Review and Measurements of Audio-gd NFB28.28 DAC and Headphone Amp

[Bluespower]

It would be less stressful! That’s been our whole point from the beginning

Why? I think it s more difficult to render a larger dynamic range. Harmonic rendering is very important for 3d sense and timbre quality that's why i would like to see curves that show how a signal with harmonics is good rendered. Maybe we would see impact of the filter type also.

 

[Krunok]

Hmm, I don’t see how it could put more stress on a speaker. What’s your line of thought on it?

I could imagine that if you send 2 tones to the speaker, say 40Hz and 55Hz, both of high amplitude, it may put more stress to the speaker than one tone, say 40Hz of high amplitude and the other at say 8kHz of low amplitude. Isn't that so?

 

[Bluespower]

I could imagine that if you send 2 tones to the speaker, say 40Hz and 55Hz, both of high amplitude, it may put more stress to the speaker than one tone, say 40Hz of high amplitude and the other at say 8kHz of low amplitude. Isn't that so?

Yes it s less stress for the speaker but more stress for the dac to render properly

 

[March Audio]

To summary, what i m looking for is a test that show ability of dacs to keep low level harmonics of instruments with fidelity. This is the dynamic. Dynamic is not ability to perform one low level tone alone and one high level tone alone. Dynamic is to perform both at the same time. This leads to realistic timbre. That is why i think the test is not pointless.
@amirm

IMD and thd are related. What you are asking for is a dac with the lowest distortion. This will maintain the integrity of the low level signals the most.

A high level IMD test tells us what we need to know about the distortion. Using lower levels tells us nothing different, just makes it more difficult to see.

 

[Bluespower]

IMD and thd are related. What you are asking for is a dac with the lowest distortion. This will maintain the integrity of the low level signals the most.

A high level IMD test tells us what we need to know about the distortion. Using lower levels tells us nothing different, just makes it more difficult to see.

Well how do you know it doesn't show anything if nobody make the test? I think test are biased and selected to have good results on modern delta segma dacs. Multi bit (5 bit if i remember) delta sigma dacs have different behaviour when you reach the multi bit level. I think with the 12db differences in imd you re within the multibit part of the musltibit delta sigma dac

 

[March Audio]

Not at all. With respect you have a misunderstanding of the mechanisms. Experience and understanding is how we know.

All you want is a dac with the lowest distortion so there are the least spurious tones to interfere with the musical ones you refer to. The distortion products will be the same if you use high or low level test signals. However using low level signals will make the spurious tones at a lower level and more difficult to see, and may even disappear into the noise floor.

I think you are conflating time domain and frequency domain. Those small harmonics aren't separate signals in the time domain, theycare part of the large overall waveform. I think you are also conflating distortion and resolution.

 

[Bluespower]

Not at all. With respect you have a misunderstanding of the mechanisms. Experience and understanding is how we know.

All you want is a dac with the lowest distortion so there are the least spurious tones to interfere with the musical ones you refer to. The distortion products will be the same if you use high or low level test signals. However using low level signals will make the spurious tones at a lower level and more difficult to see, and may even disappear into the noise floor.

A test is better than experience and thinking to,proove. And i never see such a test. Maybe i missunderstood but that would help me to see that in curves. Input: one sine With 2nd order and 3rd order low level harmonics and see how the dac render it. I think it s more easy to understand. I think you simplify too much. One dac that can render a -80db tone perfectly and a -6 db perfectly maybe will not render them perfectly when they are mixed. I'm curious to see if the test would show what you expect. Why are you not curious of the results ? If theory and experience is all then no needs to test anything?
All i want is a proof test that show that i m wrong. @amirm

 

[March Audio]

So you want to add harmonics into the signal?

That will just confuse the issue, and make it more difficult to quantify the effect of harmonic distortion.

Yes you are describing intermodulation distortion. For which we already have a test.

 

[Bluespower]

So you want to add harmonics into the signal?

That will just confuse the issue, and make it more difficult to quantify the effect of harmonic distortion.

Yes you are describing intermodulation distortion. For which we already have a test.

I would like a test like IMD but more than 12db differencies (at least 20db) between the tones and maybe more than two tones. For me that would be the dream if you got 16 bit signal you got 95db range then we should test differencies more than 12 db between two tones. Then we should check that the level'of the tone are unchanged at the dac output to validate the performance. I see no tests for that so @amirm could be the first to do that and change the way audio dac are evaluated.

 

[Krunok]

So you want to add harmonics into the signal?

I see no point in doing that. IMHO that would only mask distortion of the 1st tone.

 

[Bluespower]

I see no point in doing that. IMHO that would only mask distortion of the 1st tone.

Maybe the harmonics will be masked or altered also and that would be bad. The goal'is just to check if the level of harmonics in the output is unchanged and if there are not unwanted new hamronics that appears.
What you say is hamronic masks distortion ? Then the harmonics in nos dacs will masks the distortion lol

 

[andreasmaaan]

I could imagine that if you send 2 tones to the speaker, say 40Hz and 55Hz, both of high amplitude, it may put more stress to the speaker than one tone, say 40Hz of high amplitude and the other at say 8kHz of low amplitude. Isn't that so?

If different drivers are reproducing the tones (which they would be unless the speaker is a single driver full range design), the latter test would ofc be less stressful (no IMD between the two drivers).

Assuming the two tones are at the same frequencies for both tests, and in the second test the level of the second tone is lowered, I would still think the second test should be less stressful. Or am I misunderstanding your question maybe?

 

[Krunok]

If different drivers are reproducing the tones (which they would be unless the speaker is a single driver full range design), the latter test would ofc be less stressful (no IMD between the two drivers).

Yes, that's what I meant.

 

[andreasmaaan]

Maybe the harmonics will be masked or altered also and that would be bad. The goal'is just to check if the level of harmonics in the output is unchanged and if there are not unwanted new hamronics that appears.
What you say is hamronic masks distortion ? Then the harmonics in nos dacs will masks the distortion lol

NOS DACs do ofc create huge amounts of distortion. Perhaps that’s why ppl like them...

What’s the theoretical basis for your belief that reducing the level of the higher frequency tone in a two-tone IMD should be more stressful. I’m still struggling to understand this.

 

[Krunok]

NOS DACs do ofc create huge amounts of distortion. Perhaps that’s why ppl like them...

What’s the theoretical basis for your belief that reducing the level of the higher frequency tone in a two-tone IMD should be more stressful. I’m still struggling to understand this.

I believe he doesn't necessarily think it would be more stressful but that it would, together with having more than 2 tones, more ressemble real music and thus be more realistic. @Bluespower , is that correct?

 

[Bluespower]

NOS DACs do ofc create huge amounts of distortion. Perhaps that’s why ppl like them...

What’s the theoretical basis for your belief that reducing the level of the higher frequency tone in a two-tone IMD should be more stressful. I’m still struggling to understand this.

We don't know exactly how oversampling filter works. In the shanon theorem the frequency should be infinite. They are limited by the power of calcul of the dac. So i would like to see the impact on the harmonics. Maybe they are good for pure tones or complex tone with low dynamic. But for complex tones with high dynamic i'd like to see how they perform. In multibit all bit comes exactly at the same time at 44.a khz. With multibit delta sigma it,s much more complex that's why i would like to see their behaviour with high dynamic mixed tones.

 

[Bluespower]

I believe he doesn't necessarily think it would be more stressful but that it would, together with having more than 2 tones, more ressemble real music and thus be more realistic. @Bluespower , is that correct?

Yes it would be a test more real then easy'to understand and no doubt that a dac could render complex signals like'piano'or accoustic guitar.

 

[andreasmaaan]

We don't know exactly how oversampling filter works. In the shanon theorem the frequency should be infinite. They are limited by the power of calcul of the dac. So i would like to see the impact on the harmonics. Maybe they are good for pure tones or complex tone with low dynamic. But for complex tones with high dynamic i'd like to see how they perform. In multibit all bit comes exactly at the same time at 44.a khz. With multibit delta sigma it,s much more complex that's why i would like to see their behaviour with high dynamic mixed tones.

DS DACs function in the megahertz range which is high enough to perform this operation correctly with any signal within the audio band.

However, why don’t you propose a specific test and see if someone here with adequate instrumentation is willing to perform it?

Perhaps something like the SMPTE test but with the higher frequency tone at 1/128th power (-42dB)?

 

[Krunok]

What you say is hamronic masks distortion ?

Yes, harmonics of the tone mask harmonic distortion of that tone. For that reason IMD test is more relevant than THD (single tone) test as it produces non harmonic distortion tones which are easier to hear as they are not masked with anything.

 

[Bluespower]

@andreasmaaan with modern dac chips you can still have let's say 16 bit resolution at 20khz? Cause with dsd wich is dure delta sigma you loose resolution as the frequency grows. That is a theorical approach and with lower level high frequency tone in the'imd'test that will show that disadvantage of delta sigma. But i think new chips do better than older for this. For 2,8 mhz dsd to render a 20khz tone you'will have only 140 steps (2800/20)that means you have only 7 bit resolution at 20khz. Of course dither can improve but artificially. So maybe on modern dac chips the true resolution for high frequency is less than 16 bit.