The long road to a final unit

[levi]

AFFS+ is supposedly better in sunlight, which is a very good idea for something that isnt a desktop.

I don't care which technology is used, as long as ghosting is reduced compared to my Pandora, and there aren't too many overdrive sparklies (think that's only an issue with TN matrix screens). If you can read text on it outdoors that would be a massive improvement on my Pandora where I can only just see window edges and hints of icons.

 

[klapse]

My thinkpad tablet hat a BOE Hydis display. It was very nice. For sunlight reading nothing i saw beat the Nokia N810 and N900 with transflective display.

 

[Djhg2000]

No digitizer or touchlayer is transparent. Other than surface wave you no can haz an optically perfect solution.

The Wacom digitizers sit behind the screen and therefore doesn't have to be transparent. They're still too expensive but I felt that this needed to be clarified.

 

[dev-sst]

Well, you can use Android and touch-focussed stuff with a resistive screen but you cannot do Linux with a capacitive one.


So a resistive works for both, a capacitive one only for one.

Hi,

capacitive touchscreens work for Linux too, my Acer Aspire R7 works perfect with Kubuntu 13.04 and 14.04 (multitouch tested with Qt 5 demos).

It also has a touchpad, but I never use it (it is also in an odd place) and the keyboard is just in the right place so my palms can rest on the desk (not on the touchpad).

One problem is that most adjustment software (not the calibration software) is windows only for capacitive touchscreens.

But that is mostly a non issue because it's always resolved by the supplier, the customer would never worry about adjustment (calibration is still possible).

Some Pro & con of capacitive touchscreens that I've discovered during researches for my work

Pro:

-Very accurate since no physical pressure is needed, with a capacitive stylus it is easy to archive pixel perfect accuracy.

-Possibility to use real glass instead of a foil -> costs more

-Extremely low reflections are possible with optical bonding -> but that adds _much_ on the price

-Some controllers have auto adjusting capabilities that adapt to the environmental EM field on device startup (but all controller samples I've seen so far didn't, so who knows if it's not just marketing twaddle)

-Can measure the touched area instead of touch points (tough it is mostly useless to do this, IIRC no software / SDK supports this)

Con:

-Costs much more (resistive stuff is cheap, mostly made from plastic, does not require special adjustment ... etc.)

-Adjustment (which is different from calibration) requires either specially shielded layers on the glass, or is done in an EM field free location -> higher cost + it is mostly impossible to correct the sensitivity "by hand" without totally screwing up the adjustment

-Changing adjustment data (which may have to be changed due to hardware redesign) adds to the costs (IIRC prices are around 1000€ one time fee for a 9" device)

-Testing may be hard since the adjustment can only be made when the complete hardware is in place, otherwise EM fields may disturb the controller

-Thicker glass means less sensitivity, thinner glass means less robustness

-Touch pressure cannot be measured

If you really consider to get a sample, egalax controllers worked well in my tests, there are also german suppliers to get these.

BR

dev-sst

 

[TrashyMG]

-Very accurate since no physical pressure is needed, with a capacitive stylus it is easy to archive pixel perfect accuracy.

Are you talking about capacitive here? sure maybe with an expensive active stylus that this project can't afford to license sure, maybe.. but not not by itself. On my phone using VNC to my Linux desktop is an act of pure frustration as my large fingers can't press the right buttons in the menu.. trying to hit the close button instead I maximize the window.. so yeah I would rather stick to a resistive screen with a much cheaper stylus.

 

[EvilDragon]

Ever tried pressing a small X to close a window on a 5" capacitive screen reliably?


Every OS I've seen that works on a capacitive screen has either huge buttons (or a huge screen) or has some estimations what the user wanted to press (which does only work if there aren't multiple buttons in one area)

 

[ShleeDragon]

http://www.instructables.com/id/DIY-Capacitive-Stylus/

or just use a thin sausage!

 

[bzar]

Try with capacitive stylus and resistive with stylus: Make a tight 8x8 grid of 16x16px icons. Try to hit them all once in a random pre-decided order as fast as you can. Misses count as extra 2 seconds each.

 

[slaeshjag]

I just want to add, that cheap capacitive touch screens are even worse. I have issues pressing the right buttons on my android phone.

 

[thatgui]

-Possibility to use real glass instead of a foil -> costs more

Does this really count as an advantage, especially if we are talking about a clamshell device ?

 

[dev-sst]

Ever tried pressing a small X to close a window on a 5" capacitive screen reliably?


Every OS I've seen that works on a capacitive screen has either huge buttons (or a huge screen) or has some estimations what the user wanted to press (which does only work if there aren't multiple buttons in one area)

Of course it will not work on very small screens, unless the screen has a resolution like 320x240.

F.e. it could work with plasma active DE, but not with a vanilla XFCE.

MS has made a study available, it's about the minimal size of touch targets in order to get useable finger input.

The very minimum size of a button has to be at least 5x5mm (they translate it to an average of 97% accuracy).

Again with a good stylus (somethign like Nota) it could be possible even for small screens.

On the other hand a 39.6 cm (15.6") screen running a 1920x1080 resolution is large enough to hit every button with a stylus, and most buttons with finger input.

IMHO it is however out of the question that the Pyra is better off with a resistive touchscreen.

@thatgui: Glass is on the pro site because it allows to use fast swipes, which are rather hard to do when your finger needs to apply a constant pressure

BR

dev-sst

 

[GizmoTheGreen]

personally I'd appreciate an option for no touchscreen at all, for better colors. 

perhaps when the time comes to add touch, evildragon can take pictures of both in various conditions. if it's transparent enough maybe it doesn't matter.

 

[bzar]

Of course it will not work on very small screens, unless the screen has a resolution like 320x240.

F.e. it could work with plasma active DE, but not with a vanilla XFCE.

And because of that it also doesn't work with games that require similar input. I've tried playing SCUMMVM games on a capacitive screen once. Never again

 

[matzesu]

Ditnt used the Stylus on my Pandora often, until Drastic was out, was in fear for streches in my screen..

I like this Dual Nub Controll on my Desctop, because is realy easy to left an right click whit it... ,

for me, a resistive Touchscreen, whit a good Stylus (not just a Platik Stick), would be ok,

Also a Capazitive Touchscreen would cause a lot issue so the 2 Months wouldnt work..

Bye the Way: is the Touchscreen working yet?? Or do whe only have a Screen??

 

[Exophase]

Again with a good stylus (somethign like Nota) it could be possible even for small screens.

So you're saying that a stylus with a 3.7mm tip can be as precise as one with a ~1.5mm tip like in Pandora?

Not to mention, an active stylus that takes batteries and won't fit into a compartment...

 

[Granitehead]

Again with a good stylus (somethign like Nota) it could be possible even for small screens.

So you're saying that a stylus with a 3.7mm tip can be as precise as one with a ~1.5mm tip like in Pandora?

Not to mention, an active stylus that takes batteries and won't fit into a compartment...

Actually I tested this on my Android phone (Nexus 4). I turned on a dev layer which shows touch screen input. Then I moved my finger around. My results were that the touch layer was at least good for +-1mm accuracy. It doesn't really matter how big the tip of a stylus is, as long as the touch registers precisely in the middle of it.

But an active and too large stylus is still just a no go.

 

[Exophase]

Actually I tested this on my Android phone (Nexus 4). I turned on a dev layer which shows touch screen input. Then I moved my finger around. My results were that the touch layer was at least good for +-1mm accuracy. It doesn't really matter how big the tip of a stylus is, as long as the touch registers precisely in the middle of it.

I'm talking strictly about how much the stylus gets in the way of precise input. The touch controller and software may do a good job detecting centers of touch blobs, but a larger stylus is still going to make it harder for the user to see what they're touching. That's the biggest point of the stylus dev-sst linked to, a smaller and more rigid tip. They even have one much smaller than the one he recommended - it also costs much more, so you have to think they see it as a big benefit. But that tip is still larger than Pandora's (and Pyra's could and possibly should be finer, or there could be replacement options that are), and that's a small and cheap piece of plastic. I don't see how one can it's out of the question that capacitive isn't superior when it needs a $79 large and less portable stylus and much more expensive assembly to even get back to the baseline given by Pandora. Pyra is going to be expensive enough as it is, I'm not the least bit enthusiastic about finding news ways to grossly increase the price, especially for something with such questionable benefit.

Not having to push on the screen is a benefit but it's also a downside since it makes it easier to accidentally hit inputs. I think most of us would agree that so long as the pressure threshold is low enough it's fine - there was a legitimate problem with some Pandoras requiring too much force so I think ED has to be careful about controlling this however he can. But if it's like the CC Pandoras I have zero problem with the pressure threshold and I don't think I've heard of others who feel that this is too much.

The real benefit for capacitive remains decent multitouch. Outside of some Android games (mainly the ones that need gestures) there just isn't enough point, not with so many physical inputs.

 

[fusion_power]

I'm still for an improved Pyra Stylus design where you can simply change and replace the tip to whatever you prefer, should be doable without much additional effort:

 

[Granitehead]

I'm still for an improved Pyra Stylus design where you can simply change and replace the tip to whatever you prefer, should be doable without much additional effort:

Nope. More difficult to manufacture. You either need a more complicated mold or need to drill that hole. Possibly twice as expensive as the Pandora one. Not that that's very expensive, but it's still significantly more effort.

 

[Grench]

2X - the stylus and touch method on the Pandora was pretty much perfect for this type of device. It isn't a finger-touch screen so much as a stylus-touch screen. The OS is such that icons and selection boxes are going to be tiny - which dictates a finer point stylus AND actual contact.

Why argue when the -right- technology for the job is actually less expensive to implement?

Next people are going to want the thing to be white with a silvered metal band around the edges, proprietary plugs and filled with glue so the battery can't be changed without destroying the device.

I'm still for an improved Pyra Stylus design where you can simply change and replace the tip to whatever you prefer, should be doable without much additional effort:

I'd take THAT idea one step further and make an optional stylus out of solid stainless steel with replaceable tips that fits where the stock stylus does. AKA the sculpted Stylus from the Tapwave Zodiac days.