Tidal vs Qobuz: Which High-Res Streaming Service Sounds Best?

[Matthew J Poes]

Thank you, thank you, thank you!!! I have been beating my head against the wall for months... I have been checking the waveform/files created from VB-cable and audacity (I wasn't aware of MusicScope and just brought the different files into a multi-track editor to compare them) similar to your approach. No matter what service/hi-res sound quality (CD or better from Tidal, Deezer, or Qobuz) the waveforms are visually identical. Since I haven't done an exhaustive search of albums to test, I thought I might just have been unlucky, but I highly doubt I just happened to test 5 albums that represent a tiny fraction of the issue. My theory is that many of the studios are upsampling from CD masters to get the higher quality files (up to the 24bit/192kHZ) and there isn't any real difference in the quality of the underlying music. Unfortunately, creating all of the higher quality files can be easily accomplished with upsampling, while re-mastering from source to create the files may be much more time/resource-consuming. Would love to hear other peoples experience/takes.

To the comparison of services, I have been using Tidal, Deezer and Qobuz for at least 4 months to compare the differences before deciding on which to keep long-term. The easiest decision is that Deezer is the first to go as I don't see any advantages compared to Tidal and Qobuz. I haven't had any technical issues with Qobuz and both Apps (Tidal/Qobuz) work on my DAP (android based) equally well. While they both have large collections, I find that Qobuz has a greater offering of music in my listening genres (Outside of the Beastie boys I am not a Rap listener and some would argue that isn't Rap ) and will likely be staying with Qobuz. I am also testing the new Amazon HD service and have noticed a significant improvement in sound quality compared to Amazon SD (standard streaming). I investigated the waveforms and find the same issue with Amazon - they look identical to the other streaming services no matter what stated quality. I can understand why some say they hear no difference between Tidal or Qobuz as in my hands the waveforms are identical. I think those that do detect a quality difference are likely experiencing an ascertainment bias due to knowing which service a specific track is from... Excellent article!

Im glad someone appreciated the work that went into that analysis. I did a lot more analysis than I put into the article. @shadyJ felt it was not fair to CB radio so I had to let some of it go. Even the effort to avoid aliasing artifacts was intense (placing a file of a given size into a container of a higher sampling rate can cause aliasing).

In any case, most pop music I tested didn’t have any musical HF content above 15-20khz. It made testing the Hd aspect of the services pointless. The best material to test turned out to be fairly obscure classical music. Hence why I included the John Williams piece.

I bet I tested about 50 different songs and I did find differences in all of them. I noted the differences in the review. However, the analysis method I used let’s to tell if that is a difference in noise shaping or actual musical content. It was always either noise shaping or suspected provenance issues.

like I mentioned, a few albums that were listed as the original on Tidal turned out to be a remaster. I confirmed because they were identical to the CD wav rip of the remaster and often matched the Qobuz album listed as a remaster.

provenance turned into a very big and common problem. If you are hearing big differences, I would really suggest people check the musical provenance.

 

[Matthew J Poes]

I tried both Tidal and Qobuz and didn't keep either. Now there's Amazon as well at a better price too. I'm fine with Spotify premium service so far and have put a bit of time into creating playlists and a history for suggestions to be based on. I found little of the masters catalog (at least that I was interested in) during my Tidal trial and in general found many things I had on Spotify not available at all (altho someone suggested later it was more the search engine at Tidal, and it has been improved since but I don't know). I thought the interface a bit clunky with Tidal. Qobuz I thought better pretty much on all fronts. I like the ability to buy a download from Qobuz, too, as I've got internet access issues a fair bit. I agree with the comment that Tidal had placed too many offerings in the way that I wasn't interested in (particularly Jayz's own), nor am I interested in anything MQA, who needs another drm issue?

What I would like to see is some testing of provenance of each recording format offered among the services at various subscription levels.

what do you mean? Provenance means the source material in this case. I can confirm the provenance for the 50 or so tracks I tested. What I can’t tell you is much detail. Like I said, most of the pop music was only 24/44 or 24/48 even if listed as 24/96. I can’t tell if it’s really 16 bit or 24 bit, there would be no way to analyze the file to tell really. Music rarely if ever uses all the bits to know that. The dynamic range on any modern music isn’t even in the ball park. So a 24 bit file will use al bits but most of it is noise.

mad for the sampling frequency. That is easy to tell and I was able to confirm sources but the problem is, it’s possible this is the actual master too. We have no easy way to know what kind of A/D converters were used, for example. Who knows where the high frequencies were lost. Some music showed clear signs of low pass filters even. I don’t have enough knowledge on how to track down the equipment used to figure that out.

 

[Matthew J Poes]

Have you had any experience with the recently announced Amazon Music HD? It undercuts the competition at just $13/month for Prime members. Last year I dropped Apple Music for amazon because my then 9 yr old likes using his alexa for music. So I now plan on upgrading to the HD service to attempt to distance myself from the now widely-accepted "compressed streaming music" plague... For fun, we also use youtube music, which adds variety, and is included w youtube red.

not yet and I know a lot of people were asking. It just released and I read soon after they would be introducing ATMOS surround music as well. That is so huge that I decided to leave that to another article.

I will sign up to Amazon HD and investigate more of their music and sources. My guess is that it’s about the same. I am not totally sure what the highest bit rate of Qobuz is. I assume Amazon is comparable in quality. If they have as good a music selection then it could be a real bargain.

but let me tell you, if the Atmos releases end up being what I think, this could be huge. I’ve been asking about surround streaming music in a lossless format for a long time now. I listen to music in stereo mostly but I also have a solid understanding of the technical advantages of surround music. Done correctly, surround music IS superior to 2-channel stereo. The lack of good content has made that a moot point. This could change everything. That is, as long as they aren’t just upmixing the 2-channel source material. While that may not turn out to be the worst thing, actual remixing into surround would be much better.

Again, i hope to cover the advantages of surround music and Amazon Atmos music in a future editorial.

 

[OldHardwareTech]

I subscribe to both Tidal and Qobuz hires tiers. I thought I'd pass along the only track with which I've been able to hear a difference. School by Supertramp on The Very Best Of album. The one from Oct 1990 with a white album cover. There are 2, the other has a black cover. On both services they are 16/44.1 resolution. Using Tidal during the first minute I'm getting some noise mixed in with the harmonica. It does it on my Samsung S8 too so I know it's not my equipment. Qobuz serves up the same track on the same album perfectly, no noise. To be fair the versions on other albums have no noise on either service. Just a bad file would be my guess. For the foreseeable future I plan to keep both, mainly because building playlists on Tidal is easier. Qobuz requires adding each track individually whereas with Tidal I can generate a playlist using track/artist radio or suggested tracks and delete the tracks I don't want. I then transfer the playlist to Qobuz using Soundiiz. Qobuz does have some excellent curated playlists but I still enjoy my own. I also use Tidal on autoplay when I'm working on something and don't want to bother with choosing what's next, for background music in other words. I'm not a fan of MQA, maybe it's good for people who have low internet caps or only connect using mobile data, but I think that's probably a fairly small subset of people. Otherwise I see no reason for MQA to exist. Certainly most who listen at home don't need it. In most homes internet is geared to serve hires video and even the highest quality bit perfect audio doesn't come close to the data hires video consumes. I'm also not a fan of Tidals recommendations, but as my tag indicates I'm old.

Thanks for an interesting article!

 

[Matthew J Poes]

I will also mention that when I started seeing differences in my analysis, I held up this article because I really thought I was into some industry secret. I was finding audibly substantial differences in the HF content between services and HD vs SD versions.

I consulted digital experts. I re-analyzed. I wrote programs into Octave. I really dug in. This is what lead to the now obvious aliasing problem I had inadvertently introduced as well as the provenance issue. That is why I ultimately concluded that the musical content was the same. There was no musical differences. The differences I was finding could all be explained away.

but big lesson learned. Be very careful about how you upsample. The same method I used to create a common sampling container can also be used to upsample and this upsampling can introduce aliasing which can alter the musical content. It was really eye opening.

 

[Matthew J Poes]

I subscribe to both Tidal and Qobuz hires tiers. I thought I'd pass along the only track with which I've been able to hear a difference. School by Supertramp on The Very Best Of album. The one from Oct 1990 with a white album cover. There are 2, the other has a black cover. On both services they are 16/44.1 resolution. Using Tidal during the first minute I'm getting some noise mixed in with the harmonica. It does it on my Samsung S8 too so I know it's not my equipment. Qobuz serves up the same track on the same album perfectly, no noise. To be fair the versions on other albums have no noise on either service. Just a bad file would be my guess. For the foreseeable future I plan to keep both, mainly because building playlists on Tidal is easier. Qobuz requires adding each track individually whereas with Tidal I can generate a playlist using track/artist radio or suggested tracks and delete the tracks I don't want. I then transfer the playlist to Qobuz using Soundiiz. Qobuz does have some excellent curated playlists but I still enjoy my own. I also use Tidal on autoplay when I'm working on something and don't want to bother with choosing what's next, for background music in other words. I'm not a fan of MQA, maybe it's good for people who have low internet caps or only connect using mobile data, but I think that's probably a fairly small subset of people. Otherwise I see no reason for MQA to exist. Certainly most who listen at home don't need it. In most homes internet is geared to serve hires video and even the highest quality bit perfect audio doesn't come close to the data hires video consumes. I'm also not a fan of Tidals recommendations, but as my tag indicates I'm old.

Thanks for an interesting article!

Ill take a look at the album and track you mention.

 

[lovinthehd]

what do you mean? Provenance means the source material in this case. I can confirm the provenance for the 50 or so tracks I tested. What I can’t tell you is much detail. Like I said, most of the pop music was only 24/44 or 24/48 even if listed as 24/96. I can’t tell if it’s really 16 bit or 24 bit, there would be no way to analyze the file to tell really. Music rarely if ever uses all the bits to know that. The dynamic range on any modern music isn’t even in the ball park. So a 24 bit file will use al bits but most of it is noise.

mad for the sampling frequency. That is easy to tell and I was able to confirm sources but the problem is, it’s possible this is the actual master too. We have no easy way to know what kind of A/D converters were used, for example. Who knows where the high frequencies were lost. Some music showed clear signs of low pass filters even. I don’t have enough knowledge on how to track down the equipment used to figure that out.

You pretty much have it. By provenance I mean the specifics of the track's background, all relevant info as to who did what when for production/engineering as well as the technical specifics of bit rate/depth/sampling frequency. It wouldn't surprise me they may be mostly 16/44.1 versions upmixed to the higher bit depth/sampling frequency, which I've seen claimed by a coupla guys who say they're familiar with the distribution of such to the streaming services. HD Tracks has been caught doing as much. Whatever sells I suppose. That's mostly why I'd like to know specifics or provenance for everything if I were to pay additional for such services. I personally don't care beyond cd level in any case, that's more than good enough for my purposes (as is 320kbps OGG Vorbis).

 

[OldHardwareTech]

Very interesting, I was just thinking of which one is better yesterday as well. I have Tidal and can hear a clear difference from the compressed to lossless audio on my computer speakers.
I just got a Roku Premiere yesterday. When streaming Tidal on Roku to my receiver it seems to be compressed, it doesn't sound as good as on my computer speakers. I don't see any options to tell it to use the Hifi mode in the Roku app, or anyway to confirm if it is using the Hifi stream or the lossy streams. Does anyone know how to get the Tidal Hifi stream for Roku? Or should I use a different streaming device?
I played several albums from direct WAV rips of CDs and compared them to the tidal streams. On my computer they sound identical, but with the Roku they sound compressed and the stereo image is smaller.

I also have a couple of ROKU units in my home and I looked into this. I didn't find a way. Until Tidal or Qobuz makes a hires app for ROKU I don't think we'll be able to.

 

[Matthew J Poes]

You pretty much have it. By provenance I mean the specifics of the track's background, all relevant info as to who did what when for production/engineering as well as the technical specifics of bit rate/depth/sampling frequency. It wouldn't surprise me they may be mostly 16/44.1 versions upmixed to the higher bit depth/sampling frequency, which I've seen claimed by a coupla guys who say they're familiar with the distribution of such to the streaming services. HD Tracks has been caught doing as much. Whatever sells I suppose. That's mostly why I'd like to know specifics or provenance for everything if I were to pay additional for such services. I personally don't care beyond cd level in any case, that's more than good enough for my purposes (as is 320kbps OGG Vorbis).

yeah I’ve heard this too.

I really don’t know. I think it would take a lot of work to find out. The companies involved aren’t divulging this.

I too found a lot of music at standard red book quality, they basically matched that of the CD version. But I don’t know that it means it came from the CD so much as the master might be limited. A lot of the older digital equipment was pretty limited and goes back pretty far. I also found some where I found obvious evidence that it was true it high resolution, but the high frequency bandwidth limit was also clear.

there is clearly real HD music out there, but not a lot of it. I prefer that streaming services provide music in HD only to provide the necessary infrastructure to ensure it is the format of the future. However I don’t really care or lose sleep over current and older music being actually high resolution. It’s mostly not and it all sounds fine.

I didn’t want to make the article about 44.1khz vs 88.2khz+ and which sound better. I did compare within the service the streams from each quality level and found that the effect of the nyquist limit could be seen far lower in frequency than the actual limit. In other words, while the noise dropped precipitously above 22 or 24khz, there were significant differences in the amplitude or the high frequency music evident down to 5-8khz. I shared my results with a digital systems expert that Gene put me in touch with and he confirmed the finding. This is why I think the higher sampling rate is a good idea. It keeps the effect out of the audible range.

 

[baronvonellis]

I didn’t want to make the article about 44.1khz vs 88.2khz+ and which sound better. I did compare within the service the streams from each quality level and found that the effect of the nyquist limit could be seen far lower in frequency than the actual limit. In other words, while the noise dropped precipitously above 22 or 24khz, there were significant differences in the amplitude or the high frequency music evident down to 5-8khz. I shared my results with a digital systems expert that Gene put me in touch with and he confirmed the finding. This is why I think the higher sampling rate is a good idea. It keeps the effect out of the audible range.

Thanks, for finally confirming, what I've always heard with high frequency sounds such as cymbals I can hear some noisy hash in the upper frequencies with 44.1Khz, but music at 96khz it goes away and is much clearer for music with alot of high frequency content. For lower frequencies it doesn't make as much difference. This is what I was taught at the Conservatory of Recording Arts & Sciences in 2006, and it's easy for me to hear with cymbals.

Going back to the early 2000's I know music in studios would have been recorded at 96khz or 88k/24 bit in Pro Tools usually. If something was recorded digitally in the 80's or 90's it was probably at 48k. Pro Tools didn't support 24 bit audio until 1997, and they didn't support sample rates of 96khz until Pro Tools HD in 2002 at the earliest and it probably took studios several years to upgrade to HD systems.

 

[lovinthehd]

yeah I’ve heard this too.

I really don’t know. I think it would take a lot of work to find out. The companies involved aren’t divulging this.

I too found a lot of music at standard red book quality, they basically matched that of the CD version. But I don’t know that it means it came from the CD so much as the master might be limited. A lot of the older digital equipment was pretty limited and goes back pretty far. I also found some where I found obvious evidence that it was true it high resolution, but the high frequency bandwidth limit was also clear.

there is clearly real HD music out there, but not a lot of it. I prefer that streaming services provide music in HD only to provide the necessary infrastructure to ensure it is the format of the future. However I don’t really care or lose sleep over current and older music being actually high resolution. It’s mostly not and it all sounds fine.

I didn’t want to make the article about 44.1khz vs 88.2khz+ and which sound better. I did compare within the service the streams from each quality level and found that the effect of the nyquist limit could be seen far lower in frequency than the actual limit. In other words, while the noise dropped precipitously above 22 or 24khz, there were significant differences in the amplitude or the high frequency music evident down to 5-8khz. I shared my results with a digital systems expert that Gene put me in touch with and he confirmed the finding. This is why I think the higher sampling rate is a good idea. It keeps the effect out of the audible range.

I think they're making it very hard to do this deliberately. From a business viewpoint it's somewhat understandable, too. I think mostly the worry about Redbook or Hi-Res (rather than the attempt to change the moniker like Amazon is doing with HD/Ultra HD) is a needless one. Just the fact that they didn't provide such to begin with and resist now....just smells of bullshit like so much is to do with audio sales these days.

The title does say sound best, so might be expected to break it down by bit depth, sampling frequency and bitrate delivered, etc....

 

[Matthew J Poes]

Thanks, for finally confirming, what I've always heard with high frequency sounds such as cymbals I can hear some noisy hash in the upper frequencies with 44.1Khz, but music at 96khz it goes away and is much clearer for music with alot of high frequency content. For lower frequencies it doesn't make as much difference. This is what I was taught at the Conservatory of Recording Arts & Sciences in 2006, and it's easy for me to hear with cymbals.

Going back to the early 2000's I know music in studios would have been recorded at 96khz or 88k/24 bit in Pro Tools usually. If something was recorded digitally in the 80's or 90's it was probably at 48k. Pro Tools didn't support 24 bit audio until 1997, and they didn't support sample rates of 96khz until Pro Tools HD in 2002 at the earliest and it probably took studios several years to upgrade to HD systems.

well don’t overstate what I found. I didn’t correlate any of this with listening tests. It was just an objective measure of the spectral energy. I would have to delve much more deeply into the files and do more experiments. Maybe like a high passed version of the file.

all i was saying was that I was finding spectral differences in the audible range comparing the different files. For all I know the streaming services are tampering with the files. I didn’t downsample the files to test this. I changed the quality level and re-recorded.

 

[Matthew J Poes]

I think they're making it very hard to do this deliberately. From a business viewpoint it's somewhat understandable, too. I think mostly the worry about Redbook or Hi-Res (rather than the attempt to change the moniker like Amazon is doing with HD/Ultra HD) is a needless one. Just the fact that they didn't provide such to begin with and resist now....just smells of bullshit like so much is to do with audio sales these days.

The title does say sound best, so might be expected to break it down by bit depth, sampling frequency and bitrate delivered, etc....

but who says these small differences in bit rate would have any impact on sound. I mean, if they all use FLAC, there shouldn’t be any difference. They are all lossless compression right?

I honestly didn’t hear differences and that was the point of the article. I went in with an open mind. I still felt I could hear the difference between the 320kbps MP3 quality and the CD quality. Beyond that I couldn’t make heads or tails of things. All sounded the same to me.

analysis of the files showed that 99% of what I listened to was only red book quality anyway, so of course it sounded the same.

I managed to find just a small handful of albums with true high resolution quality and I still wouldn’t bet my life on any sound quality differences. I measured small differences which may be audible. I believe there is no reason not to offer this in HD, I just think they shouldn’t charge extra for it. There isn’t an extra cost to do it.

 

[John Nemesh]

I subscribed to Tidal (again) recently...and even bought a Bluesound Node 2i and an ifi Nano iOne DAC that was capable of decoding MQA to get the best quality from Tidal. (seriously, get rid of that Sonos! It's internal DAC is complete crap, and can't even support anything over CD quality, even with an external DAC!) Then, right after, Amazon launched their HD (I loathe that term for high resolution music!) music service. It has a larger catalog, with more of the music in CD quality or better, and works well with most major platforms. (High res audio was supported on day one with my Node 2i) And it's FREE for the first 3 months!

Navigation is a bit rough using BluOS, but the sound quality is just as good, if not better than Tidal! Best of all, neither Amazon nor Qobuzz require you to purchase specific MQA compatible hardware to enjoy the music in the highest quality possible!

I am going to ditch Tidal. It really offers LESS than the competition for a HIGHER price.

 

[John Nemesh]

ces in my analysis, I held up this article because I really thought I was into some industry secret. I was finding audibly

but who says these small differences in bit rate would have any impact on sound. I mean, if they all use FLAC, there shouldn’t be any difference. They are all lossless compression right?

I honestly didn’t hear differences and that was the point of the article. I went in with an open mind. I still felt I could hear the difference between the 320kbps MP3 quality and the CD quality. Beyond that I couldn’t make heads or tails of things. All sounded the same to me.

analysis of the files showed that 99% of what I listened to was only red book quality anyway, so of course it sounded the same.

I managed to find just a small handful of albums with true high resolution quality and I still wouldn’t bet my life on any sound quality differences. I measured small differences which may be audible. I believe there is no reason not to offer this in HD, I just think they shouldn’t charge extra for it. There isn’t an extra cost to do it.

I have found a TON of high res music on both Tidal and Amazon! Not sure what genre you listen to, but there is quite a bit. ALSO, you can often listen to the same album in regular or high res and compare the differences. If you have a good system, it's NOT a subtle difference. Play Toto's "Africa" in HR...you will be blown away! Or just about any song on Queen's "Night at the Opera" (my personal favorite is "The Prophet's Song"). Or Led Zeppelin's "Moby D-ick" (best drum solo EVER, fight me if you disagree!). And, on Amazon at least, if a song's not in "HD" it's usually in at least CD quality, which still sounds pretty damn good! Some audiophiles might split hairs over which service sounds best...but the reality is that ALL of these services that support high-resolution music sounds pretty damn good with the right equipment! If price is your objection, though, Amazon's $13/mo price tag is the most affordable (that price is for Prime members, non-Prime members pay $2/mo more). And that's only $3 more than the standard price for Spotify or any of the other lossy streaming services.

Edit: Why is it replacing "D-I-C-K" with "lord helmet"? Some kind of anti-profanity filter?

 

[Matthew J Poes]

I have found a TON of high res music on both Tidal and Amazon! Not sure what genre you listen to, but there is quite a bit. ALSO, you can often listen to the same album in regular or high res and compare the differences. If you have a good system, it's NOT a subtle difference. Play Toto's "Africa" in HR...you will be blown away! Or just about any song on Queen's "Night at the Opera" (my personal favorite is "The Prophet's Song"). Or Led Zeppelin's "Moby D-ick" (best drum solo EVER, fight me if you disagree!). And, on Amazon at least, if a song's not in "HD" it's usually in at least CD quality, which still sounds pretty damn good! Some audiophiles might split hairs over which service sounds best...but the reality is that ALL of these services that support high-resolution music sounds pretty damn good with the right equipment! If price is your objection, though, Amazon's $13/mo price tag is the most affordable (that price is for Prime members, non-Prime members pay $2/mo more). And that's only $3 more than the standard price for Spotify or any of the other lossy streaming services.

Edit: Why is it replacing "D-I-C-K" with "lord helmet"? Some kind of anti-profanity filter?

because lord helmet is hilarious.

I based my comment on high resolution from my measurements. I’ll check out the Toto track you mention and let you know what I find. Previously I did not find a lot, but with millions of tracks, I didn’t touch even a small fraction of the music.

now if the Toto piece turns out not to contain much of any high frequency content, then what?

 

[Matthew J Poes]

I have found a TON of high res music on both Tidal and Amazon! Not sure what genre you listen to, but there is quite a bit. ALSO, you can often listen to the same album in regular or high res and compare the differences. If you have a good system, it's NOT a subtle difference. Play Toto's "Africa" in HR...you will be blown away! Or just about any song on Queen's "Night at the Opera" (my personal favorite is "The Prophet's Song"). Or Led Zeppelin's "Moby D-ick" (best drum solo EVER, fight me if you disagree!). And, on Amazon at least, if a song's not in "HD" it's usually in at least CD quality, which still sounds pretty damn good! Some audiophiles might split hairs over which service sounds best...but the reality is that ALL of these services that support high-resolution music sounds pretty damn good with the right equipment! If price is your objection, though, Amazon's $13/mo price tag is the most affordable (that price is for Prime members, non-Prime members pay $2/mo more). And that's only $3 more than the standard price for Spotify or any of the other lossy streaming services.

Edit: Why is it replacing "D-I-C-K" with "lord helmet"? Some kind of anti-profanity filter?

Well I'll be Lord Helmets uncle, you were right. Toto's Africa on the studio albums, which are called HD are actually CD quality, but the newer live version appears to be true HD and the effect is evident.


This is the 16/44 output and you can clearly see a V shape with some spikes in the spectral response around 22khz. It's pretty pronounced.


Here is the HD version and the spectral response in the high frequencies is clearly flatter with no V at 22khz. It doesn't have much meaningful content above 30khz or so, but there is actually a sizable difference in the spectral content between the two. I have a different tool I have to use that post-processes a text file from this software to tell, but I will guess that there is at least 3dB's less HF energy above 5khz in the 16/44 file. If I were a betting man, I'd say closer to 6dB and by 20khz, more like 10+.

This is the issue I was mentioning before where I found often that while there wasn't necessarily evidence the music itself was HD in the sense of strong HF spectral content (though this has more than many). Instead, I found VERY strong evidence of a shift in tonal balance. Either Qobuz (or the studios) are messing with the files, or this is the difference.

 

[Matthew J Poes]

Here is a graphic comparison. The lines are 5dB increments and I changed it to a Log scale to make it easier to see. This is actually a different comparison than above, but the same thing is evident here. This is a 24/96 recording of the song from Tidal that has been exported as a flac file in 16/44 and 24/96. I then ran the analysis on both. That means that in this case, the source material is the same. The only difference is a difference I made and so any spectral difference is a result of that change that I made. I didn't apply any filters or EQ, I only saved the files differently.

 

[Matthew J Poes]

Ok I've done a different graph that now adds the 16/44 file out of Tidal but recorded into Audacity at 24/88.2. This shows the actual spectral difference when you shift the quality level. It looks like the most obvious difference that I would put any weight into is the differences above 12khz. At 1/4 of 44.1khz that seems to reflect a kind of magical point where they all converge and then differ substantially.

 

[John Nemesh]

because lord helmet is hilarious.

I based my comment on high resolution from my measurements. I’ll check out the Toto track you mention and let you know what I find. Previously I did not find a lot, but with millions of tracks, I didn’t touch even a small fraction of the music.

now if the Toto piece turns out not to contain much of any high frequency content, then what?

I don't measure the music, I just listen. It sounds amazing, whatever the case.