Re: Iterative user interface design in 1991
- Hi Larry,
Thanks for the feedback. I left the company in 1998, and the Qsystems
as far as I know are long gone. Screenshots would have to come from
someone's photos. The complex UI proposed was a 3D projection of what
a recording engineer would see. Sitting at a recording console, with 2
speakers in front and a 180 degree field of view. The processing
channel was to be represented by a virtual speaker whose position
(angle) and distance (relative size) were controlled by the joystick.
Today such an interface would be straight forward. We now have
libraries like DirectX and openGL, and processors that run at 100
times the speed of what we had then. The interface we actually
implemented was a top down view of the speakers and listener. The
180degree sound field was represented by a trapezoid with speakers at
the top corners, and the maximum position (-90 to +90) the bottom
corners. Channels were represent by a simple icon, a crosshair or
something, can't remember that detail. These were all vector drawn.
When we ran the system on a color monitor the system had colors,
decent looking layout, a few screen fonts, ability to draw spectral
plots, etc. However the monochrome LCD trumped the monitor in terms of
fitnesse for purpose so color and shading were not displayed.
The audio processing algorithms were invented by the 2 founders. They
started with a basic idea. Then they did a long experiment with
subjects listening to tones played with a phase and amplitude
differential. They would vary the phase and amplitude differences
until they found the optimal differential for that frequency. The
result was a complex stereo filter that would provide a sound image at
a location. We had 8 imaging filters, though we were able to simplify
to 4. These systems used FFTs and pretty detailed filters. For video
games and consumer audio much simpler filters were developed that gave
the effect but with less accuracy. Still the best sounding stereo
enhancement in my opinion.
The techmology for pro audio became a plugin for protools, and it may
still be available on their DSP farm, not sure.
On 2/14/11, Larry Constantine <lconstantine@...> wrote:
> Interesting story, but the value as a case study would be much enhanced with
> some diagrams or sketches (even if recreations) showing what was meant by
> the complicated and simplified versions of the display and the evolution of
> the design. More technical detail of the audio processing technique would
> also be interesting to at least some readers (count me in). I would guess
> the illusion of positioning beyond the speakers was achieved by throwing the
> signals out of phase, but would like to know the real dirt.
> At the end of the day, it should not be surprising that the audio engineers
> essentially relied on sound and not sight as the primary feedback. After
> all, they are designing an aural experience for listeners.
> --Larry Constantine, IDSA, ACM Fellow
> Institute Fellow | Madeira Interactive Technologies Institute |
> Professor | University of Madeira | Funchal, Portugal | www.uma.pt
> Chief Scientist, Constantine & Lockwood Ltd
Sent from my mobile device
Robin Dymond, CST
Managing Partner, Innovel, LLC.
Americas: (804) 239-4329
Europe: +32 489 674 366