Weald - A Wood Based Sound Effects "Processor" and IR Library

[michaelwayneharwood]

What would be the options for using the sine sweep (I presume it vibrates the body directly) at an appropriate location and then an open mic (not a transducer) on a series of bodies? E.g. acoustic basses? Maybe this has already been done, but then presumably feeding a wide spectrum signal from a standard bass via a piezo could be also modified by an IR pedal to provide some elements of the body sound? That sounds doable from an IR pedal perspective, but I don't have the bass or other gear to do the rest.

That's actually on my list of "to do's". I have access to two really nice contra basses that are owned by relatives. I am also going to sample the bodies of violas, violins, cellos, etc.

 

[michaelwayneharwood]

As a further comment, if an IR pedal had a limited processing capacity you might not be able to do bass->impulse 1->2->3 (e.g. 2 might be a vintage amp, 3 a cab), so a way to convolve 1,2,3 offline to load as a new preset would be useful.

That's doable as well.

 

[mapleglo]

I would imagine as well, that you could record the IR from a standup bass, and use that as a model for an electric. I could see that being quite useful.

 

[michaelwayneharwood]

I would imagine as well, that you could record the IR from a standup bass, and use that as a model for an electric. I could see that being quite useful.

Absolutely! What I have found is that setting your expectations to be reasonable is important. If you have a fretless bass with tape or flat wound strings, mute the strings a bit to remove some sustain, and use an IR from a large resonant wooden body (e.g. standup bass) you can get some very nice sounds. Here's the soundclip I used in previous examples processed using only an acoustic bass IR I found on the 'net. Notice how some playing styles accentuate the "acoustic bass" qualities, and others not so much.

 

[mapleglo]

Absolutely! What I have found is that setting your expectations to be reasonable is important. If you have a fretless bass with tape or flat wound strings, mute the strings a bit to remove some sustain, and use an IR from a large resonant wooden body (e.g. standup bass) you can get some very nice sounds. Here's the soundclip I used in previous examples processed using only an acoustic bass IR I found on the 'net. Notice how some playing styles accentuate the "acoustic bass" qualities, and others not so much.

Now that's what I'm talking about! Sounds really good. Is that from a stand up, or an acoustic bass? Sounds pretty close to a stand up.

 

[ole Jason]

In this context, the IR is just applying an EQ curve to the signal. It may be easier to work out the EQ signature from the IR (using spectral analysis, etc.) and then apply the signature via EQ instead of messing with loading literal IR's. Something like the Source Audio programmable EQ would be an easy way to do it.

 

[michaelwayneharwood]

In this context, the IR is just applying an EQ curve to the signal. It may be easier to work out the EQ signature from the IR (using spectral analysis, etc.) and then apply the signature via EQ instead of messing with loading literal IR's. Something like the Source Audio programmable EQ would be an easy way to do it.

I disagree, as the EQ curve is not the only thing captured.

*quick edit*

As I have stated previously you can certainly use EQ and reverb to get similar results, and if that's your point I agree with you. I much prefer the workflow I can achieve using IR's over EQ and reverb presets as they are generic and can be used in multiple ways, where a set of EQ settings would have to be copied to each device or plugin as there is not a set standard for moving those around to different pedals, outboard effects, plugins, etc.

 

[michaelwayneharwood]

Now that's what I'm talking about! Sounds really good. Is that from a stand up, or an acoustic bass? Sounds pretty close to a stand up.

The IR I am using is named "stirngbass.wav", and was found by scouring the backwaters of the internet looking for a stand-up bass IR. I could not tell you anything about it's origins.

 

[michaelwayneharwood]

In the midst of other projects involving cardboard, glue, and tape with the kids I managed to get the "sound isolation chamber" started. It's sole purpose is to isolate the microphone and wood samples from room reflections and get as pristine a recording as possible. I know it's a bit ghetto, but the rig I was using before was worse.

The original rig:

The "in progress" isolation chamber:

 

[michaelwayneharwood]

@ole Jason - One nice thing about what I will be providing is that if you want to capture the EQ curve and plug it into your workflow, you will have the IR's to do with whatever you wish.

 

[ole Jason]

Ya that is true. Generally for live applications, most people aren't running IR loaders but pretty much everyone has access to an EQ. It would be interesting to take your experiments beyond just wood. IMO, from what you've shown so far, the IR's really don't impart anything like what they represent in real life but they are still interesting. Maybe make an IR out of a sheet of aluminum

 

[-Asdfgh-]

Ya that is true. Generally for live applications, most people aren't running IR loaders but pretty much everyone has access to an EQ. It would be interesting to take your experiments beyond just wood. IMO, from what you've shown so far, the IR's really don't impart anything like what they represent in real life but they are still interesting. Maybe make an IR out of a sheet of aluminum

Ah... to discover the best metal for bass for metal?

 

[michaelwayneharwood]

Now that's what I'm talking about! Sounds really good. Is that from a stand up, or an acoustic bass? Sounds pretty close to a stand up.

One thing to note about this particular IR (the "stringbass" IR) - the sound clip I posted above was recorded prior to applying the IR. This is significant because if I was playing the bass "through" the IR my playing style would automatically adjust to the effect in the same way my playing style would adjust while playing a guitar through distortion. If I was looking to sound more like a stand up bass I imaging I would approach the sound much closer after a bit of practice with the IR and finding the playing styles, positions, etc, that accentuate the sound I was looking for, and perhaps putting a bit of foam underneath the strings by the bridge would be done as well.

 

[michaelwayneharwood]

Ya that is true. Generally for live applications, most people aren't running IR loaders but pretty much everyone has access to an EQ. It would be interesting to take your experiments beyond just wood. IMO, from what you've shown so far, the IR's really don't impart anything like what they represent in real life but they are still interesting. Maybe make an IR out of a sheet of aluminum

I have quote a few ideas of what could be, in my opinion, cool IR's. I have a wont to name things as befits my imagination, so bear with me on the titles:

• Voices of the Forest: Wood
• Voices of the Earth: Stone
• Voices of the Forge: Metals
• Forbidden Places: The inside of skulls, etc.

 

[-Asdfgh-]

It's now increasingly possible to model spaces or systems that do not exist, and presumably then synthetically create an IR for them. It's not computationally cheap or necessarily exact to model instrument bodies, though, so modelling a double bass body the size of a house would be challenging.

 

[michaelwayneharwood]

It's now increasingly possible to model spaces or systems that do not exist, and presumably then synthetically create an IR for them. It's not computationally cheap or necessarily exact to model instrument bodies, though, so modelling a double bass body the size of a house would be challenging.

That would be a big bass! For those types of applications I am going to be capturing actual instrument bodies...I am still considering names for that set of IR's.

 

[michaelwayneharwood]

The sound isolation box has been built, and a 5" by 8" by 1/8" birch plywood board has had a transducer attached in preparation for initial testing and final verification of the recording workflow. The isolation box looks *very* DIY, but it's purpose is to help isolate the microphone from room noises, and keep the reflections down from the wood when it's being utilized as a speaker cone. Initial tests have shown that it does help, though to do this right an anechoic chamber would be nice to have handy. I am recording into a Blue Yeti microphone, and using Audacity to capture the audio.

Here's what the workflow looks like for each species of wood:

1. Record a reference track that has both a high dynamic range and uses as much of the frequency spectrum as possible.
2. Record a sine sweep 3 to 5 times for comparison.
3. Record "taps" against the material with various implements such as a hard rubber mallet, a hard plastic handle, a wooden block, etc.
4. Convert audio into IR's.
5. Do comparisons between the original reference track with IR's applied and the recorded reference track to make sure the IR's are accurately capturing everything.
6. If there are differences in the results from #5 troubleshoot the process until the IR capture process is as fault free as possible.

 

[Mosfed]

Cool idea. More interested in the IR library that could be dialed into a pedal than the wood block idea but whatever.

In the end, @TheEmptyCell is right though- It's basically an EQ pedal. But an EQ pedal with some cool presets.

 

[michaelwayneharwood]

My tests came back solid, and the sine sweep methodology is providing excellent results. Just to be clear here's what I did to test things:

1. Recorded a sine sweep (call this the master or reference tone) through the wood
2. The recording was turned into into an IR
3. I then did a side by side comparison of :
a. The master sine sweep
b. The recorded sine sweep I made the IR from, and...
c. The master sine sweep with the IR applied

After checking with everything out with both ear and spectrogram the results are identical. The master sine sweep that had the IR applied has the same frequency response as the recorded sine sweep.

Next steps are to start running through the list of woods and generating IR's for everything.

 

[Rôckhewer]

Here's a list of the wood species that I will be able to create IR's for. I am waiting on a few generous souls to provide some of them, and as I get access to other species I will add them to the list. One of the "deliverables" of the IR library is to provide a PDF with spectrum plots of each of the IR's as a visual reference of the differences, but I am finding that listening to the way the IR's modify the signal is much more telling than a graph.

• Alder
• Ash
• Ash (Thermally modified)
• Balsa
• Basswood
• Bubinga
• Cedar (Western Red)
• Cherry
• Ebony (Pale Moon)
• Hickory
• Jatoba
• Katalox
• Limba
• Maple (Hard)
• Maple (Soft Quitled)
• Maple (Spalted)
• Mesquite
• Oak (Red)
• Pecan
• Pine (Radiata)
• Plywood (Birch)
• Poplar
• Rosewood (Indian)
• Sassafras
• Shedua
• Teak
• Walnut (Black)
• Walnut (Peruvian)
• Wenge

You can add two more to that list... they're in the box with the others:
Mahogany (African )
Oak (quarter sawn white)

Sorry for the delay.