Impulse Response Users...advice/ opinion.

Discuss VST stuff here!
Post Reply
User avatar
MrFigg
Competition Winner
Posts: 9124
Joined: 20 Apr 2018

04 Jul 2022

Your humble guitarist just bought the Melda Productions MTurboAmp plugin for $15 (down from 100 and something). It's really pretty good. They recommend their MCabinet plugin to use in conjunction but that'll need to wait for Black Friday I'm thinking. Anyway, Audio Assault have a one day sale today (4th July) and I just got ReAmp Studio for $7 and it's also pretty good :). They have an IR shaper plugin which is currently going for $6. The GAS has got me good. I have never loaded an IR in my life but...you know, just in case, it's cheap and all the other classic excuses for buying stuff.
So...only $6 but all those $6s add up to a real hardware pedal. It's def a maybe maybe maybe use it plugin so...anyone who uses IRs...is this a cool plugin?

https://audioassault.mx/collections/plu ... pulse-rack
🗲 2ॐ ᛉ

User avatar
Loque
Moderator
Posts: 11175
Joined: 28 Dec 2015

04 Jul 2022

Why not use RV7MKII? I got some nice Amps from it.

Never tried to create my own Amp Sim Cabinet IR. Probably some detailed know how about EQing such things would be helpful. I had more done in the direction of creative, rhythmic, esoteric and special frequency manipulations...
Reason12, Win10

User avatar
MrFigg
Competition Winner
Posts: 9124
Joined: 20 Apr 2018

04 Jul 2022

Loque wrote:
04 Jul 2022
Why not use RV7MKII? I got some nice Amps from it.

Never tried to create my own Amp Sim Cabinet IR. Probably some detailed know how about EQing such things would be helpful. I had more done in the direction of creative, rhythmic, esoteric and special frequency manipulations...
Actually just banged the Kuassa amps after it and turned off the head...good few cabs in there. Probably don't need to design my own IRs. Was GAS. I see that clearly now :).
🗲 2ॐ ᛉ

RobC
Posts: 1833
Joined: 10 Mar 2018

04 Jul 2022

Loque wrote:
04 Jul 2022
Why not use RV7MKII? I got some nice Amps from it.

Never tried to create my own Amp Sim Cabinet IR. Probably some detailed know how about EQing such things would be helpful. I had more done in the direction of creative, rhythmic, esoteric and special frequency manipulations...
Oooh, good that you mention this! Now that I have two microphones, I really gotta take a car's IR.

Although it would be better, if the test sound would be emitted from a speaker, for better accuracy. Not sure if the RV7000MKII can use stereo recordings. Though it has mono mode, so I can just simply use two instances for each channel.

Honestly, while I get that convolution reverb just focuses on reverberation, I think the position/direction of the sound source matters a lot, too. I'm not all that familiar with convolution, but using one cool sounding reverb, even if panned and all, won't be as realistic as it could be. I would record multiple samples, coming from different distances and directions, probably spots where musicians would be in case of a band.

Of course, it could be that it's already being done like this.

User avatar
selig
RE Developer
Posts: 11685
Joined: 15 Jan 2015
Location: The NorthWoods, CT, USA

04 Jul 2022

RobC wrote:
04 Jul 2022
Loque wrote:
04 Jul 2022
Why not use RV7MKII? I got some nice Amps from it.

Never tried to create my own Amp Sim Cabinet IR. Probably some detailed know how about EQing such things would be helpful. I had more done in the direction of creative, rhythmic, esoteric and special frequency manipulations...
Oooh, good that you mention this! Now that I have two microphones, I really gotta take a car's IR.

Although it would be better, if the test sound would be emitted from a speaker, for better accuracy. Not sure if the RV7000MKII can use stereo recordings. Though it has mono mode, so I can just simply use two instances for each channel.

Honestly, while I get that convolution reverb just focuses on reverberation, I think the position/direction of the sound source matters a lot, too. I'm not all that familiar with convolution, but using one cool sounding reverb, even if panned and all, won't be as realistic as it could be. I would record multiple samples, coming from different distances and directions, probably spots where musicians would be in case of a band.

Of course, it could be that it's already being done like this.
If you are ‘shooting’ a room, the quality depends on (among other things) the speaker used - you don’t want it to color the sound or leave out any frequency ranges. But if you are shooting a cabinet, then it’s the speaker rather than the room you are ‘shooting’.
RV7k uses stereo samples for it’s IR. Start with ‘stereo’ white noise if you want the smoothest reverb you’ve never heard IRL.
Using one cool sounding reverb CAN be VERY realistic, as always it depends…
It depends just like with sampling because they are very similar. If your goal is to sample one perfect note on a piano and play it and have it sound perfect, it’s really pretty easy. Same for reverb, if you know what you want just set up the speaker as the source, and setup the microphones where they sound best. But just like with sampling, the ‘trick’ comes when you try to create something more versatile and flexible, because THEN you need more than one ‘sample’ to capture the nuances and natural variations.
With a speaker cabinet they can mostly work, but real speakers DO behave slightly differently at different volumes, so you need to at least account for that (even if you only choose one of the samples for your convo).

But just think of it like sampling, and remember gigo (garbage in, garbage out). If the room sucks, or the speaker isn’t full range, or the microphones are not in ideal positions or of the desired quality, there is less chance you’ll get something in line with your expectations (happy accidents are always possible though!). Make sense?
Selig Audio, LLC

User avatar
MrFigg
Competition Winner
Posts: 9124
Joined: 20 Apr 2018

04 Jul 2022

Just for the record...if you buy plugins/ IRs on Audio Assault for over $30 then the code AAFIFTY will get you 50% off. Soooo...loads of stuff for like $17 or whatever. Wish I'd tried that before I bought the first one but there you go.
Edit: Actually, no minimum buy it would seem.
Last edited by MrFigg on 04 Jul 2022, edited 1 time in total.
🗲 2ॐ ᛉ

RobC
Posts: 1833
Joined: 10 Mar 2018

04 Jul 2022

selig wrote:
04 Jul 2022
If you are ‘shooting’ a room, the quality depends on (among other things) the speaker used - you don’t want it to color the sound or leave out any frequency ranges. But if you are shooting a cabinet, then it’s the speaker rather than the room you are ‘shooting’.
RV7k uses stereo samples for it’s IR. Start with ‘stereo’ white noise if you want the smoothest reverb you’ve never heard IRL.
Using one cool sounding reverb CAN be VERY realistic, as always it depends…
It depends just like with sampling because they are very similar. If your goal is to sample one perfect note on a piano and play it and have it sound perfect, it’s really pretty easy. Same for reverb, if you know what you want just set up the speaker as the source, and setup the microphones where they sound best. But just like with sampling, the ‘trick’ comes when you try to create something more versatile and flexible, because THEN you need more than one ‘sample’ to capture the nuances and natural variations.
With a speaker cabinet they can mostly work, but real speakers DO behave slightly differently at different volumes, so you need to at least account for that (even if you only choose one of the samples for your convo).

But just think of it like sampling, and remember gigo (garbage in, garbage out). If the room sucks, or the speaker isn’t full range, or the microphones are not in ideal positions or of the desired quality, there is less chance you’ll get something in line with your expectations (happy accidents are always possible though!). Make sense?
I would think, that using any speaker is still better than just moving around and clapping for example. Meaning that at least the played sample would be the same.
Can't, or shouldn't a speaker play back at the same volume though, regardless the distance (in other words, no loudness compensation - what's closer will be louder, what's further away, will be quieter)? Or isn't capturing the volume differences possible / worth it?

That said, a clean noise reverb is kind of what I was looking for. I used to create it with the vocoder. Although mostly to give sounds a bit of body/tail.
Ironically, I'm normally not much of a reverb kind of guy. I prefer things as dry as possible. Long, "cold" reverbs usually kind of remind me of hangovers and headaches.

User avatar
vectro
Posts: 55
Joined: 17 Jan 2015
Location: Sweden
Contact:

05 Jul 2022

RobC wrote:
04 Jul 2022
Not sure if the RV7000MKII can use stereo recordings.
Absolutely! You can do stereo, M/S and many modes. You would only need two instances if doing Quad/"True Stereo" where you have 4ch convolution outputs from 2ch inputs.

User avatar
selig
RE Developer
Posts: 11685
Joined: 15 Jan 2015
Location: The NorthWoods, CT, USA

05 Jul 2022

RobC wrote:
04 Jul 2022
selig wrote:
04 Jul 2022
If you are ‘shooting’ a room, the quality depends on (among other things) the speaker used - you don’t want it to color the sound or leave out any frequency ranges. But if you are shooting a cabinet, then it’s the speaker rather than the room you are ‘shooting’.
RV7k uses stereo samples for it’s IR. Start with ‘stereo’ white noise if you want the smoothest reverb you’ve never heard IRL.
Using one cool sounding reverb CAN be VERY realistic, as always it depends…
It depends just like with sampling because they are very similar. If your goal is to sample one perfect note on a piano and play it and have it sound perfect, it’s really pretty easy. Same for reverb, if you know what you want just set up the speaker as the source, and setup the microphones where they sound best. But just like with sampling, the ‘trick’ comes when you try to create something more versatile and flexible, because THEN you need more than one ‘sample’ to capture the nuances and natural variations.
With a speaker cabinet they can mostly work, but real speakers DO behave slightly differently at different volumes, so you need to at least account for that (even if you only choose one of the samples for your convo).

But just think of it like sampling, and remember gigo (garbage in, garbage out). If the room sucks, or the speaker isn’t full range, or the microphones are not in ideal positions or of the desired quality, there is less chance you’ll get something in line with your expectations (happy accidents are always possible though!). Make sense?
I would think, that using any speaker is still better than just moving around and clapping for example. Meaning that at least the played sample would be the same.
Can't, or shouldn't a speaker play back at the same volume though, regardless the distance (in other words, no loudness compensation - what's closer will be louder, what's further away, will be quieter)? Or isn't capturing the volume differences possible / worth it?

That said, a clean noise reverb is kind of what I was looking for. I used to create it with the vocoder. Although mostly to give sounds a bit of body/tail.
Ironically, I'm normally not much of a reverb kind of guy. I prefer things as dry as possible. Long, "cold" reverbs usually kind of remind me of hangovers and headaches.
The best current technique is the swept sine technique, which requires a full range speaker (ideally omni-directional!) to get a response that accurately represents the audible range of frequencies, and it should be as flat as possible as well. Imagine a snare drum in a room, radiating sound out in a sphere away from the drum. Putting a two way speaker facing up will only get you so far in re-creating the reflections that the snare can create. So it is important to start with a source that radiates like actual instruments in as much as this is possible.
From there, you are recording the room so all decisions made will affect the results - what types of microphones used, how they are positioned (what configuration), WHERE they are positioned, etc. Just like with any recording, you can send two teams into a space to record an impulse and get two very different results.
This parallels sampling IMO, because at some point you can't just keep taking more samples. Computational/algorithmic approaches will be better at giving complete control over a virtual 'space' just as modeling gives more control over instrument design than sampling.
There are tradeoffs with both approaches, but all to say convolution has similar limits as sampling IMO.
Selig Audio, LLC

RobC
Posts: 1833
Joined: 10 Mar 2018

06 Jul 2022

vectro wrote:
05 Jul 2022
Absolutely! You can do stereo, M/S and many modes. You would only need two instances if doing Quad/"True Stereo" where you have 4ch convolution outputs from 2ch inputs.
Darn, it's a bad habit of mine to just check a new or updated device, and if I think it's missing something (or during that time it seems that I won't be using it), then I tend to drop it. Sometimes, only later I realize it has a useful feature, or something I didn't discover yet...
selig wrote:
05 Jul 2022
The best current technique is the swept sine technique, which requires a full range speaker (ideally omni-directional!) to get a response that accurately represents the audible range of frequencies, and it should be as flat as possible as well. Imagine a snare drum in a room, radiating sound out in a sphere away from the drum. Putting a two way speaker facing up will only get you so far in re-creating the reflections that the snare can create. So it is important to start with a source that radiates like actual instruments in as much as this is possible.
From there, you are recording the room so all decisions made will affect the results - what types of microphones used, how they are positioned (what configuration), WHERE they are positioned, etc. Just like with any recording, you can send two teams into a space to record an impulse and get two very different results.
This parallels sampling IMO, because at some point you can't just keep taking more samples. Computational/algorithmic approaches will be better at giving complete control over a virtual 'space' just as modeling gives more control over instrument design than sampling.
There are tradeoffs with both approaches, but all to say convolution has similar limits as sampling IMO.
Looking it up, I didn't even know such speakers exist (neither did I think of the usefulness).
Reading that, I just realized how silly it is that I barely thought about how acoustics work. After all, I mentioned that I have a concept for a 3D effect processor RE, which exactly considers how sound travels.
Well, to get accurate results, I definitely would need more purpose built tools ~ the speaker, and matching microphones.
Currently, if I want to add a realistic touch, it looks like I'm most likely only gonna be able to properly make just M/S stereo recordings of a one time performance or recording some odd sound. Or a bit weird/colored binaural recordings, since my microphones are different. Heck, the old one is a USB one and limited to 16 bit. Unless I could take it apart, remove the AD/DAC and modify it to output to XLR. But that's a different topic.

User avatar
Iapetus 9
Posts: 199
Joined: 18 Jan 2015

08 Jul 2022

While purchasing cab IR's is the easiest way to get them, there are a handful of great FREE IR loaders out there in VST land. No need to pay for anything like that unless they come with some cabs that you like. I use the awesomesauce NadIR loader from Ignite Amps, and all my favorite Fender IR's to make my Kuassa Vermilion RE sound tons better.
38L > 51D every time.

User avatar
MrFigg
Competition Winner
Posts: 9124
Joined: 20 Apr 2018

09 Jul 2022

Iapetus 9 wrote:
08 Jul 2022
While purchasing cab IR's is the easiest way to get them, there are a handful of great FREE IR loaders out there in VST land. No need to pay for anything like that unless they come with some cabs that you like. I use the awesomesauce NadIR loader from Ignite Amps, and all my favorite Fender IR's to make my Kuassa Vermilion RE sound tons better.
Thanks. yeah, I remember trying that NadIR loader a while a back. In any case, I ended up getting the Audio Assault aIR Loader for something like $2.50. It not only loads IRs but also allows stacking of IRs and has a whole bunch of modules for further shaping the sound. It's pretty cool.
I also bought the Audio Assault Vox IR pack. The Kuassa VSTs have a built in IR loader. Been loading the IRs into Lancaster and it's sounding totally brilliant. Seriously contemplating the Orange and Mesa IR packs to use with Clarent and Rectifor :).took

Edit: Hmm...just took a look at Audio Assault's homepage and aIR Rack is gone from the shop. Hmmm
🗲 2ॐ ᛉ

User avatar
Iapetus 9
Posts: 199
Joined: 18 Jan 2015

16 Jul 2022

MrFigg wrote:
09 Jul 2022
Iapetus 9 wrote:
08 Jul 2022


Edit: Hmm...just took a look at Audio Assault's homepage and aIR Rack is gone from the shop. Hmmm
It's still there, on page two for 6 bucks. Looks interesting.
38L > 51D every time.

User avatar
MrFigg
Competition Winner
Posts: 9124
Joined: 20 Apr 2018

16 Jul 2022

It had disappeared. I wrote and asked him and he replied that it was back in and it’s getting an update in the next few days.
🗲 2ॐ ᛉ

User avatar
MrFigg
Competition Winner
Posts: 9124
Joined: 20 Apr 2018

17 Jul 2022

Forgot to say…code AAFIFTY at checkout should get you it for $3.
🗲 2ॐ ᛉ

Post Reply
  • Information
  • Who is online

    Users browsing this forum: No registered users and 12 guests