0

I come from a non-GUI and lower level programming environment. I have solid experience in OOP programming and algorithms, but the projects I worked on did not have any impressive user interfaces, hence my post here.

I want to understand how the graphics are updated in a mobile app.

For instance, let's take the basic example of a classic wall clock, with second, minute and hour arms. The Wall Clock app will move the seconds arm each second, likewise the hours arm each 60 mins, and minutes arm every 60 secs, problem I have is that I just cant wrap my head around creating these moving parts in an efficient way.

Am I supposed to save images of all possible combinations of second,minute and hour and just morph the 3 and display the appropriate time each second? Or is it a better idea to have the background image as a non dynamic entity, and have only the "seconds, minutes and hours" arms rotate? If this is the case, when I package my application, I will have to also include 60 + 60 + 12 = 132 images (60 images for seconds, 12 images for hours and 60 images for minutes), is this correct? If it is, this seems awfully inefficient...

Just to kind of give you an idea of what I am having problem understanding, consider another example that requires updating the graphics each time microphone hears varying noise levels: Assume that I want to create an app that measures the sound levels in an environment and it shows the dB measured from the microphone on a volume bar. If there is a song playing near the microphone with varying noise, the sound level indicator will move up and down. I understand how to update the noise values to the command prompt, serial window, or a TextView box, but I don't get how to update this info graphically.

I may be wrong, but think I must have a volume bar image that is just empty(as if the noise is at zero) and in the program, I fill this volume bar according to the dB levels acquired from the microphone, but then how can I fill the volume bar in real time depending on the noise level seen by the microphone?

These questions I have are not just Android specific, I'm sure the same approach to updating a Wall Clock app is also used in Windows application development, so I feel like if I have a good understanding of GUI development irrespective of the environment, I may be able to tackle these silly questions with ease; perhaps you can also direct me to a good book on GUI development in Java or C/C++.

2 Answers2

0

For your clock example: you would not generate images for every position a clock could be in. Your app would be HUGE! You would instead make images for: the face of the clock, the minute hand, the hour hand, and the second hand. You could then take the time and calculate the angle of each hand of the clock from some standard point on the clock (this could be any point you want). You could then rotate each image to that calculated amount. This would make the clock appear to have its hands pointing at the right time.

I don't know very much about audio bars, but it would work in a similar way. Make a formula to calculate where/how large the bar should be. Then draw a bar stretched to what the formula dictates.

Armek
  • 1
  • 2
0

GUI and custom drawing can be achieved in many different ways, each have varying efficiency.

How to draw?

For example, if you want to show a cartoon character walking, in most cases you need to store each frame as image and redraw it, probably on every tick or second.

But for wall clock, you don't actually need to

  • Store so many images for every possibility because you can put single arrow image and rotate it
  • Store any image at all if you can draw it procedurally by defining its shape

Suppose you have a canvas on easel that you can draw whatever you want on it. You can choose which order you want to draw. For your wall clock example, you can first draw your background (numbers / hours) and start adding minute hand and hour hand on top it. This is called z-order.

Operating system libraries gives you widgets that you can interact with it. Canvas (picture box) is one of them and you can to draw whatever you want that you to build your custom widgets and of course Android has this too.

Android has Views and each view provides an method that you can implement your custom drawing logic:

// Implement this to do your drawing.
// Parameters
// canvas   the canvas on which the background will be drawn
protected void onDraw (Canvas canvas){}

A very crude wallclock can be implemented with just using:

canvas.drawLine(float startX, float startY, float stopX, float stopY, Paint paint)

Beyond this, we can basically apply OOP concepts in GUI programming too. For instance, you can define every thing you want to draw as an object that implements a common drawing method. This is widely used on 3D game programming as usualy 3D rendering engine manages every object and asks them to draw themselves if "camera" sees them.

When to draw?

Normally this is up to operating system. It draws elements whenever it is required. However, for game like applications and for custom animations you might need to manage this by yourself.

On most operation systems, you can use seperate threads, a timer, an infinite loop or a notification event to force redraw your screen. You can achieve this redrawing with postInvalidate method on Android.

But not every environment is easy. For example, your application might need to redraw itself 30 times every second and depending the operation system and environment, you might need to acquire locks to display your UI updates correctly and might need to make your thread sleep to achieve desired framerate without keeping CPU busy. (This is used in SDL library.)

Android has an animation class to tackle common tasks.

Thing to remember

Reusing widgets that OS gives you is a good idea. You can implement a dB meter by just using a progressbar.

Community
  • 1
  • 1
Furkan Omay
  • 1,047
  • 1
  • 11
  • 22