Problem Orbiter 2016 Hires Earth Textures

[Urwumpe]

He outlined his ideas here.

However, I don't think that the texture/heightmap story alone is what 4throck was about. You can hardly call that single format change "many things".

Yes. I also can't understand it really. My biggest problem so far is the other software, not Orbiter. I am fed up with having to manually update visual studio solution files in SSU.

 

[Loru]

What do you mean by "integrated"?

Face pointed out already to my "request". Basically GUI tool that can extract textures and heightmaps into .msh format and then reintagrate them after editing backinto orbiter.

I could fairly quickly addapt my bases to new standard however I physically don't have time to developing new work pipeline with available tools.

 

[martins]

Face pointed out already to my "request". Basically GUI tool that can extract textures and heightmaps into .msh format and then reintagrate them after editing backinto orbiter.

I am not quite sure I understand the rationale of using a mesh file as an intermediate step. While it is always possible to convert a surface elevation patch into a mesh, it is not generally possible to convert a mesh into a surface patch, since it has many more degrees of freedom. You could probably implement a least-squares approximation, but why not avoid all those problems and edit the patch directly? Why is it easier to edit a mesh than a grid?

Also note that elevation patches cannot be considered in isolation. They have overlapping neighbours, and ancestors and descendants in the resolution sequence. So any editing always has to take into account the quadtree connections of the patch. Exporting/importing a mesh would get extremely complicated then.

 

[Face]

I am not quite sure I understand the rationale of using a mesh file as an intermediate step. While it is always possible to convert a surface elevation patch into a mesh, it is not generally possible to convert a mesh into a surface patch, since it has many more degrees of freedom. You could probably implement a least-squares approximation, but why not avoid all those problems and edit the patch directly? Why is it easier to edit a mesh than a grid?

Also note that elevation patches cannot be considered in isolation. They have overlapping neighbours, and ancestors and descendants in the resolution sequence. So any editing always has to take into account the quadtree connections of the patch. Exporting/importing a mesh would get extremely complicated then.

IIRC, the idea is to not reinvent the wheel by means of implementing various tools ("painting" lines, filling areas, sculpting, etc.) into an appropriate editor that immediately works on the format, but instead use already available tools like 3D editors. That way people can stick to their favorite tool instead of having to learn yet another 3D editor and its particular user interface.

That idea is used in the ele2png workflow of OT3 as well: get the elevation tile as PNG, edit with whatever editor you want, get it converted back to elevation tile, integrate it back into the tree.
But I agree that the 3D mesh inverted conversion will be much harder to get correct than the heightmap one.

 

[martins]

OK, I do understand the idea mapping the elevation grids to a standard image format so it can be edited inside a general image editor. And surely anybody familiar with a 3D editor shouldn't have problems with a 2D editor.

Only caveat - the image format should support 16-bit channels. Does PNG do that? If not, TIFF may be a better solution. Maybe even GeoTIFF, so that the images can be georeferenced. You could then even use standard GIS tools such as gdal to stitch the images together into a mosaic, without having to arrange the locations manually.

 

[Face]

Only caveat - the image format should support 16-bit channels. Does PNG do that?

PNG does that. OT3's ele2png also supports a LUT that maps the full 16-bit to 8-bit color, though, because many free editors don't (fully) support 16-bit color spaces.

 

[martins]

Ok, sounds like we are pretty much there already. I guess the only thing missing from Loru's wishlist is a way to display the heightmap as a 3D grid, with the surface texture painted on top of it so that a connection between surface features and elevations can be made.

Matlab can do that, but not many people have access to it. I'm pretty sure that octave and python/matplotlib can do it too. Maybe also ParaView, and any number of other data processing/visualisation tools. It would just require somebody to do a survey and come up with a recipe/tutorial for an elevation editing pipeline.

 

[Loru]

I am not quite sure I understand the rationale of using a mesh file as an intermediate step. While it is always possible to convert a surface elevation patch into a mesh, it is not generally possible to convert a mesh into a surface patch, since it has many more degrees of freedom. You could probably implement a least-squares approximation, but why not avoid all those problems and edit the patch directly? Why is it easier to edit a mesh than a grid?

My idea is to to work on terrain inside 3d editor to perfectly match terrain with static base meshes. TBH for me it's irrelevant if I get 3d terrain be exportedg via extarnal tool or having plugin for my 3d editor that can deal with elevation files directly. Problem with heightmaps, is that they're not accurate enough (in my experience) to make seamless fit in vertical direction.

As for degress of freedom I think any user is reasonable enough to don't change vertex coordinates in lat / long plane, only in elevation plane. That solves degrees of freedom problem and no aproximation is needed at all.

Also note that elevation patches cannot be considered in isolation. They have overlapping neighbours, and ancestors and descendants in the resolution sequence. So any editing always has to take into account the quadtree connections of the patch. Exporting/importing a mesh would get extremely complicated then.

I'm well aware of this. I'm not talking about major "planet shaping" but local one and it's easy to edit 3 or 4 levels elevation maps to get desirable results.

OK, I do understand the idea mapping the elevation grids to a standard image format so it can be edited inside a general image editor. And surely anybody familiar with a 3D editor shouldn't have problems with a 2D editor.

I work with 2d editors all the time (both bitmap and vector ones) and it's not an issue.

Issue is vertical precission in this process:

elv file > heightmap (greyscale or color bitmap) > 3d editor > heightmap (greyscale or color bitmap) > elv file.


I can adjust boundaries and overlaps easily in 3d editor but what I crave for is a tool that'd give me abillity to preciselly merge terrain based on elevation tiles with custom base meshes. Then we can have cave / underground bases, or very steep rock features blended perfectly with the planet.

That's my 2 cents on it and TBH it's too much off topic IMO. It should be discussed in add-on section.

---------- Post added at 11:21 PM ---------- Previous post was at 11:17 PM ----------

Ok, sounds like we are pretty much there already. I guess the only thing missing from Loru's wishlist is a way to display the heightmap as a 3D grid, with the surface texture painted on top of it so that a connection between surface features and elevations can be made.

Matlab can do that, but not many people have access to it. I'm pretty sure that octave and python/matplotlib can do it too. Maybe also ParaView, and any number of other data processing/visualisation tools. It would just require somebody to do a survey and come up with a recipe/tutorial for an elevation editing pipeline.

I think the best pipeline would be tool to grab elv file and export it to msh and then back after edit. No need for aproximation, just put vertices in lat/long plane in correct positions (I know they differ depending on lattitude) and then reimport them back with new vertical data. Reason for that is msh is simple and well understood and most 3d editors already have export import tools to work with them.

 

[ARMSTRONG]

WOW thanks a lot for the whole answers , and I already know that Dr. Martin Schweiger made Orbiter, when I said that... well I was thinking in how this community made Orbiter something big with the addons too.

I never thought that Dr. Schweiger himself could answer to me, that was a nice experience and I wish you fun while you are resting a little from Orbiter ^^

Well, actually, I am gonna make a clean instalation for 2016, that Moon mountains and craters are too marvelous to keep myself in 2010.

But I will use the default Delta Glider instead of the Dan's Delta Glider IV.

About my landings... I will do the rentrys at day, however, this is a problem to me because I never accelerate the time with Orbiter, and this means that my landings will take more time until find the correct moment to land XD. By the way, I went to the Moon 6 times in real time with NASSP 6 so, trully I never use the T and the R keys and that means that Patience is my second name, therefore it wont be a problem if I need wait for updates. Even I am thinking to go Mars in real time even... (Well, I will use the T and R keys when learning how the hell can I plan a flight between planets, that for sure, but for the real mission I really could see myself doing the whole travel in real time, flying 4-8 hours per day, little by little like I say, I hope live 25 years more at less, and more or less a travel to Mars could take that, if I want to visit Venus. But, I will use 2016 to start a period of training to fly the Moon again but with the non-espherical gravity option activated. At this point I went to the Moon with this thing desactivated :3, so its a good moment, I guess, to start this training now with the default Delta Glider I guess. ^^

THANKS again, and have fun with Orbiter.

PS: If I could be as intelligent like you guys, I really could like develope good staff like you do, but I don't know nothing about programing unfortunately, so I can't fix my little problem by myself U-U. But no worries. Your answers were a good thing to see today. BYE.

 

[marcogavazzeni]

PNG does that. OT3's ele2png also supports a LUT that maps the full 16-bit to 8-bit color, though, because many free editors don't (fully) support 16-bit color spaces.

It would be interesting to be able to interface ele2png with L3DT.

http://www.bundysoft.com/L3DT/

 

[Face]

It would be interesting to be able to interface ele2png with L3DT.

http://www.bundysoft.com/L3DT/

Well, L3DT can generate PNG heightmaps, so nothing stops you from using it with ele2png. Just let it generate 259x259 tiles in 16-bit grayscale and you should be good.
I don't know if L3DT is capable of generating the 1x1 tile-overlay, though. If not, you'd have to do it yourself, perhaps by means of generating bigger heightmaps that you slice into overlapped Orbiter tiles with some graphics editor.

 

[ARMSTRONG]

Guys I have a problem U-U I am experiencing troubles when trying to install again the Orbiter 2016, my anti-virus Bit Defender blocked without my permission the file orbiter_ng.exe I just can't create it. And I can't do nothing to unlock this file... I tryed all... now I am waiting for the Bit Defender support answer... but maybe you already know how to fix this. HELP!


EDITED: NEVERMIND, I just needed to restart my computer XDDDD

 

[martins]

Note that orbiter_ng.exe is just a launcher for Modules\Server\orbiter.exe. It may be that certain picky virus checkers take exception at an executable launching another executable. If you can't coax your virus checker into accepting orbiter_ng.exe, you may be able to replace orbiter_ng.exe with a shortcut to Modules\Server\orbiter.exe, and setting the Orbiter root directory as the working directory in the shortcut.

If Windows allowed shortcuts with relative paths, I would use that as the default setup.

 

[4throck]

I usually setup my Antivirus to ignore my "games" folder :thumbup:
Otherwise it just makes loading slower.

It still catches any suspicious files on download / install.

 

[fatcat]

I just love the Orbiter program and community. I have a group of students who are learning orbital dynamics using Orbiter 2016 on macOS. It runs excellently using WineHQ.
Thank you Martin et all.

 

[OneTinSoldier]

Hello all,

I am having a problem with getting a proper download of the Optional high-resolution textures. When I download big files from one of the US Mirrors the file size is correct, but the MD5 Checksum is not. And so of course, the texture files don't work. I tried downloading with Internet Download Manager, Microsoft Edge(which hangs at 99% complete probably because file is corrupt or something is wrong with the server), and Google Chrome. Although Microsoft Edge doesn't finish the d/l, with IDM and Chrome I get the same incorrect(bad) MD5 Checksum on the downloaded file.

Again, the exact number of bytes of the downloaded file size is correct, so you might think you got a good file, but the MD5 Checksum is not correct.

Example using MD5 Checksum for Surf.tree:
68a18a75aa9c54f6e4361760e920457b: Is what I get
564ff394fda164976a854e3ab8a1e6f4: Is the correct MD5 Checksum

The smallest Earth file was fine(Elev_mod.tree), and the files for Mars were fine. The rest seem to be bad/corrupt in some way.

Just thought I should drop a note in case others might be wondering why the downloaded files are not working. Or it could be something on my end, but I don't know what it would be. I ran a filesystem check on my system as I do about every week or two. No issues were found.

 

[Ripley]

It's your connection's fault, rather than your pc.
Corrupt MD5s only means you have to redownload.

 

[OneTinSoldier]

Hello Ripley.

I download various kinds of stuff all the time and hardly have ever have any problem. I downloaded the 'Surf.tree' a couple of times, on different days. Came up with the very same incorrect Checksum both times. I can try again though. It will be the third try. If it doesn't have a valid MD5 Checksum, then I will give up on it for now and maybe try again in a week or two. I still have to wonder if the problem is not on my end, because I really never have any problem with downloading huge files. But I also still have to wonder if the problem is something on my end. I don't know.

---------- Post added at 06:53 PM ---------- Previous post was at 03:48 PM ----------

I just downloaded the "Mask.tree" which had a bad/incorrect MD5sum in my previous attempts to download it. It's still incorrect... 5049cd7c9bedF57Bcc29daa35b673a2e.



It'd be great and I'd really appreciate if someone would download that file and let me know what you get for the MD5sum on it.

Cheers

 

[martins]

I can confirm that the http download from mirror 2 (USA) appears to be corrupt. I haven't tried mirror 3.

I'll try to figure out what is going on. In the meantime, you can either resort to the torrent download, or use the orbiter-radio mirror (mirror 1) which is ok.

 

[OneTinSoldier]

I can confirm that the http download from mirror 2 (USA) appears to be corrupt. I haven't tried mirror 3.

I'll try to figure out what is going on. In the meantime, you can either resort to the torrent download, or use the orbiter-radio mirror (mirror 1) which is ok.

Hello martins,

Thank you so much for checking on them. And thank you for creating Orbiter! I donated $60 to you the other day when I downloaded it. Hope that helps.

Cheers