The No.1 Website for Pro Audio
Why do Cab IR's seem lacking?
Old 2 weeks ago
  #61
Lives for gear
Feels like I'm screaming into the void here...

Quote:
Originally Posted by Raaphorst View Post
not from the speaker.
Quote:
Originally Posted by chris661 View Post
I did. Here's the write-up: https://www.grimshawaudio.com/guitarspeakers

IRs do not capture the harmonic distortion of a driven speaker, which ought to be very obvious in listening. At moderate levels, there was 40% THD in some areas.

Chris
There are graphs in the link.

10s of percent THD at moderate levels, and approx. 100% THD at high levels at low frequencies.
Should be very very audible.

Chris
Old 2 weeks ago
  #62
Lives for gear
 

Quote:
Originally Posted by Yuri Kogan View Post
But I did become a member of another 4 forums, much more professional environment, so I am giving it a go. Will see. People seem knowledgeable and the conversations respectful. Stuff to learn, not much agro...
Yuri, would you mind sharing what the other forums are? (You can pm me if that’s better.) I am quite the insomniac these days and Facebook and reading about gear is all I can do while laying here wishing I was asleep.
Old 2 weeks ago
  #63
Lives for gear
 
Yuri Kogan's Avatar
 

Quote:
Originally Posted by chris661 View Post
Feels like I'm screaming into the void here...





There are graphs in the link.

10s of percent THD at moderate levels, and approx. 100% THD at high levels at low frequencies.
Should be very very audible.

Chris
Your feeling is correct. I doubt he is interested in any facts. Doesn't matter what you demonstrate its of no interest.
I on the other hand am. And all respect for spending all this time to do a scientific test
Old 2 weeks ago
  #64
Quote:
Originally Posted by chris661 View Post
Feels like I'm screaming into the void here...

There are graphs in the link.

10s of percent THD at moderate levels, and approx. 100% THD at high levels at low frequencies.
Should be very very audible.

Chris
Interesting. THD might also be caused by the mic or a combination of mic and speaker. Would love to see the overtones it caused, mixture of odds and evens?

Might do some tests myself as well. Thanks!
Old 2 weeks ago
  #65
Lives for gear
It's possible, but unlikely, that the distortion was caused by the mic - the speaker was very obviously distressed when it was doing the sweeps. I had ear defenders on, but could still tell it was unhappy.

Details of which overtones are in the graphs - they show up to the 10th harmonic.

I haven't heard of an e906 being overloaded before, but I'm happy to swap a different mic in. Options include:
- Beyer M88TG
- Beyer M201TG
- Beyer TG-X930
- Sennheiser e904
- Sennheiser e935

I also have some MC930s, an original C3000, an SE X1D, and a couple of SE4400a. Those all top out around 140dB, which is roughly what this speaker was putting out at the mic position.
IMO, a decent dynamic mic (like the e906) shouldn't have a problem here, even though we're right at the edge of what most condenser mics will take.

Chris
Old 2 weeks ago
  #66
Lives for gear
 
Yuri Kogan's Avatar
 

Quote:
Originally Posted by chris661 View Post
It's possible, but unlikely, that the distortion was caused by the mic - the speaker was very obviously distressed when it was doing the sweeps. I had ear defenders on, but could still tell it was unhappy.

Details of which overtones are in the graphs - they show up to the 10th harmonic.

I haven't heard of an e906 being overloaded before, but I'm happy to swap a different mic in. Options include:
- Beyer M88TG
- Beyer M201TG
- Beyer TG-X930
- Sennheiser e904
- Sennheiser e935

I also have some MC930s, an original C3000, an SE X1D, and a couple of SE4400a. Those all top out around 140dB, which is roughly what this speaker was putting out at the mic position.
IMO, a decent dynamic mic (like the e906) shouldn't have a problem here, even though we're right at the edge of what most condenser mics will take.

Chris

Mic distortion and overload is very easily mitigated with the right mic technique. Or "improved" with an incorrect mic technique. But its not as controllable with a speaker. Can be done to a point, but the end results may be not what was expected. (anyone blew a speaker? )
Old 2 weeks ago
  #67
Gear Guru
 
monkeyxx's Avatar
Quote:
Originally Posted by Yuri Kogan View Post
True, they are a low quality speaker. These days by design. However I remember playing a Goodman's equipped AC30 (silver speakers) and it was a very different tone to say Celestion Blues (which are pretty bad actually). Speakers were selected to compensate for deficiencies of the guitar amps (not great quality either compared to HiFi). For example AC30 is a very dark sounding amp. But with AlNicos like Blues, which have a piercing hi-end they work, Try them with a more mid-range speaker and its very unimpressive. Marshall has a mid bump. So with the greenbacks it overdrives and sounds grumpy. This may be enhanced by playing dynamics, where you get random-ish distortion with harder picking. Same AC30 sounds anaemic at anything but flat-out where it over-saturated speakers. Ditto for Marshalls.
HiFi speakers are designed for high headroom, before the get into a non-linear region. You really need to push then hard to get there. It was harder with tube hifi amps as they used transformers and that interaction contributed the a smaller linear region. With guitars, players like that distortion and the speakers are purposely left unrefined. We play them at the edge of blowing up, cause we like that distortion.
When you drive a SS or digital amp into a speaker you get closer to hifi situation. Its like playing some sort of recording into a clean speaker to preserve it. That's why the dynamics are different and you don't get the tonal variations you get with speaker/PA interaction.
Speakers like any other device behave unpredictably when pushed to the limit. And guitarists like it. They say the sound is more alive. Push a small amp into a 100W speaker - no distortion, boring sound. Its because you are not in that unpredictable region of operation. Turn a hot amp down and you get the CLEAN tone- same thing you are working in the predictable range of the speaker. Its the truth which has been known since day one and is still true. I cannot overdrive my studio monitors with any guitar preamp recording and therefore 90% of the time afterwards guys want to record an amp, with all the nuances. And the monitors just reproduce the non-linearity the mic picks up from the speaker.
There are millions of papers written on this and speaker design. Any "tests" run here, by un-initiated would only open a discussion of twisting the truth. But fair enough, I presume people in quarantine have nothing better to do then try to disprove the obvious.
Ill just go and get a load of popcorn - may be entertaining after a long day of work
Your experience with AC30s has exactly nothing to do with my own experience.

My AC30 can be VERY bright and it sounds good at any volume.

Not sure what was wrong with your AC30s.
Old 2 weeks ago
  #68
Quote:
Originally Posted by nedorama View Post
I think it comes down to what will your guitar track(s) ultimately be used for? If it's solo guitar unaccompanied, that may be an issue. If the guitars are in a mix with other instruments and a good reverb is added along with EQ, dynamics and effects, do you notice? When I hear Pete Thorn's demos, either into his iso Cab or into IRs, it's always in a mix, and to me it sounds good.

Michael Carnes, who developed a lot of the Lexicon Reverbs we all love and then went on to found Exponential Audio, talks about how convolving reverbs that use IR's don't always work as well as mathematical models - because with math models you can add in variability for nonlinearity.

I think it would be interesting to compare today's IR's to different speaker emulators used in the 90s and 2000's when session players were all using rack gear - the Palmer, etc. and see if we're better today than those circuit based products. I personally used an ADA MicroCab II with either an ADA MP1 or a Mesa V-Twin Rack.
I really don't think thjat "sounding good" in the end product has much, if anything, to do with it.

"Sounding food" is in the ear of the beholder. That which sounds good to one person may not sound good to another. What sounds good to one person in a particular context may not sound good to the same person in a different context.

I think what IS important is twofold - (1) how does the sound work with the producer's vision? and (2) how does the sound make the musician react emotionally while he's playing? Different tonalities can make musicians more or less inspired, and may affect actual choices made in their playing. That's something that no "end user" can really understand.
Old 2 weeks ago
  #69
Quote:
Originally Posted by monkeyxx View Post
Your experience with AC30s has exactly nothing to do with my own experience.

My AC30 can be VERY bright and it sounds good at any volume.

Not sure what was wrong with your AC30s.
I think youi guys are "talking around" each other - I don't really see what the one statement has to do with the other.

I DO want to point out that over there years there have been quite a few versions of the AC30, many of which sound quite different from each other. In fact Vox has been known to make 3 or more different sounding models of "AC30" at the same time!
Old 2 weeks ago
  #70
Quote:
Originally Posted by chris661 View Post
Feels like I'm screaming into the void here...





There are graphs in the link.

10s of percent THD at moderate levels, and approx. 100% THD at high levels at low frequencies.
Should be very very audible.

Chris
Interesting article, better than a lot I'm seen but still woefully incomplete.

The author seems to be missing a very important part of the equation, which is that guitar signals are not steady sweep tones, and you can't measure or predict the real world behavior of a speaker from steady-state (level-wise) tones.

And you can't predict the behavior of a speaker, especially a guitar speaker, using only sine waves. You need complex waveforms.

And how a guitar speaker reacts dynamically is a very, very important aspect of its sound.
Old 2 weeks ago
  #71
Lives for gear
 

Quote:
Originally Posted by chris661 View Post
I did. Here's the write-up: https://www.grimshawaudio.com/guitarspeakers

IRs do not capture the harmonic distortion of a driven speaker, which ought to be very obvious in listening. At moderate levels, there was 40% THD in some areas.

Chris
Appreciate your contribution here!
Definitely two separate deals to find and present relevant info/facts, and have people actually accept those facts.
The later seems to be much more challenging these days.


Devs and engineers have surely added harmonic distortion algorithms to sims/modelers and the like by now. It’s just that the IR, which is what is in question, is apparently not able to represent Harmonic distortion like a real speaker/speaker cab.
At least as we currently know it.
The dynamic IRs possibly could.

The majority of amp sim plugins I’ve used have way too much noise from the distortion.
Basically all of them at high gain settings are way too noisey and even with a good gate if you’re just rolling the volume on you’re guitar or playing single note or a particular style, you still hear this overwhelming blanket of white noise.
I don’t get this from hardware modelers I have used near as much.
I wonder if this has to do with harmonic distortion being added.

No clue but I wouldn’t be surprised.

I think it’s all too easy for us to forget.
We are living in a very small sliver of time and this stuff will only get better.
Where we are with these technologies and more importantly the ability to run it with little latency as possible is pretty amazing.
And we’ve come a long way.
The Axe fx sounds pretty good to my ears.
Im seriously considering buying one.
I use a little modeling amp that sounds good to my ears.
But I do love me some tube amps.
It’s like a little vacation for me to play my tube amps.
Old 2 weeks ago
  #72
Quote:
Originally Posted by Tnevz View Post
...
Basically all of them at high gain settings are way too noisey and even with a good gate if you’re just rolling the volume on you’re guitar or playing single note or a particular style, you still hear this overwhelming blanket of white noise.
I don’t get this from hardware modelers I have used near as much.
I wonder if this has to do with harmonic distortion being added.
...
But I do love me some tube amps.
It’s like a little vacation for me to play my tube amps.

Likely because the DI is a little noisy and too wide band....

...or the hardware units have a good gate?

The transients often sound different in modelers. That maybe to due to intermodulation in the gain stages and the speaker in a tube amp. The hardware units likely model the V-I curve on, at least, the preamp tubes. That will help a lot with attacks and that spit then die down - like the tube combo.... Then you don't need the white noise.

I think the good digital models I've heard are more impressionist than realist. Meaning that the overall sound of a particular amp and cab is modeled in a way that makes it sound like it, not assembled from well copied parts. The extra tweaking really is an art.


-tINY

Old 2 weeks ago
  #73
Lives for gear
 
Yuri Kogan's Avatar
 

Quote:
Originally Posted by monkeyxx View Post
Your experience with AC30s has exactly nothing to do with my own experience.

My AC30 can be VERY bright and it sounds good at any volume.

Not sure what was wrong with your AC30s.
Well, I have a bunch of them from several years of manufacture. Starting with 1963. So I would say its a pretty extensive experience.
The reason the Blue speakers are popular is because they make the amp sound bright enough. Try it with some other speaker and the standard sound is quite dark. That's why the top-boost was invented, to make them more palatable.
If you are talking about the modern C versions - well its a different thing, which sounds pretty thin at anything below 80% of full. Not overly bright and sounding pretty ho-ham with the basic speaker. With the blues it gets more into the classic top-boost tone, but still not there. The HW versions are a different beast, sounding like a cross of Matchless DC30 and a later 60's AC30. I would not call it very bright at all.
If you want bright, you are looking at vintage BF Princeton reverb, or SF Twin reverb (unless you drive it real hard).
The best balanced AC30 i have is the 63 and the version made by Bruno amps on bright channel. Coupled with blues it sounds nice/fizzy because the speakers have very little headroom. If you can find an old AC30 with Googmans you will really know how dark it is.
Old 2 weeks ago
  #74
Gear Guru
 
monkeyxx's Avatar
I have a Vox AC30 HW head. I play through a 2x12 of "Red Fang" Eminence speakers. I have never owned Celestion Blues they are on the long list of things to try.

Also I have very limited experience with vintage AC30s. The one I played at a shop was indeed slightly dark, darker than the JTM45 next to it. I preferred the JTM45.

I am also open to the idea that people hear things differently, and that's all good.

Just responding to your generalizations in the first post, because my "specific" experience with the amp I just listed is not like what you originally said.

not that it matters, I'm not debating, just discussing.

On the topic of the thread I am really pleased to have a Physics perspective on IR modeling. Will continue to read into that tomorrow. Great thread.
Old 2 weeks ago
  #75
Lives for gear
 
Yuri Kogan's Avatar
 

Quote:
Originally Posted by monkeyxx View Post
I have a Vox AC30 HW head. I play through a 2x12 of "Red Fang" Eminence speakers. I have never owned Celestion Blues they are on the long list of things to try.

Also I have very limited experience with vintage AC30s. The one I played at a shop was indeed slightly dark, darker than the JTM45 next to it. I preferred the JTM45.

I am also open to the idea that people hear things differently, and that's all good.

Just responding to your generalizations in the first post, because my "specific" experience with the amp I just listed is not like what you originally said.

not that it matters, I'm not debating, just discussing.

On the topic of the thread I am really pleased to have a Physics perspective on IR modeling. Will continue to read into that tomorrow. Great thread.
All god. You have also explained why you hear ac30 as bright - Red Fang has a big emphasis in the 2.5-5khz range (the presence region).
I also prefer the Marshalls to the Vox. They have more presence and guts. But Vox's work better with pedals if you can stand the fizz. And yes AC30s have changed a lot over the years, so the fact that they are dark applies more to the classic early versions and the standard channel on the ones made in the 60's and 70's. After that - it was different every year.
So I guess there isn't much of a disagreement here. Discussion is always good
Old 2 weeks ago
  #76
Lives for gear
 
Yuri Kogan's Avatar
 

Quote:
Originally Posted by tINY View Post

Likely because the DI is a little noisy and too wide band....

...or the hardware units have a good gate?

The transients often sound different in modelers. That maybe to due to intermodulation in the gain stages and the speaker in a tube amp. The hardware units likely model the V-I curve on, at least, the preamp tubes. That will help a lot with attacks and that spit then die down - like the tube combo.... Then you don't need the white noise.

I think the good digital models I've heard are more impressionist than realist. Meaning that the overall sound of a particular amp and cab is modeled in a way that makes it sound like it, not assembled from well copied parts. The extra tweaking really is an art.


-tINY

Yeap. And that's why I say its wrong to treat them as a replacement of the modelled amp. But as their own entity it maybe just the ticket.
Old 2 weeks ago
  #77
Lives for gear
 

Quote:
Originally Posted by tINY View Post

Likely because the DI is a little noisy and too wide band....

...or the hardware units have a good gate?

The transients often sound different in modelers. That maybe to due to intermodulation in the gain stages and the speaker in a tube amp. The hardware units likely model the V-I curve on, at least, the preamp tubes. That will help a lot with attacks and that spit then die down - like the tube combo.... Then you don't need the white noise.

I think the good digital models I've heard are more impressionist than realist. Meaning that the overall sound of a particular amp and cab is modeled in a way that makes it sound like it, not assembled from well copied parts. The extra tweaking really is an art.


-tINY


I do need to check out my DI path. Now that I think about it I’ve been using a preamp lately that I’ve left plugged in for various things.
But the level going in isn’t hitting the red and I’ve adjusted input to the plugins as well.

Ive heard that hardware has better gates but I seem to get less noise over all with hardware modelers. Even when turning the gates off And running the plugged in modelers to my plugged in interface connected to my plugged in computer.

Yeah I can appreciate modeling for sure.
I love some old guitars and amps, but I grew up pretty poor with solid state, modelers, cheap gear overall I was lucky to get so I’m no purist about it.
Old 2 weeks ago
  #78
Lives for gear
Quote:
Originally Posted by John Eppstein View Post
Interesting article, better than a lot I'm seen but still woefully incomplete.

The author seems to be missing a very important part of the equation, which is that guitar signals are not steady sweep tones, and you can't measure or predict the real world behavior of a speaker from steady-state (level-wise) tones.

And you can't predict the behavior of a speaker, especially a guitar speaker, using only sine waves. You need complex waveforms.

And how a guitar speaker reacts dynamically is a very, very important aspect of its sound.
Thanks for taking the time to read my work.

I must point out that I did, in fact, run the sweeps at a variety of different levels, ranging from 1W to 256W.

I found this on the Wiki page about inter-modulation distortion:
Quote:
IMD is only distinct from harmonic distortion in that the stimulus signal is different. The same nonlinear system will produce both total harmonic distortion (with a solitary sine wave input) and IMD (with more complex tones).
If you could expand on why it is that complex waveforms are required (mathematically, it doesn't appear to be the case), that would make interesting reading.

Regards,
Chris
Old 2 weeks ago
  #79
Here for the gear
 

Quote:
Originally Posted by tINY View Post

For a lot of guitar sounds in recordings (or live on stage) the non-linearities of a guitar speaker are essential to the final sound. And, because impulse response modeling is a linear process, it can't capture the dynamics of an actual guitar speaker. In fact, everything about most electric guitar timbre is dependent on non-linear, low-fidelity phenomena.

Here is an actual quote by Ian White of Celestion:

"With hi-fi and pro PA speakers, we can develop finite element models and see on a screen what they're going to be doing,” Ian says. "But that's almost impossible with guitar speakers because they're so non-linear. It's difficult enough to predict how a guitar speaker is going to sound when you're feeding it a clean guitar signal, but it's the distorted stuff that really gets the cone modes going. It doesn't need to be that loud to do that — just the complexity of the signal really gets the tone humming."
I think you overstate the case by arguing that everything about guitar sound is about distortion. Frequency response is also important, sometimes more important. But yes, distortion is often very important and a digital simulation of traditional guitar speakers would need to include it.

(I'm using the word distortion in the engineering sense: with a sine wave input, any difference from sine wave at the output is distortion.)

What Ian White says is curious. He seems to say that guitar speakers are significantly nonlinear even with low input signals. That's quite surprising to me. I can believe that the distortion is greater relative to a PA speaker but it is surely progressive with the input signal level.

The other thing he says is also hard to understand "distorted stuff really gets the cone modes going. It doesn't need to be that loud to do that — just the complexity of the signal really gets the tone humming." Cone modes aren't in themselves non-linear.

I'm not sure what the quote means but assuming he's talking about distortion, what properties unique to guitar drivers introduce significant distortion with low input power and what's the mechanics of the distortion?

To get a better understanding I think characterizing the distortion would be a good first step. Measure harmonic and inter-modulation distortion with various test signals, frequency sweeps and power levels. I think REW can do that (it's free) and MiniDSP sells a calibrated mic for $75.

If the distortion level is large at normal input power to the cab then you're right, cab IR is not enough to simulate the effects of the cab.

Another kind of test would be to choose a popular traditional cab, get two of them, replace the driver(s) in one with PA drivers of similar frequency response, and do a double blind subjective listening test with matched loudness level. Probably want to run that test at more than one input power level (but without driving the speakers to break up).

Measurements can really clarify some discussions.

(I'm very interested because I'll need to buy new gear at some point and would prefer to go down the digital road for convenience.)
Old 2 weeks ago
  #80
Lives for gear
Well, if you read the link that I've posted multiple times already, you'll see that I've already done most of those tests. I took a 12" guitar speaker and ran sweeps at progressively higher levels, from 1W up to 256W.

The last sweep was particularly abusive, but even moderate levels showed fairly high THD.



There are two sorts of distortion: linear and non-linear.
An example of linear distortion is a non-flat frequency response.
Non-linear distortion is something that adds frequency components that weren't there originally - harmonic distortion etc.

It's important to read the full interview with Ian White - snippets can be taken out of context.


Chris

PS - The average measurement mic will be in full clipping if you put it near a guitar speaker going full bore. They usually top out around 120dB, and the guitar speaker I was testing was doing about that (higher in places) at 1m from the cone.

PPS - Just in case you really did miss it, here's the link: https://www.grimshawaudio.com/guitarspeakers
Old 2 weeks ago
  #81

The cone modes are resonant sub-systems of the cone. You need not excite them with the frequency they resonate at either.

A pulse as stimulus and a waterfall plot would reveal these much better than a sine sweep.

As to non-linearities: IMD is the same effect as an RF mixer circuit. Other frequencies arise besides the two or more that are input. This is due the the non-linear nature of the circuit or speaker.

https://en.wikipedia.org/wiki/Intermodulation




-tINY

Old 2 weeks ago
  #82
Lives for gear
 
grannis's Avatar
Why would we want IMD? Surely that’s ehy we use power chords, to eliminate the bothersome dissonance that is accentuated by the IMD.
Old 2 weeks ago
  #83
Here for the gear
 

Quote:
Originally Posted by chris661 View Post
Well, if you read the link that I've posted multiple times already, you'll see that I've already done most of those tests. I took a 12" guitar speaker and ran sweeps at progressively higher levels, from 1W up to 256W.

The last sweep was particularly abusive, but even moderate levels showed fairly high THD.

There are two sorts of distortion: linear and non-linear.
An example of linear distortion is a non-flat frequency response.
Non-linear distortion is something that adds frequency components that weren't there originally - harmonic distortion etc.

It's important to read the full interview with Ian White - snippets can be taken out of context.

Chris

PS - The average measurement mic will be in full clipping if you put it near a guitar speaker going full bore. They usually top out around 120dB, and the guitar speaker I was testing was doing about that (higher in places) at 1m from the cone.

PPS - Just in case you really did miss it, here's the link: https://www.grimshawaudio.com/guitarspeakers
I really did miss it. I read the OP and some following answers but stopped because they were getting confusing. Given that the OP was interesting, I gave my 2c, which I thought I gave in a constructive spirit. And no I didn't read the full interview, only what was excerpted on this forum. I apologize for being hasty.

Thanks for this and sharing the results. Great work. It's very interesting.

If you can eq it then I would call it coloration rather than distortion. If you get different frequencies out than you put in then I'd call it distortion. So what you call a linear distortion, I would call a coloration, even if it does change the shape of a non-sine wave. But this is just a personal language preference.

Your first and second graphs are striking in showing how much linear distortion (coloration) the speaker introduces at any power level.

The second shows that the coloration is largely independent of power level except at the highest drive. I don't know how high that was relative to the speaker's specified rating but your comment in the conclusions ("the speaker wouldn't survive long with sustained power input of that magnitude") and above ("particularly abusive") suggest it was significantly over driven. At that sort of level you'd expect the motor, cone, suspension... reach mechanical limits and introduce significant non-linearity.

The third plot of harmonic distortion at 4 W drive shows 1.91% TDH and it's effectively all 2nd harmonic. 4W was a little less than the second highest drive level in the set of curves in the first two plots, iiuc. 1.9% is more than a good HiFi speaker would produce but not radical. That much THD isn't exactly rare in HiFi speakers or HiFi tube power amps. And isn't 2% 2nd HD modest in the realm of electric guitar amps even for what one might call a "clean" setting?

So I would draw a few conclusions from these results.

First, the Eminence Man O War 16Ohm doesn't behave in surprising way. It's highly colored but we expect that. I guess it has has higher harmonic distortion than a PA driver. Driven to (what I assume are) its mechanical limits the picture changes but I'd expect that too.

Second, coloration (linear distortion in your terms) is clearly very important at all drive levels and is largely independent of power. Even at the highest drive you tested, the shape of the frequency response isn't drastically different. If so then I'd think IR should be able to do a lot of work in a digital sim.

At moderately high drive there's significant 2nd HD, probably noticeable but probably also something that an amp or (I imagine) a digital amp sim could also accomplish. It's normal for HD increase with signal level in tube amps in a fairly smooth way. Do amp modelers do this?

Does adding IR after adding HD approximately simulate a speaker alone? To a point I would think so but not at the highest drive where you showed that the color changed, though it's not clear if that's the important change that we hear.

So for some applications it may be satisfactory to combine amp sim with IR cab sim. Otoh, it doesn't look like cab IR without adding some kind of distortion is generally a winning idea. Also, listening to a cab IR with a truly clean input signal is probably not a good way to evaluate how useful it might be.

PS. The MiniDSP UMIK-1 USB calibrated measurement microphone is spec'd as Max SPL for 1% THD @ 1kHz: 133dB SPL @ 0dB analog gain setting. Iirc, REW reads its calibration file. Sennheiser e 906 spec sheet doesn't mention max SPL though I'm sure it's suitable for guitar speakers. It has ~5 dB peak around 3-5 KHz and bass rolls off under 300 Hz. But mentally adjusting your curves doesn't change the story they tell. https://assets.sennheiser.com/global...ication_EN.pdf
Old 2 weeks ago
  #84
Here for the gear
 

Quote:
Originally Posted by grannis View Post
Why would we want IMD? Surely that’s ehy we use power chords, to eliminate the bothersome dissonance that is accentuated by the IMD.
A non-linear transfer function produces inter-modulation. So if non-linearities in loudspeakers are important in determining their sound quality then measuring IMD may be useful in characterizing differences between speakers. It's a common measurement and a lot of test gear supports it.

If you want distortion, e.g. from an overdriven tube circuit, as guitarists often do, then inter-modulation is one of the inevitable results. High gain leads to power chords.

You can have distortion without that effect by using a divided pickup and a separate distortion circuit for each string, then mix the signals in a clean (linear) system. The Roland GR-300 did this (in addition to being a synth) but the result is completely different from plugging a guitar into a dirty amp.
Old 2 weeks ago
  #85
Quote:
Originally Posted by thefsb View Post
...If you want distortion, e.g. from an overdriven tube circuit, as guitarists often do, then inter-modulation is one of the inevitable results. High gain leads to power chords.

You can have distortion without that effect by using a divided pickup and a separate distortion circuit for each string, then mix the signals in a clean (linear) system....

This is why most rock guitar players don't play jazz chords much. And why jazz players like very different amps.

The other way around the IM issue is an array of bandpass filters with an overdrive on each output. Then you recombine the signals. Source Audio has a pedal or two that do this. It's a very different sound, but somehow sounds a lot like an actual overdrive - just doesn't sound as dirty(?). Of course the only practical way is to use a digital processor (unless you really love fine-tuning filters....).




-tINY

Old 2 weeks ago
  #86
Lives for gear
Quote:
Originally Posted by tINY View Post

The cone modes are resonant sub-systems of the cone. You need not excite them with the frequency they resonate at either.

A pulse as stimulus and a waterfall plot would reveal these much better than a sine sweep.

As to non-linearities: IMD is the same effect as an RF mixer circuit. Other frequencies arise besides the two or more that are input. This is due the the non-linear nature of the circuit or speaker.

https://en.wikipedia.org/wiki/Intermodulation




-tINY

If you take the Fourier Transform, a swept sine measurement can reveal an impulse measurement. The difficulty with using a pulse is getting it loud enough to show up significant low-frequency content, while simultaneously avoiding over-driving the speaker higher in the frequency range.
A sine sweep gets around all that, and still collects all the data. With just the sine sweep, REW produces waterfall plots, among other things.

The Wiki link you've posted says this:
Quote:
IMD is only distinct from harmonic distortion in that the stimulus signal is different. The same nonlinear system will produce both total harmonic distortion (with a solitary sine wave input) and IMD (with more complex tones).
Which suggests that a sine sweep can tell us everything that speaker is doing at a particular drive level. Directly measuring IMD isn't necessary - the information is all there in the harmonic distortion graphs - it just requires a bit of interpretation.


Quote:
Originally Posted by thefsb View Post
I really did miss it. I read the OP and some following answers but stopped because they were getting confusing. Given that the OP was interesting, I gave my 2c, which I thought I gave in a constructive spirit. And no I didn't read the full interview, only what was excerpted on this forum. I apologize for being hasty.
No problem. Apologies for being a bit snappy, but it feels like the discussion in this thread has gone around and around in circles and it's been more than a little frustrating.

Glad you found the link useful. The highest power level was a 256W sine sweep, while the speaker is rated for 120W. Since it's a (moving) pure tone, this test is particularly abusive at lower frequencies: when the speaker manufacturer does a thermal power test, they use pink noise which spreads the energy out across the range. A sine sweep puts all that energy at one frequency, which means cone excursion ends up very large, and you can quickly push into mechanical clipping.

I read a Sennheiser paper about maximum SPL testing of dynamic mics - they put a 421 in the throat of a high-power 2" compression driver capable of producing [email protected], and found the distortion was below 1%.
That doesn't test for low frequencies, though, so I might give them a little prod in that direction.

... Or just test it myself. Wouldn't be difficult to make a LF compression driver.

Chris
Old 2 weeks ago
  #87
If speakers show large amounts of distortion, this means studio monitors have this problem too.

My monitors don't distort. My Sennheiser HD600 is also free from distortion.
Old 2 weeks ago
  #88
Quote:
Originally Posted by chris661 View Post
Which suggests that a sine sweep can tell us everything that speaker is doing at a particular drive level. Directly measuring IMD isn't necessary - the information is all there in the harmonic distortion graphs - it just requires a bit of interpretation.

Yes. But, an easy process to model the cabinet for virtual instruments will not be able to do this.

Also, playing with the swept-sine waterfall tools years ago, I noticed that these systems had some severe trade-offs between time and frequency resolution. Maybe with higher sample rates and more bit depth it's not as bad.... I was using 48k/16 at the time.



-tINY

Old 2 weeks ago
  #89
Lives for gear
The next step up is amp modeling. Get's you one step closer. Kemper is the classic example.
Old 2 weeks ago
  #90
Lives for gear
 
Yuri Kogan's Avatar
 

Quote:
Originally Posted by Raaphorst View Post
If speakers show large amounts of distortion, this means studio monitors have this problem too.

My monitors don't distort. My Sennheiser HD600 is also free from distortion.
All speakers do. My $18k/each monitors have distortion. Doesn't manifest itself that much within the headroom parameters, but still there. Drive them hard and you will REALLY hear it. With inbuilt D-class PAs you dont get much of it as you are limited to the range which shows less distortion.
Guitar speakers - different beast.
Topic:
Post Reply

Welcome to the Gearslutz Pro Audio Community!

Registration benefits include:
  • The ability to reply to and create new discussions
  • Access to members-only giveaways & competitions
  • Interact with VIP industry experts in our guest Q&As
  • Access to members-only sub forum discussions
  • Access to members-only Chat Room
  • Get INSTANT ACCESS to the world's best private pro audio Classifieds for only USD $20/year
  • Promote your eBay auctions and Reverb.com listings for free
  • Remove this message!
You need an account to post a reply. Create a username and password below and an account will be created and your post entered.


 
 
Slide to join now Processing…
Search this Thread
Search this Thread:

Advanced Search
Forum Jump
Forum Jump