This is opaque to me… Qobuz has some recordings at greater than CD quality. HDTracks, same. How do they get these? And is their hi res different than me upsampling?
What bitrate and sample frequency do studios record and capture digital audio in? That seems to be the base limit of what we can acquire and what streaming services can deliver. Are Qobuz/HDTracks or anyone else receiving high bitrate/high sample rate files from the studio? Or are they upsampling CD quality data?
I’m pretty sure my DACs, internally will upsample as part of their process of converting from digital to analog. If that is the case, my guess is that happens with very close to the same quality of what may happen out of band. Or in Roon’s DSP. FWIW I have listened to CD quality Qobuz native through my Topping D90 and upsampled and can’t hear a difference. Admittedly, I have 56 year old ears.
So what does Hi Res audio really mean? And does it matter.
Agree. But what does Qobuz and HDTracks deliver? Something real? Or CD quality upsampled? I have listened to some HD vs CD versions of things I know. I do not hear a difference. Either I have bad ears or there really is no difference. If I have bad ears that may save me some money
Sorry, didn’t read your full reply before answering…. Perhaps my ears or system just can’t percieve a difference? I’m very confident in my two systems, less confident in my ears.
Unless I’m mistaken. If Qobuz offers an album as 24/192. That means it was taken from a master which was encoded to 24/192. It doesn’t offer an 24/192 album as an upsampled 16/44.1.
As far as I know, studios historically capture between 96kHz and 192kHz at 24 bits. (I guess some studios that master DSD also acquire DSD64/128/256, but post-process at a different resolution, e.g. 352.8kHz/24bit). That is done however to have a margin for post-processing, not necessarily to capture high frequency content - which depends mainly on the microphones and ADCs used. Most microphones don’t capture much above 20kHz, and ADCs may filters much of the signal above that anyway. It’s unlikely any hi-res audio has significant content above 20kHz. Even if it did, it would be inaudible (in the best case scenario).
For example, this is the spectrum of a 192kHz/24bit track by Neil Young, who is a strong promoter of hi-res:
As you can see, most of the music content is below 20kHz. At 20kHz, the signal is already well below -100dB. After that, it’s fairly flat (punctuated by 3 idle tones) and slowly dives below -150dB around 70kHz. There’s no music content there, only processing artifacts.
But we’re not listening in the digital domain.
The benefit of high-resolution, or upsampling/upscaling isn’t about frequency response as much as it’s about how we get the music in and out of digital. Ie. reconstruction filters, aliasing etc.
Edit: That wasn’t @ you Marian.
Everything you said is true.
Some recordings are mastered with equipment capable of capturing at DXD rates of 384kHz/32bit or DSD256, see i.e. MERGING+ANUBIS.
And there is musical content beyond the audible range, which can be verified with free software like MusicScope - one may debate, if that’s of any audible merit…
Then, there are simply upsampled redbook 44.1kHz/16bit releases, falsely marketed as high-res material - you won’t know, unless you test for yourself or find somebody else’s results.
I’ve got examples of both in my library…
On the other hand, upsampling pushes noise and other spurious artifacts, that were generated by the original encoding, farther away into the ultrasonic range, and the use of subsequent gentler digital filters improves passband ripple and impulse response.
Again, one may debate, if that’s of any audible merit…
In most cases, though, residential background noise, poor room acoustics, less resolving speakers and amplifiers swamp any slight benefit, that might have been made available…
Not exactly answering your question but I have found that the care taken in recording/mastering makes more difference to the sound quality than the sample rates etc. If they do take care recording and at higher rates it is usually plinky plonky boring music anyway.
I listen to music I like, the only thing that ruins it for me is when it has been compressed to sound loud and the dynamic range is almost non existent. Joy Crookes album springs to mind as being an example of something ruined by the loudness wars.
I do question the rates quoted by some sites as to their validity.
We record our gigs at 24/48… High res, and it’s far from plinky plonky (that made me laugh) I agree, listen to the music you enjoy and avoid the over compressed stuff…
Joke aside, mastering is no 1. I thin no 2 is the recording (devices/setup). There are some good piano concerts recorded in churches and the echo is annoying.
This is recorded at 96/24 (A2D is via an RME UFX+). I mix in the box and the mastering engineer also works in the box. So this has plenty of DSP processing - but it hasn’t be through a D/A or A/D step since recording.
You can see plenty of content > 20k - mostly cymbals.
