It’s not market drivel. It’s science. I know it’s hard for them to explain, and even harder for normal people to understand.
But if you look closer, MQA’s genius lies in its simplicity.
The power spectrum of an audio signal is relatively easy to compute. To avoid contamination of the output spectrum by edge-effects resulting from the finite signal transformation length, it is essential to apply a window function to the raw data. By down-weighting the points close to the edge of each transformation frame, unwanted harmonic distortion can be minimised. Add in the quantitized time domain correlation and Bob’s your uncle.
It’s really not that complicated. All you have to do is trust your ears.
“lossy” and “lossless” have agreed-upon meanings, just like, say, the word “blue”.
why would Mr Stuart expect from laypeople, which is clearly the crowd that this blog post addresses, that they have a PhD in Stuart drivel, and thus accept that when he says that the sky is red, he in reality has meant that it is blue for the past 30 years ?
Because he mistakenly assumes that the average person might have something other than sawdust in their crania?? And when an explanation is presented - over and over again, and is not hard to understand - that they might actually grasp it?
Yes, but if what you see as blue I see as what you would call red, and what I see as green you would see as blue, then we are always seeing the same thing but we are seeing it differently and don’t even know it. It’s the same with hearing. Everybody’s hearing is different, but we all hear the same thing in our own way. MQA corrects for that so we all hear the same thing in the same way. People who hear differently may be shocked at what they hear in MQA, others may say" see, this is what I’ve been trying to tell you all along!"
I think the only part I’m now looking forward to, after referring to the test conclusions as “a libellous manifesto”, is the part where absolutely no legal action is taken for fear of having the whole thing publicly get the legal and scientific scrutiny it so richly deserves.
It’s clear Goldensound was lying for personal gain. His result are not valid and the response from MQA shows why.
Not that you would take the time to read, it’s so much easier to just stay on the bandwagon I guess.
Seems you still don’t understand what they mean. If you assume the PCM is the truth then anything different is lossy. MQA improves on the PCM and will by that definition be lossy.
The issue is of course that the PCM wasn’t perfect to begin with …
Based on responses I have seen in Facebook forums it seems they are considering what more to do.
It’s not easy to respond to this and actually get people to listen.
Goldensound made a lot of damage and it’s clear he did leave out a lot of information (or flat out lied).
What I think of him is not fit for writing in this forum.
That’s the whole point.
There are a number of tools MQA uses and they’re not all valid for all scenarios but for the simple case of a 44.1/16 file that’s the point.
If your definition of lossless is maintaining all the bits in a mastered file any file that has less bits than the mastering format is lossy. Since most mastering these days in done in DXD most distributed audio files are lossless to some extent.
So it can be a valid consideration that appropriate corrections are just a best guess if it is working on the already encoded PCM and not the “mixing desk” scenario. No wonder half the batch converted stuff sounds so ordinary, the other half got lucky that they churned-out something which is not subjectively worse than the original PCM.
The first thing we learned in compute theory is “garbage in, garbage out”.
This could seem off topic, but much as I value the MQA ecosystem, when it comes to provenance, I would have hoped that Stuart would have used the Shakespeare neologism ‘all that glisters’ from ‘The Merchant of Venice’ rather than Dryden’s ‘glitters’.
I guess legal action could be anticipated by owners of sufficiently resolving systems who have fried their tweeters because they were rash enough to try to actually play one of these ‘test’ files…
