Saturday, 1 June 2019

POLL RESULTS: Music streaming service adoption among audiophiles in 2019...

Click on image to zoom.
Yeah, I know. Music streaming as a subscription model is the latest "revolution" for music lovers. This is why I decided to put up the recent poll in mid-March while we were concurrently also running the blind test as reported on in the last few weeks. Above, you see the overall results based on the visitors to this blog.

Remember that for the poll, I allowed voters to select up to 3 options. As a result, even though there were 616 respondents, a total of 809 selections were made. This means that up to ~30% of people selected more than 1 option and these are represented in the percentages above.

Here are the top 18 countries the respondents came from:

US: 210 (35.7%)         Great Britain: 57 (9.7%)   Canada: 42 (7.1%)
Germany: 36 (6.1%)      Australia: 28 (4.8%)       Netherlands: 27 (4.6%)
Sweden: 16 (2.7%)       Belgium: 13 (2.2%)         France: 13 (2.2%)
Denmark: 9 (1.5%)       Switzerland: 9 (1.5%)      Poland: 9 (1.5%)
New Zealand: 8 (1.4%)   Norway: 7 (1.2%)           Finland: 6 (1.1%)
Spain: 6 (1.1%)         Romania: 5 (0.85%)         Italy: 5 (0.85%)

That geographically covers the top 86% of all respondents (in total, there were responses from 49 countries). Since filling out a survey like this is easy to do and there are no special time or equipment requirements like a blind test, this also gives you a good idea of where the readers of this blog come from. It might also be a good indicator of where audiophiles in general reside. Note that there is an under-representation of Asians here likely due to the language barrier plus there are countries where the general populace will have no access to this blog - for example, China is a big one due to the "Great Firewall" in place unless using a VPN (which is officially illegal for citizens there). However, looking down the list, there are some Asian locations of note here - Hong Kong 0.68%, Singapore 0.68%, and Japan 0.5%.

From my vantage point in mid-2019 as I write this article, we are living in a time where internet penetration is essentially ubiquitous - over 85% in Europe and North America (source). We can see that average broadband speeds are quite good these days in the developed world - have a look at the SpeedTest averages by country. Mobile download speed of >5Mbps is common, and fixed broadband speeds are close to or exceeding 100Mbps through much of the world. In other words, probably the majority of people in the world now can have access to high-speed internet and if they want to pay for a subscription to a streaming music service, can have that as well.

As usual, I would not consider the blog readers here to be "typical" music listeners. While there were almost 600 voters, my suspicion is that these results are indicative of more advanced audiophiles who are quite up to date with computer technology and digital streaming (perhaps demographically similar to the recent blind test results).

Within this group then, I think it's interesting to see that already almost 40% use a paid lossy streaming service and about 44.5% with lossless streaming (including hi-res capability) - remember that some use both types of services. That's great "penetration" suggesting more than half of us here already have another billing item on our monthly "utilities" :-).

Since a large number of voters were North American (about 43% of respondents from US or Canada), it's no surprise that Tidal takes the lead as the lossless music streaming service of choice with 25% of the respondents using this system. As you can see, while I acknowledge that Tidal offers the claim of "high-res" MQA, as you know, MQA unfortunately does not substantially improve resolution. When the poll started, Qobuz had just made its debut in the USA and still unavailable in Canada so we could see it gain market share in the years ahead.

If you currently are not a streaming subscriber, you're not alone! 27% of the respondents interestingly indicate they don't have a streaming music account - free or otherwise.

Remember that in reality we all have "free" (advertising subsidized) access to essentially all the music we could ever really want to hear in the form of YouTube. So although the idea of having access to a streaming service in order to "preview" new music and then buying the CD, LP or digital download is not untrue, while perhaps less convenient, this can already be done freely with YouTube for the majority of music (and music videos) we might be searching for. The benefit of a streaming service in my mind is more with convenient creation of playlists, sharing these playlists, and the benefits of the curation and recommendations the services might provide.

Needless to say, ownership of music has its privileges as well as responsibilities. Making backups of our digital music libraries and metadata management are essential "hygienic" practices. The connoisseur can pick and choose different masterings and releases. There is no need to worry that albums might be here one day and disappear the next. Furthermore, one does not have to be concerned about decisions a streaming company might make that one could disagree with (eg. Tidal picking MQA as their "hi-res" streaming codec). Also, I've heard that certain music may be watermarked when streamed but the CD version is clean (some of you out there might be experts in these details). Sure, one could argue that a "cloud" service might provide "upgrades" to the music library and provide even better masterings of the same album... My caution with that is the lack of more dynamic remasterings over time rather than being more natural sounding. [BTW, I am reminded of back in the day when Pono promised us "the finest digital copy" but never could deliver.]

What's perhaps most interesting to me in this poll is that only around 5% of the respondents are "considering" subscribing to a lossless / lossy service. It's interesting because this suggests a saturation among the audiophile demographic here. It looks like most of us may have already tried a music streaming service or two and already signed up if interested in subscribing already, rather than sitting on the sidelines still and contemplating. Remember that Spotify has already been available for 10 years, Tidal lossless since late 2014, with other lossy services like Apple Music in 2015 and Amazon Prime Music since 2014 plus their "Unlimited" subscription since 2016. Technological progress and adoption can happen quite fast, especially when the entry cost isn't bad at all currently at around ~US$10/month for most lossy subscriptions and ~US$20/month for lossless.

It'll be interesting to see how this shakes out in the years ahead for audiophiles. No doubt there will be growth; the question is by how much and whether the streaming audio space gets further consolidated and fewer players.

It would be interesting to see whether lossless 16/44.1 becomes a "standard" at some point though I will not hold my breath. I suppose it could happen when one of the giants in the streaming space jumps in... For example, Apple could start streaming ALAC lossless compressed or Amazon jumps in with lossless FLAC. Doing this would be threatening to the likes of small players like Qobuz and Tidal. The question is whether the general public cares! Remember, a jump to lossless compressed 16/44.1 would imply a 2-3x increase in bandwidth needed. For enthusiasts like many of us here, this is simply trivial, but the mainstream end-user will need to consider limits to the data plan if streaming through a cellular network, and the large music services will quickly see a multi-fold increase in data transfer and possibly storage needs.

Is all this worth doing if we remember that sonic differences between high bitrate lossy (like MP3 at 320kbps) and lossless 16/44.1 are for the most part inaudible, the music is consumed as a "preview" for possible purchase, it's being played in the background, or in mobile setting? Claims of "better sound" are more of an advertising pitch than tangible reality for the majority of listeners IMO. These issues are likely why a giant like Apple hasn't switched to lossless already; the benefits are unlikely to outweigh the costs for the majority of users in the foreseeable future. At some point this might change if competition demands it but I don't believe this is a given - people will not pay more unless it's a clearly noticeable upgrade.

I listened to this interview with Roon's Rob Darling a little while back and I have to disagree that a high quality tier is coming to all streaming services (start ~45:00) since I don't think the hassles are worth it. Also, I double disagree that MQA will be the hi-res standard as he suggested at ~46:30 - no way (again, not just because MQA is a poor codec in many ways). I think hi-res streaming, just like hi-res music in general is destined to have minimal demand regardless of how much the music industry desires to convince listeners of its supposed benefits. Hi-res has failed to entice physical media buyers (SACD, DVD-A, even Blu-ray). I certainly don't see hi-res doing particularly well with the handful of digital download options over the last decade. And now why would the masses demand hi-res streaming? Human perception simply has no need for it particularly with the type of music production used these days. Let's see in the next 5 years if hi-res has any more interest/demand than today. (I suspect this is similar to the unlikely prospect of us having plentiful native 8K video for the home in the foreseeable future even if 8K hardware were to be widely available.)

A monthly music streaming subscription is a lower tier utility (below Internet providers, cell phone companies, or video streaming companies like Netflix). It's going to be about more competition between the sites as they try to secure market share as the rate of new subscribers plateaus.

Remember to have a peek at the "big picture" in terms of where the money's going as of 2018 within the music industry. Audiophiles as usual represent just a drop in the proverbial bucket... But I do think it's worth keeping an eye on what audiophiles are interested in as early adopters to see the trends ahead.

Oh yeah, regarding that interview with Roon's Ron Darling above. Yes. Please let us easily access our Roon server for streaming outside of the home (as discussed near the end of the interview). This is an essential feature already easily done with free Logitech Music Server for years. That would certainly be a compelling feature for the next version of Roon!

Happy June! Hope you're enjoying the music...


  1. Hi,

    Please check your PM.

    BTW, I'm a physical CD dude, digital downloader... if "streaming" includes visiting Youtube without login then I am a streaming user as well, I especially like to watch people who make rearrangement/transcription/remix and live orchestra/band performance.

    Those who voted MQA must be shills or trolls. Ignore them lol.

    1. Hi Bennet,
      Saw the PM tonight! Thanks :-).

      Naw, I suspect there is a good number of audiophiles here who use Tidal/MQA. That's cool... Recognizing again that the folks who visit the blog here are a unique group!

  2. Hi Archimago

    Thank you for your ongoing and fascinating blog.

    I'm one of those who subscribe to Tidal Hifi but only so as to get 16/44.1 resolution. I did have a few albums (20 - 25) in MQA but a couple of months ago after a bit of a serious comparison session replaced them all with the 16/44.1 versions and lost a digital sheen and harshness which my system revealed and happily got back sound as good as my CD rips. So, while I've ended up in your survey as an MQA user, that's not quite the case. FYI I do have a number of actual HiRez files and find that while subtle,there is a difference. MQA is nowhere near it. But as you say, the greatest difference is in the actual quality of the recording not the resolution.

    Kind regards

    1. Thanks Unknown,
      Yeah, over the years I have come across a number of comments in this regard. I was struggling in preparing this poll but it would be unnecessary I felt to differentiate whether one subscribes to Tidal for the lossless or whether the MQA "hi-res" inclusion made a difference.

      My belief is that MQA simply provides a "tick box" of sorts for Tidal to claim that it addresses "hi-res". Nothing more. I would be very surprised if MQA has added any kind of tangible boost in subscriptions at this point now that we are more than 2 years into the feature's availability (it was "turned on" early 2017 around the CES).

      I do hope that over time, subscribers like yourself will take the time to listen to these MQA releases and compare to true hi-res to experience whatever difference one might hear; especially since we know objectively that MQA is inferior.

  3. Ps
    I compared recent classical recordings available in both 16/44.1 and MQA expecting they would both be off the same master

    1. Thanks Pieter, nice to know you made sure to select proper music! As my comment above...

  4. Hi Archimago! Did you try out the "free with a cable tv subscription" streaming channels from Stringray Music available in Canada? You mentioned earlier you might have access to it. If you do, I suggest the "Jazz Now" stream. Very good current jazz music and great sound to my ears with headphones.

    1. Thanks Gilles,
      Yes, I have it on my phone and it does sound good. I keep it at maximum on WiFi 320kbps and thru the cell network "high" quality 160kbps. I assume this must be MP3 codec?

      Granted, most of the time I'm listening through AirPods on the go. So not sure I'm experiencing the "Jazz Now" in the best conditions :-).

  5. The thing is, not all lossy compression was created equal. I tried e.g. spotify VS tidal web player in firefox. According to what I found, spotify uses 256kbit/s AAC, while I haven't found what Tidal uses (I have streaming quality set to high). When playing the same song (Stravinsky's Firebird, Infernal Dance), capturing and looking at the spectrogram, spotify looks close to the flac version I have of that CD, while Tidal's spectrogram is mostly cut off at 12kz and 16kHz (with some peaks above that). I can also identify tiny "wobbles" in the drum rolls between the loud attacks.

    So, I'm left wondering if Tidal doesn't really "try" very hard to give the best lossy quality possible, and it seems logical from a business point of view: if you create different quality tiers, you have every incentive to push people to the higher quality tier, so if people try the lossless tier and notice an improvement over the lossy tier, they'll think it's due to the format and pay more for the lossless HiFi tier.

    Spotify quality info:
    Does someone know a similar site for Tidal, I can't find any details on their lossy quality so far.

    1. Thanks Steven,
      Very important comment!

      Indeed a modern LAME at 320kbps would be better than old-skool Fraunhofer MP3 at 320kbps. Of course we are not privy to what encoder these streaming services use.

      So long as no compatibility issues, I think it would be great to use high quality VBR (like LAME VBR 0) which would be less than 320kbps but the psychoacoustic algorithm can do its job to use the bits it needs.

      Yeah, I certainly would not be surprised to see multi-tier services like Tidal not try hard with the lossy service and "push" people to select the lossless / more expensive tier. That sucks I think and I hope what happens is that people just jump ship to something better - maybe Apple Music or even Spotify - if that is the case. It's rather dishonest even if still allowable...

    2. Indeed. Thanks for all the work you put in, I managed to go through a lot on your blog and discovered stuff I wasn't even aware of till now (high-res audio, mqa). Maybe I've been too happy with my music quality to follow up on what was going on with formats, I thought we were still comparing different lossy codecs for transparency until Opus came along, after that I haven't really followed any developments. Nice to see someone take the time to investigate things properly and report on it!

  6. summarized with pictures:

    1. Oh wow Steven! Great work.

      Hmmm... Maybe Tidal "Premium" (lossy) tier should have a good listen to the other $10/m services and see if they're getting the best from their monthly ten bucks :-).

  7. I'm one of those guys that subscribes to just about all the services: Apple Music, Tidal, Qobuz, just cancelled Spotify and Amazon Music. I don't see "downsizing" any more than this, unless I cancel Tidal and go with Qobuz only. Worth it (to me) to get just about 100% "coverage".

    1. Wow man! Sounds like you've had a taste of everything...

      Any thoughts about sound quality between the services? Ever run into any audible watermarking especially in the lossless services?

  8. Now, more and more players are choosing to buy WOW Classic Gold on ZZWOW. They realized that the Gold of ZZWOW is the cheapest on the market and will not be punished by the system. If you have any questions about this, I can guarantee you that is 100% safe.