Feature 11 - Pharaohs Return (C64)

[Helm]

Thanks for the very informative post.

So, from my edit, the blue is out. It can be replaced with the brown.

The yellow on the top not being available generally is unfortunate, but it can still be replaced with the light green with the same pixel composition.



the more significant part of my edit concerns pixel formations, not the hue shifts anyway.

[STE 86]

ok that's a much better grip on it. The only thing I notice that you won't be able to to is the orange highlight on the brown background.

in a square with black in it, only the colours Brown, Green and Lt. Green can be used.

in a square with no black in it, only the first 8 colours of the 16 colour palette can be used to replace the black

However to complicate things further in the squares that ARE just brown and black (character colour and background) the graphics can be in 1:1 pixel mode (hires chars), as the OP has done in his version.

This is however much more of a pain in the arse to draw outside of a real c64 because its a hardware thing.

Steve

[PypeBros]

too bad blue background can't be used... it worked pretty well. Does that mean I haven't de-constructed the screen shot and that it used black as the major background color anyway ?

Btw, I tried some alternate colors for the hero's jacket.



Imho, cyan could work pretty well... immediately seconded by lightgreen.
I managed to give him eyes, as well ... hope you like it.

Hmm ... and after posting, I just realised that changing the color jacket doesn't help alot. The brown of the trousers/hat is what definitely needs to be moved apart from background colors (imho)

[Helm]

However to complicate things further  in the squares that ARE just brown and black (character colour and background) the graphics can be in 1:1 pixel mode (hires chars), as the OP has done in his version.

I purposefully avoided that because I'm not sure what mixing resolution for background tiles achieves. For sprites the added information is crucial, but for tiles? Meh.  The only exception I'd allow for myself is a solid 50% dither patter with square pixels to do some fades.

[ptoing]

You are coming along nicely. Also good point about "would you do it for C64..."

I made some edits on the sprite.



Because your sprite is quite small, you have detail in the bg and the colours of your sprites and bg are similar (you know this does not need to be the case at all), it is very important to keep it readable.

I felt that the grey is totally wasted so I went for light red and orange and and then used black as the hires overlay. This helps to make sure that the sprite pops out nicely. As far as shirt colour goes the cyan works well here because it pops nicely. Nudging the sprite down into the ground one pixel makes the connection feel a bit more solid as well.

You should perhaps study small NES sprites more than C64 stuff when you are making small sprites like this with hires outline.

Keep it up 

[STE 86]

PypeBros, if you are referring to your deconstruction of the Soulless screenshot, then no, it doesn't use black colour ram to do what it does.

but your deconstruction is also errored. In soulless, the Multicolours on the screen you de constructed would be brown and white. the character colour would be predominantly the yellow areas and the hires lowlight areas and the background as black.

if you look at the 3rd screenshot of soulless it becomes clearer because there are multiple character colours (yellow, green and cyan colour ram) in conjunction with the same white and brown MC registers.

white is an unusual choice from an MC because its part of the lower 8 set and MC are usually picked from the upper 8 by artists.

Helm: the advantage of using hires chars is that you get fine colourful background detail for free. It can give the impression of depth and because its at a different resolution and can be much finer detail, is much less prone to "merging" with the sprites as they move over it and obscuring them than MC pixels.

Steve

[Helm]

I agree in theory. In practice it looks like resolution mixing usually looks to me, weird and disconnecting. There must be examples in the c64 library where this is done successfully, but I think it's more probable that the hi-res mask is used for highlights than for the darkest colour in the ramp before black like above.

[Carnivac]

I'm really not a fan of the mixed resolution appearance at all but I don't mind it here on the sprites so much as it draws them out a bit but I'm not sure I like it used on the backgrounds since to me the 1x1 pixels look sharper and more defined so my eyes are drawn to that on a piece like this and 2x1 looks a bit more unfocused which looks more suitable for backgrounds you aren't meant to pay too much attention too.

[Helm]

Yes! Exactly! Single pixel detail catches the eye and seems to be nearer and brighter than it should be. Which is why it works for Soulless, because it actually is nearer and brighter.

[BladeJunker]

I tend to associate increased resolution with objects closer to my eyes which I guess most humans do (lol) which is why I was never a fan of Mode7 scaling where they blew out the pixels when scaling the sprites larger to make them appear as if they were going closer to the viewer.
Anyway I should ask STE 86 if there is any way to make the foreground tiles single pixel? Idk maybe a double wide layer for base color with a single pixel overlay of outline or shadow maybe.

I can't say I'm a fan of mixed resolution either but as far as my Atari research its something I've had to deal with out of necessity, still I think its applicable to the C64.
I think a hierarchy of resolution priority tends to be the logical approach to mixed resolutions which I think is akin to truncation of subdivision in polygonal meshes IE. more resolution in the head than the body, more resolution in the face than the head, and more resolution in the eyes than the ears.
As far as 2D I think character and item sprites benefit greatly from single pixels especially in visual separation and focus to the player. Although Half-Life was 3D I thought the increased texture resolution on charge stations and elevator buttons to that of the world was a positive example of visual focus for players.

Its interesting to discuss but I tend to lean towards what ptoing said about just using double wide pixel throughout instead of adding rastersplits, it just seems hard to add to a game when compared to an art slide.