The infamous "amp in the room" vs FRFR discussion

phil92

Experienced
wouldn't it be possible to model an IR of a speaker (-cab) without the microphone?
And would that not accomplish the amp in the room sound when used with an FRFR?
Or am I forgetting something?

(I don't have a use for that personally, but since it comes up all the time...)
 
The first issue is how?

An IR is created by "listening" to a speaker. How else would it be done?
 
wouldn't it be possible to model an IR of a speaker (-cab) without the microphone?
And would that not accomplish the amp in the room sound when used with an FRFR?
Or am I forgetting something?
IR means impulse response, and the “response” part is what the microphone captures. It’s an actual “snapshot” of what comes out of that speaker at that moment, so something has to capture it, and that’s a mic.

I think you’re thinking about a frequency response chart - how it “should” perform.

Regardless, an IR sound is coming out of another speaker which has its own “output characteristics” in that room at that moment. It’s why you don’t run a cab sim into a real cab - the real cab can’t reproduce the full frequency spectrum of that IR/cab sim so it sounds strange as many frequencies are doubling up or getting phase cancelled.

“Amp in the room” is literally “a speaker loud enough to bounce off the room and back to your ears.” Volume is what creates the 3D effect because it IS 3D as it bounces all around you. You can’t simulate “sound coming from behind me” through a speaker in front of you at a low volume. It has to physically and literally come from behind you to feel like it’s coming from behind you.
 
No one wants “amp on room”, they want guitar speaker cab in room, and that essentially is a high sound pressure level coming from a directional source, or more commonly sources if it’s a multiple speaker cabinet like a 4x12, or, front and rear from open back cabinets.

Stack 4 atomic CLR making a 4x12 FRFR and it would sound pretty darn “in the room”
 
IR means impulse response, and the “response” part is what the microphone captures. It’s an actual “snapshot” of what comes out of that speaker at that moment, so something has to capture it, and that’s a mic.

I think you’re thinking about a frequency response chart - how it “should” perform.

Regardless, an IR sound is coming out of another speaker which has its own “output characteristics” in that room at that moment. It’s why you don’t run a cab sim into a real cab - the real cab can’t reproduce the full frequency spectrum of that IR/cab sim so it sounds strange as many frequencies are doubling up or getting phase cancelled.

“Amp in the room” is literally “a speaker loud enough to bounce off the room and back to your ears.” Volume is what creates the 3D effect because it IS 3D as it bounces all around you. You can’t simulate “sound coming from behind me” through a speaker in front of you at a low volume. It has to physically and literally come from behind you to feel like it’s coming from behind you.

well, I was thinking, readily available IRs seem to "not do it" for a few people. And it's the microphone part that they don't like. They seem to be after the sound of the cab as it sounds to their ears when they are experiencing it.

So if you could model or capture the cab but not the microphone, take it out of the equation, that could solve it. Maybe it needs to be done using ultra flat response microphones, maybe pressure zone microphones, maybe a combination of microphones, maybe manual tweaking by ear... I am aware that a speaker is not a point source and with a microphone you can basically only capture the sound at one point in space, but I think an approximation should be possible.
 
well, I was thinking, readily available IRs seem to "not do it" for a few people. And it's the microphone part that they don't like. They seem to be after the sound of the cab as it sounds to their ears when they are experiencing it.

So if you could model or capture the cab but not the microphone, take it out of the equation, that could solve it. Maybe it needs to be done using ultra flat response microphones, maybe pressure zone microphones, maybe a combination of microphones, maybe manual tweaking by ear... I am aware that a speaker is not a point source and with a microphone you can basically only capture the sound at one point in space, but I think an approximation should be possible.
Far-field IRs, taken using the ground plane measurement method with calibrated microphones. That's what you're asking for. Only a few FF-IRs exist online however, so you're going to have to make your own if you want results you like.
 
well, I was thinking, readily available IRs seem to "not do it" for a few people. And it's the microphone part that they don't like. They seem to be after the sound of the cab as it sounds to their ears when they are experiencing it.

So if you could model or capture the cab but not the microphone, take it out of the equation, that could solve it. Maybe it needs to be done using ultra flat response microphones, maybe pressure zone microphones, maybe a combination of microphones, maybe manual tweaking by ear... I am aware that a speaker is not a point source and with a microphone you can basically only capture the sound at one point in space, but I think an approximation should be possible.
That’s exactly how IRs used to be captured and why the cab block used to have Mics. Flat mic to capture the cab, but it didn’t yield results as “real” as using the actual mics you want in the positions you want.

Again, you can’t capture “sound bouncing off the wall behind you” with the short millisecond IR captures. That’s getting close to Reverb territory. The sound actually has to bounce off the room you are in at that moment. It can’t “bounce behind you” from a speaker in front of you no matter how you capture the IR.
 
Again, you can’t capture “sound bouncing off the wall behind you” ... The sound actually has to bounce off the room you are in at that moment

no argument there. I am not talking about modeling the room, just the speaker without the microphone.

If that were possible, and you used that IR, you'd just get the reflections in the room from your FRFR. As you always do. It's not a sound pressure thing either: the reflections are there, whether you can hear them or not.
 
I’m sure it’s coming. The amps and effects are better than the real thing at this point, but everyone admits there can be huge improvements in the IR realm. Which is why I run both a real cab and IR’s concurrently. Do a lot of the IR’s sound GOOD? Yes, they do. Do they sound REAL? No, they don’t. They do approximate a mic’d cab fairly well, but we don’t really want to jam with a mic’d cab sound. The one dimensional nature of that is a big reason why we work around it when recording by stacking and blending multiple guitar tracks through different guitar, amp, cab, mic, and eq options to make the recorded guitar sound interesting and energetic like it does when your in a space with an amp. I know when I pull up a real world rig next to an “identical “ virtual rig in the axe side by side inthe teal world and play, they do NOT sound the same at all, and it’s the IR that is the bottleneck in that situation. With all the horsepower I hear is available in the III, and the damn near perfection squared we have with the amps and effects, getting the IR side of things to parity is the logical next step. I would be surprised and even a bit disappointed if Cliff wasn’t working on that in secrecy as we speak. I don’t know HOW to go about achieving that, but the first guy to crack that egg successfully will own the market for a long time coming.
 
Far-field IRs, taken using the ground plane measurement method with calibrated microphones. That's what you're asking for. Only a few FF-IRs exist online however, so you're going to have to make your own if you want results you like.

I was not asking for anything, just wondering.

but hey, never mind. I'm happy with my monitors and headphones. Just kept thinking about this issue, not looking for an argument.
 
I think IR smoothing can reduce recording artifacts (e.g. comb filtering) this to a degree by reducing extreme peaks and notches due to phase cancellations of close micing. Still not the same as what two ears hear in a room several meters away with natural ambiance.

What Cliff says: "...there is a lot more high frequency content in a typical IR. When you listen to your tube amp through its speaker you are in the far field and usually somewhat off-axis. There is less high frequency content reaching your ears compared to what a microphone records. Go put your amp in an isolation room and listen to it through the control room monitors. You'll be astonished at what you hear."
.
Interestingly we also subjectively subtract room effects as we consciously/subconscious listen to a sound source. I recorded my high-end stereo speakers playing various music with binaural microphones (placed in the ear). The resulting playback sounded realistic except that the room was much more prominent in the recording than when I listen directly.

Would we ever want to model the sound of hi-fi stereo speakers "in the room" and then play that into the room with quality FRFR speakers? Makes no sense.
 
Last edited:
I was not asking for anything, just wondering.

but hey, never mind. I'm happy with my monitors and headphones. Just kept thinking about this issue, not looking for an argument.
Eh? I'm not arguing either.

What you're wondering out loud exists, it's called far-field IRs. I'm just letting you know, and letting you know how it's done if you want to try.
 
I’m sure it’s coming. The amps and effects are better than the real thing at this point, but everyone admits there can be huge improvements in the IR realm. Which is why I run both a real cab and IR’s concurrently. Do a lot of the IR’s sound GOOD? Yes, they do. Do they sound REAL? No, they don’t. They do approximate a mic’d cab fairly well, but we don’t really want to jam with a mic’d cab sound. The one dimensional nature of that is a big reason why we work around it when recording by stacking and blending multiple guitar tracks through different guitar, amp, cab, mic, and eq options to make the recorded guitar sound interesting and energetic like it does when your in a space with an amp. I know when I pull up a real world rig next to an “identical “ virtual rig in the axe side by side inthe teal world and play, they do NOT sound the same at all, and it’s the IR that is the bottleneck in that situation. With all the horsepower I hear is available in the III, and the damn near perfection squared we have with the amps and effects, getting the IR side of things to parity is the logical next step. I would be surprised and even a bit disappointed if Cliff wasn’t working on that in secrecy as we speak. I don’t know HOW to go about achieving that, but the first guy to crack that egg successfully will own the market for a long time coming.

I think I disagree with almost everything in this post. IR and cab block has been improved so much in the last 2 years that the mythical 'amp in the room' is a non-issue...if you know what you're doing (but that's the rub, isn't it?)
 
Cliff meantioned somewhere they are on to something about shooting some exceptionel far-field IRs. Hope they find a way, since it's not simple and the results are not known.

At the other side, riddle this: When one speaker has a small dispesion and another has a wide dispersion, could an equalizer make them sound all the same?
 
I’m sure it’s coming. The amps and effects are better than the real thing at this point, but everyone admits there can be huge improvements in the IR realm. Which is why I run both a real cab and IR’s concurrently. Do a lot of the IR’s sound GOOD? Yes, they do. Do they sound REAL? No, they don’t. They do approximate a mic’d cab fairly well, but we don’t really want to jam with a mic’d cab sound. The one dimensional nature of that is a big reason why we work around it when recording by stacking and blending multiple guitar tracks through different guitar, amp, cab, mic, and eq options to make the recorded guitar sound interesting and energetic like it does when your in a space with an amp. I know when I pull up a real world rig next to an “identical “ virtual rig in the axe side by side inthe teal world and play, they do NOT sound the same at all, and it’s the IR that is the bottleneck in that situation. With all the horsepower I hear is available in the III, and the damn near perfection squared we have with the amps and effects, getting the IR side of things to parity is the logical next step. I would be surprised and even a bit disappointed if Cliff wasn’t working on that in secrecy as we speak. I don’t know HOW to go about achieving that, but the first guy to crack that egg successfully will own the market for a long time coming.
Wow.

When done right IR's get pretty damn close to 100% realism when compared to mics. Maybe you just don't like how mics sound? But that would mean that you don't like any of the guitar sounds ever recorded.

You do realize that your ears work much like mics and the sound changes based on your ear placement. Do we need a virtual walk-around-in-the-room parameter in the cab block? :D
 
If you're a bedroom player and have an amp in the room issue, just get a combo practice amp in there :D

Headphone players don't have the issue
Gigging musicians don't have the issue
In-ear users don't have the issue
Producers and studio musicians don't have the issue

I practice / rehearse / gig / and record, and am yet to find a situation where I asked myself for this amp in the room stuff.

The axefx is a killing machine in term of mic'ed amp tone, ready for a lot of monitoring solutions.
You can get a power amp and a cab, or a subwoofer, or a big PA system, but if you play for yourself and wanna hear an unmic'ed amp, maybe you're just searching for an amp :D
 
Back
Top Bottom