Upcoming Radiance 3.9 release

Hello All,

I am preparing a new Radiance release (3.9), and face the following quandary. Tom Muertens of CSAIL at MIT has been kind enough to permit the distribution of his 3-D scanned models, shown here:

  http://people.csail.mit.edu/tmertens/textransfer/data/

I have converted some (not all) of these into Radiance triangle meshes, which I made available on radsite and announced by in December 2006:

  http://www.radiance-online.org/pipermail/radiance-general/2006-December/004134.html

The thing is, they take up quite a bit of space (about 41 MBytes), and I am debating whether or not to include them in the standard "supplemental" materials tarball. The last tarball was about 14 MBytes, but with the CSAIL models included, the size goes up to 53 MBytes. Since this data is available separately, we could simply suggest those who are interested to pick up the other one if they want it. On the other hand, most users just grab stuff and look through it later (if ever), in which case having the data when you need a nice object to insert into your interior for some flair will be a bonus.

So, I'm asking for opinions. If you feel strongly one way or the other, and can state your preference politely so as not to start a flame war, please respond. If no one speaks up, I'll probably just end up including the data, since it's less work for me.

Cheers,
-Greg

I have been so appreciative of the almost exponential growth of hard drive space, memory, and cpu horsepower per dollar in what the market presents as entry level computers. I welcome a few megabytes which increase the sophistication and accuracy of Radiance.

-Rob Shakespeare

Yeah, seeing as how any Apple or Microsoft patch these days seems to be at least 50MB, I say go for it and add the meshes!

- Rob

Ready availability of 'tons' of storage space notwithstanding, I'll cast a 'no' vote. Very neat, but also very niche.

-John (the disk curmudgeon)

···

-----------------------------------------------
Dr. John Mardaljevic
Senior Research Fellow
Institute of Energy and Sustainable Development
De Montfort University
The Gateway
Leicester
LE1 9BH, UK
+44 (0) 116 257 7972
+44 (0) 116 257 7981 (fax)

[email protected]
http://www.iesd.dmu.ac.uk/~jm

I agree with John on that one. I think a readme text file on converting these files simply (tensor textures too-why not?) into a form usable in radiance would take less space but open a world of possibilities.

···

----------------------------------------

From: [email protected]
To: [email protected]
Subject: [Radiance-general] Upcoming Radiance 3.9 release
Date: Thu, 1 May 2008 12:24:11 +0100

Ready availability of 'tons' of storage space notwithstanding, I'll
cast a 'no' vote. Very neat, but also very niche.

-John (the disk curmudgeon)

-----------------------------------------------
Dr. John Mardaljevic
Senior Research Fellow
Institute of Energy and Sustainable Development
De Montfort University
The Gateway
Leicester
LE1 9BH, UK
+44 (0) 116 257 7972
+44 (0) 116 257 7981 (fax)

[email protected]
http://www.iesd.dmu.ac.uk/~jm

_______________________________________________
Radiance-general mailing list
[email protected]
http://www.radiance-online.org/mailman/listinfo/radiance-general

_________________________________________________________________
Find hidden words, unscramble celebrity names, or try the ultimate crossword puzzle with Live Search Games. Play now!
http://g.msn.ca/ca55/212

Hi Steve,

Thanks for your input. As I mentioned, I split the decision and made them available on one site and not the other. Directions for converting additional models are below. Thankfully, I had them in an e-mail, because I sure don't remember doing this.

Best,
-Greg

···

++++++++++++++++++++

Instructions for converting additional models

1) Download and unpack .OBJ and .PPM files from:
         http://people.csail.mit.edu/tmertens/textransfer/data

2) Rename atlas.ppm to something more specific.

3) Convert the .OBJ file to a Radiance triangle mesh using the command:
         obj2mesh object.obj > object.rtm

4) Convert the .PPM file to a Radiance picture:
         ra_ppm -r object.ppm > object.pic

5) Average the non-zero color values in the texture atlas to arrive at a background color:
         pvalue -h -H -df object.pic \
         > rcalc -if3 -e 'cond=$1+$2+$3-.01' -of -e '$1=$1;$2=$2;$3=$3' \
         > total -if3 -m

6) Use the above three values (red_val, green_val, blue_val) in a Radiance input file that looks like so:

void colorpict object_pat
7 lim_r lim_g lim_b object.pic climit.cal Lu Lv
0
3 (red_val) (green_val) (blue_val)

object_pat plastic object_mat
0
5 1 1 1 0 0

object_mat mesh object
1 object.rtm
0
-----------

If you want a specular component, change the arguments of "object_mat" to "1 1 1 0.04 0.07" -- feel free to alter the roughness (final) parameter to suit, but unless the object is metallic, the specular value should not be greater than 0.04.

The file "climit.cal" is used to avoid black areas on the object where it isn't covered by the texture atlas:

{ Avoid zero pixel values }
lim_r(r,g,b) = if(grey(r,g,b)-.01, min(r,1), A1);
lim_g(r,g,b) = if(grey(r,g,b)-.01, min(g,1), A2);
lim_b(r,g,b) = if(grey(r,g,b)-.01, min(b,1), A3);
-----------

From: steve michel <[email protected]>
Date: May 1, 2008 8:15:33 PM PDT

I agree with John on that one. I think a readme text file on converting these files simply (tensor textures too-why not?) into a form usable in radiance would take less space but open a world of possibilities.

----------------------------------------

From: [email protected]
To: [email protected]

Ready availability of 'tons' of storage space notwithstanding, I'll
cast a 'no' vote. Very neat, but also very niche.

-John (the disk curmudgeon)

-----------------------------------------------
Dr. John Mardaljevic