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

[jkim]

I think more manufacturers than we assume do actually read and learn from this remarkable thread.
Sadly, as usual with almost all of these brands, all that happens in a rather conspirative fashion. Reading and more careful reading, monitoring what is typed in here and then discussing (your) assessments internally. But never ever contributing (despite asking for some cheerleading "comment what you like most about xyz" prior to releasing new products).
Not offering solutions for users who bought their (in numerous cases fixable) products within the last one or two years. Cirrus Logic first and foremost should take up a stance on that.
Just for example, Fosi Audio with it's latest dongle DAC release, the DS3 now implementing an ES9039Q2M chipset, walking away from Cirrus Logic, too. Does that all leave a bad taste when it comes to interacting honestly? ...IMO sure it does.

Thanks for the comment.

In fact, I recently realized that the BEST solution to the DRE issue of CS431xx is NOT just disabling DRE.

Just as the registers related to DRE (e.g., DRE on/off, DRE activation threshold dB) are hidden to the public, there MUST be hidden registers controlling analog attenuation on CS431xx. Because, without analog attenuation, DRE cannot work in the first place.

Combining digital and analog attenuation in user volume control, rather than signal-dependent control of DRE, it could achieve state-of-the-art noise performance at low levels. It would be on par with, or even better than, the ES9219 in Qudelix 5K. Given CS431xx's really low distortion, it would be amazing that way.

Then, why did Cirrus Logic not show these parameters in the chip's datasheet? Again, they most likely wanted to force the stupid DRE to be its default setting. They succeeded. As we know, no product with CS431xx has disabled DRE until very recently.

"DRE off" combined with the chip's analog attenuation integrated in a device's volume control will make a heck of a portable DAC at an affordable price. But could anyone think of this? I doubt it...

 

[HissingFree]

I think more manufacturers than we assume do actually read and learn from this remarkable thread.
Sadly, as usual with almost all of these brands, all that happens in a rather conspirative fashion. Reading and more careful reading, monitoring what is typed in here and then discussing (your) assessments internally. But never ever contributing (despite asking for some cheerleading "comment what you like most about xyz" prior to releasing new products).
Not offering solutions for users who bought their (in numerous cases fixable) products within the last one or two years. Cirrus Logic first and foremost should take up a stance on that.
Just for example, Fosi Audio with it's latest dongle DAC release, the DS3 now implementing an ES9039Q2M chipset, walking away from Cirrus Logic, too. Does that all leave a bad taste when it comes to interacting honestly? ...IMO sure it does.

I mean, it is simple, this issue is not known by the general public, so they are barely going to lose money over this issue. In general, you are not getting a better device for a similar price, just the DS3 you linked is like double the price of DS2, just to give an example.
CS43131 is such a successful chip for the masses at the price point they sell it, in terms of the specifications they reach and in terms of the general listening experience you get out of it, that it's just not worth it to do tweaks to it or spend more resources on it. They will probably introduce some of these changes in their next chip, for example a CS43232 or something like that.

 

[kuba.pl]

I admire you for your dedication and work on adapters.
I am asking about these models because they are widely available and relatively cheap in stores in Poland, where I come from.
I received this CS43198+CB5100 adapter (sometimes there is an error in the description CS4390, missing “1”).
I see that they are quite popular in stores, they look the same but are sold under different brand names.

 

[solawind]

Can you guys suggest any other test tracks besides https://reference-audio-analyzer.pro/en/demo-cs131.php ? I just got a Moondrop FreeDSP Gen 2 cable adapter , and no one seems to have checked it yet, so I can A/B it with the Sabre and Apple dongle.

 

[damex]

hi, i have aune n7d that has dual CS43198 (optional digital usb input).

https://reference-audio-analyzer.pro/test/cs/Song-of-the-Sisters.flac used this file with xba-n3 which should be sensitive enough for this test.
there are two occurances where it occurs when playing on low gain + high current mode on aune n7d:
- (normal volume with xba-n3 is around 3-4 clicks and anything around 7-8 is deafening levels) if i bump it to 18 (!) clicks (beyond deafening levels for regular audio but not for test file?)- it crackling around 2/3 of test track (tried wddm and asio) but that could be iems themself (are they?)
- this crackling happens so far only on this track and only when i adjust volume during playback. cracklig sound per let's say volume wheel 'click' (it is stepped volume wheel)

anything better to test with?
i have motu m6. can i do some test loop through motu m6 line in?

 

[SlowLife]

The firmware version V1.1 for HiBy FC5 with Display is available!

Changelog:
1. Added DRE switch function;
2. Added NOS option;
3. Optimized UI display;
4. Fixed several other bugs.

Connect the FC5 (with display) to your computer, download and install the upgrade tool [HiBy USB DFU Tool].
After installation, choose to upgrade locally or update online in the upgrade tool.

DFU Tool: https://www.hiby.app/usbdfu/en/
Firmware file: https://drive.google.com/drive/folders/1Xoa2kiFAQRXDXXAXWeOi6LsGsxo1fNJ5

 

[KDRN]

The Ibasso jr macchiato has a DRE switch

 

[CedarX]

The Ibasso jr macchiato has a DRE switch
View attachment 501830

The app actually gives you 3 settings: on, off, and optimum/optimized, whatever they did under the hood. Actually, based on the other features (settings, PEQ…), and the fact it is recognized by WalkPlay, it is another—very nice—TTGK-based dongle, customized for iBasso Jr.

 

[KDRN]

The app actually gives you 3 settings: on, off, and optimum/optimized, whatever they did under the hood. Actually, based on the other features (settings, PEQ…), and the fact it is recognized by WalkPlay, it is another—very nice—TTGK-based dongle, customized for iBasso Jr.

and in the marketing it could do 500mw+ shows it has opamps in it. and if its true this is a ttgk made product with walkplay. this is basically a luxury(better) version of the cinear protocol max(epz tp55)

 

[KDRN]

"DRE off" combined with the chip's analog attenuation integrated in a device's volume control will make a heck of a portable DAC at an affordable price. But could anyone think of this? I doubt it...

utilizing the analog attenuation feature of cirrus logic chips for "independent volume" would prolly give it some benefit.. maybe it could be easily implemented through the firmware. the digital audio stack is always at 0db and the hardware volume control s using the analog attenuation feature those cirrus logic chips have.

maybe we could annoy ask ttgk/walkplay to implement something like that for the black pearl. instead of a digital volume control. (that could probably finally fix their bug of the volume steps resetting when applying PEQ)

 

[CedarX]

and in the marketing it could do 500mw+ shows it has opamps in it. and if its true this is a ttgk made product with walkplay. this is basically a luxury(better) version of the cinear protocol max(epz tp55)

Per TTGK data, it uses a TT39518F01-Pro module (VID 0x661, PID 0x881). It is definitely customized for iBasso and the partial WalkPlay support indicates it’s not intended to be used with WalkPlay: for example, the iBasso app (Android) gives access to 10x PEQ of PK, LS, HS, LP, HP, BP, and AP filter types, whereas WalkPlay (WebApp) only gave me PK, LS, HS, LP, HP PEQ filter types, and no DRE settings.

EDIT: For @mc.god , iBasso Android App with "Via dei Coronari" background instead of the default Anime

 

[KDRN]

i guess recent cirrus logic based ibasso dac/amps releases are now shipping with DRE switches. even the DC04 Ultra comes with one

 

[CedarX]

i guess recent cirrus logic based ibasso dac/amps releases are now shipping with DRE switches. even the DC04 Ultra comes with one

View attachment 502081

Yes, thanks to @jkim, OEMs took notice... FiiO was the first one to react, and is progressively implementing a DRE switch on all their CS431xx devices (Snowsky Retro Nano just got a FW update), HiBy is doing the same (FC5 is the latest), now iBasso...
What I am not sure is whether OEMs using CS431xx chips for desktop DACs are also doing the same—granted CS chips are not as popular for desktop as they are for portable stuff.

 

[SoySalsaPicante]

Yes, thanks to @jkim, OEMs took notice... FiiO was the first one to react, and is progressively implementing a DRE switch on all their CS431xx devices (Snowsky Retro Nano just got a FW update), HiBy is doing the same (FC5 is the latest), now iBasso...
What I am not sure is whether OEMs using CS431xx chips for desktop DACs are also doing the same—granted CS chips are not as popular for desktop as they are for portable stuff.

Topping D70 Pro Octo received a firmware update a while back allowing DRE Disable.

 

[Ahmonge]

Topping D70 Pro Octo received a firmware update a while back allowing DRE Disable.

However, the discontinued Topping D30 Pro. did not. In my opinion, this is poor customer service for a model that was launched only five years ago.

 

[mc.god]

EDIT: For @mc.god , iBasso Android App with "Via dei Coronari" background instead of the default Anime
View attachment 502082

I suppose it is intended to couple with the italian word "macchiato" but honestly it sounds like a meme to me, in the end I don't really understand why you would want a background picture at all, resulting in more confusing and less usable interface

 

[Saponetto]

However, the discontinued Topping D30 Pro. did not. In my opinion, this is poor customer service for a model that was launched only five years ago.

Curious.
In Amir's detailed review I honestly did not read any of tremendous amounts of distortion or measurable humps, and just now I am pleasantly listening jazz from my D30 Pro without any hassle.
Is this a -real- problem or it's just some Cirrus implementation in dongles that ended not work as expected and needs any DRE control to patch?

 

[MaxwellsEq]

Curious.
In Amir's detailed review I honestly did not read any of tremendous amounts of distortion or measurable humps, and just now I am pleasantly listening jazz from my D30 Pro without any hassle.
Is this a -real- problem or it's just some Cirrus implementation in dongles that ended not work as expected and needs any DRE control to patch?

Amir's test predates the discovery of this "dynamic" and quite audible distortion.

I'm not going be cynical about how this design "error" crept in to this generation of Cirrus Logic chips, but if you don't measure what happens to low frequency tones as the input signal is reduced, these chips seem to produce superb results, with "inaudible " distortion. If you disable this "capability", their performance is just so-so, but still better than our hearing. So, the "feature" is not necessary, but makes some measurements appear excellent. Unfortunately as the signal level traverses the switching point it creates a splat, which is audible, with quieter low frequency content.

 

[Saponetto]

@MaxwellsEq
Thanks, now it's much more clear to me.

 

[Ahmonge]

Curious.
In Amir's detailed review I honestly did not read any of tremendous amounts of distortion or measurable humps, and just now I am pleasantly listening jazz from my D30 Pro without any hassle.
Is this a -real- problem or it's just some Cirrus implementation in dongles that ended not work as expected and needs any DRE control to patch?

You can listen to the test tracks included in previous messages in this thread and judge the result for yourself.