Music services review

I like many have had a lot of time over the past 12 months and I’ve been trying to decide on which music service to stick to. While the below isn’t really technical, it’s all I could do to try and justify my own reasoning. And with the everlasting MQA debate, I thought I could try and get me back on the fence with it after a post of mine was removed by @danny

Just to be clear, I am not a hater, nor mean any disrespect in this post. My views only and open to others opinions. Yes I’d like your opinions and suggestions. Or if you think I’m worng, great tell me. I won’t take offence, just be polite.

Sound quality is king, right?

With the constant debate over Tidal/MQA, I thought I’d take one last look at it as a consumer. Without test data and all that, I decided to check download (offline) files sizes as a comparison from Tidal, Spotify, Qobuz and Amazon Music. My logic is that if the files sizes for Flac are bigger then surely they retain all the information from the original mastered version. If MQA files were bigger than CD and a little less than Flac, as I interpret it, I could start to form a different opinion maybe.

I used the same Coldplay album as I know it is 24/192 on Qobuz and Amazon music. Spotify was set to 320kbps and Tidal used the Master version. I chose this album as I like it.

I then did an A/B comparison over a 24 hour period between each services CD/Hi-res version. At the end of the day it’s subjective to each of us, this will just be my opinion.

Here we go…

Tidal
CD quality 16/44.1 @ 284mb
MQA master @ 284mb
I am very surprised by this. Unless I’m mistaken, I thought MQA masters were meant to be about 20% bigger in file size when compared to CD quality. Does this suggest it’s the same as the CD version, with some MQA information to bring on my MQA logo on my DAC?

Spotify
Ogg Vorbis 320kbps @ 113mb
As you’d expect.

Qobuz
CD qaulity 16/44.1 @ 280mb
Hi-res 24/192 @ 1.79gb
Similar CD file size to Tidal, but massively different in file size when 24/192.

Amazon music
CD quality 16/44.1 @ 282mb
Hi-res 24/192 @ 1.52gb
Similar CD file size to Tidal and Qobuz, but massively different in file size when 24/192.

While these findings are only looking at file size, to a layman this suggests the 24/192 versions from Amazon and Qobuz retain all the important information and are normally considered lossless. Spotify is what it is. However, Tidal shocked me a little. I chose a further album by Twenty One Pilots (Blurryface) and this revealed the same. Both CD and MQA were the same file sizes. This may differ for other albums that I haven’t checked and stand to be corrected. Does this suggest all MQA files are upsampled from 16/44.1?

Can anyone explain this to me in layman terms as the research I have done, including listening to Bob Stuart previously, I was under the impression MQA file sizes are smaller than your usual hi-res Flac but bigger than CD quality file sizes. Yes there is this origami folding going on, and unfolding which my equipment can do fully. But if it’s that small a file size, can it become hi-res?

A/B tests for CD quality and 320kbps. My preferred in order are;
Jointly - Qobuz, Amazon music and Tidal
Spotify

I found Qobuz, Amazon Music and Tidal all sounded very good in CD quality. I may say Tidal was slightly more textured, warmer over Amazon but for whatever reason I felt Amazon sounded clearer, precise. Qobuz was very good. Spotify, bit unfair to compare really.

A/B test for MQA and 24/192. My preferred in order are.
Qobuz
Amazon Music
Tidal

Qobuz just pleased me more. Simple really. Amazon was clear but again lacked a little texture. During the few opening seconds of Coldplay’s Don’t Panic you can here some background noise. One noise sounds like the guitar on the right channel being picked up or a hand touching the strings, and then there’s a clear fret noise. On both Qobuz and Amazon these are clear. On Tidal these noises sound different, muted a little to me. I also found the tracks hiss (noise floor?) more noticeable on Tidal, maybe this is how it’s meant to be. On Qobuz and Amazon this hiss sound was less, and in this respect I preferred them.

This is my view for most albums, tracks I like that are available on Qobuz. I only really used Amazon on this occasion as a 3rd comparison.

Tidal MQA vs Tidal CD
While I can say there was a slight difference between the two, I cannot say it was that great of a difference for me to prefer the MQA over CD. I have only based this on this Coldplay album and a few others I really like. Other people may find different findings and I’m happy to listen to other people’s suggestions.

Spotify is Spotify, great for your phone and being mobile. It does sound good at home but not in the same league as Qobuz, and dare I say Tidal either.

I’m not a hater of Tidal. I’m just not keen on MQA, it’s supposed benefits which I don’t feel are that great. If I found the sound quality was a lot greater to my ears over CD I might be into it. I like the idea of smaller file sizing and the idea behind the technology of MQA (which is closed off from people). Bandwidth, nor storage are issues to me. I’m lucky to have unlimited mobile data as well as broadband.

My dislike of MQA (not Tidal) is the idea that to listen to MQA, equipment manufacturers have to pay MQA to have it in their equipment, higher costs for consumers. We, the consumer have to buy separate equipment to enable us to listen to MQA. But what if we don’t like it? An equivalent DAC etc without MQA support could have costed me a lot less. For many this may not present an issue.

For a friend of mine who knew nothing about MQA and subscribed to Tidal, they were amazed at the sound quality of MQA. I pointed out that using an Android phone with Bluetooth headphones wasn’t how you listen to MQA. I set them up with a £100 USB MQA enabled DAC and ok headphones. They didn’t want to be wired up so went back to Spotify.

As for individual music service apps go, I prefer Tidal. Amazon and Qobuz mobile apps are not the best.

As mentioned above, please share your views, opinions and any music recommendations, more importantly in MQA to see if I can be persuaded.

You brought it up and called me out by name, so I’ll address your need to be victimized:

You got your post removed because you were involved with a off-topic bickering with others in an MQA thread. The other poster did too. You didn’t add anything of value in that thread, and you went way off-topic. You did bad. You got your posts removed. Move on.

So, in the spirit of moving on:

There are many ways to explain your results:

  1. The MQA file in question should have been more than ~1% larger than the CD version, as your tests showed. MQA is usually 24bits per sample and not 16bits. That should add 50% to the raw size of the file. Compression should be worse on that extended 50% because MQA compresses worse than CD because the folding is non-audio data below the noise floor, which FLAC is pretty bad at compressing. Something may be wrong with how you measured the size of the MQA file.

  2. You were playing back FLAC files, and if all you were measuring was bandwidth, you don’t know the compression level the FLAC file was made with. My guess is that MQA files are FLAC’d at the highest compression level (at the greatest CPU cost) because part of the story is “small file”. Without seeing the FLACs, it’s hard to know what is going on.

  3. The possibility (and as far as I know, there is no proof, only hunches) that TIDAL does not actually have non-MQA content for much of their MQA music, and all they are really doing when delivering non-MQA is to deliver MQA streams with some of the MQA sync bits mangled so the file can not be recognized as valid MQA. If this was true, you’d see exactly the same sizes for CD and MQA.

  4. You seem to think that you listened to TIDAL MQA and TIDAL non-MQA (but CD quality), but you may have done no such thing. Unless you are in Australia, there is no CD-quality tier that’s limited beyond a suggestion flag. The rest of the world may or may not get “auto upgraded” to MQA, no matter the setting. This is an easy thing to test by looking at Roon’s signal path, unless #3 above is true (which we don’t have evidence of).

My guess is that any difference you may have heard here was purely placebo. Given your file sizes and more importantly, an ineffective testing technique (only looking at file sizes and not actually checking to see if the data was actually different in meaningful ways), I think you listened to the same content twice.

@danny

Many thanks for your reply. I don’t feel victimised. Just stating a fact a post was removed and gave my reasons for this new posting. All friendly here.

Although I didn’t state, I checked via Roon or other means, I can confirm I did check bit rates etc.

Via my DAC the Amazon streams where as stated on the device used to playback via an iPhone 12 with the iOS app, connected via USB.

For Qobuz I used Roon and likewise the stated bit rates etc were as detailed.

I also used my iPhone 12 via USB with the iOS Qobuz app and it too showed the same as Roon.
I also used USB audio player ppro on my Android device which to gave the same results.

With Tidal, Roon , Tidal via iPhone 12 USB, and USB audio player pro showed both CD and MQA as 16/44.1, but once unfolded the MQA was 24/192. Have I made a mistake?

I hope this makes sense.

I’m not a audio scientist with all the fancy kit others may have to do lots of other tests.

Just checking the basics.

File sizing was checked via my phone to which I offlined the individual albums per app. The file browser showed the file sizes. Unless I’m wrong, this would suggest the details I gave are ccorrect.

Do you have any other music suggestions for me to try via Tidal in MQA that might sound fantastic when compared to Qobuz or other normal Flac files?

Thanks agian for your input.

I think you might be correct. If they have replaced all CD quality Flac with MQA versions this would reveal what I found in the same sized files. But surely the unfolded same version would, should sound better. Maybe, just maybe my system doesn’t have the resolution to reveal such improvements. I have listened via good quality IEMs and still didn’t note a vast difference.

Can anyone suggest some albums, tracks to try in MQA with a non MQA CD version. I’m open to options you know. :slight_smile:

All content on Tidal now is MQA, when you play FLAC its also MQA file decoded in service (when you don’t have MQA equipment, Tidal in HiFi/Master tier give you software on-the-go decoder limited to 24/96). So, always origami. For me - very stupid idea. Identification here is useless imo.

You confirm that (same file size), same quality, when flag is on (24/…) Tidal give you CD quality. No more, no less with MQA digital processing (where is studio master here, btw?)

When I compare, like you, to Qobuz stream, always CD version is better in Q, and HD version absolutely the best. Also, Deezer HiFi (no VBR, solid 1411kbps) is better than Tidal HiFi/Master. For me more open, more airly sound, better dynamic.

1 Like

Unfortunately this is the worst thing about MQA. It is near impossible to do a proper a/b test because nobody has access to the encoder.

No, because I don’t have a opinion on MQA’s sound quality.

Might want to ask MQA fans… @Chrislayeruk comes to mind, but there are others here.

1 Like

Great MQA albums are many and various but you may want to try the recently released George Michael MQA studio version of Faith at 16/44
Ashley McBryde Girl Going Nowhere or Shelby Lynne Just A Little Lovin’

1 Like

Cheers @danny greatly appreciated

Cheers @Chrislayeruk I’ll give them a listen. Who knows, I may like George Michael.

Cheers @Antoni_Wozniak lucky enough I have the equipment to fully unfold MQA. Just tested some samples from www.2l.no as per the HB Channel on YouTube. My DAC also does upto DSD512 (not going there) and can decode MQA upto 352.8. I believe my DAC can go up to 32/768.


@danny can you assist with the below screenshot. What’s the ORFS bit and the enhanced at the top. Not seen these before.

1 Like

ORFS is the original rate… its the thing that MQA rendering can theoretically take you to. It’s a form of upsampling with certain filters and coefficients chosen for the stream. Because upsampling alters the audio, we can’t call it bit-perfect. The goal is to improve the sound, so it becomes “enhanced”.

1 Like

Try the Dead Weather -cut like a buffalo or Rage against the machine. Just released this week. I tried them last night and thought they were stunning live sounds.

2 Likes

Thank you for your suggestions. Haven’t listened to RATM for a while.