Help with sign logo!

Hi all! I'm new to making textures, etc. Here's my problem. I built a sign that has a logo on both sides (same logo) and I'm not sure where to start! I already have the logo. Do I just need to save it has a 32-bit TGA? The logo is has a white background if that matters...I have PS 7.0

I believe you can save that as a jpg file. You only need tga when you are using masks for transparency.

Thanks for your help. I got it done in about 2 mins.

You are welcome.

Yes and no. You only need 32 bit TGA if it has transparency. Otherwise use 24 bit TGA. Any image for the screen consists of red, green, and blue channels which are 8 bits each (adding up to 24). Images with transparency have a 4th channel (called alpha) which adds another 8 bits to the image for a total of 32. TGA's can be either RGB (no transparency) or RGBA (with thransparency), but jpeg's can only be RGB.

TGA is an uncompressed format; jpeg is highly compressed. You can use jpeg if you want, but keep in mind it's lossy quality. When you save as jpeg you're gonna lose a little quality through the compression. Further, every time you open it you lose a little more.

SL saves everything internally as jpeg2000, another type of compressed format. That means when you save as jpeg and then upload to SL, you're not only losing quality in the save, but also in the conversion from one compressed format to another. It's going to end up with the "copy of a copy" effect, meaning it will never look as good as if you had started with a TGA, an uncompressed format, which would mean compressing only once instead of twice.

Chances are you won't notice a whole lot of difference either way, but you very well might, so why take the risk? TGA all the way, every day, hey hey hey!

what he said, but also keep in mind the use for the image. if you dont need the BEST quality it is good to go with the jpeg- it will load faster on peoples screen and wont take a couple min of staring / waiting for it to load.

Trifen, after upload to SL every image becomes a jpeg2000. Once that happens, the original source file type becomes irrelevant. A jpeg2000 is a jpeg2000, period, regardless of what it started out as.

The only thing that will affect the load time on people's screens is the canvas size of the image and the bit depth. A 1024x1024 size image will load 4 times slower than a 512x512, and 16 times slower than 256x256. A 32 bit image will load 1 1/2 times slower than a 24 bit image.

So, if you've got 2 textures in SL, one of which started out as a jpeg that happened to be 1024x1024, and the other of which started out as a TGA that was 512x512. The 1024 one will still take 4 times as long to rez as the 512 one, regardless of the fact that the 1024 one started out as a jpeg. Once again, they are both jpeg2000 once they are in SL.

ah


i stand corrected.

Chosen, as I understand it, there's a little more to it than that. JPEG2000, like the JPG we're all more familiar with, is a compression algorithm. For most of them, the busier the texture, the larger the file (and thus the longer the load time.)

So a file that has a many small areas of sharp contrast could, in theory anyway, take longer to load than a file that has larger pixel dimentions, but very few color variations.

The trade off, of course, is that you don't need as many pixels for the second image to look clear in the first place. In other words, you don't need the large size unless you have a complex imagel and the very complexity is going to make it even slower to load.

Once again, as with everything else, "you pays your money, and you takes your choice."

Yes Robin, it's true that "busier" images take longer to load than "plainer ones", as can easily be seen when you watch one load and you see it go through a few stages of increasing clarity before you see it in full detail. That however, is not something I felt was relevent to this discussion, which is why I didn't mention it. No artist should ever be told to tone down the complexity of an image for speed or for any other reason. Art is art. However EVERYONE should always be encouraged to use the smallest practical canvas size and bit depth. Those are the things over which we have the most control, and which, generally speaking, do not impact image quality. No one can quantify what might make a picture "too busy", but we certainly can quantify what makes it too big.

Responsible texture management in SL is a huge issue. Go to any mall, and you'll see gigabytes worth of textures, which is why malls lag to all hell, overwhelming graphics cards & CPU's. The reason is not because people make too many nice looking, high contrast images; it's because they make them too big. It's not unusual at all for boneheads to put a 1024x1024, 32-bit texture on a piddly little .5 meter sign where a 128x128 would have worked just fine. No matter how plain or busy that texture may be, making it that big is just plain irresponsible.

The point is that advising people to use less colors or less contrast to speed up rendering is time is not practical. Telling them to use smaller canvases, and to only use alphas when necessary is.

Anyway, all of this, while interesting, is unrelated to the 24bit vs 32bit/jpeg vs tga question, which is what I was responding to here in the first place.

Chosen, I wasn't saying that people should use larger, plainer textures!

I was trying to add information to your post, not contradict it.

But, in my experience, there are people designing things in SL who aren't used to designing things for the web. They can be trying to decide between a gradient background, and one with a field of brightly colored flowers, and not realize that the choice affects more than esthetics; it can also have a real impact on download time.

I'm not about to tell people which background to choose. I'm not about to tell them what size to choose, either. But I think it only fair to provide the information that both of these things will impact the viewing experience of their prospective clients, and then let them decide how to balance things.

From my point of view, it's all about choice. But the choice should be informed. If, knowing that, a person chooses to place a high-resolution texture on a tiny sign, so that it looks good to him when he zooms in on it so it fills his entire screen, then that's his choice. Personally, I don't find it a wise choice, since it ultimately means that most won't see his sign at all; but he should, I feel, be free to make that choice.

I agree with you that advising people to use the smallest size that will show decent resolution on the finished object at normal viewing size is a Good Thing. But I admit that I don't really understand why advising them to eliminate unnecessary clutter isn't practical.

100% quality progressive scan jpeg, so little image quality lost and your 2mb tga just got droped to 200kb then sl gets them depening on how bad they compress the texutres your looking at 100kb or less with maby 3% quality gone i doubt you will even knotice it unless you strech a 128x128 pixel image 5+m.

Also do you have to work in 24-32 bit
yes you have to upload it as 24-32 bit, but if your making a white sign with out a lot of wild colors or super detail make it like 32k 16k or (god help it,still looks great) 256 colors. then when your done up it to 24-32 bit color, drop half the colors you drop half the size (mostly)

the inverse could be done also, and while you will see the pixelzation in ps gimp or psp sl does a good job of antilasting textures (not really jpeg 2000 blurs a bit)

the idea is.. if the picture is small it will (and does consideribly) res faster

below is a screenshot of one of my dance floor scenes each it is animated each frame is 128x128 pixels streched 10x10 m it has 12 frames total spritesheet is 512x512, 256 color pallet, jpg upload size was 96kb, it takes 10 seconds to res inside a fully loaded dance club, or just under 3 seconds to res everywhere else

Everytime I upload a straight JPG, I get a blank. I am using Photoshop 6. Any ideas?