[sdiy] Radfio Shack

[Roy J. Tellason]

On Thursday 21 October 2004 12:22 pm, Scott Gravenhorst wrote:

> >>With analog, you have a
> >>single design that performs a single function and interaction can usually
> >> be easily eliminated, whereas with digital, the temptation is to force
> >> multitasking to provide more functions than one from a single device -
> >> adding to the design complexity.

> >This is true. But the pay off is potentially you get a lot more bang for
> >your buck. And a much richer feature set too.

> Right, but how many of us are prepared to code a DCO, LFO, VCA and 4 pole
> filter all running in the same chip?

Why would we want to?

> These things are difficult at best when allowed to hog the chip, throw in
> multitasking and you're in a new world.  I'm not saying that more richness 
> isn't available, just that it adds to complexity which may be offputting to 
> many of us.  In an analog world, bang for the buck has traditionally been 
> based on the fact that the parts we need are cheap as dirt. So to attain the 
> richness desired, we add a few bucks and a small amount of time.  The same
> cannot be said for creating the same functionality using multitasking.

No,  but this "computers are expensive" stuff is something that's been highly 
forced by marketing people -- you can get a CPU these days for around a 
dollar and change,  and the peripheral stuff isn't that much more expensive.  
What's expensive is when you're talking whole systems,  which costs because 
of packaging and distribution and *THE OPERATING SYSTEM* and all sorts of 
other nonsense that we just don't need to deal with.

Multitasking has a bad reputation because of what was attempted to be done 
under DOS -- only it *can* work -- I still run a BBS on a dos box with 
Desqview.  And that's a kludge,  no question about it,  but we don't have to 
start out with something like DOS for a simple microcontroller.

> >>Many of the SDIY crowd were baptized in analog, it's
> >>concepts are hard to forget.  When I look at digital as a means to an
> >> end, I sigh heavily and say "perhaps not" since it requires things I
> >> either can't do or don't desire to do - such as either making or paying
> >> for a circuit board professionally etched.  If I need a VCO or a filter,
> >> I can cobble one up in a couple of hours on stripboard.  For digital, I
> >> need to do many hours of work before I can even think about the code
> >> problems I'm about to have.

> >I've breadboarded digital and it's worked fine. A PCB always helps, but
> >it's never essential.

> >I'm reading what you're saying as 'I'm used to analogue and I don't have
> >the interest in learning a new skill now.' Which I think is partly what I
> >was saying.

> I don't believe that's the case.  In fact, I didn't know a whole lot about
> analog really, until I joined this list.  My career formative years
> involved nothing but digital systems, so I actually started with a greater
> understanding of them than I did analog.  And I spent some time attempting
> to create musical devices with microprocessors, but I was never as
> satisfied with the results as I have been with what I've done with analog. 
> Add to this that I _am_ a coder, I never was much of an analog designer.  I
> have a better than average understanding of math and have no problem
> conceptualizing sound creation algorithmic details and I can code them.

Cool.  Now I know who to ask when I get stuck on some of that stuff.  :-)

> >>So then there's a digital SIDY crowd - where are the projects?  Why are
> >> there so few?

> >There are very few *hardware* projects because most of the people who are
> >interested in music DSP code VST and Direct X plugins instead of building
> >things with a soldering iron. One problem is there's a huge gap between
> > the neat and tidy world of code and the hands-on world of solder flux,
> > and there's no easy path between the two. PC and Mac architectures are
> > both closed and trying to do something like making your own USB or
> > Firewire peripherals is a *major* project.
> >
> >I think if the architectures were more open, there would very possibly be
> > a lot more interest.

> Perhaps, but I think that there should be or could be more in the way of "I
> did this all myself" and there isn't.  If creating DIY digital systems is
> more or less trivial (I quote you here:)
>
>    I've breadboarded digital and it's worked fine.
>    A PCB always helps, but it's never essential.
>
> Then why are there so few of these?  To use a PC or Mac may simplify things
> in some ways, but in others complicates (eg. either on is a general purpose
> processor that does everything, but nothing well).  USB, for example, by
> it's very name "Universal" indicates that it's more of the same PC
> mentality, can be used for anything, but does nothing particularly well. 
> I've also noticed from banter on this list that there is a stubborn concern
> over standards.  Why?  Why not just make it work?  If it does, you will
> have created a standard that might easily be replicated.  In analog, the
> only standards seem to be those of module interconnection.  I've seen more
> than one digital synth project get tanked because we can't agree on a bus
> type or some protocol.  These arguments never stopped any of us from tying
> to design yet another CCO core, ADSR or VCF.

Good point!  I don't see any need for standards beyond how many bits of 
parallel input or output you might need to accomplish some particular 
function,  stuff like that.  It *can* be done simply.

Having this major pile of chips laying around here,  I've been poking around 
some on the 'net to see what I could find.  Too darn much of what I'm seeing 
is people trying to build whole big general-purpose computer systems,  out of 
parts.  Why?  If that's the challenge that one wants,  I suppose that's okay,  
but after I got over that -- I have a LOT of machines now -- then I mostly 
decide on what I'm going to _use_,  and the homebrew DIY stuff can get more 
focused toward specific functions.  And that doesn't have to be that 
complicated.

> >>I would think that many SDIY people on this list would have developed
> >>many interesting things, and there are _some_, but certainly not the
> >> plethora that came from analog designs.  I don't believe that's because
> >> those people thought that it has to be analog or has to sound like
> >> analog.  And if it is, I suppose it's their own fault.  It may be that
> >> there's more of a sheep mentality
> >>at work than we want to admit.  There is a good reason why so many synths
> >> have those "boring" presets that copy the sounds we've heard on the
> >> radio.  It may just be that it's less common to "think outside the box"
> >> than we might wish to believe.
> >
> >That's pretty much where I am with it. (That and having the ideas but no
> >resources to develop them.) For example - hybrid approaches make things
> >like analogue control memories fantastically easy. I don't mean in the
> >patch memory sense, but in the sense of providing 8 channels of control or
> >audio voltage recording for playback or even editing later. (Like a BBD on
> >steroids. :-) But without the noise or other limitations.)
> >
> >There are whole classes of potential modules like that that no one seems
> > to be interested in.
>
> I'm not sure what you're pointing at here, but I'm certainly interested.

Me too!