Jump to content

How Did They Accomplish That?!


Recommended Posts

Capture-Face.PNG

             That took a lot of patience to do.

                     32 - 8m x 256m strips

       sitting at the magic height of 300m or so

        with that picture ribbon sliced 32 times

              on the top surface of the strips

            The apparent resolution is 1x1m

                65536 individual color cubes

 

Kewl Beans!

Link to comment
Share on other sites

That sim is diagonal from The Corn Field, so it might not be done "the usual way".

I'd just note that Tyche's Grid Survey page for this sim has UUIDs for Objects and Terrain maps, neither of which currently show the picture, although those may be from the most recent survey, done a week ago. The map for that region that's currently returned by subnova, UUID="66e95ef0-02ae-7aa3-cdc6-a04b5ffc8f9d", looks quite different, just one of the standard-issue island raw terrains.

Link to comment
Share on other sites

You may be right, Qie. I seem to recall a notice about a piece of performance art that was being planned which involved 256 chairs being placed at the intersection of four sims, with coloured pieces of card underneath the chairs, so that when 256 avatar attendees sat down and, at a signal, used an AO built in to the chairs to hold up the pieces of coloured card, they formed the mosaic image of Troi.

Then, at another signal, they turned over the pieces of card to form another image:

Double-facepalm.jpg

Link to comment
Share on other sites


KarenMichelle Lane wrote:

            The apparent resolution is 1x1m

                65536 individual color cubes

Not so fast. Only get 15K prims per sim. IIRC, the map data averages color across all sides of the prim (or mesh), but if those are independently addressable (and they very well may be), then that resolution would be well within 8 * 15,000.

(FWIW, the vertical striations in the image do seem to me like artifacts of the "usual" object pixel approach.)

Link to comment
Share on other sites


Qie Niangao wrote:


KarenMichelle Lane wrote:

            The apparent resolution is 1x1m

                65536 individual color cubes

Not so fast. Only get 15K prims per sim. IIRC, the map data averages color across all sides of the prim (or mesh), but if those are independently addressable (and they very well may be), then that resolution would be well within 8 * 15,000.

(FWIW, the vertical striations in the image do seem to me like artifacts of the "usual" object pixel approach.)

Oh I agree with you. The 1m x 1m color cubes [based on the map image pixelization] I suggested would be laid out on each of the 8 - 256m long strips. I can't resolve from the picture what the actual number of supporting objects is nor the actual texture resolution on the surface of the platforms. I assumed that the vertical striations defined the width thus the 32 - 8m wide strips estimate. I can imagine that the texture on each of the supporting objects has no less than 8 - 1m x 1m color cubes each. We can go there and see it in person to settle this small quandary.

64 - 8m x 8m platforms would do it nicely with 64 color cubes per texture making the total of 65536 individual color cubes for that apparent image resolution.

I have done something less ambitious with 32m x 32m platforms in the past. 

 

Link to comment
Share on other sites


LlazarusLlong wrote:


KarenMichelle Lane wrote:


 We can go there and see it in person to settle this small quandary.
 

Forum trip!

[More like a flash mob actually...]

Alas, Aleph & The Corn Field remain locked away with no access to either that is not granted by the powers that be. The water surrounding each is non-navigable so no sending in the Marines either. The shroud of mystery shall forever remain.

We can Flash Mob somewhere else perhaps?

 

Link to comment
Share on other sites


Qie Niangao wrote:


KarenMichelle Lane wrote:

            The apparent resolution is 1x1m

                65536 individual color cubes

Not so fast. Only get 15K prims per sim. IIRC, the map data averages color across all sides of the prim (or mesh), but if those are independently addressable (and they very well may be), then that resolution would be well within 8 * 15,000.

(FWIW, the vertical striations in the image do seem to me like artifacts of the "usual" object pixel approach.)

Correction: You can get 30000 prims on a sim, if you link them in pairs and set them to convex hull.

Just a small nitpick. :)

Link to comment
Share on other sites


Jenni Darkwatch wrote:


Qie Niangao wrote:


KarenMichelle Lane wrote:

            The apparent resolution is 1x1m

                65536 individual color cubes

Not so fast. Only get 15K prims per sim. IIRC, the map data averages color across all sides of the prim (or mesh), but if those are independently addressable (and they very well may be), then that resolution would be well within 8 * 15,000.

(FWIW, the vertical striations in the image do seem to me like artifacts of the "usual" object pixel approach.)

Correction: You can get 30000 prims on a sim, if you link them in pairs and set them to convex hull.

Just a small nitpick.
:)

:D - Love it - Well do ya think they did this for this project? It's been my experience that LL Employees will usually try to do things the easy way with the least amount of steps. Then again that's my experience. Others mileage may  vary.

Link to comment
Share on other sites

  • 3 weeks later...

I'm a bit late to the party on this one, but I've just had chance to go in-world and look at ways to do do this. My best was a cut & hollowed prism, which give 5 faces visible per prim. Linked in sets of 8, they have an LI of 4 (no physics or convex hull). That gives 10 separate visible faces per LI, so you could do a whole sim at 1 pixel/sqm for 6554 LI. Well within budget.

I've no idea if a mesh could be built with more visible faces and a lower LI. The next question for me would be: did they colour each face manually or use a script to send the RGB values? 

Link to comment
Share on other sites

I'm not sure I've tried a 5-faces-up prim (there are at least a couple ways to do that, one with a prism, one with a twisted cube), but in my tests an 8-faces-up mesh seems to get colored as a single thing when shown on the map -- but I already forget whether it's the "average" of all the faces as one would get with llGetColor(ALL_FACES) or that of a single face.

Link to comment
Share on other sites


Kelli May wrote:

I'm a bit late to the party on this one, but I've just had chance to go in-world and look at ways to do do this. My best was a cut & hollowed prism, which give 5 faces visible per prim. Linked in sets of 8, they have an LI of 4 (no physics or convex hull). That gives 10 separate visible faces per LI, so you could do a whole sim at 1 pixel/sqm for 6554 LI. Well within budget.

I've no idea if a mesh could be built with more visible faces and a lower LI. The next question for me would be: did they colour each face manually or use a script to send the RGB values? 

Here's how I might do it.

I'd construct a 5 face prim as you described, dupilcate it and link them as a pair. That's 10 faces and LI=1. I'd drop a script in the root prim that receives an llGetStartParameter() from a rezzing script that encodes the X/Y location it should move to on the sim. llRezObject() doesn't reach far enough to allow rezzing across an entire sim, so the rezzed prims must move themselves to their target positions.

I'd drop that linkset into the contents of a rezzer prim along with a script that calls llRezObject() and passes the location parameters to each prim as they're rezzed. In this way I can rez a grid of objects across an arbitrary portion of a sim.

Once all the receiving linksets are rezzed, I can start llRegionSay()ing the image as messages that contain grid location and RGB values. Each of the rezzed receiving linksets will respond only to those messages matching its location.

That's a solution that requires three prims and two scripts.

If, as Qie states, the 5 face trick won't work, that doesn't affect my method of deploying the prims and the image data. The real question remains... how did they get that many apparent pixels?

Link to comment
Share on other sites

I tested the five-face prisms and they did get picked up by the map, albeit with some distortion.

prisms_001.png

The lines on the lettering were all one metre square single faces on flattened prism prims. The upright on the K has vanished somewhat in the distortion, but it shows that the map images read single faces more or less indiviudally. It's not like the Aleph image was flawless in that respect.

For reference, the grey-green splodge is a mesh island, and the red blur is a seaplane.

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

 Share

×
×
  • Create New...