Vector to .png

I am using Corel Draw 11. I have drawn a grey/b/w tattoo.

When I export the drawing to .png it pixelizes like crazy.

I started off with the drawing being over 2100 pxls and downsized during export.

I downsized it within Draw to 1024 and exported it one to one.

But no matter what I do - it pixelizes.

I've tried exporting in a higher/lower dpi, greyscale/24bit, etc.

It's not the first time I've created something - but the first time I've had this problem.

Even my normal workaround of exporting to .psd and then importing the .psd into Gimp and then exporting to .tga there (Yes, Corel exports as .tga too, but the transparency never works *sigh*)

I am stumped.

I've never used Corel Draw, but have many years of experience with other vector drawing programs, so I'll try to help the best I can...

If you can export your drawing to .eps format, PhotoShop will usually accept it with minimal degradation. But, I think your best bet is exporting the file as an actual Illustrator file (preferably Illustrator 5.0 or higher). I know Macromedia FreeHand can do it, but I'm not sure about Corel Draw.

Also, if Corel Draw has a feature that allows you to control the dpi of what you're exporting, then aim for the highest setting... I've found that this helps a great deal also.

Anyway, give those things a try or provide some info on what Corel has in its arsenal as far as exporting options... perhaps we can figure it out from there.

When you do your TGA export, are you selecting Normal or Enhanced?

I just made a drawing in CorelDraw 12, with a compound shape with a hole it in, and exported it as TGA. I selected the shapes, did file export as TGA, selected transparent background, and Normal mode. The transparency works in SL. The background outside the shapes is transparent, and you can see through the holes.

I also exported a png and don't have an unduly pixelated result. I used pixels as my units in Coreldraw, 2408 by 3508. When I exported it I set the width and height for exporting to the same number of pixels.

For use in SL, one should always be thinking about the powers of two condition, either designing in powers of two sizes, or converting them yourself to powers of two larger than the original. So it might be best to set your background size in CorelDraw to 1024 by 1024 pixels or 1024 by 512 or whatever the appropriate powers of two might be.

I can't remember for sure at the moment but it seems like SL has a size limit of 1024 pixels, if not that then I think it's 2048, so using an non-powers of two size larger than the limit guarantees that SL will have to shrink the original to the limit, which may make the image more pixelated or blurry or both.

I don't Coreldraw 12 is very different than 11 in how it exports. Sorry I can't be clearer.

Is the Gimp capable of producing an alpha channel? That seems to be where the problem lies. You need to ensure that a mask is in place and can be exported to the PNG format.

DPI by and large should not matter. Also vector imagery isn't supported in SL so by default, it must be rasterized. I'm sure you already knew this but I wanted to make sure we're all on the same page.

I haven't used CorelDraw in many years so I can't advice you on what to do with your current iteration of CorelDraw. Do you have access to Photoshop and Illustrator?

What I would recommend is that you save your CorelDraw image to Illustrator (.ai format) or as an .EPS file. I don't know how faithfully CorelDraw will reproduce your artwork as a .PSD file which is why I recommend either of the two aforementioned formats.

Either way, when you open the files in Illustrator, simply copy the artwork, load Photoshop and then go to File -> New and hit enter as the dimensions should automatically be set for the image size. With a new file open, paste your artwork via Edit -> Paste. Next, control-click (command click on Mac) the artwork layer to create a load selection, create a new channel in the channel's palette and in that new channel, fill your selection with white. There's your alpha mask.

From there, save the image as a 32-bit TGA file (Targa format). Why? Because unlike .PNG, your image won't be compressed. When you upload to Second Life, it will be converted to a JPEG2000 format. No point in double-compressing it

Let us know how that works out.

The PNG compression is lossless, isn't it?

Yep. It uses deflate, the same as gzip and PKZIP.

Um yea, the gimp can

and you can turn off the lossless compression if you want on png files too