Take the 2-minute tour ×
User Experience Stack Exchange is a question and answer site for user experience researchers and experts. It's 100% free, no registration required.

I love physical dials and knobs, they are great for adjusting something just the right amount without looking at it. They are even great for controlling an absolute value like volume because they stop (hopefully) when you can no longer adjust it. However software knobs are a pain because:

  • You are not holding on to a physical knob so it can't tell how far you've turn it
  • there is no physical feedback letting you know when you've gone too far

I don't like to tell someone to never to use a type of control but I can't find a decent use for a software knob controlling an absolute value besides the rare use case of a software knob matching a connected hardware knob. Is there any other use case?

share|improve this question
add comment

8 Answers

up vote 9 down vote accepted

I guess they provide an intuitive interface when replacing something which is used by expert users, and is traditionally using dials.

Examples would be audio boards, cockpits, other professional-only electrical devices.

Meego Harmattan (the OS of Nokia N9) used it to set the time:

enter image description here

Although it's great and provides a natural mapping (as it is recommended by Norman's Design of Everyday Things), you actually hide the numbers with your finger while dialing, so it's absolutely necessary to have actually two displays for the same number in order to be effective.

The good thing with a circular motion on a touchscren is though, that you don't have to leave the screen in order to have an arbitrry length motion. Perhaps such a gesture could be used to seek forward and rewind audio and video materials...

Sorry, I have to run, have to quickly build a prototype music player for the iPhone which actually has an iPod-like wheel....

;)

share|improve this answer
    
+1 because, reading the question, I was about to scream "No!!!" until I saw your example, which is actually quite elegant. Conversely, try using the knobs on the Korg iMS-80 without feeling like you're trying to thread a needle and failing (youtube.com/watch?v=jCN-cnxSFhk) –  msanford Sep 11 '12 at 1:49
    
+1 great example and interesting because time can expressed in both relative and absolute values. –  Steve Moser Sep 11 '12 at 13:19
    
Actually I still want to scream no :P It's a much less compact version of the iOS time picker with the same "off by one input, off by 59 output" problem. I don't see this design as solving anything really. A better example is Convert Bot where the dial is cute but doesn't get in the way. –  Ben Brocka Sep 12 '12 at 13:28
add comment

Virtual dials are missing the "graspable" affordance that makes them effective.

So in the virtual realm the area users interact with in the physical realm (the sides) is missing and replaced with an area that they don't use (the top) and whose surface area is small in comparison. This breaks the metaphor completely.

One area they are ubiquitous is music software, especially soft-synths and mixing desks that model their physical counterparts. However physical control surfaces with actual dials/knobs/pots are usually mapped to control these virtual dials so the virtual dials end up serving a feedback and status function.

share|improve this answer
add comment

One reason why you might want to consider the rotary dial is if the physical space doesn't allow you to implement a slider or spinner control (e.g. if the space is square rather than long and narrow). I would hardly advise people to design based on the space as a first concern, but if you have a couple of interface controls that are circular (e.g. a clock or a gauge that goes from 0 to 100%) then it is also good to have a consistent appearance for aesthetic purposes. I have to say that from my experience physical control on household appliances are being phased out by digital controls, because now that people are more used to software interfaces some of the designs are moving in that way.

share|improve this answer
    
And I find this trend disturbing. :) I have muscle memory with a knob. I can turn a dial on the oven to 450 degrees F without looking at it while a digital control requires my complete attention and takes more time. –  Steve Moser Apr 8 '13 at 20:12
add comment

What makes physical knobs so good?

The core attributes of a physical knob are resistance and inertia.

Inertia smoothes the "user input", supressing smaller shaking, allows fine-tuning at low speeds and easy turning at high speeds.

Resistance adjusts the force required to something comfortable to humans, avoiding accidental changes, providing tactile feedback of "something happening". "A little bit stronger" also seems easier than "a little bit less strong" (I'm not sure how to put that better).

Low-resistance, low-inertia knobs are flimsy, which in modern lightweight, small and mobile devices can be countered by at least giving low tolerance in the rotation axis.

Low-resistance, high-inertia knobs allow a wide range of values with high precision. If the knob allows "spinning" over multiple revolutions, you can easily span multiple orders of magnitude without losing precision.

High resistance knows are a kind of safety feature: "Do you really want to turn this up to 11?", they make sure you are aware you are doing somethign important. High inertia makes them even more like confirmation dialogs, but you lose precision.

(for completeness: aspects to avoid are hysteresis and a significant difference between static and dynamic friction).


All these things transfer poorly to a digital user interface - at least I can't imagine a way. Touch interfaces make rotation at least physically bearable, resistance might be simulated by a vibrating surface, and inertia can be simulated in both the visual feedback and the effect.

Still, you are losing the tactile feedback of a high-inertia knob, which in my personal opinion is the most satisfying thing about those vintage devices.

share|improve this answer
2  
If you are as childish as I am you would have also lol'd at "What makes physical knobs so good?" –  Captain Sep 11 '12 at 10:33
    
Touch screens don't provide tactile feedback on input. We love touch screens. –  Kris Sep 11 '12 at 12:52
    
@Kris: IIRC there are some touch screen concepts that vibrate to simulate something akin to resistance. –  peterchen Sep 11 '12 at 16:08
    
@Captain: now I did :) –  peterchen Sep 11 '12 at 16:09
add comment

There are several reasons (some are posted here but I'll sum them up anyways :) )

  • They can be made to scroll to infinity without requiring an extended space or a graphical visualisation that needs to alter its appereance during use (for example - infinity scroll bars that decrease the handle size).

  • It's a really fast way of using a small area to scroll through large amounts of data. It can be further improved (in some cases) by letting the knob be controlled by the mouse wheel.

  • They can be made to represent a finite turnable area by using virtual boundries (for example make a needle stick out and a track that it runs in).

  • The space that a knob uses is different from the space that other controls use. Knobs uses a space where the side is as long as the height. Sliders, for example, needs another kind of space where the height is greater than the width because of the nature of how they look.

  • There is software (in music production for example) where hardware can be connected to control the software. In those cases, it's good mapping that knob hardware controls software that looks like knobs.

  • The way that a user interacts with the knob can be altered to fit the users needs. For example, by dragging the mouse around it or by dragging the mouse up and down. In the latter, it is controlled like a slide, but the visual area can still have the advantages that a knob can have.

Knobs aren't always the best alternative, there are plenty of controls to choose from, but sometimes they are really usable!

Ps. I know it's like saying a curse word in a church to mention this on an interaction design forum, but sometimes it's just so much cooler with a vintage knob than anything else :)

share|improve this answer
2  
+1 for first point. Other controls would have to increase in size but you don't need a bigger knob to do the job. –  Captain Sep 11 '12 at 11:59
add comment

I often see software dials on music software, for example for adjusting the volume of a lot of channels or audio bands. Software dials are easy for the mouse or touch pad than sliding buttons. Rubbing motion in a small circle is much easier than having to move in a straight line along a certain direction. Also regarding the resolution, stopping at a certain angle on a dial is much easier than stopping at a certain length on a slider.

share|improve this answer
    
I would contest that statement that virtual knobs are easier to manipulate than virtual sliders for no other reason than sliders are typically much larger as well as more ubiquitous. –  DA01 Sep 11 '12 at 16:14
    
@DA01 Sliders are much larger because its harder to accurately stop at a certain point. Dials/knobs are smaller because its easier to stop at a certain angle, and hence are more ubiquitous. –  sawa Sep 11 '12 at 18:07
1  
Where you you find knobs being more ubiquitous in software? That may be true with audio software, but would argue that it's less about usability and more about clinging to the physical devices the audio software is emulating--which can be a valid argument for their use. But in software in general, one rarely sees a knob used as a virtual control. Even with simple audio and video players, the slider is the more accepted UI control. –  DA01 Sep 11 '12 at 19:29
    
Sliders can be implemented so that they have tick marks that can be snapped to, or you can also add an input box or spinner control that gives you better precision compared to the dial. –  Michael Lai Apr 7 '13 at 22:52
add comment

Aadaam talks about the iPod, which for a long time used a physical 'dial' and later a touch 'dial'. Not quite a knob, and not quite purely software, but an interesting use of it.

I do think that a knob is a difficult virtual control given that it doesn't easily map to the hardware input device. Even on a touch device, it's a bit difficult as you have to track multiple angles while looking at the screen while doing it. Perhaps a more virtual 'gesture' based dial could work in some cases.

But, in general, what makes knobs so great is their physical interaction. They take up little space, give immediate feedback, and are rather intuitive. Alas, you lose those features in a virtual software based UI.

share|improve this answer
    
+1 for point about mapping. Handling a physical knob is much easier with hands than with a mouse. Maybe gesture based is the way to go. –  Captain Sep 11 '12 at 15:22
add comment

When you want to enable endless cyclic turning e.g. to control the angle of something. In this case adding a dent e.g. this search result + a numeric value at the side to make the current angle clear is important.

If you want to limit the angle of the dial, you can draw a dial with a position indicator that sticks out and use a track around the dial with limited angle to visualize the posible range in which you can turn the dial (that way it seems like the indicator gets stack at the ends of the range thus physically limiting the range).

share|improve this answer
add comment

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.