Testing Touchscreen Devices

Testing Touch Screen Devices

Testing Touch Screen Devices

Many of the electronic devices today utilize a touchscreen interface, and for some that is the pretty much the only way of communicating your commands to the device. These days, this particularly includes ‘smart’ phones and tablets, which have enough computer power to require a lot of interfacing by the user in order to perform the wide range of functions they can be programmed for.

Touchscreen technology has been with us for quite a while and has utilized several methodologies. For instance, one early version was not really a touch screen at all; it had a frame which projected a web of light beams across the screen and detected ‘touch’ by an intersection of horizontal and vertical beams being interrupted. Another pseudo touch technology used an optical ‘pen’ device wired to the monitor to detect where on the screen was being ‘touched’. More recently, I’ve got a ‘Palm Pilot’ which seems almost like it has a ‘flexible’ screen which detects touch by the deflection of the screen where it was being touched.

Of course, the current crop of touch screen devices seem to be mostly ‘capacitive’. This means that touching the screen at some point with your finger (or other device possessing capacitance) produces an electric charge differential which the device can detect. This is pretty much state of the art currently, but there is the ‘Blue Tiger’ stylus project for the iPad, which incorporates Bluetooth somehow and may become a step forward in touch technology.

Note that testing touchscreen devices will have a ‘hardware’ aspect of having the touch or gesture be recognized, and a ‘software’ aspect of getting the expected or an appropriate response to a detected touch or gesture. When testing devices, the hardware aspect (detection of touches) will be of major significance. Tests of apps, on the other hand, particularly generic apps for many devices, may consider the hardware aspect to be ‘out of scope’, and only be interested in the software aspects of touch testing, that is, what is done with a detected touch.

In order to test touchscreen devices, you have to, well, touch them. A lot. Since most current ones are designed to be touched with your fingers, a major portion of your testing should be using your fingers. Most people use their forefinger on these devices, so this should be the focus of your touch testing, but for hardware testing, it would be wise to include a little testing of usage of a smaller finger (little finger) and a bigger finger (thumb). You might even want to try something other than a finger, like a knuckle. There is probably no need to try something really wild, like your nose (go ahead, try it; you know you want to), since if you can’t see what you are touching, it is unlikely that a real user will expect that to work.

For any testing, cover all the supported gestures – touching and holding, tapping, touching and dragging, swiping (to change pages), SWYPEing (Android function of typing by sliding your finger from letter to letter without leaving the screen, with auto space insertion when you lift your finger), and the 2 finger gestures to shrink or enlarge an image (with fingers on the same hand, and for hardware testing, also fingers on separate hands).

You probably want to try some ‘unsupported’ gestures as well to ensure they are handled appropriately. In addition to ones above which are not supported by a particular device or app, a couple of good general ones are to touch 2 control areas simultaneously or to touch an active area while already touching a non-active area.

Note that many devices have a small screen (phones) and some apps have really small or crowded control points. Other devices and apps allow you to draw or write on the screens to generate text or drawings. Of course, you should test these functions with your finger to ensure that they can be used with nothing more than fingers. But be aware that these functions can be much more user friendly when using a stylus.

Thus, for testing any hardware, or apps which would be benefited by stylus use, some testing should be done with an assortment of styluses (or styli). There is a wide selection available, so the best choice of what to use can get complicated. Unless the device or app comes with a stylus, or the people commissioning the test define which stylus or styli to use, of course.

Many users will run out and buy the first cheap stylus they find, so you ought to have one of these in your test toolbox. You can get them on eBay for a few dollars, or perhaps in a local computer accessories department although they may cost a lot more in some of these stores. Generally these have a rubber tip of about 8mm or 9mm in diameter. You might want to get at least one backup, since the tips tend to tear or wear through, which could be a problem if it happens in the middle of a test.

If what you will be testing (or even your personal use) will involve a lot of stylus work, you might want to add a quality ‘fine point’ stylus to your toolbox. These generally have tips 5mm or 6mm in diameter and are a bit harder to find online since they are not reliably labeled. Look for ones which claim to be ‘fine point’ and then check with the seller to find out the actual tip diameter before ordering. These may be rather more expensive, $15 to $30 or even more. Unless specified or indicated otherwise, a good test might include a ‘cheap, large tip’ stylus, and your own personal favorite ‘fine point’ stylus. If you do hardware testing, you will want to have a selection of styli with differing tip materials.

Note that the size of the tip has NOTHING WHATEVER to do with what results from a (detected) touch. Any touch detected at a location is treated exactly the same. The smaller sized tip simply allows you to better see where exactly you are touching, which can become important if several active areas are close together or you need to make precision marks.

There is at least one stylus tip out there which is smaller than ‘fine’, about 4mm (HAND stylus). This may not work well with all devices, since many require a contact area of 4mm in diameter in order to detect the touch, so you would have to really mash this one down to get enough contact to be reliably detected. Also, you should avoid using a ‘pencil point’ stylus unless you are specifically required to test with these. These tend to not work well on most devices (the tip is much smaller then 4mm, so bends or ‘mushes out’ if it can be detected at all) and can even cause damage to the screen.

There are some weird styli out there. One has a tiny metal ball, nestled in a clear vinyl disk (Adonit Jot; the disk has tiny wires in it to expand the touch past the detection limit). This provides very accurate touch location because you can see exactly where the tip is touching the screen through the clear disk. But I suspect if that disk were to pop off, metal to screen contact might be disastrous and would at best put you out of action until you can get a replacement tip/disk. Another strange one is a ‘brush’ (Sensu Brush or Nomadbrush) rather than a stylus; these are particularly useful for ‘paint’ apps. One of these oddities should be utilized in testing only for specific devices or apps which require such testing, unless it meet the qualification of being ‘your favorite’.

As mentioned, stylus tips are often hollow rubber. These compress to meet whatever diameter is required to be sensed, but can be fragile and in some cases cause a lot of drag. Other options include solid rubber (can require lots of pressure to be detected), ‘Nerf’ foam (can be flimsy and may not be detected reliably) and microfiber (such as the LYNKtec Truglide, no commonly reported problems with this material that I could find). Microfiber and other cloth based stylus tips seem to be generally well regarded due to their more reliable performance and sturdiness, not to mention not leaving any smears on the screen like rubber tips usually do. They may crowd out the classic rubber tip from the market.

Physical attributes of the stylus other than the tip don’t generally factor into testing since your test environment can be controlled so that the effects of the physical attributes can be minimized, but these factors do have a big impact on customer satisfaction, and thus how common the usage of that stylus is likely to be. So physical differences should have some input into your stylus choice, since you want the styli you test with to be ‘commonly used’ as well as at least one of them being ‘the one you would choose to use’.

There are ‘capped’ and ‘uncapped’ model styli. Capped styli have the tip covered when not in use, which provides them protection usually needed for ‘pocket’ carry, but requires you to uncap them or do some other action to reveal the tip for use Then you have to take some action again to return it to protected status (if you can even find where you put the cap). Uncapped styli are less of a hassle to use, but often need to have the tip protected via external means when transporting them.

A ‘multifunction’ stylus, such as one with a pen or laser pointer or LED light has no impact on the test, but may increase the popularity of the stylus. And price. Generally, these should be avoided unless you really like that sort of thing.

Barrel size, shape and grip, as well as overall balance and weight are personal preferences. Length can be important. Shorter styli can be easier to carry and are generally good for navigation. Longer styli can be more effective to use for writing and drawing or long sessions of use. Obviously, ‘both’ short and long might be best, but would seem to be contradictory in nature. Not necessarily.

‘Stylus-R-Us’ has a line of telescoping styli which can be effective for both types of uses; they also use a tip material one reviewer likens to a ‘ball of fuzz’ which seems to require almost no pressure and provides no drag. This company also offers a good selection of capped and uncapped fixed length styli.

Most stylus reviews include and end up with fairly good opinions of the Kuel H10 and Wacom Bamboo among the ‘fine point’ stylus choices. Keep in mind that these are various reviewer’s personal opinions, and even if they ‘love’ one of these, each has negatives listed. The Kuel’s main disadvantage is usually said to be it’s short length, and the Bamboo is a little too ‘mushy’ for some reviewers. Reviews which include the Jot vary from ‘love’ to ‘hate’. Not many comparison reviews include any Stylus-R-Us models, but individual reviews of this brand are positive unless the tester tried using one of these styli with a heavy touch.

If you have an appropriate store available locally, you can try out several styli to determine how they work for you (bring your own devices if practical). For others, you may have to actually order them to try them out. Ebay and Amazon would be the first places I would check, then online stylus specialists and the brand’s web site. I’ll report on my experiences with some of the top choices in a later blog.

As mentioned, these touchscreens work by ‘capacitance’, which is an electrical characteristic of your body. Optimal touch reliability will be realized by holding the uninsulated device in one hand and controlling it with the other hand. Your hardware testing, with fingers and styli, should include the ‘optimal case’ mentioned, as well as having the device on a table, or in a holder and/or an insulating case, with no contact with your body other than your controlling finger or the stylus. Screen protectors and cold hands and fingers can also reduce capacitance and thus reliability of touch sensing, and might be considered for hardware testing as well.

Software testing is best done using the ‘optimal case’ for greatest touch detection reliability. It is best if you can eliminate hardware based marginal touch detection as the culprit if you do a touch and the app seems to ‘ignore you’.