ESS THD ‘Hump’ Investigation

[amirm]

Another designer who has conquered the ESS IMD hump: https://www.audiosciencereview.com/...ents-of-c1da-9038s-bal-portable-dac-amp.8424/

 

[diegooo1972]

In one of their talks they chatted about listening to audiophiles and making changes to the design. I wonder if this is one of the consequences.

Why that reminds me of the infinite harmonic distortions in Totaldac D1 six ?
Seems to me that highend audio world is considering specifics harmonic distortions as something good in the perception of music.
For example intermodulation distortion is used in guitar overdrive and the theory of power chord.

From wikipedia.

When two or more notes are played through a distortion process that non-linearly transforms the audio signal, additional partials are generated at the sums and differences of the frequencies of the harmonics of those notes (intermodulation distortion).[2] When a typical chord containing such intervals (for example, a major or minor chord) is played through distortion, the number of different frequencies generated, and the complex ratios between them, can make the resulting sound messy and indistinct.[3] This effect is accentuated as most guitars are tuned based on equal temperament, with the result that minor thirds are narrower, and major thirds wider, than they would be in just intonation.

However, in a power chord, the ratio between the frequencies of the root and fifth are very close to the just interval 3:2. When played through distortion, the intermodulation leads to the production of partials closely related in frequency to the harmonics of the original two notes, producing a more coherent sound. The intermodulation makes the spectrum of the sound expand in both directions, and with enough distortion, a new fundamental frequency component appears an octave lower than the root note of the chord played without distortion, giving a richer, more bassy and more subjectively 'powerful' sound than the undistorted signal.[4] Even when played without distortion, the simple ratios between the harmonics in the notes of a power chord can give a stark and powerful sound, owing to the resultant tone effect. Power chords also have the advantage of being relatively easy to play (see "Fingering" below), allowing fast chord changes and easy incorporation into melodies and riffs.

Is that the "secret" behind Totaldac ?
While this can be subjective I recognize that power chords are amazing in rock guitars but I don't think it really make some sense in audio fidelity.
Seems to me the same kind of subjective assumption high end audio world love and I don't want any kind of trickery in fidelity.

 

[DonH56]

As a brass player those subharmonic "beat frequencies" or tones are critical to ensuring we are all in tune. I do not want my sound system adding them on its own, and as indicated in the Wikipedia article the difference tones are only "pleasant" when things are in tune (justly). Having a DAC or anything else add IMD and claiming it improves the sound is mind-boggling. Maybe for certain players/instruments/songs at the right time and chord but in general? Ain't seein' it...

IMO - Don

 

[diegooo1972]

@DonH56
I concord.
But my doubt about the fact that Totaldac for example is trying to add this intermodulation distortion on certain notes is quite strong.
I fear that ESS has tried to do something similar on purpose following audiophile world.
For some reason Amir had similar doubts. But honestly I don't know why.

 

[daftcombo]

@DonH56
I concord.
But my doubt about the fact that Totaldac for example is trying to add this intermodulation distortion on certain notes is quite strong.
I fear that ESS has tried to do something similar on purpose following audiophile world.
For some reason Amir had similar doubts. But honestly I don't know why.

That would be very interesting to investigate.

 

[johan]

I am pleased to say that our new DAC got rid of the ess hump. This was taken on single ended output at 2.2Vrms

 

[Veri]

I am pleased to say that our new DAC got rid of the ess hump. This was taken on single ended output at 2.2VrmsView attachment 32560

Yay! Good job!

 

[johan]

PLease note how close (equal) we are to Okto .Matrix seems to be better (by a hair) but of course its tested of 4Vrms balanced.

 

[johan]

Obviously, you don't understand the nature of this issue.
This issue has nothing to do with the chip reversions, only related to the specific circuit design.

That was my finding as well , it is related to circuit but I want to add that DAC ic is rather sensitive to the output stage

 

[diegooo1972]

@johan
Wow. That seems to me a very generous information.
Best regards.

 

[Zoide]

I am pleased to say that our new DAC got rid of the ess hump. This was taken on single ended output at 2.2VrmsView attachment 32560

Nice! When is it coming out? What is the name of your company?

Thanks

 

[johan]

I will keep this for tech discussions not for advertising .

 

[flowjm]

Nice! When is it coming out? What is the name of your company?

Thanks

https://www.allo.com/sparky/index-dac.html

No affiliation

 

[scott wurcer]

This is a plot of an ESS DAC design exhibiting the IMD hump. I don't see any issue with concluding there is some kind of pathological behavior. I also don't see the reasoning about how this behavior is difficult to measure, etc.

 

[maxxevv]

Did the ES9018 chip have such an issue too ?

I was under the impression that it manifested after the introduction of the 9028 and 9038 versions ?

 

[scott wurcer]

Did the ES9018 chip have such an issue too ?

I was under the impression that it manifested after the introduction of the 9028 and 9038 versions ?

No the ES9018 had an issue.

 

[eslei]

hi, there,

i follow the thread for some time, most mentioned measurements are for dacs from commercial products. myself would like to focus on diy dacs, so i developed a volumio plugin to control ess chip's harmonic distortion compensation.

my system is,

digital file -> rpi ( volumio ) -> I2S -> smp v01 es9028q2m board.
/////////////////// | ///// ///////// /////////////////////////////^
/////////////////// |______volumio plugin________i2c _________|


i never do this kind of analyze measurements before, any guidance and suggestions are appreciated.



cheers

 

[eslei]

the plugin is on hex, each harmonic has 4 bits, each bit has a drop-down menu from "0" to "f", so we have a choice of "0x0000 = 0" to "0xffff = 65535".

 

[amirm]

That's pretty cool. For testing you need an ADC/Sound interface. If you have one, then you need to generate the IMD signal which is just a combination of 60 Hz and 7 kHz. Play that at different levels and see if there is a sudden rise as I show in my charts.

 

[amjosh]

Anyway to fix the hump in the old D50 (non S) version? . I was so happy with with my D50 and THX AAA setup before reading this thread