Acoustica 7.1.8 - question WAV/FLAC saving

[jumpingjackflash5]

Hello,

first let me state that Acoustica is really great product and I'm very satisfied customer.

I have just a question to the Acoustica's behaviour when saving files that have peaks above 0 dbFS in 32 bit float, such as those created by Normalize feature to e.g. +1 dB . I know that is not recommended, that we should Normalize e.g. to -0.2 dB, but sometimes only very small portions of the recording clip.

When I save such a recording to 24 bit PCM in FLAC, everything is OK, Acoustica says after saving that the files contain samples above 0dbFS, I click OK and done. Maybe there is very slight distortion at the clipped areas but nothing remarkable.

When I save such a recording to 24 bit PCM in WAV, Acoustica says after saving that the files contain samples above 0dbFS, I click OK and done, but the created files have loud cracks and imperfections on playback in those areas that clipped.

And so I ask - what this behind this behaviour? FLAC and WAV are both lossless, so they should be the same?

Thank you

Jan

[jumpingjackflash5]

Just to make sure that it will be understood well, when there is no clipping, everything is the same (and OK) regardless of the saving format (WAV/FLAC).

[jumpingjackflash5]

Just one more addendum as I was thinking about the question. Even if it is interesting what cause the difference between WAV and FLAC save, would it not be better in the cases when there is only few peaks use the "limiter" tool, which will keep those peaks under 0 dB and left the rest of the audio normalized to the desired level?
And will it still be transparent outside those peaks, e.g. only the volume of the recording will be changed as a result of limiter processing? (I generally prefer as transparent digital processing as possible, therefore I don't use any compressors or dynamic range reduction).

[stian]

Hi Jan!

Hello,

first let me state that Acoustica is really great product and I'm very satisfied customer.

Thanks! I'm very glad to hear that!

I have just a question to the Acoustica's behaviour when saving files that have peaks above 0 dbFS in 32 bit float, such as those created by Normalize feature to e.g. +1 dB . I know that is not recommended, that we should Normalize e.g. to -0.2 dB, but sometimes only very small portions of the recording clip.

When I save such a recording to 24 bit PCM in FLAC, everything is OK, Acoustica says after saving that the files contain samples above 0dbFS, I click OK and done. Maybe there is very slight distortion at the clipped areas but nothing remarkable.

When I save such a recording to 24 bit PCM in WAV, Acoustica says after saving that the files contain samples above 0dbFS, I click OK and done, but the created files have loud cracks and imperfections on playback in those areas that clipped.

And so I ask - what this behind this behaviour? FLAC and WAV are both lossless, so they should be the same?

Thanks for reporting this issue. The WAVE encoder does indeed seem to "wrap around", thus causing much more severe distortions than necessary. I've added a bug entry for that and we'll get it fixed ASAP.

Best,
Stian

[jumpingjackflash5]

Thank you very much! I am glad that I identified it correctly. While I in majority or cases try to normalize (or limit) at -0.2 dB, sometimes just a small spike remains and when saving as WAV, distortion that is not bigger than necessary is welcomed.