DAC types and their sonic signature

[solderdude]

Oh no ... please do not regurgitate all that nonsense again.
You sound like a Calexico errm.. broken record.

For 'some technical reason' reconstruction filters are essential for OS and NOS DACs.
Leaving them out at the playback end (a 'filter' is always present at the encoding process anyway) is well .... wrong in may ways.

The ASR audio designers that communicate here do not agree with the local audio designers you communicate with.

NOS R2R is fine ... filterless NOS DAC is not, unless you upsample the 44.1 content 4 or 8 times (properly) and then reproduce it with a NOS filterless DAC.

The best part with R2R is that it's not that expensive to make a DIY filterless NOS R2R and then experiment. Finished boards are available with more than one option. One might make a blind test on his own, then decide whether he preferred such DAC or some sigma-delta with those perfect measurements.

This I do agree with... someone preferring a sound signature.

 

[zalive]

For 'some technical reason' reconstruction filters are essential for OS and NOS DACs.
Leaving them out at the playback end (a 'filter' is always present at the encoding process anyway) is well .... wrong in may ways.

So, what harmful happens when you remove the filter on a NOS DAC?

This I do agree with... someone preferring a sound signature.

I use such filterless DAC. I also have few sigma-delta OS DACs (Micromega MyDac, Hifiberry DAC+ Pro on Raspberry Pi). I cannot tell any sound signature on any of them, all three are neutral when it comes to tonal balance. It's just that filterless DAC sounds superior to those two in terms of dynamic and naturality of the reproduction.

I have as well one battery DAC based on TDA1543 which does have a sonic signature. Plus a reduced dynamic response. It does sound nice and somehow natural as well, but it has a roll off on both frequency ends. So I know what a sound signature of a DAC is.

 

[solderdude]

So, what harmful happens when you remove the filter on a NOS DAC?

That has been shown a few times already... the signal is not anything remotely similar to the recorded signal.
Specifically at higher frequencies. That's what happens.

I use such filterless DAC. I also have few sigma-delta OS DACs (Micromega MyDac, Hifiberry DAC+ Pro on Raspberry Pi). I cannot tell any sound signature on any of them, all three are neutral when it comes to tonal balance. It's just that filterless DAC sounds superior to those two in terms of dynamic and naturality of the reproduction.

That's how it should be. The fact that you and some others prefer the sound of a filterless NOS DAC is a completely other matter.
It is not 'better', to some it sounds better, they prefer it.

You need to differentiate between preference and objectively better.
What some prefer does not mean it IS better just because you hang some nonsense theory on it.

 

[abm0]

So, what harmful happens when you remove the filter on a NOS DAC?

https://en.wikipedia.org/wiki/Aliasing#Online_audio_example

 

[zalive]

That has been shown a few times already... the signal is not anything remotely similar to the recorded signal.
Specifically at higher frequencies. That's what happens.

You're aware that any filter in DAC will remove solely the inaudible (ultrasonic) content, right?
So what sonic difference do you expect?

 

[zalive]

That's how it should be. The fact that you and some others prefer the sound of a filterless NOS DAC is a completely other matter.
It is not 'better', to some it sounds better, they prefer it.

But the thing is...in audio, for a personal choice of equipment, preference matters more.
And when it comes to ringing, this particular part/aspect is clearly not better, it's objectively worse on filtered DACs than on non-filtered.

 

[pkane]

But the thing is...in audio, for a personal choice of equipment, preference matters more.

Your preference matters more to you. To anyone else it is irrelevant. Yet you keep pushing it on to others. Why?

 

[Veri]

The benefit in filterless NOS R2R is in avoiding producing ringing. Filter produces ringing.

It's perfectly possible to have a filter that produces little to no ringing. It's all about trade-offs... gain something, lose something.
NOS has a perfect-looking impulse response. Great, right?! Further testing reveal that NOS probably has the most defects of all, look at these sines:

I've listened to NOS DACs. They sound quite rolled-off on top, which I'm sure some people don't mind. But there is no way you can claim this DAC does anything at all "better", it is defects/imperfections that you hear and you not minding them is all.

As @solderdude already said, some might even prefer this kind of output, but it will not be because of high fidelity reasons..

 

[zalive]

Your preference matters more to you. To anyone else it is irrelevant. Yet you keep pushing it on to others. Why?

I never said or implied my preference should matter more to anyone. You misunderstood.
I'm certainly not pushing it to anyone. Again, you misunderstood.
And the third thing, I'm not the subject of the thread.

 

[zalive]

It's perfectly possible to have a filter that produces little to no ringing. It's all about trade-offs... gain something, lose something.

This is what finer delta-sigma designs do. They try to find a fine line between sufficient filtering and less ringing (no ringing, impossible). By filtering gradually in more than one stage, for the example.

NOS has a perfect-looking impulse response. Great, right?! Further testing reveal that NOS probably has the most defects of all, look at these sines:

Stepped part belongs to ultrasonic range harmonics of those fundamental frequencies. Harmonics which show on a waveform are inaudible frequencies because those DACs are designed this way. Waveforms may look ugly but they sound the same as a pure sine waveform.

I've listened to NOS DACs. They sound quite rolled-off on top, which I'm sure some people don't mind.

Not every NOS DAC. It's a usual imperfection of R2R multibit chips. But guys from the Analog Design seems to have done a great job with their AD1865 chip. No audible roll off at all. At least it's so with the implementation I have, which uses its current output while doing the I/V conversion outside of the chip.

 

[solderdude]

This is what finer delta-sigma designs do. They try to find a fine line between sufficient filtering and less ringing (no ringing, impossible). By filtering gradually in more than one stage, for the example.

Stepped part belongs to ultrasonic range harmonics of those fundamental frequencies. Harmonics which show on a waveform are inaudible frequencies because those DACs are designed this way. Waveforms may look ugly but they sound the same as a pure sine waveform.

they are quantization errors that should NOT be there and are NOT harmonics.
And no they do NOT sound the same ..

Not every NOS DAC. It's a usual imperfection of R2R multibit chips. But guys from the Analog Design seems to have done a great job with their AD1865 chip. No audible roll off at all. At least it's so with the implementation I have, which uses its current output while doing the I/V conversion outside of the chip.

You really have no idea how DA conversion works it seems, nor what I/V conversion is, why it is needed and how different chips implement this.

 

[solderdude]

You're aware that any filter in DAC will remove solely the inaudible (ultrasonic) content, right?
So what sonic difference do you expect?

I expect a reconstruction filter to cut off everything above Nyquist (with sufficient attenuation).
I do not use filters that cut into the audible range (< 20kHz) with more than 0.3dB.
Without any filters ... ah well we have seen the pictures and how much it cuts within the audible band.

Like Calexico you do not seem to understand the whole reconstruction and post filtering thing at all.

 

[pkane]

I never said or implied my preference should matter more to anyone. You misunderstood.
I'm certainly not pushing it to anyone. Again, you misunderstood.
And the third thing, I'm not the subject of the thread.

You are very active on this thread, arguing for personal preference over measurements. I can provide many quotes, if you'd like. One of them was the one I directly responded to. But I guess I misunderstood your argument or the fact that you are trying to push it here. Don't bother to explain, I probably wouldn't understand it, either

But the thing is...in audio, for a personal choice of equipment, preference matters more.

 

[solderdude]

when it comes to ringing, this particular part/aspect is clearly not better, it's objectively worse on filtered DACs than on non-filtered

That is the ONLY aspect of it where it is 'better'.
It's a good thing you can still hear up to 20kHz so you can hear the dreaded ringing.

 

[watchnerd]

My modestly priced AT33EV cartridge is rated up to 50khz.

If inaudible high end frequency response is the most important thing, go vinyl.

 

[solderdude]

My modestly priced AT33EV cartridge is rated up to 50khz.

If inaudible high end frequency response is the most important thing, go vinyl.

Darn... my modestly priced UMC204HD is only up to 43 kHz (0 / + 0.3 dB) but it least it can record up to 50kHz.
Where have my upper 7kHz gone... holy ***t they are filtered out and ring like a X'mas bell.

 

[zalive]

That is the ONLY aspect of it in which it is 'better'.
It's a good thing you can still hear up to 20kHz so you can hear the dreaded ringing.

Ringing shows nicely at 10kHz.

I expect a reconstruction filter to cut off everything above Nyquist (with sufficient attenuation).
I do not use filters that cut into the audible range (< 20kHz) with more than 0.3dB.
Without any filters ... ah well we have seen the pictures and how much it cuts within the audible band.

Your understanding is that it cuts within the audible band because waveform looks ugly.
But if frequencies which make a waveform look ugly all belong to the ultrasonic range, how can they change what's audible?
Is there a difference in sound of a 12kHz sine wave and a 12kHz square wave?

https://www.audiomasterclass.com/ne...ference-between-a-square-wave-and-a-sine-wave

 

[zalive]

You are very active on this thread, arguing for personal preference over measurements. I can provide many quotes, if you'd like. One of them was the one I directly responded to. But I guess I misunderstood your argument or the fact that you are trying to push it here. Don't bother to explain, I probably wouldn't understand it, either

What you quote obviously means that own preference is what should matter the most to anyone.
So if your preference is completely neutral, let it be neutral. If your preference is not neutral, let it be the the way you prefer.

 

[pkane]

What you quote obviously means that own preference is what should matter the most to anyone.
So if your preference is completely neutral, let it be neutral. If your preference is not neutral, let it be the the way you prefer.

And we agreed on this days ago. That doesn't invalidate measurements, in fact it ensures that they matter more to everyone else. What you like in the privacy of your listening room is up to you, nobody can tell you what to like. Just like you can't tell others what to like because you do. NOS sounds better to you? Great! It sounds worse to me. I can explain exactly why NOS is objectively worse if no filtering is used. It's been done by others here already, many times. Explain to me why I should care that you like NOS.

 

[solderdude]

Ringing shows nicely at 10kHz.

Yes it does with an illegal test signal.

http://archimago.blogspot.com/2013/06/measurements-pulse-response-5khz-10khz.html

What you need to understand is that (on my request) Archimago did that test but... and here is the kicker... is only true for test signals that start and stop on the 0 line.
By definition that signal is illegal and contains a frequency content (at the start and stop) that goes far beyond Nyquist and can not exist in a recorded music signal as that would have been filtered out by the anti-alias filter in the ADC chain.
You can 'see' this when you look at the waveform and zoom in around the 0V line.
When you do, you will see the signal rises and stops instantly which is squarewave and thus a signal that is not recorded.
That ringing is NOT at 10kHz but at 22kHz... inaudible.. and it is because the US content of the sudden rise and stop 'triggers' the 20kHz ringing.

Your understanding is that it cuts within the audible band because waveform looks ugly.

No that is not my understanding... it's what you think is my understanding of it.
It's the consequences of the waveform in the amplitude domain over time.
Have a look at the amplitude of the 10 and 16kHz signal in post #310 and look at the amplitude which varies and it should not do this.
the amplitude thus drops and rises (varies) constantly depending on synchronization of the frequency in question with the sample frequency.
That dropping of amplitude is the 'roll-off cutting in the audible band'

But if frequencies which make a waveform look ugly all belong to the ultrasonic range, how can they change what's audible?
Is there a difference in sound of a 12kHz sine wave and a 12kHz square wave?

The 12kHz square and sinewave comparison is not what this is all about.
That has to do with related harmonics.
What if the amplitude of that 12kHz, that should be constant, varies in amplitude between nominal and way below nominal constantly and the average of it is lower than what it should be ?