[Blackwoof]

They don't even realise DVD video, TV, Video games all use 192 ~ 448Kbps Lossy audio. They somehow can only tell when they get there Flac/CD encoded to a 256kbps File, Many did this with PS3 which had PCM stereo for everything but couldn't tell 96 ~ 160kbps stereo versions on Xbox 360. Lossless is highly inefficient many Rock/metal/pop average 1mbit just 5 ~ 8% less than Wav file on the CD?.

Vorbis/Opus/xHE AAC can do 384 ~ 512kbps VBR for hard to compress samples, Even MP3 does by using bit reservoir to allow 325 ~ 640kbps frames if needed.

Also Hydrogen audio worse in other direction the whole place is a clown show of "Only use 80 ~ 160kbps, Why would u use over 256kbps?". Then why does Opus, Vorbis, AAC, MPC & etc all allow 512kbps stereo with no lowpass filter, Did they forget with lossless many common genres will be 900 ~ 1390kbps and with most samples 320kbps is not enough. The joint can be Reddit level ignorant acting like lossless compression would give 40 ~ 70% compression on everything, 330 ~ 500kbps is much better than having a single 1+ hour song be 1.1 mbit?.

 

[71 dB]

Lossless is highly inefficient many Rock/metal/pop average 1mbit just 5 ~ 8% less than Wav file on the CD?.

CD audio bitrate is 44 100 samples/s * 16 bits/sample * 2 audio channels = 1 411 200 bits/s. If a lossless encoded song averages at 1 000 000 bits/s, the lossless file is about 71 % of what the wav file ripped from the CD is.

 

[bigshot]

And lossy might be a small fraction of that while not sounding any different to human ears.

 

[PhonoPhi]

And lossy might be a small fraction of that while not sounding any different to human ears.

True, the difference may be undistinguishable for the best compressed/lossy files, so if one to enjoy them directly - it can be perfectly fine.
For any post-processing, lossless is better to assure that no artifacts of double-processing can occur, so I always buy and use .flacs since I do not want to have two libraries.
With more testing runs, my tentative boundary for distinguishing lossy is ~192, but then using parametric equalizers and up-conversion, the boundary is going up to ~256 (or to state more precisely with my limited evidence (and in the last couple of months I spend at least 20+ hours trying to test), many 192 noticeably "fall apart", while only some 256, and I have not tested any "in-between").
Double-blind tests are really tricky but insightful, with some treble-rich files, 128/160 can sound more appealing but then still largely distinguishable.

 

[bigshot]

I have one library for listening. It's AAC 256 VBR. That is no different for the purposes of listening to music while applying normal equalization and DSPs than lossless. I suppose if I was going to remaster something, I'd need more, but for that I would prefer a pro tools session with 24/96 files tracked in separate channels.

 

[71 dB]

And lossy might be a small fraction of that while not sounding any different to human ears.

Of course, but something looked off about Blackwoof's math so I "fact-checked" the numbers.

 

[Jazzme]

Flac is way more ahead than 320k, flac is lossless and all mp3 has loss

 

[71 dB]

Flac is way more ahead than 320k, flac is lossless and all mp3 has loss

Yes, but the question is can you hear the sonic artefacts caused by lossy encoding? Where is the bitrate threshold of no perceptible difference between lossy and lossless? There seems to be some sort of consensus that when the bitrate of lossy formats go higher than 192 kbps, it becomes really challenging to hear the differences.

 

[bigshot]

Things your ears can't hear just don't matter. Audiophools spend more time worrying about sound they can hear than they do the sound they can actually hear.

 

[peskypesky]

Flac is way more ahead than 320k, flac is lossless and all mp3 has loss

But most people, probably 98% or more, can't hear a difference.

 

[bigshot]

I still haven't found anyone in the 1 or 2% who can. I suspect if they exist, even they will admit that the difference is too small to matter.

 

[Blackwoof]

I still haven't found anyone in the 1 or 2% who can. I suspect if they exist, even they will admit that the difference is too small to matter.

Yup the only content that trips up 256kbps lossy seems to be Noise based music, But Vorbis/MPC/Lame(Allshort), Opus seem immune to those samples while AAC isn't. But even then I can just encode the problem cases at 256kbps Vorbis with the rest being AAC at 256kbps if needed. With how brick walled, much modern music is when encoded in Flac the average bitrate is 1000 ~ 1415kbps. With TAK codec some albums even reached 1950kbps despite being 16bit at 44.1KHz!.

 

[bigshot]

AAC should be able to handle noise pretty well if you do 320 VBR. It's the only codec I know where VBR allows the data rate to exceed 320. MP3 doesn't do that. But compression artifacts in noise might actually be an improvement!

 

[Blackwoof]

AAC should be able to handle noise pretty well if you do 320 VBR. It's the only codec I know where VBR allows the data rate to exceed 320. MP3 doesn't do that. But compression artifacts in noise might actually be an improvement!

FHG AAC seems to fare even better at 256kbps VBR with the updated encoder used in the EZ converter.

 

[bigshot]

I'll have to pull out my Throbbing Gristle LPs and rip them.