• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. 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!

Master Thread: Are measurements Everything or Nothing?

Can different filters explain why we feel inconsistent sensations when listening to different DACs explain (at least partially) when theoretically are measured linearly?
No, because the effect of the (default) filter is measured on the ouput. Two DACs measured with their default filter that both measure as transparent (inaudible noise, disotortion, and flat FR) will sound the same. Obviously if you change the filter to one that alters the FR (for example) then that can change - but then you are essentially selecting a broken filter.

Or perhaps interaction with non-linear amps explain these inconsistencies because output levels of the DAC can’t produce exactly the same sound even if the preamp volume is used to compensate them?
Audibly identical DACS fet into the same non linear amp will still sound identical. But then why are you using a non linear amp in any case?

I mean, in my particular case my studio monitors have an input gain: one of my two DACs is playing louder than the other so to compensate that I low down the volume of the DAC output or I reduce the gain on the monitor input circuit.

Is this causing some effect? Despite being convinced that both DACs are well made, one of them always cause me fatigue with highs very enhanced. I cannot feel it immediately but soon or later I will have to put the other.
I've seen you mention this a number of times. I can't see what you are perceiving as being based on reality. I think your monitors are genelec? Their gain control is not going to be so awful that they alter the sound in any audible way at different gain levels. I suspect what you are perceiving is due to insufficiently accurate level matching (obviously you cannot measure the amp output with a voltmeter - generally required if you need to accurately level match). Or, equally likely good old sighted cognitive biases. In other words - insufficiently controlled comparisons.

The one that please me more (or at least don’t produce me fatiguing) is the cheapest one (by far) so is difficult that I have psychological predisposition to believe is better; in fact I bought it for recording as it has a ADC and never thought to use it to play music.
That is not how perceptive/cognitive biases work. They are subconscious, and have nothing to do with your conscious expectations.


So I continue to search an explanation from time to time
So if you really want to know, I think the first step of your search - if you are really motivated - is to set up a properly controlled blind test.

(assuming you've checked the measurements of the two devices, and there isn't an obviously audible FR difference)

Level matching will be difficult but you could probably do it well enough by putting the speaker gain to max, and using the preamp mode of one of the DACs to lower the output to match the other. This could then be measured at the output of the DAC using a volt meter. The few DB of digital volume control is not going to make an audible difference. Then find a way to switch between the two DACS blind without any tells.
 
Last edited:
No, because the effect of the (default) filter is measured on the ouput. Two DACs measured with their default filter that both measure as transparent (inaudible noise, disotortion, and flat FR) will sound the same. Obviously if you change the filter to one that alters the FR (for example) then that can change - but then you are essentially selecting a broken filter.


Audibly identical DACS fet into the same non linear amp will still sound identical. But then why are you using a non linear amp in any case?


I've seen you mention this a number of times. I can't see what you are perceiving as being based on reality. I think your monitors are genelec? Their gain control is not going to be so awful that they alter the sound in any audible way at different gain levels. I suspect what you are perceiving is due to insufficiently accurate level matching (obviously you cannot measure the amp output with a voltmeter - generally required if you need to accurately level match). Or, equally likely good old sighted cognitive biases. In other words - insufficiently controlled comparisons.


That is not how perceptive/cognitive biases work. They are subconscious, and have nothing to do with your conscious expectations.
I assume psychological expectations can be an explanation, probably the best of them.

Why about the sound level explanation? If I listen some time to the same repertoire, I tend to adjust the volume to confort level, so this will not explain the fatiguing effect.

The only other way that is the suggested psychological path so I imagine the ancient beliefs stay long time after accepting placebo or nocebo effect, as in drugs trials (it has been shown that even if doctor explain to control group that are taking placebo most of people continue to experiment the effects at least 6 months after explanations. Was conducted in UK some years ago).

Humm, yes are Genelecs but input gain change quite audibly the sound, hope one day I can be able to measure that (or conversely to discover is false). I set the gain knob to the minimum, the sensibility of the Genelecs is so high, even on the + 6 dB adjustments I rarely take the master volume over halfway.
 
Yes some filters can be audible, any HF roll off that starts in the audible band.

TEAC make a DAC that starts rolling off at 12KHz. I'd probably hear that compared to a DAC that is flat to 20KHz. I can't hear any difference in the 6 less radical filters on a Topping E30 but then I'm 55 years old. Not impossible someone much younger could perceive them.

I would just use the DAC you prefer and not worry too much about why that is. It could be in your head or it could be an actual difference in the soundwaves, but it isn't worth losing any sleep over. There's much bigger fish to fry in the quest for good sound - loudspeakers and rooms.
Not worry about it, is just curiosity. Thank you for the explanation, I’m 48 and my threshold limit is about 16 kHz.

The “bothering” DAC is an Ifi one, it uses a proprietary filter called GTO or something similar but never found a graphic of what it is. Because is a recent inclusion in their DACs firmware I think may have changed some frequencies to cause a slight sound signature (the DAC is by random or not called “Zen DAC Signature :) )

The only reason I keep it is that decodes the f***ing MQA on Tidal, hope they will convert all of their files to 16/ 44.1 Flac and stop bothering with software implementation on devices.
 
Not worry about it, is just curiosity. Thank you for the explanation, I’m 48 and my threshold limit is about 16 kHz.

The “bothering” DAC is an Ifi one, it uses a proprietary filter called GTO or something similar but never found a graphic of what it is. Because is a recent inclusion in their DACs firmware I think may have changed some frequencies to cause a slight sound signature (the DAC is by random or not called “Zen DAC Signature :) )

The only reason I keep it is that decodes the f***ing MQA on Tidal, hope they will convert all of their files to 16/ 44.1 Flac and stop bothering with software implementation on devices.
Just a small change in FR can alter the perceived tonal balance by a surprising amount. So could well be that.

I try to keep the internet well away from my music system ;)

I've a CD transport that will decode MQA CDs but I don't own any. Quite a surprising number of them, but the prices, as you'd expect from a format that was a complete rip-off right from the start, are risible:

 
Just a small change in FR can alter the perceived tonal balance by a surprising amount. So could well be that.

I try to keep the internet well away from my music system ;)

I've a CD transport that will decode MQA CDs but I don't own any. Quite a surprising number of them, but the prices, as you'd expect from a format that was a complete rip-off right from the start, are risible:

I see no advantage on MQA, at least the tracks I listened aren’t better than usual CD quality and causes some issues on the DAC (it pops a few times when changing from FLAC to MQA and viceversa).

Why you try to avoid streaming services? Ethical reasons?

Personally without streaming I would have a lot of difficulties to find some recordings that are out of edition… I listen to classical mainly and discs have often not enough buyers to keep them on the market for long time
 
Why you try to avoid streaming services? Ethical reasons?
No! I prefer to browse a physical library, and have guaranteed access to the mastering I like the best.

Also I am stuck in my ways :)

I don't listen to any classical music so wasn't aware that there was a problem with availability of some programme.
 
Humm, yes are Genelecs but input gain change quite audibly the sound,

Unless you are doing properly controlled comparisons, you cannot know if that is in the sound reaching your ears, or is being constructed in the wetware between them.
 
Unless you are doing properly controlled comparisons, you cannot know if that is in the sound reaching your ears, or is being constructed in the wetware between them.
Maybe, is always possible, are you also sure that is no possible technical explanation?

Both systems that have volume controls (pre-amp output and monitor input gain) are after conversion so not subjected to what I recently learned about DAC transparency.

Are the combination of pre-amp and amp totally linear? I mean, not linear on the flat sense some people use here as “flat response”, linear on the sense as multiplication by a factor in one side and division in the other give the same result as on linear operators.
 
Maybe, is always possible, are you also sure that is no possible technical explanation?

No, it is not possible to be certain - but it is pretty unlikely. Making linear preamps is not at all difficult.

Are the combination of pre-amp and amp totally linear? I mean, not linear on the flat sense some people use here as “flat response”, linear on the sense as multiplication by a factor in one side and division in the other give the same result as on linear operators.

If two devices are linear, combining them won't make the result suddenly non linear.
 
No, it is not possible to be certain - but it is pretty unlikely. Making linear preamps is not at all difficult.



If two devices are linear, combining them won't make the result suddenly non linear.
This only states if and only if both are perfectly linear. Addition is not the operation what worry about, but multiplication.

If both devices are slightly non-linear but imperceptible then its composition may contain perceptible nonlinearities.

More concretely, suppose that some operator A(V) = a(V) + e(V) contains an a(V) linear term and a nonlinear negligible term e(V) and same to another B(V) = b(V) + e(V) then the multiplier (amplifying) operator A*B contains the linear term a*b (V) but also a(V)e(V) and b(V)e(V) which are fairly greater than e(V) and nonlinear by definition. The square term e^2 (V) is negligible as is smaller than the supposed negligible original e(V).

This is why two approximately linear functions can combine in an non-linear one in multiplication operator.

As a corollary, if you join enough close-to-linear amplifiers you will end with a beautiful Gaussian :)

Oh, probably you’re right nevertheless and in practice is not so important the volume combinations of the things unless I rise one of them so that there is no margin to avoid clipping.

But is a pleasure to talk with trained people, I don’t do it so often …
 
If both devices are slightly non-linear but imperceptible then its composition may contain perceptible nonlinearities.
Well yes - but only in as much as that statement applies to all audio chains. Point is non linearities of electronics are so far beyond the capabilities of our ears, that even when combined, the threshold is rarely (if ever) breached. If the non linear distortion of each is down at -110 dB, then combined they will still be 104dB. And while that figure is based on addition, multiplication will be just a tiny bit worse because the numbers you are multiplying by are tiny. The additional distortion is -110dB down from -110dB - effectively lost in the noise floor.

Further, the non linearities of your speakers (the transducer part) will easily swamp any non linearities of all the electronics combined.
 
Well yes - but only in as much as that statement applies to all audio chains. Point is non linearities of electronics are so far beyond the capabilities of our ears, that even when combined, the threshold is rarely (if ever) breached. If the non linear distortion of each is down at -110 dB, then combined they will still be 104dB. And while that figure is based on addition, multiplication will be just a tiny bit worse because the numbers you are multiplying by are tiny. The additional distortion is -110dB down from -110dB - effectively lost in the noise floor.

Further, the non linearities of your speakers (the transducer part) will easily swamp any non linearities of all the electronics combined.
Thanks for your patience! :)

You’re correct in your intuitive calculations, because dB is logarithmic unit and addition in that scale is equivalent to multiplication in power (adding decibles is proportional to multiplication of power, I believe is 3 dB= 2 times power or something).

Nevertheless what take into account is practical applications and not mathematical abstract terms, so I trusted your experience and supposed something else was happening on my signal path.

In fact was the iphone again, I don’t know why but once connected my Ifi DAC to the computer or even to the streamer it works perfectly!

I have the bad habit to use the phone for everything, and was convinced that this iteration hand properly class compliant USB but apparently not. My streamer is same brand as the DAC and from the same Zen line so apparently they use the correct digital input signal.

Thanks to your insistence my Genelecs sound now way better, I was touching all possible analogue combinations and analyzing minuscule effects on sound but finally your advices made me think the problem was in the digital source. I should remember that desktop DACs are designed to computers or streamers, compatible to phones operative systems but don’t made to them.

This time I don’t have to think is psychological effect: difference is very huge in sound quality to suppose is just an influence. I don’t know the reason of that, probably the iphone (and the ipad also) don’t give half the power to proper handle the DAC circuits to preserve battery duration.
 
Thanks for your patience! :)

You’re correct in your intuitive calculations, because dB is logarithmic unit and addition in that scale is equivalent to multiplication in power (adding decibles is proportional to multiplication of power, I believe is 3 dB= 2 times power or something).

Nevertheless what take into account is practical applications and not mathematical abstract terms, so I trusted your experience and supposed something else was happening on my signal path.

In fact was the iphone again, I don’t know why but once connected my Ifi DAC to the computer or even to the streamer it works perfectly!

I have the bad habit to use the phone for everything, and was convinced that this iteration hand properly class compliant USB but apparently not. My streamer is same brand as the DAC and from the same Zen line so apparently they use the correct digital input signal.

Thanks to your insistence my Genelecs sound now way better, I was touching all possible analogue combinations and analyzing minuscule effects on sound but finally your advices made me think the problem was in the digital source. I should remember that desktop DACs are designed to computers or streamers, compatible to phones operative systems but don’t made to them.

This time I don’t have to think is psychological effect: difference is very huge in sound quality to suppose is just an influence. I don’t know the reason of that, probably the iphone (and the ipad also) don’t give half the power to proper handle the DAC circuits to preserve battery duration.
I see, like me, you are in EU. I don't know if this is connected to the problem you have, but Iphones in EU must limit power to earphones. When you connect a device they often don't know if it is earphone or speaker, so they limit the power anyway. In bluetooth settings you can telll the phone a device is a speaker so it doesn't limit power. I don't know if/how this is possible for USB devices.

But anyway - I'm glad you've found a reason for your issue.
 
Not gonna lie, the more I read this forum, the more I believe people spend their time looking at graphs instead of actually listening to the damn thing.

I've started out listening to ESS or AKM Topping, FiiO etc. DACs, including the oh so praised Topping DX3+ Pro - because people won't shut up about them. Got them and... they sound boring. All of them. DX3+ especially, as expected from a company hyper focused on funny graphs rather than making their product sound somewhat decent, while ignoring QC completely. Yeah, I said it.

I then listened to iFi stuff, hesitatingly at first, because people here love to hate for using "old Burr Brown DACs", or for having MQA marketing, misreporting power numbers, or other incessantly spouted FUD - but then I decided to take the risk and bought into iFi. No real risk, can always return. Except ...

... I was instantly hooked. My ZEN Signature stack, which is apparently "horrendous" by this place's standards, is doing a fantastic job at making my music lively and warm while maintaining detail and clarity, and I have no doubts that my upcoming NEO iDSD 2 will do the same. The analogue filters like XBass and XSpace do their job, and they do it really well.

iFi products, from what I've experienced, are great at dishing out the one thing that matters most in music - actual musicality. Even their portable GO link, which uses an ESS DAC I expect to sound dull, sounds warm and punchy. All music genres sound like they're supposed to - but more lively, emotional, intimate. I'm not a producer, I don't mix and master music. I don't need every single bit of detail, at the cost of my music sounding extremely dull. I buy an external dac/amp to improve my music listening experience, and that's exactly what the ZEN stack does over my onboard audio. The analogue filters are just the cherry on the top - not necessary, but nice to have and makes a remarkable difference in some cases.

Measurements, distortions, SINAD, whatever the hell - who cares if it tops charts on HERE, when in the end I don't like how the thing sounds. I don't buy audio equipment in order to hear the producer scratch his behind in the studio and knowing my dac/amp has 0.00000001% distortion doesn't lift my spirits when the music continues to sound hollow.

I rest my case. And will most likely avoid this place when it comes to any actual reviews. Just saying Amir, you have a really bad reputation everywhere else for exactly this reason. Hyper focusing on pointless things, panning some equipment for simply not listening to it enough or because it's not Chi-fi. I mean, really? Please focus on what actually matters - the music - not completely inaudible differences in distortion. But I digress, it might have SOME merit - 0.0000001% distortion might impress your dog, who knows.

For everyone else, do please starting calling me an iFi shill, or things of similar caliber. I fully expect it from you lot. I'm sorry that I actually spend time enjoying my setup rather than worrying about completely trivial nonsense and telling naysayers how they are oh so very wrong on online forums for days on end. Hell of a life that must be...
Well, you've wasted your time creating an account just to post that, here of all places.

It's fine enjoying your system. I'm happy for you. Just don't waste our time, trying to convince people that your sighted listening impressions mean anything to anyone except you.

Your ears are not golden no matter how much you think they might be. And they are certainly not measuring instruments. And just like every other human, the subconscious part of your brain is quite able to lie to the conscious part about what your ears are actually detecting.
 
Last edited:
Well, you've wasted your time creating an account just to post that, here of all places.

It's fine enjoying your system. I'm happy for you. Just don't waste our time, trying to convince people that your sighted listening impressions mean anything to anyone except you.

Your ears are not golden no matter how much. you think they might be. And they are certainly not measuring instruments. And just like every other human, the subconscious part of your brain is quite able to lie to the conscious part about what your ears are actually detecting.
I doubt that you will get a reply. Call it intuition ;)
 
Well... that was quick.
When they go to that effort (8-9 paragraphs) to tell us how much they hate this place. We facilitate their exit…
 
Back
Top Bottom