Icy walls of bleh

A little before and after pictures. There are a total of 5 ICE dungeon texture SETS. As you can see in the below images, they are mostly just recolored variants with different symbols. SIDENOTE: I presume they are Daedric symbols commonly seen in later games.

Here are three of the originals….

It took a bit to decide how much different, I wanted each to be. Ultimately, I just decided on to make them vary enough so that they would have their own appeal but maintain the overall theme. I haven’t seen these dungeons in game yet (never played far enough I presume). Once I complete all the SET files, I will do a test run through the game to see if I can find them (and others).

And the same there afterward…

Where do I start….

Design Document


Texture mod for The Elder Scrolls 1: Arena



  • Create the (first) TES1:Arena texture mod replacing the down-sampled blocky artwork with smoother less pixelated textures that is hopefully appealing and unique.



  • Create a new unique art style for game
  • Improve texture variety by replacing duplicate textures and ensuring that each texture set and image is unique and stands out from the others.
  • Replace all textures.
  • Create a more cohesive theme amongst textures.



  • TES1: Arena is a very old game and uses unconventional file formats and require special programs/methods to open.
  • Most textures are only 64 x 64 pixels (or in the case of SET files, groups of images with those dimensions) and cannot be resized. This limits the amount of fidelity that can be achieved.
  • Textures are limited to a (external) 256 color palette.
  • In game, many textures in the same SET are placed side by side in seemly random order. That means most textures from the same SET (and their door images) have to be seamless with one another.
  • Some textures are compressed in an unknown way that has yet to be deciphered.
  • The game engine assigns textures based on existing INI files for each type areas but the texture sets aren’t matched very well in multi-level areas. Many areas use different texture SETs for each level but they appear almost arbitrary in how they were selected.  It would take meticulous remapping of all the INI files to create more cohesive appearance.



  • Hallfiry’s Arena Suite– It is a set of tools that allows bulk extraction and reinsertion of assets from the Global.BSA (Arena’s game asset container file). It also converts all the non-compressed textures into PNG files for easy editing and then converts them back when reconstructing the BSA file.  This program allows me to have one working directory. When I need to test my work out in the game, I just create a new GLOBAL.BSA from that directory and replace the one in the game directory.
  • GIMP – Freeware alternative to Adobe Photoshop



  1. Complete all SET files to have “Alpha” status. (95% complete)
  2. Complete all Wall and Ground IMG files.
  3. Complete remaining IMG files to “Beta” status.
  4. Tweak textures that don’t show well in game or don’t match.
  5. Fix any individual texture errors (e.g. rogue pixels or misaligned textures).
  6. RELEASE 1.0 (not set but likely on Tesnexus and ModDb.

The tale of two sets

Two more textures in timelapse. One very easy and the other still easy but a little more work.

– Martin

A little bit of paint

This is a timelapse of me retexturing one of the Mage Guild exterior sets. This one stayed pretty close to the original and I liked how it turned out. I did later on decide to make the brick darker red for a more striking look.

I have a couple more that I made but haven’t posted. I’ll put those up shortly.

– Martin

Who’s steering this thing?


One thing I noticed recently on my work w/ Arena is that I don’t really have much of a game plan other than “finish Wall SETS then IMGs”. Considering how much project management is part of my work, I found it odd that I didn’t implement it here. After thinking about it awhile, I came up with two reasons: 1. When I’m home I don’t want to think. 2. This is my hobby so I have never been too worried about how fast or slow it goes…I just do it for fun. Incidentally, for my Torchlight project, I didn’t have this problem (well not to the same degree). In that project, I tackled one levelset at a time. I marked textures to see where they fell in the game ahead of time and work one of them till I was happy how it looked in game. Only then would I complete all the others in that level set. I even had some of the templates saved for similar creatures to minimize duplication of effort.  There was still room for improvement though. I was terrible about writing down and special techniques or tricks that I used. Because of this, if I held off and came back later, I forgot how it was done.


Realizing this now, I feel their are definite ways to plan a little more accordingly. Hopefully in the next week, I’ll enough time to work out a little design document. It might be the next post, not sure yet. I plan to work out the vision, objectives, goals, etc. It might not make too much difference but I think it will help keep me on track with what I want from this project (and I can develop the habit now so the next project will go more smoothly).

Line them up and knock them down

Start strong

As I have stated in the past, one of the most taxing aspects of editing the TES:Arena textures is coming up with a unique take on the image within the constraints of the limited pixel and color count. While I was thinking about my methods and ways to be more efficient, it occurred to me that I don’t appropriately leverage my motivation. For the last 1 1/2 I have been working one texture at a time. This would exhaust my creativity for the day too quickly. I would waste a bulk of my time “finishing” said texture after I decided how I wanted it to look. All of the “fresh mojo” that I had when I started for the day was worn away by the time I got through a few (sometimes one) SET files. This is especially true with the most of the remaining SET files that have no defined look or image (i.e. barely recognizable as any thing but random colored pixels).  Not only was that detrimental to the quantity of work completed in any one sitting, but it also deterred me altogether. There were days that I just didn’t have the energy or initiative to try and figure out a new take on what looked like white noise so I just wouldn’t decide to spend my efforts elsewhere.

Shotgun effect

To better capitalize on that “creative fronted”, I tried an experiment. I loaded up 18 SET files in GIMP. Then one by one, I worked on just getting the look I wanted for a small portion, enough to establish the look. I didn’t waste time completing the whole texture as that is really the easy and mostly mundane part. It worked quite well. Using this method, I was able to solidify my design for 16 SET files in just one day (really just a couple of hours). As I said the design is the hardest part with this project, and I wiped out almost half of the remaining SET textures left. Now, if I don’t feel particularly creative but still want to make progress, I can just finish up some of the ones I have already started (template on itself) and if I do…I may just be able to finish the other 16 or so.

– Martin

If I am Ahab….

White Whale

I have noticed on my various projects that there are always one or two things that cause problems (besides technical issues). For my texture mods, this tends to be a (or “a few”) texture that I just can’t find the mojo to complete satisfactorily. Usually when I work on such textures, it just turns into a series of mis-starts and redos. Sometimes (in the case of Arena), it’s that I have an idea of what I want to do but either I struggle with how to make it happen or it doesn’t materialize as I wanted.

My bane for over a year
My bane for over a year

Not without a fight

The one texture in Arena has cursed me for almost a year is DWJ02. Don’t let the plain innocent name fool you. It taunted me with it’s ambiguous rocky look. During the many hours I toiled on it (sadly not an exaggeration), I aborted multiple attempts to complete it and in several instances started completely over. It should be noted that this highly specific artist’s block is in no way logical. I can complete several similarly abstract textures in a single day and still would not make much, if any, progress in DWJ02. I think the problem was that I locked in how I wanted to have it look pretty early on but every time worked on it, I just didn’t feel quite right or I felt it looked too similar to other textures.

Final...it's done.
Final…it’s done.

At last, victory

Last week, I barreled down finally and finished it (despite my son accidentally shutting down my computer when I was 2/3 done with it…curse….curse). The final result is not even a particularly sexy texture, although I do like it, but really that wasn’t the point. The point, obviously, was that I could not let it beat me. I faced my demon and I struck it down. Now what it this DWO02………………


– Martin

Ooh…Shiny keys

WARNING: Potentially boring technical stuff below. Oh and it’s probably unnecessarily long

Whenever I’m feeling a lack of drive to work on a particular project, I tend to stray to something else till that motivation returns. Although I strive to stick to one project, I do need brief diversion to replenish my enthusiasm now and then. Sometimes, that means working on one of my other projects for awhile. Like this week, I took a brief detour back to my Minecraft project. But other times, I just let my more analytical side run wild…..


Besides art, I really enjoy the technical aspects of computer programs. I’m perfectly at home trying to decipher file formats and learn out the hows and whys “behind the curtain”. Sometimes it can give insight into the decision-making that led up to some design choice or the other (e.g. how SET files are mostly walls and floors but the outside ground textures are individual IMG files even though they are used in a “SET”-like fashion). Other times, it just gee-whiz knowledge such as  Darkstone can handle high resolution texture or that TES:Arena’s root directory is also the file override directory (overrides the files for the game). I freely admit that I just enough to know I don’t know enough. I can analyze files (e.g. using a hex editor to examine the file byte by byte) but not enough to find the answers I want.


That is how I discover odd things here or there about games that I’m working on. Quite awhile back, I discovered that the Arena executable file (a.exe) was compressed. This immediately led to me wanting to uncompress it. It didn’t take me too long to find a dos unzip tool then I just loaded Dosbox then ran the tool from within the dos environment. It was fun for me just to have gotten that far. The executable still worked but it wasn’t as if that opened a treasure box of discovery. The main thing I found was text inside the executable that is used in game however I don’t know if editing it would cause something to break (might be worth trying sometime). Sometimes compression isn’t about size but speed. The computer can load more into memory then uncompress it on the fly. However, that isn’t really relevant to TES:Arena anymore 😉


Occasionally  when I’m bored, I load the Dos debugger version of Dosbox. It’s basically the same program that emulates Dos but with a debugging window so you can see “what’s going on in the background”. While I don’t have nearly enough skillz to understand how to copy memory from registers and other such jazz, it has shown me what files were loaded when  particular dungeon or city was entered and that has proven helpful a few times already as I tested my mod. I know it’s possible (in theory) to use the debugger to find the routine for the IMG files that use that difficult compression. Also, one might be able to wait till one of those IMG files are loaded in memory (and thus already uncompressed) and then pull it out of the memory register….no idea how to do it though. (I told you I know enough to know I don’t know enough). I would really like to see if the IMG files would still work “uncompressed” but don’t know how to make that happen.


All TES:ARENA levels are detailed out in the MIF files while the textures, sounds, etc. that they load are listed in the INF files. Think of a MIF as locations outlined out on graph paper. They determine the layout of all locations: cities, dungeons, stores, etc. INFs are more like the map key dictating what the dungeon will look like aesthetically.  With preliminary exploring with a hex editor , I found each MIF has a specific INF listed in it’s “code”. So when a MIF is loaded, it searches for the INF and loads all textures and sounds from that particular file. If we ever manage to decode the MIF format, we could theoretically create new levels and hand design them too. Occasionally, I’ll load one up but honestly, but I don’t get far (it doesn’t mean I can’t have fun trying).


INFs on the other hand, are all plain text. You’d think that would make them easy to decipher but the format is only partially obvious.  I have fiddled with changing pieces here and there to see the effect in game but only a little and I haven’t found much. There are some odd switches that didn’t seem to have any (obvious) effect on the starter dungeon. I will probably map the INF file structure out at some point when I have more time (ha ha). The bigger problem is that I think the INFs maybe loaded with extraneous junk that doesn’t get used in game. For example, the starting dungeon only has one floor and one exit with no entrance but the file not only list an entrance file but also 6 alternate files sets as well a listing of all the monsters in the game even though only a few are used in that level. My theory is that they used a base template to start with then altered it from there. This confuses the matter if a texture is used in game or not if each level list many textures that aren’t used for that level. It could be trimmed down if I we knew what the structure of the MIF files and could see if those listed IMG files are actually used in that level. Maybe in time.



Kicking down the door

Interior doors


Last week when I was lacking of motivation to work on the walls sets, I decided to fiddle with the interior doors and more specifically the wooden doors. Those are the ones that connect rooms and hallways and not ones leading out of buildings or dungeons. These particular doors are one single texture that is a door (the entry exit ones have borders that match the interior walls). That door occupies one square of dungeon space (all dungeons are square chunks of one texture). While in-game it appears that there are only two or three different wood door textures, there are actually over a dozen. It’s just that there are only 3 different  textures among them. Here’s a sample of one image among multiple IMG files:

5 Doors don’t have to equal 1

What I don’t understand is why Bethesda used multiple files with the same IMG when space and memory was a concern back then. They could have easily used one file through the INF files (the ones that decide what textures to use on each level). However, this does give me an opportunity to add some variation to the ingame art. As such, as with walls and etc., I will make each door appearance different.


Where is Bob Ross when you need him?


Color balancing

After wandering around in the game, I started feeling a need to go back and tweak some textures. I was just not happy with some of the combinations. Sometimes it’s just best to strike while the inspiration is strong. So I have spent quite a bit of time adjusting colors, cleaning up some of my more “questionable” decisions, and generally making them all play a little better together. I especially tried to tone down the floors and ceiling so they don’t clash or draw away from the walls. Some of the texture combos are much easier on the eyes now. For some of them, I removed the splotches or other weird marks that I had left to stay close to the original…the resolution/texture sizes are just to low to be that craz…er creative.

Breaking away

I have stopped trying to adhere strictly to the source material in interest of keeping the textures unique and fun. Additionally, I plan to add some “detail” to the wall sets that are just a group of plain walls  (e.g. a small object on the wall). I want each wall image in a set to be unique but I’ll be sure to keep unadorned walls too to balance the aesthetics.


Floors darkened to not match the tables and added minor definition to the ceiling
Floors darkened to not match the tables and added minor definition to the ceiling


Removed splotches and added shelves
Removed splotches and added shelves


Floor tamed, darkened walls and trim at top to better see the gold (now just need to do the ceiling)
Floor tamed, darkened walls and trim at top to better see the gold (now just need to do the ceiling)