Time To Dump Tidal? Bluesound / BlueOS Now Has Deezer HiFi

[wushuliu]

I just compared Deezer HiFi desktop to Tidal and they most definitely do NOT sound the same. VERY different presentations to me, and unfortunately the Deezer is not as good to my ears. Yes I made sure it was on 16/44. I compared several albums back and forth and the Deezer sounds slightly muted or rolled off like it has a dip in mids somewhere. This would sound smooth and pleasing on speakers with excess or poor quality treble/upper mids. Sins of omission, etc. But low level details and depth is absent. Maybe this is an issue with the desktop version.

 

I went through listening comparisons of the top streamers some years ago and they most definitely varied in presentation, with Spotify being the absolute worst. MOG though, was the best... too bad they went under.

[wushuliu]

2 minutes ago, Cebolla said:

 

The Windows version of the Deezer HiFi/beta desktop app doesn't allow you to set the audio output device and in particular to one in exclusive mode, unlike the Windows TIDAL desktop app. So the Windows Deezer desktop app forces you to use the default Windows audio device with its preconfigured audio settings (meaning you have to make sure that too is set to 16/44.1kHz to avoid resampling by the Windows audio mixer) and can unfortunately be shared by any sound producing application (including Window's system sounds) - not really ideal for 'hifi'.

 

Yes, all that was checked. It's the kind of presentation that some people would definitely prefer especially as most speakers/headphones, especially wireless/portable have unrefined upper mids and treble, which is fine. It just highlights though how little transparency there is when it comes to anything streaming. Clearly some kind of tweaking or processing was done. I'm sure Tidal does it too, but in the opposite way that maybe highlights rap/pop. Kind of like Kodak vs. Fuji film stock back in analog days, or ATI vs. Nvidia video cards. There is a subtle 'house sound' or look to fit branding and differentiate.

 

 

[wushuliu]

12 minutes ago, Abtr said:

 

I am very sure that Tidal does not in any way 'tweak or process' the sound, and I presume Deezer doesn't do that either. Both (HiFi) services stream *lossless* audio files, but a Deezer stream must go through Windows audio mixer and may no longer be lossless when it arrives at your DAC, and the extra required processing introduces electrical noise that may ultimately interfere with analogue signal reconstruction in the DAC..

 

1. You have absolutely no idea what these companies are doing. So any kind of speaking from authority on that is pointless. That's why we are having dust-ups like with MQA. There is zero transparency when it comes to streaming. Zero. Unless of course you know something we don't, in which case any illumination would be helpful.

 

2. What I'm hearing has nothing to do with the Windows mixer. I have listened to these albums on multiple formats both lossy and lossless, including Google Play and rando Youtube uploads - the audibility of compression varied (some in mp3, some lossless) but the presentation/tonality/eq/whatever were more or less the same and not like Deezer. As I said, it's possible the desktop App is to blame, but the windows mixer is not the culprit here. My PC is custom builty (by me) runs on LT3045 powered SSDs with an dual mono ES9018q2m dac with XMOS u208, itself powered by an LT3045 and IFI iPurifier. I keep track of my OS installs and sound quality with both fancy stuff like Jplay or basic playback software like MPC-HC, etc. etc. etc.

 

I've been doing this a long time.

 

At any rate, this just MY feedback. Not an absolute statement. Anyone can take a few minutes and compare for themselves.

[wushuliu]

10 minutes ago, crenca said:

 

Wushuliu,

 

Does not what you say about zero transparency in #1 also apply to #2?  If I am not mistaken (I could be - do you have information?), there is "zero transparency" as to what the Windows mixer does, even when set to what (to the end user) appear to be the optimum and non-interference settings.  My personal sense is that when set up correctly, it is relatively non-interfering but that is subjective...

 

You are correct in that we also have zero clue about the windows mixer and what it does with its funky resampling, which is unfortunate. What I'm saying is I've done playback via the mixer plenty of times so I know what to listen for. It doesn't roll off frequencies or smoothen out tracks. Heck, I'll go so far as to say as of the Creator's Update the W10 usb drivers and whatever resampling they've done sometimes sounds *better* than when I bypass with software like Jplay/Minimserver, etc. So I agree, especially these days, stock Windows playback quality is the best it's ever been. Hence my saying it's not the culprit.

 

I suggest people just compare for themselves. Maybe there's something I've missed. Maybe the desktopp App (which is beta?) is doing something screwy. Anyway, it's good to do these comparisons and to do them in as ways as possible to get a sense of what you hear and how you hear. I do it all the time.

[wushuliu]

11 minutes ago, Abtr said:

 

It can actually be tested whether an audio stream is lossless (bit perfect relative to the original) or not, and I assure you that Tidal hifi is lossless (not tweaked or processed as you said). And it seems very, very unlikely that Deezer purposely rolled off the treble of their entire catalogue, or (alternatively) processes/tweaks all downloads on the fly..

 

I never questioned whether or not they were lossless. Being bitperfect has nothing to do with what I'm talking about.

 

As for the latter, yeah sure, who knows. Maybe if they were all transparent with their process we wouldn't be having all these silly threads. There is nothing to stop them from doing so. The same goes for all streaming, frankly, including Netflix, Hulu, etc. 

[wushuliu]

1 minute ago, Abtr said:

if the streams are bit perfect (a perfect copy of the original) when they arrive at your DAC, then there shouldn't be any sound difference.

 

And yet... there is. Duh Duh Duuuuuuh.

 

Again, bitperfect has nothing to do with it. These files don't come from some Platonic Master Pure Source Matrix. We have no idea where these files come from, how they are delivered, what the masters are. We know nothing. Nothing.

 

 

[wushuliu]

4 minutes ago, Abtr said:

 

Listen wushuliu (and crenca), first, this has nothing to do with MQA. Second, one can simply capture a lossless redbook stream from Tidal (or Deezer) and verify that it is bit for bit the same data as is on, e.g., a CD of the same master. Period. There *can't* be a sound difference between the same losless files from the same master. Any perceived sound difference must be caused by the use of different masters or by Windows mixer or by specific galvanic issues of different hardware used. Your suggestion that Tidal 'tweaks' the sound of downloads is absurd. Duh.

 

You're getting a bit worked up and cherry picking my statements to fit a particular strawman. But let's play anyway. So what you're saying is that I can identify whatever is streaming as having come from the exact same master as a CD? Regardless of pressing? That there is indeed a Master Source from which all streaming and compact discs can be traced via bit analysis? 

[wushuliu]

Actually, even better, what you're saying is I can trace the streaming data to pin point exactly which

13 minutes ago, Abtr said:

 

There *can't* be a sound difference between the same losless files from the same master. Any perceived sound difference must be caused by the use of different masters

 

Also isn't that one of the issues I brought up? That we don't know what masters are used? So you're implying that yes we can determine what masters were used. In which case, I would like to know how so I can confirm that on each platform. That would be very helpful.

[wushuliu]

7 minutes ago, Abtr said:

Yep. What you call pressing, is the master. And note that an album may have multiple (re)mastered versions which will sound significantly different.  

 

And how do I do that. How do I determine what master/'pressing' was used from say, a Tidal track. What are the steps?

[wushuliu]

15 minutes ago, Abtr said:

Digital recording of streaming audio can be done with e.g. Audacity (free software). See, Windows WASAPI loopback recording:

http://manual.audacityteam.org/man/tutorial_recording_computer_playback_on_windows.html

 

 

Aren't there more steps than that? The goal here isn't to verify if my output is bit perfect. Plenty of info out there on that. The goal here is per your earlier to posts, to trace back which CD pressing/release was used as a means of determining partially, the 'master'.

 

Trying to get a handle on what you've outlined because I don't see anything about this online.

[wushuliu]

11 hours ago, Rexp said:

It's not very very unlikely, if you bothered to compare, you would agree its highly likely Deezers output is rolled off. Perhaps someone with a scope can check it.

 

18 hours ago, Abtr said:

Bit perfect is a relative term. Your output is bit perfect (hopefully), but a file is bit perfect or losless relative to the master. To determine if a file streamed from Tidal is bit perfect, you need to capture the stream from Tidal and compare it to e.g. a file ripped from a CD that is known to be from the same master.

 

But this is not answering the current question of what the process is for verifying. I've been trying to take the approach of a neophyte so that the explanations would be clearer but that's not working. So:

 

What you're saying as far as I can tell is that if I say, use Audacity, that it has the means to capture a track from say Tidal. You don't say if this involves some kind of metadata pulling or if this is just old school same time recording while another program is playing.

 

Then I would take that captured track, buy whatever CDs have that track, rip the tracks from those different releases (if any additional) and then proceed to compare something, I'm not sure what, in order to determine that the captured track is bitperfect and also thereby verify which release or related 'master' the track was pulled from. Or perhaps use something like dbPoweramp to match what pressing/release/'master' whatever? That is the only way what you have outlined makes sense, and I have never read anyone anywhere use those methods  - or any method - to trace back a streaming track. If there's a link that would be super helpful.

 

To top it off, this digression wasn't even the main point of my earlier post. But I kept with it in hopes there was some new development I missed. I don't think there is. I think we are talking about two different things. 

[wushuliu]

18 minutes ago, Abtr said:

The point is, as far as I'm concerned, that you stated earlier that Tidal tweaks the sound of its losless (bit perfect) downloads. People have of course checked if losless audio streams from Tidal indeed contain the same bits as the original recordings (e.g. on a CD) and they do. So you are wrong. And I didn't so much pointed out a method to check this for yourself; I just indicated that it can be done and has been done. If you're interested, google the subject..

 

Okay, well, I googled. So far I have found two offhand references - one by RexP actually in another thread saying to just compare tracks like you've been saying - and then a reference by someone on SuperBestAudioFriends or whatever that they compared and had a match. No specifics on the process, nothing. That's it.

 

HOWEVER

 

During my googling I DID come across something interesting. Many, many articles on this. Turns out Tidal does loudness normalization on all of their tracks. Tons of articles on it, totally verified. 

 

I think normalization by any measure qualifies as processing : )

[wushuliu]

Just now, Yorkie said:

 

Wouldn't playback volume information be added as meta-data?  If so, volume normalisation wouldn't affect the music bit-stream but might affect how it comes across in comparisons because the amplifier volume setting is not consistent.

 

Don't know. But it sounds like you're saying a stream can have processing and yet still be 'bit-perfect' : )

 

There is an additional link in the article I referenced above to a study done by the person Tidal consulted on implementing what sounds like a proprietary loudness normalization method. I'll read it later when I have time. 

[wushuliu]

8 minutes ago, Abtr said:

Normalization can be turned off for Tidal's less than losless streaming options. For losless streaming (Tidal hifi/master) in exclusive WASAPI mode, normalization is of course disabled.

 

So a normalized track from Tidal should fail the bit perfect test, correct?