Pro Tools Fundamentals - Managing Latency In Native Systems

07-12-2014 10:00 AM

In last week’s fundamentals article I discussed latency and how it is largely inescapable in native audio systems. The vast majority of this round trip latency (from input to output) is introduced by the processing done by the computer. The low latency of DSP systems such as HDX is desirable but remains expensive. This has resulted in a variety of workarounds for this problem. Nearly all of these rely on avoiding monitoring though Pro Tools itself and thereby avoiding the latency introduced by the computer. The disadvantages of working this way are that all of these solutions are either more complex, less flexible or involve additional hardware. The advantage is that all of them are less expensive than an HD system. So what are the options?
Running A Low Buffer

With a fast enough computer it is possible to monitor through Pro Tools with a buffer as low as 32 samples. In this scenario the latency introduced can potentially be under 2 milliseconds, roughly equivalent to the time it takes sound to reach your ears from some nearfield monitors. The thing which is far from guaranteed in a system like this is the reliability of the system. On the performance/latency/determinism triangle this solution is sacrificing determinism in favour of low latency. That being said I have successfully tracked bands using just this setup with no problems.

Dual Hardware Buffers

Pro Tools 11 introduced a clever new feature to address one of the inherent compromises involved in monitoring through software in a native system. Previously if you lowered the hardware buffer to minimise latency when monitoring your input, the low buffer setting would increase the CPU load and limit the amount of plug in processing available. What PT11 introduced was separate buffers for input and output. In this scenario the input buffer can be set low to minimise latency while the output buffer remains appropriately high to minimise CPU load - Genius!
Using A Mixer

If the latency is introduced by the computer, why not organise all your monitoring outside of the computer? This is exactly how monitoring was done in the days of tape machines and in studios with an analogue mixer it is still a very effective way of working. By plugging inputs and outputs from Pro Tools into a mixer and setting up the monitor mix there it is possible to avoid input latency altogether because the input signal is being monitored at a point before it has entered the computer. The obvious restriction with this method is that it is not possible to monitor through plugins on the input track. This bothers some people but personally I think dusting off an old hardware effects unit for some confidence reverb is all I need most of the time.

Using Mix Control

In a very similar way to using a mixer many USB audio interfaces have a mix control which allows the user to listen to the output of the USB device (the computer’s output), the input into the interface (directly from input to output, not going via the computer) or a mix of the two. Using this system it is possible to listen to the live input into the interface blended with the output from the computer. If the track which is being recorded to isn’t muted then there will be a distracting echo on the output as both the latent and non-latent versions of the input will be audible.

Using Mix Software/Built In Monitor Mixer

Many larger interfaces tend to use proprietary mix software to set up more complex monitoring but the principle is exactly the same as using a mix control with the input signals to the interface being mixed with the output of the software (not including the live input signals) all within the interface. This can be done using on board DSP in the interface or sometimes using a reduced latency software monitoring path built into a bespoke, non class compliant driver. While the DSP solutions are inherently lower latency, they are also more expensive.
Low Latency Monitoring Mode

Pro Tools has always had a low latency mode available accessible from the options menu. This mode allows reduced latency monitoring with some restrictions. It is only possible to monitor inputs assigned directly to outputs 1&2 and all plug ins on monitored inputs will be bypassed.

DSP Accelerated Systems

Systems such as the UA Apollo offer a halfway house between the relative affordability of native systems and the convenience of HDX though still at a premium.
HD Native

HD Native systems while expensive are significantly less expensive than a full HDX system. As well as offering HD software, they allow the very low latency monitoring of HDX but only on one monitoring path.
Not monitoring at all

Its probably worth noting that there are lots of occasions when its is not necessary for the person being recorded to hear the themselves in any way other than acoustically in the room. Many ensembles are used to performing without headphones and unless a drop in or overdub is required then for many scenarios monitoring can be dispensed with altogether.
On the subject of drop ins, it is important to point out that the major disadvantage of all of these hardware only monitoring solutions is that if you are monitoring the input only, it becomes difficult to drop in to record part way through a previous take because you are only monitoring the live input, not the previously recorded take. Of course workarounds for this can be devised but its another illustration of the additional complexity introduced by managing latency. Life would be so much simpler with an HDX system. I’m off to check down the back of the sofa.