T
tojo
New member
can someone please explain what the concept of real time is? real time editing, or real time effects. i read it everywhere, but i haven't gotten an explanation
thanks
thanks
SouthSIDE Glen
independentrecording.net
The term is often used, misused and abused, so it kind of depends upon who's saying it
.But the classic textbook definition means that it's an effect or an edit that takes place "on the fly", in the time in which it happens; i.e. in "real time". Turn the reverb on and it's there, turn it off and it's not.
Versus non-real time, which is something that has to be renderd (generated) and saved before it can be used, because the amount of time it takes to generate it is longer than the effect itself (e.g. if it takes 1.5 seconds to generate a 1 sec effect.) When the computer just can't keep up with the live stream of audio or video and generate the effect on the fly, then that effect needs to be rendered seperately and added in later.
The faster computers get, the more sophisticated the audio or video effect can be executed in real time rather than having to be rendered off-line first.
G.
D
danny.guitar
Guest
Adding on to what Glenn said, I've noticed that good sound card/interface drivers (ASIO) are actually most important when it comes to monitoring real time effects (ie: Plugging your guitar in, turning on some amp sims or other effects and listening while you play), usually more so than actual CPU power.
CPU is also extremely important though, but even my crappy 1GHz can perform well with no audible latency using solid ASIO drivers. Until the effects get to be too much then it will drain the CPU. But good drivers should be a higher priority when trying to fix latency. CPU secondary (unless it's some POS from 15 years ago).
CPU is also extremely important though, but even my crappy 1GHz can perform well with no audible latency using solid ASIO drivers. Until the effects get to be too much then it will drain the CPU. But good drivers should be a higher priority when trying to fix latency. CPU secondary (unless it's some POS from 15 years ago).
SouthSIDE Glen
independentrecording.net
Danny, what you're describing is certainly true. However - last time I checked, anyway - latency itself does not necessarily make the effect non-real time. Granted, there is real-time, and then there is latency-free real-time, but as long as the effect does not require off-line rendering and does it's thing on the fly, it's still considered a real-time effect.
Or has the bar of that definition been changed and I didn't get the memo? Wouldn't be the first time.
You think 1GHz is crap? Hell, I used to run real-time video effects on a dual Pentium 90MHz board (with video accelerator expansion cards, but still...) with no latency. Of course the best of the effects still required rendering. Of course we didn't have crap like Vista that did nothing but suck up all the CPU themselves for no real good reason back then either. Maybe the bar *has* changed...
G.
C
cusebassman
Freakin' sweet
The term is often used, misused and abused, so it kind of depends upon who's saying it
But the classic textbook definition means that it's an effect or an edit that takes place "on the fly", in the time in which it happens; i.e. in "real time". Turn the reverb on and it's there, turn it off and it's not.
Versus non-real time, which is something that has to be renderd (generated) and saved before it can be used, because the amount of time it takes to generate it is longer than the effect itself (e.g. if it takes 1.5 seconds to generate a 1 sec effect.) When the computer just can't keep up with the live stream of audio or video and generate the effect on the fly, then that effect needs to be rendered seperately and added in later.
The faster computers get, the more sophisticated the audio or video effect can be executed in real time rather than having to be rendered off-line first.
G.
Of course, you don't want to forget that as computers get faster and more powerful, the plugs that are designed to create such effects become infinitely more complicated and still push those systems to their limits. It is rather obnoxious that they should try to strive for perfection at the cost of my ability to run a freakin multi-band compressor in realtime
SouthSIDE Glen
independentrecording.net
Very true.Of course, you don't want to forget that as computers get faster and more powerful, the plugs that are designed to create such effects become infinitely more complicated and still push those systems to their limits. It is rather obnoxious that they should try to strive for perfection at the cost of my ability to run a freakin multi-band compressor in realtime
The doube-edged sword is that the commercial side of the insudtry also sees increased computer horsepower and capacity as an excuse for software bloat. For example; my previous DAW platform (the computer I'm typing on now) is a 1999 Intel PIII/450 w/384MB RAM running Win2KPRO. Prehistoric by 2007 standards, but state of the art just 8 years ago, and I have pushed many complete audio and video projects thorugh this machine in those years quite happily and successfully.
Yet when I try loading what should be a fairly simple application into it - emulating a vintage three-oscillator analog synthesizer in the form of the Artura 2600V Arp emulator, the whole system literally crawls to a stop - not a lockup. mind you, just a slowdown to a handful of free cycles per second. All the computer can do is wiggle the synths's virtual patch cords around when I run the cursor over them. There's absolutely no reason why the softwre need be that demanding; it's just not that tough of a job.
And of course nowdays you need a couple of billion cycles per second and a meg of memory just to run the damn operating system at real-tiime. Think about that for a second; you need three times the horsepower and capaciity just to run the damn operating system than what I used to be able to use to fully produce multi-roll, multitrack, broadcast quality audio and video.
G.
