It’s also best to remember that the 818 and 808 [and soon 218] essentially have TWO different paths for MQA to follow…Analog and Digital
When using its own DAC and outputting over Analog outputs…then the 8x8 will act as what MQA call a “Full MQA Decoder”, as per the diagram below
When outputting a Digital Stream, then the 8x8 will act as what MQA term as an “MQA Core”…while the DSP SE’s will as an “MQA Renderer” again as per the diagram below
All sounds reasonable.
When I’m playing MQA to the DSPs, this may very well be happening: there is no visible indication if it, but it’s a possible architecture. I can see why M would prefer not to send 352 over the MHR wire, but the engineering purpose of splitting it is not clear, if the Speakers can do the upper range, why not all.
But in the case of pre-decoded content, I definitely do not get the second stage indicated on the speakers. If you see that, any suggestions for what is wrong? Firmware upgrade for the speakers?
By what means are you passing data from the Tidal app to the 818??..Is it via USB??
When the Tidal app passes the partially decoded 24/96 stream to the Explorer 2 or Brooklyn, then the Explorer / Brooklyn will detect the presence of the MQA metadata and will then fully decode the MQA to the maximum output possible on the device…i.e. in this case, the Explorer / Brooklyn are acting as an MQA Renderer as per the diagram above
Just to be specific…if you use the Tidal app with Exclusive Mode ON…but MQA Passthru OFF…and output from Tidal to the Explorer 2, then the Explorer will light all all three lights [one blue and 2 white]…indicating that it is playing a 24/192 MQA Track [obviously this applies to the approx 66% of Warner content that unfolds to 24/192]
And the Explorer does this in spite of the fact that it is only receiving a 24/96 stream from Tidal, as that it is all it’s capable of
So, in a likewise fashion, your 818 should also display that it is receiving MQA data…and then display the Rate [2x, 4x or 8x] of the data in that MQA Stream when Fully decoded / unfolded
Perhaps your 818 needs a firmware upgrade to avail of this functionality??
Yes, I understand how it is supposed to work.
But it doesn’t.
I’ll look at firmware.
Sending 24/192 data over Speakerlink AES / EBU Balance Digital was always going to be a challenge…never mind sending 24/352 8x data!!
Set against that, I can see why Meridian have leveraged the existing Speakerlink cabling structure and kept the data transfers between the 8x8 and DSP SE’s to 24/96
It also enables partially decoded MQA 24/96 streams to be sent to the non-SE DSP’s over SPDIF and Speakerlink cabling…thus enabling those users to benefit from the High Res content and Time domain Deblurring of MQA tracks…all while waiting for Santa to bring their DSP SE upgrades 
There’s a detail hypothesis paper to support deeper understanding on the technical aspects of MQA. A must read!
on my Brooklyn both files played at 24/352
PassThrough showed a blue light, meaning all decoding carried out by the DAC.
SoftDecode showed a red light, meaning DAC is carrying out 2nd/3rd unfold and render only.
Happy to test anything else you are after.
I think the idea of the test was to see if the second sample, already first stage decoded, could have some dsp applied to it and still be recognized as mqa and second/third stage unfolded by the dac. If you could apply some dsp to the second sample in Roon, eg some basic PEQ, does your dac still unfold it to 352??
with simple PEQ the DAC now only shows 32/88 and no sign of any MQA decoding.
Ok thanks for trying that. Disappointing result.
This suggests that MQA in band signaling in the decoded 88.2/96 kHz audio does not survive intermediate DSP. In order for any subsequent unfolding/optimization to occur, it may have to take place within a single component. Or a separate signaling channel, such as available via USB, may have to convey to downstream components that the audio is MQA.
AJ
Thanks very interesting read
It can’t. The MQA diagram mentions some sort of ‘side loaded’ DSP, which short of any further supporting info from MQA, we have to assume will need to be specifically written around MQA possibly via their API.
They’ve been pretty clear that just altering the bit-perfect data ad-hoc will make it non-MQA. DSP code will probably have to check-in with their internal integrity/security code.
The fact that al the data for this isn’t published makes me think they want to keep it under wraps outside of commercial interests, or they haven’t got all the pieces to the puzzle sorted out yet.
See here also https://community.roonlabs.com/t/mqa-and-room-correction/22634
Thanks for posting the Andreas Koch interview. He’s one of the best guys in the industry. The MPS-5 player can still hold its own against anything out there, regardless of price.
He is one of pioneer of DSD technology back in late 90’s that resulted in Super Audio CD. His products are a benchmark for others to follow. Highly respected audiophile.
Hey all - I deleted some posts in this thread yesterday while we discussed internally. We’ve talked and they’ve now been restored. As you were 
Any news on MQA in Roon? Thx.
Looks like the Tidal Desktop app is organizing MQA a little better than before. Nice to see a [M] for Master
I think that was a good clarification.
