FREE Custom IRs for Your Guitar and Its Pickup (Acoustic Guitar IR Impulse Response)

[Cuki79]

Originally Posted by -GF-

I don't, unfortunately.

Then you should probably not go the IR loader way.

Originally Posted by -GF-

I had hoped I could find some IRs I could load onto a pedal

Might or might not work great for your uke... noone knows

Originally Posted by -GF-

Have you seen the Xvive Mike pedal I mentioned? It appears to have 9 preset acoustic IRs you can select.

Did not try. There is no reason for them to sound bad with the guitar they were created for.

Originally Posted by -GF-

I wonder how they match up with the stuff in your database?

It does not make sens really. The IR from the Xvive were made with 9 guitars. For those 9 guitars, none of my IR database will sound better.

My IRs were captured from very specific guitars, for those, the Xvive will not compete favourably...

Nobody can tell he made the most universal IR because there is none. That's why Fishman from the very begining with the Aura proposed an AURA Image database from which people could pick his best match.

Everybody who had the chance to compare a "best pick" to a "custom capture" from Fishman said the "custom capture" was far superior.

There was the same feedback when Tonedexter came out.

Aaron Short compared recently different IR pedals that "can" capture the IR:


I am afraid in your case, there is no cheap solution. You can try the Xvive but it's going to be a hit or miss.... You should definitely try before buying.

[-GF-]

Originally Posted by Cuki79

Then you should probably not go the IR loader way.

Might or might not work great for your uke... noone knows

Did not try. There is no reason for them to sound bad with the guitar they were created for.

It does not make sens really. The IR from the Xvive were made with 9 guitars. For those 9 guitars, none of my IR database will sound better.

My IRs were captured from very specific guitars, for those, the Xvive will not compete favourably...

Nobody can tell he made the most universal IR because there is none. That's why Fishman from the very begining with the Aura proposed an AURA Image database from which people could pick his best match.

Everybody who had the chance to compare a "best pick" to a "custom capture" from Fishman said the "custom capture" was far superior.

There was the same feedback when Tonedexter came out.

Aaron Short compared recently different IR pedals that "can" capture the IR:


I am afraid in your case, there is no cheap solution. You can try the Xvive but it's going to be a hit or miss.... You should definitely try before buying.

Thanks for such a detailed reply.

What's the cheapest audio interface and microphone you think would make something good enough for your database? I think this is an excellent project.

Are IRs truly individual? Or do you think there will come a point where a product will be available that can generate an IR based on presets the user selects like pickup type, instrument make, model etc?

[Cuki79]

Originally Posted by -GF-

Thanks for such a detailed reply.

What's the cheapest audio interface and microphone you think would make something good enough for your database? I think this is an excellent project.

Are IRs truly individual? Or do you think there will come a point where a product will be available that can generate an IR based on presets the user selects like pickup type, instrument make, model etc?

Focusrite and presonus are the leader as far as small audio interfaces. They are both great.

Product based on presets existed first before individual Capturing IR were made.

They are cheaper and less efficient at reproducing your instrument acoustic tone
Here are some examples:

http://acousticir.free.fr/spip.php?rubrique11

Careful I am not sure they have any Uke image.

Both pickup and instruments are important for best match.

[Cuki79]

Note that it is not worth buying an audio interface for that if you don’t have the use.

If you have a mic, consider buying a Tonedexter or Baggs VPDI.

If you have none. Try the fix IR pedals first before buying.

[-GF-]

Originally Posted by Cuki79

Note that it is not worth buying an audio interface for that if you don’t have the use.

If you have a mic, consider buying a Tonedexter or Baggs VPDI.

If you have none. Try the fix IR pedals first before buying.

Thanks Cuki.

Your latest post made me remember that, while I have none of this kit, a friend of mine does and I expect he would be willing to help me record some IR tracks to use.

I understand why you strongly recommend each user makes their own recordings, but I started looking at IR pedals because they held out the promise that I could actually improve my uke's sound - not just make the amplified sound closer to what my uke sounds like naturally.

I'm still interested to play around with IRs.

I've been hunting around and come across the Sonicake IR pedal, which seems even lower cost than the NUX mini studio...



It looks like I can pick that up for £35 ($50) new.

Are all IR loaders equal? Is that pedal worth considering?

Sorry for so many questions. It's a really interesting subject.

[Cuki79]

The sonicake has 8-9ms latency (really bad)

and here is a feedback from JazzyJ (check in the forum his review)

Sonicake Sonic IR doesn’t work....
Just to say that the Sonicake Sonic IR pedal didn’t work for me. Using IR’s made from Cuki’s algorithm I loaded them into the Sonic IR pedal and it does nothing to change the sound. The IR patch and a bypass patch sound exactly the same - unprocessed!

Spent about 3 hours trying different IR’s (mono, 24bit, 16bit, 2048, 1024 etc, etc.) and NOTHING works. You can load IR’s but they do nothing to the sound. You can play and change though 3 or 4 different IR’s that you’ve loaded in and the sound doesn’t change - it’s just your normal piezzo sound!

I bought the pedal new for £29.99 GBP on Ebay and am returning it. I think I’ll give the Nux Mini Studio a try.

I’m a Tonedexter user, but looking for a mini pedal for a simple set up too....

[knelt]

Originally Posted by jonfields45

EDIT: You will find up to date documentation and the ability to download the IR generation script directly on this website:

http://acousticir.free.fr/spip.php?a...ar_mode=calcul

Check out this AGF post for Doug Young’s demo of my IR (and Cuki’s):

https://www.acousticguitarforum.com/...3&postcount=12

The Easy Way to Generate a Custom IR for Your Guitar:

Email ([email protected]) me a Dropbox or Google Drive link to your recording.

If your .WAV file is less than 25 MB it can be sent as an email attachment. If you Zip up your recordings you can include several in a less than 25 MB email attachment.

Generating the IR takes only seconds in Matlab (longer in Octave). If you want, send a few recordings with different mic placements. I like 8” from the neck/body joint, but a more distant mic placement might make what is a better IR for your usage.

1. .WAV file at least 60 seconds long (my program only uses the first 60 seconds of the recording)
2. Pickup left, Mic right, set your DAW for the two to be about equal in volume
3. No clipping
4. Try to keep your peaks within 15 dB of clipping (decent SNR)

I like open position strumming followed by “E” bar chords going from F to C followed by a little fingerpicking. Play whatever you think best exhibits your playing style and instrument. I’m not sure it really matters. Doug Young’s recording was a fingerstyle tune with occasional strums. I won’t be listening to your recording so don’t worry about impressive playing :~).

I will send you back three IRs for each recording. They will be at the same sample rate as the recording you sent me. The one with 5050 in its file name will be 50% IR and 50% pickup for IR loaders that don’t have a mix function. The one with 1024 in its name is 100% IR and truncated to 1024 samples for cheaper IR loaders and experimentation with a shorter IR. The other IRs are truncated to 2048 samples.

The three most common errors users have encountered are:

1. To forget to pan the pickup left and mic right when exporting the sample from their DAW.
2. The sample WAV file is less than 60 seconds.
3. If running the script themselves (see next section) they forget to enclose the name of the wave file in single quotes, or include the extension instead of just the wave file name.

The Do It Yourself Way:

EDIT: If using Octave seems unintuitive, check out this post for help: https://www.acousticguitarforum.com/...76&postcount=1

Download the free GNU math development environment Octave and follow the installation instructions:

https://www.gnu.org/software/octave/download.html

Download the Octave version of my IR generator:

https://drive.google.com/file/d/1mKO...ew?usp=sharing

Place your guitar pickup/mic recordings in the same directory as my script, jf45ir.m, and double clip (run) jf45ir.m. The GUI version of Octave will open in that directory. Alternately, after installing Octave, run the GUI version and change the current directory to the one in which you placed my script and your recordings.

If your recording is called mgit.wav, for example, then type the following in the Octave command window followed by a carriage return (don’t forget to include the single quotes before and after the file name and don’t include the ".wav" extension):

jf45ir('mgit')

Octave will then generate four files, three IRs plus a frequency plot of the 100% IR.

jf45irmgit.wav
jf45ir5050mgit.wav
jf45ir1024mgit.wav
FFTjf45irmgit.jpg

Let me know how you make out with the IR. To my mind it is a pretty brute force implementation, but it seems to produce useful results.

Thanks,
Jon

Are you still offering this? I would be absolutely delighted to get my guitar done.

[Cuki79]

Originally Posted by knelt

Are you still offering this? I would be absolutely delighted to get my guitar done.

Jon's algorithm is still available on the website.

Note that Aaron's demo I used my own program which is different. There is an free app with a more user friendly interface to generate IRs using either my algorithm (a light version) or Jon's.

http://acousticir.free.fr/spip.php?rubrique15

If you don't want to bother making the IRs just send us the recordings. There are instructions at the links on the bottom of the webpage above. You can pm me the link of your files.

[Lid555]

First off- thanks a ton to Jon and Cuki both for all their research, knowledge, and resources.

My situation is a little different as my main IR goal has been to make my clean electric guitar sound more like a mic’d acoustic (and particularly to erase piezo “quack”). Though I’m still working on controlling the overall balance of the IR, I’m elated to say I’ve transformed the quack into a more natural sound.

Since part of my process requires mic’ing an unamplified electric guitar (and therefore requiring a hefty amount of bass EQ after the fact), I’ve found using Jon’s Octave code along with the process of editing and analyzing the results in a DAW to be immensely helpful.

I’m hoping someone could add to my knowledge of using different sample and bit rates throughout the process. So far, I’ve been using 16bit, 44.1kHz throughout the process- with the sole exception of recording into Cubase at 32 bit floating. But would it be beneficial to use a higher bit rate and/or sample rate at any other point along the way, for example, into the IR generator?

- Brett

[jonfields45]

Originally Posted by Lid555

But would it be beneficial to use a higher bit rate and/or sample rate at any other point along the way, for example, into the IR generator?

- Brett

IMO no.

The double precision floating point Octave defaults to is useful for my script as there are issues in taking very large data sets through an FFT (long data sets reduce quantization noise in the frequency domain but require extra precision for the transcendental arithmetic). For your purposes with audio, 16 bits (96 dB) is plenty to take you from the noise floor of any reasonable room (not an anechoic chamber) to the threshold of pain.

Higher sample rates than 44.1 KHz waste more of the frequency domain data points on content higher than what any guitar can produce. Also in real time, the final FIR filter is shorter when you run it faster which trades off bandwidth for frequencies you can't hear for low frequency resolution you can hear.

Probably too technical!

I tried what you are doing once and could not get an acceptable mic recording.

[Lid555]

Thanks for the thorough response Jon! Ok so then I’ll generally just keep going the way I have been: keeping the audio for the IR process at 16bit/ 44.1kHz.

For the mic’d recording of my unamplified electric:
- my electric guitar is set up as much like an acoustic as possible- higher action, high gauge/“heavy” strings, and I’ve found using a harder pick is louder and clearer
- room treatment in my home studio
- decent audio interface and preamps (Steinberg UR44)
- recording at 32bit floating into Cubase- my prior research pointed me in this direction- adding more values to play with between max and the noise floor (mainly here preparing for compression) - though it sounds like 24bits is sufficient, I figure since my system can handle it easily enough, and since it’s been my default for years, I might as well continue using 32floating- though I admit I haven’t really experimented much with other bit depths
- compression (two stages- but surprisingly less than I was expecting)
- eq- not surprisingly, the signal needs quite a hefty boost of low frequencies

So my setup is definitely not for every guitarist

Probably due to the higher noise floor(?), I’ve also had to add a process for each IR I generate. I add about a 15-20ms fade to each IR, or I get a faint distortion through at least some IR loaders. I’m mainly using Thafknar on iOS and Voxengo Boogex on macOS.

[Cuki79]

Originally Posted by Lid555

- compression (two stages- but surprisingly less than I was expecting)

In principle you should not use compression… but in principle you should not use a magnetic pickup either. Both make the system not adapted for IR.

The whole IR convolution is base on linearity…

What the guitar world calls « IR » is a Finite Impulse Response convolution. This process belongs to the Linear Time invariant Filters family.

Compression is a process in which gain changes with the attack… by definition it is not invariant in time.

So you are training your filter to perform a process it can not do…

To conclude:
* is it forbidden to use compressor and magnetic pickup? No
* will it work? To a certain limit…
* will it explode? No
* will it be accurate? For sure No.

The telecaster was not designed with rock n roll in mind… as long as you think the result is musical…

[jonfields45]

There are IR systems based on magnetic pickups, such as the Roland Guitar Synth, which when not being used as a midi converter driving a standard synth, can run a per string IR to emulate different electric and acoustic guitars. In my opinion it does not do a great job emulating acoustic guitar compared to Line6's piezo system.

To make the magnetic a better fit for an IR, placing it right at the bridge as Roland did makes the pickup more linear (or more playing position or time invariant). As you play up the neck, the pickup is still in relatively the same position regarding the vibrating string. For example, the neck pickup gets more into the center of the string as you play up the neck which reduces the harmonic content.

For your project, I would see if the bridge pickup is usable, maybe too much treble to recover from... Or I would train the IR for the neck pickup in the open position with the idea that it will work best in the open position.

[Cuki79]

Originally Posted by jonfields45

There are IR systems based on magnetic pickups, such as the Roland Guitar Synth, which when not being used as a midi converter driving a standard synth, can run a per string IR to emulate different electric and acoustic guitars. In my opinion it does not do a great job emulating acoustic guitar compared to Line6's piezo system.

I don’t understand what you say Jon.

A guitar synth by definition is not IR based. A synth means the sound is triggered? Whether it is analog (VCO) or digital, the sound is still triggered…

When you mention that individual IR are run per strings do you have any reference of that? It would require 6 times the processing power of an IR pedal which is a lot …

Then you compare this idea to line 6 piezo… I think the variax is probably something close to what you said but probably mostly based on IIR I guess to sav3 computing power. [update: it is indeed FIR but not as long as what we usually do with our processes]

[Cuki79]

From line 6 patent: String instrument with built-in DSP modelling

« As will be discussed, in one embodiment of the invention, an aspect of the emulation of the corresponding string of the selected guitar is achieved using a finite impulse response (FIR) filter. The emulated digital audio signal is then converted to analog form by the D / A converter 215 implemented to produce an emulated analog audio signal for output to an amplification device. »

For me, Roland trigger sounds while Line 6 do signal processing.