[old tech]

It would certainly get rid of subjectivism...

But also customers.

Edit: just noticed - my Headfi birthday!

Also my actual birthday!

And yes, I know no one wil believe me given the date, but I'm used to that in Sound Science.

Happy birthday!

 

[Whazzzup]

My car needs it

 

[taffy2207]

you're one to talk, Metal is famously evil!!!!
or was it rock?
no wait... I know! it's punk. no it's that gothic stuff and new wave. no it's Jazz, or maybe ... well basically, music is evil. we should entirely ban it, end of story.

Indeed it is. You'll find the Devil in the 'details'

 

[WhatToChoose]

I see the term "detail" thrown around a lot when describing sound, like "DAC A sounds more detailed than DAC B".

I reason if, say 20 kHz is as high as one can hear, and the system can play past 20 kHz, it is able to play the most detailed sound possible, because this sound contains the theoretical maximum information per time that is audible.

Is this thinking wrong? What am I missing? It seems like almost every review and comparison on here throws around that word.

I'd say detail is how closely the audio reproduction system produces perfect sinusoidal waves, which are directly correlated with audio tones. Infinitesimally accurate sine waves are synonymous with more details, where as jagged/rough sine waves are not as detailed.

 

[PointyFox]

I'd say detail is how closely the audio reproduction system produces perfect sinusoidal waves, which are directly correlated with audio tones. Infinitesimally accurate sine waves are synonymous with more details, where as jagged/rough sine waves are not as detailed.

Interestingly, I've always used sine waves for a benchmark, but I think it's harder to produce square, triangle, and sawtooth waves. I've noticed a huge difference between the sound of these on different equipment, but little to none using sine waves besides loudness.

 

[castleofargh]

Interestingly, I've always used sine waves for a benchmark, but I think it's harder to produce square, triangle, and sawtooth waves. I've noticed a huge difference between the sound of these on different equipment, but little to none using sine waves besides loudness.

square waves(or any signal with straight lines involved), contain an infinite amount of frequencies. so obviously the chances for gears to show some amount of difference is drastically multiplied compared to using a single tone. I'm not sure how that relates to the perception of details though.

 

[WhatToChoose]

Interestingly, I've always used sine waves for a benchmark, but I think it's harder to produce square, triangle, and sawtooth waves. I've noticed a huge difference between the sound of these on different equipment, but little to none using sine waves besides loudness.

Ah, the square and triangle waves most likely cause different sounds due to the high frequency component of the transition edge of the waves. What I was referring to was the rough sine wave caused by the discretization process. Essentially, the sine wave looks more like a staircase, and from my understanding, the goal is the get those stairs infinitesimally small (e.g. 4 stairs per unit distance vs. 10,000 stairs for the same distance) to simulate the perfect sine wave, which takes more expensive equipment that can handle these calculations

 

[PointyFox]

Ah, the square and triangle waves most likely cause different sounds due to the high frequency component of the transition edge of the waves. What I was referring to was the rough sine wave caused by the discretization process. Essentially, the sine wave looks more like a staircase, and from my understanding, the goal is the get those stairs infinitesimally small (e.g. 4 stairs per unit distance vs. 10,000 stairs for the same distance) to simulate the perfect sine wave, which takes more expensive equipment that can handle these calculations

Are headphone drivers even capable of moving in a "staircase" fashion? I'd think they'd have too much inertia.

 

[castleofargh]

Ah, the square and triangle waves most likely cause different sounds due to the high frequency component of the transition edge of the waves. What I was referring to was the rough sine wave caused by the discretization process. Essentially, the sine wave looks more like a staircase, and from my understanding, the goal is the get those stairs infinitesimally small (e.g. 4 stairs per unit distance vs. 10,000 stairs for the same distance) to simulate the perfect sine wave, which takes more expensive equipment that can handle these calculations

that doesn't happen. there are imperfections compared to the original signal caused by noise and distortions occurring along the playback chain, but the staircase thingy is a myth because any half competently built DAC will have a reconstruction filter removing high frequency content that could still remain from quantization(or most likely nowadays, the spiky noise from delta sigma pulses). and because we recorded a band limited signal, we know that all those frequencies above should be removed as they have nothing to do with the original music. what remains should be pretty clean even on measurements(depends on the magnitude you're looking at and the quality of that filter).

 

[WhatToChoose]

Are headphone drivers even capable of moving in a "staircase" fashion? I'd think they'd have too much inertia.

that doesn't happen. there are imperfections compared to the original signal caused by noise and distortions occurring along the playback chain, but the staircase thingy is a myth because any half competently built DAC will have a reconstruction filter removing high frequency content that could still remain from quantization(or most likely nowadays, the spiky noise from delta sigma pulses). and because we recorded a band limited signal, we know that all those frequencies above should be removed as they have nothing to do with the original music. what remains should be pretty clean even on measurements(depends on the magnitude you're looking at and the quality of that filter).

My mistake, I forgot about the LPF filter in the DAC that removes the HF components of the stairs.

In that case, the only information aside from the pure signal itself would be noise introduced, proportional to quantization error from the discretization process, which probably gets removed anyways from noise shaping.

I guess I would need to do more research on the sources of error in the sine wave. The only thing I know for a fact is that the sine waves that are output from the audio chain are not ideal (since no machine can be made to perfect mathematical precision), and I would be interested in learning what that could be....

 

[bigshot]

Whether or not a difference is audible to human ears or not is pertinent to consider too.

 

[tansand]

Small changes in pitch and breath in voices would be a good example of detail, imo.

 

[gregorio]

I'd say detail is how closely the audio reproduction system produces perfect sinusoidal waves, which are directly correlated with audio tones. Infinitesimally accurate sine waves are synonymous with more details, where as jagged/rough sine waves are not as detailed.

All sound waves are made of pure/perfect sine waves, regardless of how "jagged/rough" (including square waves and sawtooth, etc.) and we've known this beyond any doubt for about 200 years. So I don't really understand what you're trying to say because "jagged/rough sine waves" are effectively "infinitesimally accurate sine waves" and therefore the level of detail must be the same.

Essentially, [1] the sine wave looks more like a staircase, and from my understanding, [2] the goal is the get those stairs infinitesimally small (e.g. 4 stairs per unit distance vs. 10,000 stairs for the same distance) to simulate the perfect sine wave, which takes more expensive equipment that can handle these calculations

There's two different (but related) errors with your understanding:

1. The sine waves "looks more like a staircase" simply because that is the convention of how audio editing/analysis software graphically represents the digital audio data. In other words, the "staircase" you see when you zoom in (in audio software) is due purely to the limitations of the graphical displays/interfaces but in reality there is no "staircase"!

2. Yes, more "stairs per unit distance" will make the graphical display more perfectly "simulate" the visual appearance of sine waves on your computer screen when you zoom in. However, that pertains ONLY to simulating the visual appearance of sine/sound waves on your computer screen, it does NOT pertain to the digital audio data (or resultant audio) itself! The digital audio data itself is not trying to "simulate the perfect sine wave", that's what analogue audio tries to do but digital audio is completely different. Think of it like the old telegraph system: We have a message, we convert that message into a binary code (Morse code, a series of dots and dashes) and then convert that Morse code back into the message. The Morse code itself doesn't look anything like the original message and it's not supposed to, it's not trying to "simulate" the message and only works because it isn't! This analogy might not appear pertinent to digital audio but in fact few analogies would be more pertinent because the fundamental theory of digital audio was born out of the telegraph system, developed by an engineer (Harry Nyquist) working on the telegraph system.

The only thing I know for a fact is that the sine waves that are output from the audio chain are not ideal (since no machine can be made to perfect mathematical precision) ...

While your statement is true, the whole point of digital audio (and digital information theory in general) is that it neatly bypasses this fact and thereby makes it irrelevant. Going back to the telegraph system: It wasn't possible to transmit dots and dashes perfectly but the reason the system worked is because it didn't make any difference, the telegraph system neatly bypassed that issue. It didn't matter how much noise/interference existed in the system or how badly it distorted the dots and dashes, provided the dots and dashes weren't distorted so badly that they couldn't be recognised/differentiated, then the message could be recovered absolutely perfectly! This is the identical principle upon which all digital systems work and if you think about it, your smartphone, laptop, computer, etc., is moving around and operating on many billions of bits of data every second, if only one of those billions of bit of data was not recovered and processed perfectly then your smartphone/computer/etc., would crash once every second!

Of course, once we're out of the digital domain and into the analogue and then acoustic domains, we ARE trying to "simulate the perfect sine waves" and then the imperfections inherent in all machines becomes relevant again and has an effect on our sound wave "simulations".

G

 

[WhatToChoose]

All sound waves are made of pure/perfect sine waves, regardless of how "jagged/rough" (including square waves and sawtooth, etc.) and we've known this beyond any doubt for about 200 years. So I don't really understand what you're trying to say because "jagged/rough sine waves" are effectively "infinitesimally accurate sine waves" and therefore the level of detail must be the same.



There's two different (but related) errors with your understanding:

1. The sine waves "looks more like a staircase" simply because that is the convention of how audio editing/analysis software graphically represents the digital audio data. In other words, the "staircase" you see when you zoom in (in audio software) is due purely to the limitations of the graphical displays/interfaces but in reality there is no "staircase"!

2. Yes, more "stairs per unit distance" will make the graphical display more perfectly "simulate" the visual appearance of sine waves on your computer screen when you zoom in. However, that pertains ONLY to simulating the visual appearance of sine/sound waves on your computer screen, it does NOT pertain to the digital audio data (or resultant audio) itself! The digital audio data itself is not trying to "simulate the perfect sine wave", that's what analogue audio tries to do but digital audio is completely different. Think of it like the old telegraph system: We have a message, we convert that message into a binary code (Morse code, a series of dots and dashes) and then convert that Morse code back into the message. The Morse code itself doesn't look anything like the original message and it's not supposed to, it's not trying to "simulate" the message and only works because it isn't! This analogy might not appear pertinent to digital audio but in fact few analogies would be more pertinent because the fundamental theory of digital audio was born out of the telegraph system, developed by an engineer (Harry Nyquist) working on the telegraph system.



While your statement is true, the whole point of digital audio (and digital information theory in general) is that it neatly bypasses this fact and thereby makes it irrelevant. Going back to the telegraph system: It wasn't possible to transmit dots and dashes perfectly but the reason the system worked is because it didn't make any difference, the telegraph system neatly bypassed that issue. It didn't matter how much noise/interference existed in the system or how badly it distorted the dots and dashes, provided the dots and dashes weren't distorted so badly that they couldn't be recognised/differentiated, then the message could be recovered absolutely perfectly! This is the identical principle upon which all digital systems work and if you think about it, your smartphone, laptop, computer, etc., is moving around and operating on many billions of bits of data every second, if only one of those billions of bit of data was not recovered and processed perfectly then your smartphone/computer/etc., would crash once every second!

Of course, once we're out of the digital domain and into the analogue and then acoustic domains, we ARE trying to "simulate the perfect sine waves" and then the imperfections inherent in all machines becomes relevant again and has an effect on our sound wave "simulations".

G

That last part right there was the whole point of the posts lol. You have to keep in mind, theory and reality are different.

Obviously digital storage encodes the data in binary format exact representation, and things like dithering and whatnot can reconvert the signal to a perfect sine wave without noise...in theory. But if you have ever worked with any circuitry (or machine) in real life, you'd understand that there is always some....funny business that limits you from achieving absolute mathematical precision.

The important takeaway that you need to keep in mind is that, unlike mathematical models, real machines do not exhibit infinite precision. For our discussion, this is where our perception of detail comes in: how close are those compressions and rarefactions at the transducer output to an ideal sine wave.

Try to see past the mathematics and theory, and try to understand the real world phenomena that we cannot necessarily measure directly, and thus need to predict with statistical models (like surface roughness, machining tolerances, etc). Read up on some datasheets of real (not ideal) components, like transistors and ball bearings. Then, you will understand that omnipresent separation between theory and reality. Then, try to apply that to the conversation to see the asymptotic approach towards mathematical precision.

Basically, the takeaway concept is this: you cannot manufacture a 2x2x2 cube. See if you can apply this concept to our discussion, and all will be made clear.

 

[Davesrose]

This is the identical principle upon which all digital systems work and if you think about it, your smartphone, laptop, computer, etc., is moving around and operating on many billions of bits of data every second, if only one of those billions of bit of data was not recovered and processed perfectly then your smartphone/computer/etc., would crash once every second!

A digital file does not have to be bit perfect to prevent a computer from crashing. There are data correction schemes for not having any perceptible difference in audio/video quality, or a player can continue to read data. Another example is a corrupted image file: a computer won't suddenly crash if you open it. You'll just see portions of it look garbled.