Sil-Q Tileset

[MicroChasm]

I have been working on a tileset for Sil-Q for the past few months, and Quirk has been very patient in helping me put it together. I wanted to share some preview images with everyone. Let me know if you have thoughts on it!

Here is link to some screenshots:
Sil-Q Screenshots

I have two goals for this tileset: readability and accuracy.

By readability I mean that I want the tiles to be clean, clear, and recognizable even in your peripheral vision. No two tiles should look too similar, and they shouldn't bleed or blend into one another when adjacent. If you are quickly attacking through a group of orcs and Orcobal appears in the corner of the screen, you should see him immediately.

By accurate, I mean that the tiles should reflect both the Tolkien lore that they are based on, as well as reflect the experience of playing Sil-Q. For example, more difficult monsters are larger (more or less), and monsters with draining attacks are all purple to signify this attack type. Special monsters are more intricate and slightly brighter. This isn't meant to be a general tileset for Angband games; it is specifically a tileset for Sil. I reserve the right for artistic liberty of course.

None of the tiles should be considered to be final. I have made a tile for just about everything, but I am consistently tweaking, reworking, or redoing them.

I am open to critiques and criticism. Some of you have been playing Sil for a lot longer than me, or have different artistic sensibilities, and I'd like to hear what you have to say.

[mibert]

Cool stuff, thanks. Looking forward to giving it a try.

[Infinitum]

I like the color scheme! I'm having some trouble figuring out what the player character is looking at in the lower left though. A leather armor, a tangletorn (?) and 4.. wolves? The perspective seems a little bit off.

Also the orcs seem generally as tall as and bulkier than the player sprite, which doesn't seem quite right to me. I always imagined Tolkien's orcs to be more similar to "modern" fantasy goblins (which is also a synonym for them in Bilbo and other sources). The Uruks are described as taller (almost the height of a man) and more powerful, but those don't appear until the late second age iirc.

[tangar]

looks great. Is `stairs down` are from DCSS? (it's fun, as it's CC0; I use stuff from there too in my tileset)

please could you share some technical details? what the size of this tileset? which licence?

[MicroChasm]

I like the color scheme! I'm having some trouble figuring out what the player character is looking at in the lower left though. A leather armor, a tangletorn (?) and 4.. wolves? The perspective seems a little bit off.

Also the orcs seem generally as tall as and bulkier than the player sprite, which doesn't seem quite right to me. I always imagined Tolkien's orcs to be more similar to "modern" fantasy goblins (which is also a synonym for them in Bilbo and other sources). The Uruks are described as taller (almost the height of a man) and more powerful, but those don't appear until the late second age iirc.

I appreciate the feedback, and I'm glad you like the color scheme. The player is looking at studded leather armor, a grimhawk, and four wolves. There are some tanglethorns above them.

The studded leather armor is an early sprite I made... I will reconsider it if it is not clear. I've never loved it but I haven't figured out how to improve it yet. I wasn't happy with my wolves either, so I redid them right before posting. They are better than before for sure! Overall I am fairly happy with them, but I do think that some tweaking of the shading will make them more clear.

I think you are right about the orcs, especially the ones you meet early in the dungeon. I like this suggestion.

[MicroChasm]

looks great. Is `stairs down` are from DCSS? (it's fun, as it's CC0; I use stuff from there too in my tileset)

please could you share some technical details? what the size of this tileset? which license?

Sharp eye! The stairs are very very close to the Crawl stairs. I'd started with another design, but my stairs were difficult to tell apart at first and I felt that the Crawl stairs are very clear.

The tiles are 16px by 16px. I've been interested in tile art for a while but this is my first attempt at making a tileset, and anything larger than 16x16 felt a little intimidating. Constraints inspire creativity anyway, right?

If you want more details, I'm making the tiles in Photoshop, which I feel has worked well. I'm trying to limit the palette, so no more than five shades of any particular color on a tile. Also, a lot of reusing colors between monsters.

I'd like the tiles to be free and available for anyone to use. I guess the question is whether I want to ask for attribution or not. I'm more or less aware of the creative commons licenses, but I'd like to hear your opinions if you want to share.

[tangar]

Cool

Do you plan to resize it to 32 later on? It's always possible to downscale tileset in game settings...

[MicroChasm]

I don't plan to upscale the tileset at the moment. It's an interesting point, I think that I will experiment with up-rezing and seeing how I feel about refining the tiles from there. At the moment, I like the limitations of 16x16 px.

[Mechapede]

Hey, folks!

Great job on the tileset, MicroChasm!

I do share some of the criticism of the other fella, though - some of the tiles seem a bit unfocused, for lack of a better word (but perhaps this is just the result of not being used to the tileset).

In any case, I wanted to ask: will this be Sil-Q exclusive or available to vanilla Sil as well?

Cheers!

[MicroChasm]

I will have to ponder how to focus the tiles more-- it may be a lack of contrast, or just refining their forms more. It may be that I should let them go for a couple of weeks and come back with fresh eyes. I am very acclimated to the tiles right now.

The question of getting the tiles to work with vanilla Sil is a little complicated. I'm fairly new to Sil if I'm being honest. Do a lot of people still play vanilla?

Sil wasn't developed with tiles in mind, and many things were changed in Sil from NPPAngband. These changes created issues that need to be resolved in order for the tiles to work.

There are two parts to the current tile implementation (other than actually making the art): changes to .prf files, and changes to source code. I am changing the .prf files, which determine how objects (monsters/items/etc) are routed to tile locations, and Quirk is changing the source code. I can't speak much for Quirk's part, other than I know that he is working on fixing a crash bug with tiles on Linux. Also, artifacts and special items are handled differently than regular items and need additional back-end work to route them to a tile.

You could technically just replace a few files here and there in your vanilla Sil installation and the tileset would mostly work. In fact I tested it today to answer your question. There will be a few issues related to differences between vanilla and Sil-Q though. New tiles would have to be made along with slight modifications to the .prf files to accommodate those differences. This would be very achievable. However, without changes to the Sil source code, the tile implementation wouldn’t be complete.

Hopefully that answers your question.

[Quirk]

To add a little to MicroChasm's answer:

- Sil tile support is so broken on Linux that hitting an enemy crashes the game on my installation. I have traced this and fixed it but for Sil 1.3 there is no practical prospect of Linux tile support as I don't intend to continue maintaining 1.3 as a separate project.

- Listen won't have tile support on either Windows or Linux in 1.3.

- It's likely as testing happens that more code changes are going to be necessary. One thing I'm looking into at present is how to signal enemy status in tiles, particularly awareness status.

- Since Sil 1.3 won't be fully functional with tiles, it's a little questionable as to whether it's worth putting the effort in for cases where 1.3 and Sil-Q differ e.g. filthy rags.

[Mechapede]

I will have to ponder how to focus the tiles more-- it may be a lack of contrast, or just refining their forms more. It may be that I should let them go for a couple of weeks and come back with fresh eyes. I am very acclimated to the tiles right now.

The question of getting the tiles to work with vanilla Sil is a little complicated. I'm fairly new to Sil if I'm being honest. Do a lot of people still play vanilla?

Sil wasn't developed with tiles in mind, and many things were changed in Sil from NPPAngband. These changes created issues that need to be resolved in order for the tiles to work.

There are two parts to the current tile implementation (other than actually making the art): changes to .prf files, and changes to source code. I am changing the .prf files, which determine how objects (monsters/items/etc) are routed to tile locations, and Quirk is changing the source code. I can't speak much for Quirk's part, other than I know that he is working on fixing a crash bug with tiles on Linux. Also, artifacts and special items are handled differently than regular items and need additional back-end work to route them to a tile.

You could technically just replace a few files here and there in your vanilla Sil installation and the tileset would mostly work. In fact I tested it today to answer your question. There will be a few issues related to differences between vanilla and Sil-Q though. New tiles would have to be made along with slight modifications to the .prf files to accommodate those differences. This would be very achievable. However, without changes to the Sil source code, the tile implementation wouldn’t be complete.

Hopefully that answers your question.

Hey, mate,

It sure does answer my question - I dig tiles and am yet to try Sil-Q (in fact, I have barely devoted the time I think Sil deserves because life gets in the way so often), so I guess your tileset will likely make me jump into Sil-Q as soon as it is deployed.

Having slept on your tiles, I came back to provide proper criticism, as merely mentioned that tiles are "unfocused" is not nearly enough lol

So:

Some tiles feel unfocused - again, perhaps I feel this upon looking at the tiles as my eyes are yet to acclimatize** to what they are actually seeing (reading). In any case, yes, I shall emphasize that some delineation of the models (let's call monsters and the PC this way) could do, but perfectly I understand how tricky this is on a 16x16 canvas.

**I say this for today I had a far better time looking at most tiles and making the connection to what they are in reality, going as far as realizing that the reddish model found on the penultimate picture is a dragon. I know this sounds like it was an issue on my end as it now seems pretty obvious that it's a dragon, but yesterday for the life of me I could not figure that out. Also, it is relevant to mention that players will very likely be reading (as in, literally) whatever shows up as tiles, by means of the log. I won't ever be able to experience this scenario as I realized that the model referred to is that of a dragon, but if I were to play the game not knowing that said model is a dragon and [l]ooked at it, upon realizing that it is a dragon, I am fairly sure I would be able to immediately recognize it, as the brain often does these things.**

Wolves - read your post claiming that you did some work on them, which is good. I admit they were rather unconvincing as wolves, as they were. Perhaps you could post a picture of the up-to-date rendition, so that we could provide feedback?

Grimhawk - perhaps cutting a bit on the angle of its right wing (thus, the one we see on the left, as I take it the creature is facing west, right? Well, in case not, I mean the wing which is slightly higher than the other) could have the desirable effect of making it better resemble a bird. Another idea which would likely render similar results would be to simply make it look more bird-y and less monstrous (have a look at Infra Arcana's tile for the raven - quite an efficient rendition of a bird, if you ask me, so perhaps you could draw some inspiration from it?). And again I'll refer to the observation inbetween ** anyways as yesterday, upon reading that this was a Grimhawk, I could finally perceive it the way I did.

PC - I would not mind sacrificing some of its... attempt at all-encompassingness/neutrality/defaultness for the sake of some flair. Meaning, I feel as though the player character would benefit tremendously from wielding anything on their hands - be it a sword and shield, a mace and shield, even a single weapon without a shield (think dagger) would do. Or perhaps some tan/brownish colors on its body to convey a sense of armor/clothing being worn (any other color could do, so long as it matched the universe). So yes, there is the question of "why is my Elf wielding a mace and a tower shield?" or "why is my dwarf wielding two daggers (being bold here lol)?", but again, I feel as though the end result of such a push would be more beneficial than detrimental. As of now, it resembles something like #player.dummy.01 for any given 3D game, still lacking Texture&Lighting.

Tanglethorns - I like them as they are, honestly. They look effectively thorny, which is how they should look lol

Object to the right of the Tanglethorns on the first picture - what am I looking at? Greaves? Or a bladed instrument of sorts? Perhaps this could be reworked (given my question lol), but it seems far closer to "greaves" than "bladed instrument" to me.

Walls - I like them as they are, but perhaps making the bricks a little less aligned/symmetrical would be interesting, for the sake of giving them some of that rough-hewn flair that a place like Angband would possess. Perhaps adding a single black dot here or a little dash there could achieve an interesting effect here (again, I understand how a 16x16 can limit one's scope).

Arrows - Those are arrows on the floor, right? I am fairly sure they are, which is in itself a compliment to how they look right now. In any case, I... am not really digging their look, despite almost immediately recognizing them as so. Perhaps enlarging the left end of the arrow or trying to make the its head a tad more angular would be good.

Broken doors - I feel they look as much as broken windows as they do doors. I think this is caused by the representation of the damage it suffered - that is, a hole in the middle. Maybe reworking them so they show signs of depredation on their right side (as it is the side opposite to the hinges, thus more susceptible to damage) rather than as a hole in the middle would be interesting.

Object to the left of the dragon and the PC, penultimate picture - what am I looking at? I do not have the slightest clue as to what that might be... :S

Color palette - How are we in this department?

Volunteering - the last time I went to GIMP to try anything was in late 2018, but the end result satisfied me and most people who saw it, included the game's developer. This is what I am talking about:

https://imgur.com/sjPSTPG

It is my rendition of Infra Arcana's church. I am no expert by any means, but if you feel as though there remains a plethora of small things to tackle on most tiles (as often happens), I could help tackling some of those for you. No worries on rights/acknowledgements/credits - they obviously remain yours. Still, I have a full-time job and a nine-month old, so assistance coming from my actual hands could be... sporadic lol

Cheers!

[Mechapede]

To add a little to MicroChasm's answer:

- Sil tile support is so broken on Linux that hitting an enemy crashes the game on my installation. I have traced this and fixed it but for Sil 1.3 there is no practical prospect of Linux tile support as I don't intend to continue maintaining 1.3 as a separate project.

- Listen won't have tile support on either Windows or Linux in 1.3.

- It's likely as testing happens that more code changes are going to be necessary. One thing I'm looking into at present is how to signal enemy status in tiles, particularly awareness status.

- Since Sil 1.3 won't be fully functional with tiles, it's a little questionable as to whether it's worth putting the effort in for cases where 1.3 and Sil-Q differ e.g. filthy rags.

Infra Arcana does this by means of coloring the background of each model's tile, and it is quite effective once one learns the color coding: blue is for unaware monsters, purple is for stunned/confused, green is for allies, etc.

[MicroChasm]

Wow, thank you for the in-depth critique. It helps to see these images from someone else’s eyes. I started writing a reply which addressed each concern individually but it was turning into a novel. I do agree with most of your input, especially concerning the player tile and the doors.

The walls are a little tricky because any added visual interest can look strange when repeated over and over. There is currently some slight tile variation, but I will definitely experiment with it more.

You were wondering about the tile to the left of the player in the penultimate image-- it is a pile of rubble. As for the tile to the right of the Tanglethorns-- it is a curved blade. I think a couple of extra pixels in the handle might make that more clear.

The color palette is an interesting question. I set out to keep the palette limited and desaturated. As it turns out, we need to use an indexed 256-color .bmp in order for the tiles to work correctly on Linux. While this is a limitation, I’m actually glad because it forces me to stick to my original ideals.

Speaking of ideals, I mentioned the idea of clarity in my original post. My feeling is that very detailed tiles are at risk of blending together; I want tiles with areas of solid color that form lines and curves that are easily identifiable. The grimhawks are a good example. Basically I’m trying to capture some of what people like about ASCII. I have nothing against more detailed tiles, or less detailed tiles, I’m just making the tileset that I personally want to see in the world.

My tiles are a bit of an experiment really, and this may be the cause of them feeling unfocused. Now, having said that, I don’t want to use this as an excuse to make ugly tiles. I believe that they can be made both clear as well as beautiful with enough work.

Here is an example of four roguelikes that I think are good examples of how form, color, and large vs small frequency detail can affect clarity: Tileset Comparison. You probably recognize DCSS, TOME, and ADOM, but the one in the top left is called Rescue Timmy. I really like the style.

I appreciate the offer of help! Your rendition of the Infra Arcana church is actually very impressive. I think it may be helpful to let you take a pass. But I don’t think the time for that has come quite yet. Although I’ve been enjoying my time with the tiles I do have other projects-- I’m planning to let them go for just a bit and see if there is other feedback. Major changes to artwork are always less painful after some time away. After the second iteration I will let you know.

[Bill Peterson]

When Sil-Q was starting to stabilize I made an attempt to get the David Gervais tiles, as I've preferred that tileset since the days of AngbandTk. After making some good progress with the .prf files I ran into a roadblock. The person who did the original Sil was apparently not interested in tiles at all and put definitions of ASCII characters in various source files with no connection to the .prf files. Not really wanting to try to change these source files, especially while Quirk was modifying, I gave up on the project.