Forum Replies Created
-
Posts
-
Wow – what an excellent discussion! I do appreciate your creative input, as well as your considerations from all regards – musically and programmatically.
From a conceptual view I have, bronswerk’s statement makes a great point:Quote:Forget scales, major, minor whatever. […] try to define a set of rules so there is always a musical result.I think the innovation here is to make good sounding stuff easily accessible to people like me
, who do not think in scales and chords. I do happen to think in numbers, but that’s not the point here.No matter the representation (numbers or chords), we want to know what to do about a bunch of notes we have playing, given a bunch of notes we receive (via MIDI). bronswerk puts it very crisply:
Quote:So, the situation to deal with is: sequence: 0-3-12-7-10-12-7-3 and (incoming)chord: 0-4-7. A set of rules is needed how one set infuences another set.Basically, we have so far two possible solutions (from what I understand):
a) force the material to the "0-4-7" grid ("Hold" issues to be worked out later).
b) extrapolate some scale from "0-4-7" and then force to that scale.
Do I miss out on any other important aspect?Personally, I tend to prefer a) for a few simple reasons:
– you are always in control, no constraints by pre-conceptions
– immediate cause-effect relationship – no magic happening in the box
– free way for many usability aspects (like Lars’ two way mode).Some background info: the Octopus logic defines scales as a base + an interval pattern. Hence the two things you can change about scales: base, and the interval pattern, which is really the note composition when you "Select" the notes in and out.Btw. the scale buttons Min Maj etc. are just macros for intervals.
Finally, one point I picked up from the discussion with zinoff is that "re-flagging" a note as base in the same scale, which has no effect because there is no change to the note grid, should be re-built as "selecting a new base should also carry the interval pattern over". I have it running here and it is fun!
Gabriel
Post edited by: gseher, at: 2008/04/13 02:26
LDT, about that "no keys, no sound" issue, I meant you are forced to hold down a key, so you don’t have your hands free for other things, like improvising some part on a keyboard. I still find that’s "not preferable" if that’s the only mode it should work, but I also said that a Hold feature is needed. So in essence you have both ways.
I’m not trying to be negative here and all ideas are welcome, but for me the main principle is it should work as musically and logically as possible.
In the past I had several experiences with a program called Artwonk. I’ve made stepsequencers with all kinds of bells and whistles possible. One of the things that was high on the list: How can I make a sequence pattern less rigid so that it follows my chordinput in "some way"? I do have some ideas and they come from experimenting, not just some wild thoughts.
I don’t have an Octopus, but I do believe that the Octopus is flexible enough to go a step further without eleminating excisting features. If some sort of chordrecognition is possible, then you have the best of both worlds. The power of a stepsequencer and still be "musically free" because you can change scales on the fly.
EDIT
I just erased this one as I think it was rubbish:unsure:
Post edited by: LDT, at: 2008/04/13 11:56
btw, you can get sort of an arpeggiator out of the chord strum patterns in case you haven’t tried that yet.
That’s another area of the Octopus which has great potential for further development.
Quote:I don’t like the idea of hitting whatever minor chord in the right hand and that the left hand sets the root note as earlier suggested. Also when no keys are pressed, then there’s is no sound, it’s not preferable. At least there should be a "Hold" feature. More creativity is needed here.Hi Bronswerk. What we are trying to do here I think, is to have some kind of creative flow. What I hear is a lot of "saying no". I am sure you have the best intentions, but a careful phrasing of words can make a big difference. E.g I understand that it will be nice to have a hold feature, and that it will be your preferred way of using this chord-follow-stuff, but when you state that
Quote:when no keys are pressed, then there’s is no sound, it’s not preferablethen you are excluding that for other people it wil be a super cool feature that you have "gating" of a page, built into the chord-follow mode. Think of e.g. one track of step chords playing a rhythm (including rests) sending notes to a page in chord-follow mode via a virtual midi channel.
Quote:Chord recognition is *not* the same as a standard arpeggiator.You are right (I only said they were related.
About the rules: I believe ther already exists a set of rules in the octopus. Think of how it works when you have a page with many notes and you force it to, say a triad. The more we can make use of building on top of already existing features, the easier (more logic) it will be to use (and to program, I reason).
Hey, i am no programmer, so I don’t really care what’s most logical for one, i know that you can represent a scale as a set… but what I know is that scales are already in the octopus and i want to take the most out of them.
/R
The concept of sets as above described is probably from a programmers point of view the most logical way. Forget scales, major, minor whatever. Just deal with numbers and try to define a set of rules so there is always a musical result. I don’t mind if there are some limitations, if it’s musically interesting, then it’s fine. Experimentation is the key.
I remember this very old organ from Viscount before chord recognition where you would just hit a combination of keys to get your chord. the base note would pick the key.
So if you hit:
C and E it will pick C Maj
C and Eb it would pick C min
C and D it would pick C 7
and so on.However, ideally I’d like something that I can use on the octopus without an external keyboard.
More than that I would like to take distance from the arranger concept (chords) and to think in scales and scales modes, ie. know your dorian from your aeolian, I think it’s easier to get something useful out of improvising in this way.
Anyway I’ll have to think a bit more about it…
Cheers,
/RChord recognition is *not* the same as a standard arpeggiator. With an arpeggiator only the keys that are pressed are played in some predefined order and possible streched out over several octaves, while chord recognition is nothing more than an analysis of the incoming chord and determination of what key and what scale that chord is. For eg. pressing C-E-G at the the keyboard stands for key = C and scale = major. Problems could arise when more exotic jazz chords are used and the use of inversions and bass notes, but it is do-able. A perfect example is how Karma does chord analysis, but Logic has this feature also. Using internal lists of possible key combinations is one solution. I don’t like the idea of hitting whatever minor chord in the right hand and that the left hand sets the root note as earlier suggested. Also when no keys are pressed, then there’s is no sound, it’s not preferable. At least there should be a "Hold" feature. More creativity is needed here.
The big question is: How should an incoming chord affects the programmed sequences?
Assume a sequence has the following pattern:
0-3-12-7-10-12-7-3. (numbers represents semitones relative to the key transpose note)
It should be obvious that the above sequence is in a minor scale because there’s a minor third (the number 3) but also there’s a seventh (the number 10). Now what to do if we hit whatever chord?
In fact we have two sets of numbers to work with; one set is the programmed set of numbers of the
original sequence and one set of numbers of the incoming chord. If I press the keys C-E-G on the keyboard then their relationship can be described as 0-4-7.So, the situation to deal with is: sequence: 0-3-12-7-10-12-7-3 and (incoming)chord: 0-4-7. A set of rules is needed how one set infuences another set.
The Chief is positive, let´s elaborate!:cheer:
I am sure that the "chord regognition" is when a (arranger-) keyboard responds to the chords played, by transposing its patterns to the matching scale/chord. There is several ways of how this has been implemented, but it has always been the highlight of home organs (-says the proud owner of a GEM H7000 and Farsisa Professional 110:P ).
Very closely related is the classic polysynth arpeggio (which – I think – was introduced by Roland with Jupiter 4 in the late 70´s), and the way these work would be a good starting point.Actually I believe that technicly speaking we are very close with the current Octopus mode of opereation. But where one note pressed, toggles the scale note on and the following note toggles the scale note off, what we are looking for is simply that the midi note-on toggles the scale note on, and the midi note-off toggles it of again.
But where the current mode is that there is always at least one note on (the base), we should have zero scale notes if no key is pressed: No keys no sound.I am not sure if there needs to be a base note in this mode. I havent played so much with this yet, so I am not sure what difference it makes to have, say three notes in the scale selected and then moving the base note within these notes.
However I am thinking about maybe also having a "two handed mode": With your right hand you finger you chords (preferably in a two octave zone to make voicings more intuitive) and with your right hand you transpose the whole page. So in this mode the scale notes is always in relation to the key of c. So selecting C-Eb-G with your right hand and playing an F in your left hand will force the page into F-minor.I have more going on in my head now, but I have to attend to some kids. I´ll be back.
Hi Wilson,
Interesting point.
It seems the issue is not the 14 bit resolution, but rather, if I understand you correctly, that you have 2 incoming MIDI CC streams "at the same time". Is that correct for an assumption? Is that some sort of output from the touch pad of the voyager, or how is it generated?
Octopus is able to record 14 bits resolution for the controllers that need it, btw. but it knows when it needs 14 bits and when 7 will suffice, according to the MIDI spec.
Anyway – if you are trying to record this kind of stuff, I may have to think of a way to not make the tracks lock to the auto-sensed CC coming in. That’s good in some cases and not others. Maybe allow CC auto-sense only if one track is set to record, as opposed to more than one?
In that case you could set more than one track to record, lock in their CCs, and the input would get filtered across the tracks, recording your touch pad movement accurately. This can of course be extended to up to 10 tracks in a page.
What do others think?
Let me know some more details.
Cheers,Gabriel
I must say that sounds highly interesting.
Although I am not sure what you mean by "chord recognition" precisely. What Lars is describing sounds actually quite nice, if it’s the same thing. Now what about when you don’t hold any keys pressed on the keyboard? No playing? No scale active, etc? Let’s talk some more.
I know that the effect of improvisations at that level is very fun, and if you see ways to boost it, I’m all ears.
Cheers,
GabrielPost edited by: gseher, at: 2008/04/12 05:33
Actually the Octopus does this already: You can select which notes should be in your scale by pressing notes on your keyboard in realtime. However when in this mode, each key works in the same way as the switches on the Octopus. This means that you press a key and the note is added to the scale, and you press it again and the note is removed from the scale.
This is not the way your arranger keyboard or typical arpeggiator works, but it is not far off.I must admit that I would l-o-v-e to have a "classic arp" mode where the notes of the scale is selected by only the notes you hold. This would quite fantastic, and I hope that with a little sweet-talking we can get Gabriel to make it:kiss:
