On the Distortion of Cirrus Logic CS431xx-Based Devices: A Comparative Review

[Luciferium]

In the list of devices ONIX Alpha XI1 not producing hump, but we don't know firmware version, which one of this versions 1.0.3, 1.1.3, 1.2.3.
And can someone please give me the DAC measurements of Onix Alpha XI1. Can't find any, there is even no PCB pic)

 

[jkim]

In the list of devices ONIX Alpha XI1 not producing hump, but we don't know firmware version, which one of this versions 1.0.3, 1.1.3, 1.2.3.
And can someone please give me the DAC measurements of Onix Alpha XI1. Can't find any, there is even no PCB pic)

The Onix Alpha Xl1 was tested by RAA---use a VPN in case you cannot see this webpage (it seems that some US ISPs block Russian websites).

As for its Cirrus hump distortion, here's my best informed guess:

• Firmware versions should not matter. So far it appears that no manufacturer has deliberately addressed this distortion issue.
• Although it was tested (by RAA) not producing the "Cirrus hump" as defined in my review, it should produce the distortion demonstrated by the additional tests in Part II of the review.

 

[zany.ninja]

This is a nice alternative. Measurements by RAA look solid---Amir's measurements were limited by an ASIO driver issue he had before. I agree, for the current price, one may seriously consider it.

Yeah, I suppose it merits inclusion on your recommended list, especially for the price (actually is down to $45 atm). The AKM AK4377 DAC + Ti/Burr-Brown INA1620 op-amp combo in the X1 offer a good and unique alternative for anyone looking to avoid getting a Cirrus chip and/or the RT6863/SGM8261 op-amps you mentioned. Plus it has a customisable 10-band EQ too.

 

[jkim]

Yeah, I suppose it merits inclusion on your recommended list, especially for the price (actually is down to $45 atm). The AKM AK4377 DAC + Ti/Burr-Brown INA1620 op-amp combo in the X1 offer a good and unique alternative for anyone looking to avoid getting a Cirrus chip and/or the RT6863/SGM8261 op-amps you mentioned. Plus it has a customisable 10-band EQ too.

Where can I find information about its customizable 10-band EQ?

 

[zany.ninja]

Where can I find information about its customizable 10-band EQ?

The support page for the Creative app has a comprehensive overview and user guide of all the features, including EQ.

 

[dvsv]

i got tired of opening chrome(i use firefox) just to access the web app. so ide revered engineered the driver and made myself a little program to change the settings on my Black pearl. its its very bare bones. but it works

View attachment 458696

EDIT: yeah, ide made one for android too.. 80% of the time was getting monet to work lol

View attachment 458736

Hi!
I got the same DAC and idk how to control some of its parameters.
Can you upload this software?
Thank you!

 

[jkim]

Hi!
I got the same DAC and idk how to control some of its parameters.
Can you upload this software?
Thank you!

That software was created to change the parameter more conveniently. You can still use the WalkPlay android or web-based app to control the same parameter on the TRN Black Pearl. However, as mentioned in this post, using the NOS filter is not recommended due to the extremely high out-of-band distortion.

 

[pangcukaipang]

i got tired of opening chrome(i use firefox) just to access the web app. so ide revered engineered the driver and made myself a little program to change the settings on my Black pearl. its its very bare bones. but it works

View attachment 458696

EDIT: yeah, ide made one for android too.. 80% of the time was getting monet to work lol

View attachment 458736

Would you mind sharing both software? Thanks in advance.

 

[KDRN]

Would you mind sharing both software? Thanks in advance.

currently i could only give the android app right now. since the windows version is a python script and sorta kinda made windows defender unhappy and deleted it.(idk why, maybe it doesnt like seeing code that interacts with usb devices). ill try to recover or remake it.

Download black pearl control | LimeWire

Download black pearl control on LimeWire

limewire.com

maybe ill just post them in github if i ever stopped procrastinating
edit: heres the source code for the android app: https://github.com/cheesyserg/blackPearlControl-Android

 

[jkim]

As time has passed and more tests have been done, I feel the need to update my take on the audibility issue of this Cirrus hump distortion. To avoid a verbose description of this topic, I put this content here in a separate post.

In my review I initially stated it is difficult to determine its audibility in real audio content. True, but at the same time I wanted to be conservative since there should be many people who own those affected devices. Another reason is that it is impossible to draw an exact percentage of the time it is audible out of infinite possibilities. But it IS audible. There's a reason why Roman at RAA used a clip from the Dune soundtrack. I found that the Cirrus hump distortion is most noticeable when audio consists of somewhat complex low-frequency signals with not much high-frequency contents. The Dune soundtrack clip is such an example. And an ideal example is the C Major test signal used in Part II of my review. To hear, simply play the original (not recorded) signal at -14 to -15 dBFS through an affected device and listen to it using an IEM with suitable sensitivity. I used the 7 Hz x Crinacle Zero:2. The distortion is very, very clearly audible.

Below are the spectrograms of the recordings of the Dune soundtrack and C Major clips, respectively, played from the KA15:

Distortions in both cases are clearly audible through the Zero 2 IEM.

How about the JCally JM20 which does not produce the Cirrus hump from steady signals but still distorts when a signal begins (as shown in Part II of my review)?

Although less often, these distortions are still audible from the Zero 2.

 

[mc.god]

I can confirm tha audibility in specific conditions, like the C-Maj test mentioned by @jkim, with all my CS dongles:

-Sonata BHD (non pro): from measurements it doesn't produce the mentioned "hump" but playing C-Maj at certain volume level I can hear three faint clicks at the beginning of each iteration of the sample, much like the JM20 graph by jkim;
- Kuang Pai Player 3 and Ugreen Hi-Fi Pro: from measurements both are affected by the "hump", and indeed you ca hear faint clicks for the entire iterations of the signal, much like the KA15 graph poste by jkim;
- JM20: here we have a peculiar result, while I'd expect a behavior similar to the jkim's JM20, I actually can't hear any faint click, but instead a faint background noise that quickly fades away in the brief instants between the three iterations of the signal and then come back again. It's not a hissing but more like a pink noise, an it correlates with the fact that my JM20 also produce a brief pink noise when it ramps-up from idle state, ulike the jkim a othe users units. This makes me conclude my JM20 is differently implemented, maybe firmware settings in earlier units since i own mine since January 2024.

All listening tests made with 7Hz Zero.

 

[DSJR]

I'm finding this rather troubling, as 'we' tell newbies here that dacs are a done deal and transparent to the source material, yet herer we are with a measurable and sadly audible distortion/noise issue that detracts from *true* neutrality and an otherwise great performance...

 

[mc.god]

I'm finding this rather troubling, as 'we' tell newbies here that dacs are a done deal and transparent to the source material, yet herer we are with a measurable and sadly audible distortion/noise issue that detracts from *true* neutrality and an otherwise great performance...

I agree, even though these are practically inaudible issues in most real music listening scenarios, they draw a veil on this CS chip that otherwise performs and sound so good, and put your mind at unease just for knowing they exist.

 

[CedarX]

I'm finding this rather troubling, as 'we' tell newbies here that dacs are a done deal and transparent to the source material, yet herer we are with a measurable and sadly audible distortion/noise issue that detracts from *true* neutrality and an otherwise great performance...

I also think that DACs are done deal and transparent must be accompanied with a few caveats.

Modern chips tend to pack much more than a plain vanilla DAC. The CS431xx family is a good example, but ESS DACs are probably as “sophisticated”.
My favorite example is the KTmicro range of super-cheap SoC’s: not sure if they are completely transparent, but they include DAC… USB bridge, DSP w/ PEQ, noise gate & limiter, adjustable gain, and a HP/IEM amp… All of this to be integrated in <$10 dongles and USB-C IEM cables by OEMs who appear to have no clue these features are even available to them—or choose to ignore them and let the default parameter settings.

I think there is an open void, besides Amir’s standard set of measurements (these shall remain!) to develop tests aimed at identifying these extra features and what they do (DeltaWave-based?). @jkim did that for the CS431xx DRE, but what if an OEM decides to leverage the programmable filter of an ES9039Q2M to implement a weird sound “signature”… Would we detect this? How can we confidently say that the audiophiles reported differences are jut in their head?

It’s not about saying that “absolutely transparent” is the only correct answer, but more about understanding what is implemented in these DACs and why.

 

[oleg87]

I also think that DACs are done deal and transparent must be accompanied with a few caveats.

Modern chips tend to pack much more than a plain vanilla DAC. The CS431xx family is a good example, but ESS DACs are probably as “sophisticated”.
My favorite example is the KTmicro range of super-cheap SoC’s: not sure if they are completely transparent, but they include DAC… USB bridge, DSP w/ PEQ, noise gate & limiter, adjustable gain, and a HP/IEM amp… All of this to be integrated in <$10 dongles and USB-C IEM cables by OEMs who appear to have no clue these features are even available to them—or choose to ignore them and let the default parameter settings.

I think there is an open void, besides Amir’s standard set of measurements (these shall remain!) to develop tests aimed at identifying these extra features and what they do (DeltaWave-based?). @jkim did that for the CS431xx DRE, but what if an OEM decides to leverage the programmable filter of an ES9039Q2M to implement a weird sound “signature”… Would we detect this? How can we confidently say that the audiophiles reported differences are jut in their head?

It’s not about saying that “absolutely transparent” is the only correct answer, but more about understanding what is implemented in these DACs and why.

Being able to load custom filter coefficients isn't really going to let you do anything that would escape the standard panel of tests.
But yes, I agree that such tests may be inadequate to characterize various 'clever' optimizations that rely on level-dependent switching and hysteresis.

 

[dvsv]

As time has passed and more tests have been done, I feel the need to update my take on the audibility issue of this Cirrus hump distortion. To avoid a verbose description of this topic, I put this content here in a separate post.

In my review I initially stated it is difficult to determine its audibility in real audio content. True, but at the same time I wanted to be conservative since there should be many people who own those affected devices. Another reason is that it is impossible to draw an exact percentage of the time it is audible out of infinite possibilities. But it IS audible. There's a reason why Roman at RAA used a clip from the Dune soundtrack. I found that the Cirrus hump distortion is most noticeable when audio consists of somewhat complex low-frequency signals with not much high-frequency contents. The Dune soundtrack clip is such an example. And an ideal example is the C Major test signal used in Part II of my review. To hear, simply play the original (not recorded) signal at -14 to -15 dBFS through an affected device and listen to it using an IEM with suitable sensitivity. I used the 7 Hz x Crinacle Zero:2. The distortion is very, very clearly audible.

Below are the spectrograms of the recordings of the Dune soundtrack and C Major clips, respectively, played from the KA15:
View attachment 461330

View attachment 461329

Distortions in both cases are clearly audible through the Zero 2 IEM.

How about the JCally JM20 which does not produce the Cirrus hump from steady signals but still distorts when a signal begins (as shown in Part II of my review)?
View attachment 461333

View attachment 461332

Although less often, these distortions are still audible from the Zero 2.

I tested this Dune soundtrack here with TRN Black Pearl and Tangzu Fudu on a Win 11 desktop pc and set -15 dB preamp on Peace Equalizer and i couldn't hear nothing wrong.
How should i test it? I mean where i set the -15 dBFS? (there's the dac volume and windows volume settings).
I sent an email to TRN and they just ignored me, should i return the dac for a refund? (idk if this can tarnish the vendor's reputation).
Thanks!

 

[jkim]

I tested this Dune soundtrack here with TRN Black Pearl and Tangzu Fudu on a Win 11 desktop pc and set -15 dB preamp on Peace Equalizer and i couldn't hear nothing wrong.
How should i test it? I mean where i set the -15 dBFS? (there's the dac volume and windows volume settings).
I sent an email to TRN and they just ignored me, should i return the dac for a refund? (idk if this can tarnish the vendor's reputation).
Thanks!

It's much easier to hear it when you first try the C Major clip because the distortion (clicking) occurs in rhythm.

Again, play the original (not recorded) signal at -14 to -15 dBFS through an affected device and listen to it using an IEM with suitable sensitivity. Not sure how sensitive your IEM is. In any case, set the DAC and Windows volume to max. If you use a player that supports volume control on dB scale, use it---foobar2000 supports it. If not, you can set the EQ-APO's preamp cut to -15 dB and play it at your player's max volume.

If you can hear and get used to the distortion (clicking), then try the Dune soundtrack. You will be able to hear it there, too.

 

[dvsv]

It's much easier to hear it when you first try the C Major clip because the distortion (clicking) occurs in rhythm.

Again, play the original (not recorded) signal at -14 to -15 dBFS through an affected device and listen to it using an IEM with suitable sensitivity. Not sure how sensitive your IEM is. In any case, set the DAC and Windows volume to max. If you use a player that supports volume control on dB scale, use it---foobar2000 supports it. If not, you can set the EQ-APO's preamp cut to -15 dB and play it at your player's max volume.

If you can hear and get used to the distortion (clicking), then try the Dune soundtrack. You will be able to hear it there, too.

I tested again both flac files on Tangzu Fudu and KZ Zsn 2 Pro iems (106dB/mw and 108dB/mW respectively) and now i could hear some faint clicks.
But there's a catch: I did the same test on all my DACs that i got here (Jcally JA56 ALC5686 chip, Graveaudio CX31993 and Fiio JA11) and they all exhibited some clicks one way or another and imho by far the worst was the Fiio JA11 and the cleanest sounding was the cheap Graveaudio, the TRN Black Pearl was in the "middle" and still way better than JA11.

Should the average consumer be that worried about this issue or are we making the purchasing decision needless difficult?
Anyway i bought this TRN Black Pearl DAC to power a kinda hard to drive 94 dB/mW KZ PRX planar IEM iirc this distortion issue should be even more hard to hear right?

Edit: I found a DAC and Windows volume setting that basically zeroed the issue on the TRN Black Pearl DAC + Tangzu Fudu IEM and now is the cleanest sounding DAC on the C Major test idk if i'll change my mind later with the PRX planar IEM and maybe there's still some very faint clicks on the Dune soundtrack but prob it's noise from my USB cuz i was using the GPU on Stable Diffusion when i heard a click, in the end of the day i really think this issue is overblow and i immediately canceled the refund process.

Keep in mind that most ppl buy these cheap DAC to power cheap IEM or headphones (like KZ PRX, Artti T10, etc) and they're not using with U$20K audiophile gear, if we gonna reach the conclusion that most average consumer DACs are flawed in some way or another (heck even expensive Macbooks) then i guess ignorance is a bliss!

Thanks and sorry for my english!

 

[jkim]

I tested again both flac files on Tangzu Fudu and KZ Zsn 2 Pro iems (106dB/mw and 108dB/mW respectively) and now i could hear some faint clicks.
But there's a catch: I did the same test on all my DACs that i got here (Jcally JA56 ALC5686 chip, Graveaudio CX31993 and Fiio JA11) and they all exhibited some clicks one way or another and imho by far the worst was the Fiio JA11 and the cleanest sounding was the cheap Graveaudio, the TRN Black Pearl was in the "middle" and still way better than JA11.

Should the average consumer be that worried about this issue or are we making the purchasing decision needless difficult?
Anyway i bought this TRN Black Pearl DAC to power a kinda hard to drive 94 dB/mW KZ PRX planar IEM iirc this distortion issue should be even more hard to hear right?

Edit: I found a DAC and Windows volume setting that basically zeroed the issue on the TRN Black Pearl DAC + Tangzu Fudu IEM and now is the cleanest sounding DAC on the C Major test idk if i'll change my mind later with the PRX planar IEM and maybe there's still some very faint clicks on the Dune soundtrack but prob it's noise from my USB cuz i was using the GPU on Stable Diffusion when i heard a click (every DAC that i tested click a lot when i'm using the GPU on my desktop so i made sure to also test with no GPU usage), in the end of the day i think this issue is overblow and i immediately canceled the refund process.

Thanks and sorry for my english!

This topic---audibility of the distortion in real audio content---is not straightforward, as mentioned in my review. So, it is ultimately up to you whether you keep your DAC or not. There are quite a few factors (headphone sensitivity and audio content) that determine its audibility in real use cases.

It is very possible that those cheap DAC/dongles rely on some noise reduction mechanism akin to dynamic range enhancement (DRE) in CS431xx and hence are affected by similar distortion. If you really want to avoid it, whether you hear the distortion or not, you want to spend more and get ESS- or AKM-based devices, as listed in the review.

 

[KDRN]

As time has passed and more tests have been done, I feel the need to update my take on the audibility issue of this Cirrus hump distortion. To avoid a verbose description of this topic, I put this content here in a separate post.

In my review I initially stated it is difficult to determine its audibility in real audio content. True, but at the same time I wanted to be conservative since there should be many people who own those affected devices. Another reason is that it is impossible to draw an exact percentage of the time it is audible out of infinite possibilities. But it IS audible. There's a reason why Roman at RAA used a clip from the Dune soundtrack. I found that the Cirrus hump distortion is most noticeable when audio consists of somewhat complex low-frequency signals with not much high-frequency contents. The Dune soundtrack clip is such an example. And an ideal example is the C Major test signal used in Part II of my review. To hear, simply play the original (not recorded) signal at -14 to -15 dBFS through an affected device and listen to it using an IEM with suitable sensitivity. I used the 7 Hz x Crinacle Zero:2. The distortion is very, very clearly audible.

Below are the spectrograms of the recordings of the Dune soundtrack and C Major clips, respectively, played from the KA15:
View attachment 461330

View attachment 461329

Distortions in both cases are clearly audible through the Zero 2 IEM.

How about the JCally JM20 which does not produce the Cirrus hump from steady signals but still distorts when a signal begins (as shown in Part II of my review)?
View attachment 461333

View attachment 461332

Although less often, these distortions are still audible from the Zero 2.

with black pearl with NOS filter. i dont hear the clicking at all