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: Volumax weird results (@john33) (Read 20680 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Volumax weird results (@john33)

Hello,

I'm using this little gem to prescale files I do later encode to .mp3s. The command line is something like:
volumax -a -y -l -0.1234567 -b 4 *.wav

... and yes, I do have 7 digits after the comma. The files already have the desired output format, so -b 4 may be dispensable but shouldn't be a problem. I have a batchfile containing the same command line two times. In both cases the source file is exactly the same. Sometimes (not always) the resulting files are different. I'm neither a programmer nor is my english good enough to describe it better (sorry for that). Maybe I did something wrong (too many digits after the comma) or my PC is too fast (a timing problem, though I don't think so). It would be very nice if you could give me a hint or have a look at the source; maybe a recompile is all that is necessary.

Thanks in advance, Surfi

Volumax weird results (@john33)

Reply #1 –
Wow, it's been a while since I looked at this!  Nice to know someone actually uses it!

Having said that, I can't give you a definitive answer except to say that the input is converted to float values and converted back to the desired format after processing. Using floats can lead to differing results and that's the best answer I can give you, I'm afraid. While the outputs may differ on occasion, I would guess that you are unable to discern any difference by ear? If there is an audible difference, please let me know, but, otherwise, I wouldn't worry about it.

Volumax weird results (@john33)

Reply #2 –
Hi John,

let me thank you first for WaveGain and Volumax.
You're right, I didn't notice the differences by ear. What my batch does is to reduce the volume of a .wav-file, encode the .wav to .mp3, test if the .mp3 is clipping and if it does reduce the volume of the .wav a little bit more and so on ... until the .mp3 isn't clipping anymore. If this happens I do the final enode. Sometimes the batch finds a non-clipping value for -l and after final encoding the .mp3 is clipping though. I compared the files and could verify differences after using Volumax. WaveGain is used before Volumax too but never produced any differences.

Anyway, if this behavior is by design I'll have to find a different solution.
Thanks again, Surfi

Volumax weird results (@john33)

Reply #3 –
Just for the record, Volumax was compiled using what is now a pretty old version of libsndfile for I/O and would almost certainly benefit from some attention. Whether this would resolve your issue I'm not certain, but I'll take a look when I have some time.

Volumax weird results (@john33)

Reply #4 –
... you made my day ... 

Volumax weird results (@john33)

Reply #5 –
Quote from: john33 on
Wow, it's been a while since I looked at this!  Nice to know someone actually uses it!
I still use it too, usually to bring archived music back to the highest possible non-clip album level before burning for friends.
Quote
Just for the record, Volumax was compiled using what is now a pretty old version of libsndfile for I/O and would almost certainly benefit from some attention. Whether this would resolve your issue I'm not certain, but I'll take a look when I have some time.
Great! I will now track this topic and snag any updates. Thanks, john33.

Regards,
Madrigal

p.s. I just use it in the WaveGain frontend, after renaming volumax.exe ... I prefer GUI to CLI.

edit = the p.s.

Volumax weird results (@john33)

Reply #6 –
Well, here you go guys. Volumax 0.4 can be downloaded from: http://www.rarewares.org/files/others/volumax0.4.zip

I've done a quick test and the results seem to correspond to the previous version but there is better resilience and error checking. One or two options have changed slightly and there are more of them. 

Testing would be much appreciated, thanks, as would feedback. 

If there are no problems, I'll release over the weekend, replacing the one currently at Rarewares.

Volumax weird results (@john33)

Reply #7 –
Quote from: john33 on
Testing would be much appreciated, thanks, as would feedback.


I'll do (quick) testing and post results, thank you so much. 


EDIT
Didn't test functions yet but the -h switch. Please compare the output of:

1. Volumax -h
2. Volumax -h >Volumax2.dok
3. Volumax -h 1>Volumax3.dok
4. Volumax -h 2>Volumax4.dok

... and you could use an other default log-filename (VMLog.txt)
/EDIT

Surfi

Volumax weird results (@john33)

Reply #8 –
Quote from: john33 on
Testing would be much appreciated, thanks, as would feedback.
Thank you, john33.

I have never had any audible issues with Volumax, and wouldn't have the slightest idea how to test it at the CLI level and feed back to you. Please do not take my silence in this area as laziness, lack of interest or lack of appreciation. Volumax has made my life quite a bit easier on many occasions, I thank you for it and will await the final results of this latest go-round with Surfi.

Best regards,
Madrigal

Volumax weird results (@john33)

Reply #9 –
Quote from: Surfi on
......
EDIT
Didn't test functions yet but the -h switch. Please compare the output of:

1. Volumax -h
2. Volumax -h >Volumax2.dok
3. Volumax -h 1>Volumax3.dok
4. Volumax -h 2>Volumax4.dok

... and you could use an other default log-filename (VMLog.txt)
/EDIT

Surfi

Thanks. I've fixed this and uploaded again on the same link.

Volumax weird results (@john33)

Reply #10 –
Me again  ,

-g - option doesn't work (--level not tested shouldn't this be --gain too?). If you did change the default log-filename (not tested) please modify the help-text also.


... until then ...

Volumax weird results (@john33)

Reply #11 –
Quote from: Surfi on
Me again  ,

-g - option doesn't work (--level not tested shouldn't this be --gain too?). If you did change the default log-filename (not tested) please modify the help-text also.


... until then ...

Hmm, thanks. 'More haste, less speed', methinks!    Fixed and fixed, and new upload on same link.

Volumax weird results (@john33)

Reply #12 –
Quote from: john33 on
Hmm, thanks. 'More haste, less speed', methinks!    Fixed and fixed, and new upload on same link.


... no problem, slow down. I'll do my tests and post back tomorrow.

EDIT
... oops  , faster than I thought:
-g (or --gain) calculates right but doesn't write changes to the file (and yes, I did use -y)
/EDIT

Thanks again, Surfi.

Volumax weird results (@john33)

Reply #13 –
Quote from: Surfi on
Quote from: john33 on

Hmm, thanks. 'More haste, less speed', methinks!    Fixed and fixed, and new upload on same link.


... no problem, slow down. I'll do my tests and post back tomorrow.

EDIT
... oops  , faster than I thought:
-g (or --gain) calculates right but doesn't write changes to the file (and yes, I did use -y)
/EDIT

Thanks again, Surfi.

And, again, new upload on same link.  (Continued thanks for your patience!  )

Volumax weird results (@john33)

Reply #14 –
Quote from: john33 on
(Continued thanks for your patience!  )

... it's me who asked   

... but no change in results. Seems to ignore -g completely. Commandline:
volumax.exe -a -y -g -1.0000000 "*.wav"


Best regards, Surfi

Volumax weird results (@john33)

Reply #15 –
Quote from: Surfi on
Quote from: john33 on

(Continued thanks for your patience!  )

... it's me who asked   

... but no change in results. Seems to ignore -g completely. Commandline:
volumax.exe -a -y -g -1.0000000 "*.wav"


Best regards, Surfi

And, yet again, another upload on the same link. I do believe this one is OK!!

Volumax weird results (@john33)

Reply #16 –
Quote from: john33 on
And, yet again, another upload on the same link. I do believe this one is OK!!

Looks promising    ,

starting intensive tests now. Don't expect anything before this evening (have to change my batches too).


Surfi

Volumax weird results (@john33)

Reply #17 –
Quote from: Surfi on
Quote from: john33 on

And, yet again, another upload on the same link. I do believe this one is OK!!

Looks promising    ,

starting intensive tests now. Don't expect anything before this evening (have to change my batches too).


Surfi

That'll be great, thanks very much.

Volumax weird results (@john33)

Reply #18 –
... a bit later than expected but not too late I hope:

1. nothing in logfile (looks like this):

Command line:
   volumax.exe -y -l -f G:\.__test\_output\VLog.dok G:\.__test\_output\*.wav

Analyzing...

  Peak  | Scale | New Peak | Track
-------------------------------------


Volumax Processing completed normally


2. bit depth (sample rate is preserved):

Same commandline - assume you have files of different bit depth in your folder. Modified files do not have original bit depth anymore. They all have the same of the first processed file.


... continue testing

Surfi

Volumax weird results (@john33)

Reply #19 –
Quote from: Surfi on
... a bit later than expected but not too late I hope:

1. nothing in logfile (looks like this):

Command line:
   volumax.exe -y -l -f G:\.__test\_output\VLog.dok G:\.__test\_output\*.wav

Analyzing...

  Peak  | Scale | New Peak | Track
-------------------------------------


Volumax Processing completed normally


2. bit depth (sample rate is preserved):

Same commandline - assume you have files of different bit depth in your folder. Modified files do not have original bit depth anymore. They all have the same of the first processed file.


... continue testing

Surfi

Not too late at all, thanks.  I believe both of these are now fixed in new upload on same link as before.

Volumax weird results (@john33)

Reply #20 –
Volumax help says:

  Usage: volumax [options] input.wav [...]

but I've seen it's also accepting:

volumax [options] "PathToWavFiles"

even with no -z-switch though it doesn't work recursive then (which is intended).
This should be taken to the help text.


Put two .wavs to a folder named _output and compare the following command lines:

volumax.exe -f ".\VLog.dok" ".\_output\*.wav"  (this is working in any case)
volumax.exe -f ".\_output\VLog.dok" ".\_output\*.wav"  (this is working in any case)

volumax.exe -f ".\VLog.dok" ".\_output"  (this is working too)
volumax.exe -f ".\_output\VLog.dok" ".\_output"
gives you:
 Unrecognized file format for VLog.dok.
but files are processed correctly. Correction: Only analyzing is done!

I think you should limit file parsing to supported extensions and make sure logfile hasn't any of them.


volumax.exe -f ".\VLog.dok" ".\_output\"
gives you:
Couldn't go to folder '.\_output"': Invalid argument

I think this should work too but it's your choice.


All other things seem to be fixed
Correction: -b 5 not working (WaveGain too), only analyzing is done and volumax.tmp isn't removed!
  but I'll do a final test anyway.

Surfi

Volumax weird results (@john33)

Reply #21 –
Thanks for your testing and input. I hear what you say regarding input filtering and I will look at it, although I'm not sure I'll be able to that until after I return on 29 July, from a short vacation. The expectation, currently, is that input will be specified either as a specific filename, or '*.wav', etc. I will look at this though, after I return.

Volumax weird results (@john33)

Reply #22 –
Quote from: john33 on
Thanks for your testing and input. I hear what you say regarding input filtering and I will look at it, although I'm not sure I'll be able to that until after I return on 29 July, from a short vacation. The expectation, currently, is that input will be specified either as a specific filename, or '*.wav', etc. I will look at this though, after I return.

OK, all the best for you and a safe journey!

Surfi

Volumax weird results (@john33)

Reply #24 –
Quote from: john33 on
, input filtering is as it was and will probably remain unchanged.
No problem at all, thanks again for the new version. It's the perfect solution for all my needs now (and with nice-looking output too). Even the specific problem (start post) is definitely solved, thank you very much.

Quote from: john33 on
I'll be correcting the -b 5 issue in WaveGain as well.
Assume I feed Volumax with an output generated by WaveGain with -b 5 option. Will there be two conversions less?

-> is conversion

without options:
16 bit in -> float (internal) WaveGain -> 16 bit out -> float (internal) Volumax -> 16 bit out

with WaveGain -b 5 and Volumax -b 2:
16 bit in -> float (internal) WaveGain - float (internal) Volumax -> 16 bit out


Will you correct the WaveGain 1.26 or the Beta 1.27?


Kudos to you!
Surfi