How do you know what you want from your system?

[Keith_W]

9 minutes ago, davewantsmoore said:

I think these end up being one and the same thing.

 

They probably are, but I should have prefaced my remarks by saying that sometimes I don't know what the measurements are. So my only recourse is to listen. That is what I listen for. 

[Satanica]

47 minutes ago, Keith_W said:

I do NOT include distortion figures in here because some types of distortion are very difficult to hear, and even if they are present I have found that they do not negatively affect transparency. If anything, distortion can add richness to the sound. As you know, in my system I am able to dial in as much distortion as I like. I may have even demonstrated it to you. Everybody who has heard it seems to prefer some distortion dialled in. I do however have a preference for how distortion is added to the system. I would prefer that my electronics and speakers don't add it. I prefer to add it in myself via a VST. 

Hi Keith, I'm listening to the PKHarmonic plugin as I type this. The problem I'm having is that the sound is significantly louder with it engaged, even with modest settings. Is this just the added harmonics actually increasing the overall preceived level or do you know if the plugin is doing this? I don't think it's really possible for me to assess it as is, because with it engaged everything sounds louder and of course I prefer that one.

Edited July 17, 2023 by Satanica

[Keith_W]

1 minute ago, Satanica said:

Hi Keith, I'm listening to the PKHarmonic plugin as I type this. The problem I'm having is that the sound is significantly louder with it engaged, even with modest settings. Is this just the added harmonics actually increasing the overall preceived level or do you know if the plugin is doing this? I don't think it's really possible for me to assess it as is, because with it engaged everything sounds louder and of course I prefer that one.

 

I asked the author the exact same question because it sounds louder to me as well. His answer is that PKHarmonic automatically calculates the RMS of the output with the distortion added, and lowers the volume accordingly. Any perceived increase in volume is subjective and due to the added distortion. 

You could confirm if this is true or not by playing some pink noise and measuring the electrical output at the DAC. I can't do it because I don't own a volt meter

[tripitaka]

1 minute ago, Keith_W said:

 

I asked the author the exact same question because it sounds louder to me as well. His answer is that PKHarmonic automatically calculates the RMS of the output with the distortion added, and lowers the volume accordingly. Any perceived increase in volume is subjective and due to the added distortion. 

You could confirm if this is true or not by playing some pink noise and measuring the electrical output at the DAC. I can't do it because I don't own a volt meter

 

Your video was the first time I'd heard of PKharmonic

 

I'm all up for a way to enjoy valves without buying a valve amp... But you already use valve Amps!? So is this not just about even-order harmonics in your case?

[Satanica]

5 minutes ago, Keith_W said:

 

I asked the author the exact same question because it sounds louder to me as well. His answer is that PKHarmonic automatically calculates the RMS of the output with the distortion added, and lowers the volume accordingly. Any perceived increase in volume is subjective and due to the added distortion. 

You could confirm if this is true or not by playing some pink noise and measuring the electrical output at the DAC. I can't do it because I don't own a volt meter

Well, I don't think it's lowering the volume nearly enough.

[Keith_W]

26 minutes ago, tripitaka said:

 

Your video was the first time I'd heard of PKharmonic

 

I'm all up for a way to enjoy valves without buying a valve amp... But you already use valve Amps!? So is this not just about even-order harmonics in your case?

 

Valve amps are for the horns only, the rest of the frequency spectrum has no valve amps. And as mentioned, at less than 1W output, there would be hardly any distortion. 

 

23 minutes ago, Satanica said:

Well, I don't think it's lowering the volume nearly enough.

 

Yep, I agree. I should mention this to him. 

[tripitaka]

1 minute ago, Keith_W said:

 And as mentioned, at less than 1W output, there would be hardly any distortion.

 

Ah! I has missed that part

 

So a silly Q, do you think your lovely valve amps still do add something else that the PKharmonic isn't able to? (Could you get by ok even with SS amps given your processing options)

[andyr]

2 hours ago, Keith_W said:

I do NOT include distortion figures in here because some types of distortion are very difficult to hear, and even if they are present, I have found that they do not negatively affect transparency. If anything, distortion can add richness to the sound.

 

Absoloodle, Keith - and you already have a lot of H2 from your Cary amps.

 

2 hours ago, Keith_W said:

As you know, in my system I am able to dial in as much distortion as I like. I may have even demonstrated it to you. Everybody who has heard it seems to prefer some distortion dialled in. I do however have a preference for how distortion is added to the system. I would prefer that my electronics and speakers don't add it. I prefer to add it in myself via a VST. 

 

But I think you said that some people have preferred the sound without your added VST distortion?

 

Because they already hear the H2 from your Cary amps.

 

Maybe borrow Con's Purify monoblocs and try the experiment again - I would venture to suggest that no-one will prefer the sound without the added VST H2.

 

[frednork]

1 hour ago, davewantsmoore said:

If you mean distortion of the frequency response, then that basically the only remaining explanation.... but you implied the this was matched (how careful were you really?.... did you actually compare SPL overlaid?)

No @davewantsmoore, I never thought I would need to "prove" what I am saying. What I can say is that over many measurements and corrections and curves with both the D's and A/B's,  the A/B's are just noticeably better and I no longer suffer from the "distortion" I was unhappy with.  I accept that this may not be enough to convince you though.

[Keith_W]

31 minutes ago, andyr said:

Maybe borrow Con's Purify monoblocs and try the experiment again - I would venture to suggest that no-one will prefer the sound without the added VST H2.

 

Yeah, could be an experiment to try. 

 

1 hour ago, tripitaka said:

So a silly Q, do you think your lovely valve amps still do add something else that the PKharmonic isn't able to? (Could you get by ok even with SS amps given your processing options)

 

Maybe they do. But at the moment I am not inclined to investigate because I am really happy with them and they are not going anywhere. Also - in my system, DSP produces far more dramatic changes than any change in components could hope to do. And I am talking about orders of magnitude audible difference, almost as much difference as changing a speaker. A few 0.1% of THD in the amp does not bother me, especially since they already perform to my satisfaction. 

[Grizaudio]

1 hour ago, frednork said:

No @davewantsmoore, I never thought I would need to "prove" what I am saying. What I can say is that over many measurements and corrections and curves with both the D's and A/B's,  the A/B's are just noticeably better and I no longer suffer from the "distortion" I was unhappy with.  I accept that this may not be enough to convince you though.

I understand exactly what you are saying. Because I have heard and experienced what you have described. 
 

I dare not hypothesise about why this is the case, because I’m not a masochist. 

 

In listening isolation, the impact is difficult to resolve, but in direct comparison it’s audible..

Edited July 17, 2023 by Grizaudio

[davewantsmoore]

1 hour ago, frednork said:

I never thought I would need to "prove" what I am saying.

Only to yourself.... <shrug>

 

1 hour ago, frednork said:

I accept that this may not be enough to convince you though.

I was only saying that if the FR was not matched extremely closely, that that would be the suspect.    Like, actively matched between the two amplifiers, with the same target response..... ie. really try to make the difference go away by matching the FRs.

 

I'm not saying you didn't hear anything.... just talking about the cause.

[davewantsmoore]

3 hours ago, Keith_W said:

You could confirm if this is true or not by playing some pink noise and measuring the electrical output at the DAC. I can't do it because I don't own a volt meter

Use your microphone.

 

3 hours ago, tripitaka said:

So a silly Q, do you think your lovely valve amps still do add something else

I've used all manner of custom made valve amps on my horns, from simple ones to complex ones, to OTL, as well as some interesting/unusual SS amps .... they produce different frequency responses, due to how they interact with the speaker load.    If you correct for these different responses, the difference in sound largely goes away..... so I really think the answer to your question is, in general, no.

 

The differences between amps can be even larger on other types of speakers (especially lower efficiency and more varying electric load).

[Grizaudio]

9 hours ago, davewantsmoore said:

I've used all manner of custom made valve amps on my horns, from simple ones to complex ones, to OTL, as well as some interesting/unusual SS amps .... they produce different frequency responses, due to how they interact with the speaker load.    If you correct for these different responses, the difference in sound largely goes away..... so I really think the answer to your question is, in general, no.

 

The differences between amps can be even larger on other types of speakers (especially lower efficiency and more varying electric load).

 

I agree with this statement, but one experience is no more valid than the other. 

The characteristics @frednork is communicating are not FR related. This is not the delta. 

 

I use class D, I've used Firstwatt class A, and a good number of vintage class A/B. 

All offer pro's and con's... but the differences are not just apparent in FR. 

 

To get back on topic.... 

 

I think we all want our system to sound and look right for us. 

This is likely very different person to person. 

At the end of the day, for me its all about the music and music discovery. The better my system sounds, the more I listen, and the more music I discover and enjoy. 

 

[Bass13]

11 hours ago, Keith_W said:

 

Also - in my system, DSP produces far more dramatic changes than any change in components could hope to do. And I am talking about orders of magnitude audible difference, almost as much difference as changing a speaker.

 

Could you please elaborate on your DSP setup, you can PM me...if you don't want to discuss here.

Edited July 17, 2023 by Bass13
typo

[davewantsmoore]

10 minutes ago, Grizaudio said:

I agree with this statement, but one experience is no more valid than the other. 

For sure.   Results can depend a lot on "what you did" .... as opposed to "the way things are".

 

10 minutes ago, Grizaudio said:

All offer pro's and con's... but the differences are not just apparent in FR. 

But what else is there?

 

10 minutes ago, Grizaudio said:

At the end of the day, for me its all about the music and music discovery. The better my system sounds, the more I listen, and the more music I discover and enjoy. 

I can't imagine a person who wouldn't identify with this....

[frednork]

9 hours ago, davewantsmoore said:

Only to yourself.... <shrug>

 

yeah, thats cool. I just hadnt considered it at the time and it would have been easy as the mic was already out.

 

But on that, what level of difference would be required to say they are def different, especially at high frequencies . Might be of use for others as well as saving me some time if  trying to show a difference. 

 

For instance: here are some measurements below. At what point can we confidently say they will sound different.

Is this different?

 

Is this different?

 

 

 

 

9 hours ago, davewantsmoore said:

 

I was only saying that if the FR was not matched extremely closely, that that would be the suspect.    Like, actively matched between the two amplifiers, with the same target response..... ie. really try to make the difference go away by matching the FRs.

 

I'm not saying you didn't hear anything.... just talking about the cause.

All good,  was just acknowledging you would need to see the data,  thus ensuring something else obvious wasnt happening.

[davewantsmoore]

43 minutes ago, frednork said:

For instance: here are some measurements below. At what point can we confidently say they will sound different.

There is too much diffraction in these measurements to see.

 

If we were trying to measure the diffraction.... then that would be fine .... but if we are trying to answer the question of "what is the movement of the transducer" (ie. before we get over run by everything that happens after it moves) .... then we can't see anything much clearly.

 

For your actual measurement we need to appreciate that the frequency resolution is not infinite... and so where we have very big dips, and they look different between the two measurements, that we can;t really know if they are --- for example if there was 5 (x-axis) data point making up the dip (relatively low resolution), then it is just that the points are slightly different, making the dips looks like they don't extends as far, or whatever.   (in practise, these dips are literally cancellations - and with an infinite resolution would extend toward zero, and are likely the same for each..... assuming the mic has not moved at all) .... but this is not relevant to answering the question mentioned.

 

The first chart is only covering six percent of one octave .... and so, it's impossible to say if the cause is simple measurement variation, or if the mic moved a tiny amount, or....

The second chart is similarly "can't say anything" ... if we are wanting to answer question about how the transducer moved, then all this diffractions occludes that.

 

 

Not that I would apply this thinking to your charts here (as discuss above) ..... but we should remember that if there is 1dB difference... then this is an error of 12% in amount of movement.    It is a resonance, which extends in time (distorted frequency response is simply another view of distorted time response, in a speaker).    Things that people routinely call totally insignificant is very much not.

[davewantsmoore]

58 minutes ago, frednork said:

All good,  was just acknowledging you would need to see the data,  thus ensuring something else obvious wasnt happening.

Sure.

The thing is, that if it is not the frequency response.... then what is it?   There is precious little else.

 

 

Anyways... don't get me wrong.  I'm not trying to tell you you're stupid because all amplifiers sound the same.    Just why you (or anyone) got the result they did....   Carver is correct  (I think many people would surprised to hear me side there).

 

[frednork]

14 minutes ago, davewantsmoore said:

There is too much diffraction in these measurements to see.

 

If we were trying to measure the diffraction.... then that would be fine .... but if we are trying to answer the question of "what is the movement of the transducer" (ie. before we get over run by everything that happens after it moves) .... then we can't see anything much clearly.

 

For your actual measurement we need to appreciate that the frequency resolution is not infinite... and so where we have very big dips, and they look different between the two measurements, that we can;t really know if they are --- for example if there was 5 (x-axis) data point making up the dip (relatively low resolution), then it is just that the points are slightly different, making the dips looks like they don't extends as far, or whatever.   (in practise, these dips are literally cancellations - and with an infinite resolution would extend toward zero, and are likely the same for each..... assuming the mic has not moved at all) .... but this is not relevant to answering the question mentioned.

 

The first chart is only covering six percent of one octave .... and so, it's impossible to say if the cause is simple measurement variation, or if the mic moved a tiny amount, or....

The second chart is similarly "can't say anything" ... if we are wanting to answer question about how the transducer moved, then all this diffractions occludes that.

 

 

Not that I would apply this thinking to your charts here (as discuss above) ..... but we should remember that if there is 1dB difference... then this is an error of 12% in amount of movement.    It is a resonance, which extends in time (distorted frequency response is simply another view of distorted time response, in a speaker).    Things that people routinely call totally insignificant is very much not.

Ok, no worries, so obviously these sweeps are taken at the listening posiition so reflections and diffraction are unavoidable unless we have our own anechoic set up. 

 

So then how? should I be windowing? nearfield? both? something else? Just trying to figure out a viable method of what to do if I want to do it. Otherwise I  show something and then it may be , "but you should have done "this"  to show a diff"

[davewantsmoore]

37 minutes ago, frednork said:

So then how?

Yes... if you want to answer the question of how the transducer moved..... then you can measure very very close to the transducer, to eliminate all the diffraction.

 

The "window" should always be set to eliminate unwanted reflections (sometimes you might want them, sometimes you might not).   So in this case you would think to "make the window short" .... but reducing the window reduces the frequency resolution....  so, by measuring very very close, you should have no diffraction within the frequencies of interest... and should not need to have a short window, but as always we can look into the data, and see if there is any reflections being measured - and then think about what impact they have or if anything can be done about them.

 

48 minutes ago, frednork said:

so obviously these sweeps are taken at the listening posiition so reflections and diffraction are unavoidable

Yes    : ) 

To answer the question of is the speaker moving the same with different amps .... measuring in the listening position is problematic.

 

48 minutes ago, frednork said:

Otherwise I  show something and then it may be , "but you should have done "this"  to show a diff"

I understand.   It is not an easy thing..... and many many times, data tells us nothing relevant about the question, or leads to wrong conclusions.

[Keith_W]

3 hours ago, Bass13 said:

 

Could you please elaborate on your DSP setup, you can PM me...if you don't want to discuss here.

 

Sure, you can follow the link in my signature to have a look at my system thread. You can ask any questions you like over there. 

[Keith_W]

2 hours ago, frednork said:

 

yeah, thats cool. I just hadnt considered it at the time and it would have been easy as the mic was already out.

 

But on that, what level of difference would be required to say they are def different, especially at high frequencies . Might be of use for others as well as saving me some time if  trying to show a difference. 

 

For instance: here are some measurements below. At what point can we confidently say they will sound different.

Is this different?

 

Is this different?

 

 

 

 

All good,  was just acknowledging you would need to see the data,  thus ensuring something else obvious wasnt happening.

 

You know, looking at your graphs reminds me of an experience I had with @sir sanders zingmore. @aris had brought over a valve amp and we listened to it back to back with his SS amp (I think it was a Sanders Magtech). I know it is an audiophile trope to say "night and day" difference, but the difference between the two amps was pretty unbelievable. The entire tonality was different, to me the magnitude of difference was the same as switching between DSP filters. I could barely believe it so I asked if he could take some sweeps. I was quite surprised to see that the sweeps looked the same between the two amps. 

[Grizaudio]

Not withstanding my previous comments....  

But... if a frequency sweep is done before and after with everything the same, a difference of 1-2db will be significant enough.

Especially through the Upper mid > Presence range.  

 

Edited July 18, 2023 by Grizaudio

[davewantsmoore]

1 hour ago, Keith_W said:

I was quite surprised to see that the sweeps looked the same between the two amps. 

Something went wrong.

"Tonality" is a function of the frequency response (distortion from electronics or speakers doesn't do that) .... and so there must have been a (significant) difference.

 

This can be said especially confidently when you are describing a test, where the predicted result is "there will be a difference in the frequency response".