Skip to main content ↑ ↓

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: ATRAC mono file conversion - which channel to preserve (Read 5590 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

ATRAC mono file conversion - which channel to preserve

Hi,
I need help on understanding the following.
I have a lot of old MiniDiscs, some of them contain Single Play Mono (Atrac) files (the old ATRAC format, not ATRAC3). Since I playback music only through my computer I transferred the MiniDisc-files (*.oma) to my computer and converted to wav in one go with Sonic Stage CP 4.3.01.14050. Afterwards I want to convert to flac (since I couldn't find a convenient way to playback ATRAC files with foobar and any other lossy reconversion would degrade the quality further).
Anyway, no problems with the ATRAC stereo files, but the ATRAC mono files got recoded to wav stereo files and I would like to convert to mono-files. My question is:

Does anybody know which of the two channels of the stereo wav files is the original ATRAC mono channel?

This question might seem weird at first sight, but I did the following:
1. import a stereo wav file that was converted through Sonic Stage from a ATRAC mono file into Audacity
2. Split stereo track
3. a) delete right channel and change left channel to mono. Then export as wav.
3. b) Begin at 1. and delete left channel and change right channel to mono. Then export as wav.
4. compare the two wav files of 3.a and 3.b in foobar with right-click: utilities > bit-compare tracks

Result of 4.: In all the tested mono-wav-files (former stereo-wav-files, former ATRAC-mono-files) that were created like described above there are peak-differences. Therefore there are peak-differences between the left and right channel in the stereo-wav-files which were converted from a ATRAC-mono-file. Hence the question, which of the channels is the original one?

The other possibility would be to convert the stereo-wav-files (former ATRAC-mono) within foobar with downmix channels to mono. But before doing this, I would like to know why there are differences between the channels from a stereo-wav-file that was converted from a ATRAC-mono-file. To precise the question:

1. Does SonicStage fill the second channel in the stereo-wav-files (former ATRAC-mono) with a simple copy of the existing ATRAC-mono-channel?
(This does not seem to be true, except the conversion of SonicStage is always faulty and therefore these differences in bit-compare in foobar (see above).)
or
2. Do ATRAC-mono-files still have two channels?
or
3. Does SonicStage do some coding on the mono-files which leads to different left and right channels?
or
4. Do the differences in left and right channel depend solely on the Audacity conversion?

Anyhow, the question is: Take left or right channel or downmix both channels to one channel?

Or does anybody know how to embed ATRAC in another container to have it played-back without great issues in foobar and other media-players?
- I do know that the foobar-plugin command-Line Decoder Wrapper can playback ATRAC via ffmpeg, but it opens up a new window and the playback controls do not work - no help.
- I do know about the foo_input_vgmstream plugin, but it only plays back ATRAC3 or ATRAC3plus, definitely not ATRAC.
- VLC does not playback my ATRAC-files

Thanks in advance

ATRAC mono file conversion - which channel to preserve

Reply #1 –
Doesn't/didn't audacity dither by default? If so, this means identical files don't come out identical when re-saved. Hence any comparison through audacity with dither enabled is flawed.

You don't need to worry about these dual-mono files. FLAC will compress them almost as efficiently as true mono (single channel) files. You will gain almost nothing by reverting them to their original single channel format. Stereo files have slightly wider compatibility than mono files too.

Cheers,
David.

P.S. Sonic Stage may dither the conversion from ATRAC to WAV. If so, the channels in the WAV won't be identical, and that's OK, and neither is more "original" than the other.

ATRAC mono file conversion - which channel to preserve

Reply #2 –
Thanks, David!
Yes, dithering was activated in Audacity. After repeating steps 1-4 mentioned in my first post - this time without dithering - the resulting files of left or right channel were bit-identical. Ok so far, but:

You should be right that I won't gain nothing in terms of quality by reverting to mono, but I want the files to be mono-files because I'd like to see that they are mono. If I save them as stereo, I will forget about there original source. I could insert a tag, right, but then I always need to look for this tag. Instead, if the files are saved as what they are - mono - than it's always visible at first sight in my setup without any further effort.

Second, I thought, mono files should be:
1 Channel 22.05 kHz, right?
This would decline filesize. But, when exporting from Audacity after deleting one channel of a dual-mono wav-file I get:
1 channel 44.1 kHz!?

Changing the project rate inside Audacity leads to:
1 Channel 22.05 kHz but cuts of everything above 11 kHz resulting in a decline of quality. Same thing when converting inside foobar a dual-mono 44.1 kHz wav file to 1 Channel 22.05 kHz (Downmix channels to mono, Resampler (PPHS)).
Am I wrong? Stereo is 2x 22.05 kHz = 44.1 kHz and mono should be 1x 22.05 kHz, right? Is there a way to save the files as what they are (mono, 22.05 kHz) without cutting all frequencies above 11kHz?

thanks again
Daniel

ATRAC mono file conversion - which channel to preserve

Reply #4 –
Right...  You can think of the sample rate as the "clock rate". 

With a stereo file there are actually 88,200 samples per second, and you can calculate file size of an uncompressed WAV file as as:

(Playing time in seconds) x (Sample Rate) x (Bit Depth/8) x (Number of Channels).

ATRAC mono file conversion - which channel to preserve

Reply #5 –
Just curious...cd audio is 1411kbps, so 1411000/16/2=44093.75, not quite the advertised 44100khz. Anybody know precisely why it works out that way?

ATRAC mono file conversion - which channel to preserve

Reply #6 –
Quote from: Mach-X on
Just curious...cd audio is 1411kbps, so 1411000/16/2=44093.75, not quite the advertised 44100khz. Anybody know precisely why it works out that way?


Rounding error.  The actual bitrate is  1,411,200 bits per second according to wikipedia.

ATRAC mono file conversion - which channel to preserve

Reply #7 –
Quote from: schallquant on
You should be right that I won't gain nothing in terms of quality by reverting to mono, but I want the files to be mono-files because I'd like to see that they are mono. If I save them as stereo, I will forget about there original source. I could insert a tag, right, but then I always need to look for this tag. Instead, if the files are saved as what they are - mono - than it's always visible at first sight in my setup without any further effort.


You might run into problems should you save to mono. Many (most?) soundcard drivers won´t be able to handle mono files (or files with only one channel). I don´t know what hardware interface you´d be using, but WASAPI or ASIO in my experience don´t work at all with true mono files.

So I´d go for a tag stored within stereo FLAC files too. It might seem inconvenient at first but it saves you the hassle of converting your desired mono files to stereo whenever you want to listen to them on anything else than your PC.

BTW, the MiniDisc contains two mono tracks: the left channel and the right channel. Joint stereo, which would have increased efficiency of the ATRAC codec, wasn´t used for MD. The mono mode MD recorders were equipped with means nothing more than that only one channel is encoded during recording instead of two. That way, the MD was able to double recording time. The first MD recorders equipped with mono recording used the left channel of any input. Later models used both channels, which were downmixed to mono within the ATRAC IC and then encoded & stored as a mono file.

Do you use a HiMD model for uploading to PC? If so, SonicStage will just upmix the mono file by copying the single channel to the other available channel. I don´t know where copying takes place, it is irrelevant anyway as both channels are the same. And as David has said: SonicStage might dither the output which would then create visible differences (which probably won´t be audible).
marlene-d.blogspot.com

ATRAC mono file conversion - which channel to preserve

Reply #8 –
Quote
... Many (most?) soundcard drivers won´t be able to handle mono files (or files with only one channel). ... So I´d go for a tag stored within stereo FLAC files too. ...

Well, didn't know about that. I already converted to mono, giving it a try and baring in mind that it might not work... Thanks! At least now I know before what the problem might be if some files won't play (properly) in some setups...

Quote
Joint stereo, ..., wasn´t used for MD.

There were different recording modes. "Long Play 4" (LP4) on Net-MD-Recorders used Joint-Stereo. All the other Recording modes used separate channels (see: https://en.wikipedia.org/wiki/MiniDisc#MDLP).

Quote
The first MD recorders equipped with mono recording used the left channel of any input. Later models used both channels, which were downmixed to mono within the ATRAC IC and then encoded & stored as a mono file.

This was the information I initially was looking for, thanks. Even though, now it doesn't matter anymore since both channels are identical, see below.

Quote
Do you use a HiMD model for uploading to PC? If so, SonicStage will just upmix the mono file by copying the single channel to the other available channel. I don´t know where copying takes place, it is irrelevant anyway as both channels are the same. And as David has said: SonicStage might dither the output which would then create visible differences (which probably won´t be audible).

Yes, I used MZ-RH1. SonicStage does not seem to dither because:
The left and right channel of the resulting dual-mono-wav-files are identical if not dithered inside Audacity! Further comparing with foobar fingerprints (foo_biometric.dll) shows that there is no (detectable) difference in music between the left- and the right-channel of a dual-mono-file (produced like mentioned in the first post) and not even between the dual-mono-file as a whole compared to one of the single channels.
Using Audacitys Function "Stereo Track To Mono" (without dithering) generates a mono-file that yields 100% Match confidence when comparing with foobar fingerprint, therefore I do not expect any relevant differences.
The differences between left and right-channel, stated in my first post, that I found when dithering was activated inside Audacity, might or might not be audible, I haven't tested because that's too time-consuming. If I had known before, that the resulting peak-differences I initially found, where just introduced by dithering, maybe I would not have looked further into it... My problem was, that I did not know, if my encoding chain was inserting some noise or something else...

Therefore, thanks to all, you helped finding a solution!