[m2258734a]

This is a question I should have asked earlier. I have saved the Deneb image as several different types, because I was unsure of what type is necessary for Freespace SCP. So which file type would you like? As for the RGB color value, Deneb has values of 199 214 255. Thanks.

[Herra Tohtori]

Preferably DDS files these days... I think DXT3 format would be the correct one in this case.

TGA files also work, though they can be significantly larger than DDS files and also don't have the benefits of mip mapping.

[Alan Bolte]

IIRC, the visible galactic plane is at most 20 degrees wide. So if you're looking at it dead on, a band of somewhat brighter and more colorful background will fill at most half your view. Pull up maybe 35 degrees, you'll be looking at the same view as ever. I really can't see how it could get irritating. It seems silly to leave that out and put fictitious nebulae in.

[m2258734a]

One of the saved file types was .tga, so I assume that I should be able to convert my image from .tga to .dds. I just downloaded DDS Converter 2.1. Is this application recommended, and what options should I take into consideration when converting the image, such as JPG quality/compression ratios, filter types, number of mipmaps, etc.?

EDIT: I think what Agent Koopa was getting at is that the Freespace skybox, as it is right now, is very detailed in terms of nebulae. This nebulae filled sky, along with the large galactic plane might overcrowd the amount of celestial objects in the sky given the limited amount of free space available, and overall seem overdetailed. This is what might become irritating. What I mentioned is the fact that the galactic plane's appearance is not going to noticeably change as you travel from star system to star system in our stellar neighborhood. While I think it would be nice to see the galactic plane in all missions, there are just too many nebulae at the moment which could possibly obscure the band of light, or as mentioned earlier, overcrowd it. Since the galactic plane would be an unchanging background image, the overcrowding nebulae and the overall scenery would just seem repetitive as the storyline progresses. I have mixed feelings about the idea for those reasons, and maybe a decrease in the number of fictitious nebulae may solve the problem. I am all for realism as long as everyone else is content with the idea.

[Herra Tohtori]

I would recommend nVidia's DDS tools. It gives you a batch converter (command line utility) for converting files into variable DDS formats.

http://developer.nvidia.com/object/dds_utilities.html


...of course, I don't know if that program is just software or does it require a nVidia GPU or something like that. Probably not.


When you have installed the utilities, open Start menu and run cmd. Navigate to the directory where your file to be converted is. Then use the tool by typing command line

nvdxt -file <file.name> -<format> -<quality> -<other hundred cmdline options>

In this case, you would choose the appropriate format: There's about 20-30 of available formats to select from in nvDXT, but FS2_Open uses dxt1a for colour only maps like most ship textures, and dxt3 for transparent textures... mostly, anyway. Background images should AFAIK be in DXT3 mode. But I'm not really sure about what kind of blending mode the background images use, you should ask someone who knows the code.

I recommend you use the "quality_highest" flag (without dashes). And of other cmdline options, perhaps Gaussian mipmap filtering (I don't know what the default is).

Just typing nvdxt command will open a rather comprehensive documentation, you should be okay with it.


I find it especially handy when you need to convert many textures - like frames for animation. You can just use

nvdxt -file *.tga <cmdline options>

which converts all tga files in the directory into selected DDS format.

[Bobboau]

"and dxt3 for transparent textures"

not for transparent textures, for textures that NEED and alpha chanel, if you don't requier an alpha chanel do not use dxt3. anything that is blended with additive alpha and thust does not use an alpha chanle, like the nebulas in the background, should NOT use dxt3, anything that uses alpha blending, like planets in the background should use dxt3.

[Agent_Koopa]

EDIT: I think what Agent Koopa was getting at is that the Freespace skybox, as it is right now, is very detailed in terms of nebulae. This nebulae filled sky, along with the large galactic plane might overcrowd the amount of celestial objects in the sky given the limited amount of free space available, and overall seem overdetailed. This is what might become irritating. What I mentioned is the fact that the galactic plane's appearance is not going to noticeably change as you travel from star system to star system in our stellar neighborhood. While I think it would be nice to see the galactic plane in all missions, there are just too many nebulae at the moment which could possibly obscure the band of light, or as mentioned earlier, overcrowd it. Since the galactic plane would be an unchanging background image, the overcrowding nebulae and the overall scenery would just seem repetitive as the storyline progresses. I have mixed feelings about the idea for those reasons, and maybe a decrease in the number of fictitious nebulae may solve the problem. I am all for realism as long as everyone else is content with the idea.

Yes, that was what I meant. Nice to know sometimes that my opinions are not wholly unfounded.

[S-99]

Going as far as realism goes. If people lived on planets in a supergiant system. Say like humans, planet conditions would at least have to be characteristicly comparable to earth for life. Which means planets to sustain human life in a supergiant system would have an extreme distance away from the supergiant for it's orbit, otherwise you roast the planet quite easily.

Pretty much i'm saying the beautiful deneb star that was created in this thread wouldn't like take up the whole skybox because all the ships in the said mission would probably be going back and forth between planets as the standard lanes of traffic and in fs2 standard lanes of assault perhaps. Fs ships, hell not many ships could get very close to a supergiant at all. Fs ships would roast, sort of like that heat wave from hitting those two corvettes before the actual nova happened(a super giant probably gets a lot hotter than that heat wave). The supergiant in a skybox would probably be the average size star used for lighting in all missions.

[Admiral Nelson]

Oh yes, this is true.  However, the distinct color and light ray pattern he has created will remain unique to Deneb.  Note that one exception to your comment may be Betelgeuse, given that it is relatively cool, but incredibly large (perhaps 500 times the size of Sol).  Once we have Deneb done, m2258734a should give Betelgeuse a go...

[m2258734a]

You read my mind. I was planning on attacking the supergiants first, since they are at the very top of the luminosity class hierarchy. I chose to do Deneb first only because I had mentioned it towards the beginning of the thread  .

I have finished Deneb, and it's appearance is exactly the same as the image on the first page. I'm very happy with the luminous appearance, but I would appreciate if someone would finalize the image. What I plan to do is provide the new images of the stars as I finish them, along with stating the RGB color values used for each star. I can only go so far as to create the basic outline of the star, such as what I have done with Deneb, and hand it over to someone with more experience who can really give the stars the artistic touch that they need. In the case of Deneb, I would like to give the star a "supergiant" appearance without having to resort to making the star take up half of the sky box. This is a little difficult, because I think I would need some form of reference (i.e. the Sun) in order to show just how large Deneb is, and I don't see how this would work in FS. I think the luminous appearance is just right, but the surface of the star looks a little on the grainy side IMO. I'd be more content if Deneb had a more prominent, plasma-like surface..... what do you all think needs improvement?

As for the size of the image, I do not have the original FS star images to do comparisons, so my Deneb image is very large (2000 X 2000 to be exact). I could easily resize and crop the image without losing clarity, so if anyone could provide the FS star images I would greatly appreciate it. Unless the media vp stars are too small in FS, the new stars are not going to be any bigger than the stars they are supposed to replace. This idea will become unrealistic for supergiant systems with habitable planets, such as Cygnus Prime, but I wouldn't want to make the stars appear so small in the background that the new features go unnoticed. With Deneb's luminosity of about 160000 solar luminosities, we can estimate the average radius of the circumstellar habitable zone (CHZ) for a planet in orbit. This means that Cygnus Prime would have to be about 400 AU away from Deneb. At this distance, given Deneb's +200 solar radius, Deneb would appear about 3 times smaller as observed from Cygnus Prime than the angular diameter of the Sun as observed from Earth.

To an inhabitant of Cygnus Prime, Deneb would actually appear just as bright as the Sun would from Earth with an apparent magnitude of -27.3. However, I don't know how much improvement of detail a FS user would experience with a star slightly less than one third the angular diameter of the Sun. So I think it would be fine just make the new stars no larger than their predecessors unless the predecessor is too small to begin with. What do you all think?

As for converting the images, I'm still having trouble. I think I am going to send the original .tga file, and hopefully if someone does take the image to improve it, they can convert the file, as well. I know it's a simple task, but I'd feel comfortable with someone else doing it who has actually done this before. I wouldn't want to hinder the quality of the image without knowing.

Until then, I will have to wait till I see how large the original Deneb image was so that way I can resize my image and attach it to a future message... (Betelgeuse and Antares come to mind.... )

[Admiral Nelson]

The original images are just 128 x 128 pcx files.  There are two images, one for the star itself, and one for its glow.  As you see, there is no detail whatsoever.  Your work would need to be resized to some intermediate level, as these are much too small to be useful.

[m2258734a]

Are those the very original images Volition created for FS, or are these the ones from the media vps. I do recall this image before I was ever introduced to the SCP, but ever since then I remember the stars looked different: the stars had no light rays, they came in all sorts of colors like cyan and violet, and had large halos. There was no other detail that that. Let's see if I can find an example...



Here is an image from the "Screenies" thread from the general FS forum. These are the stars I remember seeing. Might you have those images available? Thanks for providing what you had.

I'm just about done with Betelgeuse.....

[Admiral Nelson]

I extracted these from sparky_fs2.vp, so these should be the originals.  This is the extract from stars.tbl of all the sun bitmaps.  You will see that there are only 5 of them.  Moreover, since the colors are defined as RGBI,  the luminosity value has to be supplied in stars.tbl.

Code: [Select]

; sun bitmaps and lights - NOTE all $Suns must have a $Sunglow and a $SunRGBI!

$Sun: SunWhite
$Sunglow: Sunglow01
$SunRGBI: 1.0 1.0 1.0 1.0

$Sun: SunRed
$Sunglow: SunglowRed
$SunRGBI: 1.0 0.1 0.1 1.0

$Sun: SunGreen
$Sunglow: SunglowGreen
$SunRGBI: 0.5 1.0 0.5 1.0

$Sun: SunBlue
$Sunglow: SunglowBlue
$SunRGBI: 0.2 0.2 1.0 1.0

$Sun: SunGold
$Sunglow: SunglowGold
$SunRGBI: 0.74 0.76 0.44 1.0

$Sun: SunViolet
$Sunglow: SunglowViolet
$SunRGBI: 0.66 0.44 0.54 1.0

#end


[S-99]

In that case betelgeuse would have closer orbiting planets.

That starfield is definitely from the 3.6.7vp's, the vanilla fs2 starfield is very starless, and the suns in the mediavp's have already replaced the vanilla fs2 ones.

[m2258734a]

Yup. Assuming that Betelgeuse does have habitable planets, they would probably be within 200 to 316 AU from the supergiant. This all has to do with how luminous the parent star is.

I understand the star bitmaps now, and thanks for the information Admiral Nelson. I now see why I was asked about the RGB color value used for Deneb. However, are my images going to be compatible with the stars.tbl since I don't have a separate image for glows, or can the .tbl be modified to fit my progress. Also, is it more beneficial to create separate images for glows or to have one single image as I have done?

Well, Antares is next on my list and the star system got me thinking.... Antares A is about 550 AU away from Antares B. Considering that Antares B is a spectral class B2.5V star, we can assume that the radius of the star is somewhere in between 5-10 solar radii. So, from Antares A we can calculate that Antares B would appear about 613 times smaller than the Sun would appear from Earth. From Antares B, the +700 solar radii Antares A would appear about 1/5 the size of the Sun as viewed from Earth. I even went ahead and took into consideration that the user is going to be about 50 AU from Antares A in order for the supergiant to appear twice as large as the Sun, yet no matter where the user is positioned, Antares B will always be a very bright speck in the star field.

Assuming that we have multiple missions in one binary system, why not change our perspective of the binary system by resizing and repositioning the stars in each mission? Let's say we are in Antares for the first time, and the main star in the background is Antares A. For another mission we can venture close to Antares B, and have it has the prominent background feature. Since there are several multiple star systems that have made their way into the FS universe, would you all agree that this might be a good idea to opportunity to enhance the scenery?

If you all think this is a good idea, then I am going to need information about the missions and the location each mission takes place. That way, I can come up with some sort of stellar template if that's possible. If this is accepted, I think I'm going to hold out on resizing the images until I can understand how I should reposition the stars according to each mission in FS and FS2. I have four more supergiants to finish: Adhara, Enif (Epsilon Pegasi), Mirfak, and Polaris. When these are finished, I will provide an attachment for all supergiants in FS for anyone to improve or take possession of.  I would say to hold back on resizing and converting the images to .dds until we know how large each star should be in each mission... assuming you all like the idea. 

EDIT: Tell me what you think about Betelgeuse...



RGB: 255 193 104

[miskat]

I would blur the edges of the inner sphere, fight now it looks like the glow is just a backlight of some sort.  blur the hell out of it all for a more realistic "youre looking into the sun, and your eyes cannot deal with it" effect.

Just my 2cp

[jr2]

TABS!!

[Admiral Nelson]

There is no need to resize your star image based upon how the star will appear in mission.  FRED has a setting to alter the scale of the star in each mission.  You merely need to select a size that is a good balance between graphics memory use and detail. 

The glow is required; you cannot use any star in FS2 without one.  Are the "rays" emanating from your star on their own layer?  That could be exported as a separate image to serve as the glow.

Betelgeuse is looking good!

[m2258734a]

Oh, ok. That's good to know.

When I made the stars I created everything on a single layer. Since this is the case, and stars are required to have separate glows in FS, it looks like I am going to have to start all over; and I was almost done with the supergiants...

It's ok though, I've learned from my mistakes so hopefully I should be able to create more promising stars in the future. So, how should I separate the star from its glow? Should I create a single image of the star with its own luminous atmosphere, including flares, and then create another image consisting of light rays? Maybe if I illustrate it....

So let's say I wanted a FS mission to have a pulsar. Altogether, the image would look like this:



So these images would be the star and glow:



If this is what I should do, do I need to do anything to make the light ray layer transparent or does FRED take care of that, as well?

[Admiral Nelson]

Based on the original examples above it wouldn't appear that there is anything special done wrt transparency.  Perhaps the best approach would be to produce a simple ordinary star like Alpha Centauri A that can be completed quickly and get that working in game before moving on to all the big stars.