Working with images..file size & jobtime

I’ve been looking at software like GIMP for image touchup, greyscale, colour/b&w croping and file compression. I’m wondering if these programs just mask over the unwanted area’s and the data remains unseen.

I thought that ‘cropping’ and a little ‘File compression’ before lasering would reduce the jobtime.

Is using ‘Trace’ in LB to crop a background, a better option?

Also, if I take an image can I select an area within the overall image and make it a layer, Then can I apply sub layers and choose settings for those that are different from the layer settings and the overall image settings?

Also, is an algorithm a particular set of settings…ie for a particular material and image appearance.

I would try ImagR

An “image” is rectangular array of pixels, with each pixel having a specific color. Ignoring some complexity, that color has three components: red, green, and blue.

LightBurn converts each pixel’s color into its grayscale equivalent, so each pixel has a single number representing the gray value: 0 = black and 255 = white. All of the pixels in the original image remain present, so the image size does not change.

Graphic editors alter the color of each pixel, but the rectangular image remains the same size. When you “mask” part of the image in a graphic editor, it changes the color of those pixels, but the pixels remain present.

Cropping an image in a graphic editor reduces the number of pixels by discarding those outside the crop rectangle. That’s different than masking an image, because the cropped image has fewer pixels in its rectangle.

File compression has no effect on the number of pixels, because compression only reduces the size of the file when it’s stored on disk. When the file is read back for further editing, it returns to the same rectangular size.

JPG compression (or any lossy compression) may slightly change some of the pixel colors when the file is decompressed.

Trace produces vectors at the borders of pixel regions with the same grayscale value. The result is basically a set of lines along the outlines of those regions, with no reference to the grayscale values. If the region looked like an island, the traced vectors will form a closed path.

You can do multiple tracings of the same image with different thresholds and put the resulting vectors on different layers.

You can trace the image to produce the vector path equivalent of the regions. Those vectors will be in a specific LightBurn layer with all its settings. You can apply all the usual vector operations to remove various parts of the vectors.

All of the pixels of the original image remain on a different layer (the Image layer) and will be processed by those layer settings. You can mask that image by suppressing some of the pixels, but the layer settings apply to whatever remains un-masked.

The sublayers define successive processing steps for all of the vectors on that layer; they do not apply to specific regions within that layer.

The vector layer remains separate from the image layer, with any sublayers applying only to the vectors or image on their main layers.

An algorithm is a specific way of solving a problem and is unrelated to the layer settings that define how the laser will behave as it travels across the material.

You may apply an algorithm to decide which settings to use on a particular layer: “This image will look best using Halftone dithering with 50% power at 1000 mm/min” or “These vectors should get 25% power at 100 mm/min”.

Keeping all the concepts and terminology straight definitely poses a challenge!

2 Likes

A nice simple and well-structured explanation as always.

Thank’s Ed.
Very clear on all points esp ‘sub layers’
I was cross referencing LBf topics, Doc’s and in this case ImageR tutorials concerning colour/bw conversion,
greyscale, considering how to delete as much information as possible before transferring to next process.
I considered that a blank background may still be considered as information and duly processed…esp if its value is close to…but not blank…So rather than croping out a background or masking, rather have only data of required subject. But think your stating that there will still be a ‘rectangle/perimiter/ and area of & around’ regardless…and therefore a value which is data and has to be processed.

Lot’s of interesting points in your reply…So thank you for taking the time to do so.

Back in the day, “Paint” was the Windows graphic editor and that’s still a good name for all of them: you’re choosing the color of each dot in a rectangular frame.

Conversely, 2D CAD programs dealt directly with shapes: a circle was just the location (coordinate) of its center and its side (radius or diameter). Unless you did something special, a circle didn’t have an interior that could be colored.

If you start with a Painted image of a circle, a Trace tool tries to lay a circular shape around its edge; you may be able to simplify the result into an actual circle, once again defined by its location and size.

If that circle is on a Fill layer, LightBurn will treat its border as a boundary to be filled in with dots at interval defined in the layer settings. In effect, LightBurn will paint that circle on the material by having the laser draw each dot as it travels back & forth.

And then your head explodes … :grin:

1 Like

This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.