Point optimisation: how to

[colouredmirrorball]

If you know how to read format 0/1, format 4/5 should not be hard to reverse engineer.

[james]

Uhhhhh....

An ILDA file ends with the extension .ild, so anyone looking for what that is will find that it came from ILDA.

That's recognition.

Giving credit where credit is due costs nothing.

That still fits well within the definition of "volunteered" information.

James.

[james]

If you know how to read format 0/1, format 4/5 should not be hard to reverse engineer.

True. But figuring out that the RGB data is actually BGR is a little weird.

That part doesn't really match the RGB data found in the palette section 2.

Plus, you would have to know that there is such a thing as format 4 & 5 to even know to look for it.

Fortunately 4 & 5 use the same 32 byte section header as 0, 1 & 2. That would give you a clue.

James.

[james]

Getting back to points optimization.....

One of the features I put into LaseBoy is a function that "explodes" a frame.

This finds every individual lit segment in a frame and makes a set of frames with one lit segment in each frame.

You can also "collapse" a set of selected frames into one frame.

If you are using this feature manually in the running application you can take a frame apart into its individual lit segments, reorder the resulting frame set and glue the pieces back together in whatever order you want.

Once you have a frame in the order you want you can convert all the blank to lit black so that optimizing the frame will not change the order of the lit segments.

It is also useful if you are automatically looking for equivalent vectors.

Another function in LB "fractures" all lit segments into individual vectors, by placing a blank vertex at the end of every lit vector.

When you explode that, you get a frame set with each frame containing only one lit vector.

Then you can count through all the frames comparing them for congruencey.

You have to look at the lit vectors from both ends, because lit vectors can be stroked in opposite directions and still be the same vector.

Yet another function reverses the direction or stroke of a lit segment. So I compare two segments first as-is and then again with one of them reversed.

If I find two frames with the same lit vector in them, I only use one of them and discard the other.

I have found this is particularly useful for cleaning up DXF imports.

This is some of the stuff I am reviewing and improving right now.

There are other uses for exploding a frame into a frame set as well.

Some of the effects that work on individual lit segments use this function.

James.

[Laser Wizardry]

" 15 characters"

[mixedgas]

Digital scanner amps have been shipping for years in marking ,medical, and drilling systems. I know one show company that has over two dozen of them.

They, and the marking and drilling companies still use point based laser art.



Steve

[james]

Steve, Cool! How much?

How do you communicate with them?

Do you know how they work?

Are they digital all the way through or are they just an analog system with a very smart controller / DAC?

LW, what kind of deflection angle?

James.

[Laser Wizardry]

" 15 characters"

[james]

What effect does that have on the beam diameter / quality?

[colouredmirrorball]

Digital scanner amps have been shipping for years in marking ,medical, and drilling systems. I know one show company that has over two dozen of them.

They, and the marking and drilling companies still use point based laser art.



Steve

But I take it the points don't require optimisation? As the driver would correct the images in real time.