Posts Tagged ‘ Analysers ’

Visual feedback in plugins

When you’re starting out, it’s useful to use plugins that have numeric values and visual feedback. Big frequency graphs in EQ and transition diagrams on compressors are extremely valuable in helping you understand how these tools work. Bonus points if the tools have animated meters and graphs that dance along with the music. It’s a great way to learn how the sound is being changed. It’s a great way to learn how the parameters control how the sound changes.

But if you’re doing real work? Forget it.

Unfortunately, our eyes trump the ears. We hear what we see. Our perception of sound is so strongly influenced by our sight that sometimes even being aware of it doesn’t counteract the effect. It’s true.

It’s bad enough that your listeners don’t have a studio exactly like yours. They hear your sound differently to how you hear it. And that’s just considering the physical space. Now factor in the difference between what you’re seeing and what they’re seeing. Not only is their physical listening environment different to yours, but their visual stimulus is different to yours. Not only do they hear your sound different, but they perceive it differently.

It’s a losing battle, but we fight anyway.

We treat our studios acoustically. We purchase ridiculously expensive and over-engineered speakers and headphones. We do this even though our listeners will hear hear it differently anyway. No matter what we do. But we do it anyway – to try to hear the sound as plainly and as truthfully as possible. And sometimes it works pretty well.

But we should also strive to perceive the sound as plainly and as truthfully as possible. And that means controlling the visual stimulus in our studios.

We already attempt to create monitoring environments that are as neutral as possible. Maybe we should make our studios look as neutral as possible as well? Drab grey walls or sterile white doesn’t sound like much fun. Our studios are our workplace, and they should be comfortable and inviting. They are a place to be relaxed and focussed and creative. There should be a balance. And for the most part, it’s ok. Our studio environment is mostly static – it becomes a constant factor that our brains adjust to.

Dynamic visuals, however, are different. When your compressor is telling you that your kick drum is being compressed by 12dB, you’ll hear those 12 decibels. And you’ll be strongly influenced by how that 12dB looks. If the gain reduction scale goes from -15dB to 0dB, those 12dB will look like a lot of compression. And it’ll sound like a lot of compression too. On the other hand, if the gain reduction scale goes from -30dB to 0dB, those same 12dB will look like much less. And they’ll sound like much less too – if you’re watching the gain reduction meter.

The same goes for EQ. That 6dB cut looks (and sounds) like a lot when the frequency analyser’s graph scale is +/-9dB. Change the scale to +/- 24dB and suddenly everything changes.

But doesn’t the same apply to on-screen controls (such as knobs and sliders)? Certainly – but to a much lesser extent because the controls don’t respond to the music. Without visual feedback, you perceive the music with your ears only. There’s nothing visual that’s telling you what the music sounds like. To go back to the monitoring analogy, your perception will be plainer and more neutral.

There’s certainly a place for visual feedback. Ridiculous dancing graphics probably help car lovers enjoy their sound system. Full-screen iTunes visualisations are great for parties. Visual feedback in plugins are good for learning how they work and identifying what to listen for (it’s hard to listen for compression if your threshold is too high!)

But if you’re doing real work? Forget it.



Frequency analysers and mastering

Sometimes frequency analysers can be used in mastering. Of course your ears should be the ultimate decision maker, but an analyser can be useful as a ‘second opinion’. It can help sway you one way or another if you’re unsure about something.

There are, however, some issues to keep in mind when using frequency analysers when mastering:

  1. There are differences between analysers. Different analysers have different options and defaults for frequency tilt, time constants, resolution, etc. What one analyser shows as a straight line may look like a gradual rolloff on another. One analyser might show show short-term peaks differently to average level, whereas another might only show the peaks, and yet another might smooth everything out to show only averages. Configurations options might appear similar in some cases, but it can be difficult to know exactly how they’re implemented ‘behind the scenes’. Solution: If you’re going to use a frequency analyser (or, really, any kind of analyser) make sure you pick one and don’t use any others. Get to know how it responds to different signals. Don’t get distracted by comparing its readings to the readings from other analysers.
  2. There are differences between songs. Just because your reference song has a certain shape in a frequency analyser, it doesn’t mean your song must have the same shape. Different voicings and dynamic behaviour will cause two songs to sound different with the ‘same’ frequency balance, and sound the same with a different frequency balance. Further, non-technical aspects of music (such as structure, pace, harmony, etc) will also have an impact on how the audio sounds, which affects the frequency balance that is required in order for the song to sound balanced. Solution: Recognise that frequency analysers only measure some technical aspects of audio, and that music is much more than what can be revealed by a frequency analyser. The analyser is not a source of truth.

Ultimately, the only ‘analyser’ you should trust is your ears. Your tools can be helpful in some situations, but only if you understand their limitations.


Spectrum Analysers

Are spectrum analysers a useful tool or a distracting diversion?

It’s been written in many words and at many places that spectrum analysers are a Good Thing(tm). They let you see what you can’t hear, help you overcome deficiencies in your monitoring environment, and help pinpoint problem areas.

Having said that, I don’t use spectrum analysers, and I’ve advised others not to use them. The justification I usually provide is that it’s better to mix with your ears instead of your eyes, and even if you’re not very experienced, it’s better to train your ears as much as you can instead of learning to rely on visual aids. After all, this is music, right? Audio? Your listeners aren’t listening with their eyes…

But am I right?

Does using frequency analysers really result in poorer mixes? Do they really slow down ear training?

This is my opinion, but it’s an only an opinion. A good friend once told me “An opinion is what you have when you don’t have all the facts. When you have all the facts, you don’t need an opinion.” I’ll admit it: I don’t have the facts to support these assertions, and I don’t think I ever will. All I have is based on my own style of working, and my own beliefs and values.

I still maintain that spectrum analysers are to be avoided, but I don’t have any supporting evidence to show you. You have to try them out and make up your own mind.