Get More From Your DAW

Learning to manage CPU resources effectively will make you a better mixer.

The amount of audio processing power on tap in the latest laptop and desktop computers is astonishing these days, and this has certainly expanded the capabilities of the software we use. We’ve all seen masterclasses where our favourite engineers break down hit mixes that have 10 or 20 plug‑ins flexing away on each channel. However, not everyone has access to the same resources, and many of us can’t justify the expense of a cutting‑edge machine for working on projects at home.

It’s always useful to know how to get the most out of a modest system, and even if you’re lucky enough to have the latest ARM processor or an external DSP interface, there are still benefits to working in a way that conserves resources. What follows are approaches to using a DAW that are rooted in old‑school analogue signal‑path sensibility, and their advantages are universal.

Monitoring During Tracking

Minimising latency is usually a priority when recording audio, because you often want to monitor the input signal through the recording path. Too much latency will mean that the artist hears a noticeably delayed version of their performance, which is distracting and makes it hard to play or sing well. Unfortunately, the actions you can take to bring down latency, such as reducing audio interface buffer size or using features like Logic Pro’s low‑latency mode, usually place significant demands on your computer. This means that, without resource‑friendly management of your project, you can easily overload your CPU at this point, especially if you’re trying to monitor with effects such as compression and reverb during tracking.

Bear in mind that you are not usually making a permanent commitment when choosing these plug‑ins. They can always be changed later, so it is worth exploring your plug‑in folder to find the ones that don’t impose either a heavy CPU load or additional latency of their own. As a general rule of thumb, the plug‑ins supplied with your DAW are likely to be pretty economical in this respect. Note also that if the tracks you’re recording to are routed through group, master or monitoring channels that have their own plug‑ins, any processing delay introduced by those plug‑ins themselves will be added to the latency. It may be worth temporarily disabling them, or routing the tracks you’re recording directly to the output during tracking.

These are approaches to using a DAW that are rooted in old‑school analogue signal path sensibility, and their advantages are universal.

Bouncing Benefits

Virtual instruments and processing plug‑ins can consume significant CPU and memory resources, as well as sometimes adding to the latency of the system. The load of each one may not be significant in isolation, but as your project grows, the cumulative burden can compromise performance during playback and recording, until eventually your system becomes sluggish or refuses to play back altogether. The solution is to bounce or render some of these instruments and plug‑ins, generating new audio files which have the plug‑ins’ output printed or ‘baked in’. The instruments and plug‑ins in question can then be disabled to free up resources.

Here, three virtual instrument tracks have been rendered to stereo audio files; the original tracks are still present with all their MIDI data intact, but they are deactivated.

This will give you more power to dedicate to the task at hand, but it can benefit other areas of your workflow, too. When you’re working on a collaborative project or sending files to an engineer when it’s time to mix down, bouncing your tracks ensures compatibility across DAW systems and between plug‑in libraries. This process also commits you to the sounds and channel strip settings you’ve created, and provided you aren’t bouncing anything with plug‑ins requiring automation at a later stage, represents a firm step towards the completion of your project.

The Commit function in Pro Tools is a neat alternative to conventional freeze and bounce‑in‑place options.Rather than using your DAW’s batch export feature to output all the tracks as rendered audio files, it’s better to get into the habit of bouncing tracks as you move through each section of your project. As you work, the bounced audio tracks become way markers showing you which sections have been processed. Adding bouncing to your workflow during the arrangement phase encourages you to think proactively about the progress of your project. Facing up to decisions like which plug‑ins to leave active and which to bypass when bouncing tracks to audio can be a positive step in becoming a better engineer and a more effective DAW user.

You’ll also need to decide which channels to process in mono or stereo depending on instruments you’re working with and their orientation in the mix, bearing in mind that some might require panning at a later stage. Thinking about what flexibility you’ll need at a later stage in the mix can help you make decisions that move the project forward.

Start Making Sends

When mixing in the analogue domain, channel counts are always limited, and you can’t simply conjure up additional effects units wherever they’re needed. Getting the most from the desk and outboard you have involves imaginatively sharing resources. For this reason, reverb, delay and even modulation effects are typically addressed from auxiliary sends rather than channel insert points; this allows a single reverb or other effect to be applied to multiple sources simultaneously.

Here, a conventional multitrack drum recording has been organised according to time‑honoured principles. Two kick and two snare mics are each combined on separate mono groups, which are then routed along with all the other drum tracks to a single stereo group. The same reverb is being applied to multiple drum tracks using auxiliary sends, and this too gets returned to the main drum group for processing along with the dry tracks.The same approach is equally valid in DAW‑world, so make a point of using effects plug‑ins on a send/return basis where possible, instead of loading up your channels with inline effects. This reduces CPU load, and having the same reverb applied to many instruments can help your mix feel more coherent — it’s still possible to vary the amount of reverb on a per‑channel basis by changing the send level. Where effects busses are being used mainly for a single group of instruments — such as, for example, a reverb that is being applied to multiple drum tracks — you can route both the source tracks and the effects return to a single group channel for processing with compression and EQ. This can help make the effects feel like they are part of the sound of that drum kit.

The point where your CPU can be tested the most is often towards the end of a project, when mixing down the multitrack or finalising tracks to send elsewhere. With arrangement and tracking complete, the project’s channel count will be high, and the final mix process will require even more system resources for tasks such as automation. Summing your channel outputs in groups will make the job easier, both for you and your CPU.

Individual tracks should be bounced to audio by now, with any additional effects applied using sends. Working with groups such as drums, vocals and guitars, and subgroups within these such as backing vocals or rhythm guitars, can help you achieve a holistic and cohesive mix faster. Instead of working with individual channel levels on 80‑120 faders, you can focus on moving seven or eight group faders to achieve the right overall mix balance.

Asset Management

Each audio file bounced within a project demands storage on your drive, and file counts will often increase as you work, until you reach the final phases where the stems are the entire length of the song. Giving your files detailed names can avoid any future confusion, both for yourself and anyone you work with.

Before entering each phase (arrangement, tracking, mix) of your project, save alternate versions of your project and consolidate its assets to ensure that no unused files are taking up unnecessary space. Keeping these alternate project versions backed up on the cloud and a separate drive will avoid duplicates being found when opening the project. This way, you only have the latest, optimised version on local storage, giving you the fastest load times and maintaining the availability of your system’s resources.

Summing Up

Every DAW has its own idiosyncrasies regarding system resources and plug‑in management, and the ‘pain points’ won’t always come in the same place for every project. For example, music styles like rock and folk might use predominantly live recorded audio, whereas electronic music can be entirely comprised of soft synths and sample‑based sound sources. The greater the understanding you have of the tools you use and the job at hand, the easier it is to find hacks and loopholes that will help you out in a given situation. Implementing a CPU‑efficient workflow is an ongoing project rather than a quick fix, so be patient, and don’t be dogmatic about it — there’s no point in saving a few CPU cycles if it means doing something that hinders your creativity.