User Details

Username: OspreyInstruments
Website: PeterCorbettEE.com
Facebook: None provided
Twitter: None provided
Soundcloud: www.soundcloud.com/peter-corbett-370535512
Account Status: enabled
Member since: Aug 12 2022
About OspreyInstruments: Hi!

I'm a recent graduate of Electrical Engineering and a long time musician/audio engineer/producer. I have recently started to dabble in hardware design (synthesizers and effects) as well as DSP. I hope someone finds some value in what I've made so far :)
(report this profile)

Devices by OspreyInstruments

PM Synth Version 1.0
Optical Version 1.0
Panner Version 2.0
ModBox Version 1.0
FM PM Demonstration Version 1.0
Formant Version 1.0
VOC-201 Version 1.0
ZenEQ Version 1.0
B-314 Version 1.0
ADT-2 Version 1.0
Scanner Vibrato Version 1.0
Scanner Vibrato Mono Version 1.0
Spring Reverb Version 1.0
Rotary Speaker Version 1.0
ADT-1 Version 1.0
LoFi Version 1.0
FocusEQ Version 2.0
JChorus Version 1.0
Mono Bass Version 1.0
LA2A Version 1.0
Kick Machine Version 1.0
Partials Version 1.0
Scanner Vibrato Version 2.0
Rotary Version 2.0
Harmonics E-1 Version 1.0
Hysteresis Version 1.0
Snap Version 1.0
Trilby Version 1.0
Clapper Version 1.0
FM PM Demonstration 3OP Version 1.0
FlexSat Version 1.0
FlexSat Version 1.0
JackedBrat Version 1.0
FlexComp Version 1.0
Tonewheel Organ Version 1.0
Lunchbox EQ Version 1.0
VOC-201 Version 2.0
Chebyshev Version 1.0
Spectacle Version 1.0
VibeStrip Version 1.0
Hysteresis V2 Version 1.0
Cambridge Bloatationizer Version 1.0
Echotron Version 1.0
Twin Reverb Version 1.0
Boba Version 1.0
F-Coder Version 1.0
Feedback Chain Version 1.0
Mid-Side Version 1.0
Phase Distortion Version 1.0
Kick Machine 2 Version 1.0
Tanh plus Version 1.0
Model D Version 1.0

Total Downloads: 4,717


Comments by OspreyInstruments

Comments

@JohnyRingo I finally made a portable version (on the KoFi downloads page) that includes the library in the device file. Hopefully this resolves the compatibility issues.

You still do need java unfortunately. There's no easy way I could think of to split the stereo buffer besides a java based buffer operations object :s

@JohnyRingo

Your M4L user library is pointing towards an external drive? Or your other ableton libraries? Sorry, yes the file path thing is super finicky, and I'm still building a picture of what does and doesn't work for different people.

Did you install Java then? I finally had a breakthrough on this one. There was a single mxj object I'd used for processing one of the IR buffers (and totally forgot about). This one object does require java it turns out... not something I realized at first because I'd already had java installed.

https://cycling74.s3.amazonaws.com/support/java_install.html

This is the cycling 74 page that points towards the java download. Theres an M1 (arm) version that you'd use in your case. I'll add a note to the instructions regarding the java prompt, I know it scares a lot of people to be asked to install it. Maybe I'll come up with an alternative way to perform this process without using a mxj block.

@Torley Randomization is a good idea, especially for the IRs and noise. (could get dangerous in the effects section, especially the feedback and distortion hahaha). I'll think if there is a good way to implement.

Hey all! Thanks for the feedback!

Yes there was some problems with the load sequence it turns out, and the IR and sample libraries weren't loading correctly. It's my first time taking this approach of "auto scanning" these relative directories, and there was some unforeseen challenges.

I posted an update that should hopefully fix it for everyone. There is a separate LoFi.amxd file outside of the archive that you can download and replace the original with, so you don't have to redownload the libraries. Please let me know if you are having problems though still after this update!

Thanks for your comments! Glad to hear it is working well for you guys.

I'm trying to track down bugs that a couple users were experiencing so its valuable to hear its working well for some/most people :)

Thanks Freddy :)

Hi! Thanks for the comment!

I actually explored making it different lengths like you mention, and I think it would be a powerful feature. After playing around with the idea, I had to scrap it, as it would be A LOT of work (this already was a ton of work) and it looked very difficult to get it stable. As it is working now, this triggers the envelopes once at the start of the bar and calculates all the envelope timings as a fraction of beats, so it quickly gets unstable and looses timing, especially when dragged out longer than 1 bar. :S

I will think over whether there might be a practical way of implementing a randomizer, but my initial thoughts are there probably isn't. I based everything around the mc.function modules and that leads to quite a few restrictions. If I come with a way to do it, I will definitely post an update though!

Thanks for checking it out, and for the feedback! :)

Thanks for the comment, telkmx! I'm glad you like it.

A randomise button would be a really nice feature, but I shutter at how much work that would be to implement! haha (although there may be some methods for working with the matrix objects that I am not aware of).

The probability would have to be weighted at the lower end of the value structure too, especially on for the diagonal (feedback) cells. I'll have to think this over and see if I can come up with an efficient way to structure it. :)

Panner was updated to include a Phase mode, just like the original. This is a phase rotation with a center frequency of 114.5Hz, and depending on the pan, the left/right channel will sweep a frequency range of 55-177Hz (phase, not magnitude), as measured by PluginDoctor on the original. The Q also changes slightly, but unlike the original, the Phase level knob controls the mix with the dry signal as well as the amount of phase divergence, so that at a Phase level of 0, there is not a static phase shift at 114.5Hz.

Unfortunately, to keep the math simple, I used linear scaling, but given the quite limited frequency range, this shouldn't affect the parameter scaling too much, and is still perfectly functional.

Thanks for checking it out Calagan! I'm sorry you were having trouble getting them to work on M1. I had my friend beta test them on his machine and I might have a work around. I made a Pack with the organ sample library as well, and that worked for him. I can't upload .alp files here though, so if you are interested, I shared it on Pianobook (https://www.pianobook.co.uk/packs/hammond-l102/) and my personal website too (https://petercorbettee.com/projects/tonewheel-organ-dsp/ at the bottom of the page).

If its those 2 effects, its a problem with the convolution. Its either not including the external library or the IRs. I was having trouble exporting it correctly, maybe because I don't have the full Max MSP license. I think the pack / .alp should work though!