Recent Tidal Masters

[coliny]

Recent Tidal Master give error message when I try to play them:-
File Allocated but cannot play...........

I found that the track .flac.mtt files do not contain a value in MQARATE :
Looking at MQA albums that do play I saw that the MQARATE : should be twice the rate displayed by XXHE when you add an album to Play list.
I edited the .flac.mtt files to include the appropriate MQARATE and album decoded & played OK.
Example: https://tidal.com/browse/album/228798472

See attached edited file

Colin

[PeterSt]

Hi Colin,

I think we saw this maybe 6 months ago already. Or I saw it myself - not sure any more. So I solved this for myself (due for public upgrade of course) but things keep changing out there and I never get around to finalize it. And Yes, the other day I really worked to get it done finally, but something else was the matter again which didn't motivate me to now sort *that* out again. Anyway I recall that the MQARATE was not obtained in the FLAC file any more, so I had to change half of the code to get that to work again.

Meanwhile ...

Meanwhile I am not motivated into MQA much any more because Tidal is changing more and more to even fake MQA with as key element that the originals disappear, which in the end leads to losing interest in Tidal as a whole.  Now *that* in itself did not happen yet, but I had to revamp the whole Tidal searching because *that* changed over there and finishing that part seems to be something which is too much for me.

Summarized, these external "cloud tools" just do not work out for me, because they change things when they like it and don't even tell about it (we, the user, just run into it and must Google to find out what the h*ck happened).

From a very long time ago, people may recall that one could buy albums from Amazon. But the same thing : they kept on changing the interface (or API if you will) and each time caused many days of work. In the end I eliminated the functionality.

And so I keep adding new or corrected stuff and never create an upgrade for you guys, because each time something else is working "half" (can be avoided when I am faster - haha).

Regards and thanks a lot for sharing, as always !
Peter

[acg]

Peter, streaming Qobuz sounds better here than streaming Tidal...it is not really even close, and I am considering giving up my Tidal subscription because if Qobuz has the track I will just stream it from there.  Crazy, I know, that different streaming services sound different, but Tidal has really dropped the ball on sound quality.

[Robert]

I agree "acg" gave up on Tidal last year. Qobuz definitely better than Tidal.

Mind you I'm not listening to any streaming for serious listening now. Only to find new music which is getting harder these days with DR going down to 3 and 4 on some new electronic music. How low can you go!!!

[coliny]

In defence of Tidal/MQA I have been going through my previous downloads and adding the MQADATA rate to the .mtt files (see first post of this topic) and they all unfold and play OK. Had to do this to all downloads in past 12 months. Amongst them were 44.1, 48, 88.2 and 96K before unfolding files.

I did find this album
https://tidal.com/browse/album/60714019
 which I downloaded 17 months ago and already had
MQADATA d44100 played but did not unfold. change to d88200 and files unfolded and played OK.

So for the time being at least I am happy using Tidal & editing MQADATA if necessary .

Colin




 

[coliny]

To speed up editing the MTT files I have created 2 macros using Macro Toolworks free edition.
Download and install it:- https://www.pitrinec.com/download
Copy the 2 attached macros to the Toolworks macro folder.
Start Toolworks and open both the macros.
Attached jpeg MQA Macro Start shows what it should look like.
Open the first .MTT file and position cursor below last line of text (as in jpeg).
When you stream a Tidal album note sample rate, the MQARate will be twice this.

If MQARate is already there without a value just highlight it before running the macro.

ctl+8 runs the 88200 macro or ctl+9 runs the 96000 macro.

Result should be like attached jpeg MQA Macro Done.
Close MTT file and repeat for remaining MTT files.

I have streamed a lot of MQA albums and all play well after editing the MQARate. Even 44/16 MQA albums work.

Good Luck & Happy New Year
Colin

[PeterSt]

Colin, man, invaluable !

I wish I could check it thoroughly, but I can't anymore because it was solved for myself. That is, "half" or so, because what I see from Tidal can't be trusted any more. Explanation :

Say half (if not more) of MQA albums are now totally fake. This means that they don't have an unfold at all, but are still denoted MQA. This is not what I see back in your workflow, so I can't see how that must be dealt with (by you, by your scripts, by us users who (apparently ?) must deal with this by eye. ... I hope I make sense so far ?

I recall that the last thing I did - 4 months or so ago - was skipping the MSQ 44.1 (or 48 would they exist) which don't unfold, at all. This is because a. it is too vague to deal with and b. it is literally useless. With the later I mean : I am not enthusiastic about a changed album (processed by whatever means) which does not even have a higher sampling rate than how it is presented at "download". Thus, I avoid those as the plague. XXHighEnd thus avoids those (I forgot for this moment how this shows - probably not at all (the album does not appear in the Library area).

Coincidentally I started to explicitly play MQA's again, lately. What I should do now, is obtain MQA's again, because I stopped with that, say 4 months ago (this is because XXHighEnd was half-finished for them). That way I can see better what is and what is not good - but probably that job is finished already.
On a side note : that leaves the selection (Search) for the albums in general, which is finished for Artists (and works more than great) but must still be done for the other means (Album Name, Track, Album ID). This is because Tidal changed things (a year or so ago) without even telling about it.

Anyway Colin, what you did looks great. Maybe you can elaborate a little on the 44.1 albums which don't unfold anyway, and how you deal with those then.

Cheers,
Peter

[coliny]

Hi Peter, try this one:-
https://tidal.com/browse/artist/9985127

Its 44/16 unfolds to 88 and plays fine after editing the MTT files.

Quite a few recent additions to Tidal are 44/16. Apparently CD's with MQA are available now so I think they are genuine.
https://www.stereophile.com/content/mqa-encoded-cds-yes

I have worked through my library of Tidal downloads and edited the MTT files and they all play.
Just a few would not play with a custom filter but they all play with Arc Predict.

I use the Tidal App to find albums & copy/paste the share link into XXHE TT search.

Regards
Colin

[PeterSt]

Hi Colin and all - Happy New Year !

Its 44/16 unfolds to 88 and plays fine after editing the MTT files.

It is a technical thing; You can't "unfold" to double the rate without adding a bit (I am referring to the bits in bytes). Thus would you do that, then level information is chopped off. This is not easy to grasp ...

When the 16/44.1 is transferred to 24/88.2 (sorry about using my own bits/sampling rate language), then you'd say that a sufficient amount of bits have been added (namely . But this has to be approached from the other side : how can the 16 bits contain information about MQA while not FIRST consuming bits for that.
This is how a 24/44.1 is different, because the required headroom for level information is still only 2x for 2x more sampling rate. Very indirectly I am now saying that a 88.2 native file, only needs to be 17 bits and not 24 as how it is presented to us. But *that* is because how bits and bytes and converters work - they jump by the 8 (one byte) although 4 would also be possible. But not 1 or 2 or 3 or 5 etc.

From MQA which really unfolds it is known that only 17 bits real data are used. Maybe mr. MQA / Bob Stuart was not really stupid, and he apparently knows a few things.

Now to the conclusion :
When 16/44.1 would be liable to fold, then first at least one bit has to be removed from the 16 bits, if not three or whatever the reality is.
When 16/.44.1 would be encoded as MQA, in those original 16 bits (from the orginal album) at least one bit is not original any more. And since it is encoded MQA and since this is not only obtained in the FLAC file any more (as you discovered, like me - something like over a year ago), then it *has* to be the file itself which is molested. And since 16 bits is not already all that much it is, well, molested.
Apart from these more vague conclusions, the main conclusion is that when 16/44.1 MQA is transferred to 24.88.2, nothing unfolds as such - it is upsampled only.

Maybe now it is clear better why I simply avoid these 16/44.1 - they are fake as fake can be, plus technically worse.

Additionally to all there's this :
The official MQA decoding standard as of the last 2 years (don't pin me down on that period of time) is that no MQA denoted album will ever play at 44.1 or 48. It will only play at double the base rate (the rate in the file). The only player which does not comply to that is XXHE. So the only player where you can still see the truth (whatever that's worth) is XXHE. You could also say that from the start I did my stinking best to show all the various truths within MQA (try that with Roon or the Tidal app etc. and you will get nowhere).
The kind of verdict of this little story is that nobody is able to see any more what's 16/44.1 because ALL is upsampled to begin with, whether it is justified or not. Thus, whether you let "unfold" your 16/44.1 MQA to 24/88.2 or let e.g. XXHighEnd do that, it is no less than upsampling - and it is up to you whether you apply that or not (sampling rate slider). However, what happens with MQA now, is that good 16/44.1 albums are transferred to (processed with) MQA fake stuff, and next they are upsampled. A more worse situation does not exist. Better would be your beloved 16/44.1 being upsampled by good means to 24/88.2, or what we are used to : to 24/705.6.


Last thing : *Because* XXHighEnd's MQA does not contain an unfold from 16/44.1 to 24/88.2 (it really hasn't), all what would be happening is that by the means of your tweaking of the .MTT file, XXHighEnd sets the sampling rate to 2x the original, and it will upsample herself to that, prior to your sampling rate slider being involved and upsampling further (like to 705.6).

Maybe it starts to be clear somewhat how it virtually has become impossible to make something out of all this mess, with Tidal in the background also changing things (and a next dimension being that FLAC file which *was* some last resort which its header data - but which got killed).
It is not undoable (for me), but it requires reasoning around false logic and untruths, coincidentally Bob S. in the front row with that again.

I will (kind of) repeat : there is no logic in what to do with a 16/44.1 (or 16/48) because whatever the logic, it is a lie. And thus in my version over here I just don't present them for search result. That avoids the problem, right ? ehh ... Yes. But because Tidal is changing about all (or at least a LOT) to 16/44.1 MQA, the albums now become unavailable. And it is me who decides that for you. :-( Understand ? this does not feel good at all and it implies infinite postponing of an update (of XXHE) I sure could have.
But one day I will do it anyway ...

Kind regards,
Peter

[coliny]

Thank's for the explanation. The justification of editing the 44/16 MTT files is just that they do play in XXHE with MQA decoding enabled, so I don't have remember whether I have to turn off decoding.
I fully accept the 44/16 albums in Tidal and marked as MQA will not have best SQ but if I like the music I can always look elsewhere for a better version.

Regard
Colin

[rudolf]

  thanks coliny, your workaround works fine and is also quite important to me !!!
  thanks peter, i keep coming back to xxhighend all the time (whether there is a 2.12 version or not... )!!!
P.S.: MQA files (at least the 'real' ones) sound good in xxhighend for me ...

[coliny]

Hi Peter

You established that the 44/16 Tidal Masters don't unfold in XXHE and I agree with that finding.

Do we know for sure that XXHE will unfold genuine MQA16. If there is doubt about this I thought it might be possible verify this by buying an MQA-CD and ripping it for replay with XXHE. There are MQA-CD's which can be purchased on ebay.

Best Regards
Colin

[PeterSt]

Hey Colin,

The situation is IMO more vague and therefore more complex to explain / reason out;

A CD will never upsample in a normal CD Player. And if it is MQA adaptive, I think it still will not do that. If I am correct in this, there will also be no unfold.
Additionally I already know there is no unfold. Why ? because it is known that it is a 16 bit original. So what could happen is that it can be upsampled, but THAT is only fake and in the end the very same as most software these days do - including XXHE. One thing : this upsampling is for a very good reason : digital (reconstruction) filtering.

Combine this with the new rules for all MQA adaptive software (like Roon) : it will upsample by default. This - as how I see is - was literally sneaked in 3-4 years ago, and only later MQA & Bob claimed to have 16 bit MQA. I say : BS. But the nasty thing is : you can not see this any more when the software (like Roon) is dictated to upsample 2x by standard (read : it is not allowed under the MQA regime to bypass that).

And so you are fooled.
But not by me.

I have no doubt that you, Colin, are able to tweak the MQA 16 bits in your current version of XXHE (I can't do that any more) so you have it "unfolded". OK, now ask for a Spectogram of it. Show us that Spectrogram (after trying a few albums/tracks) and/or give your judgment of what you see.
One thing : I can show you numerous (if not about all) Spectrograms of MQA 24 bit unfolding to 88.2 or 96, which all show proper HiRes (while these ever back were very hard to find on e.g. HDTracks).

Best regards,
Peter

[coliny]

Hi Peter

Updated 2.2.23:-

I generated some spectrograms 48/24 & 44/24 MQA files do show data above 24/22 KHz but no 44/16KHz MQA files I have tried show any content above 22KHz.

Bob Stuarts info does imply to me that MQA16 should also decode (unfold).

https://bobtalks.co.uk/a-deeper-look/deeper-look-mqa-16b-in-the-last-mile/

Best Regards
Colin

[PeterSt]

Hi Colin,

You now seem to turn things upside down. :-)
If you tweak that .MTT file with btw :

Even 44/16 MQA albums work.

(your own text)

It *is* decoded. But nothing will happen because ... I say so ?
I am afraid you listen too much to Bob.

Obviously I asked you for a reason to make that Spectogram. This is because it will show nothing above 20.050 - at least to my expectation.
Now upsample them (like Roon or anything does). Now it will show ...
... data which is rubbish.

You can do that yourself by upsampling 2x and grab the UnicodeTrack0001.wav file (in your XX folder). Now put that in a normal music folder and make a Spectrogram of it.
This is nothing worth much and only for your reference so you can see the "fake" happening. This is no unfolding of course. This should have (if so) shown by the Spectrogram you made.

Kind regards as always,
Peter (not fed up by this at all !)