Point oriented vs Vector oriented

[soforene]

I've made a few frames for a show but the point count was a tad high (I've read that 500 is the ceiling figure to aim for).
By converting them from point to vector oriented it has reduced the point count (which is nice) and the graphics still look the same (on the preview screen at least).
Is there any trade off in doing this?
i.e. are point oriented frames "better"?

A Noob question also popped into my head just now.
If we all have spiffing 30K scanners in our projectors why is the best figure for frames approx. 500?
Why not 30K?
(I know I should know this after all these years but I blame alzheimers).

[Xytrell]

30,000 points scanned at 30,000 points per second gives you a frame rate of 1fps. That's incredibly low. 500 points gives higher frame rate and therefore minimal flicker.

[JohnYayas]

I've made a few frames for a show but the point count was a tad high (I've read that 500 is the ceiling figure to aim for).
By converting them from point to vector oriented it has reduced the point count (which is nice) and the graphics still look the same (on the preview screen at least).
Is there any trade off in doing this?
i.e. are point oriented frames "better"?

A Noob question also popped into my head just now.
If we all have spiffing 30K scanners in our projectors why is the best figure for frames approx. 500?
Why not 30K?
(I know I should know this after all these years but I blame alzheimers).

Every scanner out there is capable of scanning 1,000,000 pps as long as the points are spaced very close together. Is there any difference scanning a 10" line with 10 points at 10 points per second or using only 2 points at 2 points per second? If motion was linear the answer would be no. But, galvos are sort of like kids in a sports car, then try to go as fast as possible between the stop lights regardless of how far apart they are. Lets say you have a line that is projected to be 10 foot long. If there are only two points, one at each end, the galvo will go balls to the wall from one end to the other. The line will probably not be as bright as you like since it is being draw so fast. Throw in 10 or so points and the line will brighten up because it slows down the galvos. Add 1000 points and the galvos will slow down to the point where they pause at each point and instead of a nice line, you will see a line with a bunch of bright spots.

A point oriented frame is properly drawn when there are exactly enough points so that the line is nice and bright but without any hot spots. Too fast and you lose laser intensity. Too slow and it looks dotted. A vector oriented line only defines the start and end. If the galvos drew it as presented it would appear faint. But, either the application software or DAC firmware (iShow is a vector base DAC by the way) will fill in the gaps with points to make it look bright. The smarter the point placement algorithm is, the better the image will look.

So, there really is no difference between point and vector oriented frames as far as the scanner amps are concerned. By the time it gets thes the data it is point oriented.

Anyway, I jumped around with some concepts without really answering the question but if you think about what I have said you will hopefully figure it out.

[taggalucci]

Nice explanation John Thanks for that. It's certainly confirmed a few things for me and given me a few things to consider in my own vector/point handling code.

[masterpj]

There are trade offs.
1. File size, yes vector oriented is smaller in general, except with round corners, they contain loads of data in vector orientated compared to point orientated. (because LD2000 doesn't have something called splines)
2. Vector oriented is setting dependant, when it can look identical on ur pc it can look different on an other projector because of possible variation in settings.
3. Despite it didn't affect your drawing in this case, it can sometimes influence the image after all. (not sure but i think out of my memory, i'll get back to this one)

Reason why the image became less flickery:
your point oriented image had an much higher point density and the vector oriented was on a lower one, you had simple straight lines which the conversion can properly convert to a lower density.

Point oriented frames are in a sense better, why you will off course ask:
- Point oriented frames give you more control over spacing in each part of the image, allowing you to in fact get an even better scanning image then when you would say draw something in vector only.
Though it costs some work yes.
- Point oriented frames give you the ability to use the very awsome point pulling method to get the best out of your projector scanners.
- You can control brightness in the image more, by adjusting the spacing in certain parts of the image, you can create dark and bright spots and that can give you extra drawing shading effect.
This is especially usefull on DPSS greens which modulate less great, offcourse recoloring will work too yes, but this extra ability is still good.
- Working with point oriented learns you best on what will scan properly and what now (because sometimes it's hard to make out or it is close to being scannable).
- Can give you sharper corners at parts where its needed and more efficient roundings on parts where sharp turns are not needed.

The reason why the point count meter drops:
It isn't the exact amount of points the scanners go's to, the spacing is set to a global setting that is set in your vector orientation settings in LD2000.

[soforene]

Cheers guys !!

[Zoof]

Pointplacement is the name of the game for frame creation.
When I created OLA (that converts vector oriented SVG to point oriented ILDA), the majority of time was spent on cracking the point placement nut. I believe that OLA achieves very good frames with minimal point count.

That kind of performance was needed to create Laseroids so that a larger number of asteroids in the filed would not introduce flicker. That race-car analogy holds up nicely. The algorithm for point placement in OLA actually computes the `braking' needed to make the next corner nicely and allows the galvos to accelrate out of the curve. A with max speeds set for curvature and straight lines, control is given over the brightness of the lines. I did not go as far as to modulate the beam brightness to the computed speed since I believe that varying brightness adds to the charm of laser graphics but that would be a matter of taste.

- feel free to try OLA, it is still a free download:
http://www.amelink.net/lasergl/