Is your “groundbreaking idea” supposed to make any sense? How about allowing you to only post in this thread, when you have something to add on topic?
Those things could be added to the Modulation Sources. It’s less about the sources, but more about the actual parameter routings. The sources themselfes are actually extendable by any type. I’ve just taken the sources, that are available right now.
Like you moving on to FL, who cares? Like that is going to speed up your wished implementation. I’m with you on getting rid of redundancy, but you come across as an emotional diva.
If this is kept, I’m gonna move to FL. That’s a simple statement. There is nothing emotional in there. It’s a simple decision because of lacking features. Anyway, nice try.
As much as I like your logic on this, Bit Arts, I’m not totally convinced that the current r3 system absolutely needs to be torn down in order to make it work with plenty of flexibility.
I am convinced that you need to go ahead and buy FL, since you obviously want it anyway, and maybe it would help you spread your frustrations around a little. 
You can keep up with both pieces of software and maybe have more fun with your time.
You know, I’m still using Renoise, because I really love the modular work with it, broken down to the tiniest elements. I’ve been waiting for the chance to automate instruments elements, since I’ve done my first XRNI. And now it’s all there, but implemented in a way I seriously can’t believe. Sound design is working with parameters in a flow, adjusting them all the time, till things fit. Serious sound design often enough takes hours for a single sound on a well-ellaborated synth, with all parameters synced. And now I’d have to do a vast amount of actually redundant syncing by hand for countless parameters.
Just sampling some VST and putting a volume envelope on it isn’t sound design to me. Then I could also use some VST and simply skip the sampling. What’s the point of calling this “sound design”?
So, when there’s no chance for Renoise purity, what’s the alternative? If I have to use VSTs anyway, it’s all about achieving the best end result only. And regarding this - the sound quality of DSPs and the bundled VSTs/FLIs -, FL is lightyears ahead of Renoise. Once I switch, I’m gonna use Renoise in best case as a sampler, to feed VSTs with the waves.
You see, there is nothing “emotional” here. It’s a pretty dry decision between Renoise purity or everything else.
+1 for a modulation matrix in 3.1
Please excuse me if I’ve overlooked a comment to this effect already, I started skimming on page 2 when the FUD hit.
My initial impression of Renoise 3b3 is consistent with that of the original poster. The main conceptual problem is that a system that is naturally modeled as a Graph is being flattened into a series of unique lists. Decoupling the modulation sources from where they are applied is a logic first step, but it may not be far enough. The A + B + C + D style modulation arithmetic also makes certain kinds of modulations impossible, like A + B +(C*A), which could be an envelope with an offset, with a LFO scaled by the same envelope. Furtermore, it’s not possible to have parallel paths within a single DSP chain (think: parallel reverb and delay without a bus). Most of Renoise is built around the concept of “many many serial chains”.
Where I’m going is, if you want to open up the modulation sources, you may need to do the same with DSP sources as well. By making Modulation Sources patchable, we approach a slippery slope.
I would actually vote for some kind of “patchable” or routable system with formalized graph connections and common utilities for dealing with aggregating many signal paths into single inputs etc (scalers, offsets, simple filters, simple math, integration, max, min, etc).
If this is off base, a more classic modulation matrix with shared sources would be fantastic. The A3000, A4000, A5000, and Emu E4 series all did this VERY well (with per-sample customization).
Yep got an a5000 to , amazing effetcs too …
I really hope this thread won’t derail .
Maybe taktik can chime in and let us of know if there is some chance of overhauling the mod.section (within reasonable time )
The left to right only (stomp box ‘modular’) system has some interesting strengths but is troubling for sure in some ways, and phasebash’s static free explanation makes for a more concise criticism. To me, anyway.
There are some important missing abilities that I have trouble imagining how to fit into the system in a clean way, keeping with the aesthetic. Mostly things having to do with modulating the modulators. This is not desirable simply for crazy modular synth stuff, either, but regular old stuff such as scalable envelopes times controlled key pitch and/or velocity. Or controlling lfo speed.
There are probably ways around some of this. Maybe by tricking out a couple modulators to do particular important functions, such as building a key and pitch track function directly into the ahdsr. That would be great. Or modulation control ‘taps’, which is more or less the stomp box solution of expression/cv inputs. These things would be great, and are not uncommon, but don’t promote a unified feel, necessarily [not that i think that’s so important].
I think that pitch key tracking exemplifies one usability problem with the current system, in that it is workable, but works in a way that seems backwards to almost anybody and makes it hard to wrap your head around. It doesn’t really bother me so much, personally, but it does feel odd. Again with the guitar thing, it is like a guitarist wanting different tunings/intonation or micro tunings or some such and his solution is having someone constantly turn the tuners on his guitar in a particular way while he plays different frets in order to achieve this. It seems a bit insane compared to just twisting a key-tracking amount knob and very few people are going to figure this out without quite a bit of fiddling. Perhaps the current system has some power in it that makes it worthwhile but so far I’ve only gotten so far as finally making it do more simple things.
It makes me wonder how many other things will be like this.
Whether ‘hard to wrap your head around’ means it’s broken or just quirky is to some extent left up to the user’s judgement, though. Maybe if a system can achieve something, by whatever method, it isn’t broken. But if someone is going to refuse to use it, for that person it might as well be.
[nice tutorials and a user manual with lots of ‘tip boxes’ will be indispensable]
if you really miss your “modular approach”, why dont you just link your modulators via macros or midi cc or is your hassle all about saving cpu cycles? talking of purity: there is no “right” way to do things from a programmer perspective, hacking is a bliss. you sound more like a frustrated musician rather then a happy programmer.
big up for developers and happy new year!
cheers
Could this be the most inane reply of 2013? Hurry now, the clock is ticking.
or could this one ???
I’m in agreement with Bit Arts thinking here. When I first started messing with 3.0 instruments it seemed like linking one envelope to both pitch and volume, for example, would be possible or the obvious way to do things. When I figured out that wasn’t possible I thought “hmmm…strange…OK?” Been dwelling on this for a while and can’t think of an advantage that the currently implemented system has over a more…er… modular approach like Bit Arts proposes. If someone could point out a clear advantage or unique aspect of the system we have now I’d be interested.
I’m not the most tech-savvy renoiser so perhaps I misunderstand this discussion.
Well, then the only thing I can really hope is, I don’t sound a frustrated as you sound clueless.
The “Keep the mod concept”-Renoise user buying a car:
Automobile Manufacturer: “Hey, Sir! Are you interested in a test drive with our nifty new car? We’ve developed a brandnew gear for it and you could help us to make it even better. Beside that we also have a nifty new Roadster in the pipe, based on this setup.”
User: “Man, it looks really great! Let’s take a test drive! … Wait… Sorry, what’s that? It only has a reverse gear!?”
Automobile Manufacturer: “Wut? Oh, crap. Uhm… well, that is our very own concept! Isn’t that great? Driving backwards all the time is soooo crazy and freaky! That’s what makes our car so unique and individual. Isn’t it really great!?”
User: “Any plans for fixing this in the future? A reverse gear is still fine, of course!”
Automobile Manufacturer: “Why fix it? It makes the car soooo unique. Well, maybe when you put your car in for the next service. Could be. Doesn’t have to be. Meanwhile just enjoy your crazy, freaky new way of driving backwards all the time!”
User: “Yeah, you’re totally right! Of course I’m gonna buy it! Who needs a car driving forward? I’ve always loved the cars you built and I’m sure this is the right step into the future! And of course I’m also totally looking forward to your upcoming roadster! I’m sure, it’s gonna rock!”
And then there of course was that nasty, frustrated Bit Arts guy, who didn’t want to buy a car driving backwards only, bitching around with the poor car manufacturer all the time, because he thought that car had to be a joke. What a stupid idiot he is, isn’t he!? Thank god there are so many customers more intelligent than him… 
C ya guys! 
Got my FL running since a few hours. :the_finger:
Developers what is gonna happen to the mod.section.
Will it be overhauled or is this the final stage .?
Maybe I’m wrong but I thought that the actual Renoise 3 modulation system ensures a good “backward playback compatibility” with older .xrns (?).
It’s clear that Bit_Arts is right. Indeed, there’s a “redundancy risk” with the way the actual modulation system globally behave. The actual modulation system looks like a hybrid thing, between the old static “one way” modulation scheme used in previous Renoise versions, and a future one, (maybe in renoise 3.1?) where modulators are elements of a “modular signal processor” (i.e. composed of independant modules that have both sources & targets).
Let’s face it : we can’t expect something like FlowStonefor the next release. Powerfull, flexible, requiring intelligence, adapted to prosumers, and far too complex for the simple average musician.
But we could make a step into a better direction with :
another limitation of the per-sample approach I’m facing with is with phrases:
I really would like to be able to set up modulations on a per-phrase basis