Spell scrolling & Scaling
DMZ
Member Posts: 39
Spell Scrolling
Current behavior:
When casting a spell, player has to scroll/click on arrows quite a long time if they have a lot of different spells memorized.
Desired Behavior:
When using higher resolutions than 640x480, enable the spell list on the bottom to take advantage of the extra length, so less scrolling is required. Increase the number of available quick spell slots accordingly as well. This will prevent some irritation when a player has a large spellbook, like the case with Cleric/Mages and high level casters.
Scaling
Current behavior:
When increasing resolution, screen simply zooms out.
Desired behavior:
When resolution increases, up-scale the sprites and terrain instead of zooming out. Makes the game look better. Another possibility is to use a part of the newly acquired pixels as anti-aliasing.
Current behavior:
When casting a spell, player has to scroll/click on arrows quite a long time if they have a lot of different spells memorized.
Desired Behavior:
When using higher resolutions than 640x480, enable the spell list on the bottom to take advantage of the extra length, so less scrolling is required. Increase the number of available quick spell slots accordingly as well. This will prevent some irritation when a player has a large spellbook, like the case with Cleric/Mages and high level casters.
Scaling
Current behavior:
When increasing resolution, screen simply zooms out.
Desired behavior:
When resolution increases, up-scale the sprites and terrain instead of zooming out. Makes the game look better. Another possibility is to use a part of the newly acquired pixels as anti-aliasing.
2
Comments
The resolution thing will also be solved by BGEE having a zoom feature, though its unlikely that the sprites will display increased resolution.
Wait i am evil :x
Same as natively high resolution graphics? No, not even close.
Better than vanilla graphics with no upscaling? Yes, visibly so.
As for the spell bar, I'm not sure how much the UI is going to change, but I'd much rather keep the BG1 style with a few adjustments (customizable buttons, spell&scroll list.)
As for the UI, it is now completely soft-coded and, as per Cameron Tofer, it will be completely overhauled. It will most likely be natively HD (and ingame fonts are very likely going to be re-rendered in high resolution, since Cameron Tofer once stated on Twitter that he could locate the source files) and the skin is going to be moddable / customizable. I have already requested the option to enable the original stony skin, so hopefully it will be there.
Otherwise I don't know of any upscaling algorithm that can improve image quality. At best, upscaling an image doesn't lose information; it cannot create information that's not there in the source material.
New characters would look better than old stuff, but won't new area art that they are making from scratch as they said look much better than the old area art as well? And won't the characters that stand in those areas look weird and dated also? They already looked worse compared to surroundings before, especially in BG2.
This question bothers me as i'm trying to figure out how come it's ok to have new art for areas that much better, but having characters better than monsters or commoners would be a problem. I don't understand why it had to be a all or nothing job, since it could be done slowly in the backround and releasing it in parts. We never asked about better monster models anyway, but monsters like dragons look better in quality than players anyway so i don't see the problem.
In my humble opinion always, if you really wanna do something, you do it. Let's hope they work something out eventually, although if that eventually is BG3, well, you never know if that happens anyway.
To use an analogy, you don't get more pie by splitting the pie in multiple pieces. At best, you retain exactly as much pie as you originally had (and in practice you lose some because some crumbs fall off, some of it gets stuck on the knife, on the plates, etc)
Hq3x is a special case - it assumes a particular art style and creates new information based on that heuristic. I'd be curious to see how it fares on BG art, but I don't have high hopes.
However, if you know some specific features in your image (high density of squares, for exemple) then you can apply specifics algorithms to regain them (those are generally called "filters"). While the reconstruction is not 100% perfect, it can still have very efficient results with 90% less information. A simple version of that is morphomaths, but it works only on very specifics features, so it isn't very handy to use. More complicated filters takes the hessian of the image to search for images edges, for example (I won't go into details, but images widdened this way are very neat on the general proportions, and suffers from less noise, search articles on "edge preserving filters"). It works quite well in medical imagery, at the cost of time. So yeah, basically @AndreaColombo is right.
BG being so low-resolution as to being close to pixel art, I think the only enhancements that might give good results throughout are the pixel art upscaling algorithms like hq3x or 2xSal (http://en.wikipedia.org/wiki/Pixel_art_scaling_algorithms#Pixel_art_scaling_algorithms). That's what emulators like NeoGeo use to make very old games look good on high-res displays. They tend to make the picture look like a watercolor painting if you go too far, but perhaps a combination of these and a more general upscale (bicubic/lanczos)?
So... Awesome BG:EE Team... impress us with impressive technical skills