an open letter

Discussion in 'REAKTOR' started by herw, Jul 8, 2019.

  1. herw

    herw NI Product Owner

    Messages:
    6,260
    Dear NI,

    i like REAKTOR because it is a graphic programming language.
    I like signal flow and events which follow such a signal flow. I have much fun to program with REAKTOR.
    I know some text programming languages, but graphic programming is nice, quick and descriptive.

    I don't like ...
    • two different paradigm of primary and core
    • some features in primary are not possible in core and some in core but not in primary
    • some good ideas are not finished
    I am missing a consistent and consequent concept since R5.
    I feel that i am treading water since years.
    Suggestions which are made seem to be lost to no man's land.

    I know that NI is not willing to answer because it is not wise for a company to answer.
    But i would like to see that wishes are really heard for some small progresses.

    best wishes herw

    PS: please don't add any hate comments here, i only want that this letter will be read.
     
    • Like Like x 11
  2. Paule

    Paule NI Product Owner

    Messages:
    4,765
    Der Brief ans Christkind?
    A letter to Santa Claus?
     
  3. Roman Smirnov

    Roman Smirnov New Member

    Messages:
    21
    Yup, many thing could be done to the interface usability also...
     
  4. thesmithsonian

    thesmithsonian New Member

    Messages:
    24
    psst (he wants control and display elements reprogrammed into core, from scratch, and it looks like he wants primary removed as well, don't do it he's from the east germany)
     
    • Dislike Dislike x 3
  5. Paule

    Paule NI Product Owner

    Messages:
    4,765
    If you know were the river Emscher flows it isn't in east Germany.
    Without primary you can't use panel elements.

    I thought you've never take a look into herw's work, isn't it?
     
    • Like Like x 1
  6. Laureano Lopez

    Laureano Lopez NI Product Owner

    Messages:
    135
    i'm ok with all the interfacing to the "outside world" being done on a different level, and core cells being compiled, independent units. even all the "leftover" stuff in primary doesn't hurt anyone. the basics of primary are the problem -the audio/event divide, the event system itself where nothing is synchronic. then having to fight against the border to do anything iterative, or duplicate like crazy just to find a mistake in 20 instances. it's very apparent that many have got tired of fighting the language, so it's become limited to making discrete units for single processes with little or no repetition. it's unfortunate, because there's no good replacement for reaktor, unless you work in ableton. all the alternatives are either more technical or more academic, and very cumbersome for making music in the daw.
     
  7. herw

    herw NI Product Owner

    Messages:
    6,260
    „tief im Westen” mit einem „Pulsschlag aus Stahl”

     
    Last edited: Jul 10, 2019
    • Like Like x 1
  8. krabbencutter

    krabbencutter New Member

    Messages:
    16
    What Herw said!

    I bought R6 over a year ago, because I wanted to get deeper into DSP and figured R6 would be a good starting point. I mostly skipped primary and got into core early on.
    Reaktor got me really excited, as soon as I realized all the ways I could contribute to the community around it. But all the potential stuff I'd like to build and teach felt more and more meaningless, the deeper I got down Reaktor's rabbit hole and realized how fragmented the platform was. I would have loved to really get invested in Reaktor but at the same time it felt like it simply wouldn't be worth all the time and effort.
    Once I came to the conclusion I would be doing most of the stuff for myself anyway, I switched to Max :D

    So here's to the builders :thumbsup:
     
    • Informative Informative x 2
  9. Roman Smirnov

    Roman Smirnov New Member

    Messages:
    21
    Yup...NI should really do something to gether community more around dsp learning and science. If they really want to get Reaktor out of very special box. It is not about advertising but about communication and social movement around platform... Max has strong scintific community where a lot of very advanced topics can be studeed. Doing more for teaching people Reaktor also will improve consistency of a library, thats what i think....
     
    • Like Like x 2
  10. Roman Smirnov

    Roman Smirnov New Member

    Messages:
    21
    As a person who studied dsp himself, i went through Max, but i prefer Reaktor`s way to program and present. Also Reaktor is a perfect performing toolbox. But really science is a social trend...
    In Max patch i loose orientation more easily than in Reaktor.
     
  11. krabbencutter

    krabbencutter New Member

    Messages:
    16
    Agreed. The ZDF toolkit for example is a pretty awesome feature, but the documentation is only explaining what the macros are doing. Why is there no easy tutorial to get people into this? Like "This is how you build an easy ZDF filter from Scratch".

    What also bummed me a lot was, that there's just no way to start building simple stuff in primary and then being able to dive deeper Most people will start in primary, because it's very easy to play around and get quick results. But once they want to get deeper, they realise that they have to 're-learn' everything in core.
    It would be so much easier from a usability point of view, if all primary blocks (related to signal processing) were actually core macros.
    That way people could simply patch stuff up and once they're curious how a certain block works, they can simply double click it and have a look at its internals.

    In theory it should actually be possible for NI to recreate all primary macros in core and then create some scripts to automatically "rebuild" primary ensembles. Given appropriate resources I bet lots of people from the community would actually be happy to help in such an effort (I would be!).
    But of course that's a huge 'What if...' scenario and an effort like this would not create much financial value. ¯\_(ツ)_/¯ Additionally all old ensembles would see a 700% increase in cpu :D
     
    • Like Like x 1
  12. krabbencutter

    krabbencutter New Member

    Messages:
    16
    Thinking about my last post ... If NI would release some "Convert to Core" tool, even if it just starts out as some experimental "use at your own risk" kind of software, this could make life a lot easier for a lot of people.
     
    • Like Like x 1
  13. Roman Smirnov

    Roman Smirnov New Member

    Messages:
    21
    Is difficult to develope i think... Also more detailed primitives library, self explanatory and systematized by level of abstraction, could help also. Just more tools for every level. A lot of things i was searching through tones of instruments, or was inventing a bicycle myself :) UL contents everything, and from existing instruments it is possible to add huge amount of usefull macros and core cells.. Also... it is pity - 99% of ens does not have comments on logical structure. Bad habbit for programming language...Always try to remind myself to do notes.
     
  14. Laureano Lopez

    Laureano Lopez NI Product Owner

    Messages:
    135
    most of the stuff in primary actually is available in core, except things that are not possible (iteration) or architecturally pertinent (interfacing to the outside world). i don't think it's necessary to have all primary cells replaced -you can just not use them. maybe yours and colin's point about starting in primary is true -it makes things harder later. actually NI's own approach since the introduction of blocks has been limiting primary to interfacing.
    i've thought about that. thing is, reaktor's focus and user base was never around learning and science -that's max's area. reaktor is used by people who "just" make music. what i don't agree with is this idea that directing development to people who don't make their own stuff is sort of "democratic". i have a very different idea of "democratic". anyone can buy and use stuff, from NI or anyone else -there's nothing limiting that, there's no shortage of options. for me, it's more "democratic" to give people a good, powerful, well integrated tool so they can learn to make their own stuff without having to be engineers or researchers or composers in the academic circuit. max is very limited in that sense, unless you work in ableton. i've read people in max forums saying why would you want to work in the daw, when max can do all that by itself. that's quite a way to ditch all people who, say, make beats at their home studio but need/want things that are not out there to buy. reaktor actually fills that gap, and that's more "democratic" to me than having more things to buy. i don't suggest that it's just an excuse for a marketing decision -i think the team actually believes it, and i don't agree.
     
    • Like Like x 1
  15. Quietschboy

    Quietschboy NI Product Owner

    Messages:
    354
    Btw, what is wrong with eastern Germany? Except that it seems to be closer to the USA than western Germany in one particular point?
    // leaning far out of the (Reaktor) window :D
     
    • Like Like x 2
  16. Paule

    Paule NI Product Owner

    Messages:
    4,765
    I'm from Berlin - in the middle of eastern Germany. NI is from Berlin also (stay in Kroizberg).
     
  17. krabbencutter

    krabbencutter New Member

    Messages:
    16
    Yeah, I gotta give kudos to NI for having a very specific and well documented paradigm for Blocks, while keeping all signal processing in core.

    Let's just assume for a second this paradigm (Core = DSP, Primary = Interface & Visuals) should be adopted to Reaktor as a whole.
    In that case NI would have to ensure all legacy content could be ported over to core. This means that every single dsp-module in primary needs to have an exact substitute in Core. And by exact I don't mean just feature-wise but also mathematically!

    Just take a look at how different even a basic Sawtooth Oscillator is:
    - the primary module accepts pitch values instead of frequency
    - the primary module also has a volume control built in
    - at 48KHz sampling rate the old primary oscillator (blue) shows high amounts of aliasing, whereas the module in Core (red) is band-limited. Therefore even at the same frequency and volume both signals will differ significantly in tone and level.

    reaktor_saw_core_primary.jpg
     
  18. ehdyn

    ehdyn NI Product Owner

    Messages:
    461
    What I would like to see is a behind the scenes expose of how an ensemble like TRK-01 was created.. I've been using Reaktor a very long time and I can tell you it would be difficult if not impossible for an ordinary user to create a super-ambitious and planned out masterpiece such as this.

    It would be so great if we could get a forensic examination of the design and practices that went into creating an instrument such as this.
    This would be extremely instructive for the community and newcomers to understanding Reaktor at a very deep level and how to avoid common pitfalls that usually stop people in their tracks. It also has the benefit of having two smaller sound generators already broken out.
    With NI's help we could easily convert all the different aspects of this instrument to Blocks/Racks format and everybody would learn a tremendous amount in the process.

    More to the point that I think Herw was getting at ..there's a frustration brewing because Reaktor is an unparalleled DSP environment and it just needs a few nips and tucks to make it tremendously more powerful for instrument creators... but the resources and support need to be there from NI.

    Please do some videos explaining the creation of popular instruments such as TRK-01, Monark, Kontour, Form, Razor, anything..

    It's too much for people to figure out through trial and error.

    Reaktor is ready for the next ten years, she's the only ship to have done the Kessel Run in less than 12 parsecs.. just needs some minor additions to exponentially increase it's power and flexibility!
     
    Last edited: Jul 12, 2019
  19. Moujik

    Moujik NI Product Owner

    Messages:
    1,565
    I would actually disagree with this for exactly the reason you've stated: it would be difficult if not impossible for an ordinary user to create a super-ambitious and planned out masterpiece such as this. It's the same as why it's relatively pointless to reverese-engineer library ensembles.

    IMO NI would be better off putting the resources they might use for this into more fundamental building resources & tutorials like Max has.

    (All IMHO, obviosuly. But I've been using Reaktor a long time too ;))
     
    • Like Like x 3
  20. ehdyn

    ehdyn NI Product Owner

    Messages:
    461
    There's already a number of basic tutorials and resources that don't connect together or go anywhere.. you want it again?