The Producer's Forum

Production Talk => Sound Design => Topic started by: Helius on February 27, 2018, 10:51:43 pm

Title: How do I make this sound
Post by: Helius on February 27, 2018, 10:51:43 pm
So I don't know how to describe the sound I want to make but i know it's used fairly commonly in dance music. David Guetta, Martin Garrix and Brooks' recent single "Like I Do" has it and it comes in at approx 2:02 in the song and you can probably hear it best at about 2:12ish. It sounds to me like a robotic voice sorta thing.

I've attached a youtube link for reference:

https://www.youtube.com/watch?v=WfPu9Jrcpuk

Any comments would be appreciated
Title: Re: How do I make this sound
Post by: Mussar on February 28, 2018, 02:06:05 am
There are two possibilities, just from a quick listen through:

1. Synth lead (probably a monophonic saw wave with some portamento) + Vocoder and Human Voice

2. Synth Lead (as above) + Formant Filters.

In either case, you want to impart the spectral quality of the human voice by using some filtering that imparts the shape of human vowel sounds onto a source sound - in this case, a synthesizer lead. Most DAWs have a built in vocoder, so you could experiment by recording yourself going "Oooh" and "yeah" and "Woah" and whatnot to get the formants you want, or you could use a third party formant filter like sugarbytes Wow, or the KiloHearts formant filter (or just draw your own - look up formant EQ curves and experiment!) and achieve a result that I believe is what was actually used. But vocoding might still be fun.
Title: Re: How do I make this sound
Post by: Helius on February 28, 2018, 07:28:45 am
There are two possibilities, just from a quick listen through:

1. Synth lead (probably a monophonic saw wave with some portamento) + Vocoder and Human Voice

2. Synth Lead (as above) + Formant Filters.

In either case, you want to impart the spectral quality of the human voice by using some filtering that imparts the shape of human vowel sounds onto a source sound - in this case, a synthesizer lead. Most DAWs have a built in vocoder, so you could experiment by recording yourself going "Oooh" and "yeah" and "Woah" and whatnot to get the formants you want, or you could use a third party formant filter like sugarbytes Wow, or the KiloHearts formant filter (or just draw your own - look up formant EQ curves and experiment!) and achieve a result that I believe is what was actually used. But vocoding might still be fun.

Thanks Mussar this is a great help