More importantly, perhaps you can talk about what went into your decision of when to apply a dial control, and when to use a slider. Often I see graphic designers making these decisions based almost entirely on what looks best within the layout. As I’m sure you know, the distinctive interactions of various controls are often very subtle but important for overall usability and consistency.

The only real problem I’ve ever seen with this kind of interface design is that some users take the analogy too seriously. Example: I once knew a former pilot (in her 60’s), who set out for the first time to learn how to use a desktop computer a few years ago. She had one *heck* of a time understanding that the mouse didn’t work like a pilot’s yoke (i.e., inverted). Controls aren’t automatically intuitive.. designers should always take the clueless vote into account when thinking out interfaces.

Another point to consider is that most of the time, the original real-world physical interfaces were designed that way because they were easy to manipulate; turning a knob, for example, is the natural motion of rubbing index tip with thumb. Using knoblike controls in an GUI interface only preserves visual resemblance - the physical simplicity of the real-world knob is now gone. It becomes a cute graphic toy rather than a useful technology control.

But because it has a visual resemblence and we can link its use from the interface to a real world example, then it fortifies and underlines it’s use…funny story about the former pilot, definitely made me chuckle ;-)

Quite simple, the sliders next to the preview area take up a majority of the interface and are for the primary controls (steps, spacing, opacity, rotation), the dials were used for the secondary controls (modulators) to provide visual tweaks to the primaries

I liked it, at first, as everything looked very familiar and comfortable. However, knobs workwell in the physical world because they exploit an easy movement of the fingers (twisting/rolling). Unfortunately, there is not a corresponding analogy in mouse movement (one can’t twist a mouse), so the “affordance” of a rotating knob doesn’t translate well.

I think we have not yet discovered all of the interface widgets that we will eventually come to depend on, and looking at controls from the physical world would seem a reasonable way to leverage user’s existing knowledge/familiarity.

But one should confirm that what makes a control work well in the physical world is translatable to existing computer interfaces.