Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Humbledore]

Hi!
I recently posted about that Volume Level and Dynamic Range values are changing after conversion from WMA lossless to FLAC (and other lossless formats):
http://yabb.jriver.com/interact/index.php/topic,107659.0.html

Today I also noticed that a conversion from WAV to FLAC changes its values as well.

I thought WAV files and lossless files were identical concerning audio quality. So why are the values changing? Is it a bug hiding here or...?

Thanks!

[Humbledore]

I tested to rip one of my CD again with the following steps:

1. Rip to WAV
2. Analyze Audio of WAV files
3. Convert Format of WAV files to FLAC
4. Analyze Audio of FLAC files

See attached image of the result. The analyze values are differing! For instance, take a look at the last track 'Nothing's the same' which has huge differences.

EDIT:
When adding the original WMA files to the list you can notice they more or less have the same values as the WAV files, see attached image. So the conversion tool probably is the problem here.

Thanks for any response to clarify this issue!

[blgentry]

I'm running MC22.0.30 on Mac.  I just took an album that had been ripped as FLAC in my library, and converted it to WAV, importing the WAVs after.  Then I ran audio analysis on the new WAV files.  All columns for audio analysis were the same for all pairs of FLAC and WAV files.

I know this is a slightly different experiment than the one you did.  I started with FLAC and converted to WAV.  But it certainly seems that this experiment, on this platform, worked fine.

Brian.

[Humbledore]

I'm running MC22.0.30 on Mac.  I just took an album that had been ripped as FLAC in my library, and converted it to WAV, importing the WAVs after.  Then I ran audio analysis on the new WAV files.  All columns for audio analysis were the same for all pairs of FLAC and WAV files.

I know this is a slightly different experiment than the one you did.  I started with FLAC and converted to WAV.  But it certainly seems that this experiment, on this platform, worked fine.

Thanks for trying to help! Do you have any theory why I get a different result?

I am running the latest version (22.0.34) on Win 10 x64. The only settings change I have made to MC lately is to enable (and disable) 'Use SoX...' and 'Play files from memory...'. No difference though. The analyze differences consist...

I just now converted a recently ripped FLAC file, analyzed it, converted it to WAV, analyzed it - and the values differs! It must be some kind of bug...

[DJLegba]

I know I mentioned DSP before, but can you post a screen shot of the "Audio Conversions Options" dialog box that comes up when you click the Options link in the "Convert Format" section of the screen?

[Humbledore]

I know I mentioned DSP before, but can you post a screen shot of the "Audio Conversions Options" dialog box that comes up when you click the Options link in the "Convert Format" section of the screen?

[DJLegba]

Thanks. And in return I'll post my screen shot. Do you have another computer you could try this on?

[Humbledore]

Do you have another computer you could try this on?

No, not for the moment.
I can just hope someone else will detect this bug as well. I'll stop converting until further notice...
Thanks!

[Humbledore]

I installed MC 21, restored the latest MC 21 library and performed the same steps. It's working as expected, i.e. Analyze Audio keep its values after conversion from WAV to FLAC!

So nothing wrong with my PC.

Is there a way to detect any installation/status failures of MC 22?

Thanks!

[Matt]

For whatever it's worth, I just analyzed some FLAC files in the latest MC22 then converted them to WAV and analyzed again.  Exactly the same results.

[Humbledore]

For whatever it's worth, I just analyzed some FLAC files in the latest MC22 then converted them to WAV and analyzed again.  Exactly the same results.

Sorry Matt, doesn't help me.

I just need to know how to get rid of this issue. If I re-install MC 22, is it something that I need to prepare/consider?

Thanks!

 

[rayooo]

Just did a conversion to WAV from FLAC.  (Original WAV was purchased from HD Tracks)

Converted to FLAC, (Compression 0)  Windows 10 x64, MC22 latest version)

Some small variations shown.  hmmm.  'guess I'd have thought this would have shown exact same for the two files.
oh, sorry for the black on black!

[Humbledore]

Just did a conversion to WAV from FLAC.  (Original WAV was purchased from HD Tracks)

Converted to FLAC, (Compression 0)  Windows 10 x64, MC22 latest version)

Some small variations shown.  hmmm.  'guess I'd have thought this would have shown exact same for the two files.

Similar small variations that I have noticed, although I also I've got some larger ones.
Important to note, as I previous said, when I did conversions in MC 21 (and in previous versions) I always got the exact same values.

[RD James]

I'm not seeing this. Perhaps try re-analyzing all files at once?
And if there are differences, I would expect it to be an issue with the analyzer.
I tested a number of conversions and then ran the files through dBpoweramp's Audio CRC tool.
 

Code: [Select]

MD5 Filename

8BFB1CA87DD5A0810497087BBF2ADDBF \ORIGINAL.flac
8BFB1CA87DD5A0810497087BBF2ADDBF \JRiver.aif
8BFB1CA87DD5A0810497087BBF2ADDBF \JRiver.wav
8BFB1CA87DD5A0810497087BBF2ADDBF \JRiver.flac
8BFB1CA87DD5A0810497087BBF2ADDBF \JRiver.m4a
8BFB1CA87DD5A0810497087BBF2ADDBF \JRiver.wv
8BFB1CA87DD5A0810497087BBF2ADDBF \JRiver.ape

EDIT:
Having JRiver set to TPDF dither changes the output.
Bit-exact dither or no dither match.
TPDF dither (with no other DSP applied) changes the checksum - though not the analysis result.
Seems like a bug.