Which audio processing techniques lead to soft, warm, velvety, ASMR-like kind of sound? [closed]

Question

The goal is to create as soft music as possible.

It might be EQ, tube preamps, saturation, compression, recording on tape, tape delay, certain mic positioning, or something else. I just don't know from where to start this beautiful quest of making the softest music I can.

Might you please suggest some essential tricks that work for you or some fundamental reading?

Also, might there be any research on which frequencies are perceived as being the softest ones?

Thanks a lot!

Answer 1

I suspect that may aspects of sound that might subjectively be thought of as warm will (as so much in life) lie not in the extremes, but somewhere in the middle of the spectrum of various possible sonic characteristics. For example....

• People often associate having strong low-end frequencies with warmth, but some very strong bass might seem unnatural or uncomfortable
• a little pitch instability/vibrato is often seen as 'warm', but again, very extreme vibrato might seem mechanical and might also disrupt the harmony of the piece.

A lot will also be in the realm of the subjective. If you love old analogue synths you may be very used to a resonant filter sweep being associated with warmth, but if you are more used to orchestral sounds, that kind of sound might seem cold or artificial. And the kind of harmonies that you consider 'warm' will depend a lot on your exposure to different genres.

So I think Todd's comment is spot on - you will need to seek music that sounds the way you want your music to sound, and try to work out what its characteristics are. Even if you can pin down your formula for 'warmth', you'll still need to consider how to make those warm sounds work in a mix - perhaps your warm strings will seem muddy unless they're in a mix with some other crystal clear sound to provide a contrast.

In fact if you want to be a bit focused with your of research, I'd suggest reading some books / doing some courses on mixing, which will teach you a lot about EQ ranges, how sounds are perceived, and how to get them to work together.

Answer 2

The most obvious thing is consistent levels. You don't want jumps in volume, and you don't want plosives.

Start with the recording. You need a good pop filter, and you want someone with the right kind of voice who knows how not to pop their plosives too much.

Then a large-diaphragm microphone is your friend, ideally a dynamic mic like an SM7 (the classic radio mic). There's no reason you can't make do with any other kind of mic, but a larger diaphragm mic tends to roll off the higher frequencies in a smooth, natural way, and a dynamic mic does that even more. With a cardioid mic you also get proximity effect which lifts the bass and low-mids, which can help the "depth" of the voice. Too much makes it boomy though. If you want to stop them doing this, put the pop filter where you want their mouth to be.

Then look at EQ. Most people don't have anything going on below 100Hz vocally, so a high-pass there is good to get rid of ambient noise in the recording from aircon, footfall, and so on. If the low-mids are sounding muddy, you might want to roll out a bit (although ideally just get the talent to move away from the mic a bit). You might want to roll out a bit gently around 3-5kHz (the "baby crying" resonant frequency of our ears) to make it less harsh, although if you take out anything significant then you're probably losing clarity on the voice so be cautious. And a gentle roll-off above 10kHz will smooth out the edges and reduce recording hiss. Beware with all of this that things still need to be clear and audible though.

Then look at compression. This will even out jumps in volume. A two-stage process is good here: one stage to react quickly to transients above some level, maybe even a limiter; and a slower stage to even out the general volume. On speech, if the talent knows what they're doing then you'll likely only be looking at about 3dB on each compressor.

And then decide where you want this spatially. Do you want it ultra-dry, like they're whispering into your ear? Do you want a touch of short, well-damped reverb for a bedroom? Or do you want to add some low wind and nature sounds, with a subtle well-damped echo to suggest trees around you?

Mainly though you need to start with the voice. With a good voice, all you need to do is add some sparkle and try not to kill it. If you haven't got that, you can polish the turd as much as you like, but all you've got at the end is a sparkly turd. :)

Answer 3

I'll add a tip, with the disclaimer that I'm a performer not a recording engineer: Pay attention to envelopes. In particular, if the goal is to avoid anything too jarring—crash, bang, click, pop—then you want to avoid "plosives," and probably want to soften the attacks of impulses.

At the same time, my (limited) understanding of ASMR is that it's not just about a wash of new-agey pads that slowly fade in and out, but about impulses—gentle sounds like the tap of fingers, mouth sounds, paper sounds, etc. So you probably want to use some impulse sounds, but smooth out their envelope. Again, I'm not 100% sure, but I imagine this is more about recording very soft sounds with very close mic.

Answer 4

One thing that tends to make almost anything sound "soft" is compression, possibly with just a touch of reverb.

So, try applying a compression to the different instruments in the mix, testing different levels and settings, and you will probably get something useful.

Of course, there's a lot more to it, and various other answers contain important advice that you should study and apply as well.