Flac 1.3.1 has been released

[Kurtnoise]

General:

Improved decoding efficiency of all bit depths but especially so for 24 bits for IA32 architecture (lvqcl and Miroslav Lichvar).
Faster encoding using SSE and AVX (lvqcl).
Fixed bartlett, bartlett_hann and triangle functions.
New apodization functions partial_tukey and punchout_tukey for improved compression (Martijn van Beurden).
Retuned compression presets to incorporate new apodization functions (Martijn van Beurden).
Fix -Wcast-align warnings on armhf architecture (Erik de Castro Lopo).

FLAC format:
(none)

Ogg FLAC format:
(none)

flac:
Help output documentation improvements.
I/O buffering improvements on Windows to reduce disk fragmentation when writing files.
Only write vorbis-comments if they are non-empty.

metaflac:
(none)

plugins:
Fix symbol visibility in XMMS plugin.

build system:
Many fixes and improvements across all the build systems.

documentation:
Document new apodization windows.

libraries:
Fix CVE-2014-9028 (heap write overflow) and CVE-2014-8962 (heap read overflow) (Erik de Castro Lopo).

Interface changes:

libFLAC:
(none)
libFLAC++:
(none)

Sources (http://downloads.xiph.org/releases/flac/flac-1.3.1.tar.xz) | Windows binaries (http://downloads.xiph.org/releases/flac/flac-1.3.1-win.zip)

[hello_hello]

Thanks!

[Thunderbolt8]

does this update have any practical relevance when using winamp to play flac files or eac3to to create flac files?

[the_weirdo]

does this update have any practical relevance when using winamp to play flac files or eac3to to create flac files?

In case of using eac3to to create FLAC files, yes. It'll be faster.

[Thunderbolt8]

but madshi has to update libflac.dll for that, right?

[Sparktank]

but madshi has to update libflac.dll for that, right?

I believe he said it's okay to update the library manually. It was asked before so if you do a search, you should see the post, recently this year, iirc.

Libav/ffmpeg is the the only thing with custom code, I believe.

[the_weirdo]

but madshi has to update libflac.dll for that, right?

As there's no API change in this new version, you can just replace it in place. (Download the new DLL from here (http://www.rarewares.org/lossless.php#flac-dll), rename libFLAC_dynamic.dll to libFLAC.dll).

EDIT: Note that they've retuned compression presets (including -8 which is used by eac3to), so sometimes, the compression of new version may be worse than older versions a bit (usually less than 1%).

[Thunderbolt8]

I guess I need the 32-bit version of libflac.dll for eac3to? or does that depend on my OS?

[Groucho2004]

I guess I need the 32-bit version of libflac.dll for eac3to? or does that depend on my OS?
1.Yes
2.No

[tebasuna51]

EDIT: Note that they've retuned compression presets (including -8 which is used by eac3to), so sometimes, the compression of new version may be worse than older versions a bit (usually less than 1%).

Before (1.3.0)
-8, --compression-level-8, --best Synonymous with -l 12 -b 4096 -m -e -r 6

Now (1.3.1)
-8, --compression-level-8, --best Synonymous with -l 12 -b 4096 -m -r 6

Seems the difference is
-e, --exhaustive-model-search Do exhaustive model search (expensive!)

Or there are other difference?

[the_weirdo]

Or there are other difference?

http://git.xiph.org/?p=flac.git;a=commitdiff;h=02891daa7a88bf963fab6511aa0e1270b850b0a9

Also, I said sometimes the compression of new presets is a bit worse than before so that people don't surprise when that happens. In most cases, however, this version should perform better than old versions (or at least the same; compression-wise).

[tebasuna51]

@the_weirdo
Thanks for the info.

[Music Fan]

Is there a way to use 1.3.1 with Flac frontend ? I installed the last GUI (v 2.1) but it use FLAC 1.3.0.
And there is no dll in FLAC Frontend's folder.
http://flacfrontend.sourceforge.net/

[LoRd_MuldeR]

Is there a way to use 1.3.1 with Flac frontend ? I installed the last GUI (v 2.1) but it use FLAC 1.3.0.
And there is no dll in FLAC Frontend's folder.
http://flacfrontend.sourceforge.net/

Replacing the flac.exe in the "tools" sub-folder doesn't work? :confused:

[Music Fan]

Yes, thanks, I forgot the exe.:) I was focused on dll because just before I read something about the Flac dll for eac3to.

[Overdrive80]

Hey, somebody could tell me max. frame size in ms for flac? Thanks

[tebasuna51]

Hey, somebody could tell me max. frame size in ms for flac? Thanks

It's not easy to know that, using info from https://xiph.org/flac/documentation_format_overview.html and https://xiph.org/flac/format.html#frame_header

...Like most audio codecs, FLAC splits the unencoded audio data into blocks, and encodes each block separately. The encoded block is packed into a frame and appended to the stream. The reference encoder uses a single block size for the whole stream but the FLAC format does not require it.

...In order to simplify encoder/decoder design, FLAC imposes a minimum block size of 16 samples, and a maximum block size of 65535 samples. This range covers the optimal size for all of the audio data FLAC supports.

Currently the reference encoder uses a fixed block size, optimized on the sample rate of the input. Future versions may vary the block size depending on the characteristics of the signal.

but:
FLAC specifies a subset of itself as the Subset format. The purpose of this is to ensure that any streams encoded according to the Subset are truly "streamable",...

The Subset makes the following limitations on what may be used in the stream:
- The blocksize must be <=16384; if the sample rate is <= 48000Hz, the blocksize must be <=4608.
...

Like FLAC support:
Sample size in bits: 8, 16, 20 and 24 (1 to 3 bytes)
Channels: 1 to 8
Samplerate: 8 to 96 KHz

A theoric max framesize in ms. can be for a 1 channel, 8 bits (1 byte) and 8 KHz (8000 samples/s, 8 samples/ms):

Max framesize in ms: Max_BlockSize / (Channels x ByteDepth x Samples/ms)

Max framesize in ms: 65535 / (1 x 1 x 8) = 8191,875 ms

But blocksize is "optimized on the sample rate" and with limits in "Subset format" maybe we can have for a 2 channel, 16 bits (2 bytes) 48 KHz ( 48 samples/ms):

Max framesize in ms: 4608 / (2 x 2 x 48) = 24 ms

For 6 channels, 24 bits (3 bytes) and 48 KHz:

Max framesize in ms: 4608 / (6 x 3 x 48) = 10,667 ms

[Overdrive80]

Ok, thanks Tebasuna51, I am testing time-stretch option in matroska with opus and like I knew that with flac playback fine, I need config --framesize option properly. Regards.