iPhone vs. Android via USB-C sound quality difference is huge

[lindijones]

It might be an impedance issue if you are taking analog line out using the camera adapter. IEMs can have perverse amping requirements.

I’m positive that apple devices don’t alter the original digital signal unless you turn on the sound check, EQ or enhancer functions. Apple uses its own DACs, which measure stone flat and clean to far beyond audible levels. Every app calls on the same sound decoding. If something else sounds different when you’ve got all the signal processing turned off, it isn’t the Apple product that is colored. The coloration would have to come from the other device.

Did you do level matching? Level imbalance could be the reason you’re hearing one as “better” than the other.

What kind of impedance do you mean?
The setup with the Dusk IEMs connected to the DSP cable is unchanged.
We simply connect the USC-C DAC dongle to different devices.
That‘s it.

 

[gregorio]

The frequency response of the 9$ Apple dongle measures nicely (kind of).

I’m not sure what you mean by “kind of”?

But it still sounds crappy.

If it measures “nicely”, then the sound must also “nice”. Of course, someone might not like “nice” sound and actually prefer sound reproduction that’s less “nice”. Although in some cases it may really not be so “nice”, if for example you’re using HPs that require more power than the dongle outputs.

The setup with the Dusk IEMs connected to the DSP cable is unchanged.
We simply connect the USC-C DAC dongle to different devices.

You don’t know if the setup is unchanged. How do you know if the DSP cable is applying some DSP when connected to a device with its own software and then reverting to a different DSP setting when connected to an iPhone? It is after all a DSP cable/dongle, so the only way we can know what it’s doing is if its software tells us or the manufacturer does.

G

 

[bigshot]

The Apple dongle has a high degree of fidelity- it sounds flat and clean. That may not be what you want, and if that’s the case it will sound crappy to you. But there’s nothing wrong with the dongle. It does the job it’s intended to do perfectly.

 

[lindijones]

The Apple dongle has a high degree of fidelity- it sounds flat and clean. That may not be what you want, and if that’s the case it will sound crappy to you. But there’s nothing wrong with the dongle. It does the job it’s intended to do perfectly.

The Apple dongle is ok-ish for 9$.
All the tons of other dongles also measure flat but sound way better.

Mostly they are just so much more powerful, precise and more stable which leads to better dynamics.
That‘s no question.

 

[castleofargh]

Apple getting in the way of compatibility for non apple products is almost always the explanation for the issues people have. At this point it's not a track record, it's a rule.

 

[2leftears]

The Apple dongle is arguable their product with the best performance/price ratio. I rarely have a need to use it but to me it sounds fine. I have more of a problem with the fragile cable on it and Apple's continued obstinance in learning what a properly designed strain relief looks like, but there is nothing wrong with the sound.

Apple getting in the way of compatibility for non apple products is almost always the explanation for the issues people have. At this point it's not a track record, it's a rule.

I do like their MacOS and so far I have stuck with their architecture, but man they sometimes make some boneheaded choices when designing their new products and/or software.

 

[OhGosh]

You don’t know if the setup is unchanged. How do you know if the DSP cable is applying some DSP when connected to a device with its own software and then reverting to a different DSP setting when connected to an iPhone? It is after all a DSP cable/dongle, so the only way we can know what it’s doing is if its software tells us or the manufacturer does.

Just to explain: there's no software required to use the DSP cable. The Android App is used to change the EQ settings that go on the cable and apply them. After that the cable does it's own thing regardless of where you plug them, no software needed.

All it needs is for the smartphone/computer/music player to send a digital signal of the music through it's USB C port, which usually happens automatically.

Only device I know for a fact that was sending analog data thought the USB C port instead of Digital is my Steam Deck (I use this IEM for games too, it's pretty good).

There, I had to go to audio settings and change it to send digital instead of analog. The difference is immediately noticeable, as the analog signal seemed to not be using the cable's EQ. Once it changed to digital it was working just fine.

 

[bigshot]

The Apple dongle is ok-ish for 9$.
All the tons of other dongles also measure flat but sound way better.

Mostly they are just so much more powerful, precise and more stable which leads to better dynamics.
That‘s no question.

No problem with dynamics at all. Flat and clean is what DACs aim for. The Apple dongle measures and sounds as good as many high end DACs. If you think it sounds inferior, it’s probably expectation bias. I don’t think you know what you’re talking about. Do you have an iPhone?

 

[bigshot]

Apple getting in the way of compatibility for non apple products is almost always the explanation for the issues people have. At this point it's not a track record, it's a rule.

You mean “non-Apple certified”. Apple stuff works with plenty of third party products, just not ones that haven’t been specifically designed to work with Apple.

 

[2leftears]

Apple getting in the way of compatibility for non apple products is almost always the explanation for the issues people have. At this point it's not a track record, it's a rule.

You mean “non-Apple certified”. Apple stuff works with plenty of third party products, just not ones that haven’t been specifically designed to work with Apple.

Spoiler: Off-Topic

Apple does get in the way though, although mostly unintentionally, and mostly by changing their architecture too often, such that third party providers give up on supporting Apple.

E.g. Apple's computer architecture:

• 1984 - Motorola 68000
• 1994 - PowerPC
• 2006 - Intel x86
• 2020 - Apple Silicon
• Between 2000 and 2024 there were (about) 22 major OSX & MacOS releases, on average 1 major new release every 1.1 years

Compare that to Microsoft Windows:

• 1978 to date - Intel x86 architecture
• Between 1985 and 2024 there were (about) 11 major Windows releases, on average 1 major release every 3.5 years

Given the much higher market share of Microsoft Windows, and the much more stable* architecture & OS environment, it is hardly any wonder that many third party suppliers look at Apple and think: sod that; we are not going to support that. In terms of R&D, the 'cost of carry' for third party providers supporting Apple is far higher.

This battle is currently playing out in the arena of printer drivers after the Intel x86 to Apple Silicon switch, with more and more printer manufacturers saying: sod this, no more native driver support for Apple, we'll simply allow Apple to provide basic AirPrint support for our printers (even though it totally sucks at colour management and support for many other print features).

Disclosure: I use Apple at home because on balance I do prefer the user experience (I have used both Apple and Ms Windows extensively over the years).

* stable as in fewer major releases. The stability of the OS itself is a different matter entirely.

 

[lindijones]

Just to explain: there's no software required to use the DSP cable. The Android App is used to change the EQ settings that go on the cable and apply them. After that the cable does it's own thing regardless of where you plug them, no software needed.

All it needs is for the smartphone/computer/music player to send a digital signal of the music through it's USB C port, which usually happens automatically.

Exactly. That. Thx.

Only device I know for a fact that was sending analog data thought the USB C port instead of Digital is my Steam Deck (I use this IEM for games too, it's pretty good).

There, I had to go to audio settings and change it to send digital instead of analog. The difference is immediately noticeable, as the analog signal seemed to not be using the cable's EQ. Once it changed to digital it was working just fine.

You can send analog audio through USB-C? Interesting. Never heard about it. Do you have more information on that? What standard is it?

 

[lindijones]

No problem with dynamics at all. Flat and clean is what DACs aim for.

Exactly. Depending on the load used during the measurement, the frequency response might be flat. (Which is of course the aim.)

The Apple dongle measures and sounds as good as many high end DACs.

Don't agree. Sorry.
If you are satisfied with the Apple dongle, it's good for you. Enjoy!

If you think it sounds inferior, it’s probably expectation bias.

So you think everyone on this planet using a different dongle than the Apple 9$ dongle has expectation bias? Interesting.

I don’t think you know what you’re talking about.

That's ok. I can live with that.

Do you have an iPhone?

I don't understand this question. Did you read the initial post?

 

[XTF1]

Just to explain: there's no software required to use the DSP cable. The Android App is used to change the EQ settings that go on the cable and apply them. After that the cable does it's own thing regardless of where you plug them, no software needed.

Don’t want to go too deep in a rabbit hole…

Although these DSP-cables offer what is considered “HW-based” PEQ, they are often built around some type of ARM microcontroller and firmware PEQ filters. So, it’s actually a low-level SW-based PEQ (libraries), implementing biquads filters or something similar onto the onboard MCU.

If you dig into the biquad filter coefficients calculation, you’ll find that the sampling frequency (Fs) of the signal is a parameter. So in theory, these DSP-cables chips should recalculate each one of the biquads coefficients (4 or 5 per filter) whenever the sampling frequency of the signal changes. In my experience, they don’t…. And when you set a filter in the Android app for example, the biquad coefficient are calculated (app or onboard chip) assuming a fixed Fs—usually 48kHz.

What it means is that the actual EQ’ed response changes depending upon the sampling frequency of the signal. It’s really not that bad, as it is mostly noticeable above ~10kHz… where you should not EQ anyway!

Now, standard Android resamples everything to 16-bit/48kHz. Android specialized app (UAPP, Neutron, HiBy Music…) have options to not resample. Apple iPhone… may do its own thing.

So, if you want to compare the source impact on these DSP cables, disable all your PEQs first (all 0-gain).

And if we really want to compare Apple vs. Android sound via USB… a DSP-enabled cable is a bad choice: too many uncontrolled variables.

 

[OhGosh]

I do like their MacOS and so far I have stuck with their architecture, but man they sometimes make some boneheaded choices when designing their new products and/or software.

Yup Mac OS is pretty good, especially if it's a machine for a job/production. Super clean, never gets in my way. The thing is, some of the things some people like about apple are usually what is bothersome to others, like the privacy aspect of everything requiring access permissions, among other things.

I used to prefer Windows but that changed a long, long time ago. If I had to stop using Mac for whatever reason I'd probably start working with some Linux distro.

Don’t want to go too deep in a rabbit hole…

Although these DSP-cables offer what is considered “HW-based” PEQ, they are often built around some type of ARM microcontroller and firmware PEQ filters. So, it’s actually a low-level SW-based PEQ (libraries), implementing biquads filters or something similar onto the onboard MCU.

If you dig into the biquad filter coefficients calculation, you’ll find that the sampling frequency (Fs) of the signal is a parameter. So in theory, these DSP-cables chips should recalculate each one of the biquads coefficients (4 or 5 per filter) whenever the sampling frequency of the signal changes. In my experience, they don’t…. And when you set a filter in the Android app for example, the biquad coefficient are calculated (app or onboard chip) assuming a fixed Fs—usually 48kHz.

What it means is that the actual EQ’ed response changes depending upon the sampling frequency of the signal. It’s really not that bad, as it is mostly noticeable above ~10kHz… where you should not EQ anyway!

Now, standard Android resamples everything to 16-bit/48kHz. Android specialized app (UAPP, Neutron, HiBy Music…) have options to not resample. Apple iPhone… may do its own thing.

So, if you want to compare the source impact on these DSP cables, disable all your PEQs first (all 0-gain).

And if we really want to compare Apple vs. Android sound via USB… a DSP-enabled cable is a bad choice: too many uncontrolled variables.

Thanks for the thorough answer! I didn't really know how the DSP worked. All I meant was that there's no need for an app or something to make the DSP work, but I didn't know that the DSP might have it's own firmware. I figured you might have been unfamiliar with it but you're way more familiar than I am XD.

I believe whatever is happening to create a difference is probably related to what you just said. Apple isn't always super transparent about how it does these things. If anyone is aware of how iPhone resamples the sound or if does anything different, I'd be curious to know.

 

[OhGosh]

Exactly. That. Thx.


You can send analog audio through USB-C? Interesting. Never heard about it. Do you have more information on that? What standard is it?

Unfortunately I don't, not about the Deck's audio standards . When you go into desktop mode, you can choose to activate a few audio drivers/boards. When you want to connect to a TV for example, you must activate a particular audio board that is usually inactive in order to make sound go through HDMI.

On this same panel is would allow me to pick between 6 different audio options for the Moondrop May, when a specific drive/board is being used: Analog + mono input (for the mic on the may cable), analog, digital + mono input, digital, surround + mono input and surround. I can send a few screenshots and check out the actual names of the boards and drivers if you're really interested, I can't really remember them right now.

When I tried Surround it very much didn't work XD, the audio went all crazy. It's probably because it's actual surround and not software surround made for headphones (or that's what I guess). Analog and Digital had a big, instantly noticeable difference between them. Still sounds super good on Analog to me though, as I really like how these IEMs sound even on analog cables, but especially on the top end there's a big difference when using digital, as is usually the case when using the DSP cable. So right now I leave it on digital + mono input.