So a fellow dev has shown me how powerful the image showing tools are in Inform. By using transparency and some clever base images, you can create a neat single image of a person.
The proof of this can be found in this gblorb. So I could potentially use this technique with either lineart or real images. I could probably end up doing both, but I'm trying to work out which to prioritize (assuming people think it's worth my time).
Please vote above if you haven't done so already!
Aika
Hey Aika, I am wondering what the chances are, in the future, of having images to represent hips/butt as well? Keep up the awesome work!
ReplyDeleteBe sure to take a look at the layered image option before voting! I think it looks pretty convincing and better than I would have expected based simply on reading "Real images that connect up okayish, with slight clashes on skin shade, etc.".
ReplyDeleteI did vote for "Real images that connect up okayish".
ReplyDeleteI'm not entirely convinced with it yet, but I'm willing to give it a chance.
There is still a "weird" look to it, because the images don't connect that well yet.
The other problem I have is that the layers most likely can't interact with each other.
Absurdly sized breasts would hide your pregnant belly for example.
Maybe stuff like this could be fixed with a "side-view", but that would double the images needed.
(And breasts would look like torpedoes in order to stay above the belly)
So my final "vote" would be: I'd personally give it a chance, but issues like that need to be dealt with somehow.
In the end I'm pretty satisfied with the current solution as well.
I don't think line-art is going to cut it now, though. We are already used to real-life images and the line-art still has to be produced by someone. And it basically has the same problems as "Real images that connect up okayish". Minus the "trying to find pics that fit" and plus the "producing pics that fit".
Yorekani, your concerns about interactions are valid, and something I worried about when putting together the demo! For what it's worth, I was very careful to only use source images for which a profile (side) view of the same woman was also available. I didn't include these in the demo (as it would have meant exactly double the number of images required) but if it becomes a problem, that option is definitely available.
ReplyDeleteHere's an example: http://i.imgur.com/zEEL9LM.png
On the left is the image included in the demo, on the right is the same image in profile. So the potential is there; while it would require double the images, I think we can get pretty good consistency at least.
Well... as I see it there are currently 150 people who'd like your idea to be implemented. That is quite impressive. (or generally the amount of people voting)
DeleteIn my books it would be best to have it as a toggle when starting the game.
But I'm not actively developing the game so I'll let aika decide if it's worth the trouble and how to implement it.
In case it does get implemented I hope you'll assist aika with implementing and "upgrading" the images further down the line.
Of course, I'd be happy to provide any support and assistance required. While I can't do everything imaginable (the variety of clothes in trap quest, for example, are too daunting a task for me to actually add onto all the variations in the figures) I'll definitely try to assist in any illustrations or other systems Aika requests. I'm a pretty fair hand with Inform 7, or at least can seem that way with enough stumbling around. Either way, it's something I enjoy doing and I'd be happy to work on any "upgrades" requested.
DeleteHas anyones else had a problem with the image not staying up? Mine very quickly displays the image then goes black.
ReplyDeleteWhat interpreter are you using?
ReplyDeleteA black window is just the background of the image window--it appears only when the graphics-window has been cleared but nothing new has been drawn to the graphics-window. This happens, for example, when you resize the interpreter. The demo is simplistic, but it should work in gluxe and Gargoyle at the least. Do images display properly for you in the actual Trap Quest game?
I have the same issue. I'm using Zoom on a Mac - images in the actual game work perfectly.
DeleteIn this, the image pops up for like 1-2 frames worth and then just black, without me touching anything.
Interesting; I don't have a mac, so that's not something I can test on my end. The thing is, the game shouldn't *have* frames--things drawn to the graphics window should remain there until something else replaces them, or the window is altered. That's how the layering works; I have the game draw each layer to the graphics window in turn, without clearing it in between. It sounds like zoom is doing something else that is causing the window to refresh, though I couldn't say why.
DeleteHave you tried another interpreter? The demo *should* work in gargoyle, which also has a mac version. Alternatively, perhaps try expanding to full-screen? I don't know if that will help, but it's worth investigating if the option exists.
Sorry about that; ordinarily I'd see about reproducing the error myself, but without access to the interpreter in question that's tricky.
Ah yeah, Gargoyle seems to do fine. Seems to have been an issue with Zoom. Thanks!
DeleteIt's an awesome concept. Wouldn't worry about clothes since they don't show up on the current system anyway, but the different faces are something to think about. Also the male/female PC options might make it difficult too - how do you slowly show a male body changing to a female one. Easier when you're just looking at parts (I assume).
ReplyDeleteIn either case, as cool as it is, hopefully it doesn't delay release 6 too much ;)
aika already stated that if this gets implemented it won't be implemented for the initial release 6. So work on this (if it will be implemented) most likely won't begin after release 6 is out.
DeleteVery simplistic lineart might not be very good but something highly detailed from a good artist (know any? :D) could come out quite well, even with all the complicated options involved.
ReplyDeleteI fear the real photos would fall apart once you add more options to them.
Jacques00 in particular has loads of experience in expansion and could easily do all the variants for sizes. But still, all artists require being paid...
DeleteI suspect it would not be cheap to get a good artist to produce a ton of highly detailed line art.
ReplyDeletewhat would be a example of the line art?
ReplyDeleteWould it be possible to have more than one option in game? I don't think it's fair to alienate 50% of the playerbase.
ReplyDeleteI know this is unrelated, but is there possibility of releasing this game on z-machine?
ReplyDeletez-machine can't handle quite a few things unfortunately :(
DeleteI'd rather have pregnancy create more creature spawns when you give birth at "beyond octuplets"(you give birth to a fully-equipped royal guard, who later rapes you, for example) as a birthing mechanic.
ReplyDeleteGetting lineart done will require the same artists to upgrade their art every-time the game gets upgraded, which is such a hassle that it can be described as futile and impossible, because I haven't seen ANY mass-production game successfully do this, never mind indie games. Besides, how many people REALLY try to win this game, anyway? I -always- turn my character into an extremely pregnant bimboslut and consider that a win.
In regards to wanting to beat the game, I do.
DeleteThe art would only be used for player models to "accurately" represent body sizes, something that the current system fails to do, it doesn't have to be upgraded.
Kill the minotaur. BAM, you win.
DeleteIt's primarily a text adventure. Your imagination should be all the lineart you need. Now, all you need to do is name ONE game on the entire planet that uses lineart from the same artist, an artist who continues to upgrade their art as a game gets upgraded.
Just -one-.
It doesn't happen. Ever.
I meant actually ascending levels and escaping the game when the game is complete, I never even bother with the minotaur.
DeleteI'm fully aware lineart isn't an option but that's due to the reason no one can afford the costs, it's strange you're so anally against it and assuming the art has to be "upgraded" in some way.
A lot of indie games I wanted to play ceased development because the artist who swore up and down that they would continue to work on the project flaked-out and quit, or just disappeared.
DeleteTQ is a game that will continue to get developed by Aika. As more details get added, the lineart will just get more and more out of sync with the game.
It's strange you're so ignorant of how this stuff works.
The current system itself is just random pictures of bodyparts... Not really any "sync" at all.
DeleteQuick question, was that artist being paid? Cause I'm sure as hell not suggesting that someone attempt to get by with some shady "promise"
Keep in mind the only way this could happen is by actually hiring a professional, there is no lesser alternative I'm implying.
tl;dr my entire argument is basically just lineart is impossible to implement for economic reasons, not gameplay nor artists "quitting" Aika himself can quit any time he wishes as well.
On another note, I wonder when it'll get to the point TQ has it's own mini-community and people argue in these comments everyday.
My prediction is that it will be constantly be about whether hardcore dungeon-crawling/bimbofication is the intended purpose of the game. ...Kind of already happening though, not much of a prediction. Also, hardcore all the way.
Take it easy, guys. It's just a game.
DeleteI like the way it is now. When you transform there is an appropriate picture to show it.
ReplyDeleteDrawn art won't really be a same thing, I really like how it now uses real pictures. I won't be as immersed.
The problem with the full body pictures is that it will be harder to show all the diversity, and I don't think merging the images will look as good and take away from immersion.
I'd actually miss having a variety of faces (and other things) to look at. Personally my brain does a fine job of aligning and layering the (well chosen) current pictures. Line art would make me sad. Thought the breasts looked odd in the real layered example; technically impressive though.
ReplyDeleteI like the way how the image is displayed in the demo (to the side, press space to dismiss) much more than the way it is displayed now. I think the whole system would work better with the "paperdoll" displayed on the right, description of the player to the top left and maybe even pictures of worn clothing below it.
ReplyDeleteAs long as the body pictures are somewhat consistant, I'm not that worried about the style they're displayed. Simple lineart would be cool, but would probably get boring after a while. The complete body image as used in the example seems like it would take even more work than lineart, but what do I know. Honestly the current system of using separate pics is totally fine, and totally overhauling it would just not be worth the effort - just tune the images to be of the same size, have the same background color etc.
for self published games like this the best answer is whatever is easiest for you to do.
ReplyDeleteif you have someone helping you make the images or layouts a paper doll style character with a simple equipment menu like this would be a fun addition. the pictures that are already in the game would just be laid into the corresponding boxes over the body.
this is an example of what i am talking about.
http://f5games.com/wp-content/uploads/2011/08/EquipmentScreen.png
i could see you doing the same thing as a way of handling appearances. just use the photos you have now and place them like the equipment over the corresponding boxes over your simple paper doll.
i would not recommend you try making linework for all the body variants. fenofox attempted to create linework for all the variations of character mutations for corruption of champions but even with artists helping him he was never really able to keep up with it.
Kind of prefer the current style, looked weird when I looked at the link.
ReplyDeleteSorry about that! Anything in particular I can work on improving?
DeleteNot the OP, but I'd guess it's the Uncanny Valley problem - it's not 100 % perfect, so it seems off, and that just weirds some people out. If you're not familiar with it, Google it.
Delete960 votes so far? Wow. Way to go Aika!
ReplyDeleteI really like the style in the link - but I almost think the pics should be an afterthought or handled on a different timeline - the writing and gameplay are my favorite parts
ReplyDeleteWhat if you use the lineart style when you use the "x me" command but when you type something specific like "x breasts" or "x hair" it comes up with a real life picture like we have now. I think that would be the best method but I'm not sure how easy that would be for you to implement.
ReplyDeleteBasically doubling the workload, since aika still has to get both images. It doesn't really matter how often they are displayed. (Other than coding them in so that they are shown)
DeleteWould it require significantly more work than if he just did the lineart style? The lineart will need some work but, I wouldn't have thought that the real life pictures would add that much work since he could just use what he has now. He'd need to add new commands for them though.
DeleteNot immediatly, but the point to "changing" artstyles was mainly to either improve "quality" or reduce workload from aika (iirc).
DeleteSince aika tries to select the "best" pics it actually takes some time.
Probably not that much for a single picture, but finding 10 hip pictures with each one bigger than the last while all need to look "sexy" is quite hard. Oh and it also has some standards like "no clothes".
The new artstyle should be implemented with "expansions" in mind.
aika would probably be less open to change the artstyle if everything was already done for example.
Combining 2 artstyles is a neat idea but I'm not sure if its worth the effort in the long run, since you'll have to update pictures for both.
Hey there, I have tried gluxe and gargoyle and neither will show any pics for the gblorb. Here is an example http://i.imgur.com/XRKvNoc.png Any thoughts? I'm running Windows 7.
ReplyDeleteSorry, this may seem silly, but... did you examine yourself? You have to examine yourself with a command like "X self" same as you do in Trap Quest. Just using the body modification commands alone won't bring up the image, the way it is currently set up.
DeleteYou are 100% correct. I am a derp. Thanks!
DeleteAnd BTW...how does this dark magic work? The image-changing is AMAZING!
Serious uncanny valley on the real images being fused together. Glad that option's losing.
ReplyDeleteLineart would be best, but it may be a pipe dream. Would be nice to be able to show more of what's happening to your character, like what you're wearing.
Sorry about that. Anything in particular I can improve in the future?
DeleteGetting over the uncanny valley with anything synthetic is really, really hard - so not really your fault. The only feedback would be "make it look more natural", as if that helps. Or, much easier solution: make it look much less natural - that helps as well!
ReplyDeleteunnatural beauty would be more thematic in my opinion...
DeleteHey Aika, it's looking like the porn images are losing the poll, so if they get replaced, would you mind releasing them as a pack or something? I really liked trying to get all the combinations of hair color, cum-on-face, and the makeup in particular. Thanks for the great game!
ReplyDelete^ Agreed. It adds a lot of character to our character.
DeleteIt's just a poll to get a general opinion. Even if lineart wins, it might not be possible to actually implement it for several more releases. Or people might have to start paying for it to cover the costs of an artist.
ReplyDeleteI'm worried that the two real-image based options are acting as spoilers for each other in the poll. Lineart is winning, but looking at it another way, the results say that 58% prefer real images. Although, maybe it is the other way, and real images that connect up okayish is actually taking votes away from lineart.
ReplyDeleteNot really sure if I have a solution or if this is actually a problem, but I thought it was just something to think about as you are weighing your future image decisions.
Either way, no one option has a clear majority. Probably won't change anything soon. I'm more interested in the fact that the poll's closed, we haven't heard from aika in a week and we're a week into Feb. Could release 6 be coming soon?
DeleteIndeed. 65% want images that line up better than the current setup. 58% prefer real images. And line art came out with the majority, of the three options offered.
DeleteI guess a more fair way would have been to separate the two issues. That is, something like:
Poll1: Do you favor porn, consistency, or somewhere in between?
Poll 2: Do you want real images, drawn art, or 3D?
Something along those lines.