15

I'm building a drawing application that let's users annotate images and diagrams in a freehand style.

Not unlike a glorified MS Paint program, but there is heavy background processing going on.

The question:

Have there been studies done to quantify how fast "drawing" style interactions need to update in order to seem "smooth" to users? (It terms of both input lag and update rate)

I'll certainly bring in users for testing, but I'd like to know where to set the initial goal posts.

Update to clarify some points raised below:

  • Users will care about accuracy as tight as 1-3 pixels, so tight freehand drawing tool control is necessary.

  • "input lag" and "frame rate" are issues that both need to be addressed, but they are not the same.

  • Not a touch interface. Users would be using a regular mouse.

Improve this question

5 Answers 5

Reset to default
20

For the 'input lag' part of your question, I still use the rules of thumb found in Nielsen's Usability Engineering:

The basic advice regarding response times has been about the same time for many years [Miller 1968; Card et al. 1991]:

  • 0.1 second is about the time limit for having the user feel that the system is reacting instantaneously
  • 1.0 second is about the time limit for the user's flow of thought to stay uninterrupted
  • 10 seconds is about the time limit for keeping the user's attention focused on the dialogue. For longer delays, users will want to perform other tasks while waiting on the computer to finish.

Miller, R. B (1968), Response time in man-computer conversational transactions, Proc. AFIPS Spring Joint Computer Conference Vol 33, 267-277

Card et al. (1991), The information visualizer: An information workspace, Proc. ACH CHI'91 Conf. (New Orleans, LA, 28 April - 2 May), 181-188.

Improve this answer
4
  • Almost wonder if that last point - on 10 seconds- is relevant anymore, given how fast bandwidth has gotten and how easily users can abandon sites if they're loading for too long.
    – firedrawndagger
    Apr 26, 2013 at 16:51
  • 1
    Nielsen kind of addresses this in his 2010 update to the above: "No matter the implementation...we are discussing user experience, not coding. Therefore, the response time guidelines for web-based applications are the same as for all other applications. These guidelines have been the same for 37 years now, so they are also not likely to change with whatever implementation technology comes next."
    – calum_b
    Apr 26, 2013 at 17:09
  • Ahh good link, what I was referring to was this article from NYT which shows that technology (like mobile) is indeed shifting browsing behaviors towards impatience.
    – firedrawndagger
    Apr 27, 2013 at 21:47
  • To be honest, I wouldn't be surprised if modern technology does eventually have some sort of measurable effect on these numbers, despite Nielsen's assertion. But even if it does, for something that innate the timescale will more likely be measured in generations than years.
    – calum_b
    Apr 28, 2013 at 16:16
14

US Department of Defense Design Criteria Standard - Human Engineering MIL-STD 1472F Section 5.14.9 and Table XXII requires that "Sketching" have a response time of 0.2 seconds "from input of point to display of line." That's a minimum standard of performance, so it should correspond to your worse-case conditions.

Like a lot of standards, I believe much of MIL-STD 1472 is the result of operational experience of experts in the field in addition to academic research. Someone made an app with excessive lag, and it was unusable, so they wrote a standard to prevent that from happening again.

0.2 s is pretty close to the minimum human reaction time (e.g., Kosinski RJ, 2010, A literature review on reaction time. Clemson University), so it probably has something to do with the fact humans can’t respond to substantially faster updates on the screen even if they wanted to.

Improve this answer
2
  • That's a great resource. Do you know where to find the studies that support the conclusions reached?
    – nonot1
    Mar 1, 2011 at 19:36
  • I added the last two paragraphs, but they're not too much help.
    – Michael Zuschlag
    Mar 2, 2011 at 14:22
3

What you are asking is how fast of a frame rate should you be able to maintain. I think 24 frames per second is generally considered good enough for "smooth motion", but there is some debate. Some people will say 30 frames per second, others more. It somewhat depends on what sort of graphical change is happening on screen.

This is a standard wikipedia page, but it's a good place to start: http://en.wikipedia.org/wiki/Frame_rate

Improve this answer
3
  • Or, he may be asking what the maximum acceptable "delay" between freehand movement and response on the screen is?
    – jensgram
    Mar 1, 2011 at 7:33
  • Yeah, vision is just one component, responsiveness to user input (e.g. mouse movements) is the other. For a drawing application, frame rate likely isn't that important because there's usually only a single object moving. --- The "Visible frame rate" section of the article linked has some good insights, still.
    – peterchen
    Mar 1, 2011 at 9:02
  • I took a film class in college. I learned that 15 FPS is the threshold between "moving pictures" and a true movie.
    – JoJo
    Mar 1, 2011 at 20:40
3

Just a hypothesis, but I would suggest that if there is such as thing as an acceptable lag then it is related to how precise the task being performed is.

For example; there were numerous reports when Microsoft Kinect came out about there being a lag between the person moving and the action being replicated on screen but as the actions being performed do not require precise inputs this is less of a problem than it would be for a task such as pixel-accurate drawing.

The more pixel-perfect the required action the less lag is acceptable.

However there is also the suggestion that if the system you are designing is less usable than a paper and pen alternative then users will just print out the image and annotate it by hand.

Improve this answer
1
  • I disagree there: the lack of response for Kinect (or anything else) doesn't matter for accuracy. What matters is the user ability to use it - and with lag, the user will get pretty fed up with it very quickly. I have some digital headphones that has some lag and they are useless for playing music as the slight delay between strumming a chord and hearing it makes the whole experience unsatisfactory. The reason they're ok for listening to mp3s is there is no user feedback at all.
    – gbjbaanb
    Feb 12, 2012 at 13:58
2

Slightly off topic (as OP asked specific to touchscreens)

In kinect-style devices, when the "ping" rate drops below 30 per second, that is about where the lag becomes noticeable. That works out to roughly 30ms acceptable delay.

For touch screens I would target a similar acceptable delay.

Improve this answer
1
  • See clarification. Not a touch screen.
    – nonot1
    Mar 2, 2011 at 0:40

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

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