Conference napalm::commusic_v1

    Len and I were talking about this last week....
    
    Isn't this wonderfull? Now you can mix all night, save your program,
    and come in in the morning with 8-10 hours of aural fatigue all
    built into your mix!!!
    
    The mind boggles....
    
    Edd (in_jest)

    
    No need for motorized sliders. Just a VCA stage. and some
    feedback ( led's or equiv) to tell you where the slider
    is relative to the current ACTUAL fader level...
    
    Can believe someone actually built motors into them. 
    Brute force, 1st generation, noisy, slow (I bet), ecchhhhh.
    
    cute to watch tho? big deal. You sure they're MOTORIZED?
    

    Yes.  They are motorized.  When you call up a program they all go
    >>ZIP<< to the correct position.  I don't think they are slow, nor
    noisy.  Is a floppy head slow?  I assume (I haven't actually played
    with it) that it has a stepper attached, and it keeps moving the slider
    until the position is equal to the desired value.  Servo motors
    don't cost that much either.  (But I'm sure you PAY for the novelty.)
    
    Personally, I don't think leds are accurate enough.  You can't have
    128 leds beside each slider, and with some parameters, a variability
    of 16--assuming that you have 16 leds, which is still a lot, this
    is the precision you get--can result in profound changes in the
    sound.  Note, I'm not talking about using sliders for just mixers.
    I'm especially talking about using them for synth programming.
    
    I've always hated editing a patch on a programmable synth (say a
    Juno-106) where you call the sound up, and the sliders are all way out
    of whack with respect to the current patch, and you want to adjust one
    of the parameters, just a touch.  You slide the slider, and blam, you
    have to spend time figuring out where you were before you can get where
    you want to go.
    
    Can't be THAT bad...
     

    Juno 106?  By the nature of the sound of the patch you can pretty
    much guess where any given slider is located in its range of 10
    positions (sure, they can be tweeked in between each position).
    
    I definitly think motorized sliders would be a waste on something
    like a Juno 106  -- actually, I think they are a waste on just about
    anything.  There are musch better ways of achieving the same result:
    
    For example, Roland's alpha dial.  I'm not sure if it works this
    way, but a continuous slider such as a dial could assume the current
    position of any parameter when it is called, and that parameter
    could be incremented/decremnted accordingly relative to the initial
    position.
    
    The new programmer (PG1000) from Roland for the D50 is another good
    example.  A patch is loaded into the programmer, and the current
    value of any parameter can be checked without moving the slider.
    
    Anyway,  motorized sliders sound like De-evolution to me!  :^)
    
    Ron

    
    so are 128 leds...
    
    Look. 2 leds, maybe 3 PER slider (this is on the board at
    Downtown Studio in Boston) : 
    
    		o slider is hi
    		* slider matches current point
    		o slider is low
    		
	This allow automation with almost any level of resolution,
    	and provides enough FEEDBACK. I mean isnt that what we're
    	talking about? The moving sliders just provide expensive 
    	feedback. 
    
    	SIDES! you dont move these all that much in mixing!
    
    	What we REALLY want is automation of panning and EQ and Effects.
    
    	Thats bang for the buck. And still doesnt need moving parts...
        

Disagree on 3 LED solution.

My solution:  Abandon sliders and use "pots" that turn forever in one direction
with no obstruction, putting three 7-segment LEDs (known as LED group) above
each "pot" that read out from 0 to 512.  When calling up a preset mix, each 
LED group above the "pot" knob changes to the remembered value.  The "pot"
knob can then be used to tweak values in real time.

Lost is the intrinsic feedback that a slider on a mixing board provides,
which is why I think the beastie of .0 has motorized sliders.

Other solution is to use the "pots" mentioned above sans the LEDs, and hook 
up to a PC that graphically maintains a picture of a traditional mixing
board, "moving" the slider in the picture as the "pot" gets turned.

I think everybody agrees that the concentration of features offered by 
current technology is making it increasingly difficult to get information
into and out of a computer.  And I think people are sick of seeing 3 LED's
multiplexed for 100 functions.  I also think people are getting sick of
looking at computer displays.

I'm glad to see Roland addressing traditional human interface on this 
product.  I think I like motorized sliders.

My Reel-to-reel and cassette decks are servo controlled, and I havn't had
either of them fail yet (knock on silicon)...

Todd.

    I haven't played with Rolands D-50 programmer, but would be interested
    in knowing how it works.  I know that it has million sliders, and only
    one display.  You can't possibly get feedback on all those sliders at
    once. 
    
    You guys are picking apart the details without attacking the concept.
    So the Juno has simple sounds?  BFD.  I was really thinking of a
    DX7 programmer with a set of sliders representing all the parameters
    for one operator.
    
    I HATE going clikity-clikity to find the single parameter that I want
    to change, but worst still is trying to hopscotch between two or more
    parameters (ie, turn down an EG level, and turn down the rate so that the
    time is still a constant).  Fine. I'd rather massage dem sliders all
    at once. (must be a repressed freudian urge.)  But if it was a
    traditional (ie Juno) system, it would be pretty easy to blow the sound
    away, and have to hunt for a while to find it again because you have no
    way of knowing where you started with a parameter.  You can't go wrong
    if the slider assumes the correct position in the first place.  Just
    hack away. 
    
    A small number of leds is fine for fairly linear ``understandable''
    quantities, like volume and pan, but for synth programming parameters,
    I still maintain that it would be great to know where you ARE. The
    ability to visualize the sound using the sliders seems very useful to
    me. That is why computer patch editors are so worthwhile.  You could
    certainly argue that there is no point in motorized sliders when you
    can have electronic ones which are much cheaper, and more reliable.
    Granted.  (Screens are still too small to make these things ``totally
    primo'' however). 
    
    And further, with respect to mixing, the motorized part is not really
    for ``animation'' in the middle of a mix, but rather when you are
    calling up different mixes for totally different pieces. You DO move
    things ``all that much'' when you are recording different pieces.
    Also, the sliders are modal, and can represent several different
    quantities, such as the effects send/return levels, as well as fades,
    so they can assume a broad spectrum of positions. 
    
    The problem with automated fades and pans is that you can't generate an
    arbitrary function (ie exponential, logarithmic, or the patented ``Dave
    Dreher fade'', etc.)  With the DMP-7 you can record the control changes
    that you make when mixing using a sequencer.  You can automate any fade
    or pan or whatever that you like. 
    
    I'm not denigrating auto fades, etc..  Very useful are these.  Have
    them the DMP-7 (or any mixer) should.
    
    But seriously, you guys are really slashing away at motorized sliders
    for ``marketing-type'' reasons (cost and reliability).  Try assuming that
    they are perfectly cheap and reliable (I still don't know that they
    aren't), and then see what you think.   (I'm not selling them, I just
    wanted more information). 
    
    Steph
    

    The DMP-7 is made by Yamaha (not to be pedantic).  It is also rumored
    to cost around $4K.  (All right, I'll call).
    
    Steph

    The best part about sliders is that you can take a pencil or a ruler
    and instantly move a dozen of them in perfect synchrony- a very
    useful feature in a live or mixdown situation.
    	
    However, the Yamahaha DMP-7 foils this- the slider handles are recessed
    into grooves in the face, so you can't use the pencil trick.  BOOOO!!!
    HISSSSSS!  <<<tomato>>>
    	
    Now, if there were a way to "gang" the sliders electronically (i.e.
    move one and the others track it, in proper dB format) that would
    help- but it's not as instinctive or quick than a pencil fade.
    

    Without having seen the DMP, it seems logical to believe that you
    could set a "rate" and "level" for each "channel" allowing each
    slider to hit the same point at the same time, regardless of where
    they started at.
    
    Edd

    When I played with it I didn't have that much time to look thru
    the menus on the LCD.  The optiom might well be there.
    

	Re: .3 - .7 etc.
    
    As long as we are designing the "ideal" programmable mixer,
    how about this:
    use "pots" that turn forever as Todd describes in .6,
    and next to each pot locate an LCD bar graph display
    indicating graphically the current value of that pot.
    When a new "program" is called up, the LCD's assume the new 
    values, and we go on from there.
    I think that graphical displays like this, or like the 
    traditional sliders are much more intuitive...
    
    Not sure about the "pencil interface" though... have to give
    that one some more thought...
    
    Kevin

    
    In .4:
    
    >     Anyway,  motorized sliders sound like De-evolution to me!  :^)

    I couldn't possibly agree more. This is what computers are for,
    folks!
    				- HBM

    It'll be forever before I get eyestrain looking at sliders.  CRTs
    get me in about 5 hours (and I work 8 hour days.  sigh.  The things
    we do to pay for performance electronics.)

    
    Adding the electronics to control fader position will have to be
    in the audio chain, adding noise and distortion. At least the motorized
    faders have nothing to do with that chain. To me, that is it's best
    selling point.
    
    CdH
    

    But, what about that nasty ol' wiper on the pot?  Ever had a pot
    get noisy?  'Course amps can get noisy, too...
    
    Steve_anxiously_seeking_COMMUSIC_III

    new rumor according to EU Wurlitzers in Newington..Roland and Alesis
    are to announce digital mixers soon.
                                                                 
    dave

    
    Pots (faders) get noisy whether operated by hand or motor.
    
    CdH
    

    Yah, but these faders are just setting digital values, so they
    shouldn't be affecting the audio channels, right?
    
    Richard
    

    "Digital slider" sounds like some kind of baseball pitching
    technique...
    
    len.
    

    re .19	If the pots/sliders get noisy they'll send out inconsistent
    (dirty) digital values that will then go on to get converted back
    to analogue again (eventually) and have the same effect as a direct
    dirty pot.
    
    	QED
    

    Objectively, that's a load of bologna.
    
    You can (and do) debounce potentiometers just like switches.  That is,
    you assume a certain speed of operation, and noise transients are
    faster than this speed of operation.  You just low-pass filter the
    input (try doing that with an analog signal path.), digitially, or in
    the analog domain.
    
    Of course when they get so dirty that you not only get transients, but
    also erroneous steady state values, then you will get the well known
    ``wiggle control'' type failures.  Knowing that you don't have to
    worry about noise in the sliding action, you could probably design
    a slider that would fail in this way less often.  Using discreet
    contacts is one way of doing it.  That way, you wouldn't even have
    to do a D to A.  Yah, that's the ticket.  That's probably what they
    do.  Cheaper and more reliable.
    
    Steph 

Even better is an optical grating attached to the knob that interrupts
two light beams.  The microprocessor (that is bound to be lurking about
somewhere) watches the pattern of light flickerings and adjusts the
digital value accordingly.  Nothing to wear out, no electrical contacts
at all.

(You need two beams so you can distinguish up from down motions.)

    Maybe you could get a similar effect (for cheaper) with Hall-effect
    devices?

    no. a piece of film exposed from clear to dark that
    moves inbetween a light source and sensor creating
    a variable output.
    
    But I go with the debounced pots. And that can be ALL
    software by the way... no hardware low pass filter or
    anything involved....
    
    ron
    
    

    Debouncing is nothing more than low pass filtering.  By whatever
    algorithm you use, the transformation is some type of (non-linear)
    low-pass filter.
    
    All neat ideas, though.
    
    The problem with the slots idea (like what is in a mouse)
    is, what happens if it get's out of whack?  Like dust across one
    of the slots.  Absolute positioning is better.
    
    Do people REALLY use the ``piece of film'' scheme.  Really bizarre.
    
    
    Steph
    
    

Some of DEC's disks use 2 graticules with 2 lights to position heads over
tracks.
Tom

    Yep, film and phototransistors is the way to go.  We had some 16-bit
    rotary film-based grey-code encoders at GE.  They worked great,
    never failed.
    
    

    re .28
    

    
    more never than anything else.
    
    I believe quite a few big-budget mixing boards use(d?)
    this technique....
    
    nice and quiet and durable....
    

    Well, ok, not "never"
    	
    But we dropped 60 lb. used_to_be_precision linear slide-carrier units
    on them, and tho the case dented, the encoder never missed a tick.
                                                                   
    Ahhh, robotics and cartesian manipulators.
    

    The October 1 1987 issue of EDN
                       has an article about digital potentiometers based
    on the Xicor chip, a digital linear pot with 99 steps and 3 control
    inputs.
    Some traditional audio circuits can have this incorporated without
    other changes, where the rotary pots were.page 177.
    Xicor X9MME
    Tom

    Steinberg=Jones has a DMP7 running program for the Atari but no
    the Amiga (I called for a friend).
    Tom

    My friend wants to control a Yamaha DMP-11 mixer with an Amiga.
    What would you do?  Can a standard program such as Dr. T KCS be
    used to do this?
    Tom

	re: Dr. T's

	As long as the controllee speaks either poly mode, a unique channel,
	or SYSEX, Dr. T's can manage it. I regularly use multiple control
	tracks for program changes, song initialization, etc. If I had
	a midi controllable mixer I'd definitely be sequencing my mixdowns.
	As it is, though, it works fine.

    I think the DMP-7 implements extensive machine-specific (SYSEX)
    commands, especially controlling EQ & FX parameters - Dr. T's may have
    the *capability* to manage it, but I don't know just how easily this
    will be to implement. 

    Reminds me of trying to program a DX7 from Dr. T's ... might very well
    be doable, but I don't think I'd want to do it. 

-b