Author Topic: FLAC Conversion Settings  (Read 6714 times)

bitseq

  • Newbie
  • *
  • Posts: 12
I have some AIF files at 44.1 @ 1411k

Could someone help me with the correct syntax to give me the same sample rate or at least not lose any quality when converting to flac?

Cheers

MeeMeeMee

  • Full Member
  • ***
  • Posts: 233
Conversion to FLAC (as long as you don't use the lossyWAV option) will not affect the audio in any way. I'd suggest using -s 8 option. It just costs some more CPU resources while compressing in order to produce a smaller flac file.
MusicBee 3.5.8516 / Windows 10 (64-bit) / Intel i5-3470 / 8GB RAM
Media on NAS (CIFS share)
20K+ tracks, predominantly FLAC ; converted to mp3 (lame -V 0) when synced to a Cowon D2+
________________________________________________________________________________________________
Get the latest patch: https://getmusicbee.com/patches/

bitseq

  • Newbie
  • *
  • Posts: 12
Thanks but I really want to avoid compressing the flacs, size is not an issue for me, When I use a compression of 0 the sample rates on the flacs are lower, around 800k.

Currently using this format (-s -0 -o [outputfile] -)

hiccup

  • Sr. Member
  • ****
  • Posts: 7790
Are you sure you are seeing the word 'sample rate' somewhere in your settings or results?
I am assuming you are referring to 'bit rate'.
With flac, that only means the stream volume of bits the flac decoder will have to handle.
If you compressed flac to the max, you'll have a lower bit-stream, but the decoder has to work a little harder.
If you use the lowest compression setting, the bit-stream will have a higher 'volume', and the decoder has to do a little less work.

All this has nothing to do with sound quality or sample-rate.
As MeeMeeMee said, flac is always lossless, sound quality wise.

MeeMeeMee

  • Full Member
  • ***
  • Posts: 233
If you compressed flac to the max, you'll have a lower bit-stream, but the decoder has to work a little harder.
If you'll look at the decoding CPU vs Compression graph here you'll see that there's almost no difference in FLAC when decoding high or low compressed files (=the graph is almost vertical). And as a side note - it's the fastest (=least CPU-intensive) lossless codec out there.
I don't see why one would use anything but -8.
Last Edit: January 06, 2016, 08:32:15 PM by MeeMeeMee
MusicBee 3.5.8516 / Windows 10 (64-bit) / Intel i5-3470 / 8GB RAM
Media on NAS (CIFS share)
20K+ tracks, predominantly FLAC ; converted to mp3 (lame -V 0) when synced to a Cowon D2+
________________________________________________________________________________________________
Get the latest patch: https://getmusicbee.com/patches/

hiccup

  • Sr. Member
  • ****
  • Posts: 7790
you'll see that there's almost no difference in FLAC when decoding high or low compressed files (=the graph is almost vertical). And as a side note - it's the fastest (=least CPU-intensive) lossless codec out there.
I don't see why one would use anything but -8.

It's not really relevant to the question the OP was asking about his concern about the sound quality, but to answer your question why anyone would use anything but -8.
Have a look at the wav to flac test here: http://z-issue.com/wp/flac-compression-level-comparison/
You will find there that -8 takes 3 times longer to encode than -5, and results in a space saving of 0,1% (yes, I got the comma in the correct place)
That's why the general advice still is to leave the (usually) default setting at -5

bitseq

  • Newbie
  • *
  • Posts: 12
Sorry yes I was referring to bit rate.

I guess it shows that I know little about FLAC files and encoding in general.

I think I understand more clearly now. Quality wise I should be more concerned with sample rate and bit depth, when it comes to bit rate this is essentially the factor of lossless flac compression, like a zip file for example.

The reason that the aif files had constant bit rates was that it was raw uncompressed data.

hiccup

  • Sr. Member
  • ****
  • Posts: 7790
I think I understand more clearly now. Quality wise I should be more concerned with sample rate and bit depth, when it comes to bit rate this is essentially the factor of lossless flac compression, like a zip file for example.

Correct. The most 'damage' to digital audio is usually done when converting/resampling the original bit depth or sample rate.
So you are wise to keep an eye on it if those are changed, either at converting, or at playing back. (which by the way can easily happen on a Windows PC depending on your audio and soundcard settings)
If you want to learn more about that, search for terms such as bit-perfect, asio, wasapi, exclusive mode.

And back to the question as your header states:
Just leave flac at -5. Or possibly set it to 0 or -1 when doing bulk encoding. This will cut the time required for encoding more than in half, with only a very minor effect on the file size.
In all cases, 16bits audio will remain 16bits, 24bits will remain 24bits, etc.