Right now there’s a lot of hostility about the Quad Cortex and its development timeline, and a great solution would be for Neural DSP to craft more programmability into CorOS: allow the community to develop and maintain their own add-ons.
I believe it would help to alleviate the tension of waiting for new features for many, because there are a lot of people who have an engineering background or development skills who would be willing to code their own add-ons, which could be released and maintained in parallel to Neural DSP’s own official features.
Borrow from the success of editors like Vim, NeoVim, and VSCode by creating more of a community around feature development. I believe Neural DSP would be very surprised at what people would come up with if given access to be able to run code and advanced features on CorOS.
This would probably be best served via email to support@neuraldsp.com. With NDSP creating and managing their own code base, they are able to provide a one-stop shop for support vs If the code/access was made public, it would likely become a support nightmare IMO.
Thanks for the response! Actually, many, many pieces of software, especially in the open source communities, handle a plugin ecosystem without any issues: support requests go to their respective plugin developers. The main platform defers support to individual plugins to those of the creators of those plugins. It actually pushes developers to do better by free market competition.
i think the obvious solution is they just deliver.
i´m all for community “mods” and seeing what everyone does but i don´t think the community adding independent stuff is the solution and it just piles to a bunch of complications that could happen. not to mention i don´t think most would like to do free work, neural is literally hiring so if anything people would apply there
The hidden premise in your response is that only Neural DSP could write quality code for their device, but in reality a large number of the feature requests here could be tackled by talented developers who are willing to come up with and maintain their own extension because they think it’s cool. Just look at the synth/sampler/groovebox Synthstrom Deluge: it was all done by one person. Or look at the Poly Beebo. It’s a one-man operation, done by Loki Davidson.
There are many people like myself who have the skills, who would be more than willing to create an add-on if the platform were made programmable. There are many users who contribute to others projects on GitHub or GitLab or BitBucket, for free, at very high quality. We like making contributions to projects we use.
Just take a look at GitHub as an example: Topics on GitHub · GitHub Pick a language and scroll through the tens of thousands of projects there. All of them are open source, mostly from volunteers, and many of the code bases are used in software or hardware you probably use every day.
Not at all, i know there is a lot of talent and a lot of people outside of neural to do it. it just makes managing it more difficult and i dont think the solution to their timeline is making the community do work even if you guys are willing to work for free
i would love to see this implemented id just rather they focus the resources on what they promised and then do this type of thing
I think you both make good points. Neural needs to deliver and Open Source would be grand but has support and stability implications.
I would love to see an Open Source modeler as well as one that can run VST plugins for example, without requiring porting and with low enough latency to be usable. This would open up infinite possibilities. There are so many free VSTs available already, some of them quite good, that it is insane.
Along those lines it would be nice to have a standard protocol for addressing floorboard hardware so you could swap in Neural, Line6, Boss, Headrush, or whoever’s software and not have to swap out the floorboard every time. More of a “modeler as an application” approach.
I’m glad we’re on the same page. I think actually making it programmable would is a high level of delegation, which would make them more focused rather than distracted, because there are more hands to meet the demand. There are some niche/edge-case feature requests that, for we are in that niche, we can probably never expect Neural DSP to entertain the feature request, but we could develop our own add-ons. Overall, Neural DSP could then truly focus on what matters.
I believe that’s what the Mod Devices Dwarf was meant to be. I ordered their kickstarter campaign version around the same time the QC was announced. I eventually received mine (a few weeks before the QC) but the company went bankrupt around the same time. I think they’ve rebooted, so the device isn’t dead, but I have to say in comparison to the QC the open-source modules/vsts that it runs just don’t sound anywhere near as good as the QC. I had high hopes for the Mod Dwarf and loved the conceptual flexibility of it and that you could design your own ‘plug-ins,’ but the execution fell pretty short unfortunately. I’d be a little surprised if another company is willing to go that route seeing how difficult Mod’s path was.
Yes, I agree, that is one example of which we should be mindful. There are others: ZOIA by Empress Effects has somewhat of an open ecosystem where just anyone can create a patch and throw it on PatchStorage, importing it into the ZOIA, which is actually fairly limited: it only lets you load up 64 or so patches, really. However, it’s quite a success, and one can develop both their own synthesizers and their own effects, with ZOIA as the platform. There’s already a thread feature request about the ZOIA and the Quad Cortex, and this one is complementary but different because it would focus on the Quad Cortex providing a platform to develop features to the normal user behavior and not just one’s own effects. I don’t think allowing add-ons would threaten the profitability of the QC because it’s already a commercial success. It would actually stimulate community engagement and keep people invested in the QC as a platform.
yeah but i think the legal aspect of it (since i think their code is private), adaptability and whatever unexpected issues between add-ons interacting with the base of the QC presets ( and having to implement the add-ons in themselves as an update which requires designing and testing)
i really do would love to see it but i think it´s more of a fantasy than something they can achieve right now.
I was just in another topic regarding a metronome which was heavily requested since 2021, if they can´t get to that i doubt they´ll get through the trouble to get something like this started. hopefully they prove me wrong but i´ve adjusted my expectations regarding neural
Might be a great way to flesh out the Morph category on the QC;
NDSP could concentrate on the basic/overdue features, community could go nuts with the ‘anything goes’ Morph category
(that’s essentially what I’m trying to do with my presets)
I agree! And if it’s from among the group of people who use Morph, they’ll likely be more satisfied with the end result as the group understands the nuances needed
Everyone’s going to hate me for this, but whatever: Open Source on QC is a fantasy.
I highly doubt Neural would ever consider opening up their code base to allow outside programs/apps to run on the QC, no matter how many talented people outside of NDSP are capable of doing so. And this isn’t the first time people have requested Open Source either.
Go ahead and @ me; this is the only time I’m speaking on this topic.
Just a minor re-steering: The request was not to open up their existing code base. It was to make it programmable such that it can execute add-ons/plugins/extensions downloaded by their users. The add-ons would be open. There’s no request in this thread to make the Quad Cortex open source.
I don’t ‘hate’ you at all for your opinion, or find it irritating; thanks for contributing, in fact! Based on my years of experience with software development, it would only be a fantasy as long as Neural DSP chooses not to do it. A forward-thinking company that cares enough to implement this could make it a reality without any issues.
There are other companies with very monumental engineering tasks behind the scenes – take Proton as an example, who recently published GopenPGP, a cryptography library for native Go applications. They built a scalable, working encrypted calendar (which sounds simple but is hugely complex) that rivals Google Calendar, with a small staff. They also are the same company that built a programmable open API into Proton Mail.
The question isn’t if it can be done, it’s more about whether Neural DSP is willing to do it.
Not at all brutha, it’s totally okay you state your opinion, I for one cannot see any negativity in your words. Not least given your big heart for this place
