Is the gain change of the Shanling UA1 Plus (CS43131) in the digital or analog domain?

[zany.ninja]

Using the Shanling UA1 Plus (dual CS43131, no discrete op-amp, 1.6V @32 Ohm, 80mW), I can seamlessly switch between high and low gain using the companion app during playback - with no pauses, pops or anything. I was wondering if this gain switch is likely to be a toggle for an analog gain stage change, or would it be doing some sort of digital volume remapping / PCM value scaling?

I was wondering because when I try to listen at the same volume levels across both gain modes, I do get a sense of lower/compressed dynamic range in low gain mode.

I suppose this could be due to the lower max output level in low gain, but I'm wondering if digital attenuation might also be involved, because I'm not using a particularly hard to drive headphone - just a 60 Ohm Koss Porta Pro/KPH30i, for which low gain should be sufficient.

Thanks in advance!

 

[zany.ninja]

@Frankie-Shanling perhaps you could confirm please? Thanks.

 

[phofman]

IIUC CS43131 has several features to control power of its output hadphone class-H amps (HV_EN, OUT_FS). Taking into account the small size of the device, I would expect that app setting just modifies those CS43131 registers. IIUC it would be analog power control (changing supply voltage of the output power stage) https://statics.cirrus.com/pubs/proDatasheet/CS43131_DS1155F2.pdf

 

[MCH]

OP mentions it doesnt have op amps, but in any case, for reference, the meizu hifi pro does have an analog switch that can be used to bypass, or not, the opamp output stage. I dont know what is the exact purpose, just wanted to bring up that a switch at the output of a tiny CS43131 dongle would not be unheard of.
See a teardown of the non pro (same device with the op amp and switch not populated) here:

Teardown and (maybe) mod of Meizu Hifi Non-Pro CS43131 dongle DAC

Some background: These last weeks i have been thinking on building a headphone amplifier as part of a larger project (USB interface). I thought that i could use the CS43131 inside my Meizu Hifi Non Pro. But i wanted to do it with output opamps to have more power available, as this is going to be...

www.audiosciencereview.com

 

[antcollinet]

I do get a sense of lower/compressed dynamic range in low gain mode.

I would guess it is almost certainly digital gain control.

But what I can tell you is that digital volume change does not change the dynamic range of the sound in the way you are desribing. There is no "compression" (compression. being reducing the level of quieter sounds less than that of louder sounds)

Dynamic range is reduced - but only in terms of reducing the signal to noise ratio. That is only a problem if you are hearing actual digital noise (something like tape hiss) at your normal listening level.

Obviously if the low gain settings result in the maximum volume level being lower than your normal listening level, then that can result in the sound feeling less dynamc. But that is just a result of lower volume, regardless of that being from analogue or digital control.

 

[antcollinet]

IIUC it would be analog power control (changing supply voltage of the output power stage)

But (if I am understanding correctly) that wouldn't result in a gain change - just a limit to maximum output voltage/power before clipping?

 

[phofman]

IMO a screenshot of the actual app showing what is being changed would be handy. Nevertheless IMO this looks like output gain (output full-scale voltage):

 

[zany.ninja]

I would guess it is almost certainly digital gain control.

But what I can tell you is that digital volume change does not change the dynamic range of the sound in the way you are desribing. There is no "compression" (compression. being reducing the level of quieter sounds less than that of louder sounds)

Dynamic range is reduced - but only in terms of reducing the signal to noise ratio. That is only a problem if you are hearing actual digital noise (something like tape hiss) at your normal listening level.

Obviously if the low gain settings result in the maximum volume level being lower than your normal listening level, then that can result in the sound feeling less dynamc. But that is just a result of lower volume, regardless of that being from analogue or digital control.

In high gain, playback with my 60 Ohm Koss has quite a bit more crunch and dynamism, better imaging and soundstage, etc. compared to low gain (after matching the volume levels between the 2 gain modes as best as I can).

But the difference is not as marked between the 2 gain modes (barely any actually) when I use my 28 Ohm Moondrop IEMs, which are a fair bit more sensitive. So I suppose that indicates analog gain? @Frankie-Shanling liked @phofman 's first reply too, so I guess that confirms it.

IMO a screenshot of the actual app showing what is being changed would be handy.

Not sure I understand. You want a screenshot of the mobile app? It's just Shanling's Eddict Player app, with a switch labelled Gain, and you toggle between High or Low. It's not very informative.

---

I asked my original question because low gain + 40% volume is a comfortable listening volume for me, and because the general advice is that low gain is preferred for better fidelity - if you get sufficient volume that is. But that was proving not to be the case for me - high gain (30% volume) was noticably better, and I was curious why. I speculated it could be because low gain was digital gain, and crushing the dynamic range, but it seems it isn't. I guess my headphones benefitted from the extra power of high gain after all, even though they are easy to drive, and counter to the prevailing wisdom out there.

I guess that's that. Thanks to everyone for all your feedback!

 

[antcollinet]

(after matching the volume levels between the 2 gain modes as best as I can).

This is almost certainly the reason for the perceived differences. "As best I can" is not good enough to eliminate perception of different sound quality caused by small level differences that are not perceived as a volume difference.

Level matching needs to be accurate to <1% (0.1dB), and can normally only be achieved by measuring the output of the amp with a voltmeter while playing a fixed test tone.

It is particularly problematic when you have a digitally controlled volume, with 0.5dB steps - which are not fine enough for the level of accuracy needed.


I'm also intrigued how you are changing your volume between 30% and 40% with this device - because as far as I can see you can only to that on the devices the DAC is plugged into. In which case that is also digital volume control.

If I am correct, then there is literally ZERO difference in dynamic range between low digital gain and high digital volume, vs high digital gain, and low digital volume - if the level achieved is the same.

 

[zany.ninja]

This is almost certainly the reason for the perceived differences. "As best I can" is not good enough to eliminate perception of different sound quality caused by small level differences that are not perceived as a volume difference.

You might be right. But I did try to take this factor into account and during some tests deliberately listened at a significantly lower volume on high gain than it was with low gain. Even at a lower listening level, on high gain everything felt more expansive and lively.

But like you said, I could be wrong. My tests are hardly scientific. I have to do it by ear since I don't have any measuring equipment to do it more accurately.

I'm also intrigued how you are changing your volume between 30% and 40% with this device - because as far as I can see you can only to that on the devices the DAC is plugged into. In which case that is also digital volume control.

For the tests, I kept the device volume at max, and change the DAC volume the same way I toggle between high and low gain - using the app. All on Android btw.

For general listening I prefer my iPad (100% volume) with DAC at 30%.

 

[phofman]

IMO that app option toggles between two hard-coded settings of the OUT_FS register for the CS43131-based Shanling devices.

 

[zany.ninja]

IMO that app option toggles between two hard-coded settings of the OUT_FS register for the CS43131-based Shanling devices.

I'm not so sure. As I understand it, OUT_FS can only be changed when the amp is powered down with PDN_HP (to avoid pops and clicks). The fact that I can toggle between high and low gain on the fly, with no pause, pop, or click, suggests to me that OUT_FS is not being altered by the Shanling app gain setting.

 

[phofman]

Well then there is the 1dB_EN setting which is just 5mW -> 6.8mW. Apart of that, I see only digital volume control (either in the app or in the chip, equivalent).

 

[antcollinet]

For the tests, I kept the device volume at max, and change the DAC volume the same way I toggle between high and low gain - using the app. All on Android btw.

For general listening I prefer my iPad (100% volume) with DAC at 30%.

So, like I say both the gain and volume control operate as a digital multiplication of the sample values. It makes literally no difference if it is 30%x40%xSample Value, or 40%x30%xSample value (or whatever the gain/volume values are)
It is the same thing.

So if it is not a level shift causing your perception - it is perceptive bias since you are not comparing blind.

 

[zany.ninja]

So, like I say both the gain and volume control operate as a digital multiplication of the sample values. It makes literally no difference if it is 30%x40%xSample Value, or 40%x30%xSample value (or whatever the gain/volume values are)
It is the same thing.

So if it is not a level shift causing your perception - it is perceptive bias since you are not comparing blind.

You could be right.

But I guess it's best I stop speculating, as that way madness lies. The only way to confirm one way or another is to do proper tests and measurements.

Thanks to everyone for your feedback!