DAC types and their sonic signature

[solderdude]

Do you realize this is not a graph of a 20kHz signal, but instead a graph of a complex waveform consisting of more than one frequency?

yes.

Why is in bold a false statement? Because you don't see a 20kHz waveform on the graph. And you can't actually tell the amplitude on the 20kHz part of the waveform on graph either. Because it's mixed with the rest of frequencies in single graph from which you can't visually separate a 20kHz from the rest.

Actually I can .. IF it were a 'mix' of frequencies the peak of the 20kHz would be higher than the few peaks you see there. That peak actually is the -1dBdB FS and is what the original signal SHOULD be and is when a proper reconstruction filter would be used behind the R2R conversion.

But what we do know? Physics. Well, at least I know it and physics tells the continuous tone of 20kHz, which is a fundamental tone in this graph, well...won't have its amplitude changed because of presence of the rest of frequencies (which are higher than a 20kHz tone). Because amplitudes of various frequencies don't add.

If it were an addition (superposition) of 2 frequencies then the average level would be -3dB. Which is what we see on a frequency plot.
So If the extra frequency would be removed by a sharp filter the average would be -3dB because you remove the extra signal.
THERE is the roll-off which you insist does not exist.

You keep insisting that your ear 'removes' the ultrasonic content and thus the 20kHz which is 'hidden' in the complex waveform re-emerges and the amplitude is correct again. This would mean that amplitudes do not add but only subtract.
And we do know that superposition (mixing) of signals is an addition of signals not a merging.
Leaving out the (obligated) reconstruction filter and assuming the ear takes care of this by 'ignoring' everything above 22.05kHz is a novel idea.
Let's just say that you haven't come with compelling evidence you are correct in this.

 

[solderdude]

Using dithering, I'm following

Dither is only 1 LSB in size and is used to increase the effective dynamic range of the used bit depth and thereby decreasing quantization errors, at the expense of noise though. but this is less objectionable.
It is not of any help here as step sizes are much larger than 1LSB and vary in amplitude


We could ask Fury but I think he has gained a little more respect for his Flerken.

 

[luisma]

We could ask Fury but I think he has gained a little more respect for his Flerken.

It is just a scratch!! Nooo it is not

Dither is only 1 LSB in size and is used to increase the effective dynamic range of the bit depth used, at the expense of noise and thereby decreasing quantization errors.

I thought it was used "mainly" to hide quantization errors on non-audible frequencies, guess have to go back there again, did not know it was related to LSB but added algorhytm. Thanks

 

[solderdude]

Ian Shepperd has a nice way of demonstrating proper usage of dither

 

[solderdude]

If a DBT is someone's personal way to determine which sound he likes the most...so be it.

A DBT or BT is not intended to determine which sound he likes the most.. it is merely to determine IF there are audible differences under controlled circumstances.
It has nothing to do with taste nor with personal insecurities of 'objectivists' (you know what I am referring to)

 

[zalive]

Actually I can .. IF it were a 'mix' of frequencies the peak of the 20kHz would be higher than the few peaks you see there. That peak actually is the -1dBdB FS and is what the original signal SHOULD be and is when a proper reconstruction filter would be used behind the R2R conversion.

Graph the way it's drawn shows the amplitude of the waveform, this is how it looks like. Such thing doesn't exist in the nature. Each frequency has its amplitude, but there's no such thing as the amplitude of the complex waveform. Waveform as this is not a wave at all. It represents sum (well when we see a graph like this it's correct to call it a sum since its representation actually does sum everything up, while the superposition as happening in nature is not visible. Waveform doesn't exists in the nature, it's just graphics. Waves exist in the nature. And each wave has its own frequency and its own amplitude.

Again from the physics we know how to recognize whether a waveform represent a single frequency or a mix of frequencies. If it's a sine, it's a single frequency. If it's different than a sine, it's a mix of frequency. How do we know this? Because physics tells that wave (a single frequency) can only have a sine waveform. It represents its natural oscillation from being a wave, and the natural form of its varying amplitude is a sine.

Do you know what you get if you superpone (add) two waves of a single frequency (two sine waveforms), only being at a different phase? You get another sine. No matter which is the phase on which of them, as a result you will always get a sine waveform. Oh, there is one exception. If phase difference is 180 degrees while amplitudes are the same, you'll get...a flat line. Waves will completely cancel one another and there will be no sound/no signal. As for the other options, the phase difference will determine which will be the amplitude of the resulting wave.

 

[zalive]

A DBT or BT is not intended to determine which sound he likes the most.. it is merely to determine IF there are audible differences under controlled circumstances.
It has nothing to do with taste nor with personal insecurities of 'objectivists' (you know what I am referring to)

Nothing stops one from using a DBT (or a SBT) to determine which sound you prefer. What's invalid with that?
As for the taste, there's no difference between objectivists, subjectivists and all the rest we all have our own taste.

 

[solderdude]

Graph the way it's drawn shows the amplitude of the waveform, this is how it looks like. Such thing doesn't exist in the nature

It shows the exact signal as it goes to the transducers and is converted to soundwaves in the same amplitude.

there's no such thing as the amplitude of the complex waveform

Do you know what you get if you superpone (add) two waves of a single frequency (two sine waveforms), only being at a different phase? You get another sine. No matter which is the phase on which of them, as a result you will always get a sine waveform. Oh, there is one exception. If phase difference is 180 degrees while amplitudes are the same, you'll get...a flat line. Waves will completely cancel one another and there will be no sound/no signal. As for the other options, the phase difference will determine which will be the amplitude of the resulting wave.

That is 100% correct...and what if the frequencies differ and they aren't mixed at all nor is one a carrier wave ... what if one is a (fixed) sampling frequency and the other one a recorded one which can have any frequencies between a few Hz and Nyquist ?

You assume the 44.1kHz sampling frequency is a sinewave in disquise and mixes with the audio frequencies (like a carrier wave) and can simply be removed.... ehhh.... does not even have to be removed as we can't hear it so why remove it at all ?

 

[solderdude]

Nothing stops one from using a DBT (or a SBT) to determine which sound you prefer. What's invalid with that ?

Sure you can use such a test that way, to determine which one likes better. You don't need a blind test for that. you can do that sighted which gives better results as the eyes gets stimuli as well.

If the test is to determine if one DAC sounds different to another then taste has nothing to do with it ... nor before mentioned insecurities of objectivists..

You keep running in circles during the last 300 posts or so ... and keep coming back to the same things that have been explained in detail to you, over and over again, but which you refuse to believe.

 

[zalive]

It shows the exact signal as it goes to the transducers and is converted to soundwaves in the same amplitude.

But that's the thing. It doesn't show a signal as it is a nature. 2D in a single layer can't do it.
This is how it should look like (at least) to portrait how it looks like in the nature:

Resultant as shown in the last waveform is a graphical representation. Because there's no interference between different (frequencies) waves. Waves which are not sines don't exist. It's just graphics.

You see, when you're engineer but don't deal with the physics, when you look oscilloscope shots but don't understand what you're looking at, I understand if you end up with a misconception like you do. Only you need to quit being stubborn on this but instead study and learn, upgrade your engineering knowledge with the understanding of physics. Trust me, I'm trying at the same time to upgrade my knowledge in electronics so I can undesrtand basic design concepts well.

That is 100% correct...and what if the frequencies differ and they aren't mixed at all nor is one a carrier wave ... what if one is a (fixed) sampling frequency and the other one a recorded one which can have any frequencies between a few Hz and Nyquist ?

Frequencies will interfere only when frequency is identicall on both/all of them.
If frequencies are different there will be no interference.
Do you understand what it means?
It means is that they will remain completely separate. Its amplitudes won't sum. Waves won't join. They will simply coexist, they'll share the same space and that's all.

You assume the 44.1kHz sampling frequency is a sinewave in disquise and mixes with the audio frequencies (like a carrier wave) and can simply be removed.... ehhh.... does not even have to be removed as we can't hear it so why remove it at all ?

No, of course, that's what you assumed about me . Sampling frequency is a sine wave of course because it can't be anything else, but it's not present in or with the signal.

 

[zalive]

Sure you can use such a test that way, to determine which one likes better. You don't need a blind test for that. you can do that sighted which gives better results as the eyes gets stimuli as well.

Well I'd say it's helpful in a same manner as it is helpful to test for the audible difference. Same principle applies, you remove bias.

Just imagine yourelf doing a sighted test of filterless NOS vs filtered NOS (or any filtered DAC), with all the present negative opinion you have on engineering filterless DACs. This would not create any bias in you?

 

[solderdude]

But that's the thing. It doesn't show a signal as it is a nature. 2D in a single layer can't do it.
This is how it should look like (at least) to portrait how it looks like in the nature:

AGAIN... The waveform I showed is NOT an addition of waveforms AT ALL. It is a SAMPLED waveform with NOT enough samples to reconstruct it WITHOUT a filter... you either need more samples OR a proper reconstruction filter.

Frequencies will interfere only when frequency is identicall on both/all of them.
If frequencies are different there will be no interference.
Do you understand what it means?

I do... but you obviously don't.

Sampling frequency is a sine wave of course because it can't be anything else, but it's not present in or with the signal.

THAT is exactly what the mistake is in your line of thinking. It is NOT a 44.1kHz sine added/superimposed/mixed with audio. It is a sample RATE.

Just imagine yourelf doing a sighted test of filterless NOS vs filtered NOS (or any filtered DAC), with all the present negative opinion you have on engineering filterless DACs. This would not create any bias in you?

It would as it is sighted, would it not in your case ?

In a blind test there is no bias and know I can pick it being different. You probably can too. There is no mystery or magic there.

 

[Blumlein 88]

@zalive so what do you think the signal looks like coming from a microphone?

 

[zalive]

AGAIN... The waveform I showed is NOT an addition of waveforms AT ALL. It is a SAMPLED waveform with NOT enough samples to reconstruct it WITHOUT a filter... you either need more samples OR a proper reconstruction filter.

This is where you fail with your (non)undersanding of physics. Anything in the graph is constituted from sines. But you again forget that sines can be pulses. Even their amplitude can change dynamically (though it won't be the case in this aside of rise time/fall time of the reconstruction wave which you can't avoid). Only signal tone is being here continuous. Reconstruction waveforms can be square wave pulses, meaning they already consists of various frequencies pulses.

What you furthermore reiterate wrongly is that it's not the signal wave that is being reconstructed. Reconstruction waveform leftovers (which served for sampling purpose) are being removed. Signal wave is perfectly sampled with two samples per cycle. Or you count Nyquist and Shannon had no clue what they were calculating?

THAT is exactly what the mistake is in your line of thinking. It is NOT a 44.1kHz sine added/superimposed/mixed with audio. It is a sample RATE.

That's what I said in other words (or at least what I tried to say, possibly I failed to be clear), so I absolutely agree with this.

It would as it is sighted, would it not in your case ?

Of course, bias is included. Still I trust myself I can hear what I prefer if sighted. But someone who is prone to do blind tests, I don't see why he would do sighted tests to determine his preference.

In a blind test there is no bias and know I can pick it being different. You probably can too. There is no mystery or magic there.

If you can pick a preference in a blind test, you picked a difference too - same statistics applies, right? so blind testing for preference icludes testing for a difference, too. However testing only for a difference don't necessarily include testing for preference as well.

But, wait. What did you test exactly, related to NOS DAC? Can you describe tested setups?
Could you blind test a difference in sound between a filtered NOS vs filtered delta-sigma?

I do...

I'd love if this was the case.

 

[zalive]

@zalive so what do you think the signal looks like coming from a microphone?

No idea about microphones.

 

[Krunok]

@zalive so what do you think the signal looks like coming from a microphone?

Based on his famous statement ("Waves which are not sines don't exist. It's just graphics. ") the signal is obviously a series of sine waves.

You guys simply cannot win with arguments like this because this guy obviously thinks that complex waveforms don't exist so Fourier transform is not a way to REPRESENT a complex waveform with a bunch of sine waves but bunch of sine waves IS a complex waveform.

He was once shown a ball rolling over a circle and he was told that given the constant speed of the ball ordinate projection of the ball vs time is a sine wave. Then he was shown a Fourier transform and now he thinks only sine waves exist and that everything else is a "superposition" of sine waves.

Of course, if ball would not be rolling over a circle but over some more complex geometry form, like elipse or something even more complex, the resulting wave would not be a simple sine but a complex wave. You would still be able to REPRESENT that complex wave with a bunch of sine waves but the point here is that y-axis projection of the ball moving over an elipse is NOT a sine wave, thus proving that non-sine wave exists.

Of course, hoping that he will understand and accept an argument like this is absolutely an illusion but I wish you all the luck in your future efforts.

 

[zalive]

the signal is obviously a series of sine waves.

Yes, the signal is exactly a series of sine waves which happen simultaneously.

Of course, if ball would not be rolling over a circle but over some more complex geometry form, like elipse or something even more complex, the resulting wave would not be a simple sine but a complex wave. You would still be able to REPRESENT that complex wave with a bunch of sine waves but the point here is that y-axis projection of the ball moving over an elipse is NOT a sine wave, thus proving that non-sine wave exists.

You can make various models but those are not waves.

 

[mansr]

Should we tell him about wavelet transforms?

 

[Krunok]

Uf, not yet.. I think he should first accept a few basic facts, like the one that not all waves are sine, that Santa doesn't exist etc.

 

[SIY]

@solderdude You do realize that you're just being trolled at this point?