-
Welcome to ASR. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!
On the Distortion of Cirrus Logic CS431xx-Based Devices: A Comparative Review
- Thread starter jkim
- Start date
Where are these measurements?
Hello. I registered with ASR because I wanted to report this issue.
Fiio KA5 has received firmware version 1.70, which includes a "DRE optimization function," resolving previous DRE issues.
However, a crackling noise very similar to the DRE problem has now appeared at low volumes.
This new noise is audible when playing "Original_CMaj.flac" at low volumes of approximately Vol 25-45 (High gain/4.4mm balanced output).
It also occurs when increasing the hardware volume and lowering the software volume.
Changing the digital filter to "NOS" completely eliminates this noise.
This indicates that this noise is related to the DRE.
Since the KA5 does not have a function to turn the "DRE optimization function" on or off, I decided to downgrade the firmware to v1.60 to circumvent this problem.
The same noise is also audible on the Tanchjim SPACE, which also has the "DRE optimization function".
I contacted Tanchjim support, but they didn't acknowledge this issue.
It seems to me that the "DRE optimization feature" offered by Fiio and Tanchjim doesn't actually resolve the DRE-related noise, but rather makes it more audible.
(Vol25-45 are the volume levels I normally use.)
Unfortunately, I don't own any other models that claim to address the DRE issue, so I can't do any further testing.
Have any of you noticed anything if you own those models?
Fiio KA5 has received firmware version 1.70, which includes a "DRE optimization function," resolving previous DRE issues.
However, a crackling noise very similar to the DRE problem has now appeared at low volumes.
This new noise is audible when playing "Original_CMaj.flac" at low volumes of approximately Vol 25-45 (High gain/4.4mm balanced output).
It also occurs when increasing the hardware volume and lowering the software volume.
Changing the digital filter to "NOS" completely eliminates this noise.
This indicates that this noise is related to the DRE.
Since the KA5 does not have a function to turn the "DRE optimization function" on or off, I decided to downgrade the firmware to v1.60 to circumvent this problem.
The same noise is also audible on the Tanchjim SPACE, which also has the "DRE optimization function".
I contacted Tanchjim support, but they didn't acknowledge this issue.
It seems to me that the "DRE optimization feature" offered by Fiio and Tanchjim doesn't actually resolve the DRE-related noise, but rather makes it more audible.
(Vol25-45 are the volume levels I normally use.)
Unfortunately, I don't own any other models that claim to address the DRE issue, so I can't do any further testing.
Have any of you noticed anything if you own those models?
I don’t remember if @jkim showed the actual measurements in some post, but he did check this—see this note on the first post:
Note. This distortion behavior does not depend on the device's playback sample rate, impedance load, or any firmware settings like digital filters, Class H/AB modes, etc. (if available)---it always occurs regardless of such conditions. The Fiio BTR13 was not included here due to its 16-bit resolution masking distortion to some extent (instead, see below for its TDFD bass distortion measurements).
One of the idiosyncrasies of the CS431xx is that the NOS setting also turns DRE off (Cirrus may have some good technical reasons to do that, I don’t know).
Some OEMs hinted that their “optimized DRE” setting acts like a noise gate, turning DRE on/off depending on the volume. I don’t know how FiiO & Tanchjim have implemented their DRE feature.
There are dongles with a 3-position settings: DRE on, “optimized”, and off.
I really searched hard but I couldn't find any.
Bamboszek
Active Member
- Joined
- Mar 17, 2018
- Messages
- 218
- Likes
- 405
Although the FiiO K11 is a desktop device, it is based on the CS43198 chip, commonly known from dongles.
Unfortunately, it appears to suffer from the "Cirrus DRE hump".
I'll try to ask FiiO if they could implement the ability to disable DRE in the next firmware. In the meanwhile, I'm using it with NOS filter workaround.
Measurements:
FiiO K11 24/96 Optical In - RCA Line Out
measured using MOTU UltraLite-mk4 (ESS version) with RCA to TR Jack adapters
512k FFT, 8x AVG
Blackman-Harris 7, 87.5% max overlap
Bandwidth: 20 Hz to 20 kHz
31 multitone TD+N vs level
Dual tone 41 & 89 Hz 1:1 -16 dBFS
16x AVG
Unfortunately, it appears to suffer from the "Cirrus DRE hump".
I'll try to ask FiiO if they could implement the ability to disable DRE in the next firmware. In the meanwhile, I'm using it with NOS filter workaround.
Measurements:
FiiO K11 24/96 Optical In - RCA Line Out
measured using MOTU UltraLite-mk4 (ESS version) with RCA to TR Jack adapters
512k FFT, 8x AVG
Blackman-Harris 7, 87.5% max overlap
Bandwidth: 20 Hz to 20 kHz
31 multitone TD+N vs level
Dual tone 41 & 89 Hz 1:1 -16 dBFS
16x AVG
Thank you for reporting. I've previously listed K11 as having the issue reported in my FiiO devices compilation post, however that was actually a misedit on my side - @IsaacOscar pointed out KA11 was reported and linked to the source post, and then I just edited K11 on my list instead of the KA11 by mistake. I however accept no blame for this - it is all on FiiO for using similar designations
I can't count how many people I saw completely mistaking K11 and KA11.Yeah, they likely just changed some parameters but the DRE function is still on. I was suspicious of the wording since they didn't mention a switch to turn the feature off. Thanks for confirming.
The measurements are what you see in the OP. The quote:
The result is simply unaffected by switching between Class H and Class AB.
If you come across a device that is affected, then you are welcome to post your results. The more we know the better.
Last edited:
pangcukaipang
Member
Maybe because completely turning off DRE would increase the noise floor and decrease the SNR/DNR and they don't wanna do that?Yeah, they likely just changed some parameters but the DRE function is still on. I was suspicious of the wording since they didn't mention a switch to turn the feature off. Thanks for confirming.
The measurements are what you see in the OP. The quote:
The result is simply unaffected by switching between Class H and Class AB.
If you come across a device that is affected, then you are welcome to post your results. The more we know the better.
Ibasso mentions that on Macchiato and DC04 Ultra user manual:
FiiO BTR13 now has the DRE setting! https://www.head-fi.org/threads/fii...r-is-officially-released.973995/post-19138482
My Luxsin X8 uses CS43198 (8 of them lol) and came with DRE disabled by default. Will be interesting to see how it measures… I wonder if they found some way to reduce the noise floor without using DRE by using 4 DACs per channel. I don’t think it’s the first octo DAC configuration, but I don’t know enough about DACs to know why one would use 8. 2 makes sense… one per channel, but 8? No idea.
Edit: AI says “The primary mathematical reason to run 4 DACs in parallel is to drastically lower the noise floor.” There were some other benefits too, but that’s the primary one.
Last edited:
See this thread: https://www.audiosciencereview.com/forum/index.php?threads/paralleling-dacs.2212/
In theory, paralleling two DACs cuts the (uncorrelated) noise in half, or -3dB.
reallystate
Member
- Joined
- Nov 6, 2025
- Messages
- 18
- Likes
- 28
I measured the new BTR13 firmware for Cirrus Hump and I found some surprising findings.
- Joined
- Mar 1, 2021
- Messages
- 3,097
- Likes
- 6,631
Typical Fiio software update then?, they can design pleasing looking devices that attract the eye but when it comes to software and device management/apps they are utterly woeful.
I had a BTR17 for a few days but all of the above was true so sent it back and I continue to use my Qudelix 5k (but would appreciate a refresh of the design/button placements
RadioRat
New Member
- Joined
- May 23, 2025
- Messages
- 2
- Likes
- 0
Has anyone recently tested the Jcally JM20 and JM20 Max with the artifacting test? I've heard so many conflicting reports.
I'm glad they've improved the humps, but I'm very sensitive to the artifacting issues present in many of these dongles. Ideally I'd want the JM20 since I don't need the extra power. But I'll go with whichever model is artifact free, if either is.
I'm glad they've improved the humps, but I'm very sensitive to the artifacting issues present in many of these dongles. Ideally I'd want the JM20 since I don't need the extra power. But I'll go with whichever model is artifact free, if either is.
That's unfortunate :<
Edit 1:
I actually tried the CMaj test audio from the main post with my Truthear Pure at a reasonable volume with APTX Adaptive. I can hear like very obvious plastic crumpling when DRE is on (kinda like putting an object on top of a bass speaker as it plays and it creates a vibrating sound), but it's reduced when DRE is off. Not sure how it was in the former firmware - I will have to test when I get the chance.
Edit 2:
I also tested the IEM with the laptop's headphone jack - no such crunchy sound can be heard at a similar volume.
Edit 3:
Tested with the previous firmware, it's actually better than new firmware + DRE on. But it does have some audible crunchy bits, arguably the same as with the new firmware + DRE off.
Last edited:
Similar threads
- Poll
