It is silent - what's being developed for Spring right now ?

[AlexMax]

I know that you guys are on the graphical tangent, but are any of these engine improvements going to make TA:Spring as a whole more stable? A bunch of my friends found TA:Spring, and they're addicted as all get out, but it has problems crashing for a portion of them. Heck, even I got a crash/hard reset when playing a 4v4 Shore to Shore.

Just wondering, not knocking the project, and we appriciate the work you guys have put into it already.

[jcnossen]

Bugs are fixed when they are found, but they are usually quite boring to discuss in a thread
I know tvo has recently fixed a bug in the projectile code that is used by the current windows version too, so that might improve stability...

[Neuralize]

I say we drop the old map format, it's good to purge the maps completely every once and a while.

[Caydr]

I'm almost tempted to agree. There's a lot of crap maps...

[smoth]

You know, a rating system where it would grab stats on what map is played and how often would probably help deal with alot of the crap maps.

[Buggi]

Wow did I ever start a fire on this thread.

All I said was work was being done with a new map format... I didn't think Zap would go the extra mile about it.

Also possible with this new format are maps of Epic proportions, i.e. 4000x4000. Although again, that's a bit on the steep side of things

Wait till you all see the demo's... wow!

[smoth]

no, that is good props to you guys and your work. If I wasn't soo busy with stuff other then my mod I would help. I really admire they work that has been done.

[Forboding Angel]

better yet, use a polling system on fileuniverse and delete all the crap maps.

[Masse]

Wait till you all see the demo's... wow!

trying to excite us ? well i cant be more exited ! im not eating until i see that demo !

[Forboding Angel]

You can have 24 bit heightmaps... All mine are...

[jcnossen]

screeny here:
http://taspring.clan-sy.com/phpbb/viewt ... 1245#51245

[Weaver]

You can have 24 bit heightmaps... All mine are...

How do you achieve this?

[Maelstrom]

Just make a regular bitmap. Then draw in grey scale. 24bit height map, that only uses 8 of the possible 24 bits. Kinda usless really.

[Weaver]

Thats what I was thinking, in my tests only the red channel was used.

However if can find out the 16bit raw format mapconv was once reported to accept I have an idea for a converter.

[Forboding Angel]

no no no

draw in full rgb

manually type in the rgb colors

example

r 81
g 81
b 81

then save as a 24bit bitmap. Do not convert to greyscale mode. Quite simple really. Just make sure you do everything in rgb mode and you will end up with an actual 24 bit bmp. I don't do anything in greyscal mode. yuk

[FizWizz]

still, if you keep everything in the grays, then you still only have effectively an 8-bit level of variety in your height values, do you not?

[Forboding Angel]

honestly I'm not 100% sure, I was under the impression that 16 bit was the highest possible, I could be quite wrong though.

[smoth]

still, if you keep everything in the grays, then you still only have effectively an 8-bit level of variety in your height values, do you not?

One would think so, after all colors range from:
000,000,000 white
256,256,256 black

Of course when you convert it to grey scale maybe the program some people use applies a compression algorithm. Just a guess, but maybe.

an 8bit color system:


For those who are noobs to binary... I'll try and be somewhat legible..

works like this....

These are two 8 bit unsigned strings....
(don't ask why just understand it is unsigned)

0000 0000 = 0
0000 0010 = 2

It works like this:

there are 4 bits to a byte(yes there is a thing called nybble), each bit represents a number to the power of 2. So thik of it like this:

[128][64][32][16] [8][4][2][1]

So a string of: 1111 1111 is: 128+64+32+16+8+4+2+1
or a string of: 0000 0001 is: 0+0+0+0+0+0+0+1
So our range is 0-255 giving us a total of 256 height values. I do not know what the system for greater then 256 colors would be but the 16 bit system is known as RGB color because it is REDGREENBLUE values...

I.E. 1,255,1 would be:
0000 0001, 1111 1111, 0000 0001

Breaking an image down to greyscale ONLY converts the pallete to a 256 color one with ONLY the values 0,0,0 to 255,255,255 in it is what you are saying. There is no real difference between rgb and greyscale heightmaps because of the the way the color range in defined in rgb. but we are still talking about an 8 bit image which leaves us with an 8 bit pallete meaning ONLY 256 colors

That a 16 bit system has a MUCH larger possible range. I am sure most of you are painting in 8bit colors but that is not the color that it is when painted on, because your are not using a solid brush, it has translucency, with that you will have blends of your 256 colors. So why does converting it to a greyscale image not reduce the color quality? It is because this is still not saved to a color system. You see as I say above an 8bit uses the RGB system and it uses ONLY 256 colors from that system..

This that means that any map(assuming an 8 bit color system)is going to have a base range 0-255 for colors.

However a 16 bit color system:

I am not really sure about a 16 bit color system but in real life grey scale should go from (-infinity,+inifinity) 16 bit color would grant a higher range of grey.

16 bit utilizes the color range more like this:

11111 11111 111111

Which is:
5 bytes for red
5 bytes for blue
6 bytes for green

16 bit images have a MUCH larger range of colors and the amount of colors stored with the image.

So 16 bit colors would be the way to go unless you have a want for only 0-255 on you image's range? Reguardless of image conversion to greyscale what you save your map at, be it 8 or 16bit will determine the difference between having good height range and barely any.

I do not know enough about the color system to tell you in detail about 16 bit but here is an wiki article about color systems:http://en.wikipedia.org/wiki/Bits_per_pixel

Does this answer your question FizWizz?

[FizWizz]

still, if you keep everything in the grays, then you still only have effectively an 8-bit level of variety in your height values, do you not?

One would think so, after all colors range from:
000,000,000 white
256,256,256 black

Of course when you convert it to grey scale maybe the program some people use applies a compression algorithm. Just a guess, but maybe.

an 8bit color system:


For those who are noobs to binary... I'll try and be somewhat legible..

works like this....

...(Rocket Science)...

Breaking an image down to greyscale ONLY converts the pallete to a 256 color one with ONLY the values 0,0,0 to 255,255,255 in it is what you are saying. There is no real difference between rgb and greyscale heightmaps because of the the way the color range in defined in rgb. but we are still talking about an 8 bit image which leaves us with an 8 bit pallete meaning ONLY 256 colors

...(Rocket Science)...

I do not know enough about the color system to tell you in detail about 16 bit but here is an wiki article about color systems:http://en.wikipedia.org/wiki/Bits_per_pixel

Does this answer your question FizWizz?

ahhh..... no. You had me up to the point where you started explaining 16-bit, and then you lost me. I'll check wikipedia though, hopefully I'll have better luck understanding it then.

[Weaver]

[ahhh..... no. You had me up to the point where you started explaining 16-bit, and then you lost me. I'll check wikipedia though, hopefully I'll have better luck understanding it then.

Thats because he explained 16bit 2 ways first where you use 16bits to define all the colours and then he mentioned 16bits for each colour! Which is 152,625,957,830,656 colours! Then there are palette based formats which allow a limited number (eg 256) of colours from a choice of 16,777,216