Old Pictures from Celestia (locked)

[BobHegwood]

I got some strange comments about my Avatar system from people and I want to answer here because I can't add my answer in the add-on site.

Milos,
Thanks very much for doing this here. The ML comments section is really
no place for people to learn/argue about/discuss an add-on.
Methinks that we really need to be able to direct people here in order to argue the relative
merits of add-ons.

Again, many thanks...
- Brain-Dead Bob

[Dollan]

PlutonianEmpire,

Ok, let me rephrase that. Earth like meaning a terrestrial about the size of Earth, not necessarily exactly Earth like. The term Earth like is getting used to much as far as I am concerned. I think it may be time to start bringing in terms like those used here. http://www.onewest.net/~dollan/ARCpclindex.html

Oh, that is somewhat out of date. The Onewest site has long since been dropped, but for whatever reason, I cannot get the company to remove my old site. The most recent version of my PCL is here: http://arcbuilder.home.bresnan.net/PCLMaster.html although it needs a LOT of work.

While we really have no idea what the conditions in the early Centaurus system were like, and whether it is a totally proven fact that all of Earths water came from comets, as it is chemically different in everything I have seen and read, if the early conditions on any of the planets was at all like it was in our own system, then we may find something similar to Earth. But personally I doubt we will find another Earth right next door. We simply could not get that lucky. Most likely we are going to find a more like Mars analog there, either hot and dry or cold and dry. Or a few Venusian worlds.

See here for some of the most recent information regarding Alpha Centauri: http://www.solstation.com/stars/alp-cent3.htm Right now, I would say the odds are fifty-fifty, based on what we know thus far about planetary exploration, for there to be terrestrial worlds of any size present. But I'm hopeful, being something of an Alpha Centauri-phile!

I personally think James Cameron didn't do enough research. He could have found many star systems that could of housed his system. Good old HD 28185 would have been a great candidate, just a bit far away at 120LY away. I think he was worried by all the recent talk about faster than light travel and it may or may not be possible at least through standard space. I am sure there are other factors in there as well.

I agree wholeheartedly. heck, there are other systems far closer than that, even, which would have worked very well. 18 Scorpii, 37 Geminorum, Eta Cassiopeiae A, maybe (just maybe) even Tau Ceti.

By the way, what ever happened to your HD 28185 add-on? Now that I have a rockin' computer and can actually use those wonderful add-ons of yours, I can't find it anywhere.

...John (gone a long while but trying to make a comeback)...

[Dollan]

Milos,
Thanks very much for doing this here. The ML comments section is really
no place for people to learn/argue about/discuss an add-on.
Methinks that we really need to be able to direct people here in order to argue the relative
merits of add-ons.

Again, many thanks...
- Brain-Dead Bob

I don't even bother reading the comments anymore. People ignore the readme files, they don't understand the purpose of posting the systems (sometimes, they are posted as WIPs), and most post before they try to learn about the conditions in the system.

That said, I'm trying to get myself gathered back together here, so I can start working on my systems again. I have a much better system, and I have access to some nice software toys now. All that remains is to relearn what I've forgotten about coding in Celestia, as well as to learn all about what the latest version has to offer!

...John...

[Hungry4info]

Alpha Centauri is system where planets are impossible.

Normally people don't say things like that without a lot of conclusive backing evidence. From papers on the subject such as this, this, and this, the most you could say is that it is inconclusive.

Binary systems with planetary companions are known with closer stellar separations than Alpha Centauri.

Impossible is a word people use far too often.

[Fenerit]

.
.
.

That said, I'm trying to get myself gathered back together here, so I can start working on my systems again. I have a much better system, and I have access to some nice software toys now. All that remains is to relearn what I've forgotten about coding in Celestia, as well as to learn all about what the latest version has to offer!

...John...

Welcome back, John.

[Dollan]

Welcome back, John.

Thanks muchly!

Erm... do you happen to have the link handy for the latest version of Celestia? I need to reconstruct all of my bookmarks, too....

...John...

[Fenerit]

http://www.shatters.net/celestia/index.html

"basic" site for latest official version

http://www.shatters.net/forum/viewtopic.php?f=11&p=121769#p121769
Latest copiled SVN (exe only, so mind at shaders and so on...) version offered by Vincent

http://en.wikibooks.org/wiki/Celestia
Wikibooks Celestia's pages with the latest features addition

P.S.

Ops, for exe patch I've in mind Windows, for others OS I do not know. For the rest, explanation of features added are here and there in the forums (scripts, idea first of all).

[John Van Vliet]

on a different note ( 43 pages ?? this is getting long - maybe time for a new one??? )
the cleaned up LOLA and Kaguya normal map i have been working on
this is a merger of the two .After i destriped the LRO lola data
the first one is with the moon vt that i remapped and the second is with a 50% gray image





this level 0 to level 5 32 k normal will soon be uploaded.

[t00fri]

on a different note ( 43 pages ?? this is getting long - maybe time for a new one??? )
the cleaned up LOLA and Kaguya normal map i have been working on
this is a merger of the two .After i destriped the LRO lola data
the first one is with the moon vt that i remapped and the second is with a 50% gray image





this level 0 to level 5 32 k normal will soon be uploaded.

John,

a posteriori manipulations of ANY kind with normalmaps do need a renormalization of the gradients, such that r^2 + g^2 + b^2 =1 ! How did you do that after cleaning?? Normally ALL manipulations must be done BEFOREhand, i.e. at the 16 bit integer level and not afterwards.

One such possibility would be to import the Kaguya data and the LOLA .IMG/.LBL pair for the elevation data into ISIS3 cube format ('pds2isis), then use the ISIS3 highpass/lowpass/...merging routines for cleaning etc and then save the result as a cleaned 16 bit elevation map with 'isis2raw'. If you then run my nmtiles tool on the result it should be OK.

Fridger

[Milosz21]

Alpha Centauri is system where planets are impossible.

Normally people don't say things like that without a lot of conclusive backing evidence. From papers on the subject such as this, this, and this, the most you could say is that it is inconclusive.

Binary systems with planetary companions are known with closer stellar separations than Alpha Centauri.

Impossible is a word people use far too often.

Thanks, I’m not normal because I have different opinion, you must be the person who knows everything, right? Your source where you get information’s is some forum where anyone can write what he want. So I didn’t call this good source.
Someone on this forum saying that Alpha Centauri planets may be too dry if there exist, for me planets there are impossible. Gravitational effects of the ingredients (stars) makes the dust in the system does not connect with each other to form something bigger, I thinking about something like our asteroid belt between Mars and Jupiter, Jupiter makes asteroids in belt never connect in the planet.

[Hungry4info]

Thanks, I’m not normal because I have different opinion, you must be the person who knows everything, right? Your source where you get information’s is some forum where anyone can write what he want. So I didn’t call this good source.

My point was that no one knows if planet formation there is impossible. All we've got is computer simulations and planet formation models. Some say it's easy, some say it's hard.

If you actually visited the links I showed you, you would find that it's not a forum where anyone can post what they want.

[John Van Vliet]

When the lola data came out i already had a version of the kaguya 5760x2880 data "bumped up" to 32k
using two programs to do it
"resynthesizer " to add "missing" details to it ( noise based on the rest of the image)
then GMIC to also "bump it up " by running it's PDE to solve for the missing pixels
[img=http://img41.imagevenue.com/loc499/th_86017_Screenshot_122_499lo.jpg]

----- for the LOLA image ---
then used GMIC to remove the top 5 % of the hi noise pixels in the 16 bit image
it works MUCH better than ISIS
the PDE used for this dose leave an interesting pattern in the image ( but it is at about 0 to 100 value out of 65538 tones) - looks like "Celtic knot-work"
extracted this and changed the tone for viewing

Screenshot.jpg

then added some ( 4 %) back in to sharpen the less noisy image

do to the fact that the kaguya image was worked on as a 24 bit rgb the 16 bit needed some help
this is where a "multiresolution enhancement" comes in to help
shrink the 32k to ( 16k,8k,4k,2k,)
a 2k enlarged to 4k overlayed on a 4k
that 4k bumped to 8 overlayed on a 8k ....
and so on

then the two images added together and divided by two ( done with the math functions in NIP2 )

[t00fri]

When the lola data came out i already had a version of the kaguya 5760x2880 data "bumped up" to 32k
using two programs to do it
"resynthesizer " to add "missing" details to it ( noise based on the rest of the image)
then GMIC to also "bump it up " by running it's PDE to solve for the missing pixels
[img=http://img41.imagevenue.com/loc499/th_86017_Screenshot_122_499lo.jpg]

----- for the LOLA image ---
then used GMIC to remove the top 5 % of the hi noise pixels in the 16 bit image
it works MUCH better than ISIS
the PDE used for this dose leave an interesting pattern in the image ( but it is at about 0 to 100 value out of 65538 tones) - looks like "Celtic knot-work"

then added some ( 4 %) back in to sharpen the less noisy image
( i dont have a copy of this at the moment )
do to the fact that the kaguya image was worked on as a 24 bit rgb the 16 bit needed some help
this is where a "multiresolution enhancement" comes in to help
shrink the 32k to ( 16k,8k,4k,2k,)
a 2k enlarged to 4k overlayed on a 4k
that 4k bumped to 8 overlayed on a 8k ....
and so on

then the two images added together and divided by two ( done with the math functions in NIP2 )

John,

this sounds like black magic! One can "hardly" gain physical detail by "bumping up" a "tiny" 5.6k texture to 32k. Yet I am aware that certain sharpening algorithms work by adding in a certain amount of noise...

You forgot to mention whether you did all these manipulations BEFORE the normalmap was generated or AFTER. Since you mentioned 24 bit rgb, it rather seems that you manipulated the normalmaps NOT the initial binary elevation graymaps. If true, how did you then do the required normalization r^2+g^2+b^2=1 AFTER your normalmap manipulations??

I never had any complaints using the ISIS high/lowpass tools. They do what they are made for.

Fridger

[bh]

Blahblah...

Saturnrise from Titan... still can't get over the atmospheric effects!

[John Van Vliet]

before the normal map was made

the 24bit kaguya image was because i was working in Gimp
i then converted it back to 16 bit

this sounds like black magic!

this is artistic in-painting
orig LALT_GGT_MAP.cub


bumped up



over the years i have liked Greyc ( now Gmic) the heat flow PDE dose a nice job

once i have a 16 bit image i then rescale it down to the sizes i need ( 16k,8k,4k,2k) and make normal maps from those . resizing a normal map dose NOT work

I never had any complaints using the ISIS high/lowpass tools. They do what they are made for.

Gmic just dose it differently than isis or PhotoShop dose
http://cimg.sourceforge.net/
-- pde_TschumperleDeriche2d , greycstoration ---
http://gmic.sourceforge.net/gimp.shtml

[Dollan]

. From papers on the subject such as this, this, and this

...Your source where you get information’s is some forum where anyone can write what he want. So I didn’t call this good source...

What forums? These are peered reviewed papers, and well worth the read, if one is willing to bother.

By the way, Hungry4info... thanks for those links! I'm reconstructing my archives, and these papers were nice to add to them!

...John...

[t00fri]

before the normal map was made

the 24bit kaguya image was because i was working in Gimp
i then converted it back to 16 bit

But sorry, once you truncate to 3x8 bit RGB the elevation detail that was previously contained in 16 bit Kaguya data is definitely lost. Also if you formally re-convert to 16 bit later, you cannot get the detail back. GIMP is unable to convert to 16 bit and to work at 16 bit level. So once GIMP is added into the pipeline, the all-important 16 bit precision is irretrievably lost. That was one major motivation why I wrote my nmtools...

Also I don't understand your motivation for "bumping up" smaller textures by such large amounts. This does NOT increase the information content in the picture whatsoever, but instead may introduce incorrect extrapolation errors. While interpolation procedures (down-scaling) are reasonably safe, extrapolation (up-scaling) is NOT.

I am really confused about your procedure...

Fridger

[John Van Vliet]

GIMP is unable to convert to 16 bit and to work at 16 bit level

true but nip or cinepaint ( gui), or the cmd line vip ,gmic ,gdal and imagemagic can.

[t00fri]

GIMP is unable to convert to 16 bit and to work at 16 bit level

true but nip or cinepaint ( gui), or the cmd line vip ,gmic ,gdal and imagemagic can.

Sure, but above you stated that you used GIMP for Kaguya, which inevitably destroys the 16 bit precision at elevation level. That's why I don't understand ...

Cinepaint is an old attempt to make GIMP 16bit proof, but it essentially failed (lack of interest...). ImageMagick is full of bugs and doesn't work well for large files (for obvious reasons given it's memory management). I would never trust it. Gdal is terribly slow for larger files. I wonder why you prefer to use such poorly tested and/or badly performing tools instead of ISIS3 and/or my tools?

Fridger

[Hungry4info]

By the way, Hungry4info... thanks for those links! I'm reconstructing my archives, and these papers were nice to add to them!

Any time