FIIO Warmer R2R DAC (with tube buffer)

[foden2025]

It has already been established that this DAC has a different sound signature by design.

What is being said here is that all well designed DACs that measure transparent can not be told apart in a controlled blind test.

This is something entirely different than stating all DACs sound the same. They don't and is easily shown with measurements.

The DAC in question will sound fine to the vast majority of people despite the relatively poor measurements.

I can totally agree with that, although I didn't do that test. Yes, this DAC is different than DS DAC I listened to. It's what I'm after in my system, non fatigue long sessions. BTW, I paired it with Apollon Class D Purifi amp, BAT tube preamp no harshness nothing but clean sound from that amp, of course I'm adding tube harmonics in preamp and DAC.

 

[Jeromeof]

So, I really need to dig out my Cosmos ADCISO for some proper measurements, but since I was doing some measurements of the FiiO FT13 headphone and @solderdude made an interesting point about the 'measurable differences' in the treble between NOS and OS. I thought I would do some 'headphone' measurements via some of the FiiO stuff I have setup.

So, first I did a measurement of the FT13 headphone (with my KB501X with a 711 clone coupler) directly via the K17 using its AK4499EX + AK4191 in low gain, then without moving anything I routed the measurement through the Warmer in both OS and NOS mode ( level matching the audio using the K17 ):

So first thing to note there is a bass roll-off with both the OS and NOS modes via the Warmer compared to the K17:

Since as there is a little too treble with the FT13 anyway, I was thinking that maybe the Warmer might just be the tweak the FT13 might need, so zooming in on the treble. So first the OS via the SS ( I did 2 measurements before and after of the OS mode), but in OS mode you get slightly more upper treble with OS mode:

But then comparing the OS and NOS modes - you can see NOS does lower the treble by 1-2 dB from 8K onwards:

And here are the 3 options:

Looking at the 3rd harmonic distortion differences

Anyway, I thought these unconventional measurements might be interesting to people, if I get some time (unlikely as it is Christmas Party season here). I will see can get pure DAC measurements.
:

 

[oleg87]

What's going on with the bass? That's a huge difference
fiio's spec sheet shows a bit of rolloff, but 4db at 30hz? no input impedance specs for the K17 I can find.

 

[mhardy6647]

.
I wouldn’t consider 88-100 degrees farenheit as hot. Warm, yeah.
.

more or less snuggly/cuddly warm, with respect to the standard human body temperature.

 

[Jeromeof]

What's going on with the bass? That's a huge difference

Yes, did them a few times to verify and the headphones stayed connected to the K17 and where not moved, to only differences where I routed the usb either directly through the K17 or via the Warmer and the Warmer output into the RCA input of the K17 ( I might measure the RCA input on the K17 tomorrow separately).

 

[Curupira]

So, I really need to dig out my Cosmos ADCISO for some proper measurements, but since I was doing some measurements of the FiiO FT13 headphone and @solderdude made an interesting point about the 'measurable differences' in the treble between NOS and OS. I thought I would do some 'headphone' measurements via some of the FiiO stuff I have setup.

So, first I did a measurement of the FT13 headphone (with my KB501X with a 711 clone coupler) directly via the K17 using its AK4499EX + AK4191 in low gain, then without moving anything I routed the measurement through the Warmer in both OS and NOS mode ( level matching the audio using the K17 ):

So first thing to note there is a bass roll-off with both the OS and NOS modes via the Warmer compared to the K17:
View attachment 496433
Since as there is a little too treble with the FT13 anyway, I was thinking that maybe the Warmer might just be the tweak the FT13 might need, so zooming in on the treble. So first the OS via the SS ( I did 2 measurements before and after of the OS mode), but in OS mode you get slightly more upper treble with OS mode:

View attachment 496434

But then comparing the OS and NOS modes - you can see NOS does lower the treble by 1-2 dB from 8K onwards:
View attachment 496437
And here are the 3 options:View attachment 496438

Looking at the 3rd harmonic distortion differencesView attachment 496439

Anyway, I thought these unconventional measurements might be interesting to people, if I get some time (unlikely as it is Christmas Party season here). I will see can get pure DAC measurements.
:

Superb work.

 

[foden2025]

So, I really need to dig out my Cosmos ADCISO for some proper measurements, but since I was doing some measurements of the FiiO FT13 headphone and @solderdude made an interesting point about the 'measurable differences' in the treble between NOS and OS. I thought I would do some 'headphone' measurements via some of the FiiO stuff I have setup.

So, first I did a measurement of the FT13 headphone (with my KB501X with a 711 clone coupler) directly via the K17 using its AK4499EX + AK4191 in low gain, then without moving anything I routed the measurement through the Warmer in both OS and NOS mode ( level matching the audio using the K17 ):

So first thing to note there is a bass roll-off with both the OS and NOS modes via the Warmer compared to the K17:
View attachment 496433
Since as there is a little too treble with the FT13 anyway, I was thinking that maybe the Warmer might just be the tweak the FT13 might need, so zooming in on the treble. So first the OS via the SS ( I did 2 measurements before and after of the OS mode), but in OS mode you get slightly more upper treble with OS mode:

View attachment 496434

But then comparing the OS and NOS modes - you can see NOS does lower the treble by 1-2 dB from 8K onwards:
View attachment 496437
And here are the 3 options:View attachment 496438

Looking at the 3rd harmonic distortion differencesView attachment 496439

Anyway, I thought these unconventional measurements might be interesting to people, if I get some time (unlikely as it is Christmas Party season here). I will see can get pure DAC measurements.
:

Thank you, I think Robbie in his youtube review mentioned that bass was not as impactful as in the Fiio K13 and another reviewer on youtube mentioned the same so this was the first thing I noticed too, in my SVS subwoofer I used before room correction gain at 40 Hz 6dB slope, with the Fiio I had to disable that since bass was already not poweful and increased gain volume in the sub.

 

[JSmith]

I'll be gone for good now.

Thanks for coming;

JSmith

 

[raif71]

Warmer R2R arrived today and it sounds amazing.

Don't listen to these old jokers that DAC's don't have sound signature
I'll be gone for good now.

 

[theREALdotnet]

They should have used Nixie tubes for the VU meter. Missed opportunity...

 

[Soria Moria]

So, I really need to dig out my Cosmos ADCISO for some proper measurements, but since I was doing some measurements of the FiiO FT13 headphone and @solderdude made an interesting point about the 'measurable differences' in the treble between NOS and OS. I thought I would do some 'headphone' measurements via some of the FiiO stuff I have setup.

So, first I did a measurement of the FT13 headphone (with my KB501X with a 711 clone coupler) directly via the K17 using its AK4499EX + AK4191 in low gain, then without moving anything I routed the measurement through the Warmer in both OS and NOS mode ( level matching the audio using the K17 ):

So first thing to note there is a bass roll-off with both the OS and NOS modes via the Warmer compared to the K17:
View attachment 496433
Since as there is a little too treble with the FT13 anyway, I was thinking that maybe the Warmer might just be the tweak the FT13 might need, so zooming in on the treble. So first the OS via the SS ( I did 2 measurements before and after of the OS mode), but in OS mode you get slightly more upper treble with OS mode:

View attachment 496434

But then comparing the OS and NOS modes - you can see NOS does lower the treble by 1-2 dB from 8K onwards:
View attachment 496437
And here are the 3 options:View attachment 496438

Looking at the 3rd harmonic distortion differencesView attachment 496439

Anyway, I thought these unconventional measurements might be interesting to people, if I get some time (unlikely as it is Christmas Party season here). I will see can get pure DAC measurements.
:

Bass roll-off is the last thing I want a dac to do. I was semi-interested in this product but now I no longer am. Thanks for this!

 

[Mnyb]

So, I really need to dig out my Cosmos ADCISO for some proper measurements, but since I was doing some measurements of the FiiO FT13 headphone and @solderdude made an interesting point about the 'measurable differences' in the treble between NOS and OS. I thought I would do some 'headphone' measurements via some of the FiiO stuff I have setup.

So, first I did a measurement of the FT13 headphone (with my KB501X with a 711 clone coupler) directly via the K17 using its AK4499EX + AK4191 in low gain, then without moving anything I routed the measurement through the Warmer in both OS and NOS mode ( level matching the audio using the K17 ):

So first thing to note there is a bass roll-off with both the OS and NOS modes via the Warmer compared to the K17:
View attachment 496433
Since as there is a little too treble with the FT13 anyway, I was thinking that maybe the Warmer might just be the tweak the FT13 might need, so zooming in on the treble. So first the OS via the SS ( I did 2 measurements before and after of the OS mode), but in OS mode you get slightly more upper treble with OS mode:

View attachment 496434

But then comparing the OS and NOS modes - you can see NOS does lower the treble by 1-2 dB from 8K onwards:
View attachment 496437
And here are the 3 options:View attachment 496438

Looking at the 3rd harmonic distortion differencesView attachment 496439

Anyway, I thought these unconventional measurements might be interesting to people, if I get some time (unlikely as it is Christmas Party season here). I will see can get pure DAC measurements.
:

This DAC seems to be one of the outliers that really have potential to sound slightly different especially in NOS mode , thanks for the testing .

 

[Angst92]

So, I really need to dig out my Cosmos ADCISO for some proper measurements, but since I was doing some measurements of the FiiO FT13 headphone and @solderdude made an interesting point about the 'measurable differences' in the treble between NOS and OS. I thought I would do some 'headphone' measurements via some of the FiiO stuff I have setup.

So, first I did a measurement of the FT13 headphone (with my KB501X with a 711 clone coupler) directly via the K17 using its AK4499EX + AK4191 in low gain, then without moving anything I routed the measurement through the Warmer in both OS and NOS mode ( level matching the audio using the K17 ):

So first thing to note there is a bass roll-off with both the OS and NOS modes via the Warmer compared to the K17:
View attachment 496433
Since as there is a little too treble with the FT13 anyway, I was thinking that maybe the Warmer might just be the tweak the FT13 might need, so zooming in on the treble. So first the OS via the SS ( I did 2 measurements before and after of the OS mode), but in OS mode you get slightly more upper treble with OS mode:

View attachment 496434

But then comparing the OS and NOS modes - you can see NOS does lower the treble by 1-2 dB from 8K onwards:
View attachment 496437
And here are the 3 options:View attachment 496438

Looking at the 3rd harmonic distortion differencesView attachment 496439

Anyway, I thought these unconventional measurements might be interesting to people, if I get some time (unlikely as it is Christmas Party season here). I will see can get pure DAC measurements.
:

Ive heard that tubes can change characteristics over an initial usage period, how many hours roughly would you say youve used the warmer for before taking these measurements?

 

[solderdude]

What's going on with the bass?

Possibly it is intentional to 'mimic' all tube amps which generally have low bass roll-off because of the output transformers.

Could also be that a relatively low output coupling capacitor is used (2.2uF) if the actual output stage is the tube buffer.
In that case the roll-off could will be load resistance dependent.

It is very clear that the Warmer has a deliberately different 'tone' than your standard DAC.
The differences are audible and very measurable and explain the 'reports' on the sound.
The influence of the DAC section and/or tube section are visible in the 30dB higher distortion versus the K17.

I still find it funny that one can f-up technical performance (signal fidelity) so much it is even audible and people still like it (and even prefer it).
To me it is testimony of how poor the hearing is in gauging 'signal fidelity' and how lenient the hearing is for signal degradation (things like tonal changes and added IM products).

Ive heard that tubes can change characteristics over an initial usage period, how many hours roughly would you say youve used the warmer for before taking these measurements?

This is not a tube power amp where one might need a few minutes or so to get all tubes to get in its intended bias range.

This is a tube buffer using ECC88 meaning less than a minute is required and there will be no change after that.

 

[Angst92]

This is not a tube power amp where one might need a few minutes or so to get all tubes to get in its intended bias range.

This is a tube buffer using ECC88 meaning less than a minute is required and there will be no change after that.

Do you know of any studies that have tracked the changes over time to definitively show how long it takes for changes to stabilise? I'm finding a lot of opinions but no evidence of anything so far.

 

[solderdude]

You are touching on the problems of tubes.
Aging..

Tubes age from the moment they are turned on. Some do so quickly and some have years of life on them.
I have an old tube radio from my dad. It is older than I am and only had to clean the switches.
Original tubes, original capacitors and still works (I have it on a few hours a month).

Yet, I had a new JJ tube (ECC88) that gave up after a few years and have old pulls that still work.

So there is no general rule as to how quickly a tube ages in such a degree it becomes problematic.
In some cases new tubes may change properties in the first hour before they 'stabilize' in performance.
In other cases it might be months and you may have to re-bias it after some time.
Mostly true for power tubes as they need to emit a LOT of electrons.

For low voltage/low signal tubes such as the ECC88 this is rarely an issue. Certainly not when used as a buffer or 'tubify' stage (with an op-amp as output buffer).

Emission depends on the cathode being heated and this can be pretty quick (within a minute) and depending on how good the coupling between cathode and heater is and what current needs to flow is circuit dependent. This means that in circuit A it might take longer before 'stable' operation is reached than in circuit B using the same tube.

They don't have to break-in for x-hours either.
Needless to say there is tons more of subjective babble about tubes than there is about facts/measurements.
All of this info on tubes has been known since the 50's.

 

[Jiraya369]

I won't argue the noise point, but I will offer three observations.
1) Maybe it's just me, but I think of dither as replacing non-random noise (is that an oxymoron?) with random noise. Is dither still a thing in signal processing?
2) In the 90s, there seemed to be a faddish popularity to adding some LP-record-style surface noise to pristine CD tracks - e.g., Sheryl Crow's Riverwide.
3) The best-ever evocation of old-school noise, arguably, dates back to the analog era... and... the Monkees.

Magnolia Simms, in case the video isn't available in everybody's neck of the woods.

Only in recent years have I stopped to think how this 'effect' (noise + lo fi) was achieved. I am guessing the basic track was recorded and cut on an acetate, and then the acetate was played a number of times, with some physical abuse in between. Needless to say, I've never tried to unleash the power of the interent -- I reckon there're probably making of essays, videos, and podcasts. Heck, Rick Beato's probably interviewed the cutting lathe!

monkees mentioned!!!

 

[Sliphorn]

.
Robbie Khan’s latest Youtube video of his Warmer DAC states that the bass is much improved after several hundred hours of listening.
.

 

[Angst92]

You are touching on the problems of tubes.
Aging..

Tubes age from the moment they are turned on. Some do so quickly and some have years of life on them.
I have an old tube radio from my dad. It is older than I am and only had to clean the switches.
Original tubes, original capacitors and still works (I have it on a few hours a month).

Yet, I had a new JJ tube (ECC88) that gave up after a few years and have old pulls that still work.

So there is no general rule as to how quickly a tube ages in such a degree it becomes problematic.
In some cases new tubes may change properties in the first hour before they 'stabilize' in performance.
In other cases it might be months and you may have to re-bias it after some time.
Mostly true for power tubes as they need to emit a LOT of electrons.

For low voltage/low signal tubes such as the ECC88 this is rarely an issue. Certainly not when used as a buffer or 'tubify' stage (with an op-amp as output buffer).

Emission depends on the cathode being heated and this can be pretty quick (within a minute) and depending on how good the coupling between cathode and heater is and what current needs to flow is circuit dependent. This means that in circuit A it might take longer before 'stable' operation is reached than in circuit B using the same tube.

They don't have to break-in for x-hours either.
Needless to say there is tons more of subjective babble about tubes than there is about facts/measurements.
All of this info on tubes has been known since the 50's.

Thats the thing though, im finding lots of talk about this but no actual measurements demonstrating it.

 

[Soria Moria]

.
Robbie Khan’s latest Youtube video of his Warmer DAC states that the bass is much improved after several hundred hours of listening.
.

Improved meaning what? Less roll-off?