-
-
Notifications
You must be signed in to change notification settings - Fork 257
NeoBuffer object
The NeoBuffer is basic image storage object. Based upon the "method" object it is defined with, it can store the image data in RAM or reference image data stored in PROGMEM.
It provides direct x,y access to get and set pixels along with a render method Blt() that will copy bits to the NeoPixelBus or other Raster objects.
The important part of creating one is to match the "ColorFeature" between your NeoPixelBus and the buffer since the buffer will store/reference data in the same format as the bus.
Here you can see that they both are defined with NeoGrbwFeature.
NeoPixelBus<NeoGrbwFeature, Neo800KbpsMethod> strip(PixelCount, PixelPin);
NeoBuffer<NeoBufferMethod<NeoGrbwFeature>> image(myImageWidth, myImageHeight, NULL);
There are two template "Method" objects that used in the construction of the NeoBuffer that define where the data for the image is stored. There is no need to call these objects directly and they should just be used to construct the two raster objects.
- NeoBufferMethod - This method object will define the raster object so that it uses RAM to store the image.
- NeoBufferProgmemMethod - This method object will define the raster object so that it uses PROGMEM to store the image.
If you plan to dynamically modify the image through source code, then you have to define it to use RAM by using NeoBufferMethod . You can still initialize the RAM stored image from PROGMEM.
If you plan to just consume the image and never modify it, your best approach is to use PROGMEM by using NeoBufferProgmemMethod as it will store in flash like source code and save precious RAM for your sketch.
// stored in RAM, but initialized by the "myImage" from PROGMEM
NeoBuffer<NeoBuffermMethod<NeoGrbwFeature>> image(myImageWidth, myImageHeight, myImage);
// stored in PROGMEM
NeoBuffer<NeoBufferProgmemMethod<NeoGrbwFeature>> image(myImageWidth, myImageHeight, myImage);
To render the buffer image you use one of the BLT (block transfer) methods. They will copy the image data from the buffer to the target.
The BLT methods all take a generic target for the first parameter as a NeoBufferContext<>. This buffer context is a mechanism to allow for different types of targets to be used. You can pass the NeoPixelBus directly for this. You may also pass any of the other Raster objects that allow writing to.
If you have both an image that is one pixel high, and you have just a string of pixels on your NeoPixelBus, and you want speed; then you can use Blt(dest, destIndex) method. This method is limited but is very quick.
If your image has more than one row or column of pixels and/or your pixels are arranged in a matrix, then can use Blt(dest, xDest, yDest, xSrc, ySrc, wSrc, hSrc, layoutMap). This is the most common way to render.
The complexity with this method is that the target buffer doesn't explicitly know how you have them laid out. You maybe using NeoTopology or maybe NeoMosaic. To solve this you need to pass a callback as the "layoutMap" that will be called so you can provide the mapping. Inside this you can provide your custom layout mapping or you can use the MapProbe() method of any of the matrix panel support objects like NeoTopology, NeoTiles, or NeoMosaic.
If you are providing your own mapping; you must return PixelIndex_OutOfBounds if the coordinate is outside the range you support.
NeoTopology<RowMajorLayout> topo(16,16);
uint16_t LayoutMap(int16_t x, int16_t y)
{
return topo.MapProbe(x, y);
}
Once of you have defined this callback, then you can use to render parts of the image.
image.Blt(strip, 0, 0, 4, 4, 3, 4, LayoutMap);
This will render the image from location (0,0) to the strip at location (4,4) consisting of a rectangle of pixels that is 3 pixels wide and 4 pixels tall.