Loading ...
Sorry, an error occurred while loading the content.

RE: [cc2-dev-l] XP development

Expand Messages
  • L. Lee Saunders
    Actually if we worked together on REDN, maybe we could generate a tutorial or two. Sound good? Lee _____ From: cc2-dev-l@yahoogroups.com
    Message 1 of 7 , Oct 12, 2006
    • 0 Attachment
      Actually if we worked together on REDN, maybe we could generate a tutorial or two.  Sound good?
       
      Lee


      From: cc2-dev-l@yahoogroups.com [mailto:cc2-dev-l@yahoogroups.com] On Behalf Of bidmaron
      Sent: Wednesday, October 11, 2006 7:07 PM
      To: cc2-dev-l@yahoogroups.com
      Subject: [cc2-dev-l] XP development

      Lee, as you know, I've done all your tutorials on XPs, and I'd love to try my hand at doing
      something useful and modifying your previous XPs as an exercise.
      --Dale--

    • Charles Sykora
      Sure! The thing I d like to tackle someday is to write an XP that fractilizes a path and then eliminates any crossovers that might occur. As for the joint
      Message 2 of 7 , Oct 12, 2006
      • 0 Attachment
        Sure!  The thing I'd like to tackle someday is to write an XP that fractilizes a path and then eliminates any crossovers that might occur.

        As for the joint project, in case you misplaced my direct e-mail, it is charlesdsykora (at) mac (dot) com.

        --Dale--

        On Oct 12, 2006, at 10:34 AM, L. Lee Saunders wrote:


        Actually if we worked together on REDN, maybe we could generate a tutorial or two.  Sound good?
         
        Lee


        From: cc2-dev-l@yahoogroups.com [mailto:cc2-dev-l@yahoogroups.com] On Behalf Of bidmaron
        Sent: Wednesday, October 11, 2006 7:07 PM
        To: cc2-dev-l@yahoogroups.com
        Subject: [cc2-dev-l] XP development

        Lee, as you know, I've done all your tutorials on XPs, and I'd love to try my hand at doing
        something useful and modifying your previous XPs as an exercise.
        --Dale--



      • Simon Rogers
        Roof shading for CD3 works in a similar way to varicolor PNGs symbols, that is, two pngs are stacked on top of one another. With varicolor pngs, the top png
        Message 3 of 7 , Oct 13, 2006
        • 0 Attachment
          Roof shading for CD3 works in a similar way to varicolor PNGs symbols, that is, two pngs are stacked on top of one another. With varicolor pngs, the top png appears in the current color, with whatever transparency that png has. For png building symbols, we are going to take a similar approach. The underneath png will be the building, the top png will mark the areas' downward slope. Each roof section will have its downward slope direction marked with a color. So, if a roof section was facing down towards the north, it would be 90 degrees, if the south, 270 degrees.
           
          My question is, how should I map this angle to a color? Is there a standard way of doing this. I want to use rgb, possibly greyscale. An example would be R value is hundreds, G value is tens and units, B value first two decimals. Ideally, such a scheme would mean that it wouldn't be impossible to do a dome or cone using a paint program and some gradient fills.
           
          Please let me know if this makes any sense at all, and if you have a suggestion as to the mapping, I'd like to hear it.
        • Joseph Slayton
          The traditional way to encode angle information is via a surface normal (a unit XYZ vector) encoded into the RGB channels of an image, with the image being
          Message 4 of 7 , Oct 13, 2006
          • 0 Attachment
            The traditional way to encode angle information is via a surface normal (a
            unit XYZ vector) encoded into the RGB channels of an image, with the image
            being referred to as a "normal map". The dot product of the normalized light
            vector direction and the pixel from the normal map at the location gives the
            lighting intensity at that point. These calculations only work in something
            called tangent space, however. If you transform the light direction by the
            rotation matrix for your object then the images will be in tangent space (or
            maybe it's the transpose of the rotation matrix, I forget right now). There
            are many discussions of this sort of operation out there on the web (key
            phrases "normal map", "tangent space", and/or "pixel shader" should give the
            required references).

            The fun part about normal maps is that if you use single colors for the
            whole image then you can encode a single direction for each polygon, but if
            you use a normal map of varying colors then you can encode things like
            shingles, roughness, and so on into the image. There are filters available
            for Photoshop and the like that make it easy to construct normal maps (see
            the http://developer.nvidia.com and search for PhotoShop Filters).

            The hardest part about this implementation is that you need to pass the
            rotation matrix for your entities to the renderer, but that's another issue
            entirely.

            Joe Slayton


            >From: "Simon Rogers" <simonwork@...>
            >Reply-To: cc2-dev-l@yahoogroups.com
            >To: <cc2-dev-l@yahoogroups.com>
            >Subject: [cc2-dev-l] Roof angle
            >Date: Fri, 13 Oct 2006 15:40:40 +0100
            >
            >Roof shading for CD3 works in a similar way to varicolor PNGs symbols, that
            >is, two pngs are stacked on top of one another. With varicolor pngs, the
            >top
            >png appears in the current color, with whatever transparency that png has.
            >For png building symbols, we are going to take a similar approach. The
            >underneath png will be the building, the top png will mark the areas'
            >downward slope. Each roof section will have its downward slope direction
            >marked with a color. So, if a roof section was facing down towards the
            >north, it would be 90 degrees, if the south, 270 degrees.
            >
            >My question is, how should I map this angle to a color? Is there a standard
            >way of doing this. I want to use rgb, possibly greyscale. An example would
            >be R value is hundreds, G value is tens and units, B value first two
            >decimals. Ideally, such a scheme would mean that it wouldn't be impossible
            >to do a dome or cone using a paint program and some gradient fills.
            >
            >Please let me know if this makes any sense at all, and if you have a
            >suggestion as to the mapping, I'd like to hear it.
            >--
            >Simon Rogers
            >http://www.profantasy.com <http://www.profantasy.com/>
            >Livejournal - http://sjrlj.notlong.com <http://sjrlj.notlong.com/>
            >
            >
          • Simon Rogers
            Thank you. We were considering using a fixed roof angle for simplicity. I (or some other lucky person) needs to go through every png symbol and add the
            Message 5 of 7 , Oct 16, 2006
            • 0 Attachment
              Thank you.
              We were considering using a fixed roof angle for simplicity. I (or some
              other lucky person) needs to go through every png symbol and add the
              information to each one, so I need a fairly straightforward way of doing it.
              Maybe the filters will make this more straightforward. The option to add
              textures is tempting, though, even if we don't specify everything. I'll look
              into it.

              Simon Rogers
              http://www.profantasy.com <http://www.profantasy.com/>
              Livejournal - http://sjrlj.notlong.com <http://sjrlj.notlong.com/>

              The traditional way to encode angle information is via a surface normal (a
              unit XYZ vector) encoded into the RGB channels of an image, with the image
              being referred to as a "normal map". The dot product of the normalized light

              vector direction and the pixel from the normal map at the location gives the

              lighting intensity at that point. These calculations only work in something
              called tangent space, however. If you transform the light direction by the
              rotation matrix for your object then the images will be in tangent space (or

              maybe it's the transpose of the rotation matrix, I forget right now). There
              are many discussions of this sort of operation out there on the web (key
              phrases "normal map", "tangent space", and/or "pixel shader" should give the

              required references).

              The fun part about normal maps is that if you use single colors for the
              whole image then you can encode a single direction for each polygon, but if
              you use a normal map of varying colors then you can encode things like
              shingles, roughness, and so on into the image. There are filters available
              for Photoshop and the like that make it easy to construct normal maps (see
              the http://developer.nvidia.com <http://developer.nvidia.com> and search
              for PhotoShop Filters).

              The hardest part about this implementation is that you need to pass the
              rotation matrix for your entities to the renderer, but that's another issue
              entirely.

              Joe Slayton
            • Joseph Slayton
              Silly me. I said tangent space. The normal maps for this sort of thing should be in model space because we don t have the surface normal available, just the
              Message 6 of 7 , Oct 17, 2006
              • 0 Attachment
                Silly me. I said tangent space. The normal maps for this sort of thing
                should be in model space because we don't have the surface normal available,
                just the normal map. The same principles apply but you need a normal map
                with encided base direction for each of your basic directions (N, NE, E, SE,
                S, SW, W, NW).

                Joe Slayton

                >From: "Simon Rogers" <simonwork@...>
                >Reply-To: cc2-dev-l@yahoogroups.com
                >To: <cc2-dev-l@yahoogroups.com>
                >Subject: RE: [cc2-dev-l] Roof angle
                >Date: Mon, 16 Oct 2006 11:31:39 +0100
                >
                >Thank you.
                >We were considering using a fixed roof angle for simplicity. I (or some
                >other lucky person) needs to go through every png symbol and add the
                >information to each one, so I need a fairly straightforward way of doing
                >it.
                >Maybe the filters will make this more straightforward. The option to add
                >textures is tempting, though, even if we don't specify everything. I'll
                >look
                >into it.
                >
                >Simon Rogers
                >http://www.profantasy.com <http://www.profantasy.com/>
                >Livejournal - http://sjrlj.notlong.com <http://sjrlj.notlong.com/>
                >
                >The traditional way to encode angle information is via a surface normal (a
                >unit XYZ vector) encoded into the RGB channels of an image, with the image
                >being referred to as a "normal map". The dot product of the normalized
                >light
                >
                >vector direction and the pixel from the normal map at the location gives
                >the
                >
                >lighting intensity at that point. These calculations only work in something
                >called tangent space, however. If you transform the light direction by the
                >rotation matrix for your object then the images will be in tangent space
                >(or
                >
                >maybe it's the transpose of the rotation matrix, I forget right now). There
                >are many discussions of this sort of operation out there on the web (key
                >phrases "normal map", "tangent space", and/or "pixel shader" should give
                >the
                >
                >required references).
                >
                >The fun part about normal maps is that if you use single colors for the
                >whole image then you can encode a single direction for each polygon, but if
                >you use a normal map of varying colors then you can encode things like
                >shingles, roughness, and so on into the image. There are filters available
                >for Photoshop and the like that make it easy to construct normal maps (see
                >the http://developer.nvidia.com <http://developer.nvidia.com> and search
                >for PhotoShop Filters).
                >
                >The hardest part about this implementation is that you need to pass the
                >rotation matrix for your entities to the renderer, but that's another issue
                >entirely.
                >
                >Joe Slayton
                >
                >
                >
                >
                >To Post a message, send it to: cc2-dev-l@...
                >To Unsubscribe, send a blank message to: cc2-dev-l-unsubscribe@...
                >Yahoo! Groups Links
                >
                >
                >
                >
              Your message has been successfully submitted and would be delivered to recipients shortly.