How to generate a static waveform presentation of an audio file.

[Kevin Camp]

i think this will get you what you want.

first you’ll want to create a new comp, add a file that has audio and then convert the audio to key frames.

then add a new solid with the width that you want each line to be and the height that you want the maximum height to be (ex: 4 x 100), and drag it bellow the audio amplitude layer (that was created by convert audio to keyframes).

on that solid you’ll want to add these expressions.

for position:
target = thisComp.layer(“Audio Amplitude”);
offset = 1;
x = (index – target.index) * (width + offset);
value + [x,0]

and for scale:
lines = 100;
maxAmp = 18;
target = thisComp.layer(“Audio Amplitude”);
amp = target.effect(“Both Channels”)(“Slider”);
smpl = amp.numKeys / lines;
smplT = framesToTime((index – target.index) * smpl – (smplT / 2));
avg = amp.smooth(framesToTime(smpl)/2, smpl, smplT)
s = linear(avg,0,maxAmp,0,100);
[100,s]

the position expression will offset the solid (and copies of the solid) horizontally by the width of the solid + the number of pixels set by the offset value (default 1 px).

the scale expression then divides up the audio amplitude ‘both channels’ value into the number of ‘lines’ that you set (default 100). you’ll need to determine the ‘maxAmp’ value by looking at the ‘both channels’ keyframe values.

so once you have those values set, just duplicate that solid the number of times based on your ‘lines’ value and it should build your audio wave form.

to update the wave form, add another file with audio and convert the audio to keframes. by default the new audio amplitude layer will be at the top of the comp and the expression will look at the first ‘audio amplitude’ layer that it finds, so the wave form should change immediately. but since both expressions assume that the lines (the solids with the expressions) will be immediately bellow the audio amplitude layer, you’ll want to delete the old amplitude layer and probably the actual audio layer, or at least move it to the bottom of the comp.

different audio layers may have different max audio levels, so you may need to adjust that value in the expression. make that adjsutment on one layer, then select that scale property and choose edit>copy expression only. then choose all the other ‘line’ layers and paste.

Kevin Camp
Art Director
KCPQ, KZJO & KRCW

[Kevin Camp]

to clamp the audio amplitude scale expression to a max of 100%, you could change the s line to this:

s = Math.min(100,linear(avg,0,maxAmp,0,100));

that will make it better resemble the example you posted, and it may make it so you don’t need to adjust the maxAmp value as often.

Kevin Camp
Art Director
KCPQ, KZJO & KRCW

[Mac Smiechowski]

Hello Kevin. Thank you for such a quick and indepth reply! This is amazing.

There is but one problem with the code below:
lines = 100;
maxAmp = 18;
target = thisComp.layer("Audio Amplitude");
amp = target.effect("Both Channels")("Slider");
smpl = amp.numKeys / lines;
smplT = framesToTime((index - target.index) * smpl - (smplT / 2));
avg = amp.smooth(framesToTime(smpl)/2, smpl, smplT)
s = Math.min(100,linear(avg,0,maxAmp,0,100));
[100,s]

When I use it in that form, it drops an error on me:

After Effects warning: Class ‘global’ has no property or method named ‘smplT’
Epression disabled.

Error occured at line 6.
Comp: ‘Comp 1’
Layer: 2 (‘Deep red solid 1’)
Property: ‘Scale’

I’m no programmer, but is that due to the smplT variable being used at the end of line 6 (smplT / 2) before it has any value? I mean in line 6 we set a variable smplT, and part of it is being controlled by smplT itself (that has no value yet). After changing the variable name to ‘smpl’ at the end of line 6 it seems to work alright (atleast looks alright). But I’m too much of a beginner to assess if it’s ‘doing it right’ 🙂

I will test it out when I’m at work and fiddle with the expressions a bit.

Thank you so much for your help Kevin, I’m in your debt!
So glad I stumbled upon CreativeCow 🙂

I will keep you posted with the updates.

—–
Mac Smiechowski

[Kevin Camp]

yep.. that line had a typo…

smplT = framesToTime((index – target.index) * smpl – (smpl / 2));

Kevin Camp
Art Director
KCPQ, KZJO & KRCW

[Aivaras Sutkus]

Hello guys,
really good post, that’s exactly what I was looking for. But I faced one problem as I tried to generate waveform in AE using this method. When I played the song I noticed that waveform, which I generated from that song, doesn’t really match in some parts. It seemed that peaks and low parts just moved a bit from their original position. To better understand what I’m trying to say I made this picture:

https://i.imgur.com/X5mndN3.jpg
The black waveform is the one I generated with AE and the red one I made using online waveform generator and it also shows how waveform of this song really needs to be. I marked worst parts with the arrows.

I tried to generate this waveform many times with different options, but every time I get similar waveform. Any ideas what could be wrong and how to solve this ?

View post on imgur.com