Screen Colormap

Do we have a consensus on what grey levels best represent the 4 available shades on the empeg display?

I'm writing a Gimp plug-in to load and save logo-screen files, and I'd like to have a suitable palette for it. I currently have {0x00, 0x20, 0x80, 0xFF} but wanted to know what others use.

I've attached my current state of the plug-in. This compiles on Linux with gimp1.2, and will read and write logo files like those distributed with "download.c" . I'm planning to add support for linking separate AC and DC images, Empeg/Rio personalities, generating blank 128x32 4-colour images (currently done in script-fu), and an integrated uploader.

Toby,

What's the chances of extending this to be able to generate all the relevant image files (vu background, escher image pan, boot logos etc etc) and zap them up to the empeg (a la LogoEditor)?

In reply to:

---

What's the chances of extending this to be able to generate all the relevant image files (vu background, escher image pan, boot logos etc etc) and zap them up to the empeg (a la LogoEditor)?

---

That's an eventual goal (I'm taking one step at a time...) It's the first time I've done any Gimp work, so I'm learning a lot as I go. I've just re-organised the logo stuff to use layers on a background image, as that's what Gimp does with multi-frame GIF images.

Uploading anything other than the boot-logo is going to require a Hijack-enabled player, though.

amne for hijack, its a pretty safe bet that anyone wanting to do anything such as logo editing and what not is going to have hijack

I eventually worked out a good set of search terms and found http://empeg.comms.net/php/showthreaded.php?Board=empeg_tech&Number=11510

So I'm about to try (0x00, 0x20, 0x40, 0xFF) as an approximation.

Remember that the percentage of time the pixel is "lit" is not necessarily going to translate mathematically into perceived brightness. The grays are much brighter than the post you linked would imply.

If you want to know what I think the proper gray values are, look at the INI file of your copy of Empeg Logo Editor sometime.

In reply to:

---

Remember that the percentage of time the pixel is "lit" is not necessarily going to translate mathematically into perceived brightness. The grays are much brighter than the post you linked would imply.

---

You're right - I realised why when I got home. Persistence of the VFD means that the pixels stay lit a bit longer than they are powered.

In reply to:

---

If you want to know what I think the proper gray values are, look at the INI file of your copy of Empeg Logo Editor sometime.

---

Isn't the Logo Editor a Windows program? Or are you saying that the distribution comes with a configuration file that I can read?

Isn't the Logo Editor a Windows program? Or are you saying that the distribution comes with a configuration file that I can read?

Well, I think the config file only gets written on first-run of Logo Editor, so if you don't have a Windows box, then no. There is a COLORS.TXT file that discusses the color values, though.

Your best bet is to look here for my discussion on the colors.

In fact, let me strengthen that statement. If you are writing code that does any kind of color conversion display onto the empeg screen, don't go any farther until you've read this thread. Please click here.

Thanks for that thread, Tony. It didn't turn up in any of my searches (probably because you used terms like "screen shades" where I tried searching for "greyscales").

I've put these values in my palette:
{ 0, 91, 115, 255 }

Colour-reduction's not my problem - the Gimp can reduce to that palette using fixed dithering, Floyd-Steinberg dithering (recommended - with the low-bleed option - for photographs), or no dithering (good for cartoony images).

The code's looking quite good, though I've now moved to Gimp 1.2. I'll post it when I resolve some issues with layer composition.