Re: Real-time category

[Travis Hartnett <travishartnett@gmail.com>]

I don't think Wikipedia is a great source of definitive technical
data, but in the RTOS entry therein I do see where the two examples it
cites are in range of microseconds.  The matter that I was discussing
was actually related to the period that elapses between when a MIDI
event is received by the EDP (a "Record" message for example) and when
the EDP can actually begin recording.

Again, my understanding is that the EDP was designed so that the lapse
between command and execution is constant, and I believe that constant
is 1.5ms.  A constant 1.5ms seems to be plenty accurate enough for
musicians needs in this context (looping).  If you're doing engine
management on a rocket, it may be woefully slow.  I'm not talking
about MIDI jitter in this case, but the response time between user
commands and execution.

And the RTOS vs. non-RTOS tool debate that I believed I was
participating in concerned the issue of personal computer operating
systems being a bit of obstacle to providing a constant command
latency figure.  Most people are aware that it's a good idea to turn
off your network connection, quit out of all other programs, disable
energy saving features, etc. when you're running many music
applications.  What I was referring to was the potential of
unanticipated delays in processing commands within an audio
application which are a result of OS design that are possibly beyond
the ability of a regular user to address.

Changing control panel settings and preferences wouldn't affect these
problems because they happen at lower levels of the OS than the
standard user interface can affect, and are possibly tied in with the
particular hardware design of the computer in question, hence the
importance of checking for "preferred chip sets" and such when buying
a computer for audio purposes.

Earlier someone mentioned that modern computers are so fast that this
isn't much of an issue, and I'd say that potentially it's not an issue
for a given set of tasks, but many people have the experience of
loading down HD recording software with a too-heavy load of plug-ins,
tracks and edit points no matter how fast the computer is.  The
display stops keeping pace with playback, recorded audio may get
choppy, MIDI and audio tracks lose sync, and eventually you get an
error or a crash.  With something like the EDP, the UI has been
optimised and stripped down to a set of possibilities that esentially
prevent this from happening given the overall performance of the
system.

Again, I've always been referring to looping-type tools, not soft
synths or amp simulators or what have you.

TravisH

On 12/8/05, Stefan Tiedje <Stefan-Tiedje@addcom.de> wrote:
> Travis Hartnett wrote:
> > My understanding of the main practical difference for loopers between
> > RTOS and non-RTOS is that in the first case there's a guarantee on the
> > maximum time the system will take to process an event.
>
> This is correct.
>
> > In the case of the EDP I seem to recall 1.5ms as the magic number.
>
> The technical term RTOS in the computer world is used only if it where
> in the range of microseconds, this is 1000 times faster as you claim it
> would be a RTOS. Otherwise an old Apple II with its old operating system
> would have had already a realtime operating system. (I did nice realtime
> Midi stuff on such a machine ;-)
>
> > Everything is read in 1.5ms, all the time.
> > In a non-RTOS, the overall response lag
> > may average at 1.5ms (although I seem to recall that Windows and Mac
> > OS take slightly longer than this), but there are times when another
> > task might have precedence, and so the looping app response drops down
> > to something like 20ms.
>
> You might refer here to computer keyboard or mouse clicks, I don't know,
> I never assume these as part of my instrument, they are only an aid for
> navigation, not for playing the music.
>
> If the timing jitter of an EDP is really 1,5 ms, then its much, much
> less accurate than my software looper. If I need, I could have it sample
> accurate (trigger with audio), this would be more than 50 times more
> accurate than that specific hardware looper. But I guess other loopers
> are not that bad ;-)
>
> Before using technical terms it might be a good advice to check the
> Wikipedia to be sure what you are talking about. Here is a quote from
> the RTOS entry:
>
> "On a 20MHz 68000 processor, task switch times run about 20 microseconds
> with two tasks ready. 100 MHz ARM CPUs switch in a few microseconds."
>
> Thats the time it needs to switch from an interupt (incoming Midi event
> for example) to deal with this event. 20 microseconds on a 20MHz 
>68000....
> I run a 1,5 GHz Powerbook with a PPC, I guess its probably more in the
> range of nano seconds...