right clicking on it in the asset bin

Is it just a linear sequence of 37 seconds straight from the source ? or a compilation of various cuts at different sections ? You can do basic rough edits with ffmpeg as well. 37 sec of WAV audio will be well under the limit (ie. there is no reason to convert the whole thing if you don't have to, or at least keep that in mind for future projects)

[ShadowEdit]

I haven't been able to find any replace option.

There are various cuts at from the same section of the video. At one place I cut out a few seconds to remove a word from the dialogue and I also added a picture into the video at multiple places, so I don't think that this is something doable in ffmpeg. Not easily, at least.

Maybe converting the audio to Windows Media Audio Lossless could work?

[poisondeathray]

WMA lossless encoding is only available through Microsoft . There are no open source implementations like ffmpeg/ libav. I've never used it but I guess you could try it through WME, or Expression Encoder. You probably have to use the AVI as input - it probably won't be able to decode THD, or MKV

What I meant by rough edits in ffmpeg, is you copy/cut clips like small 30 second or 1 minute clips from various sections and import those into PD to do the fine editing. Since the likely explanation is the 4GB WAV limit, those smaller clips or clip will clearly be under the limit . At least for future projects, there is no reason to import the entire video, if you only need a few sections or seconds

[poisondeathray]

Does PD support WMA Lossless variant ?

It's possible to do it through avisynth + wmnicenc (since it uses WME)

Here is the official THD sample encoded to WMA Lossless - can PD import it ? If not, try renaming the extension to WMA

[ShadowEdit]

Ok, I'll convert the thd file to avi using ffmpeg -i input.thd -vn -c:a pcm_s16le output.avi and then try Microsoft Expression Encoder and Windows Media Encoder.

What I meant by rough edits in ffmpeg, is you copy/cut clips like small 30 second or 1 minute clips from various sections and import those into PD to do the fine editing. Since the likely explanation is the 4GB WAV limit, those smaller clips or clip will clearly be under the limit . At least for future projects, there is no reason to import the entire video, if you only need a few sections or seconds

Oh, that's a good idea.

[ShadowEdit]

Does PD support WMA Lossless variant ?

It's possible to do it through avisynth + wmnicenc (since it uses WME)

Here is the official THD sample encoded to WMA Lossless - can PD import it ? If not, try renaming the extension to WMA

PowerDirector can import it and recognizes it as an audio track.

Which tool would you recommend: avisynth + wmnicenc, Microsoft Expression Encoder or Windows Media Encoder?

[poisondeathray]

Double check - is the audio actually "audible" in PD ?

I've never used WMA Lossless before. So I don't know what to recommend. EE is probably the easiest to use, or at least is a nicer GUI

But I'm just doing a quick MD5 check to see if it's truly lossless and I'm getting some differences between the THD (converted to WAV by eac3to) and the WMA Lossless (converted to WAV through directshow and the official MS decoder) . Part of the reason might be the original audio was 16bit, but the WMA Lossless was 24bit and there are different dithering algorithms used in the up/downsampling. I didn't see a 16bit WMA Lossess option for 5.1 48Khz in wmnicenc, but there might be in EE or WME . Theoretically both WAV samples should yield identical checks if it was lossless

[ShadowEdit]

I just thought of something. Maybe if I had ffmpeg cut away all the extra stuff after the clip and then converted the remaining audio to WAV?
1:12:57 - 2:01:08 could be cut away and since the audio is a little over 4 GB when uncut, this cut version should be importable.

[poisondeathray]

Sure, but you would have to re-align the audio / redo the edits slightly since the time of the audio track wouldn't be the same as the original length. Unless it's from the very beginning ? ie. from time zero to 1:12:57 ?

-ss is start time 
-t is duration 


( hh:mm:ss.ms notation )

Now was that time "cut away" ie. removed; or kept ? IT's easier in ffmpeg to specify sections that are kept


eg. convert the audio to wav from time zero (at the original beginning) to 1:12:57

ffmpeg -i input.thd -vn -c:a pcm_s16le -t 01:12:57 output.wav

[ShadowEdit]

I use none of the video or audio after 1:12:56. 1:12:57-> is disposable.

This command didn't work:

ffmpeg -i input.mkv -ss:01:12:57 -t:00:48:11 output.mkv

"Invalid duration specification for ss:01:12:57: -t:00:48:11"

By the way, could I do thd -> cut -> wav in one command?

[poisondeathray]

This command didn't work: ffmpeg -i input.mkv -ss:01:12:57 -t:00:48:11 output.mkv

By the way, could I do thd -> cut -> wav in one command?

You need to specify the codec options, and there is no colon after -ss or -t

If you wanted just the audio , add -vn -c:a pcm_s16le

This is for MKV container. Otherwise use "output.wav" instead of "output.mkv"

ffmpeg -i input.mkv -vn -c:a pcm_s16le -ss 01:12:57 -t 00:48:11 output.mkv

[ShadowEdit]

I tried this now:

ffmpeg -i input.thd -vn -c:a pcm_s16le -ss:01:12:57 -t:00:48:11 output.wav

"-vn -c:a pcm_s16le -ss:01:12:57 -t:00:48:11 '-vn' is not recognized as an internal or external command, operable program or batch file."

[poisondeathray]

There is no colon after -ss or -t . It might be a browser display issue . I've had some issues on firefox with the code boxes. Double check post # 28 again

Or how long is the entire duration of the video ? Are you sure the length is correct ? If the -t duration is longer than the actual duration, you will get an error message