Are you Euphonophile?

[Cosmik]

Am I correct in thinking that the addition of some of the instantaneous signal amplitude squared to a sine wave gives only second harmonic distortion i.e. is the 'Rolls Royce' of euphonic distortion? So if I pass some music through a squaring transfer function and listen at the relevant amplitude, this would be the pure euphonic contribution? And I can 'zoom in' by boosting the amplitude and listening to that..?

If so, it sounds like sizzling and shrieking to me...

 

[andreasmaaan]

Am I correct in thinking that the addition of some of the instantaneous signal amplitude squared to a sine wave gives only second harmonic distortion i.e. is the 'Rolls Royce' of euphonic distortion? So if I pass some music through a squaring transfer function and listen at the relevant amplitude, this would be the pure euphonic contribution? And I can 'zoom in' by boosting the amplitude and listening to that..?

If so, it sounds like sizzling and shrieking to me...

Could you post more details of what you're doing, and some audio files if possible? Sounds interesting.

 

[andreasmaaan]

Is this the basic procedure you're following? What % 2nd harmonic distortion sounds like shrieking to you?

 

[Cosmik]

Is this the basic procedure you're following? What % 2nd harmonic distortion sounds like shrieking to you?

Remember that there is no such thing as a harmonic distortion generator for a complex signal (without breaking it down into its constituent parts). All there is is a transfer function that imparts harmonic distortion on a sine wave. But music isn't a sine wave, and when you pass music into this transfer function, what you get is a complex signal derived from the intermodulation between the components.

And yes, it shrieks, sizzles, sounds bloody awful.

 

[Cosmik]

Is this the basic procedure you're following? What % 2nd harmonic distortion sounds like shrieking to you?

Thanks for the link. More scholarly than what I'm doing with a bit of C code on each sample in a music file, but basically that. I haven't yet thought about aliasing and stuff, though, so I daren't post any details just yet!

 

[andreasmaaan]

Remember that there is no such thing as a harmonic distortion generator for a complex signal (without breaking it down into its constituent parts). All there is is a transfer function that imparts harmonic distortion on a sine wave. But music isn't a sine wave, and when you pass music into this transfer function, what you get is a complex signal derived from the intermodulation between the components.

And yes, it shrieks, sizzles, sounds bloody awful.

I appreciate this But to try to quantify the % distortion you can take either the approach of representing it as a distortion % when applied to a sine wave, or the actual % distortion when to a complex signal - much like what was done in the study I posted earlier.

 

[pkane]

Here's a good presentation that discusses various harmonic distortions. Even has a schematic for building a tube-based 'distortion box'. Take a look at the powerpoint file for some measurements and observations. Admittedly not scientific research, but generally information that I've corroborated from other, more scientific sources:

http://www.pmillett.com/etf_sod.htm

 

[Cosmik]

Here's a good presentation that discusses various harmonic distortions. Even has a schematic for building a tube-based 'distortion box'. Take a look at the powerpoint file for some measurements and observations. Admittedly not scientific research, but generally information that I've corroborated from other, more scientific sources:

http://www.pmillett.com/etf_sod.htm

At least he acknowledges :

Simple music (vocals, Norah Jones, etc.): – Even distortion can be “harmonious” – Odd distortion can be “gritty” •
Complex music (big band, orchestras): – Any distortion is “muddy” – Odd distortion can mimic increased detail or brightness

If you look at Aural Exciter tricks and tips they often tell you not to apply it to the overall mix, but to apply it to individual 'stems' or tracks.

When using an exciter, it’s often best try it out on the stems of your mix rather than on the whole master bus... Most commonly, exciters or enhancers are used to bring one part towards the front by cutting through the mix. Often vocals, are the most popular application.

So best of luck to all you people wanting to creatively enhance the overall mix! The recordings where it may 'work' are limited, and it may not even 'work' over the whole recording but just a few spots where the mix is simpler.

 

[Fitzcaraldo215]

Well then like those USB cable aficionados maybe you should try it for yourself.

Preferably, sighted like they do. Then you can join their parade.

 

[andreasmaaan]

If you look at Aural Exciter tricks and tips they often tell you not to apply it to the overall mix, but to apply it to individual 'stems' or tracks.

Those aural exciters introduce odd and even harmonics out to a very high order. Their effect is different from the low order harmonics typically desired by tube enthusiasts.

But in any case, the manufacturer's recommendation on how to use a particular distortion unit is no substitute for the science, which tends to show - or at least holds the door wide open for the possibility - that most listeners have a preference for certain kinds and degrees of audible distortion applied to the signal as a whole.

 

[pkane]

At least he acknowledges :

If you look at Aural Exciter tricks and tips they often tell you not to apply it to the overall mix, but to apply it to individual 'stems' or tracks.

So best of luck to all you people wanting to creatively enhance the overall mix! The recordings where it may 'work' are limited, and it may not even 'work' over the whole recording but just a few spots where the mix is simpler.

There is no question that added distortions reduce the fidelity of audio reproduction. Even if they somehow make the sound more subjectively appealing, they are still masking or modifying or destroying some of the original content.

 

[Blumlein 88]

An Aural Exciter is nothing like a euphonic amp or source. Sorry, this is just another straw man. Aural Exciters are excruciatingly bad for music.

 

[Sal1950]

My take on the original threads premise and the follow up posts is this.
Distortion would equal anything that would effect the original signals waveform. IE if it doesn't equal a straighwire with gain, it's distorting.
Euphonic distortions would be anything introduced into the signal the makes the results more pleasing to the listener however it is accomplished.
Way back in the day I did a lot of euphonic coloration. From use of things like tone controls or loudness buttons, to add on boxes. Things like a DBX subharmonic synthesizer added some sweet purrr to the bottom end.
My Phase Linear 1000 dynamic range expander was actually quite useful for vinyl playback, reducing surface noise, removing low end rumble, and improving dynamic range. (strange and sad how back then we were upset at the compression being applied to LP's, now today we praise them because they're crushing Digital sources to death. )

 

[Cosmik]

An Aural Exciter is nothing like a euphonic amp or source. Sorry, this is just another straw man. Aural Exciters are excruciatingly bad for music.

Sorry, that's just a straw man: the thread is about preferring added distortion versus transparency.

An aural exciter is just an up-front 'euphonic' effects box that adds distortion in a more controllable - and honest - way than a boutique audio system. You're not imagining that there's some secret psychoacoustic formula lurking in garage-built amplifiers that the aural exciter people have missed are you?

 

[andreasmaaan]

Sorry, that's just a straw man: the thread is about preferring added distortion versus transparency.

An aural exciter is just an up-front 'euphonic' effects box that adds distortion in a more controllable - and honest - way than a boutique audio system. You're not imagining that there's some secret psychoacoustic formula lurking in garage-built amplifiers that the aural exciter people have missed are you?

Let's leave this diversion to one side.

What's wrong with the science?

 

[tomelex]

Remember that there is no such thing as a harmonic distortion generator for a complex signal (without breaking it down into its constituent parts). All there is is a transfer function that imparts harmonic distortion on a sine wave. But music isn't a sine wave, and when you pass music into this transfer function, what you get is a complex signal derived from the intermodulation between the components.

And yes, it shrieks, sizzles, sounds bloody awful.

See my bolded above, I just want to interject that while sound waves are not "perfect" sinewaves, they are very close to sinewaves, all sound is, air reacts to a disturbance with just a bit of second harmonic distortion built in and thus music is composed of these slightly imperfect sinewaves indeed. Pure sinewaves are used for testing because they contain only the fundamental, so they are certainly valid for testing, but more of them all at once and at the proper intervals does better exercise any piece of audio reproduction gear.

 

[Cosmik]

I think we need to consider what the signal is, and what 'distortion' is.

The signal is a system's internal representation of captured air pressure variations. It needn't actually be understandable by looking at it: it could be in scrambled form, for example, or encoded or multiplexed in some other way. It could be transported as a stream of numbers. But it just so happens that in most audio systems, for at least part of the way it is represented as a voltage proportional to air pressure.

What can happen to such a signal on the way to the speakers is purely a by-product of it being a voltage transported through cables and other electrical components. Those side-effects are not acoustic, nor 'musical', in nature; they are electrical effects i.e. in another domain from music and acoustics altogether.

The true hi-fi enthusiast understands this: the aim has got to be to get the signal through unchanged. Transformations of the content of the signal are for the people who create it or, if trying to salvage a corrupted or poor existing recording, for specialists to use deliberate techniques to break the signal into its constituent parts and process those 'scientifically'.

That messing around with cables, transistors and diodes and the composite waveform might impart some meaningful musical or acoustic transformation of the content of the signal is a fantasy. Talking about distortion in the time domain as if it is distortion in the frequency domain ("second harmonic") is an example of this wishful thinking. All the magic box has done is maybe crush up the bottom of the electrical waveform and stretch out its top - a completely meaningless transformation in terms of music and acoustics. This transformation doesn't even attempt to analyse what is being represented in the signal, nor what a meaningful transformation might be for the contents of this particular signal.

If you think you are a euphonicist, consider: are you trying to change the signal, or the contents of the signal?

If it's the former, then what in the world of acoustics and music is 'a signal'? It doesn't exist. You are trying to change a meaningless internal by-product of an electrical system in order to improve your musical experience.

If it's the latter, what are you doing to access the contents of the signal in order to process them? Nothing. You are merely distorting 'the signal' without accessing its contents, and hoping that this will translate into something meaningful acoustically or musically. Again, an exercise in blind faith and wishful thinking.

At best, you may get a buzz (literally) from some sections of some recordings leading to you changing the music you listen to in order to pander to your system's foibles. At worst you corrupt everything you listen to with meaningless 'noise' (if only it sounded as good as noise) between now and when your critical listening abilities finally leave you.

 

[svart-hvitt]

I think we need to consider what the signal is, and what 'distortion' is.

The signal is a system's internal representation of captured air pressure variations. It needn't actually be understandable by looking at it: it could be in scrambled form, for example, or encoded or multiplexed in some other way. It could be transported as a stream of numbers. But it just so happens that in most audio systems, for at least part of the way it is represented as a voltage proportional to air pressure.

What can happen to such a signal on the way to the speakers is purely a by-product of it being a voltage transported through cables and other electrical components. Those side-effects are not acoustic, nor 'musical', in nature; they are electrical effects i.e. in another domain from music and acoustics altogether.

The true hi-fi enthusiast understands this: the aim has got to be to get the signal through unchanged. Transformations of the content of the signal are for the people who create it or, if trying to salvage a corrupted or poor existing recording, for specialists to use deliberate techniques to break the signal into its constituent parts and process those 'scientifically'.

That messing around with cables, transistors and diodes and the composite waveform might impart some meaningful musical or acoustic transformation of the content of the signal is a fantasy. Talking about distortion in the time domain as if it is distortion in the frequency domain ("second harmonic") is an example of this wishful thinking. All the magic box has done is maybe crush up the bottom of the electrical waveform and stretch out its top - a completely meaningless transformation in terms of music and acoustics. This transformation doesn't even attempt to analyse what is being represented in the signal, nor what a meaningful transformation might be for the contents of this particular signal.

If you think you are a euphonicist, consider: are you trying to change the signal, or the contents of the signal?

If it's the former, then what in the world of acoustics and music is 'a signal'? It doesn't exist. You are trying to change a meaningless internal by-product of an electrical system in order to improve your musical experience.

If it's the latter, what are you doing to access the contents of the signal in order to process them? Nothing. You are merely distorting 'the signal' without accessing its contents, and hoping that this will translate into something meaningful acoustically or musically. Again, an exercise in blind faith and wishful thinking.

At best, you may get a buzz (literally) from some sections of some recordings leading to you changing the music you listen to in order to pander to your system's foibles. At worst you corrupt everything you listen to with meaningless 'noise' (if only it sounded as good as noise) between now and when your critical listening abilities finally leave you.

You touched upon it, but maybe only indirectly?

The point is, hifi is not about music. Music is just one example of what may go through a hifi system. An abstract debate on hifi should consider it for what it is: Information theory. Nothing more.

 

[andreasmaaan]

@Cosmik and @svart-hvitt, your idealism is admirable, but it doesn't account for anecdotal reports or, more importantly, the scientific evidence.

It might be difficult to palate that most listeners seem to prefer their music-containing signal to be distorted at the point of reproduction, but no amount of arguing from first principles will change that fact if it is indeed so.

 

[Cosmik]

@Cosmik and @svart-hvitt, your idealism is admirable, but it doesn't account for anecdotal reports or, more importantly, the scientific evidence.

It might be difficult to palate that most listeners seem to prefer their music-containing signal to be distorted at the point of reproduction, but no amount of arguing from first principles will change that fact if it is indeed so.

As I said before: if, in the 'science', the system deviates from neutral before the distortion is applied, then preference for distortion of the signal is not being tested. If all the tests were done with Kii Threes or D&D 8C I might be persuaded that it was attempting to be more scientific, but even they are not completely neutral.

And then there are other human factors: who chooses the musical examples; at what volume; how do you know people are not just responding to novelty, etc. This isn't science, but just casting around randomly without any theoretical justification. You may as well try scrambling the signal, distorting that, and then descrambling it. Or do sum-and-difference on the stereo, distort those, and then restore the L-R channels. Digitise the signals and pass them through a magic lookup table with 'golden ratio' properties. Etc.