10 steps to creating a punchy snare sound out of field recordings
When I started working on the 'Hits from the woods' project, I was determined to create a unique drum kit for modern music production, using only recordings made in the woods. To be able to come up with big and punchy drums and percussion hits, capturing the audio just wasn't enough. It took quite a lot of layering and gradual processing in order to make it sound full and punchy. Following is my 10 steps approach to designing a snare sound out of nothing but field recordings.
Here's a snare sample (HFW_Snare_6.wav) taken from 'Hits from the woods' sample pack. I'll talk you through it and explain the processing from start to finish.
Firstly, let's have a listen to how sounds recorded in the woods progressed to the full snare sound through the processing.
Evolution of the Snare 6:
- DryHit 1
- DryHit 1 (processed)
- DryHit 1 (processed) + Kick 1
- DryHit 1 (processed) + Kick 1 (processed)
- DryHit 1 (processed) + Kick 1 (processed) + DryHit 2
- DryHit 1 (processed) + Kick 1 (processed) + DryHit 2 (processed)
- First three layers (processed group)
- First three layers (processed group) + Crackles 1
- First three layers (processed group) + Crackles 1 (processed)
- First three layers (processed group) + Crackles 1 (processed) + Crackles 2
- First three layers (processed group) + Crackles 1 (processed) + Crackles 2 (processed)
- All five layers (processed group)
- All five layers (processed group) + parallel processing (NY compression)
- The final snare sound before mastering
- The final snare
- The final snare
- The final snare in action
1. Record percussive sounds
This one is obvious. I recorded sounds in the woods just outside Edinburgh. I captured many different percussive sounds produced using wooden sticks found in the woods as well as some other sounds like breaking branches in my hand, dragging branches on the floor, breaking tiny twigs, etc.
2. Slice the recordings to individual clips
After the recording, I listened to it all and precisely cut all the good takes, and exported them as individual files. It took a while to go through every recording I had made, but it was very useful to have individual samples as individual files ready for layering in the sound designing stage.
3. Have a vision
Serendipity in sound design is real. Many great things can happen when you freestyle and wander around by turning knobs and experimenting. However, when designing drum samples, I feel it's been useful for me to have an idea of where I want to take it. That way, I knew what kind of samples I need to layer together in order to structure the punchy snare sound. This is probably quite subjective, but the way I see it is that a snare sound has 4 essential elements to achieve the sound: A fundamental sound, bottom end, punch, and texture.
4. Pick a fundamental sound
The first layer I began with was ‘DryHit 1’ - one of the raw samples that I recorded in the woods.
The sound has its fundamental frequency peaking at 440Hz (A4), which is a musical value so I didn't need to tune it.
I wanted to keep just the initial part of the sound, so I faded it out.
I wanted it to hit a bit softer, so I backed off some of the attack and increased the release with Transient designer.
I high passed it to get rid of unwanted low frequency rumble. I put it in mono and added some clip distortion to make it richer in its low mid frequencies.
After that, I isolated the frequencies I wanted to keep by applying a high pass filter at 349Hz (F4) and low pass filter at 1397Hz (F6).
This was my first processed layer and I wanted to continue stacking it up and go towards the overall pitch of F.
5. Add mass to the bottom end
I needed to add some bottom end to the snare to make it thicker and bigger. I added 'Kick 1' which is a recording of two stumps hitting against each other.
I tuned it to the F3 note (178Hz) using Waves H-EQ, which fits the first layer nicely.
I wanted it to hit stronger and add a bit of punch to it - I brought out its attack. I love using Waves H-Comp for that. I compressed it heavily with about 12dB of gain reduction.
After that I applied some limiting - 3dB of gain reduction with L2, to control the transients.
I wanted to do another round of shaping of the attack. This time, I went with a Transient designer (+3dB on the attack knob, -1.5dB on the sustain knob).
I applied a limiter (L2) again, cutting the peak for about 5dB. Limiting at 5dB of gain reduction backed off the attack, but since this is the snare’s bottom end layer I don't need it to be the punchiest layer off all.
6. Control the transients - Comp/Lim technique
Compression can be used for more than just to control dynamics. In this case I used compression as an effect to shape the recorded sounds and make them sound punchier and more aggressive. For doing that, I applied compression with a fast to medium attack time and high gain reduction.
In my experience, this kind of compression works best at the attack time of 7ms - 20ms, depending on source's frequency information. Lower frequencies travel slower, so you might want to increase the attack time, and vice-versa for sounds in higher frequencies.
By applying strong compression like this, the beginning of a waveform (transients) gets really high in amplitude. This can cause some problems later when mixing it with music, as it tend to trigger your bus compressor and duck the overall level of a mix.
To preserve a strong attack while controlling high amplitudes, I like to use series of compressors and limiters, so I am gradually shaping the sound and keep the transients under control. I use a compressor to shape the attack of a sound as described above, while limiter after it is cutting off a bit of its peak and controlling the headroom. The trick is to cut as much as possible without taking the power of the attack away. Too much limiting will soften the attack again.
I usually lower the threshold of a limiter until I notice the attack getting softer - then, I back it up a bit to where usually the sweet spot is. Limiting with everything above 3dB of gain reduction will most probably hurt the attack of your percussive sounds.
7. Add the attack
By this stage, I added some tone and mass to the snare sound. Next, I wanted to add a punchy, mid frequency sound that gives the snare that snap, extra punch in the initial attack.
I added ‘DryHit 2’ that sounds quite punchy by itself.
It has a peak at 699Hz (F5) which fits previous two layers.
Firstly, I emphasised the F5 note range by boosting it for 9dB with a wide bell and high passed it at F3 (175Hz).
Then, I shaped the sound's attack using the same compression technique as described above. I compressed it with a fast attack and fast release, reducing gain for about 12dB.
I cut off the peak with a L2 by 2dB, added Transient designer set at +7dB on the attack knob and -5dB on the sustain knob, and finished with L2 reducing gain for 6dB.
By this point, these three layers sound okay together and somehow closer to a snare sound. There’s some bottom end and some body, but it’s still lacking something.
8. Add texture and colour
To add more body to the snare and make it colourful, I added some texture using layers of crackling sounds that I recorded with breaking multiple twigs at once.
I used ‘Crackles 1’ - the sound of breaking a bunch of dry twigs.
I was quite happy with the way it sounds, though I wanted to add some saturation and make it louder. I added some harmonic saturation with a tube overdrive emulation through UAD UA 610B.
After that, I just applied some EQ - high pass filter at 349Hz (F4) and high shelf boost for 9dB around 8kHz.
I added one more layer like this, but processed it differently - with ‘Crackles 2’ I wanted to add some width in addition to texture.
Firstly, I distorted it with an amazing distortion plugin by D16 Group, Devastor 2. I started out with a preset called ‘More than reality’ and kept it at 30% wet.
Next, I simply spread it out with Logic's Spread plugin (crackles 10 spread) in the low mid and mid area, and high passed it at (699Hz) F5.
The last two layers added some kind of a clap-ish sound to it.
By this point, I've put together 5 layers.
I achieved the snare-like sound that I'm happy with, so I stopped adding any other sounds to it. From there on on, it was all about final shaping and processing to make it sound clean, loud and punchy.
9. Parallel processing
I printed all sounds to another track in order to do some parallel processing. On the parallel track, I added 10dB of the attack with Transient designer.
I completely removed the sustain, because the purpose of the parallel track is to emphasise the attack only - to make it punchier, and then bring it in underneath the overall sound.
I applied some limiting - 5dB of gain reduction worked well in this case.
I pushed it into UAD UA 610 heavily, and did another round of the Comp/Lim technique, to shape the attack of the sound. I did that with a H-Comp and L2 combo again.
On its own it sounds a lot like snare to me - it sounds a bit thin, though. But both layers together sound better.
The final stage is processing it all together on a master bus.
Firstly, I high passed it at 175Hz (F3) where the snare is tuned to.
I emphasised the bottom end a bit, as well as the F6 area, where the punch in the mid frequencies lies.
I applied the Comp/Lim combo again. Firstly, H-Comp with 30ms attack (to leave the slower bottom end of a snare through) working at around -12dB of gain reduction, followed by L2 - reducing gain by 3dB.
Next were a Transient designer adding some attack and a bit of sustain, and L2 reducing the peak by 3dB again.
At the end I boosted some 160Hz and 2.5kHz with Maag EQ4 and ended up with L2.
This is just one of the approaches I took to sculpture the samples. I wanted to process every sample in a different way, so that I could create a diverse sample pack.
'Hits from the woods' is a unique sample pack (drums and percussion hits), carefully crafted out of recordings made in Scotland's forests. For more on the sample pack, see the video below. CLICK HERE and have your version of the sample pack sent straight into your inbox.