View Single Post
Old 04-27-2012, 03:55 PM   #46
Banned
Human being with feelings
 
Banned's Avatar
 
Join Date: Mar 2008
Location: Unwired (probably in the proximity of Amsterdam)
Posts: 4,868
Default

Quote:
Originally Posted by TonE View Post
Just have a closer look to the Keykit programming language, then tell me what you like more, pd or Keykit? I am sure pd is also great, I looked at it a little long time ago, but as I had already some experience with Keykit, pd had no chance against its coolness, but only from my point of view. Others with different experiences might see it differently. Nowadays people want to use and buy crap tools, like iBad or similar, ignoring most powerful tools like Keykit, but is it my problem? No.

If I should have anything usable one day for OSC <--> midi conversion using Keykit, of course I can share it also here. Just the lack of send level control in Reaper would be the reason to implement something like this. I would not do it, if I would not have to do it. I suppose Reaper will stay for a longer time without any appropriate send level control, via external hardware controllers. Actually in my case for this particular scenario I would not even do any OSC --> midi conversion then, doing everything only via OSC. Keykit <--> OSC <--> Reaper. No midi at all. Midi is only required if the data has to leave the computer, to other devices, e.g. hardware midi controllers.
I guess I misunderstood your (desired) use case - mainly out of curiosity, what do you want to use MIDI for then, if not external hardware controllers? Apparently it is not about controlling the sends from another app using MIDI, because you assume (wrongly, imho) that MIDI is only needed for external hardware?

So I'm confused there... I may misunderstand completely what you're intending to do, but it seems that you have things the wrong way around: if you don't need MIDI, but have some way of using OSC from wherever it is you want to control REAPER (KeyKit? KeyKit <--[OSC]--> REAPER should already work), you don't need any conversion tools *now*. You only need REAPER to support MIDI controllable sends natively if you don't want to rely on any OSC<-->MIDI conversion now. Regardless which tool you'd use for that (KeyKit, Pd, whatever).

And I really thought this thread was about workarounds for the time being, so I don't see the need to repeat the desire for native MIDI control. All of us are on your side there already.

But when you say that REAPER has a "lack of send level control" or that it remains "without any appropriate send level control," I respectfully disagree. That is simply not true. I'm happily using a very appropriate track send level control in REAPER as we speak, thank you very much Cockos.

As far as your KeyKit vs Pd question goes, first of all, I explicitly mentioned that my suggested approach to solve the issue that this thread is about is not about Pd, and that you should just use any tool that suits you. And prior experience is a major factor for me to, so I completely understand you taking that perspective, and I have no need or desire to convince you otherwise.

But, fair enough, I drew a comparison with regard to its looks, so I'll be glad to play a little head to head comparison game for a bit. It may be quite informative for some people who haven't found a tool they're comfortable with yet, and want to compare the many possible alternatives. But also keep in mind that I never implied Pd was the best or my favorite tool. To reiterate: I'm mainly mentioning it as a prime example because it is free and cross-platform. Partly because I know I can expect many bitter or hateful comments on this forum anyway (much like your "iBad" reference, which has absolutely nothing to do with the subject of this thread afaics).

So let's start there: KeyKit is also free and cross-platform (I haven't downloaded it and tried it but since Mac is mentioned I'll assume it works as well as the Windows version). To an objective newcomer, which I think is a fair standard, that arguably makes an opening score of 2-2. It doesn't look prettier, at first sight, so that's probably a tie, no points on either side for easy slick looks I guess (I think we agree that looks are secondary to functionality, or even unimportant, right? Which may not be true for everybody though, so I would have given points if either of these two apps deserved them). The most important difference, I guess, is that KeyKit is more of a programming language and less of a programming environment than Pd, i.e. low-level vs. high-level programming approach. Both approaches have their merits, and both are probably pretty powerful, so even while I think the high-level approach is probably easier to grasp for the newcomer, it very much depends on preference and prior experience, and low-level programming typically results in more efficient CPU usage,so for power, I'll give two points to both (4-4). However, Pd can be freely expanded using low-level programming languages, while KeyKit seems a proprietary project (as a lawyer who has worked in telecoms extensively, I do not take signing a license with AT&T's name on it lightly, especially to speak a language, to learn it, or to even have a look at it, sorry). I'll ignore my personal legal objections here (heck, I even gave KeyKit a point already for being available "for free", while it is free only as in free beer), but on expandability Pd must surely win a point over KeyKit (5-4). After a quick scan, it seems to me that the development/support team is tiny (a single guy?), there's not a very large or active user base, but it definitely does look very mature (developed since 1996), so I'll have to assume it's well tested and stable (although I just read somewhere that OSC over UDP is broken on Linux and has never been tested... seriously?), and portability seems good: lots of ports, but mostly to older OSs, and not to modern mobile / DIY platforms like iOS, Android, Arduino. Let's give 2 points to Pd and 1 to KeyKit, making it 7-5. Since I can't think of any other (important) criteria at the moment, I'll end the contest here. End score: a modest but fair win for Pd after against a very decent competitor. By no means kicking its butt, but a fair win nonetheless.

At least, I tried to be fair... I invite you to give me your account of the match. Especially after you've actually tried this approach a bit more in practice, I'm sincerely interested to hear about your results. And the ones of other users, too! Feel free to change the contestants as well.

Btw, for those type of people that prefer programmer-oriented low-level language workflows, you may also be interested in things like ChucK and SuperCollider. And since KeyKit apparently uses SimpleOSC for OSC stuff, perhaps combining that with ReaScript is also an interesting (Python-based) approach.
__________________
˙lɐd 'ʎɐʍ ƃuoɹʍ ǝɥʇ ǝɔıʌǝp ʇɐɥʇ ƃuıploɥ ǝɹ,noʎ
Banned is offline   Reply With Quote