Subject: Re: [linux-audio-dev] read it and drool
From: David Olofson (david_AT_gardena.net)
Date: Tue Dec 05 2000 - 06:52:25 EET
On Tuesday 05 December 2000 04:18, Chris Baugher wrote:
> On Tue, 5 Dec 2000, David Olofson wrote:
> > On Monday 04 December 2000 18:00, Chris Baugher wrote:
> > > On Mon, 4 Dec 2000, David Olofson wrote:
> > > > There are optical high resolution rotary encoders that would
> > > > be suitable for this. One of those, a counter chip and a few
> > > > gates on a bidirectional parallel port would probably be the
> > > > easiest prototype hack. (The counter wrapping at 256 is ok,
> > > > as you only deal with deltas. Just read frequently enough to
> > > > get at least two readings per full counter cycle.)
> > >
> > > This is exactly the kind of thing I've been thinking of trying.
> > > Where might one find these high resolution encoders you speak
> > > of?
> >
> > Bigger electronic component dealers have all kinds of stuff like
> > this; industrial strength joysticks, trackballs, chassis monitors
> > etc. Try
> >
> > http://www.farnell.com
>
> Ok, yeah. I was thinking you had some specific model from some
> specific place in mind that you had used before and thought might
> be good for this application.:)
Well, a few actually... But I don't have the Farnell catalogue at
home, so I can't check anything right now. Anyway, they're all pretty
similar, except for price, robustness and resolution. The physical
format is like one of those high-end/industry strength potentiometers
for panel mounting. Connect it to the electronics, put it all in a
box, and fit a nice knob on the axis.
> Anyway, one issue with encoders is whether to use the absolute or
> incremental types. The absolute encoders behave more like
> potentiometers in that they "remember" their positions and
> therefore offer some feedback as to their current position.
I don't think that's a very big deal with rotary knobs... (Which is
what I was thinking about.)
> Incremental encoders do not and that could be annoying in
> some(many?) places.
Driver issue, really. But of course, you can either use a motorized
potentiometer, or some LED or LCD indicators around the knob. Some
digital consoles have the latter as a more affordable (for the
manufacturer, at least) alternative.
> Also the rotary encoders I have seen don't have position stops like
> potentiometers do. (It wouldn't make sense for the incremental
> variety, but maybe it would for the absolute).
Well, the whole point is just that there are no stops! :-) The deal
is that you lock the current position of the encoder with the GUI
item you select, and then move incrementally from there. The GUI is
the feedback, but you turn the knob instead of fiddling with the
mouse.
I think of this controller as something you should use with the
non-mouse-hand, but that should still behave similarly to the mouse
wheel and buttons WRT GUI context. That is, just move the mouse over
the control (which might be highlighted in some way, to indicate that
the GUI is alive and in sync with you), and turn the knob. The GUI
control will follow!
Yes, only one knob, as an extension to the mouse and keyboard setup.
I just think it gets messier if you add more physical controllers
together with a GUI, at least in some situations. Going that way, I'd
rather build some kind of integrated physical and grahpical UI... I
have some idea how I'd like it to be, but I have no clue as how to
build it, if at all possible with current technology. Maybe I'll
figure out some way eventually. *thinking* Actually, it may not be
all that impossible... :-)
//David
.- M A I A -------------------------------------------------.
| Multimedia Application Integration Architecture |
| A Free/Open Source Plugin API for Professional Multimedia |
`----------------------> http://www.linuxaudiodev.com/maia -'
.- David Olofson -------------------------------------------.
| Audio Hacker - Open Source Advocate - Singer - Songwriter |
`--------------------------------------> david_AT_linuxdj.com -'
This archive was generated by hypermail 2b28 : Tue Dec 05 2000 - 07:23:59 EET