• 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!

FIIO KA15 Portable DAC & Headphone Amp Review

Rate this DAC & HP amp:

  • Poor

    Votes: 1 2.9%
  • Not terrible

    Votes: 3 8.8%
  • Fine

    Votes: 15 44.1%
  • Great

    Votes: 15 44.1%

  • Total voters
    34
Awesome thanks! (So only the model needs changing; it's annoying you have to do that).
No problem, glad top help. I hope no new issues get discovered, it seems the relatively low price on this dongle DAC comes at a high expense of poor design choices.
 
No problem, glad top help. I hope no new issues get discovered, it seems the relatively low price on this dongle DAC comes at a high expense of poor design choices.
Most of the problems I've had are with their Fiio Control software, which is apparently used by most of their digital products, including the K19 (which is about A$1,700 at the cheapest)
 
Random thought:

Seems like the ka15 supports a microphone input but fiio themselves state that it doesn’t work with all operating systems.

I was having a curious idea where I could try measuring it to see how good the ADC was, but on my Mac it seems like the mic input is not recognized. Not sure about windows but maybe it only works for mobile devices?
 
FiiO KA15, the subject of this thread... I don't have it around, but I suspect it won't be any different, perhaps glitching under a different set of conditions involving these HS filters.
In the meantime, it's probably wise to proceed with caution when using High-Shelf filters on these SPV chips... or stick with Peak filters!
I just noticed you crossed this part out. Has someone confirmed whether or not the bug occurs on the KA15?
 
I have a KA15 for 2 weeks now (Latest firmware) used only with my phone as portable.It has none of the problems some mention here.I don't like the volume control though both modes are not good,there should have been a menu option to switch the +- buttons to just volume up down.After a while you get used to it but its not convenient. Soundwise no problems ,i listen with HD25 and had not tried with any other headphone. Desktop mode is on as with the HD25 volume reaches the end very easily with out it on.I hear basically no big difference between the various filters.Software control works very good on Android but not the web app,which is not very good for me ,not very well made.I don't use EQ anyway since i have a thing to always listen to sound as it was recorded,at least in portable situations.The built quality is very good but the buttons can not be pushed easily if you use the provided case,there should have been left open.I made a 3D printed one with out covering the buttons. So far I am happy with it thought i tend to believe that it lacks a bit of brilliance in highs,not quite sure yet thought.I have to listen to many more and various recording to be more specific but just as an opinion not measuring or proving anything.Overall its a good portable unit!
 
I don't like the volume control though both modes are not good,there should have been a menu option to switch the +- buttons to just volume up down.
The Fiio Melody has the same problem. I just put the volume at maximum though, and alter my OS's volume (which is what the volume buttons on a headphone cable will alter)
I hear basically no big difference between the various filters.
Same with my melody. From reading other threads on ASR, it seams that all the filter options are basically as good (except for "no over sampling", which is basically no filter); so I'd just leave this at the default. Similarly with a class AB / class H option on my Melody.
Software control works very good on Android but not the web app,which is not very good for me ,not very well made.
It's horrible! I've complained to Fiio already with a lot of issues I've had with it. The web app however has a few features missing from the android app: you can click and drag the points in an EQ graph, you can import and export EQ profiles, the AutoEQ features has more settings, and you can upload custom frequency responses and targets to AutoEQ.
I don't use EQ anyway since i have a thing to always listen to sound as it was recorded,at least in portable situations.
That was my logic too before discovering ASR. The thing is that different headphones and speakers makes everything sound different anyway, so EQ can basically make your headphones sounds similar to other headphones. In particular, equalising to the Harman gives you a "neutral" or standard sound. (Of course I don't actually like this sound, as it feels hollow/narrow, I want to feel like the music is all around me! So I've partially equalise my headphones: I've increased the base, and made the highs follow Harman, but left the mid-range frequencies alone). Of course trying to EQ your headphones will make you addicted and spend tons of time on it... So I would at least give the AutoEQ feature on the Fiio app/website a try. (Hand made presets like Amir's or Oratory1990's are good to try as well, but they don't have your headphones)

So far I am happy with it thought i tend to believe that it lacks a bit of brilliance in highs,not quite sure yet thought.I have to listen to many more and various recording to be more specific but just as an opinion not measuring or proving anything.
From reading ASR, any competent DAC should sound the same (at the same volume). So either do a blind test or provide some measurements showing the problem; or people will just think you're imagining it (e.g. I've in fact been imagining that the volume of my DAC connected to my computer isn't as loud as when connected to my phone, but a SPL-metre test confirmed they are the same: I think I just pay more attention to the sound when on my phone, vs. when reading my big and bright monitor).
 
The Fiio Melody has the same problem. I just put the volume at maximum though, and alter my OS's volume (which is what the volume buttons on a headphone cable will alter)

Same with my melody. From reading other threads on ASR, it seams that all the filter options are basically as good (except for "no over sampling", which is basically no filter); so I'd just leave this at the default. Similarly with a class AB / class H option on my Melody.

It's horrible! I've complained to Fiio already with a lot of issues I've had with it. The web app however has a few features missing from the android app: you can click and drag the points in an EQ graph, you can import and export EQ profiles, the AutoEQ features has more settings, and you can upload custom frequency responses and targets to AutoEQ.

That was my logic too before discovering ASR. The thing is that different headphones and speakers makes everything sound different anyway, so EQ can basically make your headphones sounds similar to other headphones. In particular, equalising to the Harman gives you a "neutral" or standard sound. (Of course I don't actually like this sound, as it feels hollow/narrow, I want to feel like the music is all around me! So I've partially equalise my headphones: I've increased the base, and made the highs follow Harman, but left the mid-range frequencies alone). Of course trying to EQ your headphones will make you addicted and spend tons of time on it... So I would at least give the AutoEQ feature on the Fiio app/website a try. (Hand made presets like Amir's or Oratory1990's are good to try as well, but they don't have your headphones)


From reading ASR, any competent DAC should sound the same (at the same volume). So either do a blind test or provide some measurements showing the problem; or people will just think you're imagining it (e.g. I've in fact been imagining that the volume of my DAC connected to my computer isn't as loud as when connected to my phone, but a SPL-metre test confirmed they are the same: I think I just pay more attention to the sound when on my phone, vs. when reading my big and bright monitor).
Well that was a very detailed reply.As i said i have no problem with the android app,works well for me .It would be better for me if the made the web EQ into an app.Web does work well on me and most times does not detect the dongle.Spending allot of time with EQ while i am out enjoying music is like you said addictive and gets you hooked.Maybe while on a desktop (laptop,pc etc) and with time to play it might be a thing to do.Yes basically all dual Cirrus chips are similar ,some might be a bit louder or drive a higher ohm headphone easier but in basics are the same.I will try to do a-b blind tests with other dongles i own just out of curiosity.Like i said its a preception of what i think i heard this past weeks not a proeven fact. Well sometimes ,you might thik something is different like you said, on the other hand ,there other factors,such as external noise in the back,listen in quieter environment vs a louder,stress ...o yes it plays a role,and also if you or your ears or you are tired.Since i am into music production,sometimes i hear my mixes as great or bad and the next day to my surprise they sound different,either better or worse.Basically in all ,to get back to the subject its a nice dongle,hope they improve the volume control.
 
Maybe while on a desktop (laptop,pc etc) and with time to play it might be a thing to do.
Just a suggestion, but I've found that lying down in the dark with my phone makes it much easier to notice the effect of EQ than when sitting at my desk in the light.
 
The Fiio Melody has the same problem. I just put the volume at maximum though, and alter my OS's volume (which is what the volume buttons on a headphone cable will alter)
You're probably not getting the full dynamic range of your music by keeping the Melody at max and attenuating your source volume.

From reading other threads on ASR, it seams that all the filter options are basically as good (except for "no over sampling", which is basically no filter); so I'd just leave this at the default. Similarly with a class AB / class H option on my Melody.
The comprehensive measurements of the Melody here suggest that there are significant measurable differences between class AB and class H modes, as well as DRE on vs off. True, they're barely perceptible, but I like the fact that the Melody makes all these settings user-selectable. I'm considering getting one just for this reason.

About filters, I defer to and agree with Archimago that linear phase + fast roll-off is preferable to all other filters. This post of his (IMO) brings together all his reasoning and measurement analysis neatly in one place, and (IMO) conclusively puts the issue to rest once and for all. We can argue about the audibility of the differences between filters all we like, but the measurements are clear, and if you want the intellectual satisfaction that you're getting the most objectively accurate reproduction of audio out of your system, linear phase + fast roll-off is the way to go.

It's horrible! I've complained to Fiio already with a lot of issues I've had with it. The web app however has a few features missing from the android app: you can click and drag the points in an EQ graph, you can import and export EQ profiles, the AutoEQ features has more settings, and you can upload custom frequency responses and targets to AutoEQ.
Does the Melody have a lot of app issues? Is it a deal-breaker? I wouldn't be using the EQ features of the app, just occasionally adjusting the filter/amp/DRE options.
 
You're probably not getting the full dynamic range of your music by keeping the Melody at max and attenuating your source volume.
I feel like I've read somewhere thar the DAC volume was sent seperately to the audio data (so modifying the volume). Regardless, I use it in 32-bit mode so hopefully it doesn't make a difference. (I put my melody at the back of ny computer, so I really don't want to have to press the buttons)
The comprehensive measurements of the Melody here suggest that there are significant measurable differences between class AB and class H modes, as well as DRE on vs off.
Hmm, the default is class H, but that says that AB is better. So I guess I'll change to that.
True, they're barely perceptible, but I like the fact that the Melody makes all these settings user-selectable. I'm considering getting one just for this reason.
It's a good device, the only real problem is the volume buttons: there are two modes of operation in the settings, either pressing the buttons changes the volume, or holding them does. But there's no option to let you do either! (because the volume buttons double as nxt/previous track buttons, but that feature didn't even work for me).
About filters, I defer to and agree with Archimago that linear phase + fast roll-off is preferable to all other filters. This post of his (IMO) brings together all his reasoning and measurement analysis neatly in one place, and (IMO) conclusively puts the issue to rest once and for all. We can argue about the audibility of the differences between filters all we like, but the measurements are clear, and if you want the intellectual satisfaction that you're getting the most objectively accurate reproduction of audio out of your system, linear phase + fast roll-off is the way to go.
Thanks! The default is minimum phase fast-rollof, so I'll change that to linear.
Does the Melody have a lot of app issues? Is it a deal-breaker? I wouldn't be using the EQ features of the app, just occasionally adjusting the filter/amp/DRE options.
You can only change the settings on an Android (not iOS or a computer); you can change the EQ on a Mac/Windows/Linux computer though.
Most of my issues with the app have been EQ related, however:
  • if you disconnect and reconnect your device while the app is suspended/in the background, it gets confused and the settings don't dp anything, you just and you have to go back to the main page of the app and reconnect it.
  • the app often stops any in-line microphone/cable buttons from working, but you can just exit the app.
 
You're probably not getting the full dynamic range of your music by keeping the Melody at max and attenuating your source volume.

I recently found this wonderful little video by an ESS engineer explaining the difference between analog and digital volume:


TLDR: As long as the volume control is happening in a 32-bit environment with proper dithering, then the results will be audibly indistinguishable from the best analog volume implementation.

(I would add that a pre-SRC 32-bit digital volume is even better than an analog one because of its ability to fix inter-sample clipping).

Thanks! The default is minimum phase fast-rollof, so I'll change that to linear.

Indeed, linear phase, fast rolloff is more accurate. Minimum phase fast rolloff is beneficial if you want the least amount of latency.
 
You're probably not getting the full dynamic range of your music by keeping the Melody at max and attenuating your source volume.


The comprehensive measurements of the Melody here suggest that there are significant measurable differences between class AB and class H modes, as well as DRE on vs off. True, they're barely perceptible, but I like the fact that the Melody makes all these settings user-selectable. I'm considering getting one just for this reason.

About filters, I defer to and agree with Archimago that linear phase + fast roll-off is preferable to all other filters. This post of his (IMO) brings together all his reasoning and measurement analysis neatly in one place, and (IMO) conclusively puts the issue to rest once and for all. We can argue about the audibility of the differences between filters all we like, but the measurements are clear, and if you want the intellectual satisfaction that you're getting the most objectively accurate reproduction of audio out of your system, linear phase + fast roll-off is the way to go.


Does the Melody have a lot of app issues? Is it a deal-breaker? I wouldn't be using the EQ features of the app, just occasionally adjusting the filter/amp/DRE option

Hi you said ,linear phase + fast roll-off is the way to go.Which filter setting is that ? There are 5 filters (FAST-LL, FAST-PC,SLOW-LL,SLOW-PC.NON OS) , Fast PC seems to be Fast roll-off but no mention of linear phase.
 
Hi you said ,linear phase + fast roll-off is the way to go.Which filter setting is that ? There are 5 filters (FAST-LL, FAST-PC,SLOW-LL,SLOW-PC.NON OS) , Fast PC seems to be Fast roll-off but no mention of linear phase.
PC means phase compensation. That's the linear phase option.
 
What does LL mean?
Low latency. Unfortunately the naming conventions aren't standardized and immediately obvious.

For anyone who's interested :) minimum phase filters have a very slight latency advantage over linear phase filters, as well as requiring slightly less compute power I believe, which is likely why minimum phase is most commonly set as the default option or only option in digital audio these days. It may also be because of ringing, but the complaints about audible pre-ringing (in properly recorded and mastered music) caused by linear phase filters are just a myth - it's the (undesirable) phase distortion of minimum phase filters that cause the slight audible differences between the filters.
 
Last edited:
Hi you said ,linear phase + fast roll-off is the way to go.Which filter setting is that ? There are 5 filters (FAST-LL, FAST-PC,SLOW-LL,SLOW-PC.NON OS) , Fast PC seems to be Fast roll-off but no mention of linear phase.

FAST-LL = Fast Low Latency = Steep Roll-Off Minimum Phase (only use when latency is a concern)

FAST-PC = Fast Phase Compensated = Steep Roll-Off Linear Phase (best)

SLOW-LL = Slow Low Latency = Gentle Roll-Off Minimum Phase (inaccurate)

SLOW-PC = Slow Phase Compensated = Gentle Roll-Off Linear Phase (inaccurate)

NON OS = Non Oversampling = Very Gentle Roll-Off (crap, unless you're using your own oversampled files)
 
Back
Top Bottom