Blog

  • Blakestone HD textures sneak peak

    Blakestone HD textures sneak peak

    Recap

    I have been working on remastering the textures from Blakestone on and off for several years now, initially dumping and light upscaling them for ease of redesign. From there, I started hand drawing over the old textures adding detail and transforming a blurry pixel-y mess into what you see below.

    A key goal is to maintain the spirit of the original pixel art and only add detail that enhances or complements the composition.

    The human NPCs are a pain in my rear

    Some of the images, such as above, are fairly straightforward and are easy to interpret for the new design. However, some of them aren’t the greatest in the original format or so low resolution that they require more creative effort to bring to life. The scientist’s face, for example, was a blobby mess that lacked defined detail in key places. I had to envision what the new one would look like and that too a lot of back and forth experimenting to get the vibe I wanted.

    Setbacks

    I do most of my work on my IPAD. I somehow corrupted a couple of my sprite sheets that I had already completed (but like a dummy hadn’t exported yet). That was about 150 completed textures (several of the monsters) of the thousand or so I’m redoing. It was a brutal motivation killer. I had to put the project aside for a while till the pain of it faded and I regained motivation.

    Now

    As with the TES Arena Depixelization Project (ADP), I’m slowing picking steam back up. I am also long overdue for a few sneak peaks. Please note, that these are still alpha. The combination of trying to keep the same shape and vibe of the amateurish looking originals and the fact that my versions are drawn on the IPAD “look” may seem slightly not ready for primetime (i.e., like a kids notebook drawings) but I have some post drawing steps to take to up the quality some more.

    SIDE NOTE: Some of these monsters are incredibly terribly designed. I wanted to change a few because of how childish they looked but decided to stick to the original version. So the goofy monsters are still goofy.

    Progress

    I have quite a bit done, most monsters have been completed or final designed. The initial security NPC is done, the scientist is 1/2 the way complete, but the other three are still pending. All miscellanceous textures are complete. The walls will be last but are the least difficult to fix.

    Enjoy

    Martin

  • ADP lives….yet again…well, technically it never died.

    THE RIGHT NOW

    Let’s just ignore the fact I haven’t posted in quite a bit (or that I started this project 13+ years ago). To be fair, I have released three other Arena mods in the meantime (more on that in the future).

    I recently started digging back into the Arena Depixelization Project (ADP), my graphics replacement mod for The Elderscrolls: The Arena, and I have been hitting it hard.

    I’m doing a recorded playthrough looking for missed art in the game and general quality control. This has led to me not only finishing missing textures but also tweaking (and sometimes overhauling) 90% of the work already done. I guess over 13 years of game art fiddling has matured my style slightly.

    THE SWAMP

    I first had to corral all my work over the last decade+. I had 7 different versions saved through the year and some files went missing so I had to extract them all compare and then pull any missing forward. This project was quite a mess (no file discipline at all, the shame). Then I had to relearn the tools, but that didn’t take too long. Then I created a new working directory and archived all the rest to clean up my sloppy work.

    In just 2 weeks, I have probably edited almost 100 textures already. As I record my playthrough, I save when a missing (or ugly/mismatch) texture is found and either fix it ASAP or catalog a bunch at once to do in a surge.

    GALLERY

    Here’s a slew of screenshots I made from the most recent version.

  • Blake Stone gets a glow up (PART 3)

    InGame

    Now the naming and image dimension issues have been resolved, so I loaded the game and it worked perfectly! Drunk with power, I ran a quick and dirty (i.e. impatient) bulk upscale on every texture just to get the payout of seeing that high resolution quality in gameplay. I wish I had a picture, but I don’t sorry. So you’ll just have to settle with one of my test pictures. The ultimate result of that initial effort was a messy, uneven upscale that wasn’t very good.

    WTF is upscaling?

    I’ll take this hot second to address this topic for those not in the know. One of the first “AI” booms was image upscaling using an algorithm model called ESRGAN (it stands for something, but no one cares). There are many models now, but that one paved the way. Note, when I say “AI”, that’s really just the media branding of this technology because the technical stuff is not as sexy. It (and most of the current AI stuff) is nothing more than highly sophisticated pattern recognition software. You teach it what 200 apples look like in low resolution and high resolution. It learns that when it sees a low resolution apple what the approximate of a high resolution apple might look like and gives you that. See how boring that already got….hence “AI”.

    I use a program called CUPSCALE, which is a frontend for the tech behind it. It allows bulk upscaling, supports most common image formats, and allows switching between different models. Best part, it will install all the needed files (python and all things related) for you, so it’s as close to plug and play as possible.

    Almost a mod?

    I spent many hours trying different combinations of upscale models. Finally, I settled on a workflow that best accomplished the cartoon look I wanted. I started with 1X upscales (meaning no change in size) just to clean up details. I focused on de-dithering, which was often used in old games and extremely low rez art, and sharpening detail a little, so colors weren’t bleeding as much (refocus model). Then I scaled it by 2X using 4XNickelBack model and reducing 50% afterwards. After much trial and error, I found this was the best option for Blake Stone art as it was the least destructive to the essence of the picture. After that I hit them with a couple rounds of a FSManga model to make them more cartoony, with a 50% resizing here and here. Ultimately, I ended up with game textures that were 6X the size of the originals that had a more cartoon appearance.

    UP NEXT TIME….A new mod is born!

    Part 4 will be about releasing my first full HD texture mod. But it doesn’t stop at Part 4 (wink wink).

    – Martin

  • “Extracting art from Blake Stone” is not a very funny title (Part 2)

    “Extracting art from Blake Stone” is not a very funny title (Part 2)

    Collecting Pixels

    Step one is figuring out how the game stores the art files (textures). Normally what I do is check for the largest files, as they are usually the texture and/or sound archive. However, with Blake Stone, I knew that this game used Wolfenstein 3d-style archives. Blake Stone’s format was slightly different and used the extension BS6. But it was close enough that Slade, a popular tool for actual Doom and Wolf3D games, worked perfectly to extract the files.

    FUN NOTE: I started this many months ago. I JUST noticed that BStone has a command-line option to extract textures. #facepalm #hashtags #DOH

    You may call me…”Tim…er… I mean SPRITE”

    The BStone program looks in the “aog” folder for any textures (Planet Stike has it’s own folder, “ps”). If it finds a texture there, it will use that texture instead of the original. Unfortunately, BStone only supports sprite and wall files. That means some elements can’t be replaced currently (mostly UI and cut scenes). For Blake Stone: Aliens of Gold, there are 1017 sprite and wall textures but there isn’t a folder structure. They all just go into that one folder which makes it easier to manage.

    I extracted all textures to a working folder and hit it with a quick and dirty upscale so I could check how it works in-game. After dumping them into the “aog” folder, I booted BStone and enabled “External Textures” in the option menu. It didn’t work. Turns out that the extracted files were names were very slightly misnamed when extracted (SPR instead of SPRITE, etc.). It was easy enough to fix with a utility called Bulk File Renamer. A quick touch up and I reloaded BStone to see if that fixed it.

    Stretching them on the rack (by rack I mean in Photoshop)

    It worked…sort of. The textures were definitely upscaled but many of them looked squished (skinnier than normal). Quick glance at the readme revealed that rendering of sprites was not 1 for 1 in-game.

    After a little experimenting, I realized that every texture just needed to be perfectly square. Maybe the extraction method was flawed, not sure. To fix it, I just made a quick macro on my Logitech keyboard (frequently my savior for repetitive tasks), dumped 100 or so textures on photoshop and ran the macro (it resized the canvases to be 64×64, saved and then closed the file). I just needed to hold down the macro button and let it work through each file. After finishing all 1017 textures, I saved a backup for the inevitable time I make a drastic mistake. (I will share the reformatted originals when I get around to doing Planet Strike)

    Up next time…..Version 1 of my Blake Stone HD texture mod

  • Wait! I’m upscaling a 1993 game?! (Part 1)

    Its more an obsession at this point

    I have a weakness for mucking around trying to create HD art packs for games. The bulk of them don’t get completed, partially for technical reasons, but sometimes because something else catches my eye. Others are technically “done” but just aren’t up to my personal release standards. I can’t help it and my ADD frenzy usually kicks off when I see low resolution or blurry artwork in a game. Blake Stone by Apogee is a perfect example of that. I am not the biggest fan of pixel art, so it was an ideal candidate to be upscaled.

    I forgot it even existed

    I should note that before this, I never played Blake Stone or even remotely desired to play it. I had barely even remembered it because Doom overshadowed everything at that time. But I stumbled upon a source port for the game. The project called BStone, an open source port for Blake Stone: Aliens of Gold and the Planet Strike sequel, is a windows port for the DOS game.

    Wait it does what?!

    As I love game engine ports because they usually bring features and Quality of Life improvement to older games, I couldn’t resist checking it out. A quick search revealed that I actually owned both games on the GOG platform (I have over 2000 games…so….whoops). So using the BStone, I gave the loaded up Aliens of Gold. While tweaking the settings, I saw a few interesting options. First, you could use image filters which are OK but tend to just make the game look blurry to me. Second, the “upscale filter” applies an XBR filter that looks pretty nice for what it does, however there are limitation to how good pixel art can look that way. Last, there was an option to turn on “External textures” and…the fuse was lit. I knew that by just “merely” replacing the images, I could make the game look so much better and my obsession was triggered.

  • 2023 update…so…yeah..it’s been a smidge

    Peek-a-boo! I’m back and I have a ton of updates on the art side and just a few on the writing side. Lots of life changes slowed me down for a while. Though this might be a boring update, I’ve been gone long enough that it needs addressing. Sorry 😛

    THE WEBSITE

    It’s about time for an update to this website for 2023. Let’s hope I can make the next post before another 2 years pass. Though, I would like to transition to another web service eventually, but I’m not sure when that will happen. I need to figure out how I want to organize all the posts and what kind of menu it will have. I’ll keep this URL though. I kind of dig it 😉

    The idea is to make ArtInPinkerton.com the central hub for my various online activities and consolidate my nerdy world under one umbrella. Naturally, the key to “big plans” is the carry through and that can be the tricky part. A recent programming revelation has made me realize that the biggest barrier for me is the logistic preparations that have to take place before I get to do the thing I actually want to do. For example, I find WordPress clunky and the thought of having to wrangle it for when I write an entry often is enough to put me off.

    Right now, I want this website to highlight my various artistic endeavors and hobbies. Off the top of my head, that will include:

    1. Writing and stories
    2. My YouTube channel Martin Saves the Universe (video game play and modding)
    3. Painting and miniatures
    4. Boardgaming
    5. Podcast (Nerds and Normies, no longer active)
    6. Technology
    7. Other

  • So I want to be a writer….

    So I want to be a writer….

    Starting somewhere

    As the title suggests, I am going to pursue being an author.

    Upfront, I technically settled on this decision back in August 2020. But, I have a spectacular social anxiety coupled with a healthy dose of self doubt that delayed this “official announcement”. I also found many reasons to procrastinate, such as revamping this very website. I now acknowledge that I can’t wait till everything is perfect to begin. So here I am…my name is Martin, and I am a writer.

    I have been planning on doing this post for a long time. My original idea was to simultaneous post on this website while also releasing the first of a video log series on a YouTube channel called “I Dream of Electric Sheep” (logo seen above). I meant this blog and that channel to be complimentary pieces journaling my road to becoming a published author. NOTE: That is still my goal, the deadline has just slipped.

    Why a writer?

    I have always, and I mean always, seen things differently than many people. I tend to see the world from a different angle and built my entire “wacky/weird” sense of humor on that skill. This isn’t me trying to say “I’m not like other guys” or that I’m special; just that, it allows to me take ideas and put interesting spins on them. I have only ever shared these ideas with family and friends (and then usually only fragments). But like a writer, these stories stick with me and over the years and I developed them more in my head space. I think writing will not only allow me to share them with others but I will also get to relish seeing how my own creations grow and take a form I may have never expected.

    What can we expect?

    I have a lot of practice writing to do before I can write a full book. So up first will be smaller stuff, some from ideas I committed to for my podcast; others will be stories I spit-balled with friends. Either way, my stories will be like me, a little off and hopefully fun. I’m ok with that as I think it lets me bring something new to the table (or at least something less common). For example, my first practice story is Hogwarts Fight Club: yes, a fanfiction and it’s kind of exactly what it sounds like but even more bombastic (but also with a plot). Another, I’m calling “Rapt-her”, is an action story about a teenage girl with spliced Velociraptor DNA who fights Nazis. You get my point. They seem odd and silly, but I want to write them with heart. I feel even the most off the wall concept can be good if you treat (and write) it respecting the idea behind it.. Just warning you ahead of time so you’re not shocked when Dino-Hitler arrives on the scene complete with funny mustache.

    Why now?

    I’m approaching 50, retired from the military, and have NEVER WRITTEN A SINGLE STORY. I always had the distant notion that someday I might write a book, but the idea was a like a distant mirage that never seemed in focus. Most stories just swirled around in my crazy noggin. Though some saw daylight through other mediums (art, D&D, game modifications, etc), I never felt the urge to write them to paper.

    Even when I started pitching fanfiction ideas as part of my podcast, Nerds and Normies, it felt more like an academic exercise than a “need”. In August 2020, I was in the telling my friend about my first podcast story idea (Hogwarts Fight Club) when, mid conversation, a switch flipped and I just knew that I wanted to be a writer. It wasn’t a “stop the presses” moment, but more like a there was a familiar presence that had been hanging back for a long time, waiting for me to feel ready only to step forward when the time was right.

    And just like that, I was all in.

    – Martin Pinkerton

  • And now for something completely different

    And now for something completely different

    I recently decided to pursue being an author and am very excited by this new passion. With that, I will redesign my website to focus primarily on my misadventures in writing. However, the plan for now is to still keep room to highlight my more visual based art too. It just won’t be the front page material anymore. It is very possible that I move away from WordPress but am unsure where or what that new page will look like (structurally). However, the URL will stay the same.

    I appreciate all who followed my very niche little art blog for all these years and warmly welcome you to stick around. If you prefer to not, that’s fine too, as I did kind of pull a switcheroo.

    For anyone interested, I plan to write fiction, to include science fiction, fantasy and horror. I have a long way to go and a lot a practice to get in. You can expect at least some of it will have my trademark humor for silliness (it IS trademarked, right? no? OK copyright pending).

    – Martin Pinkerton (Art In Pinkerton)

  • I should punch myself in the face

    I should punch myself in the face

    256 color limitation

    One of the tricky aspect of working with Arena textures is that they are only 256 colors AND that color information is indexed from a palette file. That means the image doesn’t have the color info in the file, it has a reference point to a master color file (this saved valuable space in the old timey 3 1/2 floppy days). Colors in an image have to be one of those exact colors in the palette file. To work around this, I have been only using colors in whatever image I was working on. It worked, but it can limit in being creative. I did most of the project this way.

    Palette image

    DOH!

    Now, but a scant mere 9 years later, I realized to import the color palette in as a raw image similar to how I first did early in my hobby working on another project. Once imported, I could create a Photoshop swatch of all the working colors and boom. I’m blond, what can I say.

    Palette swatch

    More progress on actual files next update, although you can see a preview in the image above 🙂

    – Martin

  • Time to clean up the mess

    Time to clean up the mess

    Begin again

    So, I’ve finally got my stuff together, life has settled down, and the mojo to get back to my projects. I have a lot to post on other art projects I’ve done in the past (cough) couple years, but first, I figured I’d give an update on where I am with the Arena Depixelization Project. Yes, it’s still on.

    screenshot005

    Testing

    I have loaded up the textures into OpenTesArena. It allows very easy testing of levels without having to play the game. You can just select what level at the menu screen and load it. Right out the gate, I’m thrilled with the starter dungeon. No need to mess with that.

    screenshot003

    More testing

    However, the wall textures in the very next main quest dungeon are too jarring. So I have already toned them down for next iteration (not pictured below).

    screenshot000

    Then I skipped to the 3rd main quest dungeon and found these round blob walls that just don’t cut it. So next I will clean those up. My will systematically go through them all until there are sufficiently aesthetic pleasing (or tolerable) then I will release ADP 1.0.

    screenshot001

    Limitations of Arena

    There will be limitation right now, as the texture combination for some levels is very off. Maybe one day, OpenTesArena will allow us to edit the level files and add new textures. Then we’d be able to make matching or complimentary “tile sets” for each level, adding cohesion and variety overall. But, until then, this is what we got.

    – Martin