TIDAL to add 'millions' of Master Quality (MQA) Tracks

Yes, you keep finding the same flagging error. Nothing is being decoded, there isn’t anything to decode. Either Roon flagging is wrong or (conceivably I guess) something in the MQA bitstream is mis-set. Ask the Roon guys.

Just for Reference:

Donald Fagan Nightfly shows up as 48kHz MQA on an MQA DAC.

In the case of a true 44.1/48 k file (not meaning the Flac data rate), the Core shouldn’t be active, correct? The file should be sent directly to the renderer at 44.1k or 48 k, for subsequent upsampling within the DAC.

Have you another zone linked as you play this. I have seen this in my Meridian zone when I have my Pulse Flex and Meridian 210 zones linked.
I think the Roon unfolding takes over to support the flex and this is also repeated in the Meridian zone

Does this apply to 16 bit MQA ?
Before my understanding was that MQA only could be unfolded if 24 bit.

1 Like

Just to be clear. I don’t think Donald Fagan 24/48 is an error being unfolded to 24/96. But you indicated so.

However I accused Warner for doing upsampling of a original 16 bit recording. No 24 bit master exist. This has nothing to do with MQA. (But of cause Warner used this file to create a MQA. Maybe it unfold further in a DAC ?).

I’m talking about 16 bit MQA can’t be unfolded. Pr. definition of how MQA works.

Also I’m only talking about Roon doing fist unfold. Whatever type of MQA file is offered. If I had a MQA DAC I could tested more if there was any difference between how Roon and a DAC handle 16 bit MQA.

play the accused track in Audirvana, still the same?

I don’t know. MQA is a complex technology with many options.

Let me try to ask in a different way then.

Is it a correct behavior that Roon is unfolding a 16/44.1 to a 24/88.2 ?
Or can you have someone verify ?

1 Like

Yeah, this has puzzled me too. Haven’t gotten worried about it yet - but it does seem strange.

You keep using the term “unfolding.” That seems to be the hang up. For MQA, all unfolding is decoding, but not all decoding is unfolding. If not >48 kHz ORFS, then no ultrasonic encoding, nor unfolding.


1 Like

Andrew, you persist… What @R1200CL and others are saying that these tracks are MQA authenticated at 44.1Khz or 48Khz, they should NOT be “unfolded/upsampled/expanded” to more than the authenticated sample rate.


As I suspected.
Would be interesting if anyone using Audirvana can report same issues. I would assume the error is on MQA (the company) side.

Maybe this is your department to take necessary actions ?

This is definitely regional , i still can’t see the Pink Floyd stuff that started this thread South Africa the poor relation as ever , I’d check on Qobuz but we don’t have it , or maybe Amazon HD if we had that as well


I think you’re adding some confusion here.
But it’s interesting to see you have the CD version available. I add a circle :o: where your version has CD written. An interesting detail.
However this album is also being decoded into 24/88.2 when playing as any probably 16 bit MQA album will be. (I assume you have master available and hooked of in your Tidal subscription).

Can I try this argument once more? This album has an original rate of 44.1 kHz. (Where it says MQA 44.1 kHz). There is no high resolution information above 44.1 kHz, therefore there is no decoding or unfolding to do. The data will be sent to the DAC for upsampling from 44.1 kHz to the higher rates needed at the sigma delta modulator. What appears to be the error is that Roon is still showing a decoder flag in the path even though decoding isn’t active. Probably this is a software error in Roon.

Another possible error is that one of the signal bits in the MQA file is mis-set. However there is no unfoldeding to 88.2. I don’t have access to Pink Floyd but did check your earlier Donald Fagan NIghtfly example. That shows up as 48kHz MQA both in Tidal and in Roon on MQA DACs, so it’s not being unfolded.

It could be there is mqa bit depth metadata inside the file that is set to 24bit regardless of the file is originally 16bit or a 16bit file exported as 24bit flac (that pink floyd album was 24/44.1 before). the core decoder upsample it (44.1 to 88.2) and use the bit depth metadata to display as 24bit/88.2.

you can export a 16bit to 24bit and the file size is approximately the same, there is a difference in sound quality - the 24bit is a little smoother. (very subtle).

The core decoder doesn’t do upsampling. It just unfolds, assuming there is high res data.

The size in MB of a 24b file and 16b file are very different. If you’re seeing the same size in MB and you can calculate that the expected sizes (MB or Mbps) for the two data rates match what you’re seeing, then it’s likely that the 16b file is zero-filled to 24.

I think it would be more correct to say that the core decoder decodes to MQA Core, which is 2x the sample rate of the delivered MQA, and it does this for *all* 1x (delivery) rate MQA. I don’t know exactly what is happening in the case of 44.1/16 (and 44.1/24) original sample rate material.

1 Like