• Replies 7 replies
  • Subscribers 18 subscribers
  • Views 362 views
  • Users 0 members are here
Options
Related

visible pixel mask for Instinct 3 solar watchfaces?

6125703
16 days ago

I tried to search but couldn't find clear answers on this. Is there any available mask image or detailed geometry spec to tell you what pixels are actually visible out of the 176x176 addressable area on the Instinct 3 MIP screen?  It seems silly if everybody has to reverse-engineer this via test patterns.

Are the three rounded corners plain arcs? What is the effective radius and center for these in terms of pixel units?  To my eye, it looks like one circle concentric with the circular bezel.

For the circular sub-display, what is its visible inner diameter in pixel units?

How many pixels are obscured as you go from the main display, across the border, and into the sub-display in the scan line that bisects the sub-display?

Top Replies

All Replies

  • 16 days ago

    It's the same as the Instinct2.

    Starting with the subscreen, you can find it's location and size with  WatchUi.getSubscreen()

    You get its x and y, width and height
    That also allows you to see the width of the screen to left of the subscreen and the height of the screen below the subscreen  You need to adjust for the size of the circle itself.
    As far as the semioctagon shape, I usually have to just run a few tests and see what works
  • 16 days ago

    Based on the simulator I also think that the rounded corners are a circle around the center "pixel" of the screen. Haven't found any better resource that describes what you are looking for.

    I'll ask a counter question: why do you need this? Isn't looking at the simulator good enough?

  • 16 days ago in reply to flocsy🤠
    flocsy🤠 said:
    Based on the simulator I also think that the rounded corners are a circle around the center "pixel" of the screen

    Sorry to nitpick but there can be no center pixel on a 176x176 display (or any display with an even width and height). This has actually come up in the past as a practical concern, as in order to draw a "proper" circle using drawCircle, one would have to call drawCircle 4 times (with each circle being centred around a different one of the 4 pixels in the center of the display, and all the circles overlapping to create a "proper" circle.) (By "proper circle" I ofc mean a circle that appears to have the same thickness all the way around, and appears to be perfectly centred with respect to the display.)

    It would be more accurate to say the "center of the screen", or if you must refer to pixels, the "4 pixels in the center of the screen".

  • 16 days ago in reply to flocsy🤠
    flocsy🤠 said:
    I'll ask a counter question: why do you need this? Isn't looking at the simulator good enough?

    I think this was already addressed in the OP

    6125703 said:
      It seems silly if everybody has to reverse-engineer this via test patterns.

    I totally get it - it's annoying to have to create test apps and eyeball which pixels are visible (in the case when you want to create pixel-perfect apps, ofc).

    Obviously it's easy for rectangular or square displays. For round displays and even the old semi-round displays (shudder *), you can write a function to figure out the bounds of the display

    (* these are flat on the top and bottom, and rounded on the sides, but I *think* you can calculate the bounds if you assume the side arcs would join to form a circle if you extended them further, as the watch itself is circular, and the arcs seem to follow the curvature of the watch)

  • 16 days ago in reply to FlowState

    I added especially for your eyes double quotes around the word pixel.

  • 16 days ago in reply to FlowState
    FlowState said:
    I think this was already addressed in the OP

    Not really. Because even though I agree it would be nice to have this readily available from Garmin, I'm not sure how useful it would be. For instance it's not very useful if you have a text close to the edge, because it depends on the font, robably on the translation of the text, etc. It can be maybe useful when you draw static things though, but even then, it would only be useful in limited cases IMHO.

  • 15 days ago in reply to FlowState

    Yeah, you get me... If there was a truly interactive and accurate simulator, I could stomach trial and error more. I can spend time thinking about it while walking, and do a little napkin math to triage my ideas, but don't want protracted sessions at the Monkey grinding wheel.

    To me, this is just a hobby and I don't want it to take on all the mundane burdens of a day job in computing.