Platform Masters Title
Platform Masters - Will you be the world's next platform master?

Ronnisa Plains - new hills and narrower Keveran River

Date: February 13, 2014
Screenshot #491: This screenshot clearly shows how much narrower the Keveran River is. This width is still wider than it will actually be as I'll still need to make ground decals for the Keveran River. Even without that, this width of 16 SU (1602 feet or 488.3 meters) makes far more sense than 144 SU (2.73 miles or 4.39 kilometers). The likely final width will likely probably be more around 1/2 to 3/4 as wide as this depending on how much weaving I do. This view is maybe about how high the supersize level might get (though probably only 2/3 this height).

Ronnisa Plains - new hills at a high height

Date: February 13, 2014
Screenshot #492: In the previous screenshots, you couldn't see the clouds at all. In order to even see the clouds, you need to get much higher than what any level will get you to. This screenshot gives an idea on how high you need to get to see the clouds.

Keveran Desert - new Keveran Mountains near ground

Date: February 21, 2014
Screenshot #493: From ground level, the newly designed Keveran Mountains easily stand out compared to what it was before. Not only are they significantly taller in pixel count, they're also 25% taller physically as well. The Keveran Mountains really dominate the scene.

Keveran Desert - full set of the new Keveran Mountains

Date: February 21, 2014
Screenshot #494: From about the height of a typical jumbo level, the Keveran Mountains really stand out - the entire set is visible. Of the several changes I made, I moved the starting point to 36 SU instead of 40 making the closest ones a bit closer. I also made them rougher-looking, more jagged.

Keveran Desert - new Keveran Mountains into the clouds

Date: February 21, 2014
Screenshot #495: This screenshot shows 2 of the other changes I made. The primary reason for redoing these was to space them much closer together than they were and here, you can't really see too much of a gap between them. Second, due to being 25% taller, the tallest ones rise into the clouds making Keveran Desert now the first world you'll see this effect in (as opposed to waiting clear until Carnivalesta). The third, the second unique one, is that they are much greener than before. You can see the grass at the tops much closer and in much more detail than before.

Keveran Desert - new Keveran Mountains with more grass

Date: February 21, 2014
Screenshot #496: This screenshot better shows how different the grass is. Because of Lake Keveran in the background (yes, now officially Lake Keveran), and shorter mountains beyond, it's easier for the rain from the ocean to get to those but they begin to block moisture creating a dry spot, hence Keveran Desert. When I get at working on the overworld, you'll see Keveran Desert being really landlocked and surrounded by mountains.

Programming - automating fog generation

Date: February 28, 2014
Screenshot #497: After extensive planning on how I should go about processing the fog, I realized I needed over 200 layers, rather extreme. So, instead of processing the fog manually, I knew it was going to be a lot faster and easier to just write a program to do it. Not only that, I could also use pure dithering instead of just a checkerboard, enhancing the accuracy. This is just a part of the program, the main part of it. And indeed, when I used the Sentus Caverns as my test subject, I practically couldn't tell any difference between the old checkerboard system and the new full dithering system. Also, until this point, I didn't know the full TGA file format (I only learned enough of the format to get the image's width and height).

Keveran Forest - new hills and early fog preview

Date: February 28, 2014
Screenshot #498: The last time I really did anything significant with Keveran Forest was way back in late June of 2010. Over 40 months later, Keveran Forest is now finally starting to get some attention. Taking after the idea with Sentus Caverns for the mist, I thought Keveran Forest could use the same thing. Although it looks barren, thanks to the complete lack of trees, I'm still working on it. At least the fogging effect is very nice. The Teigram River looks rather wide, but this is the planned span though I'm likely to change this to something from 8 to 16 SU instead of 8 to 24 SU). Yes, I decided to change the spelling slightly - the pronunciation is still the same - "TEE gruhm".

The fog does have a few weak points I need to deal with. Because of unintended oversights, I need to redo the fog. The good news: that fog generation program has somewhat unexpectedly saved me a whole lot of hassle so I don't have to spend anywhere near as much time making the fixes and updates as needed. I was expecting a possible redo, though only from the uncertainty on how the opacity at the bottom should be.

Keveran Forest - fog gradient too obvious

Date: February 28, 2014
Screenshot #499: The farthest parts, for which I have the fog only going out to 112 SU, unexpectedly gave me some indications that I don't need anywhere near the 240 layers of fog I thought I was going to need (221 actual as the 19 layers of the Keveran Mountains can actually use the fog embedded in them). In fact, I really only need to go to 81 SU. From there, nothing more than a simple transparent cover is needed for the fog, which can be combined into the ground fog as this area is nothing but monotonic textures that rectangular fills cover well. This means I only need to focus on the foreground area for quality and just simplify the background area.

Keveran Forest - fog too fine

Date: February 28, 2014
Screenshot #500: A major milestone - number 500! In this screenshot, I adjusted the fog's parameters so it only went out to 81 SU and also with a lot finer spacing. I also adjusted the parameters so that the fog steps by 1/4 SU until 16 SU instead of 4 SU and stepping by 1/2 SU from there until the end. The result? 191 layers with a far too high degree of fineness and way too much memory and drawing being done. At the lowest point, I'm passing 35 million pixels drawn and hovering around 2100 draws at the peak. This is really excessive for what I'm targeting but the end result is very nice. So, another adjustment is needed apparently.