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: 'Too low' for lossless FLAC ratios? (Read 4246 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

'Too low' for lossless FLAC ratios?

Hi. I recently encoded several large (around 600- to 800-MB) WAV files that are typical CD spec. and noticed that the resulting FLACs have _low_ bitrates (almost near MP3's maximum), around 456 kbps average. Similar music files encoded around 900 to 1100 kbps. I was wondering what would happen if someone decoded a lossy file to WAV, and someone else (mistakenly) got this WAV file that unknowingly came from the lossy source [and decided to FLAC it]. Would you have really low bitrates like that? The only other place I found such low bitrates is when there's not much going on in the music (like light Classical with a lot of silence).

'Too low' for lossless FLAC ratios?

Reply #1 –
What music type is it ?
To get ~35kbps with Monkey's Audio (which encodes a bit better than FLAC) I need to have average RMS ~-40dB !! (rough numbers)
Why don't you just test it ? Encode some wavs (tested with auCDtect that they are real wavs) to various lossy codecs, decode them and encode to FLAC.
More easy, just test the CD with auCDtect.
I know that I know nothing. But how can I then know that ?

'Too low' for lossless FLAC ratios?

Reply #2 –
With Mono or semi-Mono files you also get low bitrate FLACs.
Some audiobooks will be compressed to 260kbps with FLAC.
.halverhahn

'Too low' for lossless FLAC ratios?

Reply #3 –
Quote
I was wondering what would happen if someone decoded a lossy file to WAV, and someone else (mistakenly) got this WAV file that unknowingly came from the lossy source [and decided to FLAC it]. Would you have really low bitrates like that?
[a href="index.php?act=findpost&pid=259474"][{POST_SNAPBACK}][/a]


No. The way lossy codecs encode (and thus reduce the size of) the music can even create files harder to encode for lossless codecs.

What could make an easier-to-encode file would be (between others that i don't know) a lowpass filter and/or monoish content.

'Too low' for lossless FLAC ratios?

Reply #4 –
Hmm. Thanks for the tips. I'll try that auCDtect. I never heard of such a program before. I hope they're real WAVs. 

'Too low' for lossless FLAC ratios?

Reply #5 –
Most likely the files you encoded are mono or near mono material as halverhahn said. So nothing to worry about.

'Too low' for lossless FLAC ratios?

Reply #6 –
That is, mono merely means 100% correlation between the channels, allowing flac to use less info to encode.  It could be mono (100% correlated), or close to mono (highly correlated).

Quote
Encode some wavs (tested with auCDtect that they are real wavs) to various lossy codecs, decode them and encode to FLAC.  More easy, just test the CD with auCDtect.
As far as I can tell, auCDtect is designed to detect bootlegs of studio recorded music.    It wouldn't seem to wise to recommend using it on tapes of live shows unless you know the source isn't one that it will erroneously flag as having been through compression. So it depends on what's being evaluated.

'Too low' for lossless FLAC ratios?

Reply #7 –
Quote
Most likely the files you encoded are mono or near mono material as halverhahn said. So nothing to worry about.

No or nearly no stereo info can also mean a lossy codec. Many lossy encodings reduce the stereo field ("convert to mono ? yes/no"  )
I know that I know nothing. But how can I then know that ?

'Too low' for lossless FLAC ratios?

Reply #8 –
I performed a test.
I took a sample and converted it into a true-mono version with the same average loudness. Using Monkey 3.99 and "insane", the stereo version encoded to ~60% new file size, the mono version to roughly 30% (!). Probably not the half just by chance.
I know that I know nothing. But how can I then know that ?