Resources for building a sequencer from scratch?

The Lounge General Discussion
#16

And the coding rabbit hole similar to GAS begins haha

I dunno I find that micro second latency is fine for game engines in JS I don’t see how a ms can make such a big difference but what do I know. I’ve only dealt with baud rates of 9600 when making USB controllers (not MIDI, radio/relay controllers) but I prefer to use memory safe languages like Rust/Go however C is the industry standard and can do a lot more in this world. You just have to be sure to manage memory well. And while C is static/compiled, the type system is ok.

At the end of the day, just pick something useful to get a prototype going while you learn hardware. Then refactor to low latency solutions moving forward.

I’ve heard great things about micropython for sure!

MaxMSP seems quite useful as well. Never knew about it! TIL

Have fun with it :sunglasses:

3 Likes
#17

The hardware side of things isn’t so difficult if you can find the resources of people that have already connected similar components.
Knobs an buttons are easy enough. I hooked a screen up to axoloti an it was straightforward.
A good soldering iron goes a long way - it’s really frustrating using a cheap one an the results can be pretty poor.

3 Likes
#18

I was just searching for something yesterday for doing a Euclidean Sequencer using an rPI and came across this:

I am, however, neither a hardware guy nor a coder. I knew some basics of C a long time ago and I’ve poked at Python before. I’ve always wanted to learn and figured “what better way than to learn by doing a project for something you’re passionate about!”

2 Likes
#19

I think Simon posted it somewhere on the forum a while back.

1 Like
#20

It’s not only the user experience, but clock timing internally as well as externally that is vital. If the device has a sloppy clock because of interrupts and a shared CPU with an operating system deciding resource allocation you are going to have a sloppy sequencer. Low latency and solid clocking (not just MIDI) is essential to audio/midi sequencers working well with themselves and each other :smiley:

Just search around for people not being able to get their hardware to sync with Ableton Live (on a computer with a shared CPU) if you want to see examples. If Ableton can’t make it work reliably… well you get the idea :slight_smile:

If you look at the majority of pro audio equipment, you’ll see that almost all of it is done on microcontroller or equivalent components(and for good reason). Much of it is written in Assembly or C to get as close to the machine language as possible.

5 Likes
#21

Q: **"Can you describe the testing process for a new machine? How long do you spend with a prototype, how many people use it, how many iterations does it go through before deciding on the specs for a production model?

I would call this more ‘design’ than ‘testing’ - while prototyping a lot of time is spent figuring out functionality and parametrization, i.e how it should work and how you should work with it. Testing this out at an early stage (prototyping) is crucial. An idea can be great in theory, but horrible in practice. For the Digitone I worked with Max and Gen to test everything, and when implemented on hardware we tested it further.”**

3 Likes
#22

https://youtu.be/6ArDGcUqiWM At first I found this guy really annoying but he has grown on me.

1 Like
#23

Thanks, that will be another great example for my Max for designers intro course :smile:

2 Likes
#24

Makes sense! Thanks for the detailed response!

2 Likes
#25

I haven’t tested any Elektron boxes recently, but back in 2015 MIDI Clock output from the Analog series was much less stable than what Ableton managed to generate on OS X though a cheap Behringer interface.

1 Like
#26

Following :popcorn:

2 Likes
#27

I just wanted to throw up some more axoloti love haha! Axoloti works great if you’re like me and not very good at coding as there are plenty of objects to build synths and fx units without digging deeper but if you do code then you can make and adapt objects so you get the best of both.

The SQ is pretty damn good and the microcontroller comes on a board with midi in/out micro usb an sd card and stereo audio in and out all done for you. To add physical components there are solder pads on the board for around 20 analogue and digital ins and outs and gound and power pads as well.
You can also store several patches on the sd card so it could be a synth, load another an its an fx unit. There are some of the mutable instruments open source modules on there as well.
Check this mini buchla number - the video in the first post has a sound demo : D

1 Like
#28

Midibox? http://www.ucapps.de/midibox_seq.html

I keep almost pulling the trigger on one of the new full kits from the midiphy guys https://www.midiphy.com/en/news/

I’ve built a Klee and have breadboarded a number of small 555 based sequencer circuits, but never anything from the software side. I think if I did I would start by looking at some of the teensy based eurorack projects on GitHub. Specifically the ornament and crime stuff. As a number of folks have written sequencers for that firmware.

1 Like
#29

+1 for trying out a few easy designs using 555, 4017 etc on a breadboard. It’ll introduce you to hardware and you’ll be able to interface any PIC/Arduino boards too. Do a bit of reading up on logic gates, shift registers, multiplexers etc. By reading I mean the data Sheets for the chips. Have a look around for some old diy electronics books, they’re a great reference resource while you’re learning.

3 Likes
#30

When i think of sequencers most lacking point at this time - is that general automation thing - i.e. volume curves, that lowers volume over time for X bars. I dont know why its so hard to program an automation layer, which is recording this in a linear fashion, or why its not possible to place a trig, which sets volume of the track to a certain level, or does a fade in/out after X bars.
(its ofc possible with abelton live etc.) Maybe circlon can do it - i didnt read its manual.
(Same for other midi CC is thinkable - essentially triggers that trigger mathematical functions /curves)

Also what i think would be cool, if there would be triggers, which are triggering whole musical phrases, instead of a single note. Maybe a track , which is triggering other tracks to start /stop.

1 Like
#31

That’s solid advice. I moved onto axoloti after building Arduino synths an tinkering with various other electronics so had some experience. Probably wouldn’t have gone straight to soldering an £80 odd circuit board.

2 Likes
#32

When you just want to get your hands dirty for the first time (without the need to build your own hardware first) I would highly suggest to make your own firmware for the Novation Launchpad Pro. It’s quite simply:

5 Likes
#33

It’s definitely not automatically better but it “can” be better. When you have some “unknown code (operating system)” sharing the processor you don’t have control over what gets priority all of the time.

When you write code on a microcontroller you are in charge of every piece of code running on it and what interrupts what when or what get priority over something else.

You can still make mistakes or compromises that can affect the clock quality, no doubt.

2 Likes
#34

In the audio world (read: the core processing loops) or when using micro controllers you won’t do any (dynamic) memory management at all, because it’s simply too slow and not predictable.

2 Likes
#35

Yea something like Rust comes to mind but the LLVM doesn’t have that many convenient targets for the world of controllers. So C it is for most of this I guess!

1 Like