MUFF WIGGLER Forum Index
 FAQ & Terms of UseFAQ & Terms Of Use   Wiggler RadioMW Radio   Muff Wiggler TwitterTwitter   Support the site @ PatreonPatreon 
 SearchSearch   RegisterSign up   Log inLog in 
WIGGLING 'LITE' IN GUEST MODE

Creating a eurorack filter
MUFF WIGGLER Forum Index -> Music Tech DIY  
Author Creating a eurorack filter
Anjunagasm
Repost from a different thread:

Okay so this probably will sound stupid, but I created a filter rack on Ableton that I would love to clone in Eurorack format. But physically I could make it better because I could manipulate it exactly how I want if I personally make it or help someone make it for me. The Ableton rack I made has a very unique and characteristic sound due to me using different EQ and filter plugs for making each filter type sound as good as possible. I would like advice or maybe even to commission someone for help making this filter. I’m not even positive that I will do it, but I would like to know that it’s possible.

I have no idea how much it would cost, or even if it’s practical to make it. But I will post the filter types. It’s made out of EQ 8, Pro-Q2, and ableton’s Auto Filter. It’s got a LP 12db/oct, LP 48 dB/oct, HP 12db/oct, HP 48db/oct, bandpass with multiple settings I’m not sure of right now cause I’m not at my computer, a comb where you can add up to 12 tines, a spike filter which is like the opposite of a comb with 8 tines, and a vowel filter with a butcrush. This would probably be impractical to make, but I just want to know if it’s even possible.

I’m not near my computer right now so I cant post sound demos at the moment, but I will come back later.
cackland
Definitely provide a visual representation of the features / parameters / outputs this concept has. This will enable those to envision behind the panel better as to how to design and connect individual features.
EATyourGUITAR
I'm really trying hard not to laugh or say something mean. what you have done right there is take software simulations of $10,000 of audio equipment and ask if it is possible to make one of the following

A) a eurorack module that runs software that Ableton Gmbh has the source code and ownership of. you need a very fast computer, something like an actual computer. you need an excellent audio interface. something like an actual audio interface. then fit it inside a 24HP eurorack module that runs on 200ma.

B) clone $10,000 worth of audio equipment, put everything behind a 24HP eurorack panel, write software for a small microcontroller to manage presets, routing, switching. pay someone by the hour to develop not one filter. you want them to work on 10 filters. all the presets. write the software. build and test the prototypes. design and test the production prototypes. manage the factory production and distribution. sell it for $500.

now if you actually want to make a eurorack module. you need to actually know how to make eurorack modules. your education in software is probably not going to help you at all. it only helps you as a user of eurorack, an audio engineer. not so much as an electrical engineer.

if you want to make presets for ableton and sell them on the internet that is definitely a viable business model. you may even grow that business into selling max msp plugins for ableton. you may even hire a programmer to write more efficient VST plugins in C that are crossplatform. Richard Divine has made a career out of being a preset designer without actually being a software programmer.

software engineers and electrical engineers have different constraints to work within. you can have a small system or a huge system of modules in the computer for exactly the same price of a computer. a lot of times a $1000 computer will simulate $20,000 of real equipment. for some reason, people still buy eurorack modules. a lot of them are working great with no software at all. I'm not going to get into why real analog modular synthesizers are better. but they are different and they will have different things that make them great or make them limited. the simplest forms that can be made into working audio circuits dictate the low cost design of eurorack mdules. the workflow is a result of the modules. not the other way around.
Anjunagasm
I mean I figured it was impractical, just wanted to ask. Haha thanks for you input.
cackland
EATyourGUITAR definitely has a point, however don't let that put you off.

I would suggest mapping out the individual blocks of how you've built your rack in Ableton and then break it down into components that can be re-produced in a tangible device.

Does 'this' filter need to be digital, or would you want an analog implementation?

It's always best to do some preliminary drawings on the interface and what each segment / control would do and how they would interact with each other. Then digging down in the technical aspects / elements on each.

Obviously if you don't have any electrical engineering experience, then outsourcing this project is your only way. And if 'digital' is the route which you want to take, having the tangible aspect in eurorack, then your DSP skills will have to be up to par.

Without any knowledge of all of this, its hard to give you further directions.

If curiosity gets the better of you (like it does me), then research and dive into the world of engineering.
Sandrine
The new Teensy 4 is a very powerful dev. board that's 100mA I believe.
So that and a decent codec chip / low noise amps will get you started.
Then a little bit of fancy software and it's good to go!
d'oh! hihi very frustrating applause zombie thumbs up
forestcaver
It’s not a silly question. If I was going to approach this, I’d decide whether to build it in analog (will be phyiscally large with multiple modules) but conceptually simple once you describe the filters mathematically. There are lots of example filters that you could adapt to do this as well as many open source modules whose schematics could be adapted.
Or you could model it digitally. You could do this in many ways, depending on knowledge and experience of dsp eg teensy 4 (but the adc may not be sufficient) or one of the stm32 mcus or even simpler with a higher level abstraction by using one of the pd interpreters eg based on raspberry pi. You would need to be careful with your coding and probably cut some corners/optimisations to fit your processing power.
How difficult it would be to do depends to a large extent on what your ableton rack is doing as well as your knowledge and experience level.
Essentially, conceptually it isnt difficult. How hard it is depends on your level of knowledge and experience. You’ll *never* get anyone to do this for you though and you’ll likely never be able to afford to commission it.... (unless you are super-rich)
Ps I wouldnt personally do this as the huge effort wouldnt be anywhere near worth it to me when I could just pipe it into ableton :-)

PPs it the short term take a look at terminal tedium module - it may do what you want *but* you may struggle with latency/processing power
EATyourGUITAR
so basically make a simulation of a simulation with no coding experience and expect it to sound the same as it does in ableton without using any of the source code or programmers from ableton. now do it with less DSP power and inferior DAC's. Sandrine is an expert DSP coder. when she cheers you on and sums it up she is coming from the perspective that you are just doing what is easy for her if she were doing it. Personally, I do not think it is easy to do well. it takes years of practice. If you are already doing it for many years than that is a different starting point.

qubit nebulae V1 is two arduinos and a raspberry pi. they could have done it on the ARM chip in the pi and ditched the arduino's but they were basically slapping together some crap they found on the internet browsing muffs and shit. it runs linux. do we care that it runs linux? does it matter? this is just wasting power and making heat. it also wastes CPU cycles that would be better served as audio DSP. compare this to synthesis technology running DSPic in everthing. look at the power consumption. Mutable Instruments runs natively on an ARM chip that consumes less power than a Pi. I don't think we want a repeat of someone using high level languages full of abstraction and inefficiency to release a product for sale.
lacsap
without knowing any specifics on the routing in your filter rack, I would say a dedicated DSP would be a good choice. I experimented a bit with the ADAU1701 from Analogs Sigma DSP series. It's a smaller one with already 2 adc and 4 dacs and 4 adc for analog control of your patch parameters among other GPIOs. So in a modular synth setting one would need to design proper input output signal conditioning only, matching levels, offsetting etc...
You can program them with Sigma Studio which offer pd/MaxMSP style visual blocks for each element (eq, filters etc). So if you know some bits about audio processing it's quite easy to pick up. the hardest part being to connect the dsp to Sigma studio.

I did use the breakout board from sure electronics, and played around with some headphone crossfeed patches. I think the smaller boxes from minidsp use the same chip or another one from the same series from room eqing.
thetwlo
so, if it's to be "modular" you need a module that does each of those functions. Modular is about building blocks, like you are doing, once you combine them all into one--that joy, of endless possibilities is removed.
cackland
I see this as an opportunistic challenge. Sure it may make sense to keep this kind of 'filter' in the box, however if tangibility is the aim here then why not have a go.
Sandrine
cackland wrote:
I see this as an opportunistic challenge. Sure it may make sense to keep this kind of 'filter' in the box, however if tangibility is the aim here then why not have a go.


I agree, didn't mean to be condescending, just joking.
It *is* a steep challenge either way analog or digital.

The Pro_Q 3 would be really hard to make analog (especially if you wanted the reference display) to control with knobs (i.e. how many knobs? what scheme if using encoders...)
Digital would almost need to be touchscreen

Perhaps give that part some thought first, then decide on which method would suit the task.
cackland
Absolutely.

More info from the OP would help smile
thetwlo
actually, I have no clue what these "virtual/software" modules are, Are they emulations of hardware? What are they?
ixtern
48db/oct hardware analog LPF/HPF? Voltage controlled and tracking 5-8 octaves? I would like to see such design...
MUFF WIGGLER Forum Index -> Music Tech DIY  
Page 1 of 1
Powered by phpBB © phpBB Group