Advertisement
Recording

How to Manage & Minimize Latency in Your Audio Projects

by

Whatever genre you are working in and whatever DAW you use to produce your music, it is likely you have suffered from latency issues at some point. Whether you are aware of it or not latency can be a real problem in the modern digital studio and can really effect your workflow.

To help you tackle this tricky subject I’ll break it down into easy segments, starting with what latency actually is and following onto subjects such as optimizing your system and how to tackle latency throughout your workflow.

Step 1 - What is latency?

Latency is the delay generated when a signal is routed through a digital audio workstation and connected peripherals. When a MIDI note leaves your keyboard it is sent to your computer processor and the sound from your software instrument is then generated by your CPU, which is then sent back out to your interface, converted and sent out as an analog signal to your speakers or headphones.

A similar thing occurs when you use an instrument, or microphone, to record audio into your DAW. The audio is sent into your interface, converted into digital information, processed in your DAW and then re-converted into analog and sent to your speakers. This round trip takes time, usually only a few milliseconds, but enough to create a noticeable and annoying delay.

Even if you are using a high end machine with the correct settings, a quality interface and the latest drivers … you will still experience latency. There is no way around it. Unfortunately it is a part of ‘in the box’ digital production.

Even high end machines suffer from latency.

On the up side these delay times are often extremely small, if things are set up correctly. There are also things that can be done to work around the problem and methods for minimizing the impact latency has on your session. Let’s take a look at the different methods we can use to make things manageable and techniques we can use during our project that will cut down on time spent trouble shooting.

Step 2 - Optimizing your machine

Before we get into altering any settings or discussing the technical methods for reducing latency, let’s look at the basics. In most studios, the computer is the center of operations and it is likely that a lot (if not all) of your audio is processed here. With this in mind it’s worth doing a bit of work so that things are working at their optimum level. With these alterations you will be giving your system the best chance of achieving low latency values.

Whether you are on OS X or Windows there are certain things you can do here. For a start make sure you have installed enough RAM, as this is crucial for allowing your processor to run at its full potential. This is especially important if you use a lot of multi-layered sample instruments or work with projects that have a high track count.

If you are wondering what amount of RAM is best for your machine, I would say a good guideline is to simply install as much as your machine can handle, or failing that as much as you can afford.

Although installing RAM is probably the most cost effective upgrade for your machine and usually easy to fit, some people do get one part of the process wrong. To ensure your memory is running in true dual channel mode and at its full speed, the memory slots must be loaded correctly. To make sure you have this right check your user guide and search the forums. It differs with every computer but is well worth getting right.

Installing large amounts of RAM correctly can really help to prepare your system for low latency performance.

Next up, look at your hard drive space. Ideally you should have at least 15-20% of your drive’s capacity free. If you are beyond this then it’s time to think about getting some new drives. Also, if you haven’t done it already, you should be thinking about running your audio, samples and instrument data from a secondary drive. This will result in your system drive being under less pressure. Windows users also need to make sure their drives are regularly defragmented.

Defragging Windows drives can help improve overall performance.

Installing multiple hard drives will take the pressure away from your system drive.

Again this is more important for Windows users, but keeping the amount of applications on your drive to a minimum is always a good idea. Regular virus checks and spring cleans of the registry are also recommended.

Once you have gone through the rather tedious and generic IT stuff, you can start to look at the audio specific aspects of managing and minimizing latency. Make sure you have the very latest drivers for any hardware connected to your machine and ensure that these devices are also running the latest firmware if applicable.

Also try to make sure you are using either USB2.0, Firewire or PCI based interfaces as opposed to the older USB1 models. These tend not to feature enough bandwidth for multiple streams of high resolution audio.

Step 3 - Adjusting your buffer settings

You are now ready to start looking at the nuts and bolts of your DAW and audio interface control panel. These settings are hugely important when it comes to managing your system’s latency. Every DAW approaches this area at a slightly different angle but you can usually find the settings your after under Preferences > Audio/MIDI or Devices> Audio, or something along those lines.

The key parameter you are looking for is the ‘buffer setting’. A buffer is a small amount of system RAM put aside to hold a portion of the audio stream before it is played and you hear it. Altering the amount of buffer memory your DAW uses will directly affect the latency value of your project.

Logic Pro’s buffer setting page

Low buffer settings will result in very low latency times, which is great for playing soft synths and recording over dubs etc. On the other hand high buffer settings will obviously increase the latency value, making the audible delay you hear between playing your instrument and the resulting sound from your DAW.

So why, you may ask, is there any point in ever having a high buffer setting? Why don’t we just leave it at a low value all the way through the project? This is where the catch comes in. There is a pay off for having such low buffer settings and that is a higher CPU load.

If you have a lot going on in a project and you lower the buffer settings then it’s likely you will start to hear faults in the audio and experience CPU overloads in your DAW. At this point the obvious option is to raise the buffer amount but of course this will increase latency and subsequently the lag we all hate.

Let’s look at some of the options we have here to make things a little more bearable.

Step 4 - Riding the buffer

An obvious way to reduce load on the CPU is to bounce tracks down that are using resource hungry instruments or plug-ins. This isn’t always ideal as you may want to tweak or automate parameters at a later stage in the project.

You can always attempt to keep your buffer settings low at the start of a project, bounce everything you can get way with and then slowly raise the buffer value as the project grows. This way you would aim to have the buffer at maximum at mixdown. This would get the most out of your CPU and only create high latencies later in the project when all your recording has been completed.

Step 5 - Using low latency modes and delay compensation

Some DAWs have cunning systems to get around the latency issue and keep things in time when it is present. Logic Pro and Cubase, for example, have plug-in delay compensation. Many argue that both systems have their inherent problems but when engaged they can at least bring MIDI synths into line with buffer delayed audio.

Logic also has a ‘low latency’ mode that can be engaged from the transport. This is pretty effective and actually disengages some of the plug-ins in the signal patch that are causing large delays, due to large CPU cycles. This means you can technically get away with playing a virtual instrument with reduced delay, even in a busy project.

Step 6 - Zero latency and hardware monitoring

Some sound cards feature a zero latency monitoring feature. This basically means that if you plug a synth or guitar into the interfaces inputs, it uses either a DSP-driven internal mixer or hardware throughput to output the sound to your speakers. This results in a real-time listening experience, as the audio is not having to do the return trip through the DAW.

Unfortunately this won’t work for internal software-based instruments but it is one way to bypass the latency issue if you are recording real instruments. Some interfaces even offer DSP effects on their devices so you can listen to your vocal, guitar or synth performances with some reverb or delay for extra inspiration.

Motu’s latest interfaces offer low latency monitoring and DSP effects for input sources.

Everyone has their own methods for combating latency. Feel free to share your personal experience and any solutions you may have.

Related Posts