>2016 >thinking that flac = lossless https://my.mixta

>>56035910
Do you even know what lossless means?
Get out.

There are millions of ways to fuck with audio files. If it doesn't have a log file, don't trust it.

Only public tracker plebs would download shit without logs.

>>56036053
>not ripping your audio discs into WAV lossless.

>not ripping your audio cds as iso files

>>56036053
>needing log files
>not inspecting the spectrogram using an oscilloscope
desu you're the pleb here

>not ripping your audio CD in half.

File: Capture235.png (765KB, 931x639px) Image search: [Google]

765KB, 931x639px

>>56036832
There are always exceptions with spectrograms.

File: 1433965586040.png (84KB, 471x471px) Image search: [Google]

84KB, 471x471px

>lossy
Opus>Vorbis>Mp3(LAME3.99)Musepack>wavpack hybrid>AAC

>Lossless
FLAC>OptimFROG>APE>Shorthen>ALAC>WMA Lossless

FACTS
A
C
T
S

>>56037048
I'm pretty sure lossy is Apple AAC > MP3.

>>56036971
how ?

>>56036832
>not inspecting the spectrogram using an oscilloscope
You can only see a very limited range of artefacts like that. It's pretty easy to tweak the audio in a way to make those disappear as well for most people. Is there actually a software suite that specifically analyses data to check whether it's legit or not? If there's not, I think I have a new project.

>>56035910
I was really expecting something cool there, but it's literally just downsampled to 22.05 Hz/8 bit mono. So the thing is, it IS still a lossless copy of that downsampled file.

I'd be more interesting to see someone exploiting FLAC's compression algorithm, similar to how lossy GIFs are possible by altering the source in a way to make it easier to compress. Also, new project. Inspiring thread, OP.

>>56037048
>lossy
Opus > why mention anything else?

File: p1iEFr9.jpg (105KB, 600x970px) Image search: [Google]

105KB, 600x970px

>>56036053

lol gotta have to log files for your music when you download your shitty EDM or russian jumpstyle music right

>>56035910
are you actually this retarded

>>56035910
>something happened to it, like a lowpass or 64kbps low bitrate lossy encoding, which destroys most of the data
>it's been transcoded to FLAC, because WHY WOULD YOU DO THAT (proprietary codec?)
The quality will equal what the lossy codec was, but the bitrate will increase because FLAC won't know how to encode it as optimally as the codec it came from. But there's still not a lot of information to encode, so hello Shannon - low bitrate FLAC.

Or someone fucked with the FLAC to remove some of the residuals. If you know how lossless codecs work (predictor + residuals), which say OptimFROG exposed...

>>56039487
Lowpass disabled? I think you used to be able to do that with -k, or one of the --dm-presets; certainly you can change it with --lowpass 22050. It may also be --freeformat with a higher bitrate.

If the former it probably sounds like shit, because it's been told to favour a complete-looking spectrum over actually spending bits in a scalefactor you can barely hear and MP3 isn't poor at.

If the latter, good luck playing it (you can go higher than 320, but the bitrate flags in the frame cannot understand that, so quite a few things break and due to inherent limitations in MPEG I Audio Layer 3 you're not going to do much more accurate encoding in practice).

>>56037048
WavPack is technically superior to FLAC, but we really only needed one good codec (and favour low-power efficient decoders), and later FLAC encoders (Flake, etc) can put more effort into finding optimal codes anyway, so it's not much of a difference.

>>56039591
damn, I really thought it was mono

It's funny because this is the same way I used to work with samples in a tracker. That or applying an envelope to compressed 8-bit samples.