gutsblow

an archived site about computer graphics and motion design

Audio Analysis in After Effects

If you have been using After Effects for a while, you might definitely know Trapcode Soundkeys which is an excellent plug-in to create extremely audio driven animations. But there are many disadvantages like you need to re-insert keyframes everytime you change your audio track/change the position in the time line. Not only that, you cannot control the frequency range that it samples. For example, if you want to analyze only high frequency sounds in detail, it is not possible with Soundkeys. Also, you cannot change the length of the spectrum if the amplitude of your audio is low. Although this can be achieved by increasing the decibel range of the source audio, it is not accurate. Also, one of the most important advantages of using this method is, audio spectrum shifts hue value with respect to the sound. Which actually means you can get an audio reactive sampling right out of the box. I will update with proper tutorial in the next couple of days, but meanwhile you should get an idea. Below is a complete 1 min video (instead of the 20 sec at AETUTS) in which I analyzed only two frequency ranges belonging to the main beat and the water sounds. But we can get a much more organic and hyper reactive form if 3-5 frequencies are analyzed.


Audio Analysis 2 from gutsbl'w on Vimeo.

So let me just outline a few Pros and cons using this method, in comparison to Sound Keys.
Pros:

  • You need not insert keyframes everytime you change the audio layer.
  • More frequency range.
  • You can control the frequency distribution(i.e. the number of spectrum bands)
  • You can control the audio delay to make it more spontaneous or for late reaction.
  • Audio reactive color sampling out of the box.
  • Frame rate independent.
  • Unlimited ranges.(Just duplicate the Null Object)

Cons:

  • It is a bit laborious if you don’t have the presets ready.
  • A bit cluttered workflow.
  • Difficult to obtain exponential drops and rises. (It is possible, but I am still working on it.)

Watch the Tutorial

Spectrum Analysis_preset_CS4.zip

Spectrum Analysis_preset_CS3.zip

Tests:
I tested extensively using many audio files using both SoundKeys and Audio Analysis method and the results are almost the same for regular music files. But for audio with very high frequency ranges or very low tones, audio analysis is much more accurate. Also, if you are working with complex animations I suggest you to convert the expression into keyframes, just for the sake of speed.

Conclusion:
I can safely say that Audio Analysis method is an effective alternative for Trapcode Soundkeys and you can get much more advanced features. I also think it can be easily used every day professionally.

Related