Home Page Forums MAME MAME Mattel Aquarius palette inaccuracy

compared to emulators like Virtual Aquarius, the palette from Aquarius emulation at MESS looks quite different

people at the AtariAge used to say that the gray from colour 15 is quite dark, very hard to distinguish from the black from colour 0 - at Virtual Aquarius it is a dark grey, and at MESS it is quite a bright grey - so i guess, based on this, that the palette at MESS is all messed up?

how can we get the exact colours used in the Aquarius palette? should we have the YPbPr colour codes just like some descriptions at http://en.wikipedia.org/wiki/List_of_8-bit_computer_hardware_palettes ? is frame grabing into 720x480 or 720x576 pictures (via dvd recording or video digitizing) enough? perhaps only using specific electronic equipment? which accurate methods are used in this process?

and a call: who in this forum could kindly help on fixing it?

(and meanwhile, how can we use alternative palettes on Aquarius emulation on MESS?)

another thing is that, on Mattel Aquarius, the upper leftmost character from the 40x25 text area is repeated in the whole border (the border "colour" is defined there) - at MESS nothing happens

These should be submitted as bugs at MAME Testers. For the palette issue you will likely need to produce some corroborating evidence first though - other emulators are often not a good reference for such things.

There is no way to use an alternative palette other than modifying the source code and recompiling your own build.

i guess the evidence can be find here, from the official technical documentation: http://atariage.com/forums/topic/227441-...om-the-palette/ - the info there is related to luminance, chrominance, phase, amplitude, and i don’t know how to convert them to YPbPr and RGB values

about MAME Testers (thanks about mentioning!), it targets us back to here, so i guess i'm doing the right thing posting this thread here!

Those links at MAME Testers main page are for those people who haven't signed up, or have signed up and are awaiting approval to post official bugs as a tester and "cannot wait". I guess I need to make the text bigger at the top?

Like Justin mentioned, you need some proof (video/snap/comparisons) from actual hardware so that the Devs can have a reference to go by.

Larger text doesn't help, hence that notice on the top page of the forums here now.

Based on that data sheet, the first 8 colors are more or less normal digital RGB, it's the inverted ones where you get into trouble. Also, the final luma values appear to depend pretty heavily on some voltages being passed to the chip, so we'd need someone to go VOM-ing on a working Aquarius or two. Or is there a schematic available?

there is more updated technical discussion at http://atariage.com/forums/topic/227441-...om-the-palette/

based on that, i tried to propose https://github.com/nitrofurano/mame/commit/2ddff28f6d3b5bad47f694aca05162e8afc66fad ( https://github.com/mamedev/mame/pull/86 ), but i'm seeing no changes yet at https://github.com/mamedev/mame/blob/master/src/mess/video/aquarius.c

thanks again!

I'll take care of it.

Yeah, that submit is rather more rigorous than the previous "eyeballing an NTSC screen" one so it's acceptable

Old vs. new:





Don't know how to generate the color bars.