Question from an Audio Masterclass visitor: “I am still new to computer music, and some terminologies still confuse me. Can you please explain what is a TDM plug-in?And what is the Native version? How do they differ from each other? When to use each one?”
Can you believe there was a time when computers were too slow to record audio. CD-quality stereo audio demands a data rate of around 1.5 megabits per second and once upon a time computers couldn't manage that.
Now, we expect a standard computer to be capable of handling dozens of channels simultaneously, and the processor is fast enough to be able to do that.
But we need to do more than just record and play back audio. We want to apply processes like EQ and compression, and add effects such as delay and reverb.
All of these processes and effects require the processor to make calculations in real time. So if you want to EQ ten channels, then the processor has to perform ten channels' worth of calculations as the music plays.
Fortunately it isn't hard to calculate EQ, compression or delay, and these processes don't put too much of a strain on the processor.
However some processes are different. Reverb is a good example. A cheap-sounding reverb effect can be created using minimal processing resources. But a really good quality reverb requires a massive amount of calculation.
As you add more and more plug-ins to the mix, then the processor of your computer has to work harder and harder. Eventually there will come a point where it can't keep up and playback will have to stop. The only cure for this is to buy a computer with a faster processor.
So far, I have been talking about 'native' processing. This is where the processor of the computer does all the work.
But what if all this work was handed over to chips that are specially designed for audio – DSP (digital signal processing) chips? The computer's processor can relax, and the DSP chips, which are optimized for signals rather than general computer data, can work at a much faster rate.
Digidesign refers to its upmarket systems as featuring 'TDM', which stands for Time Division Multiplexing. This in itself doesn't refer directly to the method of processing via specialized DSP chips, but it has become a synonym for that.
There is, by the way, nothing special about time division multiplexing. It is a very common method and by no means unique to Digidesign.
A system that uses DSP will be capable of more plug-ins and software instruments than a native system. There isn't necessarily a quality difference because quality is down to the algorithms used. However the additional power of a DSP system will allow algorithms to be more complex, and thus potentially capable of better quality.
Ultimately it is likely that computers will become so fast that DSP will be irrelevant. But for the moment DSP offers a worthwhile advantage.