MQA Deep Dive - I published tracks on Tidal to test MQA

[BS5711]

I don't think anybody gets you.

 

[Currawong]

And, why do you think we have tin ears

And we circle back to “I can’t help it if your hearing isn’t good enough”.

Boy it gets so old hearing that all over Head Fi, it seems all the people with hearing beyond scientific understanding are audiophiles, why is that do you think ?

You guys with this level of hearing ability need to get together and help science redefine the current data in respect of the limits of human hearing and show us tin eared losers that you were right after all.

Someone already has. Actual research has explained what we have known all along:

From toddlers to teens: auditory development and the impact of musical training

After discussing her background in music production and neuroscience, Susan went on to discuss how our brains and bodies process and respond to sound, explaining how signals travel from our eardrums to our auditory cortex. She explained how our brains have evolved to prioritise certain types of sound such as speech and how we are able to effectively filter out sounds that we don’t want to hear so we can focus on those that we do (in other words, boosting frequencies within a certain range, while suppressing others that fall outside of this).

She then explained how our hearing changes and develops throughout our lifetime. Young children, for example, have less refined hearing than teenagers and young adults – “they don’t have good high frequency resolution” – and our hearing continues to improve throughout childhood, peaking around college age.

Our auditory system also continues developing past toddlerhood, but is fully formed for most people by the time we hit adolescence.

“By age 12, it has to stop because your body is getting ready for puberty,” said Susan. “This means that unless we’re taking music lessons, our auditory processing circuitry, all of our nuclei, our auditory nerve bundle and our auditory cortex is done [developing] by age 12.”

Analytic vs synthetic: How musicians hear differently to non-musicians

Whilst most people’s auditory systems finish developing before we hit puberty, this is not the case for those who receive musical training.

As Susan explained to CanJam attendees, the auditory processing path of trained musicians continues to evolve, bringing physical changes that affect our music processing capabilities. “The nuclei get fatter and thicker, and the auditory nerves, your wiring, grow what’s called dendritic spines – more little branches.” These developments allow musicians to become better and faster at processing the subtle differences between sounds.

This, in turn, enables musicians to listen in a different way to non-musicians. Citing a test she used to conduct with students at Berklee (which you can try out online here), Susan explained how musicians are capable of listening analytically – meaning they are able to hear different frequency components individually. Non-musicians, however, can only listen synthetically – meaning they focus on the ‘global whole’ or complete sum of parts.

Through her work at Berklee, Susan found that whilst she was unable to process certain subtleties in a musical work in the same way as students who could listen analytically – lacking the neural infrastructure required to do so – this allowed her to assess the complete sound of a recording or musical work without getting “bogged down” by focusing on minute details.

As for the WAV vs FLAC issue, an ideal, actual scientific approach would be to take detailed measurements of a system to find out if there is any possibility of file decompression influencing some part of the output of that specific system, such as noise, and whether that noise is influencing the system in some way. However, there is more of an interest from self-styled "objectivists" in just putting people down, than actually doing something that resembles research.

Of course, the most sensible thing to do would be if there's a difference, buying a good streamer usually fixes that in my experience.

 

[Audiophiliac]

Someone already has. Actual research has explained what we have known all along:



As for the WAV vs FLAC issue, an ideal, actual scientific approach would be to take detailed measurements of a system to find out if there is any possibility of file decompression influencing some part of the output of that specific system, such as noise, and whether that noise is influencing the system in some way. However, there is more of an interest from self-styled "objectivists" in just putting people down, than actually doing something that resembles research.

Of course, the most sensible thing to do would be if there's a difference, buying a good streamer usually fixes that in my experience.

Salutations, Currawong, and thank you for the many reviews of AGD gear that lead me to my R8 DAC purchase (spent more than I was planning to get this, but thank goodness). Have you not uncompressed your music collection onto a nice quiet USB SSD yet? Drive sizes and prices for a big uncompressed personal collection on USB ssd is here, and any need to have to tell people that FLAC is unnoticeable is gone.
Let me ask this: If you had 20 billion dollars, and your dream audio system was only 1m + 700k for the speakers you liked best, what would you think of playing your music collection uncompressed from a 4TB usb ssd that can read ahead at 900mb/s during track jumping? The future is now, and it's sickly cheap.

 

[Currawong]

Have you not uncompressed your music collection onto a nice quiet USB SSD yet?

No need. I just use a streamer. Then, the source computer set-up becomes mostly irrelevant. Finding good-quality recordings is more of a priority these days then chasing endless tweaks. Unfortunately, a lot of the music I like isn't well mastered enough to be totally enjoyable on a resolving system.

 

[BS5711]

Someone already has. Actual research has explained what we have known all along:



As for the WAV vs FLAC issue, an ideal, actual scientific approach would be to take detailed measurements of a system to find out if there is any possibility of file decompression influencing some part of the output of that specific system, such as noise, and whether that noise is influencing the system in some way. However, there is more of an interest from self-styled "objectivists" in just putting people down, than actually doing something that resembles research.

Of course, the most sensible thing to do would be if there's a difference, buying a good streamer usually fixes that in my experience.

Thank you, that is very interesting.

I don't for a moment, however, believe the large number of folks that talk here and on every social media audio group about hearing differences in pretty much everything audio are all trained listeners or musicians, I think that would be a fairly safe assumption.

Additionally, despite that some highly trained folks might be able to dissect sounds better there still needs to be sound to dissect. If there is demonstrably no technical difference and yet a person claims there is that isn't refined hearing that is imagination.

I can certainly understand that training refines the perceptive abilities but it seems less likely that it completely rewrites human hearing as it would need to for many such folks to be correct.

Edited to add: Or maybe I am just stupid and have tin ears as our man here asserted and I am blind (deaf) the very high level that other folks operate, I guess I will never know

 

[GoldenSound]

I'm done arguing. It's not my problem that you can't hear less noise when wav, and I hope you surprisingly find it offensive that I do, I guess.

Generally speaking, if you make a claim that is unsubstantiated, it's on you to test/prove it.

I said that I could hear a difference between upsampling filters for example, others said that shouldn't be the case, so I went out of my way to show that it was the case:


Getting angry when people don't believe your claim ESPECIALLY when its not even possible to actually demonstrate it with any measurements (at least with the upsampling filter we could show that they were different, the debate was just whether that difference was audible), is not going to solve anything and this "Well I can't help if your ears are bad" argument is the same one used by people claiming their $3000 USB cable or purple fuses are making a difference

 

[GoldenSound]

Someone already has. Actual research has explained what we have known all along:



As for the WAV vs FLAC issue, an ideal, actual scientific approach would be to take detailed measurements of a system to find out if there is any possibility of file decompression influencing some part of the output of that specific system, such as noise, and whether that noise is influencing the system in some way. However, there is more of an interest from self-styled "objectivists" in just putting people down, than actually doing something that resembles research.

Of course, the most sensible thing to do would be if there's a difference, buying a good streamer usually fixes that in my experience.

FWIW I have actually looked into this myself and tried to measure any difference from my main PC, my laptop, or even a raspberry pi when playing FLAC vs WAV and was completely unable to find any difference in output noise whatsoever.
On the pi the background tasks that were running were more of a factor, so I guess this guy needs to look into running an ultra stripped down linux or something if we follow the logic to its conclusion.

Though regardless, the cleanest output was achieved just by putting an Intona 7055-C in the path even if the desktop PC it was connected to was running Prime95 and Furmark benchmark software simultaneously to stress the CPU and GPU

 

[Audiophiliac]

No need. I just use a streamer. Then, the source computer set-up becomes mostly irrelevant. Finding good-quality recordings is more of a priority these days then chasing endless tweaks. Unfortunately, a lot of the music I like isn't well mastered enough to be totally enjoyable on a resolving system.

But how could anything improve if your playback reference is uncompressed files on a usb ssd in your streamer?
I am the other way around, it's just too bad if a favorite track is poorly recorded. I still buy all the best gear I can afford, because it always works out for average sounding recordings also, anyhow. Higher res speaker demo playlists are usually well recorded embarrassing selections. I don't hunt for tracks that sound outstanding, but if a favorite track is a relatively recent recording, I melt in awe of my sound system. At least many releases are finally showing up as at least 96khz, but many of the still 44.1 have still chosen the 24bit toggle, to max out our gear's volume sensitivity tolerance.
Back to topic, have you tried your favorite recordings uncompressed on a usb ssd in your streamer, compared to anything else? I just bought "The Essential The Chicks" to better support them financially, but also to get the 96khz wav copies. As usual with local uncompressed playback, it is so free of noise, that I can fall asleep too easily because of how relaxing my playback makes me feel. Do you have any arguments against uncompressed reads from an ssd being the reference reads that all other reads have to try to prove they are identical to, for some reason?
When people can buy 1 Petabyte usb ssd's for $200, then would you consider trying uncompressed reads, so that you don't have to sit there listening to your FLAC's for crazy noise, instead of all sounds being noisy renditions all of what was there originally? Lol, yeah, FLAC is a punk rock cover of the Bach wav, now that I've found the difference to listen for.

 

[GoldenSound]

When people can buy 1 Petabyte usb ssd's for $200

These are scams

 

[Audiophiliac]

FWIW I have actually looked into this myself and tried to measure any difference from my main PC, my laptop, or even a raspberry pi when playing FLAC vs WAV and was completely unable to find any difference in output noise whatsoever.
On the pi the background tasks that were running were more of a factor, so I guess this guy needs to look into running an ultra stripped down linux or something if we follow the logic to its conclusion.

Though regardless, the cleanest output was achieved just by putting an Intona 7055-C in the path even if the desktop PC it was connected to was running Prime95 and Furmark benchmark software simultaneously to stress the CPU and GPU

If isolated audio playback was my only requirement, I would probably prefer someone's streamer over stripped down linux on a pi.
Alas, I crave the ability to do everything from a single source, movies and video games with music in the background, so it's PC for me, and because of the games, a windows PC with a huge graphics card, which only sucks tons of power during gaming. It sucks that movies look the best on my pro 4k screen, but some video games have an amusing style, like WoW. Album cover art on my 32" pro 4k screen now seems to beat how good the 12" covers looked, and can display at 15.5" tall now. It will probably take 8k to get the last little bit of wanting higher res out of me.
I believe the Intona 7055-C would improve my sound, but I believe it will be beaten on my system by purchasing the AGD DI-24, which would also give me a dedicated I2S output, to beat my current noisy usb I2S. It also has an external clock port...

 

[Audiophiliac]

These are scams

I didn't say people can buy them now, I said "in future".
What if someone says you're full of it, up sampling filters sound no different than my original 4-track copy of Neil Diamond?

 

[GoldenSound]

Alas, I crave the ability to do everything from a single source, movies and video games with music in the background, so it's PC for me, and because of the games, a windows PC with a huge graphics card, which only sucks tons of power during gaming. It sucks that movies look the best on my pro 4k screen, but some video games have an amusing style, like WoW. Album cover art on my 32" pro 4k screen now seems to beat how good the 12" covers looked, and can display at 15.5" tall now. It will probably take 8k to get the last little bit of wanting higher res out of me.

If you're running a windows desktop pc particularly one with a gpu, then the random activity is going to be more of a factor by several orders of magnitude.....

I believe the Intona 7055-C would improve my sound, but I believe it will be beaten on my system by purchasing the AGD DI-24, which would also give me a dedicated I2S output, to beat my current noisy usb I2S.

I've tested and measured the di20he before. It's a good DDC, though i2s vs usb is entirely dependent on the DAC.

Purely in terms of noise intona is quieter but Jitter performance is situational as i2s is synchronous whereas usb is asynchronous.

Additionally, the di20he has full galvanic isolation internally anyway, so you won't have any noise (including any caused by FLAC decoding) passed through in the first place.

It also has an external clock port...

Unfortunately only 10Mhz, and 10Mhz clocks actually make things worse not better. But they're expensive so people still buy them.

What if someone says you're full of it, up sampling filters sound no different than my original 4-track copy of Neil Diamond?

That was the entire point of the video. Oversampling/upsampling filters do make a difference. And you are using one already unless you're using a NOS DAC, and if you are playing NOS then you're actually incurring additional distortion anyway

 

[gregorio]

As for the WAV vs FLAC issue, an ideal, actual scientific approach would be to take detailed measurements of a system to find out if there is any possibility of file decompression influencing some part of the output of that specific system, such as noise, and whether that noise is influencing the system in some way. However, there is more of an interest from self-styled "objectivists" in just putting people down, than actually doing something that resembles research.

No, that is not the “actual scientific approach”! The actual scientific approach would be to take into account the basic obvious facts that we cannot hear digital audio signals or even analogue audio signals, we can only actually hear acoustic sound. Therefore, it does not matter in the slightest if “some part of the output of that specific system, such as noise” is measurable, what actually matters is firstly whether that noise (or interference/distortion) actually makes it’s way into the acoustic sound and (only) if so then secondly, is it of sufficient magnitude to be audible? For example, we can have a digital audio signal with relatively huge amounts of noise/interference and yet it can make no difference whatsoever to the acoustic sound produced and even no difference to the intermediary analogue signal (so the second question of audibility is moot). That of course is the whole point of why digital audio was invented in the first place.

Also, I personally don’t know any “self-styled objectivists”. Everyone I know who would be labelled an objectivist by audiophiles are at least somewhat uncomfortable with the term and would generally not “self-style” themselves as “objectivists”, except maybe in certain contexts. And lastly, what makes you think we have not done “something that resembles research”? Certainly I have very carefully objectively measured numerous aspects of audio performance under controlled conditions over the course of more than two decades, including both lossy and lossless codecs.

I can certainly understand that training refines the perceptive abilities but it seems less likely that it completely rewrites human hearing as it would need to for many such folks to be correct.

Training does not affect hearing ability and it does not attempt to, what is actually being trained is the ability to consciously dissect what is being heard. Formal music training involves the obvious, such as identifying the individual instruments within an ensemble, the notes being played and the rhythm, as well as the less obvious, such as variations in note “envelopes” (the attack, decay, etc.), fine intonation, etc. For this reason, the training is typically called “Listening Skills” rather than anything to do with hearing. Incidentally, this means that musicians with significant formal listening skills training will be able to identify certain aspects of music performance that untrained listeners will not consciously be aware of, but they will not necessarily be any better than untrained listeners at identifying other aspects of audio (not related to music performance), various artefacts or audio processing parameters for example, which require different “listening skills” training. And lastly, listening skills improve with experience/age, while hearing does the opposite and deteriorates with age. This is the reason why highly experienced audio engineers are generally better at identifying certain fine details than far younger students, despite their hearing abilities being significantly worse.

I just bought "The Essential The Chicks" to better support them financially, but also to get the 96khz wav copies. As usual with local uncompressed playback, it is so free of noise, that I can fall asleep too easily because of how relaxing my playback makes me feel.

Firstly, it is not “so free from noise”! Unless it was recorded in a world class anechoic chamber (which it wasn’t) and with microphones and mic pre-amps that break the laws of physics (which it wasn’t) then it cannot be free from noise. Admittedly, the recordings of theirs I’ve heard are relatively “very clean”, very obviously so compared with far older analogue recordings, and therefore it is not easy to identify the noise but it’s certainly there. Who has the “tin ears” now?! This noise (and everything else in the audio) is perfectly maintained (neither increased, nor decreased) by FLAC. And incidentally, this noise is within the limitations of even 16bit and is not affected by the sample rate, so 96kHz or being 24bit also makes no audible difference.

Do you have any arguments against uncompressed reads from an ssd being the reference reads that all other reads have to try to prove they are identical to, for some reason?

I have no arguments against that, the same as I have no arguments against a losslessly compressed version that reads from a HDD being the reference, because they are identical. And why on earth would we “have to try to prove they are identical” when it has already been proved decades ago, before FLAC was even released?

G

 

[Audiophiliac]

With a normal uncompressed file, it reads into ram, and gets sent to output as the clock demands. With compression, your noisy cpu plays the track into the same ram result. I'm listening to FLAC streaming now, it's not that bad after I say it is, wait, no, it is so all noise versions of the same sounds again if I say it's not bad, FLAC is awful!

 

[gregorio]

With a normal uncompressed file, it reads into ram, and gets sent to output as the clock demands. With compression, your noisy cpu plays the track into the same ram result.

No! Firstly, with a normal uncompressed file, it reads into ram, then gets routed and converted into whatever transport protocol you’re using (USB, SPDIF or whatever) through driver software by a noisy processor and then gets output as demanded by that protocol (not necessarily the clock speed of the audio). With FLAC you also have a noisy processor involved, so both compressed and uncompressed files involve a noisy processor, what’s the difference? Secondly, even if there were some difference in the amount of noise transmitted with the bitstream/data packets, what difference do you think that makes to your speaker/HP output? Digital audio is NOT analogue audio. Unlike analogue audio, noise/interference in the digital signal cannot be represented in the digital data and therefore ceases to exist, which was the whole point of why digital audio was invented. Digital audio only has two states, a one or a zero, that’s it there is nothing else, so there cannot be a noisy one or a noisy zero! The only *potential* way for noise to get through would be a particularly badly designed DAC, which allows the noise on the digital input connection to somehow find it’s way to the analogue side of the DAC and the worst example of this I know still does not allow enough noise contamination to be audible at any reasonable playback level.

If there really is noise so loud when using FLAC that you can actually hear it above the noise floor of your recordings (which you apparently can’t hear) then your computer, your DAC or both are very seriously defective!

G