FLAC still cuts out part of the signal. It’s limited to 20khz.
Bhat’s typically well above the limit of an adults hearing, especially someone old enough with enough money and equipment to be considered an audiophile.
No, it doesn’t. Digital PCM audio, as a concept, can only represent frequencies up to the sample rate used. Which can be anything. Typically 44kHz.
Going above that is pointless as humans are unable to perceive the ultrasonic frequencues that would unnecessarily include.
Lossless doesn’t mean “perfect recording”. By that logic lossless images or videos aren’t lossless, because they don’t include an infinite amount of pixels between every pixel, representing every photon that was captured.
Lossless refers to data-retention, not reality retention.
Lossless converting a CD to FLAC. But that CD was recorded at 44hkz sampling rate, which gives you a maximum frequency of 22khz. You have lost audio above 22khz. Children can theoretically hear frequencies higher than this, but typically adults cannot.
FLAC doesn’t cut anything out though. Whatever input you use, FLAC compresses losslessly. You can use 96kHz 24bit recordings and the resulting FLAC file can be decompressed back into a bit-perfect copy of the original.
In the OP, the messages in red are correct. FLAC is like a ZIP file designed to be more effective at compressing audio files. And just like a ZIP file, you can reconstitute the original file exactly. There’s no data lost in compression.
Yes if you’re transcoding a CD to FLAC it’s lossless. That’s not what I’m talking about. I’m talking about the process of digitally recording the audio in the first place.
Nevermind the fact that nobody seems to have paid any attention to the original joke which is that the boomers who can afford high end stuff can’t even hear the difference.
I dont think you understand the difference between a lossless file format/encoding algorithm and “losless” recording/storing of signals. If anyone ever speaks of a lossless encoding algorith theyy mean that avter encoding and decoding the input and output will be the same e.g nithing was lost. Why would the recording have annything to do with the lossyness of the encoding algorithm? If the music was made digitally there would be no loss in any sense since the output of your daw or midi file etc is already digital. Btw in general you just cant record any arbitrary analog signal but you can record a lot of it. You will also never in no media be able to store the exact signal. There is always noise always some variation. Even if you store your signal analog there is only so much variance of the magnetic field in a tape and only so many atoms of height difference in the groove of your vynil. The thought of lossless recording is just dumb if you think about it because you change the signal by measuring it annyway so what even is the “original” signal?
You can encode at higher bit depths and sample rates. I have music I’ve bought at 24bit 48Khz. (I know I won’t ever be able to hear the difference between that and the more common 16bit 44.1Khz.) I think you can go up to 96Khz, although I’m not sure I’ve actually seen it before.
Even uncompressed audio cuts out frequencies. With digital audio capture it is impossible to capture everything. There will always be a floor and a ceiling. In the case of flac it’s typically 20-24hkz.
Audiophiles have moved onto “high res lossless” because regular lossless wasn’t good enough for them.
The “high res lossless” you’re referring to, is still FLAC. FLAC has no downside. Whatever PCM audio you want, it can represent perfectly, while using less storage.
FLAC doesn’t “limit” or “cut out” anything unless you or the software you’re using is reducing the bit depth or samplerate of the source PCM waveform.
Which is something you might want to do, since it will reduce file size significantly to not use a higher samplerate than necessary. But FLAC itself doesn’t do or require that.
On new formats, you might be thinking of MQA, which supposedly encodes the contents of a higher samplerate PCM waveform into a lower samplerate file, but it has been proven to be largely snake oil, and lossy as hell in terms of bit integrity.
And this is because audiophiles don’t understand why the audio master is 96 kHz or more often 192 kHz. You can actually easily hear the difference between 48, 96 and 192 kHz signals, but not in the way people usually think, and not after the audio has been recorded – because the main difference is latency when recording and editing. Digital sound processing works in terms of samples, and a certain amount of them have to be buffered to be able to transform the signal between time and frequency. The higher the sample rate, the shorter the buffer, and if there’s one thing humans are good at hearing (relatively speaking) it’s latency.
Digital instruments start being usable after 96 kHz as the latency with 256 samples buffered gets short enough that there’s no distracting delay from key press to sound. 192 gives you more room to add effects and such to make the pipeline longer. Higher sample rate also makes changing frequencies, like bringing the pitch down, simpler as there’s more to work with.
But after the editing is done, there’s absolutely no reason to not cut the published recording to 48 or 44.1 kHz. Human ears can’t hear the difference, and whatever equipment you’re using will probably refuse to play anything higher than 25 kHz anyways, as e.g. the speaker coils aren’t designed to let higher frequency signals through. It’s not like visual information where equipment still can’t match the dynamic range of the eye, and we’re just starting to get to a pixel density where we can no longer see a difference between DPIs.
Flac is literally lossless in the mathematical sense.
Yeah but my ears don’t care about mathematics.
Some audiophile, probably…
FLAC still cuts out part of the signal. It’s limited to 20khz.
Bhat’s typically well above the limit of an adults hearing, especially someone old enough with enough money and equipment to be considered an audiophile.
LOL… FLAC happily handles 192kHz
And at 327,675hz (the maximum for FLAC) you can still be missing out anything 327,676hz and greater. But that won’t stop the audiophiles.
Based
No, it doesn’t. Digital PCM audio, as a concept, can only represent frequencies up to the sample rate used. Which can be anything. Typically 44kHz.
Going above that is pointless as humans are unable to perceive the ultrasonic frequencues that would unnecessarily include.
Lossless doesn’t mean “perfect recording”. By that logic lossless images or videos aren’t lossless, because they don’t include an infinite amount of pixels between every pixel, representing every photon that was captured.
Lossless refers to data-retention, not reality retention.
FLAC is totally lossless. You can rip a CD to 44kHz WAV, compress it to FLAC, and then decompress it and get a bit-perfect copy of the original WAV.
Lossless converting a CD to FLAC. But that CD was recorded at 44hkz sampling rate, which gives you a maximum frequency of 22khz. You have lost audio above 22khz. Children can theoretically hear frequencies higher than this, but typically adults cannot.
https://en.wikipedia.org/wiki/Nyquist–Shannon_sampling_theorem#%3A~%3Atext=If+the+essential%2CNyquist+interval.
FLAC doesn’t cut anything out though. Whatever input you use, FLAC compresses losslessly. You can use 96kHz 24bit recordings and the resulting FLAC file can be decompressed back into a bit-perfect copy of the original.
In the OP, the messages in red are correct. FLAC is like a ZIP file designed to be more effective at compressing audio files. And just like a ZIP file, you can reconstitute the original file exactly. There’s no data lost in compression.
Yes if you’re transcoding a CD to FLAC it’s lossless. That’s not what I’m talking about. I’m talking about the process of digitally recording the audio in the first place.
Nevermind the fact that nobody seems to have paid any attention to the original joke which is that the boomers who can afford high end stuff can’t even hear the difference.
I dont think you understand the difference between a lossless file format/encoding algorithm and “losless” recording/storing of signals. If anyone ever speaks of a lossless encoding algorith theyy mean that avter encoding and decoding the input and output will be the same e.g nithing was lost. Why would the recording have annything to do with the lossyness of the encoding algorithm? If the music was made digitally there would be no loss in any sense since the output of your daw or midi file etc is already digital. Btw in general you just cant record any arbitrary analog signal but you can record a lot of it. You will also never in no media be able to store the exact signal. There is always noise always some variation. Even if you store your signal analog there is only so much variance of the magnetic field in a tape and only so many atoms of height difference in the groove of your vynil. The thought of lossless recording is just dumb if you think about it because you change the signal by measuring it annyway so what even is the “original” signal?
You began this by saying
Recording from analog to digital is lossy, in the same way as previously described about images. But this has nothing to do with FLAC.
That’s the entire yoke.
You can encode at higher bit depths and sample rates. I have music I’ve bought at 24bit 48Khz. (I know I won’t ever be able to hear the difference between that and the more common 16bit 44.1Khz.) I think you can go up to 96Khz, although I’m not sure I’ve actually seen it before.
I’ve seen, and I have 24bit 96khz files.
They’re less common than your average 16bit 42khz, but they do exist.
Isn’t 44.1 KHz more common?
Yeah I fixed it idk why I thought 42
No its not lol
Even uncompressed audio cuts out frequencies. With digital audio capture it is impossible to capture everything. There will always be a floor and a ceiling. In the case of flac it’s typically 20-24hkz.
Audiophiles have moved onto “high res lossless” because regular lossless wasn’t good enough for them.
The “high res lossless” you’re referring to, is still FLAC. FLAC has no downside. Whatever PCM audio you want, it can represent perfectly, while using less storage.
FLAC doesn’t “limit” or “cut out” anything unless you or the software you’re using is reducing the bit depth or samplerate of the source PCM waveform.
Which is something you might want to do, since it will reduce file size significantly to not use a higher samplerate than necessary. But FLAC itself doesn’t do or require that.
On new formats, you might be thinking of MQA, which supposedly encodes the contents of a higher samplerate PCM waveform into a lower samplerate file, but it has been proven to be largely snake oil, and lossy as hell in terms of bit integrity.
And this is because audiophiles don’t understand why the audio master is 96 kHz or more often 192 kHz. You can actually easily hear the difference between 48, 96 and 192 kHz signals, but not in the way people usually think, and not after the audio has been recorded – because the main difference is latency when recording and editing. Digital sound processing works in terms of samples, and a certain amount of them have to be buffered to be able to transform the signal between time and frequency. The higher the sample rate, the shorter the buffer, and if there’s one thing humans are good at hearing (relatively speaking) it’s latency.
Digital instruments start being usable after 96 kHz as the latency with 256 samples buffered gets short enough that there’s no distracting delay from key press to sound. 192 gives you more room to add effects and such to make the pipeline longer. Higher sample rate also makes changing frequencies, like bringing the pitch down, simpler as there’s more to work with.
But after the editing is done, there’s absolutely no reason to not cut the published recording to 48 or 44.1 kHz. Human ears can’t hear the difference, and whatever equipment you’re using will probably refuse to play anything higher than 25 kHz anyways, as e.g. the speaker coils aren’t designed to let higher frequency signals through. It’s not like visual information where equipment still can’t match the dynamic range of the eye, and we’re just starting to get to a pixel density where we can no longer see a difference between DPIs.
If that’s happening you need to fix your transcoder settings.