Please note: this article assumes a really basic knowledge of composing and mixing techniques such as compression and equalization, frequencies and gain.
This is what I’ve discovered specifically for mixing sound acceptably for delivery as audio online via website.
I’m working on an online drama. It’s due to launched in January/February if we keep to schedule (touch wood) so watch this space for more details.
As well as video clips, there’s web design, images, blogs, links and flash animation.
I’m designing the sites and animation, composing and mixing all the sound for the video, putting together flash animations and scoring those too.
The first consideration when sound mixing is how the sound will be heard. Laptop speakers are really only designed for listening to speech, and music second to that.
There’s really no bass response at all below 200 Hz… laptop speakers are really only a step up from a telephone earpiece. And they’re unbelievably quiet. Anything less than -3dB and it might as well be muted. Domestic desktop computer speakers are a tiny bit better… but get it right for the laptops and the desktops will be more than happy.
Composing with restrictions
So to score: Dialogue is most important. Don’t compose anything in the music that will interfere with that dialogue, such as percussion or mid-frequency instrumentation; instead compose around it (good advice for general composition). The same applies to sound design.
Mixing Sound for Media
Sound for Media has mostly three or four lines intertwining at the same time – Dialogue, Foley/Spot Effects, Music, Atmos. At any given time one or, at the most, two of these are at the forefront and the others are mixed down.
Sound Mixing for Online Web Audio
Mixing for online audio is far more brutal! It’s better to treat the audio track as a single thread, where all the elements of dialogue, music and other sound take turns rather than interweaving up and down in importance.
Again – dialogue is first and foremost in the mix, so mix speech to a consistently high level then manually ‘duck’ the rest of the audio around any speech. If there’s an especially emotionally important bit of the music or sound at the same time as the dialogue, analyse the frequencies of the dialogue, pick the range where they are most amplified, and duck out that range on the accompanying music or sound track. Keep an ear out for these bits on the final mix to make sure there’s still no interference with the dialogue.
Gain and Compress
On the master channel, cut the bass below 200 Hz and then compress the whole mix within an inch of its life – up the gain and get that meter up in the red 90% of the time. Don’t worry if it clips occasionally and momentarily. If you can’t hear it, it really doesn’t make a bit of difference, especially if it buys you a little more gain over the average of the rest of the track.
Then test your beautifully slick mix on as many tinny little laptops and desktop speakers as you can get your hands on and hear how much more you need to pump up the volume, and roll around the eq to compensate for uneven frequency response, before it’s even audible, never mind perfected…