AFIII To All The Pundits

I wish there were more apples to apples comparisons like this. Actual amp sounds a bit stiffer dynamics-wise, so the Axe probably had a nicer feel as well.

Amps have many strengths but several weaknesses. With the Axe matching amp strengths so well now, users can dial their goldilocks amp while also dialing out some of what they may not like about the original amp.

Guitar may not be as popular as it was in its heyday, but there's never been a better time than now to get the tone and feel you want.
 
the distortion of the guitar is probably sample accurate, the only thing that lets you tell it apart from a model is the time image you get when you record the real mic. (time image is what i'm going to keep calling it until i get my PhD is musicology and can define what the hell I'm actually talking about.)

when you record a mic on an amp, even if it's mono, it picks up first the sound closest to it, and continues to pickup sound reflection from all angles, that come back to it at different times. That way even a mono signal can provide an audio image about the space it's in, based on what else the mic pics up as the recording continues. the audio will hang in the air and swirl and mix as the source continues to output the signal. like you could tell how big a room was from a drum hit even if you only had one ear, it's not stereo but there's time information available that tells you front-to-back, floor-to-ceiling, and how far away you are from the source in relation to the space you're in.

when an IR is used on a model, the audio source (model) is fed through a filter (IR) to approximate the sound of the mic on the cab, and there is no time space between the audio source (model) and the IR, it's just 1:1. In a recording, that time image lasts the whole duration of the recording, for as long as it's recording, so the time image is never broken up or cut short.

The "amp in the room" sound is audio diffusion, audio softlight, that allows you to get a perspective of space...where the source is coming from and where you are in relation to it. You could approximate the diffusion after the fact, placed after the IR, but that's not how the signal chain works in a recording setup. The audible element of TIME, where frequencies hang out and swirl and mix, even for just a few milliseconds, and hit the microphone from different angles and different amounts, once twice three times and so on, that element of TIME belongs between the source (model) and the mic (IR). When an IR is only a few MS long there isn't that time available for the sound energy to get spun up and crash into each other as much, so you only get a little bit of the spacial information.

you know i just had an idea that i wish i had an AxeFx2 to try on. if you could get the room parameter in front of the IR, that would theoretically work to approximate the time image in front of the filter. maybe the room parameter is already before the IR in the cab block. that would explain why it works so well compared to a reverb block after the cab block. with two cab blocks you could turn on room in the first one and leave it empty to effect the IR in the second one. wonder if that would do anything different than room adjustment in the IR block.
 
camilovelandiamusic said:
There's a "bottle like" thing about a real deluxe reverb.
Click to expand...

that's a time image, the mic is picking up the slightest reflections from the grill cloth, the amp, the floor, and whatever space it's in. the time image in a recording lasts the whole duration and never gets broken up or cut short, where an IR of a few hundred milliseconds can only provide an image of that length. unless...it gets fed a signal that has been defused already. but then i think you might have to get into separating the speaker simulation from the mic simulation, since an IR is effectively both of those things. brain hurts.
 
shatteredsquare said:
the distortion of the guitar is probably sample accurate, the only thing that lets you tell it apart from a model is the time image you get when you record the real mic. (time image is what i'm going to keep calling it until i get my PhD is musicology and can define what the hell I'm actually talking about.)

when you record a mic on an amp, even if it's mono, it picks up first the sound closest to it, and continues to pickup sound reflection from all angles, that come back to it at different times. That way even a mono signal can provide an audio image about the space it's in, based on what else the mic pics up as the recording continues. the audio will hang in the air and swirl and mix as the source continues to output the signal. like you could tell how big a room was from a drum hit even if you only had one ear, it's not stereo but there's time information available that tells you front-to-back, floor-to-ceiling, and how far away you are from the source in relation to the space you're in.

when an IR is used on a model, the audio source (model) is fed through a filter (IR) to approximate the sound of the mic on the cab, and there is no time space between the audio source (model) and the IR, it's just 1:1. In a recording, that time image lasts the whole duration of the recording, for as long as it's recording, so the time image is never broken up or cut short.

The "amp in the room" sound is audio diffusion, audio softlight, that allows you to get a perspective of space...where the source is coming from and where you are in relation to it. You could approximate the diffusion after the fact, placed after the IR, but that's not how the signal chain works in a recording setup. The audible element of TIME, where frequencies hang out and swirl and mix, even for just a few milliseconds, and hit the microphone from different angles and different amounts, once twice three times and so on, that element of TIME belongs between the source (model) and the mic (IR). When an IR is only a few MS long there isn't that time available for the sound energy to get spun up and crash into each other as much, so you only get a little bit of the spacial information.

you know i just had an idea that i wish i had an AxeFx2 to try on. if you could get the room parameter in front of the IR, that would theoretically work to approximate the time image in front of the filter. maybe the room parameter is already before the IR in the cab block. that would explain why it works so well compared to a reverb block after the cab block. with two cab blocks you could turn on room in the first one and leave it empty to effect the IR in the second one. wonder if that would do anything different than room adjustment in the IR block.
Click to expand...
I'm no expert, but pretty sure that this has been discussed here in the past... And that Impulse Responses do, in fact, account for this.

Ultra Res IRs are 180ms, I believe.
 
shatteredsquare said:
the distortion of the guitar is probably sample accurate, the only thing that lets you tell it apart from a model is the time image you get when you record the real mic. (time image is what i'm going to keep calling it until i get my PhD is musicology and can define what the hell I'm actually talking about.)

when you record a mic on an amp, even if it's mono, it picks up first the sound closest to it, and continues to pickup sound reflection from all angles, that come back to it at different times. That way even a mono signal can provide an audio image about the space it's in, based on what else the mic pics up as the recording continues. the audio will hang in the air and swirl and mix as the source continues to output the signal. like you could tell how big a room was from a drum hit even if you only had one ear, it's not stereo but there's time information available that tells you front-to-back, floor-to-ceiling, and how far away you are from the source in relation to the space you're in.

when an IR is used on a model, the audio source (model) is fed through a filter (IR) to approximate the sound of the mic on the cab, and there is no time space between the audio source (model) and the IR, it's just 1:1. In a recording, that time image lasts the whole duration of the recording, for as long as it's recording, so the time image is never broken up or cut short.

The "amp in the room" sound is audio diffusion, audio softlight, that allows you to get a perspective of space...where the source is coming from and where you are in relation to it. You could approximate the diffusion after the fact, placed after the IR, but that's not how the signal chain works in a recording setup. The audible element of TIME, where frequencies hang out and swirl and mix, even for just a few milliseconds, and hit the microphone from different angles and different amounts, once twice three times and so on, that element of TIME belongs between the source (model) and the mic (IR). When an IR is only a few MS long there isn't that time available for the sound energy to get spun up and crash into each other as much, so you only get a little bit of the spacial information.

you know i just had an idea that i wish i had an AxeFx2 to try on. if you could get the room parameter in front of the IR, that would theoretically work to approximate the time image in front of the filter. maybe the room parameter is already before the IR in the cab block. that would explain why it works so well compared to a reverb block after the cab block. with two cab blocks you could turn on room in the first one and leave it empty to effect the IR in the second one. wonder if that would do anything different than room adjustment in the IR block.
Click to expand...
So much about this is wrong it's hard to know where to start.
 
That "time image" is reflected sound. It's called reverb or delay, depending on how distinct the reflections are. IRs capture it in most live rooms, but it's masked by the direct sound. The room you're listening in provides its own reflections.
 
FractalAudio said:
So much about this is wrong it's hard to know where to start.
Click to expand...

start at the beginning then

chris said:
i don't understand.
Click to expand...

in real recorded life sound sources that are only a few milliseconds long last longer that a few milliseconds, because the energy continues to move around after the source stops it's output. a speaker pushing sound into air is constantly mixing the frequencies with the next ones, that mix in the air with the ones that were played before. a mic picks up all of that interaction, because the time image is before the microphone. An IR doesn't pick up that time image, because there is no time element between the audio source and the filter. You can tell the difference between holding your ear next to a laptop speaker at 1 inch, 2 inch, 3 inch. Your ear can pickup those differences. At 1 inch your ear is picking up a lot of reflection that is ping ponging back and forth from your head to the laptop. At 2 inches it picks up less of that, and your brain can tell. When you remove that time domain, and put your ear drum right on the audio source, there is no more time domain, and it's unnatural, your brain can tell, there's no space to it. In an IR the only time domain comes after the filter, it's on the wrong end. The diffusion in real life happens before it gets to the mic, and the mics sums it flat.

summary point: the IR is an approximation and a static filter, so it doesn't translate the time image after 180ms. There is information after 180ms, even if it was miced in an isoCab, due to the fact that in a real recording the sound will be continuous, and there is a lot of interaction that happens continuously that gets summed by the mic. An IR only translates most of the fundamental and misses a bunch of the harmonics, harmonics that provide spatial reference. However slight, it makes an audible difference, enough to effect spatial perception. I'm just trying to enunciate specifically what is missing from IR realism currently IMO, compared to a real recording. It's the only thing that tells me that A is the real amp and B is the model, it lacks the time image from the recording. If a time image were imposed on B even on the wrong end of the chain, i wouldn't be able to tell a difference.
 
shatteredsquare said:
There is information after 180ms, even if it was miced in an isoCab, due to the fact that in a real recording the sound will be continuous, and there is a lot of interaction that happens continuously that gets summed by the mic.
Click to expand...

I think you are misunderstanding how sound waves behave. Sound waves superimpose in a linear fashion. This means that sound waves emitted at time X don't 'interact' with reflections from a previous time Y. The sound you hear at the microphone is the linear sum of sound from time X and reflections from the earlier time Y. This can all be captured and replicated very well by an IR (provided the IR is long enough, e.g. 180ms).
 
shatteredsquare said:
the distortion of the guitar is probably sample accurate, the only thing that lets you tell it apart from a model is the time image you get when you record the real mic. (time image is what i'm going to keep calling it until i get my PhD is musicology and can define what the hell I'm actually talking about.)

when you record a mic on an amp, even if it's mono, it picks up first the sound closest to it, and continues to pickup sound reflection from all angles, that come back to it at different times. That way even a mono signal can provide an audio image about the space it's in, based on what else the mic pics up as the recording continues. the audio will hang in the air and swirl and mix as the source continues to output the signal. like you could tell how big a room was from a drum hit even if you only had one ear, it's not stereo but there's time information available that tells you front-to-back, floor-to-ceiling, and how far away you are from the source in relation to the space you're in.

when an IR is used on a model, the audio source (model) is fed through a filter (IR) to approximate the sound of the mic on the cab, and there is no time space between the audio source (model) and the IR, it's just 1:1. In a recording, that time image lasts the whole duration of the recording, for as long as it's recording, so the time image is never broken up or cut short.

The "amp in the room" sound is audio diffusion, audio softlight, that allows you to get a perspective of space...where the source is coming from and where you are in relation to it. You could approximate the diffusion after the fact, placed after the IR, but that's not how the signal chain works in a recording setup. The audible element of TIME, where frequencies hang out and swirl and mix, even for just a few milliseconds, and hit the microphone from different angles and different amounts, once twice three times and so on, that element of TIME belongs between the source (model) and the mic (IR). When an IR is only a few MS long there isn't that time available for the sound energy to get spun up and crash into each other as much, so you only get a little bit of the spacial information.

you know i just had an idea that i wish i had an AxeFx2 to try on. if you could get the room parameter in front of the IR, that would theoretically work to approximate the time image in front of the filter. maybe the room parameter is already before the IR in the cab block. that would explain why it works so well compared to a reverb block after the cab block. with two cab blocks you could turn on room in the first one and leave it empty to effect the IR in the second one. wonder if that would do anything different than room adjustment in the IR block.
Click to expand...
1) The time image you're talking about is just room reverb, an IR can also contain those room reflections but it has to be long enough to capture the entire decay.
The IRs in the Axe fx don't contain all the room reflections cuz they're too short and minimum phase transformed, but you can always use an IR loader on your PC to load long and unprocessed IRs.

2)reverb, IRs, filters and EQs are all Linear Time Invariant, this means that the order you place them in the chain doesn't matter, the result will be identical if you place a reverb before or after a cab.
The room parameter in the cab block is just a room reverb with only early reflections, you can easily replicate it in he reverb block.
 
DLC86 said:
2)reverb, IRs, filters and EQs are all Linear Time Invariant, this means that the order you place them in the chain doesn't matter, the result will be identical if you place a reverb before or after a cab.
The room parameter in the cab block is just a room reverb with only early reflections, you can easily replicate it in he reverb block.
Click to expand...

As far as my understanding, this is true only as long as as no distortion is added by the effect (such as the preamp in the cab block, etc).
 
Yes, forgot to mention those exceptions. Even modulation is not LTI but in normal circumstances the difference would probably be impossible to hear.
 
To get it sorted.

1st: Listening to an FRFR system in room to replace a guitar cab:
When you include room reflections into the IR you get the room twice when listening to it: 1st the room from your signal coming from the IR plus 2nd the room you are standing in listening to your FRFR system. So the room should not to be a part of the IR when displaying via an audio system in the room.

2nd: Making recordings/ using headphones.
When recordings get produced with a cab and a mic, the room where the cab stands gets part of the signal. The techs usually try to keep the rooms influence on the sound as low as possible, so their target is clear. It then would be best to keep the room all out of the IR. In reality many recordings contain some room, more or less, they just couldn't do it better...

Should we include room in the IR to imitate the unperfect reality or should we keep it out, because that's what the techs were after?

I'd say if you want room in your IR keep it very very low. Done.
Not worth a groundbreaking discussion whether IRs work at all, or is that little background noise that much important?
 
666was999 said:
To get it sorted.

1st: Listening to an FRFR system in room to replace a guitar cab:
When you include room reflections into the IR you get the room twice when listening to it: 1st the room from your signal coming from the IR plus 2nd the room you are standing in listening to your FRFR system. So the room should not to be a part of the IR when displaying via an audio system in the room.
Click to expand...
Unless you only listen in headphones you're always hearing the room you are listening in.
2nd: Making recordings/ using headphones.
When recordings get produced with a cab and a mic, the room where the cab stands gets part of the signal. The techs usually try to keep the rooms influence on the sound as low as possible, so their target is clear. It then would be best to keep the room all out of the IR. In reality many recordings contain some room, more or less, they just couldn't do it better...
Click to expand...
I don't think they do try to keep the room out.. . That is why the room can be very important.
 
You know in the studios they put carpet on the walls?
I guess that's to eliminate some room reflections. That's their target.
Am I wrong?
 
You must log in or register to reply here.
Back
Top Bottom