High-res audio comparison: Linn Records Free High Res Samples

[SIY]

https://bostonaudiosociety.org/bas_speaker/abx_testing2.htm

I assume you are referring to this test where a total Linn system was used with the signal sent thru a Sony PCM-F1 between preamp and amp to see if Tiefenbrun could detect it in or out of circuit. 44.1khz at 16 bit. He scored 10 of 20 in the listening test.

Exactly. And the irony of the identification by Lipshitz is particularly delicious.

 

[Blumlein 88]

What would be the point of "music content above 22 KHz" that no one can hear?

Cheers

There is no point in that content really.

Now a sharp cutoff at 20 khz usually will monkey with the phase at least between 10khz and 20 khz. By having more bandwidth you can have the phase done without change in that 10-20 khz range. It wouldn't be what is up above 20 khz that matters, it is the effect that has on below 20 khz content. OTOH, it seems human hearing is rather oblivious to phase above 10 khz.

 

[AudioSceptic]

https://bostonaudiosociety.org/bas_speaker/abx_testing2.htm

I assume you are referring to this test where a total Linn system was used with the signal sent thru a Sony PCM-F1 between preamp and amp to see if Tiefenbrun could detect it in or out of circuit. 44.1khz at 16 bit. He scored 10 of 20 in the listening test.

Actually, it was Linn/Naim, because Linn didn't yet make their own amps. ;-)

 

[julian_hughes]

This series of videos is fantastic.

Surely there is no point at all to playing back anything above 22KHz? If your system is well designed the higher frequencies won't harm it, but if it's not well designed then it might?

Why would you want the amplifier or speakers to try to reproduce frequencies we can't hear?

Qobuz doesn't seem to have an option to "downgrade" to CD quality only. I think once Spotify starts their 16/44 service I will simply use that.

Qobuz lets you set your download and/or streaming quality to 16-bit 44.1kHz if you prefer.

 

[voodooless]

Now a sharp cutoff at 20 khz usually will monkey with the phase at least between 10khz and 20 khz. By having more bandwidth you can have the phase done without change in that 10-20 khz range. It wouldn't be what is up above 20 khz that matters, it is the effect that has on below 20 khz content. OTOH, it seems human hearing is rather oblivious to phase above 10 khz.

Most DAC reconstruction filters are phase linear, so also above 10 kHz, there is nothing going on phase wise.

 

[Beershaun]

I think I need a primer on hi res. How does 24bit 192khz information differ from 16/44 and below in the audible spectrum. Give the same passage of music and looking at the same bandwidth. Is the amount of information describing that passage the same or different? If so how does it impact the final analog voltage output and in turn sound produced by the speakers. Does higher sample rate equate to more fidelity of information in the audible spectrum?

 

[amirm]

I think I need a primer on hi res. How does 24bit 192khz information differ from 16/44 and below in the audible spectrum. Give the same passage of music and looking at the same bandwidth. Is the amount of information describing that passage the same or different? If so how does it impact the final analog voltage output and in turn sound produced by the speakers. Does higher sample rate equate to more fidelity of information in the audible spectrum?

The two dimensions are different in purpose. Using non-shaped dither, the dynamic range of 16 bit audio can be insufficient at high enough playback levels to have inaudible noise floor. Conversion from 24 bit (used in production of music) must be done properly as otherwise, it can create distortions of its own.

Similar issue occurs in the the conversion of high sample rate to 44.1 may have an audible effect. This is extremely subtle if it exists.

All of these are extremely minor to non-existent effects and hence all the arguments back and forth.

 

[amirm]

Repeating again, these are content reviews, not format reviews except for DSD. In other words, other content use the format differently and in much better way.

 

[Beershaun]

Repeating again, these are content reviews, not format reviews except for DSD. In other words, other content use the format differently and in much better way.

Understand you are evaluating the production quality of the file for sale, not the format. I think, understanding the fundamentals of the difference in the resolutions and what their promise is supposed to be, would help me and others take more away from your presentation. Should I just be looking at ultrasonic hygiene and where the musical information ends in the spectrum? Or is there more I can take away, such as the dithering used and whether or not it would produce artifacts as you mentioned above.

E.g. when you review DACs and amps you start with the specs and.claims promised by the manufacturer and then tell us if they love up to the hype. I am curious to learn what the equivalent of that promise and claim is and how to determine if a given hi res file lives up to that claim and promise.

 

[amirm]

The promise is that you are getting proper musical information above 20 kHz. That promise is broken in all the files I have analyzed and posted.

On dithering, unfortunately that process cannot be teased out of these tests.

 

[Billy Budapest]

No. And they deliver the next most common idiocy:

They cannot but have people that know this is total BS. The whole thing reads like the sort of ignorant idiocy one heard pedalled back in the 80's by people with no mathematics. Back then Linn were happy to push all manner of woo, mostly revolving around the magical properties of their turntable. It seems they haven't changed.

What happens to all the information in the gaps? Nothing, because there are no gaps.

These people are engineers. They have to understand how sampling works, how digital recording works, how D/A conversion works. It’s really shameful that they are propagating falsehoods to a public that largely doesn’t understand the physics of sound.

 

[respice finem]

That said, I've come across engineers believing in strange things. We are sometimes not quite as rational as we like to think of ourselves.
The strange part, why the audio "realm" is so prone to this.

 

[Beershaun]

That said, I've come across engineers believing in strange things. We are sometimes not quite as rational as we like to think of ourselves.
The strange part, why the audio "realm" is so prone to this.

Thats what Amir refers to as "Lay audio intuition." Not hard to understand how engineering minded people get mixed up.

For example, I'm trying to understand the benefit of 192khz sample rate. My "lay audio understanding" is that the higher sample rate is giving me more information in the audible band to the tune of 4x more samples of the same sample of music. Amirs most recent response is telling me that the advertised sample rate is not actually telling me the sample rate but rather the bandwidth being sampled. So I'm currently confused because my previous understanding.was different. And now I will have to unlearn what I thought I understood.

So.now my freaking question is. Why even sample above 25+30khz? It seems like it just adds noise and data bloat to the file and doesn't give any benefit.

 

[Blumlein 88]

Thats what Amir refers to as "Lay audio intuition." Not hard to understand how engineering minded people get mixed up.

For example, I'm trying to understand the benefit of 192khz sample rate. My "lay audio understanding" is that the higher sample rate is giving me more information in the audible band to the tune of 4x more samples of the same sample of music. Amirs most recent response is telling me that the advertised sample rate is not actually telling me the sample rate but rather the bandwidth being sampled. So I'm currently confused because my previous understanding.was different. And now I will have to unlearn what I thought I understood.

So.now my freaking question is. Why even sample above 25+30khz? It seems like it just adds noise and data bloat to the file and doesn't give any benefit.

Amir is telling you correctly. The accuracy of a reproduced 1 khz tone is no different in 24/48 khz than in 24/192 khz. The whole idea higher sample rate is high resolution is actually a very unfortunate idea born from thinking about other myths wrongly. High sample rates are only helpful for greater bandwidth. The extra in between samples don't help because even with fewer samples nothing is missed at lower frequencies. It comes from thinking info between samples is lost. It isn't. It is there and is reconstructed from sampled audio values.

 

[Beershaun]

The promise is that you are getting proper musical information above 20 kHz. That promise is broken in all the files I have analyzed and posted.

On dithering, unfortunately that process cannot be teased out of these tests.

Thanks. What is the potential benefits of the musical information above 20khz? Second question, what is the actual sampling rate of these music files if 192,44.1,96khz describe the bandwidth and not the sampling rate.

 

[amirm]

So.now my freaking question is. Why even sample above 25+30khz?

Some effects can create aliasing components if the sample rate is low (bug but it is what exists). So it is safer from editing point of view to sample higher than delivery. I suspect it also looks more "professional" to sample higher than CD in the eyes of the client.

 

[dc655321]

Thanks. What is the potential benefits of the musical information above 20khz? Second question, what is the actual sampling rate of these music files if 192,44.1,96khz describe the bandwidth and not the sampling rate.

44.1, 96 kHz does describe the sampling rate. The theoretical maximum bandwidth is 1/2 of the sample rate.

 

[Beershaun]

I see. I follow that you want some room to edit stuff out in the final editing process without sacrificing fidelity. Is their an optimal bandwidth to target in the final master that we should be seeking to purchase from media producers? Such that we are getting a simple copy of the final master and avoiding the conversion step to CD/Redbook spec?

 

[Beershaun]

44.1, 96 kHz does describe the sampling rate. The theoretical maximum bandwidth is 1/2 of the sample rate.

So it describes both the sample rate and the bandwidth recorded?
Im familiar with the Nyquist Shannon theorem. I want to better understand it's practical application. And whether or not there is demonstrable musical information we could potentially hear or feel.

 

[dc655321]

So it describes both the sample rate and the bandwidth recorded?
Im familiar with the Nyquist Shannon theorem. I want to better understand it's practical application. And whether or not there is demonstrable musical information we could potentially hear or feel.

Your questions suggest the opposite of familiarity. Please don't make me drag out the link to Monty's video