• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required as is 20 years of participation in forums (not all true). Come here to have fun, be ready to be teased and not take online life too seriously. We now measure and review equipment for free! Click here for details.

Is distortion in electronics meaningful ?

o2so

Member
Joined
Sep 9, 2019
Messages
65
Likes
30
#1
Quick question.
I do not understand why we fuss so much about distortion in a DAC which might be at -100db rather than -130db when the best speakers that have been tested here have second harmonic distortion at around -70db. How does it work, do they sum up? Or because the speakers has distortion at -70db it does not really matter what the DAC does?
 

KeithPhantom

Senior Member
Forum Donor
Joined
May 8, 2020
Messages
301
Likes
208
#2
Quick question.
I do not understand why we fuss so much about distortion in a DAC which might be at -100db rather than -130db when the best speakers that have been tested here have second harmonic distortion at around -70db. How does it work, do they sum up? Or because the speakers has distortion at -70db it does not really matter what the DAC does?
Humans are pretty bad at detecting nonlinearities (distortion) by ear, so we resort to using measuring equipment that far exceeds our ability to critically hear to reveal what we cannot detect. The marginal benefits of having better measurements decrease quickly and pretty much flatten at -80 dBFS considering the use cases home audio equipment is used, so there is a limit to the audibility of distortion measurements.
 

bunkbail

Senior Member
Forum Donor
Joined
Feb 27, 2018
Messages
428
Likes
456
#3
Until a certain threshold, nope, they're not meaningful. We are just measurebating beyond that threshold.
 
OP
O

o2so

Member
Joined
Sep 9, 2019
Messages
65
Likes
30
Thread Starter #4
Humans are pretty bad at detecting nonlinearities (distortion) by ear, so we resort to using measuring equipment that far exceeds our ability to critically hear to reveal what we cannot detect. The marginal benefits of having better measurements decrease quickly and pretty much flatten at -80 dBFS considering the use cases home audio equipment is used, so there is a limit to the audibility of distortion measurements.
So I gather that any component with a distortion at - 80db will be basically indistinguishable from one with distortion at -130db? If this is the case, what is the point of measuring this difference or even caring about it?
Anyway my question was a bit more about the speaker/component relative importance in defining distortion. If the speakers go at best at -70db, who cares if the dac or the amp measure -80db rather than -130db? Am I missing something?
 
Last edited:

KeithPhantom

Senior Member
Forum Donor
Joined
May 8, 2020
Messages
301
Likes
208
#5
If this is the case, what is the point of measuring this difference or even caring about it?
Technical progress, pure engineering progress. Also, it is easy and cheap to make equipment that measures -120 dBFS, so why not?

If the speakers goes at best at -70db
Don't take this number so seriously, they are quoting distortion from a single nominal value at a single volume level, not the average (or a total presented in a graph) across the frequency range. And yes, the distortion coming from transducers is way more important than the rest, since electronics are able to measure better and being more consistent.
 

exaudio

Member
Forum Donor
Joined
Jun 11, 2018
Messages
46
Likes
71
#6
I'm speculating, but I'd say most of the frequent visitors to this forum look at measurements more for signs of excellence in engineering than from the perspective of audibility.

That said, a while back forum member @flipflop compiled a bunch of audibility research into a post titled Audibility Thresholds of Amp and DAC measurements.

At the end of his post he summarized the results into lenient and strict thresholds. What I took away from his post was the levels I use when interpreting @amirm's FFT spectrum plots showing SINAD. They are as follows:

-66 dBFS Lenient threshold for distortion
-85 dBFS Lenient threshold for noise
-120 dBFS Strict threshold for full transparency / provable inaudibility

One other thing to note about distortion is that, while everybody agrees that intermodulation distortion (IMD) is bad, some of the more traditional audiophiles out there find a certain level of harmonic distortion to be pleasant.
 

Racheski

Addicted to Fun and Learning
Forum Donor
Joined
Apr 20, 2020
Messages
706
Likes
877
Location
Chicago
#7
Technical progress, pure engineering progress. Also, it is easy and cheap to make equipment that measures -120 dBFS, so why not?
I don't think "easy" is the right adjective. We only have a handful of DACs and Amps that achieve -120 dBFS and they are all less than 2 years old I think.
 

carlosmante

Active Member
Joined
Apr 15, 2018
Messages
165
Likes
116
#8
So I gather that any component with a distortion at - 80db will be basically indistinguishable from one with distortion at -130db? If this is the case, what is the point of measuring this difference or even caring about it?
Anyway my question was a bit more about the speaker/component relative importance in defining distortion. If the speakers go at best at -70db, who cares if the dac or the amp measure -80db rather than -130db? Am I missing something?
Am I missing something?
You are missing bragging rights.
 

solderdude

Major Contributor
Joined
Jul 21, 2018
Messages
6,341
Likes
11,754
Location
The Neverlands
#9
Quick question.
I do not understand why we fuss so much about distortion in a DAC which might be at -100db rather than -130db when the best speakers that have been tested here have second harmonic distortion at around -70db. How does it work, do they sum up? Or because the speakers has distortion at -70db it does not really matter what the DAC does?
1: As KeithPhantom already mentioned people aren't as sensitive to distortion as they might think it is.
2: Noise and frequency response and some other aspects may be more important than distortion as long as it isn't reaching absurd levels for this day and age.
3: We measure because we want to know the electrical performance this includes distortion.
4: A DAC may not be as sound determining as the interaction between an amp and a complex load but still we want to know both.

I would fuss more over FR linearity and noise in a DAC more so than distortion but all aspects have importance and need to be characterized.
When everything but distortion levels are the same the -100dB distortion and -130dB distortion levels will not be of practical importance. No one claims they do, they may feel it is important to them (piece of mind) though.
 

sergeauckland

Major Contributor
Forum Donor
Joined
Mar 16, 2016
Messages
1,673
Likes
3,620
Location
Suffolk UK
#12
I'm speculating, but I'd say most of the frequent visitors to this forum look at measurements more for signs of excellence in engineering than from the perspective of audibility.


One other thing to note about distortion is that, while everybody agrees that intermodulation distortion (IMD) is bad, some of the more traditional audiophiles out there find a certain level of harmonic distortion to be pleasant.
To the first sentence, I agree, for me it's a sign of correct design, not audibility.

As to the second, I can't see how one can have harmonic distortion without intermodulation distortion, as they're caused by the same mechanism, non-linearity of transfer function. That distortion can be pleasant, of that I have no doubt. Valve amplifiers of the SET variety and/or low feedback designs have both.

S.
 

Wombat

Major Contributor
Joined
Nov 5, 2017
Messages
6,014
Likes
4,860
Location
Australia
#13
If it is audible and not meant to be there, then yes.
 

RayDunzl

Major Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
10,121
Likes
8,663
Location
Riverview FL
#15
1: As KeithPhantom already mentioned people aren't as sensitive to distortion as they might think it is.
My PC, for work purposes has an alarm clock which emits eleven cycles of a cuckoo clock sound at 11pm and midnight, among a few other times, on days of weeks. It takes about 11 seconds to play. Cuckoo, cuckoo, cuckoo, etc...

Audio Buddy came over on Monday evening. He brought some new (old) music to which to listen attentively.

The Focusrite was still set to "mix" the PC sound output with whatever else was going to the DAC.

Twice, the cuckoo clock sounded during tunes, and twice I noticed it, and twice Audio Buddy didn't.

???
 

solderdude

Major Contributor
Joined
Jul 21, 2018
Messages
6,341
Likes
11,754
Location
The Neverlands
#16
because you are 'trained' to recognize it where audiobuddy may have thought it belonged in the recording or ignored it ?
Most likely the cuckoo clock was not mixed in at -70dB either.
 

RayDunzl

Major Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
10,121
Likes
8,663
Location
Riverview FL
#17
because you are 'trained' to recognize it where audiobuddy may have thought it belonged in the recording or ignored it ?
Most likely the cuckoo clock was not mixed in at -70dB either.
I "recognized" it, yes.

It was approximately the same level as the instruments accompanying the current solo (guitar, bass drums, piano), like another player accompanied for eleven seconds or so, in neither the same key nor the same time signature.

It played, I looked at him while it was playing.
He's just bobbing along with Wes Montgomery.
"Did you hear that?"
"Hear what?"
"The cuckoo clock."
"No."

An hour later, it played again. I looked at him while it was playing.
He's bobbing to Charlie Christian (which turned out to be more like jitterbug and less emphasis on the guitar, which I found less satisfying, despite the amazing instumentalists)
"Did you hear that?"
"Hear what?"
"The cuckoo clock."
"No."

Surprised me, anyway.

The point being "people aren't as sensitive to distortion as they might think" as you mentioned above.

Carry on.
 

Harmonie

Addicted to Fun and Learning
Joined
May 30, 2020
Messages
506
Likes
229
Location
France
#18
I'm speculating, but I'd say most of the frequent visitors to this forum look at measurements more for signs of excellence in engineering than from the perspective of audibility.
I really hope that there are more reasons for attending this site
 

Wombat

Major Contributor
Joined
Nov 5, 2017
Messages
6,014
Likes
4,860
Location
Australia
#19
I'm speculating, but I'd say most of the frequent visitors to this forum look at measurements more for signs of excellence in engineering than from the perspective of audibility.

That said, a while back forum member @flipflop compiled a bunch of audibility research into a post titled Audibility Thresholds of Amp and DAC measurements.

At the end of his post he summarized the results into lenient and strict thresholds. What I took away from his post was the levels I use when interpreting @amirm's FFT spectrum plots showing SINAD. They are as follows:

-66 dBFS Lenient threshold for distortion
-85 dBFS Lenient threshold for noise
-120 dBFS Strict threshold for full transparency / provable inaudibility

One other thing to note about distortion is that, while everybody agrees that intermodulation distortion (IMD) is bad, some of the more traditional audiophiles out there find a certain level of harmonic distortion to be pleasant.

First sentence - run a poll.
 

Similar threads

Top Bottom