[StanD]

Let's approach thins from a perspective of visualization. Imagine a complex waveform of only a single instrument playing a single note. This would not be a simple sine wave but have all sorts of squiggles deviating from the fundamental tone frequency sine wave due to the various harmonics and their continual phase shifts. What do you think will happen due to the truncated 8 bit quantization? Intuitively I would suspect that much of this would be lost due to the lack of resolution made worse by truncation. What say you gents?

 

[pinnahertz]

  Let's approach thins from a perspective of visualization. Imagine a complex waveform of only a single instrument playing a single note. This would not be a simple sine wave but have all sorts of squiggles deviating from the fundamental tone frequency sine wave due to the various harmonics and their continual phase shifts. What do you think will happen due to the truncated 8 bit quantization? Intuitively I would suspect that much of this would be lost due to the lack of resolution made worse by truncation. What say you gents?

If you were to look at that waveform on something like an oscilloscope, it's unlikely you'd see any difference at all  unless you were to look at only a small portion of the waveform. There just not enough resolution in a scope display if you display a full cycle of a 0dBFS signal. 
 
The visualization part of this doesn't work well. 

 

[gregorio]

  [1] What do you think will happen due to the truncated 8 bit quantization?
[2] Intuitively I would suspect that much of this would be lost due to the lack of resolution made worse by truncation.

 
1. I'm not sure what you'd see on a scope. In a spectogram I'd expect to see a peak representing the fundamental, another lower peak for the first harmonic, a lower peak for the second harmonic and then dozens of other lower peaks again, some below the fundamental, which represents the truncation distortion. Among these dozens of peaks will be the higher harmonics of the instrument but how many you'd be able to visually identify by their amplitude would depend.
 
2. The lack of resolution itself wouldn't be a problem, you wouldn't loose anything that hadn't already been lost in the original. But how many harmonics would actually be identifiable would depend on where in the frequency spectrum those harmonics are, their amplitudes, the amount of dither and the amount and distribution of any noise-shaping of the dither. The amount lost due to lack of resolution could, under the right conditions, be very little or even none. So it could be that's it's not a case of loss "made worse by truncation" but a case of pretty much the only loss being due to truncation error, although reducing down to just 8 bits, the "right conditions" for this to be the case would be fairly unlikely.
 
G

 

[StanD]

Perhaps it's time for something you can't mention in too many places in head-fi, carefully constructed ABX listening tests. Will a stradivarius still sound the same?

 

[old tech]

  Perhaps it's time for something you can't mention in too many places in head-fi, carefully constructed ABX listening tests. Will a stradivarius still sound the same?

https://en.wikipedia.org/wiki/Player_preferences_among_new_and_old_violins
 
Perhaps use a modern violin.

 

[nvfan]

Isn't the highest resolution technically something like 22bit, and 24bit is virtually impossible with modern technology?

 

[old tech]

  Isn't the highest resolution technically something like 22bit, and 24bit is virtually impossible with modern technology?

It's not just the possibility but also pointless.  24bit resolution is below the noise level generated by resistors and other electronics.

 

[StanD]

  Perhaps it's time for something you can't mention in too many places in head-fi, carefully constructed ABX listening tests. Will a stradivarius still sound the same?

 
 

  https://en.wikipedia.org/wiki/Player_preferences_among_new_and_old_violins
 
Perhaps use a modern violin.

Modern is good, I just used Stradivarius in a humorous context.

 

[Intensecure]

Modern is good, I just used Stradivarius in a humorous context.

Actually, there is an interesting study done blind testing musicians and Stradivarius violins..Not quite what you were inferring but still quite amusing.. We are all fallible..
http://www.thestrad.com/blind-tested-soloists-unable-to-tell-stradivarius-violins-from-modern-instruments/

 

[StanD]

Actually, there is an interesting study done blind testing musicians and Stradivarius violins..Not quite what you were inferring but still quite amusing..

We are all fallible..
http://www.thestrad.com/blind-tested-soloists-unable-to-tell-stradivarius-violins-from-modern-instruments/

I'd like to see a proper blind test of a $99 Schiit Modi 2 DAC vs. some megabuck DACs that have a religious following. Even 24/32 bit on the most worshipped DACs vs 16 bit on the Schiit. I suspect may of the purists will require medication once proven that they can't tell a difference.

 

[pinnahertz]

  It's not just the possibility but also pointless.  24bit resolution is below the noise level generated by resistors and other electronics.

Only at the A/D or D/A, in other words, where analog and digital transition to or from each other.  Internal data structures can and do go to 64 bit floating point.  Adobe Audition does everything at 32 bit FP.
 
The best A/D I'm aware of has a noise floor down at about 22 bits (German company "Stage Tec").  They cascade more than one A/D to get that to happen and convert at 32 bits, then use DSP to change gain and spit out 24 bit words. But you'd pretty much classify their stuff as a bit exotic.

 

[castleofargh]

  Isn't the highest resolution technically something like 22bit, and 24bit is virtually impossible with modern technology?

and that's just looking at the DAC. for the entire recording/playback list of processes and gears, even 16bit final resolution to my ears is highly optimistic. I mean when do we get a headphone that has flat FR and no distortions at -90db? and how often do we listen to louder than 110db singers performing in anechoic chambers?

 

[pinnahertz]

  I'd like to see a proper blind test of a $99 Schiit Modi 2 DAC vs. some megabuck DACs that have a religious following. Even 24/32 bit on the most worshipped DACs vs 16 bit on the Schiit. I suspect may of the purists will require medication once proven that they can't tell a difference.

It's a dead end. The purists don't acknowledge ABX as a valid test protocol. 

 

[castleofargh]

 

  I'd like to see a proper blind test of a $99 Schiit Modi 2 DAC vs. some megabuck DACs that have a religious following. Even 24/32 bit on the most worshipped DACs vs 16 bit on the Schiit. I suspect may of the purists will require medication once proven that they can't tell a difference.

It's a dead end. The purists don't acknowledge ABX as a valid test protocol. 

true, also how hard is it to get a colored DAC and then claim it is the only transparent one because I like more songs on it? ^_^

 

[TheoS53]

Found this on Wikipedia:

24-bit digital audio has a theoretical maximum SNR of 144 dB, compared to 96 dB for 16-bit; however, as of 2007 digital audio converter technology is limited to a SNR of about 123 dB^[12][13][14] (21-bit ENOB) because of real-world limitations in integrated circuit design. Still, this approximately matches the performance of the human auditory system.^[15][16] (While 32-bit converters exist, they are purely for marketing purposes and provide no practical benefit over 24-bit converters; the extra bits are either zero or encode only noise.)^[17][18]