Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[ktroy]

Fun little game so far! Just wondering if there will be a terrain detail improvement at some point or if it is final. It is impressive and fun as it is, but compared to google earth it is lacking the sharp cliffs and proper elevation.

Below is a screenshot of the exact same location for comparison. It was taken in the Southeast USA, which may or may not even be ready.
Thanks for all your hard work on this game!


Google Earth




Anteworld

[Spudly]

I think Google uses a higher resolution data set, which Outerra could use, but it would be like 500gb.  Also, if you change your field of view in your graphics settings, it will make them look taller and sharper.

[cameni]

It would be best to complement it with a real world photo. Time of day is important as well. Google does several things - exaggerating heights because without the detail and in wide FOV they look small, sharper lighting, lack of ambient skylight .. plus it uses 30m data, while OT has 90m.

[ZeosPantera]

How dare it. We should just sue them. Them and their 30m dataset.

[ktroy]

Google may do a lot of things, but whatever they do it matches the real world almost perfectly. I'm talking terrain height map here not textures or lighting.

 I actually do have a real world picture on hand of that spot, from a different location farther down the canyon. I can retake the in game picture of that spot but I am not on my computer now.

so, the 30m data is the culprit? It would be cool if you could download 30m in areas that you plan on spending a lot of time in Anteworld. Obviously not worldwide though.

[ZeosPantera]

It would be cool if you could download 30m in areas that you plan on spending a lot of time in Anteworld. Obviously not worldwide though.

Well if everyone is going to use the streaming download of terrain why not just do 30m. OR have it do the 90M and then the 30M if bandwidth is an issue.

I would just download the full terrain dataset regardless. I think 100 Gigs is worth it for the entire earth.

[ktroy]

I agree 100%. Make it an option for 90m if bandwidth is an issue. A world at 30m would be unbelievable.
Any way this could happen?

[ktroy]

Also, where does Outera get it's information of where trees and rocks go. Sometimes it's very accurate, having a patch of rocks on a mountain side, where I know there is in real life. Other times its not accurate at all.
Interesting.

[cameni]

I agree 100%. Make it an option for 90m if bandwidth is an issue. A world at 30m would be unbelievable.
Any way this could happen?

Yes. The way of patience
A few more runs of satellite scans and maybe the global 30m ASTER dataset will become usable.

Also, where does Outera get it's information of where trees and rocks go. Sometimes it's very accurate, having a patch of rocks on a mountain side, where I know there is in real life. Other times its not accurate at all.
Interesting.

Mathematical probabilistic model, only taking the topology into account. Later it will use available data for forests to adjust the probabilities.

[foxfiles]

Yes. The way of patience
A few more runs of satellite scans and maybe the global 30m ASTER dataset will become usable.

30m ASTER does not cover the whole Earth, especially near the poles, does it?
I see that it stops at 83° North and 83° South?

[cameni]

Given that SRTM provides only up to 60° and had to be combined with other sources, ASTER is pretty good. But it contains lots of noise, that can be removed by averaging a couple of scans, a process when the noise gets cancelled out and the signal stands out.

[foxfiles]

Ok I see...  quality problems : errors, holes, artefacts...

[foxfiles]

Dear Cameni, we could write an algorithm (another one  ) for OT need,
that adjusts the missing ASTER gDEM gaps seamlessly 
with data from the SRTM XSAR. Since both data sets are very similar 
(i.e. 16bit, signed integer, and accuracies), it is the best possible 
way to get a seamless and no gapped surface topography or DTM...

It is said that it will still take the ASTER science community around 
4-5 years to fully process the ASTER data from stereo paired imagery, 
so the above methodology is what we all have to work with...

Or we find a kind of partnership with European Space Agency to get "for free"
their 30m Elevation datas (or even better resolution for some parts of the world, SvK, LU? to play a bit) 

[cameni]

The algorithm exists already

OT tool that processes terrain data works with layers of source data, in the order as they are registered. If it can't find the data in the first layer, it continues looking for it in further layers. For example, you could register a 30m tileset, followed by a 90m and 1km bathymetry layers (the last two are used now). It's a hybrid CPU-GPU algorithm and it is very fast.

However, the problem with ASTER data is not with the holes and missing data. Problem are the artifacts on the available data. Various worm-like structures in valleys and on flatter areas. That cannot be repaired.

[foxfiles]

Wahooo, you had already referred me to this link, but I had not catched your ability to work with multi terrain grids and layers" !
so let's go to ESA to "steal" them full 30m, and 10m DTM on an HD drive...