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: Apple moves to lossless audio is really making more audiofools (Read 12689 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Re: Apple moves to lossless audio is really making more audiofools

Reply #26
Audiofools can do what they want. I'm more than happy with my transparent lossy music.

How do you know it's transparent if you don't have the lossless original to compare?

Never heard of ABX testing?

You ABX every single file? Damn.

Re: Apple moves to lossless audio is really making more audiofools

Reply #27
Audiofools can do what they want. I'm more than happy with my transparent lossy music.

How do you know it's transparent if you don't have the lossless original to compare?

Never heard of ABX testing?

You ABX every single file? Damn.

I've ABX'd hundreds. Damn.

Re: Apple moves to lossless audio is really making more audiofools

Reply #28
If you go through bluetooth it turns it to lossy audio.  You'll have to hook up directly to the playback device.
 

Re: Apple moves to lossless audio is really making more audiofools

Reply #29
Audiofools can do what they want. I'm more than happy with my transparent lossy music.

How do you know it's transparent if you don't have the lossless original to compare?

Never heard of ABX testing?

You ABX every single file? Damn.

I've ABX'd hundreds. Damn.

I can't believe this is devolving into an argument against lossless audio... What has this forum become?!

ABXing hundreds is not statistically significant to the hundreds of millions or recordings produced. Also in the era of 4K streams, lossless audio isn't demanding on even the most modest systems. Every lossy codec has killer samples, and why would you ever want to process your recording through a very complex psychoacoustic model when it's not necessary? Especially if you're paying money for it. Me, I don't want to take the chance so I always prefer lossless. It's my money at the end of the day, and I want all the bits.

Christ, you're arguing to receive worse value for money from a service...

But my point is this - unless you're ABXing every single track against the lossless original you don't know that it's transparent. You can operate with a high level of confidence, but you can't know.

Re: Apple moves to lossless audio is really making more audiofools

Reply #30
I wonder what they did before settling on 44.1/16 bit for CD. ABXing every track? Not quite.
I wonder what one would do if migrating to digital end-user format today. With 2021 storage cost ... maybe just shrugged it off as "why not just deliver the DAW format?"

But if the transition were fifteen years ago? Best hunch: a lossy format, like one does with video. (Remember, converting to CDDA is lossy anyway.)

Re: Apple moves to lossless audio is really making more audiofools

Reply #31
But my point is this - unless you're ABXing every single track against the lossless original you don't know that it's transparent. You can operate with a high level of confidence, but you can't know.
Do fluorescent pink cats exists? Will the sun rise another time tomorrow? Do human lives more than 150 years?
We all have answers to these questions because most of our knowledge is based on high level of confidence. Even some scientific principles are changing with different paradigms and updated theories.

So you don't have to ABX everything to know if something is transparent to your ears: high level of confidence based on a small amount of tests and/or test done by others is enough to make that claim with confidence.

BTW, transparency doesn't necessary mean “identical to the reference on a blind comparison”. I often use the word transparent when a lossy audio file sound natural, regular with nothing weird, or when a picture or movie doesn't show any visible artifact or obvious sign of reduced sharpness. It's transparent when nothing wrong is detectable. 256 kbps AAC/OPUS/MPC is transparent and will stay transparent even if you can find a tiny difference on a couple of samples.

A window is obviously transparent but if you check carefully it absorbs a tiny amount of light and slight distortions might appear on a closer look. Everyone would agree but for everyone a window is still transparent. So even in everyday language the word “transparency” does not even mean absolutely perceptually identical :)


Quote
I can't believe this is devolving into an argument against lossless audio... What has this forum become?!
This forum is based from day one on listening tests and statistics. In other words, in high level of confidence only. Nothing has changed for 20 years  ;)


Re: Apple moves to lossless audio is really making more audiofools

Reply #33
I wonder what they did before settling on 44.1/16 bit for CD. ABXing every track? Not quite.
I wonder what one would do if migrating to digital end-user format today. With 2021 storage cost ... maybe just shrugged it off as "why not just deliver the DAW format?"

But if the transition were fifteen years ago? Best hunch: a lossy format, like one does with video. (Remember, converting to CDDA is lossy anyway.)

That is a FAR simpler model to verify than anything using these advanced psychoacoustics.

Also, CDDA isn't relevant here. PCM, yes, but CDDA seems like an odd phrase to use. I don't care about the physical media.

Re: Apple moves to lossless audio is really making more audiofools

Reply #34
But my point is this - unless you're ABXing every single track against the lossless original you don't know that it's transparent. You can operate with a high level of confidence, but you can't know.
Do fluorescent pink cats exists? Will the sun rise another time tomorrow? Do human lives more than 150 years?
We all have answers to these questions because most of our knowledge is based on high level of confidence. Even some scientific principles are changing with different paradigms and updated theories.

So you don't have to ABX everything to know if something is transparent to your ears: high level of confidence based on a small amount of tests and/or test done by others is enough to make that claim with confidence.

BTW, transparency doesn't necessary mean “identical to the reference on a blind comparison”. I often use the word transparent when a lossy audio file sound natural, regular with nothing weird, or when a picture or movie doesn't show any visible artifact or obvious sign of reduced sharpness. It's transparent when nothing wrong is detectable. 256 kbps AAC/OPUS/MPC is transparent and will stay transparent even if you can find a tiny difference on a couple of samples.

A window is obviously transparent but if you check carefully it absorbs a tiny amount of light and slight distortions might appear on a closer look. Everyone would agree but for everyone a window is still transparent. So even in everyday language the word “transparency” does not even mean absolutely perceptually identical :)


Quote
I can't believe this is devolving into an argument against lossless audio... What has this forum become?!
This forum is based from day one on listening tests and statistics. In other words, in high level of confidence only. Nothing has changed for 20 years  ;)

You've got the wrong definition of 'transparent' then. Transparent in this context means audibly identical to the original. There is no use in telling me that a "lossy audio file sound natural" - this is useless. What if the original recording didn't sound natural?

I'm not sure where your definition comes from but I do not agree with it at all.

Re: Apple moves to lossless audio is really making more audiofools

Reply #35
ABXing hundreds is not statistically significant to the hundreds of millions or recordings produced.
Huh?
https://www.tools4dev.org/resources/how-to-choose-a-sample-size/


I think you've got that the wrong way around. This article essentially says that you need 'x' number of people to verify that a single transcode is transparent. Not that 'x' number of transcodes proves codec transparency. At least that's my interpretation of it.

Also, this should not be a random selection of a population, it needs critical listeners, not Joe Public. If that was not the case we'd all be using 128kbps Xing MP3s.

My goodness, there was a sample posted on this very forum a few months ago that totally killed (I think it was LAME) at any bit rate. It was also an EXTREMELY commonly used technique - if I remember correctly it was a kick drum over white noise. There's a huge number of so called "killer samples" posted on this forum. We tend to brush them off as outliers (like no one listens to Black Sabbath or Autechre).

All I'm saying is that this is not something I need to worry about when using a lossless codec. I know 100% of all samples in a lossless codec are as they are intended to be.

I've edited this response a lot however one last thing comes to mind. If this is such a well understood topic, can someone please tell me the percentage of non-transparent encodes in Apple Music? 256kbps AAC, 75 million files. We know it's very good, but we also know there are "killer samples". This is a question that can't be asked when using lossless.

Re: Apple moves to lossless audio is really making more audiofools

Reply #36
I wonder what they did before settling on 44.1/16 bit for CD. ABXing every track? Not quite.

 It was based on the Nyquist-Shannon sampling theorem.

"The Nyquist–Shannon sampling theorem says the sampling frequency must be greater than twice the maximum frequency one wishes to reproduce. Since human hearing range is roughly 20 Hz to 20,000 Hz, the sampling rate had to be greater than 40 kHz.

In addition, signals must be low-pass filtered before sampling to avoid aliasing. While an ideal low-pass filter would perfectly pass frequencies below 20 kHz (without attenuating them) and perfectly cut off frequencies above 20 kHz, such an ideal filter is theoretically and practically impossible to implement as it is non-causal, so in practice a transition band is necessary, where frequencies are partly attenuated. The wider this transition band is, the easier and more economical it is to make an anti-aliasing filter. The 44.1 kHz sampling frequency allows for a 2.05 kHz transition band."

I'm pretty sure the human hearing range hasn't evolved above 20,000 kHz in the past 100 years, but I know mine hasn't.

Re: Apple moves to lossless audio is really making more audiofools

Reply #37
I wonder what they did before settling on 44.1/16 bit for CD. ABXing every track? Not quite.

 It was based on the Nyquist-Shannon sampling theorem.

"The Nyquist–Shannon sampling theorem says the sampling frequency must be greater than twice the maximum frequency one wishes to reproduce. Since human hearing range is roughly 20 Hz to 20,000 Hz, the sampling rate had to be greater than 40 kHz.

In addition, signals must be low-pass filtered before sampling to avoid aliasing. While an ideal low-pass filter would perfectly pass frequencies below 20 kHz (without attenuating them) and perfectly cut off frequencies above 20 kHz, such an ideal filter is theoretically and practically impossible to implement as it is non-causal, so in practice a transition band is necessary, where frequencies are partly attenuated. The wider this transition band is, the easier and more economical it is to make an anti-aliasing filter. The 44.1 kHz sampling frequency allows for a 2.05 kHz transition band."

I'm pretty sure the human hearing range hasn't evolved above 20,000 kHz in the past 100 years, but I know mine hasn't.

Exactly, this is something that can be mathematically verified. The transparency of a lossy codec cannot - it must be manually tested.

Also, it's 22.05 kHz and 20,000 Hz.

Re: Apple moves to lossless audio is really making more audiofools

Reply #38
20.000 Hz + 2050 Hz transition band?
TheWellTemperedComputer.com

Re: Apple moves to lossless audio is really making more audiofools

Reply #39
It was based on the Nyquist-Shannon sampling theorem.
[...]I'm pretty sure the human hearing range hasn't evolved above 20,000 kHz in the past 100 years, but I know mine hasn't.

Unless you are in your mid-teens I'm sure yours has deteriorated. I am not sure if your sarcasm detector ever was better, nor your knowledge on the topic. But by all means, feel free to enlighten us at what the sampling theorem says about the sufficient number of bits per sample for useful music reproduction.

Does the sampling theorem tell us that there should be sampling rates of 48k - and if not, do you think those who implemented it just were more ignorant than you are?  Does the sampling theorem tell you anything about what analogue filters they had to use how deep into the audible range for aliasing of not to be audible in actual music? Does the sampling theorem tell us that yes, scientific studies have identified subjects who can in fact detect tones as high as 28 kHz?

Does the sampling theorem tell us that you can not hear those 28 kHz notes when masked by music-level sub-20k tones - and whether that ability sticks around for long if you listen to it at the volume in question day in and day out?

No, it does not.

Even still, I bet that if the transition to digital end-user formats happened after the findings of Ashihara and collaborators - who by the way, are more famous for debunking results claiming ultrasonics were essential - were established, one would have gone to a sampling rate of say, 64k. Just because it would be an easy-peasy cheap thing to do around 2007. Not so when the CD standard was frozen.

Re: Apple moves to lossless audio is really making more audiofools

Reply #40
This topic is totally off the rails now because HydrogenAud.io is full of know-it-alls. This place is such a mess.

Re: Apple moves to lossless audio is really making more audiofools

Reply #41
You've got the wrong definition of 'transparent' then. Transparent in this context means audibly identical to the original. There is no use in telling me that a "lossy audio file sound natural" - this is useless. What if the original recording didn't sound natural?

I'm not sure where your definition comes from but I do not agree with it at all.
What is exactly the context we are talking about on the transparency word?

In a context of A/B comparison, I agree with you, transparent mean indistinguishable.

But In a context of music listening (e.g. streaming from a service that provides lossy files, or mass-encoded lossy files stored and played on a device), transparent have a different meaning. Because as you said, you can not compare (for obvious practical reason) what you're listening to any tangible reference. So a lossy file is claimed transparent when nothing wrong occurs. By wrong, I mean any defect usually caused by lossy encoders (pre-echo, ringing for audio, macrobloc or blurring for pictures). I remind that many people on this forum are listening for decades transparent MP3, MPC, Vorbis, AAC, LossyFLAC, WavPack Audio. None of them are ABXing everything they listen towards the CD Reference. I'm sure that a vast majority of them haven't even tried to ABX something. Some DACS are resampling 44.1 KHz to 48 KHz in a transparent way. It's not technically possible to ABX it  and you can note that nobody seriously question the transparency of this process. In most case, you don't have to prove that something is transparent: but you have to prove that something isn't (TOS #8).

So I believe that your definition of transparency is too restrictive and doesn't work outside the ABX workshop.


So when peskypesky claims on 2021-05-21 19:05:21, “Audiofools can do what they want. I'm more than happy with my transparent lossy music” I'm sure he exactly means that. To answer your question, which was “How do you know it's transparent if you don't have the lossless original to compare?”, transparent means excellent without perceptual issue.

Re: Apple moves to lossless audio is really making more audiofools

Reply #42
I can already see some problems in the future

> offer free lossless music
> lots of people start using it because bigger is always better
> network corporations see this
> people are using too much data grrr
> increase prices

We have already seen something similar when Corona virus started.
Whole internet got overloaded and they were forced to reduce bitrates.
Our infrastructure just cannot take it.
There is no reason to offer lossless music to people who cannot tell a difference between 128 kbit/s CBR MP3 and lossless.
Some people might say "hey, this will force them to upgrade infrastructure" and all I can say is big LOL.
gold plated toslink fan

 

Re: Apple moves to lossless audio is really making more audiofools

Reply #43
Lately MQA Bob Stuart mentioned in an interview how his revolutionary format besides many other fantastic things will save our planet.
The bandwith this lossy encrypting saves helps to reach the climate target.
In reality it saves no bandwith against a similar shrinked PCM file and needs extra power on every decode.
When BS can argue that way we must consider 256k AAC may save 2 planets against full lossless streaming.
Is troll-adiposity coming from feederism?
With 24bit music you can listen to silence much louder!

Re: Apple moves to lossless audio is really making more audiofools

Reply #44
I can already see some problems in the future

> offer free lossless music
> lots of people start using it because bigger is always better
> network corporations see this
> people are using too much data grrr
> increase prices

We have already seen something similar when Corona virus started.
Whole internet got overloaded and they were forced to reduce bitrates.
Our infrastructure just cannot take it.
There is no reason to offer lossless music to people who cannot tell a difference between 128 kbit/s CBR MP3 and lossless.
Some people might say "hey, this will force them to upgrade infrastructure" and all I can say is big LOL.

Wouldn't the solution be to offer lossless downloads in their store to purchase that cut down on overall traffic with exclusive bonus materials like vouchers that could be redeemed for concerts and other fun stuff?

Re: Apple moves to lossless audio is really making more audiofools

Reply #45
Lately MQA Bob Stuart mentioned in an interview how his revolutionary format besides many other fantastic things will save our planet.
The bandwith this lossy encrypting saves helps to reach the climate target.
In reality it saves no bandwith against a similar shrinked PCM file and needs extra power on every decode.
Both you and I know that MQA is shit. I don't even want to talk about it.

When BS can argue that way we must consider 256k AAC may save 2 planets against full lossless streaming.
Unironically, this is somewhat true.

Wouldn't the solution be to offer lossless downloads in their store to purchase that cut down on overall traffic with exclusive bonus materials like vouchers that could be redeemed for concerts and other fun stuff?
If you simply want music, use 256 kbit/s AAC version.
Are you 100% sure you know what you are doing? Buy lossless version.

From my perspective, all I can see is a smart move from Apple.
Add some clever "passive-aggressive" marketing that promotes hi-res lossless audio so people will start using it.
Guess what? You just ran out of storage space! iPhone does not have SD card slot. What should you do?
You should buy new iPhone with more storage space that is very expensive!
gold plated toslink fan

Re: Apple moves to lossless audio is really making more audiofools

Reply #46
If you simply want music, use 256 kbit/s AAC version.
Are you 100% sure you know what you are doing? Buy lossless version.

From my perspective, all I can see is a smart move from Apple.
Add some clever "passive-aggressive" marketing that promotes hi-res lossless audio so people will start using it.
Guess what? You just ran out of storage space! iPhone does not have SD card slot. What should you do?
You should buy new iPhone with more storage space that is very expensive!

I use an Android based phone as almost nothing Apple offers appeals to me.  I'm still pretty much in the Android (phone) and Windows (desktop/laptop) ecosystems.  Getting sold music in a lossy codec feels more like vender lock-in, in a way than getting sold a lossless encoded copy that can be converted to whatever I like or need as it's often wiser to leave a lossy encoded file alone.

Also wouldn't their software be able to do encodings straight to AAC / LC if the user setup it to do so, if space was an issue for them?  I had an iPod many years ago and their iTunes software could do that, although I never used the feature.  It's not hard to imagine their phones could be setup to prefer AAC / LC over Apple Lossless offered in their store and the end user could setup different devices to use whatever one they wanted to.

Re: Apple moves to lossless audio is really making more audiofools

Reply #47
BTW, transparency doesn't necessary mean “identical to the reference on a blind comparison”. I often use the word transparent when a lossy audio file sound natural, regular with nothing weird, or when a picture or movie doesn't show any visible artifact or obvious sign of reduced sharpness. It's transparent when nothing wrong is detectable. 256 kbps AAC/OPUS/MPC is transparent and will stay transparent even if you can find a tiny difference on a couple of samples.

Yup, It only took 10 files for me to stick with 160kbps Opus/MPC on my phone. But that won't stop people claiming lossy sucks because they can tell 160kbps area and refuse to check if they can do 256kbps. With how non classical music compresses with lossless that 1mbit average for rock/metal/pop/electro. I wouldn't switch to 16/44 Spotify since I fail to ABX 320kbps Vorbis & MP3.

Got locked out on a password i didn't remember. :/

Re: Apple moves to lossless audio is really making more audiofools

Reply #48
I use an Android based phone as almost nothing Apple offers appeals to me.  I'm still pretty much in the Android (phone) and Windows (desktop/laptop) ecosystems.
Same.

Getting sold music in a lossy codec feels more like vender lock-in, in a way than getting sold a lossless encoded copy that can be converted to whatever I like or need as it's often wiser to leave a lossy encoded file alone.
You are not paying for lossless or lossy copy. We are talking about Apple Music, streaming service, not iTunes Music Store. Music that you buy on iTunes is always 256 kbit/s AAC-LC.

Edit:
On this Tidal thread https://hydrogenaud.io/index.php?topic=107645.msg883286#msg883286
you can see that global moderator Kohlrabi had the same concerns as me.
gold plated toslink fan

Re: Apple moves to lossless audio is really making more audiofools

Reply #49
You are not paying for lossless or lossy copy. We are talking about Apple Music, streaming service, not iTunes Music Store. Music that you buy on iTunes is always 256 kbit/s AAC-LC.

Edit:
On this Tidal thread https://hydrogenaud.io/index.php?topic=107645.msg883286#msg883286
you can see that global moderator Kohlrabi had the same concerns as me.


Basically a move by Big Tech to try to push everyone to more profitable streaming services and to phase out and eventually kill off download services.  It's all about the money.  Apple is been doing that for years with the iTunes software.