[rlw6534]

No, there is a real difference, it's not word games.

Upsampling is adding digital samples (instances of time that were not in the original recording) to increase the sample rate, downsampling throws away samples.

MQA is lossy compression. It throws away parts of the signal per sample, not some of the samples- in order to be able to compress it. So for example a MQA-encoded song that was mastered at 96 KHz and compressed to 48 KHz has a nominal sample rate of 48 KHz before unfolding. When it's unfolded (decompressed) the 96 KHz version is recovered, albeit with some loss. There are no samples added (upsampling) or removed (downsampling). It's just lossy compression.

44.1 kHz masters that play back at 705.6 kHz and 48 kHz masters that play back at 768 kHz are clearly upsampled. I understand that the MQA process does more than that with the origami under the noise floor (for higher resolution sources) but the majority of MQA files are from 44.1 or 48 kHz masters so there was no high-res to start with. Any “unfold” beyond the first, is only up-sampling using a very short digital filter. Again, that is what the 4x, 8x and 16x terms refer to.

Maybe were are on the same page here, just not communicating well...

 

[gimmeheadroom]

44.1 kHz masters that play back at 705.6 kHz and 48 kHz masters that play back at 768 kHz are clearly upsampled.

There is no such thing, to my knowledge the highest MQA masters are 352.8 KHz or something like that. Not that it is impossible, but there are none so far that I know of.

I understand that the MQA process does more than that with the origami under the noise floor (for higher resolution sources) but the majority of MQA files are from 44.1 or 48 kHz masters so there was no high-res to start with.

I don't believe that's correct. I don't think any MQA plays back higher than the master. If it does, then sure that is clearly a case of upsampling.
There is plenty of MQA at 44.1 and especially 48 KHz that doesn't expand at all when unfolded.

Any “unfold” beyond the first, is only up-sampling using a very short digital filter. Again, that is what the 4x, 8x and 16x terms refer to.

I have not seen anything unfold higher than 352.8. I just checked the MQA pages (not very easy to find facts!) and it says there any unfold the hardware supports is possible. On Tidal there I haven't seen any 16X and only a small amount of 8X. It is reasonable since there are some PCM 352.8 masterings floating around.

Maybe were are on the same page here, just not communicating well...

I think your latest post helped clarify what we don't agree on, which is that to my knowledge no MQA expands higher than the master, while you seem to be saying it does.

 

[rlw6534]

There is no such thing, to my knowledge the highest MQA masters are 352.8 KHz or something like that. Not that it is impossible, but there are none so far that I know of.


I don't believe that's correct. I don't think any MQA plays back higher than the master. If it does, then sure that is clearly a case of upsampling.
There is plenty of MQA at 44.1 and especially 48 KHz that doesn't expand at all when unfolded.


I have not seen anything unfold higher than 352.8. I just checked the MQA pages (not very easy to find facts!) and it says there any unfold the hardware supports is possible. On Tidal there I haven't seen any 16X and only a small amount of 8X. It is reasonable since there are some PCM 352.8 masterings floating around.


I think your latest post helped clarify what we don't agree on, which is that to my knowledge no MQA expands higher than the master, while you seem to be saying it does.

As I said, MQA can be confusing (and complicated to explain). It doesn't help that they invented new terms (unfolding, deblurring) and never really explain what they mean. It's almost like they intentionally try to confuse (obfuscate) the facts for those of us that want to understand how it works.

 

[gimmeheadroom]

As I said, MQA can be confusing (and complicated to explain). It doesn't help that they invented new terms (unfolding, deblurring) and never really explain what they mean. It's almost like they intentionally try to confuse (obfuscate) the facts for those of us that want to understand how it works.

Agreed. I like the sound of MQA and it's not clear why. Some people have posited that the UMG watermark is not present on MQA albums. I remember the minidisc codec which is also a lossy, psychoacoustic codec but sounds better than it should. On the other hand, a big move by studios to offer only MQA and not Redbook or highres is bad. and we have a right to be concerned about that. It's hard to understand who benefits from compression in the 21st century. MQA made sense 40 years ago, today it's hard to understand what's behind this.

 

[rkw]

Some people have posited that the UMG watermark is not present on MQA albums.

UMG removed their watermark from streaming services during 2019-2020, and it is no longer a factor.
https://www.mattmontag.com/music/an-update-on-umg-watermarks
I can confirm it from my own experiece.

 

[gimmeheadroom]

UMG removed their watermark from streaming services during 2019-2020, and it is no longer a factor.
https://www.mattmontag.com/music/an-update-on-umg-watermarks
I can confirm it from my own experiece.

Thanks, this is important info. I guess it was you who was making us aware of this back then

 

[dangusyra]

Can you share your favourite test music Tidal playlists?

 

[gimmeheadroom]

Can you share your favourite test music Tidal playlists?

Hi, these are not mine but I tried some of them a while back.

https://tidal.com/browse/playlist/14dc73a4-27d3-4bd8-af18-644290c91509
https://www.reddit.com/r/TidalPlayl...28khz_mqa_master_albums_mixed_genres_updates/
https://tidal.com/browse/playlist/9e3435f8-7a19-410f-a58c-255cd2923d68

 

[peterinvan]

Can you share your favourite test music Tidal playlists?

https://tidal.com/browse/playlist/fa6622ac-b9ca-4482-97a4-24e6480b4449

I use this list for testing headphones, but it contains many of my favourites.

 

[dangusyra]

Hi, these are not mine but I tried some of them a while back.

https://tidal.com/browse/playlist/14dc73a4-27d3-4bd8-af18-644290c91509
https://www.reddit.com/r/TidalPlayl...28khz_mqa_master_albums_mixed_genres_updates/
https://tidal.com/browse/playlist/9e3435f8-7a19-410f-a58c-255cd2923d68

Link starting tidal.com.. does not open playlist, just redirect to tidal web page.. strange. BUT playlist from reddit is awesome, thank you!

 

[peterinvan]

Link starting tidal.com.. does not open playlist, just redirect to tidal web page.. strange. BUT playlist from reddit is awesome, thank you!

try starting the tidal app on your pc/mac, then click on above link and "Open in Tidal"

 

[dangusyra]

try starting the tidal app on your pc/mac, then click on above link and "Open in Tidal"

The link doesn't work directly from head-fi.com, but opens from a forwarded to email Thank you.

 

[justanut]

There is no such thing, to my knowledge the highest MQA masters are 352.8 KHz or something like that. Not that it is impossible, but there are none so far that I know of.


I don't believe that's correct. I don't think any MQA plays back higher than the master. If it does, then sure that is clearly a case of upsampling.
There is plenty of MQA at 44.1 and especially 48 KHz that doesn't expand at all when unfolded.


I have not seen anything unfold higher than 352.8. I just checked the MQA pages (not very easy to find facts!) and it says there any unfold the hardware supports is possible. On Tidal there I haven't seen any 16X and only a small amount of 8X. It is reasonable since there are some PCM 352.8 masterings floating around.


I think your latest post helped clarify what we don't agree on, which is that to my knowledge no MQA expands higher than the master, while you seem to be saying it does.

 

[rlw6534]

Yeah, I tried to explain that MQA rendering always upsamples to the capabilities of the DAC but as I said, MQA can be confusing with respect to sample rates. You have the source Master rate, the MQA file sample rate (always 44.1 or 48kHz), the core unfold (always 2x the file rate) and the rendering rate (4x, 8x, 16x). To complicate matters some DACs report the source (Master) rate with the MQA symbol and others report the actual final sample rate at the DAC (without a symbol).

 

[gimmeheadroom]

My Brooklyn DAC+ shows the actual sample rate and bit depth of the input stream and also the MQA unfolded sample rate and bit depth. I'll check the two songs @justanut showed but I know there is an issue of different MQA masters on Tidal based on region. We tried to compare some Metallica (I think) earlier and some guys didn't see MQA albums at all, and others had nothing but MQA. And I remember some DAP manufacturer on the site saying that the unfold sample rate shown was not necessarily correct for some reason.

@justanut would you please confirm what DAP that is and send me the Tidal links to the two songs? I'll post my update when I can. Thanks.