This forum is now locked and moved (back) to Plogue's forum http://plogue.com/phpBB3/viewforum.php?f=14
Whta's these codes eg06_cutoff=-2700, lfo07_amplitude=100.289 etc?
  • After converting some soundfont files to sfz I find a lot of op codes like

    eg06_cutoff=-2700
    eg06_sustain=4
    eg06_time0=2
    eg06_level0=0
    eg06_shape0=0
    eg06_time1=100.022

    I can't find them in the sfz specs. What do they mean? Can somebody explain?
  • They are sfz 2.0 opcodes for defining custom envelope generators.
  • Ok, thanks. Can I find a description of them anywhere?
  • Yes.  The SFZ 2.0 format is fully documented (*) in Simon Cann's "Cakewalk Synthesizers" book.  There's also a listing on the LinuxSampler site.

    (*) Note that, as ever "fully documented" means whatever anyone reading or writing wants it to mean ;)
  • Hmm, apparently there's most often a reference to Mr Cann and his book whenever sfz 2.0 gets mentioned. I see that neither http://www.drealm.info/sfz/plj-sfz.xhtml (your site?) or the Linux sampler site describes these codes (although the latter list them, just as you say).

    Even this very site, ariaengine.com, makes a point of not describing them but merely referring to that book. Shall I understand matters so that these opcodes are owned by Mr Cann and mustn't be described publically (and possible not be used in your own sfz works)?

    I just want to make sure I don't do any copyright infringements or spend my time on something that I later find out that I can't distribute to others.
  • The reason Simon's book is mentioned is that it's worth getting for the SFZ 2.0 section.  It's really not worth anyone's time trying to reproduce something similar as he's done a good job.  I put my SFZ 1.0 site up whilst René was working on it, because his "official" page (now at Cakewalk) really didn't help me understand the logic.  The LinuxSampler site exists so they can track their progress in implementation.  There's no "complete" reference here as I guess David's used Simon's book primarily before adding his own magic (some of which is documented - if I were less busy I'd try and pin more of it down and document it myself!).

    If you copy out ("significant", subject to local legislation) chunks of a copyright work, that's copyright infringement.  If you create something yourself, that's your own work and not infringing copyright.

    René made it clear when coming up with the SFZ format that it was open.  No one owns it.  Anyone can add or remove anything to it in their implementation.  The problem is also that, whilst there's an understood meaning to the opcodes, no one has to implement them the same way.  You can document them if you like, but it won't mean that's how they work.
  • I'm sure Simon Cann's book over the subject is good, and I might even consider buying a copy of it myself some day. René's site made it possible for me to implement a basic sfz-player and your site helped fill in a few gaps in my understanding. Thanks.

    It's good to hear that sfz is an open format. But is it free? Or rather, how free is it? Is it free like Linux, which you (basically, and to my understanding) can do whatever you want with, but If you change/extend it, you're obliged to make your extensions publically available (at no cost)?

    Is sfz 2.0 free?

    Did Simon denounce his default copyright of his contribution to the sfz format, i.e. the eg custom envelope generators? Otherwise, like you say, anyone's use of these opcodes would be a possible copyright infringement. Including Davids, yours, mine and any others using his opcodes.

    Of course, I might be able to define my own envelope generator opcodes and so might everyone else, but what standard would that be? Hardly an industry one.

    If I make a sfz sound font of my own, or a sfz player, I want to be able to share it with others. Possibly with a description of the used opcodes. And If anyone asks me what eg06_cutoff means, I don't want to answer
     
    "I sure can tell you what it means, but I won't. You'll have to buy this book..."
  • davidvdavidv
    Posts: 453

    All this really is is a question of how do you finance a thorough definition of an open spec?

    Who pays? What kind of returns of investment do you get? Simon wrote the book, that took him time and money, so its well within in his rights to keep the copyrights to his own _description_ of it, and to sell it. 

    But the spec is not HIS, only his description of it.

    Note, you need to buy an OpenGL book. You need to buy the official MIDI documents from the MMA.

    Without saying too much, lets just say that there is something planned which will make it easier for people interested in the spec.


  • Indeed, most "industry standards" are either actually proprietary (e.g. VST and ASIO) or you have to pay to get the written spec and pay to have your implementation approved and pay to display the "approved by" logo and pay if you want to suggest a change (e.g. nearly every industry standard you can think of).

    The reason I don't answer questions about opcodes is that I only use a relatively small number myself and, as I program drum kits exclusively, I often use them in ways most people aren't interested in.  (I couldn't even get René to make exclusive groups work the way they need to -- the Aria Engine is the first SFZ implementation I've had to play with that's "drum friendly" :).)
  • To davidv: Since this thing called Internet appeared, you don't need to buy any book at all. And in not doing so you'll save some trees and help keeping the CO2 and sea water level down.

    There are plenty of sites devoted to the Midi protocol and you most certainly don't need to buy any documents from MMA or from anyone else to implement a midi player. In fact, I've heard the official Midi specification is rather hard to obtain.

    And I don't recall any of these sites saying something like "Don't ask us about the midi spec. It's all in this book..."

    Looking forward to what's planned for "people intereseted in the spec" A thorough sfz2.0 spec maybe?

    By the way, the link where you decline any request for the opcodes specification of your product doesn't seem to point to any book by Simon Cann, if that now was your intention.

    To pljones:
    Regarding VST you can download the necc source code needed to implement it (at no cost) and you don't have to pay for the logo. It's quite the contrary - you're required to display their logo should you wan't to distribute any software containing VST.

    Sorry to hear you don't answer any questions about op codes. Otherwise I would have asked you about the exclusive groups implementation. But I guess that's a part of sfz 2.0...



  • Sorry, you appear not to have correctly read what I said.

    Nor do you understand the difference between an implementation and a specification, from your comments on VST and ASIO.
  • davidvdavidv
    Posts: 453

    "Since this thing called internet appeared much everyone expects anything to be free and when it isn't its usually stolen." Is closer to the truth imho.

    Whether you want to pay or not, you don't have to cut trees or raise your carbon footprint:

    http://www.barnesandnoble.com/w/cakewalk-synthesizers-simon-cann/1100355124?ean=9781435457034

    As far as the MIDI spec is concerned, it is not free. Even if everyone and their uncle have placed their own interpretations of it online, if you are a serious company that use more than NOTE ONs and NOTE OFFs it is  crucial to get the real spec.

    OTOH the exact maths behind CC10 (PAN), CC7 (volume) etc are very precise, NPRN's, system messages etc. Each time there's been a technical argument about how it should be done, either between devs, or devs and users, we as a company have referred to the official docs plenty of times to justify dealing with the MMA directly.


  • Well, I might not be a company, I might not even be serious. I might just have stumbled upon a promising free format for soundfonts I want to test in my unserious midi player. And to discuss this format with other users that might or might not have a copy of a certain book.

    You seem to think it's a terrific idea not to have a free specification available for everyone for a free format, whereas I don't. I think that a free specification for a free format will help to spread this format. And this would benefit all, including any book authors.

    So I stick to sfz 1.0 for the time being, it sounds reasonable ok for the soundfonts I've listened to sofar. Anyway, the conversion process of sforzando solved a problem I had on have to map the midi banks in the directory structure including where to put the drums. Thanks.
  • You seem to think it's a terrific idea not to have a free specification
    available for everyone for a free format, whereas I don't.


    There you go again, making up what you'd like to think other people have said rather than actually understanding what has been said. No one but you is saying this. Please do not accuse others of it.

    Consider this: what does "free" mean?  In terms of "free software" - i.e. open source software, which is subject to copyright law and is licensed according to a variety of licences agreed to conform to an agreed notion of "free" (see, already complicated) - there are no standards at all.  There is no standard implementation of the Linux kernel, for example.  There is no standard defining the Linux kernel.  There is a standard defining a Linux distribution but that isn't free in the same way.  It would be pointless if it were, after all.

    The SFZ format is free and open as in open source software.  As with implementations of Linux, there is no one definition that is correct, final and defining.  (Unlike a piece of open source software, there's no source code, of course - SFZ is a format not a program.)

    The common analogy has the other meaning of "free" being "free as in beer" - i.e. just wanting something for nothing.

  • What's your point, pljones, if you have one could you please be a bit more specific?

    There you go again, making up what you'd like to think other people have
    said rather than actually understanding what has been said. No one but
    you is saying this. Please do not accuse others of it.


    Hmm, a really strange sentence... where do I go again?! How do you know what I like to think? Clairvoyant? Please don't make up what you think I like to think.

    I made comment on davidvs all references to paid documentation and his blatantly statement of how crucial it is for serious companies to buy real midi specification instead of using the available ones on the internet
    and when it's not free there it's usually stolen. Whatever he meant by that.

    I don't say "You think...", I say "You seem to think..." which is short for "It seems to me you think..." while that is the impression his answer gives me. But I could be wrong, perhaps he's having nightmares over paid specifications, waking up sweating each and every night...

    Well OK, perhaps it was bit low of me to "accuse" him for not wanting to have a free sfz-specification available on the internet. Of course he wants that, why wouldn't he?

    Consider this: what does "free" mean?  In terms of "free software" -
    i.e. open source software, which is subject to copyright law and is
    licensed according to a variety of licences agreed to conform to an
    agreed notion of "free" (see, already complicated) - there are no
    standards at all.  There is no standard implementation of the Linux
    kernel, for example.  There is no standard defining the Linux kernel. 
    There is a standard defining a Linux distribution but that isn't free in
    the same way.  It would be pointless if it were, after all.

    The
    SFZ format is free and open as in open source software.  As with
    implementations of Linux, there is no one definition that is correct,
    final and defining.  (Unlike a piece of open source software, there's no
    source code, of course - SFZ is a format not a program.)

    The common analogy has the other meaning of "free" being "free as in beer" - i.e. just wanting something for nothing.

    And the sky is blue and the grass is green, but in the night it's dark.

    Once again -what's your point? If you have any, why not state it?

    You seem to be more interested in discussing what you think that I think someone else thinks than to have a view on the benefits of a free no cost sfz 2.0-specification available on the internet.

  • why wouldn't he?

    I can't speak for anyone else.  Personally, I can't think of any reasons why anyone wouldn't want this, in theory.  Indeed, the LinuxSampler people have made a start.  Perhaps you could offer to help them expand on their work.

    You seem to be more interested in discussing what you think that I think
    someone else thinks than to have a view on the benefits of a free no
    cost sfz 2.0-specification available on the internet.


    Nope, the benefits of having such a specification are clear.  The costs are also clear: substantially higher in terms of the effort involved to create and maintain than I can consider.
  • kinwiekinwie
    Posts: 61
    Hi Jones / David,

    I just take a look in the SF2-converted file. I see this: lfo06_pitch_oncc129=50
    Is that cc129 is the beta/unofficial opcode that defined as channel aftertouch?

    Can the beta, cc133 use for keyboard (note number) read by sforzando?
  • In http://ariaengine.com/forums/index.php?p=/discussion/4391/aria-specific-sfz-2-0-extended-ccs#Item_1 davidv said:
    > From 128 to 139 aria follows the official expression engine (Cakewalk synths)
    so CC129 should be as per 1.0 (as that assignment was adopted by 2.0), which is channel aftertouch, as you say.
  • kinwiekinwie
    Posts: 61
    Thanks a lot for your explanation, Jones

Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!