Re: Low Latency - What it allows
Re: Low Latency - What it allows
Great, thank you for that comprehensive explanation.
Couple more tech Qs about this too..
- What kind of latency does 'hardware' midi have to deal with? I am sure it
varies, but is, say a MIDI controller keyboard hooked up to a sound-module
going to experience, say 7ms latency?
- Will Firewire (mLan) offer at least the possibility of monitoring through
the interface via the computer? I just ordered an iMic, and I'm sure this
wouldn't be able to have acceptable speed for this, but firewire seems like
it would have the speed.
Oh yeah, is your sig from 'Vanilla Sky'?
Thanks!
>
on 22/1/02 1:11 AM, email@hidden wrote:
>
>
> Hi all, please excuse the incredibly novice and possibly off-topic question,
>
> but right now, with the scarcity of X audio Apps, this is where the
>
> knowledge is at.
>
>
>
> What I am wondering is - if Core Audio / Midi will allow for sub 1ms latency
>
> (using even built in sound-out?) - Would that effectively mean that one
>
> could use their powerbook/ibook as an externally driven sound module via
>
> midi?
>
>
Yep.
>
>
> Say, I have a hardware sequencer,and I hooked it up to my iBook via a USB
>
> midi interface with some softsynth (like Reason which is claiming sub 1ms
>
> latency) - would I be able to send 16 midi channels of data to my ibook and
>
> have it stay perfectly in sync with any other hardware connected to the same
>
> sequencer?
>
>
Yes... Except that ALL hardware has some kind of latency/response and having
>
the low msec latencies just means that you don't have to explicitly adjust
>
for it and can use it in real time.
>
>
If you're using USB for MIDI transfer, you've already added at least a msec
>
of time to transportation of the data for instance, and MIDI itself has its
>
own problems in this area.
>
>
> Could I get rid of my Akai S2000 hardware sampler, how does the latency in
>
> OS X compare to 'real' hardware dedicated MIDI devices?
>
>
This is one of the reasons people are intensely interested in CPU based
>
solutions to this problem. I know this is about to start a screaming match,
>
but it seems that the CPU based solutions fall somewhere around the 3-4msec
>
round path latency and that is fine. (In fact an overheard conversation at a
>
booth at NAMM .... "Could you ADD some latency to your piano sounds because
>
it really doesn't feel like a piano - it responds TOO fast!")
>
>
> Also, are the audio in->out latencies low enough to make real-time audio
>
> monitoring a reality? I really am sick of these 'pass through' solutions
>
> that we need to use, as I would just like to use my computer to mix
>
> everything.
>
>
Most pro's require a zero latency monitoring - I doubt that you'll see this
>
feature of hardware dropped. Its also one of the strongest reasons why USB
>
is not a particularly good solution in this area... None of the USB Audio
>
interfaces we've seen have hardware thru capabilities.
>
>
Bill
>
>
mailto:email@hidden
>
tel: +1 408 974 4056
>
__________________________________________________________________________
>
"We'll talk about it later...."
>
"When?"
>
"In a future life when we're both cats"
>
__________________________________________________________________________