I found bigben's "Angle from a reference vector" post on this forum and this setup was very helpful in achieving the nightside masking based on the Sun's position. Now I want to add the ionosphere (as suggested by user Arial DK in the same post), auroras, higher-res nightlight imagery (30k+), revamped ocean layer for better reflections, and TG clouds.
In regards to the cloud layer(s), I am still trying to figure out the best way to accomplish a very realistic look using 43200x21600 imagery I found (NASA bluemarble), with some sort of procedural input to supplement and/or custom painted cloud layers due to the lack of resolution.
It would also be great to find a better hi-res background star field (more than 16k). I had to "double-up" the one attached in the render to get something reasonable at this zoom level.
I would appreciate any input at this point as I am new to the forum and only 5-months into TG4, learning one day at a time.
This looks really good and I'm glad you figured it out. Might have to try fiddling with that setup.
Try my night sky and disable the Galaxy for a large 16k stars render. Probably could do a 360 skybox no prob. The latest setup has much larger stars that do not bunch up as much into streaks, only the occasional two stars close, which is still realistic.
I'll try and get a stripped version uploaded for you once I'm home, or try and export you a large res skymap.
Thank you WAS. I can PM you my .tgd in exchange.
Here is my background orientation and trying to setup a real Laplace's invariable plane or earth-centered ecliptic plane. If your solution is procedural based, I am super interested as in most cases that is all that is needed.
Hey, I can see my house from here!
Jokes aside, very cool! I don't know if it will work or help at all, but have you tried setting up luminance on the lights for the night area? They seem a bit underexposed to my eye (though it could be my crappy monitor) compared to the stars in the background. If you're using the sat maps for the texture, you could pull it into something like Photoshop and very quickly make a b+w mask for some kind of luminance layer. Don't know exactly how that would work in Terragen as I'm much more used to doing things like that in a program like Cinema 4D, but I'm almost positive it's possible. Maybe a masked surface layer? If you've already done that and I'm just falling victim to my practically ancient hardware, ignore and move on, ahaha!
So I woke up this morning intending to download a 15k starmap for you, but I entered my Terragen screen to find the command to render just chilling, never initiated. :P
Sooo, that's rendering. In the mean time here is the file. Maybe see how the background object works in your scene as a skybox.
The super stars seem to be too big, and their scale may need to be adjusted. But that ones a little a tricky due to the huge scale.
Here you can download excellent star maps for your own projects.
https://svs.gsfc.nasa.gov/3895
STORMLORD
Quote from: Stormlord on January 06, 2020, 04:00:27 PMHere you can download excellent star maps for your own projects.
https://svs.gsfc.nasa.gov/3895
STORMLORD
Those are nice, but I'd doctor them up. They're very noisy hubble images.
Here's a 15k export of the stars in 360. Wasn't exactly sure on resolution so I borrowed a 360 file from here and just locked the aspect ration and upped it. This doesn't have the super stars enabled so no super large stars.
https://drive.google.com/file/d/1-9iNIvi7nTyOP3ezSAoguN04-gTwVvPP/view?usp=sharing
A lot less noise from over exposure here compared to the NASA one.
Quote from: WAS on January 06, 2020, 04:45:47 PMQuote from: Stormlord on January 06, 2020, 04:00:27 PMHere you can download excellent star maps for your own projects.
https://svs.gsfc.nasa.gov/3895
STORMLORD
Those are nice, but I'd doctor them up. They're very noisy hubble images.
NOT NOISY!!! It's the Milky Way. The star charts have been created by rendering a star data base with almost 3 Million stars.
There is a natural star chart and another with the same stars, but rendered in galactic coordinates. So you have the plane of the Milky Way centered in the middle of the picture.
Based upon these maps, I created some nice animations and some own star charts which are more pleasant to see.
I wrote to the autor at NASA and requested the same images in a higher 32K resolution if possible.
He replied that would be possible, but he hadn't the time to do so. Maybe he would process his images at another time he mentioned to me.
Well... so we are here and still waiting... Or doing our own star map!
@ Jordan
Did you read my response via google mail, if not, please do so. Thank you very much for sharing your great galaxie and background star scene with me.
STORMLORD
Music Milky Way Animation Screenshot.jpg
Example: Screenshot of an animation with a star map
STORMLORD
Quote from: Stormlord on January 07, 2020, 09:46:36 AMQuote from: WAS on January 06, 2020, 04:45:47 PMQuote from: Stormlord on January 06, 2020, 04:00:27 PMHere you can download excellent star maps for your own projects.
https://svs.gsfc.nasa.gov/3895
STORMLORD
Those are nice, but I'd doctor them up. They're very noisy hubble images.
NOT NOISY!!! It's the Milky Way. The star charts have been created by rendering a star data base with almost 3 Million stars.
There is a natural star chart and another with the same stars, but rendered in galactic coordinates. So you have the plane of the Milky Way centered in the middle of the picture.
Based upon these maps, I created some nice animations and some own star charts which are more pleasant to see.
I wrote to the autor at NASA and requested the same images in a higher 32K resolution if possible.
He replied that would be possible, but he hadn't the time to do so. Maybe he would process his images at another time he mentioned to me.
Well... so we are here and still waiting... Or doing our own star map!
@ Jordan
Did you read my response via google mail, if not, please do so. Thank you very much for sharing your great galaxie and background star scene with me.
STORMLORD
They're noisy, too. You can play with levels and see all the encoded noise too, which can be brought out in render programs with lighting settings accidentally. They're based on imagers that inherently come with tons of noise, Hipparcos for example (not hubble) is defunct and no longer in use, and hasn't even been in use since 1993, which is their imager source, using Tycho 2 as their catalog mapper, or database of star points. Which is what they used to render the stars only, which is why the map is so faint, there isn't any actual difussion in it.
I'm fairly certain even in movies, they make their own skymaps in artistic style of, probably using catalogues, rather than technically low quality images from imagers. Downscaling a larger 32k image into a scene is only going to give the appearance of less noise do to down-sampling.
Also I had to search for your email, and found it in spam as "This message seems dangerous". I have no clue why. Maybe cause one of your URLs mentions FTP. That's a great animation though by and by. I do notice though, because there are so many stars creating just noise, the noise seems to "come alive" during panning, like it's fuzz or something.
Also, thanks for the source again. I'm not much into doing "the milky way" or home scenes as I am more into alien and hypothetical.
Stormlord and WAS, thank you for the links to star maps... I just found a 32k version here:
http://paulbourke.net/miscellaneous/astronomy/
Blattacker, I took your advice on better night area luminosity and actually found this link with high-res (500m, or 21600x21600x6) nightside RGB and grayscale imagery. I used the RGB in the color layer and added the grayscale as the luminosity layer as you suggested, and it does look way better and now interacts with any cloud layers above them.
https://earthobservatory.nasa.gov/features/NightLights
https://earthobservatory.nasa.gov/features/NightLights/page3.php
Is it possible to create vertical "curtains" in a cloud layer, in order to simulate auroras?
Please let me know what you think of my attached renders. I struggled today in creating the aurora and the ionosphere atmosphere model. I think they may be too high in altitude and I cannot seem to find a sweet spot in making them bright enough when viewing at the earth tangent (from camera POV). Am also struggling with the clouds v3 nodes.
As always, I would greatly appreciate any input.
Quote from: hydrodigger on January 12, 2020, 09:19:27 PMStormlord and WAS, thank you for the links to star maps... I just found a 32k version here:
http://paulbourke.net/miscellaneous/astronomy/
Blattacker, I took your advice on better night area luminosity and actually found this link with high-res (500m, or 21600x21600x6) nightside RCG and grayscale imagery. I used the RGB in the color layer and added the grayscale as the luminosity layer as you suggested, and it does look way better and now interacts with any cloud layers above them.
https://earthobservatory.nasa.gov/features/NightLights
https://earthobservatory.nasa.gov/features/NightLights/page3.php
Is it possible to create vertical "curtains" in a cloud layer, in order to simulate auroras?
Please let me know what you think of my attached renders. I struggled today in creating the aurora and the ionosphere atmosphere model. I think they may be too high in altitude and I cannot seem to find a sweet spot in making them bright enough when viewing at the earth tangent (from camera POV). Am also struggling with the clouds v3 nodes.
As always, I would greatly appreciate any input.
Thanks for the links
As for the clouds you'll probably have to try a inverted depth map (function at 1 instead of 0) with a surface layer with huge soft zone I'm altitude cutoff.
First thought.
Those look awesome already!
For aurora, try a vertically stretched noise function (e.g. Power Fractal with "Noise stretch" at 10 on Y axis and 1 on X and Z). That can be a good start, then warp it...
- Oshyan
Quote from: Oshyan on January 12, 2020, 09:42:08 PMThose look awesome already!
For aurora, try a vertically stretched noise function (e.g. Power Fractal with "Noise stretch" at 10 on Y axis and 1 on X and Z). That can be a good start, then warp it...
- Oshyan
I think he's trying to get the auroa layer like effect like in the image he provided, where it's solid with a hard falloff on top, and fades in lower altitude. I'd image a surface layer with a soft fade on huge scale with inverted depth might get around that area. If now a huge cloud v2 with a altude mask cutting off the cloud midway up a soft altitude surface layer.
I don't think a surface layer will do (unless you meant something else). What you need is a a very high cloud layer at very high altitude, with indeed a stretched Y component, a good warp, and some appropriate colors from an extra PF into direct light input (or final coverage). Plus a fade/fuzzy zone by altitude (distribution shader, or distance shader with a camera in earth's center), or another method.
Quote from: hydrodigger on January 12, 2020, 09:19:27 PMStormlord and WAS, thank you for the links to star maps... I just found a 32k version here:
http://paulbourke.net/miscellaneous/astronomy/
Thank you for the link, I got the map from Paul in 2012 and yes they're huge but I prefer the NASA Maps.
They have a 16K Version which is good for me and they provide the constellations ect. as well :-)
The only thing which they do not have is a star map with only the stars of the constelaltions.
But I made one myself.
If you have the computer power and if you're willing to programm your own star maps, then you can give it a try with "Processing 3.3".
Links are here...
Language
http://processing.org/
http://processing.org/learning/
Tycho Star catalogue
http://cdsarc.u-strasbg.fr/viz-bin/Cat?target=http&cat=I%2F259&
Star colours
http://www.vendian.org/mncharity/dir3/starcolor/
Wikipedia
http://en.wikipedia.org/wiki/Tycho-2_Catalogue
STORMLORD
I created a few maps in 2017 with Processing but ended up with a maximum size of 12288 x 6144.
Then my RAM was full... lol... (I still use my Intel Quadcore Pentium with 2.67Ghz with 7GB RAM)
2048 x 1024 - Tycho 2.0.jpg
Star Map created with Processiong 3.3 (Tycho)
2048 x 1024 - Tycho 2.0 (Galaktische Koordinaten).jpg
And the same with Galactical Coordinates.
STORMLORD
Luckily I found in my archive the Source Code which I have used to create these maps in Processing and further info's for you.
The files which you can download containing .exe files. These are a compilation of the original web sites. Just start them like a progamm, rest is self explaining.
But before you can use it, you have to download the star data catalogues and cobine them into one single file!
My original file has 501 MB (in which all single files has been combined into one big file). So it's too big to upload it for you, sorry...
Please see the links above and the attachments.
STORMLORD
Interesting. Thanks for sharing.
@hydrodigger welcome to the forum! and nice to see somebody working on this kind of scenes again ;D
Did you check this one? https://planetside.co.uk/forums/index.php/topic,20037.msg197502.html#msg197502
Is an oldy version of my setup for clouds, but it will be a good start if you don't want to waste too much time in the clouds lol
About the ionosphere, it should be a very tiny layer (like 10 mts depth), with a 90 km altitude. the tricky question is his luminance and density. i gonna search the file when i back to home later ;)
Quote from: Stormlord on January 15, 2020, 03:31:02 AMLuckily I found in my archive the Source Code which I have used to create these maps in Processing and further info's for you.
The files which you can download containing .exe files. These are a compilation of the original web sites. Just start them like a progamm, rest is self explaining.
But before you can use it, you have to download the star data catalogues and cobine them into one single file!
My original file has 501 MB (in which all single files has been combined into one big file). So it's too big to upload it for you, sorry...
Please see the links above and the attachments.
STORMLORD
These are fun and all, but they're really designed for catalogs and star mapping, not star gazing visuals. These only include stars, without any diffusion from their light or nebulae/galaxies. They aren't realistic in scenes. No offense. They would be good for leveled versions for viewing through atmospheres where only the brightest stars remain in underexposed imagery.
Even with a high exposure shot with auroras, not too many stars are going to show through that close to Earth.
I see this program used in my astronomy group to add back circular stars to composites and tracking anomalies through exposure times.
Quote from: Dune 12/01/2020, 23:16:48I don't think a surface layer will do (unless you meant something else). What you need is a a very high cloud layer at very high altitude, with indeed a stretched Y component, a good warp, and some appropriate colors from an extra PF into direct light input (or final coverage). Plus a fade/fuzzy zone by altitude (distribution shader, or distance shader with a camera in earth's center), or another method.
Oh I misunderstood. Thought we were talking about the corona thingy.
A series of surface layers with altitude cutoffs corresponding to the altitude layers of aurora colour effects can be used as the colours.
Here's a rough example. It's not fun playing around with the altitudes and fuzzy zones, and unfortunately these alts aren't realistic I don't think.
I think it looks quite close when you do a search with "aurora from space" ?
Yes, looks good. You have to get used to working with huge numbers (the e) indeed.
Quote from: Dune on January 16, 2020, 01:49:35 AMYes, looks good. You have to get used to working with huge numbers (the e) indeed.
More TGs falloff issue. Hard to get clean fades. Could achieve this with the corona because it's density is so low it hides the issue.