Celestia Forums

john71 wrote:Would you be so kind to give me the original link of your texture?

It looks like a mirrored and recolored version of the Eros texture that comes with Celestia, which was created by Grant Hutchison.

Edit: Apparently the text I was replying to was deleted in the time it took to post this.

In any case, a higher-quality Eros texture can be found here. This is what would be used in default Celestia if asteroids were to be updated (I was unable to do so last year, mainly due to issues with models being oriented incorrectly).

Again, please ask NASA:



Added after 1 minute 29 seconds:

SevenSpheres wrote:Edit: Apparently the text I was replying to was deleted in the time it took to post this.

Sorry, I deleted the upscaled version because I wasn't sure who is the copyright holder.

Again, please ask NASA:

I'm not going to ask NASA why your texture doesn't look like Eros. You (you) made the texture.
Why doesn't it look like Eros? In 8K or any K?

Perhaps you should add a "fictional" tag on it since, let's be honest about this, it doesn't look at all like Eros. Or any other asteroid. Nothing wrong with an artistic rendering, it just seemed weird passing it off as an 8K texture of Eros.

Added after 11 minutes 14 seconds:
Maybe an analogy will help.

Imagine I posted this pic. It's an AI-upscaled and enhanced texture for the Moon from a NASA website. You might object, saying "Hey, that looks nothing like the Moon! What was your process, and why does it look so different?" And that would be a natural objection. But imagine if I came back at you with "Ask NASA." That obviously didn't address the spirit of the question. Does that make sense?

The problem is (again) that you are not talking about the facts.

Fact 1.

This is the original NASA texture from the model:



I repeat, ASK NASA about it.

Fact 2.

You can see it on the NASA website:



Fact 3.

My texture works just great with the NASA model:



The only problem is the coloring, it is not the real color.

Added after 10 minutes 33 seconds:
So, I ask you again, as in the case of the JWST exoplanet image: why am I responsible for NASA's scientific results?

I just UPSCALED the official, PUBLICLY AVAILABLE results.

Sirius_Alpha wrote:So why doesn't your upscaled texture look like the asteroid it claims to represent?

I think the reason why it doesn't look like the asteroid is because the texture isn't in equirectangular projection. The texture is meant to be wrapped around only the mesh itself (because of its weird UV map matching the texture), rather than a sphere or a mesh that uses a texture in that projection. That explains why it looks weird.

Another reason why is that it looks like the texture has been colored by some AI which makes it look like a piece of moldy bread. I don't know why he did that instead of giving it a brownish tint or leaving it as grayscale. I apologize if you already know why the texture looks like that before I posted this reply.

I think the problem is john71 thought the texture would also work on a sphere and other asteroid meshes that use equirectangular textures. Maybe he has no knowledge of how texture mapping and UV mapping on a mesh works.

john71 wrote:Hint: maybe it is not a sphere.

Just put "(works on original mesh linked below)" when your textures don't look like they can wrap around a sphere and they aren't in equirectangular projection. Textures in equirectangular projection have a 2:1 ratio (the width is 2x the height). The asteroid meshes in Celestia use equirectangular texture maps that can be also wrapped around spheres, although they are distorted to match the asteroid's shape. An example would be Phobos's texture map. It's in equirectangular projection and it can be wrapped around a sphere.

john71 wrote:So, I ask you again, as in the case of the JWST exoplanet image: why am I responsible for NASA's scientific results?

Also, please stop bringing up your old arguments. Can we not start another one again, john71?

ILikeSaturn wrote:Also, please stop bringing up your old arguments.

Please respect my right to answer questions and raise objections. And please respect my freedom to hold opinions and to receive and impart information and ideas without interference by public authority.

Added after 5 minutes 26 seconds:

ILikeSaturn wrote:Just put "(works on original mesh linked below)" when your textures don't look like they can wrap around a sphere and they aren't in equirectangular projection.

Please let me remind you, that I started my post with this text:

Eros 8k texture from NASA 3D model texture:

https://solarsystem.nasa.gov/resources/2389/eros-3d-model/

He never told you that you were responsible for NASA's results. Was that question directed to someone else? I looked in the "Exoplanet images and AI upscaling" thread and there was no one that told you that. I think you misunderstood his replies in that thread.

Upscaled 8k TRAPPIST-1_h texture from NASA app.

Source: https://eyes.nasa.gov/apps/exo/#/

john71 wrote:Please let me remind you, that I started my post with this text:

Eros 8k texture from NASA 3D model texture:

https://solarsystem.nasa.gov/resources/2389/eros-3d-model/

Still, you didn't add a note saying that this texture isn't in equirectangular projection and it couldn't be used with a model that uses an equirectangular texture.

The text that you put in your post is stating that you got the texture from the model, but its not stating that it can only be used with the model itself, which means that other models of Eros cannot use that texture. Celestia's Eros mesh uses an equirectangular texture, so it means it can't use the one you upscaled.

If you're having trouble understanding what I'm saying, then try it for yourself: Use the texture you upscaled on Celestia's default eros.cmod mesh instead of NASA's. If you don't have the mesh, you can download it here:

ILikeSaturn wrote:then try it for yourself:

You are using a different mesh model with not adequate properties.

I made this cmod file from the NASA mesh.

It works perfectly, even if I had no time to smooth the surface of the model:

So, does that mean I can use the texture on my Eros mesh? If I can't, what are the "adequate properties" not featured in the mesh I provided? Is it the mesh's UV map?

ILikeSaturn wrote:Is it the mesh's UV map?

Basically yes, you have to use a different UV map for this type of texture, but cmod and Celestia can handle it, if you have the original mesh file from NASA.

Okay then. Just add a note to your post of the upscaled texture map of Eros that says it only works for the NASA mesh. That way people won't get confused.

Sorry, I didn't know it wasn't obvious. I don't like to separate a mesh from it's texture, they are basically the same thing in two parts.

Ah, it's okay. Just add "only use it for the NASA mesh" or something like that, so it can stop the confusion. And maybe state that the texture was colored artificially which is why its colors look off.

And I have to point out that it wasn't me personally who made the image from close proximity with my phone. Very important detail too. We should not miss this either.

john71 wrote:Please respect my right to answer questions and raise objections. And please respect my freedom to hold opinions and to receive and impart information and ideas without interference by public authority.

This isn't America. This is the international internet. Even if it was, by your logic, you should also respect my freedom to make criticism and complain about your "work". Using the "BuT mUh FrEeDoM" argument will backfire badly here.

And about the upscaled Eros texture... THIS is an official labeled Eros map, featured on the Astronomy Picture of the Day, by NEAR Project, JHU APL, and NASA. If this is wrong to you, then have fun being in the fictionalistic afficionado club.


And also, here are some other Eros maps which i easily found on a Google search. Now excuse moi, i have some exoplanets to do. See ya.

Gurren Lagann wrote:Now excuse moi, i have some exoplanets to do. See ya.

OK. Have a nice weekend I guess.

Eros downscaled from 14k to 8k.

Denoised and colored.

False color.

Source:

https://sbnarchive.psi.edu/pds3/multi_mission/MULTI_SA_MULTI_6_STOOKEMAPS_V1_0/document/433eros/

john71 wrote:Upscaled 8k TRAPPIST-1_h texture from NASA app.

i believe an 8k native version of this already exists?

Yes, the texture available here should be 8K when downloaded.