FIIO Warmer R2R DAC (with tube buffer)

[Sliphorn]

Improved meaning what? Less roll-off?

.
Please go to Youtube and check it out.
.

 

[solderdude]

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

Demonstrating what aspect ?

Aging of tubes is well known.
How this pans out in measurements is circuit and tube dependent.
Perhaps somewhere there is someone doing repairs on a tube amp and did before and after measurements.

 

[solderdude]

.
Please go to Youtube and check it out.
.

please post before and after measurements that shows better performance after 'hundreds of hours'.
And ... that must be when the DAC is connected to the same amp + load.

No measurements = just an opinion.

 

[Angst92]

Demonstrating what aspect ?

Aging of tubes is well known.
How this pans out in measurements is circuit and tube dependent.
Perhaps somewhere there is someone doing repairs on a tube amp and did before and after measurements.

Demonstrating the audible changes that occur is what im asking.
The measurements shown here show the warmer has a rather significant bass roll off, i was wondering if this changed over time as the tubes are used as some people are claiming that after a few hundred hours the bass roll off is reduced.
Like i said im just finding it hard to find any actual measurements demonstrating the audible measurable changes that can occur over time.

 

[solderdude]

Demonstrating the audible changes that occur is what im asking.
The measurements shown here show the warmer has a rather significant bass roll off, i was wondering if this changed over time as the tubes are used as some people are claiming that after a few hundred hours the bass roll off is reduced.
Like i said im just finding it hard to find any actual measurements demonstrating the audible measurable changes that can occur over time.

There won't be any changes due to tubes or capacitors 'breaking in'.

Besides the measurements from @Jeromeof are recent and he has been using the warmer for a while I believe.
It shows the reported bass roll-off even after it has been used before.

1: people have a tendency to get used to a sound over time which can explain the 'this many hours is needed'.

2: It looks like the output capacitors are 2.2uF (judging from pictures) and IF these are the actual output capacitors then a 2kohm load (4kohm for balanced) could cause the roll-off that is shown in the measurements.
IF that is the case/cause and this DAC would be connected to a headphone amp with say 10kohm (RCA) then the -3dB point could have dropped to 7Hz instead of 36Hz.
This would need to be confirmed with measurements.

We can't say without additional measurements.

 

[Sliphorn]

please post before and after measurements that shows better performance after 'hundreds of hours'.
And ... that must be when the DAC is connected to the same amp + load.

No measurements = just an opinion.

.
Why should I post before and after measurements? I simply reported what Robbie Khan, the reviewer, has stated. I don’t even own the DAC in question, yet. Please contact Robbie Khan with your request, not me.
.

 

[Jeromeof]

FiiO own published measurements show a bass roll-off:

 

[solderdude]

FiiO own published measurements show a bass roll-off:

View attachment 497261

Looks like the roll-off appears to fit an output capacitor of 2.2uF (7Hz -3dB).
Chances are that if you would re-measure using an amplifier with a high input resistance of >10kohm that there won't be any roll-off visible.

So bass response is amplifier input resistance dependent so best to only use headphone amps or pre-amps with Rin > 10kohm.
Mystery seems solved.
That is ... when the input resistance of the used amp (K17 ?) is around 2kohm.
This is not mentioned in the specs.

 

[ObjectiveSubjectivist]

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).

I believe people are hearing this in some cases but their brain tells them its better its warm its tube sound and they convince themselves and they hear better sound indeed

 

[mhardy6647]

monkees mentioned!!!

Did you report me to the moderation team?

 

[foden2025]

I believe people are hearing this in some cases but their brain tells them its better its warm its tube sound and they convince themselves and they hear better sound indeed

I don't care about the tube buffer in this DAC one bit since I have a tube preamp that I know it have a big impact on the sound I'm getting. I'm actually a bit skeptical if the "warm" sound I'm getting from this DAC is due to tube buffer but mainly it's due to the R2R construction. I did a comparison today multiple times between NOS with no HQPlayer oversampling and HQPlayer with digital filters and I heard a difference in most tracks with heavy orechestra works, Jazz trio works. I can call it natural, warm and analog and you can call it treble roll off which I feel it's a bit rolled off on this DAC.

 

[Sliphorn]

I don't care about the tube buffer in this DAC one bit since I have a tube preamp that I know it have a big impact on the sound I'm getting. I'm actually a bit skeptical if the "warm" sound I'm getting from this DAC is due to tube buffer but mainly it's due to the R2R construction. I did a comparison today multiple times between NOS with no HQPlayer oversampling and HQPlayer with digital filters and I heard a difference in most tracks with heavy orechestra works, Jazz trio works. I can call it natural, warm and analog and you can call it treble roll off which I feel it's a bit rolled off on this DAC.

.
About how many hours of play time do you have on this DAC?

 

[foden2025]

.
About how many hours of play time do you have on this DAC?

Around 45-55 hours.

 

[solderdude]

I don't care about the tube buffer in this DAC one bit since I have a tube preamp that I know it have a big impact on the sound I'm getting. I'm actually a bit skeptical if the "warm" sound I'm getting from this DAC is due to tube buffer but mainly it's due to the R2R construction. I did a comparison today multiple times between NOS with no HQPlayer oversampling and HQPlayer with digital filters and I heard a difference in most tracks with heavy orechestra works, Jazz trio works. I can call it natural, warm and analog and you can call it treble roll off which I feel it's a bit rolled off on this DAC.

Yes, the filterless option will be the only sound affecting aspect.
The added harmonic AND intermodulation distortion added by the tubes will mostly be masked by the music.
Besides the poorly performing (from a signal fidelity p.o.v.) R2R DAC construction, which is non-linear, has a large harmonic spray that unlike the tube behavior adds higher order harmonics and IM products.
Add to that the tube distortion will become less with when the volume of the signal decreases but the distortion from the R2R non-linearities will become relatively worse the lower the signal is.
There is a reason why the specifications are quoted at -6dB FS out... it may be much worse at 0dBFS and FiiO might not wanted to publish that number.
At some point in time it will be properly measured most likely.
The bass roll-off also won't be an issue in your case as usually tube pres have a high input resistance (usually around 100k).

Choosing OS or using HQplayer solves the 'roll-off' from the sample-and-hold errors.

Fortunately... the hearing is quite insensitive for this kind of 'changes to the sound' (depends on the recording a bit too).

The looks are great though which means half the battle is already won.

 

[solderdude]

.
About how many hours of play time do you have on this DAC?

Hours does not matter.

The bass roll-off is caused by the output capacitor capacitance and that WON'T change over time.

Tubes also will NOT change LF response over time and IF they do it will only become worse as that would mean the tube is faulty.

 

[foden2025]

Yes, the filterless option will be the only sound affecting aspect.
The added harmonic AND intermodulation distortion added by the tubes will mostly be masked by the music.
Besides the poorly performing (from a signal fidelity p.o.v.) R2R DAC construction that is nonlinear has a large harmonic spray that unlike the tube behavior adds higher order harmonics and IM products.
Add to that the tube distortion will become less with when the volume of the signal decreases but the distortion from the R2R non-linearities will become relatively worse the lower the signal is.
There is a reason why the specifications are quoted at -6dB FS out... it may be much worse at 0dBFS and FiiO might not wanted to publish that number.
At some point in time it will be properly measured most likely.
The bass roll-off also won't be an issue in your case as usually tube pres have a high input resistance (usually around 100k).

Choosing OS or using HQplayer solves the 'roll-off' from the sample-and-hold errors.

Fortunately... the hearing is quite insensitive for this kind of 'changes to the sound' (depends on the recording a bit too).

The looks are great though which means half the battle is already won.

My main concern is the ultra sonic noise, from what I read on this and please correct me if I'm wrong: A NOS R2R DAC implements a low pass filter in the output stage as in a tube buffer like in the case here and that helps not passing those noise to the amplifier and in my case (Class D Purifi amp) which is sensitive to those ultra sonic noises. If that how this DAC works I don't think I have a reason to use digital filters in HQPlayer since I like the sound better with just NOS + no digital filters, if there is a risk passing those ultra sonic noises to a sensitive Class D amp I might need to use the digital filtering in HQPlayer, is that correct?

 

[Blockader]

My main concern is the ultra sonic noise, from what I read on this and please correct me if I'm wrong: A NOS R2R DAC implements a low pass filter in the output stage as in a tube buffer like in the case here and that helps not passing those noise to the amplifier and in my case (Class D Purifi amp) which is sensitive to those ultra sonic noises. If that how this DAC works I don't think I have a reason to use digital filters in HQPlayer since I like the sound better with just NOS + no digital filters, if there is a risk passing those ultra sonic noises to a sensitive Class D amp I might need to use the digital filtering in HQPlayer, is that correct?

NOS dacs do not use low-pass filters, which is the whole idea behind their design. Instead of filtering, they rely on a simple "step-and-hold" process (zero-order hold) to reconstruct the signal. This creates strong ultrasonic images that can leak into later stages of the audio chain. While some NOS designs add a very gentle analog low-pass filter, it is generally not strong enough to meaningfully suppress this ultrasonic content.

Delta sigma dacs, on the other hand, use heavy digital filtering together with noise shaping, followed by an analog low-pass filter. In well designed implementations, out-of-band and ultrasonic noise can be pushed down to around −110 db or lower.

This is why NOS is a bad idea. Proper reconstruction filters prevent unnecessary ultrasonic energy from ever reaching the amplifier and tweeter, keeping the driver out of its breakup region. Breakup mode distortion is 100.000 times louder than "ultra sonic noise" and the tweeter can fold it back to audible range via intermodulation.

 

[DSJR]

I'm not sure I've seen a measured DACs that does something more exciting that rolling off either some amount of treble or bass, or adding faint levels of distortion. I think that calls for DSP, not swapping converters.

Personally, I think if you want to home-remaster your music, sticking it into a DAW and futzing around with the reams of eq and distortion and dynamics plugins you can find is both a more fun and more productive activity.

Late reply and a complete? aside, but my cheap-n-humble SMSL SU1 easily reproduces production and mixing differences...

As you were chaps...

 

[solderdude]

My main concern is the ultra sonic noise, from what I read on this and please correct me if I'm wrong: A NOS R2R DAC implements a low pass filter in the output stage as in a tube buffer like in the case here and that helps not passing those noise to the amplifier and in my case (Class D Purifi amp) which is sensitive to those ultra sonic noises. If that how this DAC works I don't think I have a reason to use digital filters in HQPlayer since I like the sound better with just NOS + no digital filters, if there is a risk passing those ultra sonic noises to a sensitive Class D amp I might need to use the digital filtering in HQPlayer, is that correct?

Yes.
In NOS mode there will be a LOT of ultrasonic crap in the signal. It is the 'mirrored' image of the sound (mirrored at 22.05kHz or 24kHz in most cases).
Inaudible by itself as we can't hear anything there but the ultrasonics could potentially be a bit problematic in some cases.
You can even spot some of it here (look closely at the 'noise' in the higher frequencies)
Note: headphones and the used test fixture will have filtered out most of it but it will surely be there.

There usually is a very slow (6db or 12dB/octave) low pass in there above 40kHz or so to get rid of the MHz crap but that won't do anything for the crap that might damage tweeters (when playing really loud) or might interfere with the modulator of class-D amps.
The latter (when it is a good one) should have a steep low-pass at its input to prevent this but.. you never know.

Best NOT to use NOS mode (simulated or not) when using class-D amps for sure.
In fact best to not use that mode at all. If a gentle roll-off is your thing do that filtering digitally before it hits the DAC.
It is not exactly the same effect but comes close.

Best to use OS mode in this DAC anyway.
The whole NOS thing is a gimmick but people ask for it so the manufacturers provide.
They don't care if, in rare cases, it might cause an issue. It won't in most cases anyway.
Manufacturers only care about selling as much as they can to an as wide as possible audience.

 

[foden2025]

Yes.
In NOS mode there will be a LOT of ultrasonic crap in the signal. It is the 'mirrored' image of the sound (mirrored at 22.05kHz or 24kHz in most cases).
Inaudible by itself as we can't hear anything there but the ultrasonics could potentially be a bit problematic in some cases.
You can even spot some of it here (look closely at the 'noise' in the higher frequencies)
Note: headphones and the used test fixture will have filtered out most of it but it will surely be there.
View attachment 497453

There usually is a very slow (6db or 12dB/octave) low pass in there above 40kHz or so to get rid of the MHz crap but that won't do anything for the crap that might damage tweeters (when playing really loud) or might interfere with the modulator of class-D amps.
The latter (when it is a good one) should have a steep low-pass at its input to prevent this but.. you never know.

Best NOT to use NOS mode (simulated or not) when using class-D amps for sure.
In fact best to not use that mode at all. If a gentle roll-off is your thing do that filtering digitally before it hits the DAC.
It is not exactly the same effect but comes close.

Best to use OS mode in this DAC anyway.
The whole NOS thing is a gimmick but people ask for it so the manufacturers provide.
They don't care if, in rare cases, it might cause an issue. It won't in most cases anyway.
Manufacturers only care about selling as much as they can to an as wide as possible audience.

Thanks, one question though does this applies to PCM 44Hz only? Meaning if I feed PCM 44Hz files after the digital filtering and upsampling in HQPlayer but PCM files higher than that like 192Hz or more and DSD files directly to NOS mode in the DAC those files are immune to those ultrasonic noises correct?