Spotify to launch 'Hi-Fi' CD Quality Tier.

[Pearljam5000]

My hearing cuts off at 16khz. I hear no audible differences using my HD800s.

I understand there may be some very slight differences that could be possibly be detected if someone really concentrates and really wants to hear the differences, but nothing that would make me feel I am missing out.

Bass also sounds different
It's punchier and more powerful
Overall clarity it better

 

[Sal1950]

where I will down sample a 24 bit track it to MP3, play it back to you blind, without comparison. Y

I don't see the point there. I'm sure without a comparison there's all sort of distortions that could be introduced that most still wouldn't be
able to pick out blind.
OTOH you have to really be in love with their UI to put up with a known source of lesser quality, from a company that's been stringing many users
along for 5+ years now, when others are offering lossless + some multich streams for the same or even less money ???
I hung in there with them for a number of years with their lies but finally said, enough is enough.
It's your dollar..

 

[Brian Hall]

Spotify's algorithm just found another singer I like that I had never heard of. I was listening to a generated playlist featuring Omnimar and it included some songs by Ari Mason. I doubt if I would have ever heard her music without Spotify's discovery algorithm.

I checked and her albums are available on Bandcamp. Going to buy a couple of them now.

 

[masterhw]

I watched the Amir video as I hadn't previously and it was interesting. He begins most directly addressing this topic around 21 minutes in.
My summary:

• Amir's bar for full "transparency" is transparent to all listeners on all material in all situations
• His professional background allows him to identify various types of audible artifacts which the overwhelming majority of audiophiles or musicians could not
• The only way for him to reliably do this on decent quality encodes is to A/B specific sections of a track, and presumably at high volume
• Echoic memory is under 5 seconds, so without the ability to rapidly flip back and forth (in an AB) you cannot reliably differentiate off memory alone

So I suppose you're both right, without any reference it would be exceedingly difficult to identify a decent lossy encode from lossless. But this doesn't mean it meet's Amir's bar for fully transparent either, and given how affordable it is now to distribute in lossless then Amir would choose it.

 

[PristineSound]

and given how affordable it is now to distribute in lossless then Amir would choose it.

1000%, hence I look forward to the new lossless Spotify. Because the recommendation engine and playlist is significantly better than Qobuz and Tidal.

 

[sigbergaudio]

What you described is the current Spotify, it's lossy, 320kbps is generally considered low bit rate.

Per ChatGPT, the upcoming Spotify "Music Pro" is CD quality 1,411 kbps.

View attachment 452673

1411kbps is uncompressed. It can be losslessly compressed to around 900kbps on average. So to compare 320 vs 1411 is making the difference in bitrate sound even higher than it actually is. "generally considered" sounds like an unvalidated statement.

Not at all. Nothing I said is contradictory. I am making two distinct points (1) there is absolutely an audible different (2) but without doing a back to back comparison, you simply can't tell

And I am contesting your first point.

Amir already have hard evidence in the form of Foobar ABX log that he can tell the difference between MP3 vs CD quality 16 bit vs 24 bit. With random song, Amir demonstrated that with his knowledge of audio compression and signal processing, he knows what to identify. But can he do this for every single track that ever existed? Probably not, but I think he can do it for most.

That you can tell 128kbps MP3 from CD does not prove that you can tell the difference between 320kbps Vorbis and CD. They are vastly different.

I personally have tried it on 3 random tracks and I can identify all 3, but that can just be luck.

Please elaborate.

 

[Brian Hall]

I will upgrade to Spotify's new plan when/if they provide a lossless CD quality tier. I am not opposed to lossless.

I do think "Hi-Rez" above CD quality is a useless scam for playback.

 

[sigbergaudio]

Bass also sounds different

No.

It's punchier and more powerful

No.

Overall clarity it better

No.

 

[Pearljam5000]

No.



No.



No.

You forgot to add "IMO"

 

[PristineSound]

1411kbps is uncompressed. It can be losslessly compressed to around 900kbps on average. So to compare 320 vs 1411 is making the difference in bitrate sound even higher than it actually is. "generally considered" sounds like an unvalidated statement.

I think this has morphed into some sort of confusion. Bottom line, today's Spotify is both lossy and low bit rate. The reported upcoming is higher bit rate of CD quality and lossless.

Their are audible difference under certain conditions as demonstrated by Amir between the various permutations of different bit rates and lossy, lossless.

And I am contesting your first point.

Then you are also contesting what Amir is stating on the video. Since this topic has been pontified many many times in many many places, further pontification will not benefit anyone of anything. Therefore, I recommend you experiment it yourself and judge for yourself and then leave it at that.

That you can tell 128kbps MP3 from CD does not prove that you can tell the difference between 320kbps Vorbis and CD. They are vastly different.

I didn't say 128kbps MP3 and I also said, I can tell the difference between 16bit CD quality and 24bit.

Please elaborate.

Using Fre:ac, I took a 24bit 192kHz track, down convert it to 16bit 44.1kHz and 320MP3. Used Foobar ABX comparator, I am able to pass between 9/10 and 10/10. I did this for 3 different tracks, all same results. BUT. . .I can only do this by looking for artifacts. I want to emphasize as I have multiple times, without back to back AB, there is no way, I can tell lossy vs lossless and bitrate. And I highly doubt anyone can either.

 

[sigbergaudio]

You forgot to add "IMO"

Not really. But your claims are just IMO as well, in the absence of any kind of evidence or proof. If what you are saying were true, it is strange we have seen no evidence of this. If there are clear spectral and tonal changes between 320kbps Vorbis and CD, we would all know it, and we would be able to easily pass ABX tests on the matter. This is however not the case.

 

[sigbergaudio]

Then you are also contesting what Amir is stating on the video. Since this topic has been pontified many many times in many many places, further pontification will not benefit anyone of anything. Therefore, I recommend you experiment it yourself and judge for yourself and then leave it at that.

I am not contesting anything he says in the video. I have experimented, and concluded with the same as anyone else who has experimented, it's essentially impossible to tell the difference (between Spotify Premium and Lossless).

I didn't say 128kbps MP3 and I also said, I can tell the difference between 16bit CD quality and 24bit.

Using Fre:ac, I took a 24bit 192kHz track, down convert it to 16bit 44.1kHz and 320MP3. Used Foobar ABX comparator, I am able to pass between 9/10 and 10/10. I did this for 3 different tracks, all same results. BUT. . .I can only do this by looking for artifacts. I want to emphasize as I have multiple times, without back to back AB, there is no way, I can tell lossy vs lossless and bitrate. And I highly doubt anyone can either.

There seem to be some kind of semantics in the way here, or we are discussing two different things, or you have forgotten where this started, I'm not sure which.

My initial statement in reply to @Pearljam5000 was "I assure you that you cannot reliably hear the difference between 320kbps Spotify and lossless."

Then you replied in a way that made it seem to me that you disagreed, as your claim was "With back to back, such as ABX, with training, one can very very reliably tell the difference as does Amir can."

You have yet to provide evidence that Amir (or anyone else) can do that.

 

[majingotan]

There seem to be some kind of semantics in the way here, or we are discussing two different things, or you have forgotten where this started, I'm not sure which.

My initial statement in reply to @Pearljam5000 was "I assure you that you cannot reliably hear the difference between 320kbps Spotify and lossless."

Then you replied in a way that made it seem to me that you disagreed, as your claim was "With back to back, such as ABX, with training, one can very very reliably tell the difference as does Amir can."

You have yet to provide evidence that Amir (or anyone else) can do that.

It's also the same 320Kbps SBC/AptX Bluetooth against wired lossless (Apple music). I have impromptu sighted listened a some of my peers with it and none of them can tell the difference between BT and wired at all. Spotify knows the majority (99% arbitrary guess) of their subscribers/listeners can't and don't care whatsoever about lossless. Didn't make a dent on their sales when Apple Music launched lossless and hi-res and even Dolby tracks either

You can easily hear the difference with good headphones or studio monitors.

How sure are you that the tracks are 100% volume matched? Any null test evidences?

 

[Pearljam5000]

How sure are you that the tracks are 100% volume matched? Any null test evidences?

No

 

[PristineSound]

I am not contesting anything he says in the video. I have experimented, and concluded with the same as anyone else who has experimented, it's essentially impossible to tell the difference (between Spotify Premium and Lossless).

For the benefit of everyone else, let me clarify few things. Today, Spotify has two offering. (1) Free plan (2) Premium plan (in the US at least).

The premium plan offers "high audio quality", which is very misleading as it is not high quality at all. It is of a bitrate of 320kbps. It is believed that this "high audio quality" is likely lossy, for reasons outside the scope of this conversation.

See source here: https://www.spotify.com/us/premium/


Now, in the future, it is reported by the media that Spotify will offer another plan that is a level higher than the "premium" which will have 16 bit 44.1kHz, which is a bitrate of 1,411 kbps.

What I am saying is that, there is indeed an audible difference between lossy low bit rate (such as the 320kbps currently streamed through Spotify and other formats such as MP3) vs. 16bit, 44.1 kHz vs. 24bit 44.1kHz/96kHz/192kHz. Amir in that video demonstrated such, I know it's a long video, but it's scattered all over that video (past the 20 min mark as another member has pointed out).

If you has experimented this and you cannot tell the difference, then leave it at that. There is no golden ear at play here. I do not have a golden ear and neither does Amir.

And for more context: for me, this all started when someone else made this claim online, and I laughed at it. Until I found this video of Amir, then I tried it for myself and I failed miserably. But I kept trying and trying, I used Foobar ABX comparator on two tracks second by second, and holy shit, I found an artifact, once I found that artifact, it was super easy. So no, I do not have golden ear, I just found the artifact.

There seem to be some kind of semantics in the way here, or we are discussing two different things, or you have forgotten where this started, I'm not sure which.

My initial statement in reply to @Pearljam5000 was "I assure you that you cannot reliably hear the difference between 320kbps Spotify and lossless."

First, the lossless Spotify is not out yet. Even when it comes out, there is no ABX feature on the Spotify app.

So my comment is in general: lossy vs lossless, low bit rate vs. high bit rate, one can indeed tell the difference as evident in the video with Amir. You keep saying that video has no such evidence, which clearly tells me you didn't watch it. I refuse to give you the exact timestamps of the various location on that video, I already spend time looking for this video to share, it is not my job to spoon feed you or anyone else. If you refuse to watch that video in it's entirety, then no point of further discussing and let's just leave this discussion as is.

Then you replied in a way that made it seem to me that you disagreed, as your claim was "With back to back, such as ABX, with training, one can very very reliably tell the difference as does Amir can."

You have yet to provide evidence that Amir (or anyone else) can do that.

Again, go watch the video, I refuse to spoon feed you on the timestamps on that video. It is not my job nor do I owe anyone anything.

 

[sigbergaudio]

For the benefit of everyone else, let me clarify few things. Today, Spotify has two offering. (1) Free plan (2) Premium plan (in the US at least).

The premium plan offers "high audio quality", which is very misleading as it is not high quality at all. It is of a bitrate of 320kbps. It is believed that this "high audio quality" is likely lossy, for reasons outside the scope of this conversation.

See source here: https://www.spotify.com/us/premium/

It feels like we will never agree here, but again this I think is inaccurate. How can "high quality audio" be misleading if no one on the planet can differentiate it from lossless 16bit 44.1khz audio in a normal listening situation?

Then you are saying it is likely lossy, I don't not understand what you mean by that. It most definitely is lossy.

Now, in the future, it is reported by the media that Spotify will offer another plan that is a level higher than the "premium" which will have 16 bit 44.1kHz, which is a bitrate of 1,411 kbps.

What I am saying is that, there is indeed an audible difference between lossy low bit rate (such as the 320kbps currently streamed through Spotify and other formats such as MP3) vs. 16bit, 44.1 kHz vs. 24bit 44.1kHz/96kHz/192kHz. Amir in that video demonstrated such, I know it's a long video, but it's scattered all over that video (past the 20 min mark as another member has pointed out).

It feels like you are putting any lossy format in one bucket. That is not accurate or fair.

If you has experimented this and you cannot tell the difference, then leave it at that. There is no golden ear at play here. I do not have a golden ear and neither does Amir.

And for more context: for me, this all started when someone else made this claim online, and I laughed at it. Until I found this video of Amir, then I tried it for myself and I failed miserably. But I kept trying and trying, I used Foobar ABX comparator on two tracks second by second, and holy shit, I found an artifact, once I found that artifact, it was super easy. So no, I do not have golden ear, I just found the artifact.

Here again we are either at semantics or misunderstanding each other, or I am perhaps being imprecise. My claim was: "I assure you that you cannot reliably hear the difference between 320kbps Spotify and lossless." With "reliably" I do not mean "A person who trains in a focused manner to be able to identify artifacts and then find such an artifact cannot reliably identify that artifact". I mean that they can't consistently discern if a piece of music they are listening to is Spotify Premium or lossless audio.

First, the lossless Spotify is not out yet. Even when it comes out, there is no ABX feature on the Spotify app.

When I say "lossless", I refer to any lossless 44.1khz/16bit source, not necessarily Spotify lossless. My apologies if that contributes to the confusion. With regards to conducting the test, it is possible to create or obtain 320kbps Vorbis files in other ways than through Spotify.

So my comment is in general: lossy vs lossless, low bit rate vs. high bit rate, one can indeed tell the difference as evident in the video with Amir. You keep saying that video has no such evidence, which clearly tells me you didn't watch it.

But you are apparently discussing something else than I am, no wonder we don't agree. I have not said "in general it is impossible to hear the difference between lossy and lossless. On the contrary I have defined a very specific situation with 320kbps Vorbis vs Lossless 16bit/44.1khz audio already in my first post. So I am not looking in the video for general evidence that it is possible to identify lossless audio, I am looking for specific evidence that one can identify 320kbps Vorbis vs Lossless 16bit/44.1khz audio, and so far I have found none.

I refuse to give you the exact timestamps of the various location on that video, I already spend time looking for this video to share, it is not my job to spoon feed you or anyone else. If you refuse to watch that video in it's entirety, then no point of further discussing and let's just leave this discussion as is.


Again, go watch the video, I refuse to spoon feed you on the timestamps on that video. It is not my job nor do I owe anyone anything.

It's a pretty general (written or unwritten I don't know) rule on this forum (and in general in scientific matters) that if you make the claim, it is up to you to prove it, not everyone else. "It's somewhere in this 45 minute video" is not a polite or efficient way to provide proof.

 

[voodooless]

Spotify != MP3 . MP3 has been obsolete for at least a decade now, and the others have vastly surpassed it at any bitrate.

You cannot extrapolate your high bitrate MP3 ABX to mean you’ll do just as well with high bitrate AAC or Vorbis.

 

[Joe Smith]

I used to care about this, but just don't anymore. If Spotify had moved on it years ago, I might have been willing to pay more per month, but now - no. I'm no longer chasing "perfection."

 

[PristineSound]

It's a pretty general (written or unwritten I don't know) rule on this forum (and in general in scientific matters) that if you make the claim, it is up to you to prove it, not everyone else. "It's somewhere in this 45 minute video" is not a polite or efficient way to provide proof.

So you just admitted you didn't watch the video before vehemently commenting that such evidence doesn't exist and doing so with the utmost conviction.


The video needs to be watch in it's entirety, others have watched it in it's entirety, I have watched it in it's entirety. If you don't want to watch it, that is perfectly OK. We can just leave this conversation as it is and move on.


Anyway, we are talking past each other. My bottom line is:

1. There is absolutely and scientifically an audible difference between various permutations of lossy vs lossless and bit rate with caveats.
2. Currently today, on Spotify, it is lossy and low bit rate only.
3. The caveats mentioned above is that you (a) need reference, such as AB switching, (b), you need to look for artifacts.
4. Without reference such as AB comparation, it is highly unlikely anyone can tell off the bat, what the track is lossy or lossless and it's bit rate.

 

[PristineSound]

Spotify != MP3 . MP3 has been obsolete for at least a decade now, and the others have vastly surpassed it at any bitrate.

You cannot extrapolate your high bitrate MP3 ABX to mean you’ll do just as well with high bitrate AAC or Vorbis.

Currently Spotify is lossy, yes? Lossy is lossy, you lose information, period, end of story, case close.

There is an audible difference between high res vs CD quality, let alone CD quality vs lossy format.

Anyway, I stated my point and provided the evidence, should one choose not to evaluate the evidence, I cannot help with that. My last comment on this topic.