Why do some IR's and DynaCab's seem to compress more than others?

NotGlenn said:
Now that I know some of the finer points of clipping vs. compression, I feel like it should be renamed to speaker clipping so that I am not lead astray again lol
Click to expand...
I know that Cliff and Jay have had more than a few "discussions" on various deep topics that they didn't agree on.

I'll leave it to Cliff to call it what he sees fit :D
 
side note: if you do your own IRs and want to go deeper into the rabbit hole, drive your mic preamp into distortion. It changes the result, and can sound nice once baked into the IR.
 
aziz said:
side note: if you do your own IRs and want to go deeper into the rabbit hole, drive your mic preamp into distortion. It changes the result, and can sound nice once baked into the IR.
Click to expand...
That’s a good tip. Easy to listen with headphones without preamp vs with added preamp distortion if your preamp has a headphone output. I thought the Mic+Pre IR shoot method would neutralize that, but realized that’s not the case.

Shooting your own IR’s is a lot of fun if you have a great sounding cabinet and good ear.
 
aziz said:
side note: if you do your own IRs and want to go deeper into the rabbit hole, drive your mic preamp into distortion. It changes the result, and can sound nice once baked into the IR.
Click to expand...

It changes the result, but not in the way you might be expecting. The clipping of the sampled signal is interpreted as additional higher frequency content due to the added harmonics. Those harmonics will be factored into all of the signal processed by the IR regardless of its input level. IR's are linear so the processing is the same regardless of input level. That means, the frequency response will be the same as the clipped signal, but at any input level, not just the signal above the original clipping point like in the reference signal.
 
Last edited:
There is a method to capture non-linear behavior using a series of IR like captures at various input levels and interpolating the results based on input level. That's called profiling... I'd wager that speaker systems are generally linear enough that the extra processing overhead required by the profiling method is not really worth it just for speakers and cabs.
 
randocaster said:
This is one of the posts I was thinking of:

"Speakers "compressing" is a myth. Compression is a change in gain in response to the envelope (which only can occur over several cycles) of a signal. Speaker nonlinearities are all responses to peaks that occur within a single cycle. Any such nonlinearities would therefore constitute clipping. In case anyone doesn't know this, I'll point out that clipping and compression sound very, very different."

And this:
"

No. They have a limited range of linear travel, but driving them beyond that range causes clipping, not compression. I refer you to the post of mine you quoted for the difference between the two.



Soft clipping is still clipping; it is quite different from compression.



You're unintentionally misusing the phrase "mechanical resistance." Mechanical resistance is proportional to velocity of motion, not displacement from rest. In fact, velocity - and therefore the force due to mechanical resistance - reaches zero at maximum displacement. The phrase you need is spring constant. While it is true that the restorative force due to the spring constant of the suspension increases with displacement from rest, it is also true that this force is linear and therefore causes no nonlinear behavior, as long as the cone's displacement lies within its linear range. Displacement beyond the linear range in a guitar speaker only occurs at low frequencies and causes obvious and objectionable distortion. That sound is not part of any iconic guitar tone."

YMMV and all that.
Click to expand...
Speaker compression is not a myth. The voice coil gets hot. Very hot. The resistance is proportional to the temperature. As the temperature increases the resistance increases. The tempco of copper is about 0.4%/deg. C. If the voice coil temperature increases, say, 50 deg. C then the resistance of the voice coil will increase by 20%. If you doubt this think about how hot a 20W resistor gets. Or a 50W. 99% of the power delivered to a speaker is dissipated as heat. The 1% left is acoustic power. If you put 50W into a speaker 49.5W are converted to heat in the voice coil.

For simplicity let's assume that the speaker is driven by a power amplifier with zero output resistance. The power into the speaker is Vrms^2 / R. If the resistance increases by 20% then the power decreases by about 17% (roughly 1.6dB). That's very audible. Voice coils can actually get much hotter than this, even above the boiling point of water. Some speakers are rated to 150 C or more. That's why they use Kapton formers as plastic formers would melt. Most loudspeaker failures are usually due to voice coil overheating.

This is modeled under the Dynamics page of the Amp block. You can set the amount of compression (voice coil temperature increase) and the thermal time constant. The default values are typical of a 12" guitar speaker.

Guitar speakers also exhibit nonlinear impedance characteristics. Guitar speakers are designed so that the voice coil exits the magnetic gap under normal use. This causes compression when driven by a tube power amp because the magnitude of the impedance decreases and therefore the voltage from the power amp decreases causing a gentle compression. This is modeled by the Speaker Compliance parameter.

Finally guitar speakers exhibit nonlinear distortion behavior. There are two types of distortion. These are modeled by the Speaker Drive and Speaker Thump parameters.
 
FractalAudio said:
Speaker compression is not a myth. The voice coil gets hot. Very hot. The resistance is proportional to the temperature. As the temperature increases the resistance increases. The tempco of copper is about 0.4%/deg. C. If the voice coil temperature increases, say, 50 deg. C then the resistance of the voice coil will increase by 20%. If you doubt this think about how hot a 20W resistor gets. Or a 50W. 99% of the power delivered to a speaker is dissipated as heat. The 1% left is acoustic power. If you put 50W into a speaker 49.5W are converted to heat in the voice coil.

For simplicity let's assume that the speaker is driven by a power amplifier with zero output resistance. The power into the speaker is Vrms^2 / R. If the resistance increases by 20% then the power decreases by about 17% (roughly 1.6dB). That's very audible. Voice coils can actually get much hotter than this, even above the boiling point of water. Some speakers are rated to 150 C or more. That's why they use Kapton formers as plastic formers would melt. Most loudspeaker failures are usually due to voice coil overheating.

This is modeled under the Dynamics page of the Amp block. You can set the amount of compression (voice coil temperature increase) and the thermal time constant. The default values are typical of a 12" guitar speaker.

Guitar speakers also exhibit nonlinear impedance characteristics. Guitar speakers are designed so that the voice coil exits the magnetic gap under normal use. This causes compression when driven by a tube power amp because the magnitude of the impedance decreases and therefore the voltage from the power amp decreases causing a gentle compression. This is modeled by the Speaker Compliance parameter.

Finally guitar speakers exhibit nonlinear distortion behavior. There are two types of distortion. These are modeled by the Speaker Drive and Speaker Thump parameters.
Click to expand...

Candidate for the Tech Notes subforum?
 
giantslayer said:
IR’s are more than just an EQ curve. There is also a time component. Heck, IR’s can even be used to create reverb.
Click to expand...
That's totally true, but they're still linear.
And I have some news for you: EQs have a time component as well, every type of EQ filters is created by applying group delay, i.e. some amount of delay to a band of frequencies which is what's causing the phase shift that alters the frequency response.
That's the reason why linear phase EQ can't be made without adding some latency to the whole signal.

PS: an interesting video about this

 
Last edited:
giantslayer said:
IR’s are more than just an EQ curve. There is also a time component. Heck, IR’s can even be used to create reverb.
Click to expand...
"EQ curve" is semantics. It's a filter. In this case usually an Infinite Impulse Response (IIR) filter. There are two types of filters: Infinite Impulse Response (IIR) and Finite Impulse Response (FIR). An "IR" as we use the term is an FIR. It's a filter and, hence, could be said to be an "EQ curve".

The name "filter" arose because it filters out certain frequencies and lets others pass. Both an IIR and FIR accomplish this. The difference is that an FIR has no poles, only zeros. An IIR can have both poles and zeros. An IIR can achieve similar performance to an FIR with a lower order thus making it more efficient to implement. IIRs, however, typically have nonlinear phase response whereas an FIR can be linear phase.
 
steadystate said:
Um, that's not how it wotks.
Click to expand...
Depends on how you deconvolve. If you use "classic" deconvolution the distortion products will affect the IR to some degree.

If you use matched-filter deconvolution the distortion products are pushed out in time and separated from the linear IR. Our IR Capture utility gives you the option of both methods. Classic is recommended if you're not distorting things much. You can use Matched Filter to obtain an IR from a distorting tube amp.
 
FractalAudio said:
Speaker compression is not a myth. The voice coil gets hot. Very hot. The resistance is proportional to the temperature. As the temperature increases the resistance increases. The tempco of copper is about 0.4%/deg. C. If the voice coil temperature increases, say, 50 deg. C then the resistance of the voice coil will increase by 20%. If you doubt this think about how hot a 20W resistor gets. Or a 50W. 99% of the power delivered to a speaker is dissipated as heat. The 1% left is acoustic power. If you put 50W into a speaker 49.5W are converted to heat in the voice coil.

For simplicity let's assume that the speaker is driven by a power amplifier with zero output resistance. The power into the speaker is Vrms^2 / R. If the resistance increases by 20% then the power decreases by about 17% (roughly 1.6dB). That's very audible. Voice coils can actually get much hotter than this, even above the boiling point of water. Some speakers are rated to 150 C or more. That's why they use Kapton formers as plastic formers would melt. Most loudspeaker failures are usually due to voice coil overheating.

This is modeled under the Dynamics page of the Amp block. You can set the amount of compression (voice coil temperature increase) and the thermal time constant. The default values are typical of a 12" guitar speaker.

Guitar speakers also exhibit nonlinear impedance characteristics. Guitar speakers are designed so that the voice coil exits the magnetic gap under normal use. This causes compression when driven by a tube power amp because the magnitude of the impedance decreases and therefore the voltage from the power amp decreases causing a gentle compression. This is modeled by the Speaker Compliance parameter.

Finally guitar speakers exhibit nonlinear distortion behavior. There are two types of distortion. These are modeled by the Speaker Drive and Speaker Thump parameters.
Click to expand...
So, when I want to achieve the sound of a Celestion Blue being driven really hard, where you can actually hear the speaker compressing, it sounds like, I should really be reaching for Speaker Compliance rather than Speaker Compression?
 
Smittefar said:
So, when I want to achieve the sound of a Celestion Blue being driven really hard, where you can actually hear the speaker compressing, it sounds like, I should really be reaching for Speaker Compliance rather than Speaker Compression?
Click to expand...
No, speaker compression is the right one for that. Turning compliance up simulates a speaker that has been worn down or “broken in” over time. A little goes a long way.
 
steadystate said:
Um, that's not how it wotks.
Click to expand...
For sure it does, i’ve done hundreds of those with mic in one spot and just tweaking preamp gain into different levels of distortion. It changes the resulting eq, of course it doesn't capture the distortion itself. Monitoring the actual sound is useless, as the result is unpredictable. Thats why its rabbit hole stuff.
 
Last edited:
You must log in or register to reply here.
Back
Top Bottom