The vOICe scans images from left to right, producing a stereo soundscape which pans in the same way. Pixel position corresponds with the frequency of the tone (pixels that are high up = high frequency tones), pixel brightness codes for loudness (white pixels = loudest) and pixel position relative to scan is presented within the stereo soundscape image.
Below are some samples as heard through the vOICe software, image scan speed is indicated in titles:
Banana (negative / image scan 2 sec)
A fairly simple image with distinct curvature which is somewhat recognisable in the resulting soundscape. As the vOICe encodes white pixels with greatest loudness, the image contrast has been reversed so as to ignore the white background.
Earth viewed from the moon (Image scan 2 sec)
Again an image which is fairly low in complexity, it is possible to hear the presence of the Earth as a high pitched squeal, appearing as it does in greater contrast relative to the surrounding darkness of space.
Einstein (Image scan 1 sec)
It takes some experience before being able to recognise a face through the vOICe, even then only vague features are usually detectable. Some vOICe users claim to be able to recognise faces in detail, even as far as discriminating between different individuals.
Goat (Image scan 2 sec)
A more complex image with the decking and the foliage in the background – but see if you can pick out the white goat in the foreground.