With all of this painstaking work I have been doing with restoring artwork so that it looks presentable in Mugen, I decided it was about time to make a thread highlighting what was touched up.I just finished Chun Li from her Revival style, and her results are quite pleasing. Here's her end result:**I am currently at work, so later on today I will post before and after screens of her templates**I started out with her main image that is about 550x550 pixels and began with cleaning up any black levels, rough looking patches, and what I call "jepg pixelation" that is very common in your generic Google image search. I see that a lot of aspiring portrait makers fail to realize these little details, so when they index their picture, tons of clarity is lost. I also spent a solid nine hours on her template alone, with an additional eight on her 240x280 rendition you see above.I also understand that 1.1 allows for 32bit png files. I'm going to remain true to my hard working ways and forgo that type of render. Instead of going through a bullet list of why I'm not, proof is in these below screenshots.All of what you see below is indeed indexed on 1.0:The way I see it is that I would rather make a governing template that could be used for just about any other format, whether it be a 1.1 release, a 1.0 release, or any size in between. Those MvC2 and SFxT images you see were indexed at their native size and cleaned up so that I could easily transform them for Big Boss' Nexus Screenpack and Kyokano's SFxT Screenpack. I'm a stickler for quality, so no pixel will be overlooked.As work progresses, I'll gladly post what's up. Enjoy!!
YES.I can't wait to see their glorious restoration! Thanks for all the effort you put into your portraits, Saikoro!
OK. This is my first attempt to show all of you what goes behind the process of my portrait releases. I saved the below two examples in the form of a .jpeg file, and it takes away from what I intend to point out quality wise. From the next share onward, I'm going to save in a clearer format. Silly me and my first attempt. Here is a sampling of Chun LiAnd here's a sampling of T Hawk**For best comparison results, save these to your desktop and look around with a close zoom. You'll then see what I'm talking about below**When starting any portrait, I first go in and make sure that every possible black level is clean without any distortion. When you zoom in and look closely, comparison wise, you'll see a shit ton of speckles that look atrocious throughout. Hence why I do that first. What follows is erasing/separating the portrait from it's white background while making sure it's edges look smooth and cohesive. I also make sure black levels of those edges are even as well, giving the illusion of a complete image that wasn't erased from it's background. I then go in and make sure that all fleshtones, color lines and gradients look as clean as possible. Note the complete and thorough lack of pixelation (There may be a slight amount here and there, but it's very minimal if so) in all of those areas. Also, I try to make all of the primary clothing areas, which are blues in this case, as clean and as pixel free as possible.Both Chun Li and T Hawk here took about 8 hours a piece to properly clean up. After that entire process, the portrait is free transformed to fit my signature 240x280 portrait size standard where further cleaning is done to finalize the image. I go the extra mile to ensure that colorloss from indexing and free transforming is non-existent. I'll post a comparison shot of T Hawk's indexed 240x280 image upon completion.Lastly, since my Marvel VS Capcom 2 and Street Fighter X Tekken sets, I have been saving each piece as a template so that I can transfer those images to any format if needed. Doing work this way saves a lot of time even though further indexing may be required if looking to make something for 1.0. With the release of 1.1, these templates can easily be manipulated to fit any screenpack with minimal correction outside of edging since there isn't ant colorloss. Epic win either way!! Thoughts and feedback are always welcomed. And apologies for posting in .jpeg format!!
