Friday, August 22, 2014

Elysian Shadows: Interview With The Artists Part 2 (Leandro Tokarevski)

This song has nothing to do with game, but as I already shared Connor Linning's work on the game I figured I'll share music I'm enjoying a lot from a different indie game, Electronic Super Joy: Groove City! I love this song, and the whole soundtrack. EnV did a great job.

This is a continuation of my previous interview with Patryk Kowalik, this time featuring the game's other artist Leandro Tokarevski! Read the previous interview, and get details on the Kickstarter itself, here.

The Logbook Project (tLBP): So let's start simple: who are you and what is your role in the Elysian Shadows team? I'm interested in general information as well as your history with regard to art, and specifically pixel art.
What sort of programs do you use for making the assets for the game?
Leandro Tokarevski (LT): My name is Leandro Tokarevski, I'm a pixel and concept artist for Elysian shadows, as well as a level designer. I'm half Italian and Half Russian, I was born in Rome, and now live mostly in St Petersburg, Russia. I've always loved drawing and painting, when I was 11 I entered an evening art school and then went on to entering the academy of arts of St. Petersburg, I'm about to start my fifth year studying architecture. Despite drawing for so many years, I've only relatively recently delved into pixel art, Elysian Shadows is my first real pixel art project. It's been an interesting experience, learning the specifics, but I found that most of the techniques used in traditional art were just as valid in pixel art, so it wasn't hard to adapt. I use GraphicsGale, Adobe Photoshop, and more recently Spritelamp. I find GraphicsGale is the best program for more old-school pixel art and animations, but ES has ventured so far from the classical 16-bit RPGs it was inspired by in terms of visuals, and gameplay for that matter, that its features are simply insufficient for our game. The addition of Spritelamp really helped develop ES's visual style, I'm still figuring out some of the details, but I'm pretty sure that with the help of Photoshop and Spritelamp We'll develop an atmospheric and unique art style. We're really doing things nobody really did on such a massive scale with pixel art.
tLBP: I absolutely agree that traditional art techniques are often just as useful to pixel art. When I'm working with pixel artists almost half of what I talk about is more fundamental general art concepts, with the other half being specific pixel-art techniques. I also agree about GraphicsGale's excellence as a program for standard pixel art - I don't use it personally, but along with Pro Motion it's the most robust pixelling program I know of. It's really interesting that ES is unique enough that GG wasn't enough. Having watched the video about SpriteLamp Patryk put out a few days ago though I can totally see how unlike anything else SpriteLamp is. It's a great piece of unexplored technology that I underestimated originally.

tLBP: Can you tell me something about how the process works between you and with the rest of the members of the team?
LT: I usually am given to design and draw an entire area of the game at the same time, along with all of its enemies, the architecture of the towns and dungeons, so everything in it is consistent and the designs are similar and work with each other. The team has a huge (and secret) design document in which all of the lore, characters, and environments are described, so before creating any assets I go there and read what exactly the area of the game is about. After that as long as I stay true to what's in the design doc, I have pretty much carte blanche, and can invent pretty much anything. We have a team chat on Skype, in which we discuss all of our ideas and suggestions. Once an idea is approved, I can go ahead and draw it out in pixel art. Sometimes, when the team chat is not enough, we have a team hangout where we brainstorm ideas and figure stuff out. It needs to be planned in advance, because we're all in different time zones. It isn't uncommon for us to have a hangout when for someone it's 4am.

tLBP: I don't have experience with pixelling for game development, but I always hear about how restrictive the time crunch can be in creating the art. How much time do you reasonably take on a single asset (a background, a character and their animations, whatever)? Are you often forced to work with the first draft of a piece or do you get much chance to go and revise before moving on?
LT: It's only me and Patryk creating assets for ES, and the amount of stuff that needs to be drawn is sometimes overwhelming. Most times when creating a tilesheet or a sprite we skip the concept art stage entirely and start drawing pixel art right away, speeding up the process. This sometimes backfires, however, if there's a miscommunication between team members during the discussion phase, and sometimes we end up repixelling something we spent a lot of time into. After the kickstarter we'll finally have enough time to draw concept art first, and then go into the pixel art. I'd say on average a tilesheet for an environment of the game will take about 15-20 hours to complete, it varies from area to area, though. Same with sprites, it depends on the kind of creature. A stone golem will take less time to animate then a living breathing creature. The time crunch was particularly restrictive on the last weeks before the kickstarter campaign launched. So much work needed to be done, that there was barely any time for discussion. Now the pace of work has slowed down a bit, thankfully.
tLBP: I don't have a metric for if 15-20 hours really is a huge amount in game develoment, but it certainly sounds like a lot! Glad that the pace is starting to slow a bit though, seeing more refinement will be really cool.

tLBP: How indicative of the game's look are the Kickstarter photos? I gather that the asset quality will continue to be upgraded, but is there anything style-wise thatyou think will change from the current look?
LT: I'd say the kickstarter screenshots are barely representative at all at this point of how the game is going to look like when finished. Since we decided to use Spritelamp and to draw normal maps by hand, our process for creating pixel art has changed completely. We no longer have to put shadows on to sprites and tiles themselves, we draw a normal map that when implemented in the game will create lights and shadows automatically. This is such a fundamentally different and new approach of creating pixel art, that we'll have to redraw and rework every single asset we currently have, so I'll say that 99% of what you see on the kickstarter page will look different in the actual game. Actually no, make it 100%.
tLBP: Wow! So the general aesthetic (perspective, etc) will be pretty similar but everything will be redrawn using Sprite Lamp's normal mapping? I'm really excited to see hwo the areas start looking when you get whole regions fully built with the normal maps, I bet that'll be really impressive.

tLBP: A big change to the way pixel art works in games comes from the lighting engine the game will be using, SpriteLamp. Can you tell me something about how that technology works and how it changes the creation of art assets?
In my own experience, a consistent lightsource is an important part of making a pixel piece seem cohesive. How do you make the base art that SpriteLamp will then work on while still trying to produce good assets on their own?
LT: When we first experimented with Spritelamp, creating up to 5 lighting profiles for every single asset, first of all it took way to much time, it would become impossible for us to ever finish all of the artwork for ES, and secondly the normal maps that it created weren't even that accurate and looked off most of the time. Patryk then had the idea of drawing all of the normal maps directly by hand. He explained to me how it would work, and it just made sense. We would start by drawing a diffuse map, basically a drawing without any shadows, where the light comes from every angle. Then, we would draw a normal map by hand in Photoshop, so that it would look exactly the way we want it. Then if needed we would draw depth maps and ambient occlusion maps, and feed it all into Spritelamp. The results are fantastic, and it doesn't take that much more time than drawing traditional pixel art. It's sometimes even quicker. When a dynamic light hits an object it comes to life, it looks three-dimensional, and exactly the way we want it to look, while still looking like pixel art, since we can regulate how many shades the dynamic light will be divided into when it hits the object. 5 seems to work best. As far as I know nobody ever used such a method of drawing pixel art, I'm sure we'll learn more along the way on how to perfect it, but even now we already fixed basically all of the issues traditional pixel art had with dynamic lighting.
tLBP: One of my initial concerns with using SpriteLamp was that the outputted normal maps on Sprite Lamp's own Kickstarter page looked pretty unimpressive to me, but manually creating the normal maps has seemed to really show the true power of the program, as shown in some of your recent content. Being able to regulate the shades is interesting too, I didn't know you had that control. Did having more shades just not hold enough of that "pixel art" look or something? It's also really interesting that it's not any slower to do it this way. I'll have to try it sometime!

tLBP: The game is going to be run on a variety of systems. SpriteLamp is native to Windows, and has plans to work with Mac and Linux, but are there special considerations you need to accommodate for it to function on Dreamcast/Android/iOS? Should you reach the Xbox One/PS4/Wii U stretch goals I'd imagine those consoles are strong enough to handle it, but the relative power of the mobile OSs seems like it'd be an obstacle.
LT: That's not really my area of expertise, but from what I heard from Falco, the Dreamcast is just about capable of handling what we create with Spritelamp. I really hope the programmers will find a way to implement the light engine on all of our platforms, I really don't know if that's possible and what it would entail.
tLBP: Patryk was confident that the systems were more than capable of handling it, which is very exciting.

tLBP: Still on the topic of the other operating systems, are you doing anything special to make it look great both at the small resolutions of mobile units as well as on large monitors or televisions with the PC/Mac/Linux/Dreamcast? Will it be a matter of scaling the assets, or some other solution?
LT: The resolution of the game right now is 720p. I would personally prefer it to be a little lower resolution, because right now the character sprites are really small on screen, and sometimes we show a little too much of the map at the same time. The pixel art suffers a little too, making the grid-like patterns of the tiles more apparent, but I digress. We'll scale all of the assets to match 720p, or whatever resolution we decide on. It just doesn't make sense to make a full HD pixel art game. It's better to just zoom in. Pixel art looks better zoomed in anyway. :)
tLBP: I agree in a lot of cases about zoomed-in pixel art. And full 720p doesn't actually add that much precision to pixel art. Especially if emphasizing the pixel-art look is the goal, a bit lower might be preferable. As you've said though, ES's pixel art isn't quite like any other pixel art, so it'll be interesting to see how it behaves.

tLBP: This is a long ways out, but about the Next-Gen+ stretch goal: will that high-end graphics setting be an entirely remade set of assets or just upgraded capabilities (physics etc)?
LT: It will have the same set of assets as the regular game, just upgraded features. Not sure exactly what though, it's still a little early to decide, I'd rather wait for us actually reaching that stretch goal!
tLBP: No problem, that covered what I was interested in! Thanks for all your time - these answers have been really in depth, so I can't actually think of any additional questions! I'm really excited to see how the game develops. Good luck on the final leg of the Kickstarter, and I hope that even should you miss the goal that we haven't heard the last from you yet! I'm rootin' for ya. :)

The game is only $4000 away from funding! A great burst in these last ten days has taken the game out of the scary uncertainties of the middle point of the campaign.

End Recording,

No comments :

Post a Comment