-Does the product of convolution = real reverberation?
-What are the differences? strengths? weaknesses?
-How does sound percussiveness/fluidity affect convolved sound quality? frequency range?
This project was an experiment in which I used multiple sounds and spaces with contrasting qualities, and convolved them in each possible permutation, in order to discover how their variances affected the quality and realism of sound produced. In order to have a "test" case for each example, I recorded "live" versions of each sound sample being emitted in the actual space. This allowed me to discover by comparison how the real reverberation and simulated reverberations differed. The intention of this project was to reveal both the strengths and limits of the convolution tool as applied in digital editing of various types of sound.
Sound Samples:
Ideally, I would have had a wide variety of sound sources at my disposal, that would have been portable enough to take to both recording locations, and that would have emulated the sounds exactly. But for both time and logistical reasons, I decided that using my voice in different ways would be good enough. The result is a bunch of vocal samples, most of which sound ridiculous, but which hopefully give a broad enough range of sound qualities nonetheless.
Locations:
In the attempt to try and expose the workings of the convolution tool as percievably as possible, I decided to pick two recording locations with much reverberation. The main variable between the two was resonating space, a factor that affected the echo time of the sounds. In the end, I used a tiled bath/shower room, and a concrete outdoor breezway in between two buildings in Stern Hall. Here are the impulses (claps) that I recorded at these two locations:
bathroom (wav) outdoor walkway (wav)
Dry Sounds:
First I recorded all the sound samples in as near to an anechoic chamber as I could find-- it was small room with carpeted walls and a pool table, all of which absorbed the sound nicely. The results were pretty clear samples, with almost no ambient noise or feedback from the room:
whistle (wav)
national anthem (wav)
"16 tons" (wav)
vocal beatbox (wav)
"hello" (wav)
Once I had all of these samples recorded in my two "echoic" locations, along with impulses at each spot, I began the labor of convolving each possible combination...