Anybody Out There Who Hears a Difference Between 320 kbps MP3 and Red Book CD? What Differences Do You Hear?

[Robbo99999]

I am not aware of any commercial applications of mp3 that use bitrates above 320 kbps, primarily for compatibility reasons - some players may not be designed to handle mp3 rates above that.



Is it possible Amazon is pre-loading the next track while the current track is still playing?

I've found out some more info on Amazon mp3 formats. I've found out they use the LAME encoder, and most reports from people are that their downloaded music from Amazon does use different formats, sometimes CBR, sometimes VBR, but most of the time it's LAME VBR V0 which means it's 240kbps VBR. I don't have any Amazon downloaded music, so I decided to purchase just one very recent track, which I'd also purchased in ITunes, so following in the screenshot you can see the Amazon downloaded mp3 file which I imported into ITunes so I could see the file details - so you can see that Amazon does indeed use LAME and at a VBR bit rate equivalent to that of the V0 setting in LAME.

What's also interesting is I compared the above file to the same track I purchased in ITunes, which you can see here:

You can see the ITunes version is 1MB larger in file size but 256kbps CBR vs the VBR 238kbps of the Amazon file. VBR is generally better quality than CBR, so my initial impression is that the Amazon download should be better quality than the ITunes download, what do we reckon? I know we're comparing AAC to LAME encoded mp3, but what do we know about the quality differences between the two?

As to how this all relates to my average download speed of the Alexa process of a high 536kbps average over the duration of any given track, I'm not sure. You'd think it would mean that it's using at least LAME V0 238kbps VBR given that download speed, but there are reports that LAME can work up to 640kbps, although as you say not many players support that.....but we're talking streaming of music here rather than just playing the downloaded file (if there's a distinction?), and I'm using the Alexa App in Windows 10 PC, so maybe that app has been made compatible to take advantage of higher than 320kbps LAME encoded files.......could it be that music streaming quality on Amazon Music is higher quality than the downloads, seems a bit unlikely?? Having said that I can't explain the increased overall download speed of the Alexa process during music streaming, because it happens the same on each track, so it can't just be loading up the next track, because you wouldn't see that pattern when listening to tracks back to back without pause.......so it does seem to be downloading at lot more data than it would need to in order to play even a 320kbps LAME mp3....which is all in contrast to the fact that purchased & downloaded Amazon music is "just" 240kbps VBR LAME V0 encoded. Hmmm!

 

[Julf]

I know we're comparing AAC to LAME encoded mp3, but what do we know about the quality differences between the two?

You should ask JJ, but AAC is definitely better than mp3.

As to how this all relates to my average download speed of the Alexa process of a high 536kbps average over the duration of any given track, I'm not sure.

The download speed might be very misleading, due to buffering, protocol overhead and many more factors.

 

[Robbo99999]

You should ask JJ, but AAC is definitely better than mp3.



The download speed might be very misleading, due to buffering, protocol overhead and many more factors.

Who's JJ?

Yeah, I'd considered protocol overhead. I think we can rule out buffering, because it's the same pattern on each track that is played even when no pause between tracks, so given no pause then even if there was buffering then this would equal out to the bitrate required by the music after you've finished playing the first track of the session. Regarding protocol overhead, I think this is a more likely explanation, but I noticed there were around 3 to 5 concurrent Alexa processes open and consuming bandwidth, with only one of those processes showing significant download speeds, so it was that particular process that I attributed to the music download itself....and makes me think the other Alexa processes are the ones that would deal with the protocol overhead (and so those processes weren't part of my calculation so haven't artificially inflated my figures). Unless Resource Monitor in windows is not accurate, quick pic of the program following (data not related to Amazon streaming):

 

[BDWoody]

Who's JJ

https://audiosciencereview.com/foru...sing-and-psychoacoustics-master-library.2066/

 

[Robbo99999]

https://audiosciencereview.com/foru...sing-and-psychoacoustics-master-library.2066/

Ah, thanks!

@j_j , what do you reckon re my questions & observations in my last couple of posts, can you shed any light on it with regards to your knowledge on the topics?
https://www.audiosciencereview.com/...what-differences-do-you-hear.5077/post-615711
https://www.audiosciencereview.com/...what-differences-do-you-hear.5077/post-615722

 

[escape2]

You can see the ITunes version is 1MB larger in file size but 256kbps CBR vs the VBR 238kbps of the Amazon file.

Just for the record, AAC doesn't have a true CBR option. If you watch an AAC CBR file being played in Foobar2000, for example, you will see that its bitrate still fluctuates some, but not as much as in case of AAC VBR.

VBR is generally better quality than CBR, so my initial impression is that the Amazon download should be better quality than the ITunes download, what do we reckon? I know we're comparing AAC to LAME encoded mp3, but what do we know about the quality differences between the two?

If someone can't tell the difference between high bitrate MP3 vs FLAC, then that someone is also unlikely to tell the difference between high bitrate MP3 vs high bitrate AAC, IMO.

With that said, I've seen claims on the web stating that 256 kbps AAC is equivalent in quality to 320 kbps MP3. My own ears can't tell these apart in ABX testing.

 

[Robbo99999]

Ah, thanks!

@j_j , what do you reckon re my questions & observations in my last couple of posts, can you shed any light on it with regards to your knowledge on the topics?
https://www.audiosciencereview.com/...what-differences-do-you-hear.5077/post-615711
https://www.audiosciencereview.com/...what-differences-do-you-hear.5077/post-615722

@j_j and others who are interested....replying to my own post here as I have additional information after contacting Amazon via chat, basically they're saying their Prime Music streaming service targets 320 - 850kbps, which corresponds pretty perfectly to my measured download rate of the Alexa process. Here's a couple of the screenshots of the chat:

Weirdly that leaves Prime Music streaming to be higher quality than the purchased mp3 downloads, which seems a bit off, but I did indeed ask the guy in chat and he confirmed that this meant Prime Music streaming was happening at a higher bitrate than the purchased downloads! I'm not sure how you can have an mp3 encoded at up to 850kbps which would be the logical extension of his replies? Still interested in your experience & take on this.

 

[escape2]

850 kbps is certainly not MP3.

If Amazon standard streams at 850 kbps, then what does Amazon HD stream at?

 

[Robbo99999]

850 kbps is certainly not MP3.

If Amazon standard streams at 850 kbps, then what does Amazon HD stream at?

That is weird, up to 850 kbps for Amazon HD:

Although you can see I did ask the guy specifically about the bitrate of the standard Prime Music and specifically said I wasn't referring to the HD version. Maybe with HD you'd be at 850kbps constant, whereas when I measured with Prime Music I was working out an average of 536kbps. Either way in my pretty quick subjective listening tests comparing lossless CD rips vs corresponding Prime Music streaming, the lossless rips have more dynamics and overall clarity, so I'm surprised to hear confirmation from Amazon supporting the 536kbps I measured.

 

[Kevin1956]

Your not quite in the right forum to make those types of proclomantions, there are a tons of other sites that get into making claims of hearing the impossible and improbable.
On the other hand, I don't understand why you would brag about being deaf to hearing many of the details that make the difference between a six transistor radio and a really good High Fidelity system. I also wouldn't crow about being color blind or wanting on the ability to enjoy taste or odor differences. I'm not very understanding of your post?

Was trying to be a bit sarcastic and humorous about how all this time spent discussing things that are shown to be inaudible to about 99% of the population seems to create a disconnect from the fact that listening to music should be enjoyable, even when reproduced in less than laboratory standard conditions.

My 50 years in this hobby eventually taught me to worry about the things that are audible to me, and not to worry the things that are inaudible to me. To not obsess over what I can’t discern compared to the golden eared audiophile. And that once your gear has inaudible levels of noise and distortions, switching to equipment that measures “better” doesn’t make the inaudible more inaudible. Sorry if my attempt fell flat.

 

[escape2]

That is weird, up to 850 kbps for Amazon HD:
View attachment 101817
Although you can see I did ask the guy specifically about the bitrate of the standard Prime Music and specifically said I wasn't referring to the HD version. Maybe with HD you'd be at 850kbps constant, whereas when I measured with Prime Music I was working out an average of 536kbps.

The image you posted does not support what you're seeing nor what the rep told you.

 

[Robbo99999]

The image you posted does not support what you're seeing nor what the rep told you.

Hence me saying "weird"......although that pic doesn't picture standard Prime Music streaming either....it just refers to "standard streaming audio" which is likely/potentially competitors solutions rather than standard Amazon Prime Music.

 

[Tks]

As a teenager I was ABXing high bit rate lossy vs. lossless and I have not used lossy since. There is an obvious lack of audible information if you listen carefully. Maybe my hearing has gotten worse by now. There is no reason to use lossy any more because of ample storage.

Portable sources like phones remain a sore spot if you use lossless. A 256GB card is good enough for a few 100 albums (especially if you have Hi-Res formats).

Another issue with storage isn't cost, but backup and the process of backups and swaps. Ive seen folks with almost 200,000 tracks that have been into music for a decade (I can't imagine some of those folks that used to buy the largest capacity iPods today, their library must be more than half a million tracks or something).

Having a backup and transferring that into working drives (since many computers use solid state storage rather than HDDs) is problematic. And perhaps a lossy version of your music for easier streaming or use in portable devices is even more annoying to take into account for seeing as how SSD's aren't terribly high capcity, and those that are, still aren't cheap if you want to keep swapping chunks from your main backup.

So while price wise, having a proper NAS or backup scheme isn't a cost problem to many people. It's still quite an annoyance in a few particular aforementioned ways to deal with; for those few select people that still exclusively pirchase, and have been purchasing music for years.

 

[Julf]

There is no reason to use lossy any more because of ample storage.

Bandwidth isn't free.

 

[Robbo99999]

I did some comparing of the recent track I downloaded in Amazon vs the Itunes version (briefly talked about in earlier post), listening on my K702 Harman EQ'd headphones. As mentioned earlier , Amazon downloaded track is 238kbps LAME VO VBR encoded mp3, and Amazon ITunes is 256kbps CBR AAC. Here's the Orban Loudness Meter comparison of each track, the Amazon track is 1.3dB louder than the ITunes track if I'm interpreting this right?
Amazon downloaded track:

ITunes downloaded track:

Because I identified that the Amazon version was 1.3dB louder I put in a negative preamp of -1.3dB when listening to the Amazon track to establish equal loudness. Listening to around 30 second passages & flipping between the two, I think I like the Amazon version better....there is quite a distinct difference, in the Amazon version the voice sounds a little more recessed but has more dynamics in terms of a feeling of greater depth to it, whereas in the ITunes track the voice is more in your face and seems to have less dynamic variance to the voice. I don't know if the Orban Loudness graphs say anything to support that, but I did try to compensate for what I think is the fact that the Amazon track is 1.3dB louder. It does look like for the ITunes track they've just squashed the frequency range to fit within a -1.2dBFS window whereas with the Amazon track they've allowed it to go to 0dBFS.

 

[Julf]

Listening to around 30 second passages & flipping between the two, I think I like the Amazon version better....there is quite a distinct difference, in the Amazon version the voice sounds a little more recessed but has more dynamics in terms of a feeling of greater depth to it, whereas in the ITunes track the voice is more in your face and seems to have less dynamic variance to the voice.

Sighted listening?

 

[Robbo99999]

Sighted listening?

Yes, sorry, I know

I think I'm fairly good at noticing differences though, I just close my eyes and try to pay attention to the various elements....and on short passages so I can remember some specifics, then flip within a few seconds to the start of that short passage on the other version of the track. I don't really have a vested interest in preferring Amazon, because any purchased downloaded music I have is in Itunes format, just that one track from Amazon that I have bought (as part of my quest for learning about Amazon Music now that I've recently setup music streaming with them through Prime).

 

[Julf]

Yes, sorry, I know

I think I'm fairly good at noticing differences though

And I think I am a better than average driver.

I don't really have a vested interest in preferring Amazon, because any purchased downloaded music I have is in Itunes format, just that one track from Amazon that I have bought (as part of my quest for learning about Amazon Music now that I've recently setup music streaming with them through Prime).

Unfortunately expectation bias is pretty much unconscious and not really dependent on rationally understood reasons for bias (or lack of it).

 

[hmscott]

That is weird, up to 850 kbps for Amazon HD:
View attachment 101817
Although you can see I did ask the guy specifically about the bitrate of the standard Prime Music and specifically said I wasn't referring to the HD version. Maybe with HD you'd be at 850kbps constant, whereas when I measured with Prime Music I was working out an average of 536kbps. Either way in my pretty quick subjective listening tests comparing lossless CD rips vs corresponding Prime Music streaming, the lossless rips have more dynamics and overall clarity, so I'm surprised to hear confirmation from Amazon supporting the 536kbps I measured.

Yup, you really want to add HD to Prime Music, it's a big difference. There are some recordings on Amazon that rivel Masters on Tidal, and dare I say surprisingly better than Tidal sometimes - I wish I took notes. I went through a lot of comparisons many months ago now, I don't have recall which were the standout's.

I finally pre-paid for an annual subscription to both:

Subscription Renewal
On April 22, 2021, your subscription will renew at $79.00/year.
An additional $50.00/year will be billed separately for HD access on April 22, 2021.

Well worth it even though I divide my time between numerous services. Backing up playlists on several. And, occasionally I get a nice surprise, like when I recently got another Mac after many years without - after Apple canceled the 17" Macbook, killed the Desktops, and generally mis-managed Mac OSX development (long story) - the M1 Macbook Pro was a revelation in CPU/GPU power and power usage.

The next surprise was the Amazon Music HD application on Mac OSX, it is way better than the buggy Windows version, I get "crashes" inside of playlists on the Windows version when setting them up, "Please Contact Amazon Support", worthless, but it works well enough - until I experienced the Mac OSX version, I think it even sounds better. Tidal's application is different but their Windows app is already solid.

Anyway, check out Amazon Music HD, it is a big step up from Apple Music, Spotify, and can go toe to toe with Tidal

BTW, Apple Music Streaming - the music you get to stream for "free" after paying $9.99/mo quality sucks. I compared tracks I've bought on Apple Music that are 44.1khz 256k bit streaming, vs "unknown" bit rate tracks of the same artist - that on Tidal and AMHD sound of similar quality - sound like crap from AMS if you don't "own" them.

I haven't tried Apple Music Streaming at the $14.99/mo quality though. I've got a free 1 year AppleTV + 3 month Apple Music subscription to run out first. No rush.

Here's what I mean by the quality of the purchased music you stream vs "free" streaming:

Free:

Enjoy

 

[Robbo99999]

And I think I am a better than average driver.



Unfortunately expectation bias is pretty much unconscious and not really dependent on rationally understood reasons for bias (or lack of it).

Hey, listen, I know I'm a better than average driver.....it's because I'm kinda sporty & have good hand eye coordination!