Cab Sim Features

[Boudoir Guitar]

It seems every,even very extensive, “we’ll have a UI that makes it seem like you’re swapping and moving mics when there is trickery behind the scenes that is not just interpolating between some very closely captured IRs captured with the actual mic and cab” have lost out both sonically and user preference to a headache file management system of IRs (see Helix cab block). At the end of the day, for most folks in the market for this kind of thing, as much as they may complain about IR rabbit holes and file management they are only willing to leave that burrow for something that sounds very, very good.

I think what is missing in this space is not a new cab sim, but a plugin that can actually leverage the huge library of IRs that so many folks have. OH and other file naming system is sufficiently regimented it shouldn’t be TOO tedious to do the leg work to make an app that can import those packs into a better UI than just “scroll through a bunch of files”.

 

[laxu]

I've seen you mention this request before and I still can't really grasp the benefit of it.

If I understand correctly:

The IR's would be organised in a big list, where the brightest would be at the top, and the darkest at the bottom. This would mean that as you scroll down the list, the IR's will gradually get darker, but also, the rest of the frequency response will change TOTALLY randomly depending on what the IR is captured from. If you want a progressively darker IR, is it not just better to use an EQ? What would the benefit be in organising by brightness?

The only way that would work if it's a sort order personally curated by the IR maker based on perceived sound, so it won't be just frequency analysis because you can perceive something with less bass as brighter than the same sound with more bass.

Movable mics are better for this. If you want "this, but brighter/darker" you simply move a mic a bit.

 

[laxu]

It seems every,even very extensive, “we’ll have a UI that makes it seem like you’re swapping and moving mics when there is trickery behind the scenes that is not just interpolating between some very closely captured IRs captured with the actual mic and cab” have lost out both sonically and user preference to a headache file management system of IRs (see Helix cab block). At the end of the day, for most folks in the market for this kind of thing, as much as they may complain about IR rabbit holes and file management they are only willing to leave that burrow for something that sounds very, very good.

I think a lot of users are looking for shortcuts, whether it's captures, presets or IRs. The people willing to dive deep and learn to work all the gear is far less than people who just want "gimme a good sound and let me play".

I can understand the appeal of both approaches because I'm constantly battling them myself. E.g my Mesa Mark V represents the "tweaker's dream, granular approach, super versatile but complicated", and my BluGuitar amps are the "straightforward, takes little effort to get tones I like" side. Or how a Fractal unit is super capable and great sounding but complicated to work with vs the more curated approach of my Strymons.

I think what is missing in this space is not a new cab sim, but a plugin that can actually leverage the huge library of IRs that so many folks have. OH and other file naming system is sufficiently regimented it shouldn’t be TOO tedious to do the leg work to make an app that can import those packs into a better UI than just “scroll through a bunch of files”.

I think it would have to be something quite abstract. A lot of digital gear programming is designed around real concepts like moving a mic, but if you've never miced an amp wtf does any of that mean to you?

It would likely need to be a user interface that takes away any notion of real cabs and mics, but uses all the data from those IR vendors in the background to make the final sound.

For example Cabinetron offers a whopping ton of shaping and mixing features for IRs, but it goes too deep for most users. It's dimension mixer is probably the closest to an abstract concept that might appeal where mixing IRs becomes something to play with rather than some kind of analytical process.

 

[Dom Mcsweeney]

Usually when it's a curated speaker cabinet plugin with a GUI, one can't typically load their own library of IRs, instead it being proprietary IRs.

I see the frequency analyser idea as a step towards reaching the middle ground, where it would maybe help with bridging that gap.
You could have a list view, or a visual representation of a mic on a cab, catering for people with different preferences.

 

[laxu]

Usually when it's a curated speaker cabinet plugin with a GUI, one can't typically load their own library of IRs, instead it being proprietary IRs.

I see the frequency analyser idea as a step towards reaching the middle ground, where it would maybe help with bridging that gap.
You could have a list view, or a visual representation of a mic on a cab, catering for people with different preferences.

IMO with the movable mic sim plugins, loading your own IRs is not a feature that gives any benefit. You just lose all the features that make the movable mic system great.

It would be better if there was e.g a system for sharing presets for that movable mic cab sim so you can load somebody else's mix and tweak it to your liking if needed. So far I haven't seen anything like this.

 

[Orvillain]

The only way that would work if it's a sort order personally curated by the IR maker based on perceived sound, so it won't be just frequency analysis because you can perceive something with less bass as brighter than the same sound with more bass.

Movable mics are better for this. If you want "this, but brighter/darker" you simply move a mic a bit.

You can categorize across spectral features using MIR techniques.

Spectral Centroid is an audio feature that is more or less a measure of brightness. Sine tone + IR == output data. You process the output data, extract the spectral features, calculate statistics, then you can do some sorting on it. It's actually quite easy to do. Whether it makes sense from a UX perspective is another thing.

 

[Dom Mcsweeney]

IMO with the movable mic sim plugins, loading your own IRs is not a feature that gives any benefit. You just lose all the features that make the movable mic system great.

It would be better if there was e.g a system for sharing presets for that movable mic cab sim so you can load somebody else's mix and tweak it to your liking if needed. So far I haven't seen anything like this.

Eh, just an idea.
I'm fairly sure that preset idea would be trivial, and even exists already.
I mean, you can share a preset for most plugins as far as I can tell, or else they don't have a preset system built in and you use the native DAW preset system instead.

 

[Orvillain]

I think what is missing in this space is not a new cab sim, but a plugin that can actually leverage the huge library of IRs that so many folks have. OH and other file naming system is sufficiently regimented it shouldn’t be TOO tedious to do the leg work to make an app that can import those packs into a better UI than just “scroll through a bunch of files”.

That's actually a cool thought.

 

[Boudoir Guitar]

IMO with the movable mic sim plugins, loading your own IRs is not a feature that gives any benefit. You just lose all the features that make the movable mic system great.

It would be better if there was e.g a system for sharing presets for that movable mic cab sim so you can load somebody else's mix and tweak it to your liking if needed. So far I haven't seen anything like this.

The benefit comes into play more in the "changing mic" than "moving the mic" part. It's not so hard to import a group of files with a particular mic from most libraries and move through those files. its not tooooooo hard to import a group of files using different mics in similar location and comparing -- more than that, and its a royal pain in the ass. its not about getting a "more realistic UI" here, its about getting one that just works.

 

[laxu]

The benefit comes into play more in the "changing mic" than "moving the mic" part. It's not so hard to import a group of files with a particular mic from most libraries and move through those files. its not tooooooo hard to import a group of files using different mics in similar location and comparing -- more than that, and its a royal pain in the ass. its not about getting a "more realistic UI" here, its about getting one that just works.

But if you are going to do that, a regular IR loader will work just fine. I find that way of working much less intuitive because it's basically a mystery box where you spin the wheel until you like the sound, but can't really know what you get until you learn how each IR sounds.

The Hotone Ampero 2 cab block kinda combines both because it has a limited set of mic positions you can pick from a list, then a mic you can choose from another list. It is a more intuitive than IR files because it goes from brighter to darker as the mic moves further away from the center of the cone. It's like a snapshot version of the granularly movable mic sims of e.g Helix, QC or Fractal. Much more coarse, but good enough to give you variation and greatly simplifies the selection process.

 

[Boudoir Guitar]

But if you are going to do that, a regular IR loader will work just fine. I find that way of working much less intuitive because it's basically a mystery box where you spin the wheel until you like the sound, but can't really know what you get until you learn how each IR sounds.

No. There is a loooooooooot of ground in between "doesn't have to be graphic of a mic being moved" and "mystery box". Hell, even using an IR loader I know what the file name conventions are for the packs. I just don't want to have to go 25 files down to snag one of similar mic position but different mic, to then have to scroll back up 24 files to go back to mic I had before but slightly different position.

 

[laxu]

No. There is a loooooooooot of ground in between "doesn't have to be graphic of a mic being moved" and "mystery box". Hell, even using an IR loader I know what the file name conventions are for the packs. I just don't want to have to go 25 files down to snag one of similar mic position but different mic, to then have to scroll back up 24 files to go back to mic I had before but slightly different position.

Ah so your issue is more with the management of those IR files. I agree, the "click up/down arrow to pick next/previous file" UIs are pretty crap.

 

[Orvillain]

I find that way of working much less intuitive because it's basically a mystery box where you spin the wheel until you like the sound

Devil's avocado - how is this any different when the list is turned into a microphone graphic? You still have to move the microphone around until you like the sound. It really isn't drastically different from pressing up/down buttons next to a combo box.

Fundamentally, the very task of deciding which sound to use, involves an amount of spinning the wheel until you like the sound. In most avenues of sound design and music, this is actually regarded as not only desirable, but an inherent part of the creative process.

If it's an IR loader, It would be neat to have a 'frequency analyser' feature, which then would sort through a batch of selected local IRs and give it some order from lets say 'bright to dark'.

I actually don't hate this idea. For most non-technical non-audio engineer musicians, they do think in terms of bright>dark, in a general sense. The don't generally care about the microphone or the speaker on a sonic level. They might from a brand loyalty or a "I've read X on the internet, and so I shall do X" perspective.. but really what they're looking for is a sound that sits somewhere in the bright>dark spectrum.

I think it could be a worthwhile thing to consider.

 

[Orvillain]

No. There is a loooooooooot of ground in between "doesn't have to be graphic of a mic being moved" and "mystery box". Hell, even using an IR loader I know what the file name conventions are for the packs. I just don't want to have to go 25 files down to snag one of similar mic position but different mic, to then have to scroll back up 24 files to go back to mic I had before but slightly different position.

For sure, it isn't as quick as moving a graphic around. But in principle, it is the same.

One thing to consider with these 'move the mic around' type plugins is, you're not really getting the true underlying IR. You're getting a mixture of multiple IR's, and this can sound quite different depending on the underlying DSP.

If the plugin does the morphing in the time domain, then you're going to get phase alignment issues. If you're doing FFT based magnitude+phase morphing, then it is a much heavier process, and you could end up with FFT smearing artifacts if you don't do it right. Of course if you're doing spatial IR interpolation, then you need a shit ton of IR's for it to sound decent, which requires memory.

So effectively, not all of these plugins or platforms are the same.

Personally, I'm convinced that the Quad Cortex cab block is doing simplistic shit under the hood, because it sounds really bad to me. Doesn't have ANY of the punch that a single IR does.

 

[laxu]

Devil's avocado - how is this any different when the list is turned into a microphone graphic? You still have to move the microphone around until you like the sound. It really isn't drastically different from pressing up/down buttons next to a combo box.

The difference to me is that with a list of IRs, I don't know what I get from the next one, and often there's no continuity from the previous one. The next IR could be much darker or more midrangy than the previous one. That's the "mystery box" part.

I know that if I move a mic closer to the cone it will get brighter and so on, even if you still ultimately have to figure out what position gives you the result you like. The drawback is that you have more options so it's harder to make a decision.

I actually don't hate this idea. For most non-technical non-audio engineer musicians, they do think in terms of bright>dark, in a general sense. The don't generally care about the microphone or the speaker on a sonic level. They might from a brand loyalty or a "I've read X on the internet, and so I shall do X" perspective.. but really what they're looking for is a sound that sits somewhere in the bright>dark spectrum.

I think it could be a worthwhile thing to consider.

I can agree with that. If you've never miced a cab, moving a mic around and the different mic models mean absolutely nothing to the user. But most of the time people just want a sound that is in the Goldilocks zone where it's the right brightness and thickness for their preference.

 

[MirrorProfiles]

The difference to me is that with a list of IRs, I don't know what I get from the next one, and often there's no continuity from the previous one. The next IR could be much darker or more midrangy than the previous one. That's the "mystery box" part.

I know that if I move a mic closer to the cone it will get brighter and so on, even if you still ultimately have to figure out what position gives you the result you like. The drawback is that you have more options so it's harder to make a decision.

This is the key point. Moving a mic around isn't a lucky dip until you're happy. There is a clear tonal direction that is predictable and audible as you move between positions. They sound different, but it's in a very particular way that is intuitive to listen to.

 

[Orvillain]

The difference to me is that with a list of IRs, I don't know what I get from the next one, and often there's no continuity from the previous one. The next IR could be much darker or more midrangy than the previous one. That's the "mystery box" part.

Right, but that has nothing to do with the concept of lists or buttons. That's a content problem - IR producers not naming their files properly. Which tbh, Ownhammer and York handle it very well. I've never had a problem where an IR jumped from one to the next, without some kind of logic behind the tonal change.

I know that if I move a mic closer to the cone it will get brighter and so on, even if you still ultimately have to figure out what position gives you the result you like. The drawback is that you have more options so it's harder to make a decision.

Right. You know this. Not everyone does. These things are not self-evident to someone who has never mic'd a cab before. So there is an expectation being placed upon the user; and them understanding it is by no means universal.

I can agree with that. If you've never miced a cab, moving a mic around and the different mic models mean absolutely nothing to the user. But most of the time people just want a sound that is in the Goldilocks zone where it's the right brightness and thickness for their preference.

Yeah, so the real question is - all of this UX stuff aside - how does a user quickly sort through what could be tens of thousands of "potential positions" and get to the exact right one for their scenario.

This is the key point. Moving a mic around isn't a lucky dip until you're happy.

Well I repeat, an up/down list doesn't have to be a lucky dip either. You could feasibly have a list of IR's named in the correct way so that they give you an array of -100 > 0 < +100 IR positions along that full horizontal plane you're looking for. The real issue is, a graphic takes 1 click+drag operation, and constantly updates what your hearing as you move, whereas a list and buttons takes 200 clicks, and would probably suffer from slight audio gaps as it moved from IR to IR. Which is no small difference.

There is a clear tonal direction that is predictable and audible as you move between positions. They sound different, but it's in a very particular way that is intuitive to listen to.

Intuitive once you know that it is analogous to something that happens in the real world.

But Jimmy Smith from Smithsville, USA, who is 14 and just bought his first NDSP plugin... he isn't going to know that. It could quite easily confuse the fuck out of him, seeing a 3D model of a 421 on a speaker that he has no experience with. He doesn't even know what a 421 is.

Those sorts of users... I think they're probably better served in another way. It's for this reason I think Dom's point has something to it.

 

[MirrorProfiles]

Right, but that has nothing to do with the concept of lists or buttons. That's a content problem - IR producers not naming their files properly. Which tbh, Ownhammer and York handle it very well. I've never had a problem where an IR jumped from one to the next, without some kind of logic behind the tonal change.

Right. You know this. Not everyone does. These things are not self-evident to someone who has never mic'd a cab before. So there is an expectation being placed upon the user; and them understanding it is by no means universal.


Yeah, so the real question is - all of this UX stuff aside - how does a user quickly sort through what could be tens of thousands of "potential positions" and get to the exact right one for their scenario.


Well I repeat, an up/down list doesn't have to be a lucky dip either. You could feasibly have a list of IR's named in the correct way so that they give you an array of -100 > 0 < +100 IR positions along that full horizontal plane you're looking for. The real issue is, a graphic takes 1 click+drag operation, and constantly updates what your hearing as you move, whereas a list and buttons takes 200 clicks, and would probably suffer from slight audio gaps as it moved from IR to IR. Which is no small difference.


Intuitive once you know that it is analogous to something that happens in the real world.

But Jimmy Smith from Smithsville, USA, who is 14 and just bought his first NDSP plugin... he isn't going to know that. It could quite easily confuse the fuck out of him, seeing a 3D model of a 421 on a speaker that he has no experience with. He doesn't even know what a 421 is.

Those sorts of users... I think they're probably better served in another way. It's for this reason I think Dom's point has something to it.

If the target demographic is Jimmy from Smithsville, then I’d just put 5 boxes with pictures of women in bikinis and put 5 different well chosen IR’s there. But that’s setting a low bar and IMO is a totally different product than an all encompassing cabinet engine.

Loads of products just have emulated cabinet outs where you don’t know or care what it is. That seems pointless to me in a plugin.

 

[Orvillain]

If the target demographic is Jimmy from Smithsville, then I’d just put 5 boxes with pictures of women in bikinis and put 5 different well chosen IR’s there. But that’s setting a low bar and IMO is a totally different product than an all encompassing cabinet engine.

Loads of products just have emulated cabinet outs where you don’t know or care what it is. That seems pointless to me in a plugin.

Jimmy is a mormon and doesn't like boobies, I don't think this plan would work.

I'm more thinking of ways for Jimmy to know how to best use the all encompassing cab engine plugin, rather than one plugin for spanners, and one plugin for real people.

 

[Boudoir Guitar]

For sure, it isn't as quick as moving a graphic around. But in principle, it is the same.

One thing to consider with these 'move the mic around' type plugins is, you're not really getting the true underlying IR. You're getting a mixture of multiple IR's, and this can sound quite different depending on the underlying DSP.

If the plugin does the morphing in the time domain, then you're going to get phase alignment issues. If you're doing FFT based magnitude+phase morphing, then it is a much heavier process, and you could end up with FFT smearing artifacts if you don't do it right. Of course if you're doing spatial IR interpolation, then you need a shit ton of IR's for it to sound decent, which requires memory.

So effectively, not all of these plugins or platforms are the same.

Personally, I'm convinced that the Quad Cortex cab block is doing simplistic shit under the hood, because it sounds really bad to me. Doesn't have ANY of the punch that a single IR does.

Right. I think this is where you need to decide who your audience is. I am somewhat in agreement with mirrors profiles that the target is not somebody that doesn't know (or care) what a 421 is. Seems to me it's going to be a broad cross-section of folks that consider themselves to be at least slightly "more" than just guitarist and see themselves at knowing at least a little about "production" -- one end of the spectrum being those that are on lower part of the assent up Mt. Stupid, and on the other to very experienced actual producers.

My feeling is the LESS experienced the user the more important actual graphical representation is -- I call these folks (and count myself one of them) RPG producers. The more they feel like they are doing the work of music production the better, even if there are some sonic discrepancies that they PROBABLY don't even realize exist creep in because of it.

On the other end -- you've got Mirror Profile who is annoyed at Fractal's choice to default to phase alignment -- it doesn't matter how "real" the UI is, if the sound is different than what they are expecting its just not going to work well for them.

Perhaps you could have it operate in two modes - one where there is lots of morphing between the IRs in the database and smooth sweep of sound. And a second mode that is truly just swapping between raw IRs without introducing any phase?

It doesn't seem like memory should be an issue for a piece of software like this. Yes, it could be a problem for the morons that are running a 100 track session with a million plugins and software instruments processing simultaneously and trying to "virtually move the mic" on playback of the whole shebang, but that seems like a pretty unreasonable expectation of operation? User error. Also seems like this could be mitigated with some "smart loading into RAM" -- No need to load every iteration of every mic type at every distance. User narrows it down somewhat at the start with "close, mid, far mic position" and chooses a couple of mics -- maybe even a starting point of "bright, balanced, dark" so you're only bringing in files from a subset of files from dust-cap to cone edge.