Firmware supplied IR's

The greatest short-coming, IMO, is... time invariant, linear.
That is not to fault IRs.. absolutely not!!! What they can provide now that we have DSP like this available is great. And Cliff has done a brilliant job compensating for the time variant, non-linear bits. And the entire shebang wrapped up in a nice package? Great! However.. ;)

(that last bit was me joking btw. I have been playing every day now since getting this grasp on IRs. AlbertA has been simply brilliant!)
 
When I read these papers on the mathematics behind the various techniques to create an IR that represents a linear (and time-invariant) system (something relatively very simple), then I think about the complexity of the math behind techniques to create the transfer function of a non-linear system (and a time variant one at that, at least based on what I've read about the "G2" tube models and how real tubes behave) I'm humbled.

Take that a step further and put it all in a box, pack it with virtual circuit models of tube amplifiers, and now include some type of algorithm that creates a transfer function to replicate distortion and tones simply by "listening" to a snippet of a sound...

It's a great day in age for guitar players and musicians. This is how technology was meant to be used. It's unfortunate mankind see's so many destructive uses. I hope my kid has the opportunity to experience all this one day. He's not yet 10 months old, but I suspect the capabilities of the AXE-FX II will be inside a cell phone by the time he's playing the guitar.
 
What a great thread. Most of it is way over my head, but interesting nonetheless. Seeing Spock's picture in Nikki-k's sig, I picture you all looking like that as I read your posts. "Facinating" :)
 
What a great thread. Most of it is way over my head, but interesting nonetheless. Seeing Spock's picture in Nikki-k's sig, I picture you all looking like that as I read your posts. "Facinating" :)
lol.. the real me:
nikki horns1.jpg
 
  • Like
Reactions: jon
lol.. the real me:
nikki horns1.jpg

For some reason that's not how I pictured you. I expected someone more "nerdy" looking from your perfect grammar, wide vocabulary, mathematical knowledge and overall intelligence.

I'll have to post the real me sometime. I'm in the process of attempting to grow a beard for the first time in my life, but my (long) hair is brown with no gray and the damn beard is coming up white! Yipes! WTF is with that? I look like a bum right now who lost his razor.:lol
 
lol!! That was from last year, maybe a month after my 45th birthday. I do actually wear glasses for stationary tasks, but due to how my eyes.. "operate"... they are not an option while moving. (do glasses make one nerdy? lol..)
 
According to my understanding a 20ms 1024 point IR is sufficient for a FF IR. Does this include bass cabinet IRs?

To get right to the point, would a front ported bass cabinet continue to sound at its resonant frequency for more than 20ms once the signal to the speaker is removed? If so, wouldn't the truncation of this sound at 20ms alter the sound of the IR?

A 41Hz tone has a period of about 24ms. If a bass were playing a 41Hz note through a cabinet that had a resonance near that frequency and the player changed to a dead patch to mute the sound, would the cabinet continue to sound for more than one period of the tone? If not, I have my answer. If so, wouldn't a longer IR be needed to accurately capture the response of the cabinet?

What about the same issue with a front ported guitar cabinet? If a cabinet were resonant at about 82Hz and the low E was immediately muted, how many 82Hz cycles could theoretically be reproduced by the cabinet resonance that would be audible? With a period of about 12ms, if it were more than two audible cycles would not a 1024 point IR truncate information? If the tone had a strong sub octave and the cabinet had a resonance point around 40Hz, how many cycles of that would be audible once the signal was cut off?


That's a good question, and I don't know the answer, but I'm taking a guess - I think that IRs are lacking in low frequency information for exactly the reason you describe.
 
When you capture the sound of the speaker in the farfield and then play it back, the sound is "correct" if your ear was right at the speaker, though, right?

Doesn't listening to that capture at ~2+ meters do something further to the sound to distinguish it from the sound you are trying to have?

I don't understand why it isn't possible to simply have the FRFR cabinet mimic the frequency response of a given cabinet and speaker and let the room and air do what they normally do... it obviously isn't or such discussions of farfield, nearfield, mixing, etc... wouldn't occur.

I saw in another discussion a suggestion that someone's mix of IR's was so blended from different sources that it approximated an "ideal" representation of the other speaker. I wonder if there isn't something to that.
 
When you capture the sound of the speaker in the farfield and then play it back, the sound is "correct" if your ear was right at the speaker, though, right?

Doesn't listening to that capture at ~2+ meters do something further to the sound to distinguish it from the sound you are trying to have?

I don't understand why it isn't possible to simply have the FRFR cabinet mimic the frequency response of a given cabinet and speaker and let the room and air do what they normally do... it obviously isn't or such discussions of farfield, nearfield, mixing, etc... wouldn't occur.

I saw in another discussion a suggestion that someone's mix of IR's was so blended from different sources that it approximated an "ideal" representation of the other speaker. I wonder if there isn't something to that.

The words "far field" as applied to speaker impulse responses gets confusing to me because it means different things to different folks.

For me, far field impulse response captures should not include any boundary reflections i.e. the space in which the capture is created does not affect the speaker's impulse response measurement.

In the nearfield, the physics of speaker itself can cause comb filtering effects. It is debatable whether these are desirable or not. As is the debate of using a measurement microphone and reference amplifier and preamplifier vs. a "color" signal chain like a Royer 121 mic and Neve preamp etc.

Given the accuracy of the AxeFx II amp models and efx, I personally would like more FF IR's, in the meaning above to use exclusively vs. the many near field IR's we have now.

Richard
 
When you capture the sound of the speaker in the farfield and then play it back, the sound is "correct" if your ear was right at the speaker, though, right?
No.
Explained below.

Doesn't listening to that capture at ~2+ meters do something further to the sound to distinguish it from the sound you are trying to have?
Once again, no. And, explained below.

First: the "No" to the above was not scolding... I hope it did not come across that way.

Explanation:
A *proper* FF IR that is created to provide the "sonic fingerprint" of a speaker(s) and the enclosure (cab), but to not include any environment variables (room, mic, etc), so as to provide the same experience as one might get with an FRFR *enclosure* directly substituted for said original cab... such an IR requires certain criteria to be met, and such criteria are stringent, and have strict, proven scientific foundation.

To elaborate and simplify a bit: The science and math provide the equations to determine the minimum requirements for the room that can be utilized, the placement of the cab, the placement of the mic, and the results one can expect. To provide for an uncolored capture, one must also utilize the proper equipment, which must be known to be operating properly. Finally, the person(s) performing the captures must be adequately versed in such methods. To compromise in any aspect of all of this will result in an inaccurate capture. However, as we are appreciative of *distortions and coloring*, so it is that compromise can be desirable :D .

Multiple speakers in a cab provide for certain effects, notably comb filtering. Also, the FRFR system, in use, may not provide for an "identical enough experience" for some, which is understandable, if even only considering the psychological aspect. IME, as touched upon in this thread, I believe that, when implemented properly, a properly captured and convolved IR via an FRFR system can be indiscernible from the original cab.

I don't understand why it isn't possible to simply have the FRFR cabinet mimic the frequency response of a given cabinet and speaker and let the room and air do what they normally do... it obviously isn't or such discussions of farfield, nearfield, mixing, etc... wouldn't occur.

I saw in another discussion a suggestion that someone's mix of IR's was so blended from different sources that it approximated an "ideal" representation of the other speaker. I wonder if there isn't something to that.
IMO, the issue is with capture, and reproduction.

For the majority of us, IR use is either 3rd party offerings, or self-created IRs that fall short of what I described as "proper" above to some degree. As such, we are not (as a majority, not an entirety!!!) experiencing what you described, and this is why these discussions occur :D We end up trying to create something that comes close to what we have or want. Ask yourself: "If I could have my favorite cab in the Axe, would I need to blend, etc still?" IOW, is your real-life cab already "good enough"? Or, to cite Sasha from last week, are we simply settling for that cab's tone due to our inability to tweak it, beyond replacing speakers?

Next, consider the inherent nature of the bulk of IRs out there; they provide for a similar experience of sitting in a studio control room, with you playing your guitar through someone else's rig, with a mic on the cab in another room somewhere.

I have gotten close enough with my experiments to produce IR results that are quite satisfying.. to me. I am learning where my incompetences lie, and also where my environment's shortcomings lie (room size is the prime prohibitant currently). I am confident that if I were able to hire AlbertA or Jay M, ship my cabs to them, and have them capture the IRs I desire, the result would allow me to substitute an FRFR system for my cab (same proximity and size, of course), and while blindfolded, I would not be able to pick the cab correctly 100% of the time. It took me weeks to reach that conclusion though! Weeks of deep, obsessive research, countless text exchanges, and several "experimental sessions." Anyone can do the same if they desire, but be forewarned: the investment of time and patience would likely be significant.

You bring up excellent questions. IMO, the best part is that you questioned things :D
 
Back
Top Bottom