So can I…
We have to be careful @RBM , if we keep joking around we might end up with a Dutch version (and the work) (4 votes and counting…)
That’s four brains and eight hands – done in a whim!
I’d love to see some Flemish flourishes by @stevev1 
Nice alliteration
We do seem to have a lot of Dutch members, must be the famed plain speaking common sense at work.
I strongly support a Witchhunt to find the most obstructive Alpha tester. For too long these anonymous figures have influenced events from the shadows. They must be dragged out of their basements into the sunlight, blinking and weeping.
7 Likes
Absolutely, the developers and team managers definitely deserve a lot o credit.
Maybe if you started by taking that apple off of your face it would be easier.
my wish for 1.3 release - as Tidal now streams MQA it would be great to have software MQA decoder available in the Roon
4 Likes
It’s been confirmed to be coming in the near future. Probably not in the initial 1.3 release but shortly after in maybe a subsequent update build.
1 Like
Linked to the request for MQA software decoding, can I please also request a set of “filters”, for the MQA decoding to take account of my downstream DAC.
Right now, I have an OA13 card in my Meridian 861, and will be buying an 818 at some point (unlikely to be the v3).
What sort of filters are you referring to ? MQA has some time correction DSP but that might only be implemented in hardware; we don’t know yet.
My understanding of MQA is that it:
- Applies a filter to cater for the ADC used during the first digitisation
- Applies a filter to cater for the DAC that will be used during translation to analogue
- “Folds” the music file to compress it
I believe that the Tidal encoded files will include by the ADC filter and the file “folding”, but not the 2nd filter, as it’s currently unaware of the DAC. So software decoding will only take account of the first filter and and the unfolding.
As such, I’m referring to the 2nd filter. I’m assuming that this could be created to cater for known DACs, and being a Meridian user, I’d assume that Roon would be able to easily access that information.
Correct. The 2nd stage only happens if you decode in an MQA DAC, ie you use passthrough rather than the decoding engine in Roon. I very much doubt that there will be any such profiles available to the software decoder.
Question about MQA decoding in 1.3 (and really any transcoding): Where does it take place? In the core or in the endpoint?
That is: if I send an MQA file to a RoonReady renderer (eg microRendu), is the microRendu running the decoding or is the data decoded to PCM prior to routing it over RAAT? Specifically, I have a mini running the core, controlling with an iPad, sending to a microRendu.
What happens in the case of transcoding by the way? Eg when I have a DAC that does not support DSD and I play a DSD64 file. Where does the conversion happen?
Thx.
Core. And core.
1 Like
I would think the ADC used might not always be known? Might be different for different instruments etc etc.
Right. Or there are many ADCs used. Or there are no ADCs because the music is created digitally in the first place. Or more commonly, the mastering stage gets a digital file, converts it to analog where mastering is actually done (I understand most mastering engineers do this) then back to digital - so only the very last ADC artifacts can be corrected. The claim about ADC corrections is full of hair and I find it disingenuous.
If you read this you will see that Bob Stuart fully understands the mastering process and that there a mix of ADCs etc and a final mix. He indicates that this final mix, as you suggest, is where most of the correction is and can be made so I don’t see you you can find this disingenuous.
http://www.computeraudiophile.com/content/694-comprehensive-q-mqa-s-bob-stuart/
Q&A 16 and 18 are particularly relevant.
In Q&A 19 he estimates that the provenance of about 70% of albums is well known.
1 Like
The diagrams I have seen depict a microphone, an ADC, folding, unfolding, and final DAC corrections. My point is simply that the conveyed idea that there’s one ADC in the process that gets corrected is very limited to very specific cases - it could apply to transfers from old analog masters directly to the digital file. But most modern recordings have many more steps.