RGBSB Target modele...
 
Notifications
Clear all

RGBSB Target modeler

26 Posts
7 Users
0 Reactions
12.2 K Views
smeighan
Posts: 2595
Topic starter
(@smeighan)
Famed Member
Joined: 15 years ago
n8huntsman wrote:There is a tree style that has crossed my mind that can not be modeled. Not sure if anyone has ever tried to build one like this or not, but... I was thinking about a tree that was constructed with with rings. Make a ring at the base Ø, then another 6 inches or so above that, and repeat all the way to the top. The advantage is that your resolution remains constant throughout the entire tree. You would have to use quite a few guy wires made from something like fishing line so you would end up with somewhat of a polygon rather than circles but I think it might work.

You mean cannot be modeled with what I have so far. It would be easy to. Make the modeler support this, has anyone made a tree like this?

I was thinking if you had something to make a giant cone, then you could wrap the strings around the tree the way Christmas lights are normally strung. It seems harder to build.


Reply
Materdaddy
Posts: 2368
(@materdaddy)
Famed Member
Joined: 16 years ago

I plan on making a mega wreath this year. It will only be regular incandescent lights, but in the future, I might pixelize it.

In the future I might be making something that will have concentric circles that could be similar to the tree described by N8.

I might also want pixel rows like your existing setup, but with gaps in the pixel numberings for dead space, like this:


   1  2  3  4
  5  6    7  8
 9 10     11 12
13 14      15 16
 17 18    19 20
  21 22 23 24

(I apologize for the crude ASCII drawing)


Reply
smeighan
Posts: 2595
Topic starter
(@smeighan)
Famed Member
Joined: 15 years ago
Materdaddy wrote:I plan on making a mega wreath this year. It will only be regular incandescent lights, but in the future, I might pixelize it.

In the future I might be making something that will have concentric circles that could be similar to the tree described by N8.

I might also want pixel rows like your existing setup, but with gaps in the pixel numberings for dead space, like this:


   1  2  3  4
  5  6    7  8
 9 10     11 12
13 14      15 16
 17 18    19 20
  21 22 23 24

I am thinking of supporting google sketch up as an input to my target modeler. Or maybe other vector driven files.

You can always feed me a file of dimensions


Reply
n8huntsman
Posts: 562
(@n8huntsman)
Honorable Member
Joined: 16 years ago

So if I send you an .ipt or .dwg, that's cool?


Reply
smeighan
Posts: 2595
Topic starter
(@smeighan)
Famed Member
Joined: 15 years ago

Hi;

"So if I send you an .ipt or .dwg, that's cool?"

it will be when i write the parser. i use google docs and have thought of expanding the places i can get a source file. i was thinking these three sources for target models.

1) rgbsb builds the target model
2) User provides and ascii file , they do all the work
3) take google docs, dwg files .etc.

Today, (1) is the only choice. I said i would look into google docs after i get some major milestones done

1) Build targets using rgnsb, Mostly Done. Need to allow user to delete models they dont want and to edit an existing model
2) Allow users to create effects into their library, In Progress
3) Allow user to select from a pull down of effects from their library and also select a target, this week
4) Produce an animated gof so user can see what that effect will look like on their virtual tree by next week
5) produce xml files for sequencers, 2-8 weeks
6) Create models for other rgb targets , maybe by june
7) New effects, movies, snowstorms, fireworks i will do this as a break from (5)
8) Write a user manual, before june
9) Look to using a google docs file as a target summer, fall

I am working on 2 & 3 now. I hope within the week people will be trying effect son their trees and seeingg the result on the web page.

I expect #5 to take 2-8 weeks.,

I want to be done with 1-5 by june.

thanks


Reply
smeighan
Posts: 2595
Topic starter
(@smeighan)
Famed Member
Joined: 15 years ago

Released new target builder code Feb 27

This release now has

1) editing (yeah!). Click on a model name and it will load the form, edit and save. Still dont have the ability for you to delete models.
2) a unit of measure (inches or cm)
3) There is a summary at the bottom for the form showing what sequencers the users want to support.
4) Changed fields to light blue in form. I had a hard time seeing where i should type.
5) Fixed bug where "How will you build your tree" = UP/DOWN 180 was not working.
6) modified wording of this prompt.
old: **Total number of Pixels on this string
new: **Total number of Pixels on this string (Note, this is NOT the total pixels on your tree, just the count on one string):
Some people had inputted the total pixel count of their tree. Sorry, if it was not clear.

The hyperlink for editing was not complete and caused people to have a blank user name. I believe this is now fixed since editing should work.

thanks


Reply
smeighan
Posts: 2595
Topic starter
(@smeighan)
Famed Member
Joined: 15 years ago

Hi all;

All of the original mega tree targets on the web page were 32 strands of lights, each strand had 35 pixels.
http://meighan.net/rgbsb/index.html

i am getting ready to release the next phase of rgbsb.

Phase 1) user can create a target modeling their tree. DONE
Phase 2) User will select from the effects classes (Spirals, Bands, text, Pictures, Meteors, .etc.) , customize the effect and save it in their private library. IN PROGRESS. Probably will be finished within next 7 days.
Phase 3) User select a target model and and effect and produces an animated GIF. IN PROGRESS, This is 75% complete, it will come one oine within next 14 days. This will be a big deal for the users. With this release you can try hundreds of effects on your tree and get a visual idea of what different effects look like. I think this ability to simulate your tree will be popular.
Phase 4) User now chooses to output their data into an xml file for the 3 different sequencers. 2-12 weeks.

I will also start writing a user manual when Phase 3 comes online.

So to give you an idea of what this might look like, i built a different target model. Instead of 32 strands x 35 pixels, here are effects projected onto a 48 strand by 60 pixel tree.

6 spirals, 100ms delay, only show front half of tree

s6 t1 48 60 100ms

The pictures, as expected, project with better clarity.

IMG 7965 head

picture 48 60 50ms
14578

14578 48 60

thanks


Reply
smeighan
Posts: 2595
Topic starter
(@smeighan)
Famed Member
Joined: 15 years ago

and a few more

grinch 48 60 50ms
snoopy 48 60 100ms

And text looks better

text 48 60

thanks


Reply
smeighan
Posts: 2595
Topic starter
(@smeighan)
Famed Member
Joined: 15 years ago
Steve Gase wrote:
Question/suggestion:

Should "pixel size" be introduced into your model?

With the current visualizations I see these nicely done pictures projected onto a tree and I think "that's nice!" but I suspect when this is applied to my tree that the images may not be as recognizable. My reason for suggesting this? When I look at the nutcrackers I see one tree with a pixel size that is just as big as the gap between that pixel and the next. in another nutcracker, the is very little gap. i reality the smart pixels are about .39in... and the gap between pixels is 2.61in.

In a different element -- like a coroflake, the pixels sit behind corrugated plastic, and there is no "gap"... instead light from one pixel blends into the area of the pixels around it.

there are other coro examples -- like david moore's candy canes -- where pixels are segregated into separate compartments so that there is little blending/bleeding.

...in the coro examples we might define a pixel as a much larger visible area... maybe 3" circles or 3" squares.

yep, never any lack of tasks. We get one idea, it opens 3 more.

The answer is yes to all of your questions. They can be modeled more and more accurately. as Phase 3 of rgbsb (The ability to have your tree display your effects on an animated gif) comes online, I think people will want to use the tool to do some initial sizing. We probably want these gif images to be as close to real life as we can get.

So in gnuplot you can specify how big the dots are drawn (in world coordinates).

So i removed the background skeleton that i draw in gray. This was so people could visualize the tree.

Here are 4 images drawn in different dot sizes. All have 50ms frame rates

0.3" dots

point 0 3 50ms

0.6"

point 0 6 50ms

1.0"

point 1 0 50ms

let gnuplot pick the default dot size, this has been what i have been drawing with

point 1 0 50ms

I dont think i can draw them exactly to size, because a gif image is not glowing like a light. maybe 0.6" is closest to real life?


Reply
smeighan
Posts: 2595
Topic starter
(@smeighan)
Famed Member
Joined: 15 years ago

New release of target modeler. Many people were getting some of the prompts wrong.

I have added logic to check ranges on a number of fields. I build a calculated tree given that you told me how long one string is. Your string is the hypotenuse of the triangle. If you are doing "UP/DOWN NEXT" or "UP/DOWN 180", i cut the hypotenuse in half.

Still, Alpha. Tool may break, data may be lost.
I did need to reset at least half of the models. I am backing up the user and model table every few days.

so ...

Please try to create 1 or 2 models.

I have also removed these 3 prompts
1)orientation (Where is the front of the tree . I will add this back in later this summer. really not needed for now)
2) direction of rotation of strings , CW or CCW. Again this really does not affect us getting a model out. It will only affect if you want text scrolling right to left or left to right.
3) Type of string, CCR or smart string. again , i may need to add it back this summer, but it does not affect generating sequences

The secret to getting the model tool to work is to get the number of pixels per string and the distance between the first and last string correct. You can now let the tool calculate some of the other dimensions for you.

Being RED on the tool is like the tool saying "I don't believe you but i will build the model that way". Your dimensions will always be used, not my suggestions.

If people could give some feedback on my ranges

1) Number of strings you will build a megatree from (1-100)
2) Number of pixels on a string (40-130)
3) Spacing between pixels (2-6" or 5-30cm)

thanks

member index screen shot

Reply
Page 2 / 3
Share: