Whichever way you look at it dialogue is an essential part of the cinematic experience. It’s true to say that the quality, clarity and intelligibility of recorded voices varies wildly from one to the next. Common sonic symptoms include muffled speech, boxy room sounds, boomy low tones, sharp sibilance (‘S’ and ‘Shh’ sounds) and that distracting nasal sound picked up by most internal DSLR camera mics.
The good news is: Equalisation can improve recorded dialogue inside one plugin
The Human Voice
Piques the Human Ear
Understanding the human voice is key to understanding how transducers convert it’s analogous origins into correlating digital information. How microphones work, what rooms to record voices in, how to position yourself, your actors, your crew and your equipment to best capture voices in the way we hear them in perceived reality. But do we have time for all of this? – not always. Hence, quick steps can be made to sweeten and enhance Dialogue in audio post production
No. 5 – Lows and Highs
Poorly recorded Dialogue often falls into 4 categories, which are:
- Sounding muffled, boomy and dull
- Feeling boxy, roomy (echo) and smeared
- Sounds sharp, burned and lispy
- Sounding noisy and overly dynamic (too loud to too quiet)
Identify which of the above categories your camera audio falls into
- Insert a graphic EQ plugin at the very beginning of your plugin chain on your audio channel or clip.
- Set one of your frequency bands to cut @ -6dB and whilst playing back your audio sweep the frequency control from 20kHz down to 20Hz slowly.
- Here you are ‘Cutting’ and so you are scooping out a band of frequencies that your recording has too much of. Listen carefully for the audio to lighten, open up and sound clearer as you scoop out the muddy, boxy frequencies.
- As you move higher you’ll hear the dialogue get duller and more muddy. Try to focus on just the high frequencies, listen to the clarty in the voice, the amount of sibilence ‘S and Shh sounds’ and the taming of harsh burned top frequencies.
No. 4 – Cut Out More Than You Add On
The common misconception is to lean towards boosting frequencies when trying to improve DLSR camera audio dialogue. The opposite is true, in fact. Cutting problematic frequencies to then gain the output level back up is the technique here. It’s like removing lead balls from a weighted scale full of gold. The best frequencies ascend as the problematic frequencies are removed.
- During playback of your source audio set another band on your EQ to cut @ -6dB at 20Hz and slowly sweep the control up to 250Hz. Use your ears and listen for the boomy, muddy frequencies to lighten up and ease back.
- Set another band on your EQ to cut @ -6dB at 250Hz and slowly sweep the control up to 600Hz. Listen for the boxy, room boom sound to be tamed.
No. 3 – Move in Steps of 3
When molding your dialogue recording it can be beneficial to always move in steps of 3dB. Cutting and adding frequencies proportionally results in a more natural sound, preserving the natural dimensions of the original audio, in the much the same way as when transforming an image.
- Whilst playing back your source audio set another band on your EQ to cut @ -6dB at 600Hz and slowly sweep the control up to 2000 Hz (20kHz). Use your ears to hone in on the nasal, honky frequencies of voices captured on inbuilt camera mics. Once you find this frequency sufficiently tamed, stop.
- Set one of your frequency bands to cut @ -6dB and whilst playing back your audio sweep the frequency control from 2kHz to 9kHz. Listen for the sharp, burned high frequencies to tame and lispy ‘S’ and ‘Shh’ sounds from the voice reduce
No. 2 – Add the Smile
‘The Smile’ is an audio engineering term for adding a ‘Shelf’ EQ to both the very low and very high frequencies. If you want to know how to EQ dialogue, this technique can be applied almost all of the time, as well as to any other audio source from crowd scenes, location recordings to live music. ‘The Smile’ adds warmth to the low end and airy sparkle to the high end.
- Using the same EQ plugin engage the ‘High Shelf‘ and the gain control to +3dB. Frequencies between 10,000Hz (10khz) and 14,000Hz (14Khz) are airy and give audio a ‘lift’, without effecting more harsh, sharp frequencies.
- Engage the ‘Low Shelf‘ and apply a + 3db lift somewhere between 20Hz and 200Hz to ease warmth and body back into the recording whilst at the same time compensating for cutting problematic frequencies that had left audio feeling thin and vacuous.
No. 1 – Gain Up
When taking away, give back proportionally. If after sweeping to discover your problematic frequencies and applying sizeable cuts your DSLR audio dialogue gain is not used to compensate, it can suffer. Boosting the level of your Eq’d audio by as little as +1.5dB can make it ‘pop’.
- Using your EQ plugin set the ‘Output Gain’ to between +0.5dB and +2.5dB. Use your ears. You don’t need to use the channel fader for level adjustments in processing – faders are primarily for mixing a number of audio sources together, not level adjustments after processing.
- It’s by no means a rule that follows the mantra ‘louder is better’ but it’s a tool, along with the steps above, that molds audio shedding more light on the things your ears like.
Get it Right At The Source
With EQ most problematic audio issues can be addressed, however it’s limited in it’s ability to mold only the information present, if the pleasant qualities aren’t present it can’t work miracles on damaged recordings. The rule of thumb is to get it right at the source, record dialogue as best you can. Use your ears during recording and do test runs to make sure you’re hearing as sweet an intelligible voice as you can, that way next time in post you’ll be taking that dialogue audio to the next level.
I hope you find these tips on how to EQ Dialogue useful… let me know how things work out for you!
We’re doing some cool stuff with audio for video editors find out more White Rabbit Audio Presets