Opinions on Ambika as soft synth

I am reading through the GitHub code for Ambika and I do see the avr library referenced, and some things like that, but it doesn’t look super daunting in terms of complexity.

Since I am really not experienced with hardware at all, and I am a very experienced programmer for a living, but not a very good one, I was thinking what if we just make Ambika a soft synth?

I know from some people’s point of view that might be missing the point, but it would be cool to be able to, for example, make the software version of the controller able to address say, 32 voices? and it would be up to the user to have a computer that could support that.

(you could use something like JUCE with osc control to edit the synth, so you could run that on your iPad or anything else, and the Ambika could run on its own box. that might sound weird, but for example I have a 2U rack server I am building to run soft synths and I want to control them or the VST host from a remote device and not have the keyboard, monitor, etc. But that is even more off-topic. )

is this a stupid idea?

The code base won’t get you very far in terms of sound because the “magic” of the Ambika is in the super crappy signal path (cheap 12-bit converters for audio, dirty 8-bit PWM CVs, analog filters).

So either you work hard in emulating all that, or maybe you take another approach and build a poly VA using Braids or Plaits’ oscillators (what the “Ambika Pro” would have been).

Ok, that’s funny. Ok I am peering at the legacy ambika site and looking at the PCBs for the filters…

Just listening, the Ambika has a very pleasant sound, I am still having a hard time looking at a project like this as code + PCB designs etc and forming a good opinion.

I was thinking that since Ambika sounds good, and people like it, and it seems simple enough in concept for me to get my head around, then it follows that it might be worth it, given my limitations, to make this a first attempt at translating a hardware synth that is known to be “good” into a form that I can shape.

Thanks for your comments, I am looking at Plaits, we will see if I could go forward or not.

My idea for making this soft synth isn’t a reflection of lack of respect for all the MI hardware, just the opposite. But I am still mostly in a software-based world of polyphony, and by increasing processing power, you can effect positive changes in voice count, realism, etc.


Don’t know about stupid, but it does seem like looking at things backwards.

With literally hundreds of soft synths out there with all manner of controls and GUIs, the landscape of digital in-the-box [sounding] synths are well and truly covered.
The whole purpose of building an Ambika for me is for the sound which I can’t get with any soft synth. As pichenettes said, the magic is in the imperfections in the digital and analog circuits which computers still have a hard time emulating.

I now use Ambika for all my synth sounds because it sounds firkin awesome, but I control it with the plugin interface because the Ambika’s controls get a bit tedious after a while.

So what you’re suggesting is keep Ambika’s clumsy controls, and loose the hardware that makes it sound how it sounds? That’s crazy talk!