Game Art Remastered: From Crash, to Kerrigan, to Corley Motors

[Music] hello and welcome to the game art remastered panel featuring brian sousa from blizzard entertainment talking about starcraft remastered cory from vicarious visions we'll be talking about crash bandicoot and eugene will be talking oh from shiny shoe we'll be talking about full throttle i'm going to start this one off with starcraft remastered um there's a lot of questions people have about why we would uh remaster a game for instance why would you remaster a game um i'll talk a little bit on that and mostly about how we did it and there we go this is not working there we go so most of you probably don't remember the google days of the 1990s blizzard had just hit two pretty successful games with warcraft 1 and warcraft 2 and while we did think about other games

that we could develop another rts was inevitable what we did though is we switched genres all together from fantasies to science fiction and we added a third race so we have the stoic terrans the enigmatic protoss and the swarms of zerg a little bit about starcraft is about a three year development cycle we took a little time off for diablo if you remember that and uh it was released in the spring of 1998. broodwar was released later that year and uh and this was actually the formation of the dedicated cinematic team i should know i was there i'm on the top row still bald still angry just kidding 20 years later 20 pounds later and that's me now blizzard has released many new games since starcraft had come out um but the fact is that millions of people

are still playing starcraft 1. hundreds of tournaments professional tournaments have been played the players have recreated and re-imagined our editors to create new maps and they've even set up private servers so they could play games with matchmaking and ladders this was a picture from busan when we released the remaster and around thirty thousand people showed up for our um announced that's not the announcement the uh the release of of starcraft remastered however we felt that it is actually our duty um to keep these games alive keep them up to date and accessible to all of our players and so the blizzard classic games team was formed i came on the team about two years ago and i was focusing mainly on starcraft remastered i had worked on the original so they felt that my expertise would be invaluable starcraft

was released at 640 by 480 resolution it had 256 color palette and an amazing 15 frames per second but it did have eight player multiplayer and that is what helped starcraft become what it is today so um when we when we look back at what we were going to do with bringing starcraft into our you know current battle plan uh battle platform i should say of uh battlenet we talk to the community and the professional players we said hey you know what would you like to see oops what could we fix there's bugs in the game still and what changes could we make well a lot of it we suspected and a lot of it was surprising for instance we knew we shouldn't touch the gameplay the game has been played for 20 years everybody loves the game as it is it's like chess you can change the board you can change the pieces but don't change the rules

or how it plays don't fix the bugs the bugs are incorporated into the gameplay now and these players require these bugs they like these bugs they know and they work around these bugs and that's what makes a master player there was one bug we fixed however uh early in the development mike morheim came to us and he he mentioned that there was a bug in that he introduced back in 1997 and uh it was just an animation that played one frame out of order and it really bugged him and so we actually said okay we'll fix that bug for you it doesn't have anything to do with gameplay we're good and don't change the look this was kind of surprising to us but the fact that starcraft looked a certain way was really key to the players so this led us to say well what can we do with remaster and uh

my boss came to us and he basically told us don't screw it up because this game has been played for so long so many people love this game still playing this game that anything we do we can't screw it up right which also gives us a lot to work with so we started thinking about what our options could be what if we did this well we want to bring it up to 4k resolution future proofing it so in 20 years we're not revisiting all of these assets again redoing them we put in 16x9 widescreen support integration into battlenet was key that brings you up into today's you know player base seamlessly playable with the standard version which was a huge undertaking to make that possible we wanted to put a new lighting additional wrap directions additional animation frames and my personal because i worked on the cinemax i really wanted to update them for today

however not all of these were possible but we were able to upgrade the graphics problems were we had no original assets to work from we found a cd that one of the artists had burned that had assets from early games diablo starcraft warcraft 2 and i was able to find a few models they didn't have any textures on them and they were very rough to say the least we didn't have any art tools all of the tools were gone so we had to rewrite the art tools and basically what we did was we extracted all of the original assets and then scaled them up 400 percent to see like what kind of resolution we're working with and when we did this we found out there's no consistent cameras or lighting i don't know what those artists were thinking but we had to eyeball everything to make sure that the perspective was right for the game and not right for what you would think so what did we do we went first we said

let's let's get our style down and we worked with the battle cruiser battle cruiser is big it's badass and it's iconic everybody knows the battle cruiser so this is our first test we set up a huge style guide about how we did it and what we were going for and we showed these to our our uh our group our focus group they hated it it wasn't what they wanted they already thought we were screwing up the game and so like we had to rethink and so we came up with second battle cruiser which as you can see is more in line with the original we went back then we looked at all of the original concepts sam didio was awesome he kept everything he's ever drawn ever in these folders and so i got a lot of the stuff from him you can see the zergling had little arms on his back the ghost actually has six fingers which i really forced into the game somehow but you'll never see it

so this is what we did we took the original asset and then we did a concept of what kind of details we would add to this concept where could we go with this just kind of enhancing what was already there just filling in the gaps and then our final was rendered and we we built off of that final concept so that when you look at the game you instantly recognize everything in it animations this was one that we really wanted to do if you look at starcraft the animations are very stiff there's not very many frames for any of the things some of them have four or five frames for a complete cycle and they're playing this at 15 frames a second so they're just repeating frames over and over we did a test with the uh the zealot of how it could be we had we added a new idle frame we added a new run it turns out that this was a little bit too much and it changed the game the perception of the game of how these units acted some of the professional players actually would recognize frame three of the zerglings attack is when it does the damage and then they

can move it away so we basically had to scrap our ideas of adding animation and speaking of the zergling uh originally we had him in and we had a play test and uh the professionals felt he felt too small he was exactly the same size and is all perception because we didn't change the hit bot hit test box at all it's the same one from standard but because they felt it was smaller they perceived it that way so we scaled it up a little bit just so they would have a easier time of it we were also able to go through the angles originally the game was supposed to have 36 32 angles 32 angles yeah but because of memory constraints at the time it was cut down to only 16 which is only nine flipped so frames one and two exactly the same three and four exactly the same and we

went in and we were able to put in more angles of rotation one of the biggest changes we did was for the lighting we have the standard in-game sprite which matches almost one-to-one with the originals however we added a new lighting system which adds ambient occlusion mapping depth mapping normal mapping specular map and this gives the actual we have in-game lighting that can affect these to make it a little more depth like a two and a half d representation we were actually able to put lights on to units and so they kind of glowed and they would affect the buildings and of course we had to add widescreen support now you would probably think this is not a big deal change the camera you're good the biggest problem was the consoles and the consoles were a pain in the butt we had to make them be expandable from 4x3 to 16x9 and uh

we worked on these consoles up until ship until after ship with feedback from the community because it just they just weren't feeling right so finally we got something in we have these expandable consoles and i'm happy we never have to do it again the environments as many of the units and the buildings as there are there was eight environments that we had to do each with thousands of individual cells and each of these cells had to be put together like a puzzle piece and we had no way of knowing how these things worked we actually had to create documents and we had the the engineers extract all of the numbers so you can see them on screen like which tile goes to which tile and there's like just in the tile sets of those there's 33 000 of these that we had to make sure that all fit together and we did add some new water the original water was color cycled um usually with five colors and it just

kind of rotated through them and then we added a new shader to the water and i think this was the biggest revamp for the the actual graphics the ramps there was only two ramps in starcraft uh southwest and southeast which at the time seemed like it was okay right we didn't expect this to be a huge um tournament based game that's you know been played for 20 years the community stepped up again and they actually created in the world editor their own ramps using pieces that were already existing while they were hard to make and they didn't actually fit in that well so we took our time and we put in all new directions so we added six directions and all of these are expandable so you can make whatever size ramp you want and we also put these into the standard definition so that you can play still with the with your friends that didn't get the hd version cinematics this is what made me sad

because i think if anything makes it still feel dated are the cinematics and of course there's the people out there that think nostalgia these cinematics are what make starcraft starcraft i wanted to do you know some updated versions even cutting them down because you know we had 30 minutes plus in starcraft and brood or pre-rendered cinematics and it just wasn't feasible today's age to recreate these the only thing we were able to do is the story mode we had uh basically text screens that drove the story and we went through and we added interstitials that added characters that you know from starcraft 2 mengsk and kerrigan and here's the fall of phoenix and we kind of illustrated these to be a little more interesting for the story so that's pretty much the gist of what we did for starcraft um i know that don't screw it up isn't much of a mantra to go off of for for

remastering a game but people play this game for a reason and we didn't want to screw it up everything we changed or added is an option if you don't want the rotation you can turn that off in the in the uh the menu if you don't want the lighting you can turn that off if you want to play it in standard definition you can still do that you can play it in four by three everything is optional so no matter if you like it or not it's it's up to you like how much you want to get into it um that's it um we had no original assets or tools we had self-imposed guidelines of what we could do but we still successfully updated this game without actually breaking it starcraft remastered was released with much positive feedback and we enjoyed a resurgence of uh interest in a game older than many of its new players uh thanks so much next is corey and he's going to be speaking about crash bandicoot [Applause]

did i hit something i sent it to sleep what the heck okay don't look at my password [Music] all right hey good afternoon everybody um thank you for coming so i'm corey from vicarious visions as brian said um i was one of the character artists on the crash medic insane trilogy um and i'll be talking about our approach to remastering the crash's characters um and specifically um the way we thought about our work in the context of the state of the franchise and um fans expectations um for those not familiar with these games

um the crash trilogy was created by naughty dog for the playstation one um and here's a quick set of clips of our own and the original from our reveal trailer so this is from our initial reveal things like some color tuning we're not fully dialed in yet but what's key to know is that aside from g raw geometry reference for level layouts our art was being created from scratch and by eye we didn't have a traditional concept pass for our project so we leaned on the seniority of our production artists who would need to come together and form consensus on the best approach for building our kind of very large amount of content uh i'm a character artist by discipline and some of my previous work at um vv was to help integrate bowser and crash into the skylanders games uh doing this gave me some practical experience trying to meet the goals of a new and different product uh while being respectful of a brand's heritage i personally think it's wise to recognize the meaning iconic franchise characters have to

their audience and to their fans my deck will focus on characters uh but i hope to reflect the feelings and approach of our remaster project in general and i feel this kind of work kind of begs you to yourself become a fan which is a point that i'll be returning to so what were our broad goals our game is a from scratch rebuild of the original three crash games um a high level overview of our approaches we wanted to represent the original games faithfully while building everything from scratch from the get-go both out of necessity we didn't have access to everything and a desire to achieve a modern presentation in the end we'd hope to kind of freshly reignite the love fans felt for the original by giving them something consistent with their memories but also that felt visually modern for characters we wanted to bridge the gap between the classic 32-bit uh look of the original game and somewhat of a feature animation type style blend the low poly punch of like naughty

dog's original art uh with modern expectations uh to find a kind of sweet spot of maximum nostalgia both of these things was to hopefully not disappoint anyone playing it totally fresh um or for the first time in two decades and this could mean the usual things like more rounded out volumes uh smoother deformations more detailed texture presentation but importantly we couldn't sacrifice the essential design language that would be key to triggering those kind of fond feelings we wanted our audience to experience there's two particular challenges that we dealt with that i'd like to kind of focus on uh finding faithful designs for the rebuilt characters and supporting faithfully dynamic performances for them because we were rebuilding these games from scratch part of the question becomes where do we pull our style and inspiration from this kind of can become a dilemma it could be easy to go down the path into a full remake or reimagining territory uh where changes that are not carefully

executed could lead fans to question the intent of a product a game like ours now as a personal example i you know remember as a kid watching cartoons um i would often wonder why characters would look off model from like their 16-bit kind of representations if i was watching a video game cartoon you know and you know that's a harmless effect of graphical limitations at the time but the point is like that we would want to try to avoid evoking that kind of disconnect with what we were making that we wanted people to feel comfortable that what they were looking at is what they remembered and what they cared about uh because we were um wanted to flesh out the low poly source art we needed to consider um kind of where we should pull our character design influence from at first we kind of flirted with a kind of simple idea of taking inspiration from original concept art from toys costumes later era models things that were already more kind of fleshed out and detailed in 32-bit games

while this would make sense uh seem to make sense while targeting a higher fidelity look they really kind of diverge to a degree while the various looks of crash of commonalities orange for blue shorts spiky hair people have particular opinions about which one they might personally want to see uh the crash band had been relatively dormant and the art style had changed over the years and this seemed to kind of cause a sense of confusion to a degree along the fan base and we had to be mindful of what source we were pulling from this is a discussion we had to have in regards to every character and all the assets in the game crash is just a kind of a clear good example of the variation that the brand experienced over the years the conclusion is that we eventually came to is that these higher fidelity assets marketing materials and stuff is they lack the simplicity and direct lines of the playstation 1 art

and most people were likely to be subconsciously responsive to we tried to keep our ear to the ground and observe how people responded to different different game remakes um or how they had discussed our previous crash efforts our solution was to kind of preserve the essential angularity of the 32-bit models of the vintage look get the character designs right by being faithful to the classic looks the majority of audiences would have nostalgia before for without actually creating low-poly art ourselves when i say essential angularity uh i tend to mean the major movements of the shape and silhouette so you know these are we feel people respond to on a kind of core level so artists needed to not be selfish had to let go of our previous styles and any desires to really heavily reinvent these characters we kind of had to reverse engineer them back to their core not over design or fix them these characters have a lot of appeal in their own way and we

should polish and not try to lose that kind of core essence uh for crash himself this meant keeping that basic triangular shape language that was essential um you know preserve it because it's so easy to lose those kind of strong core primitive shapes when you want to be making art that feels modern and impressive this is something we strove to keep in mind which isn't to say we were completely literal all the time you know calls had to be made like fitting crash's eyes a little bit better into his cranium kind of giving him more structural integrity kind of more flexibility for animation and consistency you can see this approach here applied on entropy as well the old model and the new model just kind of being mindful of certain shape regions and the relative relationship they have even dur during pretty noticeable squash and stretch so kind of moving on from what worked to when there were complications for us uh we found eventually that individual

perception of the retrographics could cause disagreement definitely among the fan base and even among artists working on a given task there's only so much fidelity that was visible to most people playing these games in the first place over composite cables on their crt tv videos and artists might be referencing might have been captured at a low resolution it can lead to a lack of clarity over what some characters prop pieces were um as a kid i remember trying to look through my tv screen trying to get past the color bleed of the crt you know trying to see the pixels of the characters more clearly an early unfinished version of this green character right here was shown in one of our trailers early on we were called out in some comments for not getting the colors right or for having a slightly odd look the truth was it was an unfinished asset that had slipped out into the wild like sometimes happens unfortunately um but we were also called out for not realizing what the character was some people saw a lizard character that was the actual intent

um others saw some kind of moving plant they were discussing what they remember seeing in the original game the real truth was more complicated it was some kind of cyborg lizard with a laser eye with a molten frill with bionic parts you know a pretty complex set of characteristics to illustrate in such low fidelity something that's actually a little hard to deconstruct for even a current artist you know it was it was funny it was a little silly but it was an illuminating moment for us we established that personal interpretation is a potential challenge for anyone remastering games like crash of that vintage you know luckily there was you know great rejoicing when this one asset seemed to be adequate at release um you know i think posts like this are are more kind of ingest about the amount of scrutiny that almost every aspect of a game being remastered goes through but still the point being if people care about what you're doing and if you're remastering something

probably people care about it you've got to understand that you're going to be looked at and scrutinized and it's an interesting dialogue back and forth so the other big piece for us on the character side was hitting performance consistency with that kind of zany dynamism that naughty dog's original vertex animations were known for the level of stretchy smeary deformation that they had was not something that we had been focused on doing a lot of in our skylanders work that we were doing before crash bandicoot a few examples of crash's expressiveness right here it was important to keep in mind things like his large asymmetrical eyes his kind of like independently scaling pupils and squared off jaw that can go from extreme close to ridiculous open these were all things that we would have to hit and unlike in sky we aimed for less detail on our scopes to get clean major forms only um to support this very stretchy deformable topology uh we used the same character models for

our cinematics and for our in-game characters and we didn't build any mascot like expressions in we aiming for like a pretty true neutral to support more dynamic performance and a wider range of appealing expressions it's the kind of thing that feels risky when you're doing it when you're used to making a stylized game to not build in any character but it is a more sensible cinematic workflow and for a greater range of animation um here's crash's flat shaded model with some early expression tests by our rigor jim donnelly because crash was known for such expressiveness we were anxious to see if everything would come together in a flat shaded view on our target character this was probably day one like the first past topology uh with the first rig and it was kind of a really nice relief to see that our kind of assumptions and and plans were hopefully going to work we were going to find the expressiveness we were looking for and we still had a lot of time to polish at this point

so to show you a little bit of how we vetted some of our work here's a gag test by cinematic animator eric johnson uh seeing if we could hit this kind of classic zeni animated feel as well i'd love to play it for you again but i don't think it's gonna roll the just to oh let me try oh okay move on the thing being that what we needed was a lot of flexibility a lot of bendiness we had to kind of like internalize kind of the classic like kind of sense like that topology should be even and nicely distributed we had to like really keep this in mind with the characters we were using because of the amount of stretchiness we wanted um and the amount of influences pervert uh also it played really well with the first shader we were using having this nice kind of consistent ring mesh type topology allowed the fur shells to kind of stretch uniformly and not look too

distorted to get crash's eyes to work his eyeball topology needed to be stitched directly into his eyelids for one contiguous mesh he still had a bit of a sphere under there but the lids could open beyond their widest extents and begin to stretch which is kind of like a mix of an old school kind of painted on eye and a normal eyeball um the pupils then would slide around on the surface in a very cartoony manner and could be scaled independently since we weren't completely liberal to the original geometry ideally we'd have an even greater expression expressive range getting full wacky and and hopefully with the ability to control for more nuance and subtlety on the occasions it was called for um in the end the player was meant to go this looks good that's the reaction we wanted um our arch should not been overly distracting or to the average player for its stylistic choices in comparison to the

original uh we didn't want to be overly burdened by hardcore fan reaction however it could really be an interesting gauge for us to see if our choices blended in or if we were distracting and breaking our this is how i remember it goal we wanted to give a little bit more than a straight remaster report but we didn't want to go too far and uh kind of break everyone's immersion we learned some good lessons from the path we chose um we're very happy with the fan reaction we hopefully people enjoyed it we think it was very wise to be aware of how the crash bandicoot brand was being perceived at the time we were working on our project and how it's being scrutinized and especially important to not be above our source material to look to embellish and not replace what was there but to keep clear and to keep clear targets on how to achieve that i think people love these franchises enough so that they're out there producing their own art and keeping the torch burning uh even when the series is dormant for long stretches of time i believe it pays to be respectful and to

recognize like that love if you're doing this type of work and that's it for me if you have any questions please feel free to contact me if you'd like to talk about anything covered today or the merits of playing retro games on crts versus upscalers i'd love to talk about that up next is uh eugene from chinese shoe the other words awesome hi everyone i'm eugene kim co-founder and art director of shiny shoe an indie studio located here in san francisco shanichi makes original titles and also partners with other studios on their titles we teamed up with our friends at double fine to remaster the classic lucasarts point-and-click adventure games grim fandango day of the tentacle and full throttle today i'm going to focus on our actual production experience on full throttle remastered in relation to some assumptions about remastering visuals of a game assumptions i myself have made in the past or heard during planning or budgeting

meetings or heard from consumers specifically one remastering is just a reskin it just needs a fresh coat of paint two we have better tools now it should be easier to build build assets it will take less effort and things will go more quickly three we have assets from the original so we can reuse stuff or leverage existing assets it's been my experience that these are false assumptions that can put you and the production team in a tight spot fortunately the full throttle team was an experienced group and while we did not make such assumptions outright we still faced a lot of production challenges that would have proved self-destructive had we made such assumptions many resources were invested into reviving full throttle which i hope gives you food for thought for your projects be it a remaster or even original title when shiny shoe first partnered with double fine to remaster grim fandango day of the tentacle and full throttle a core philosophy that

tim and double fine consistently emphasized was to remain true to the original experience we want to keep true to the original work of art it's a collaboration of a bunch of artists coming together the acting the writing the sound design the music and animation all these things all these people work together to make this thing and we don't want to mess with it we just want to present it in the best possible way and make it more true to the original intentions we're getting rid of artifacts compression and aging technology to make it look like it looks in our minds a key decision in adhering to this core philosophy was to port the original game engine to current platforms or to paraphrase our programmer trevor dean while the game is running for the user query the game state each frame and retrieve enough data about the game state so we can then provide the user with a more modern presentation so as far as the game is concerned is still ticking along like it had back in 1995. this meant that we had to author the

remastered assets in a similar way that the original artist did back in 1995 but of course with modern tools no problem that's just a reskin a fresh coat of paint like painting over 20 year old paint on a house sure it's easy plus we have modern spray guns and stuff right excuse me i'm confident that any professional house painter will tell you it's not that simple preparation is required before a single can of paint is open there's tons of scraping spackling sanding masking there's a process and a pipeline required to do the job properly so that it holds up over time so that your customer is satisfied the original full throttle paint was beautiful a blend of 2d and 3d assets and animations it was unique for its time but it was of its time and now is 22 years old the game's original resolution was 320 by 200 pixels the remaster had to support 4k resolution well over 10 times the original resolution in other words that 10 by 10 foot wall

that needed a fresh coat of paint is suddenly 100 by 100 feet 100 times the scraping spackling and sanding but we have better tools now it should be easier to build assets things will go much quicker than it did for the original it's my understanding that the original full throttle development team use d-paint proprietary animation and editing tools and 3d studio release two or three for production i'm not sure how many here are old enough to remember d-paint or 3d studio before it was maximized but just for kicks they look like this limited colors in d paint and no character rigging in 3d studio and if i recall correctly uv mapping meant rotating the camera so it faced the polygon and taking a screenshot or using graph paper clearly they were limited tools by today's standard but powerful at the time and the best available a nice big paint bucket and durable brush to make to paint that 10 by 10 foot wall and paint that wall they did creating full throttle's unique visual style and

universe for the remaster photoshop flash premier and maya were our modern spray guns but remember that 100 square foot wall is now 10 000 square feet aside from the pure physical scale that needed to be reworked the upper has created content issues things that could be alluded to or suggested in the original were obviously incomplete or awkwardly amorphous at 4k furthermore the now standard 16 by 9 aspect ratio created dead space around the original 4 3 aspect ratio rather than pillar or letterbox the image we extended the scenes with new content which in turn created new issues for visual composition animations that cross the screen or gameplay that relied on the original boundaries of the visual plane this created new design discussions and feedback loops to devise new guidelines to preserve the original intent of the visuals while adding new content

we're all aware that modern spray guns have shortcomings of their own custom pipelines often push tools to perform in ways that they were not designed for for example full throttle's animation engine briefly described here by oliver franskay put flash in a bad way we stuck with flash because it suited the character art style animators are familiar with it and we had existing tools and pipelines from the remaster of day of the tentacle but further into production the difference in the animation engine from day of the tentacle amplified minor pre-existing problems in authoring alignment between animation components and animation slots became problematic and flash really pooped its pants when high-res pre-rendered 3d frames needed to be incorporated granted by introducing high-resolution raster images we pushed flash in the way that it wasn't meant to be pushed but we were too far into production to make changes and authoring tools instead more resources had to be put into existing animation pipelines to shore it up

for cut scenes we went with premiere primarily because we're familiar with it but automating premiere for cut scenes rent for custom rendering proved to be challenging in the time frame that we had thus since every tweak bugger polish pass required a cut scene to be reprocessed through a fairly complex export cycle we ultimately poured a lot of resources into managing and maintaining that pipeline so while new tools are more powerful you're not using them in the same context of the original which seems to be the root of the misconception people assume we're making similar assets to the original but with new more powerful tools but new pipelines carry as much overhead as they ever did and new specifications and expectations keep pace or outpace the tools so in many ways you are in the same boat as the original team we have advantages over a full development cycle because game mechanics concepts and designs may remain the same but a remaster involves a brand new production cycle that is rife with production challenges of its own

but we have assets from the original so maybe we can reuse stuff or leverage existing assets there might be an argument for this if the original game is within a generation or of current or so of current technology for full throttle however the 22 year old games archives with a cardboard file box of cd roms and binders the cd-roms of the source art contained scans of peter chan's original concepts and incremental and redundant copies of individual artists hard drives from 1995. within those hard drive backups were file folders of incremental source and intermediate 3d and 2d files basically there was no master source control project that we could use and there was no tracking documentation that we could find so we had to sift through all the data and create tracking spreadsheets just to figure out what and where everything was in terms of 2d assets aside from the scans of peter's original concepts which were really cool we found no higher res authoring source files instead

oliver extracted frames from the shift shift source code to create authoring files in photoshop and flash and artists and animators work from low res reference images scaled up to 4k resolution there's a lot of room for misinterpretation with that much resolution difference on a 3d front we had source files but since there was no source control we could not determine which files were actually used to render the game's final assets going by latest date was pointless because there was no guarantee that the file was actually used in the final product trevor ended up creating a special tool just to isolate the best or closest 3d animation a 3d animation file used in the game it was a viewing and sorting tool that played thumbnail previews rendered out from each source 3d file as well as the original intermediate output frames from 2d files alongside the final footage of the original game we filtered by file format and frame count ranges and cherry pick the files that looked correct it was a fantastic tool but we still spent weeks

distilling the archives down to the files needed for remastering check out trevor's blog on yamasutra detailing the process and tools involved after we isolated full throttle source 3d files custom scripts and tools were created to import the relevant original data mostly cameras and keyframes into maya set up the render globals to remaster specs and allowed artists to quickly reference in remastered models over the original and set them up for the render farm the full remastering and re-rendering of 3d files were only part of the process we were unable to find about one-third of the 3d source asset so we resorted to rotoscoping in maya using the original cut scene frames as reference after that we still have to composite the 2d assets with the rendered 3d assets into cinematics and in-game cutscenes masking and match moving as well as the logistics of tracking and timing file dependencies was extremely time consuming once again a hope or an assumption of production gains is negated by new

production overhead thus in estimating our production for remaster consider that even if the core game mechanics and design remain the same the production aspects are always a new undertaking and will pose as many challenges as the original title the original core r production team for on full throttle apparently comprised of two painters five animators three 3d modelers and an additional four animators and five artists in support working over a 20 to 24 month period by comparison full throttle remastered involved approximately 14 painters about a dozen 2d animators three 3d animators and three modelers ramping in and out over roughly 12 months there were over 15 000 unique hand animated frames nearly 200 remastered backgrounds and 445 individual fmv cuts that made up 134 fmvs my special thanks to double fine the double fine artists and our production partners at yatta super genius and laksha i'll close with a few general thoughts

i'm of the opinion that game development like nature abhors a vacuum production overhead and expectations always rush in to fill that vacuum or another way to look at it is that the game is never done it's just taken away unless it's a mobile game then it's just never done thank you uh so that's it for us but if you guys have any questions we'd love to answer them so um we'll open it up to you guys do these mics work oh yeah they do everybody's muscles your mics work i think there's one here there's one there any questions yeah i'm jeff from clay um so for the starcraft not redoing the cinematics was that just a production issue or was that because they wanted the core cinematics to stay the same um it was a little bit of both uh

initially when we reached out i think because of the the people that we talked to were mostly professionals um the people that are continuously playing they're not playing the single player campaigns and so it wasn't important to them and so we kind of like that was the feedback we got we're like well they're not really important nobody watches them anyway um so initially that was kind of our thing but then once we started getting into it and we started showing what we've done with the multiplayer they're like are you guys redoing the cinematics and i was like yes let's do that and but by that time we were kind of six months in and um it was just we would have to if somebody come in and actually rewrite everything we'd have to go through lore there's a lot of roadblocks on it and eventually we just said it's it's just too much for for our time that we wanted to do yeah thank you hi okay i guess this kind of applies to all three but um that's pretty loud um what was qa like did you guys have like aesthetic nostalgia qa for like checking

in whether or not you're hitting home and the original character is feels or just the environments feels everything like that we uh speaking at least to our project uh qa was absolutely headed up by an enormous fan um so key qa people on the project were people who were very big fans of the franchise and the expectation was to be checking it against the original game and being mindful of things like that absolutely so for us we actually had an embedded qa team um nobody from back when but one of the things we had incorporated was an f5 key for testing so that we could switch back and forth because to us it was really important that nothing really jumps out as different and so like one of the main things we tested was like you know what while you're playing push f5 play it in standard definition go back to hd if something pops out then it's wrong and we had to redo it and uh like one of this this isn't even with gameplay but one of the things we had is uh we had a protos portrait

and his eyes were supposed to be blue and i think we changed him to red because in starcraft 2 his eyes were red and immediately like in a focus group i was like how do you even know what the protoss eyes colors is but they're like hey it's supposed to be blue and so we actually re-rendered it all and and put that in there because it's one of those things if somebody notices it then we did it wrong that was the main focus was not to do anything that people noticed [Laughter] could you make the most subtle art in the world but also make it impactful thank you welcome more of a question for crash there was a lot of speed runs on the original game yeah and that means the speed runners were using a lot of glitches or exploit has there been analysis to reproduce them in the remaster or you're asking was their intention to reproduce them or were they reproduced yeah exactly i mean if a speed runner

get on the remaster and try yeah the exploit to get the same speed runs uh has there been an analysis on yeah yeah pre-production so i can't speak too much to stuff like that because it hews more on the design end of the spectrum but i do know that we had players who were doing like internal speed runs and time trials and testing things like that i can't say whether we kept bugs in or intentionally removed them we kind of did a mixture of polishing features and keeping features so sorry i can't give you a really clean answer but hopefully that's a little bit awesome thank you yep you're welcome hey guys um this is for corey um first of all when i played it i felt like i was eight years old again so you definitely all right um i was actually just curious uh did you guys how did you guys render the fur on on all the characters sure yeah so it was a kind of a

procedural shell solution that was made in-house originally it was part of my deck but for time we can't cover like everything so the fur used kind of uh kind of detail maps to kind of push out shells like almost like traditionally like from the from the uh mesh and we'd use varying height maps to kind of get different amounts of kind of like variation in the height along the fur we could scale the grain we could push at the height basically we had a suite of sliders we also had a normal brushing solution for doing the directionality of the fur so we would create additional more vertex normals and and brush those and that would give us our direction on top of the shells that were just being pushed out from the fur we had an anisotropic feature on it as well and the direction of the an iso would go with the brushed direction of the normals too yeah it was maybe at some point somebody will show that too awesome thanks yeah you're

welcome hi i'm lizzy from uh pixel x and gamesmart and uh i was just wondering if you guys anticipate remastering any of your games in the future again is that for everybody um so so as the classic games team we're basically taking each game as a case-by-case basis um obviously starcraft is one because it's such an old game and it had never been incorporated into battlenet and um we saw it as an opportunity to do that um but in general the classic game team is just supporting all of our older games if we're going to remaster any that's kind of like up in the future like does it require it you know do people still play it um in general we just want to make sure that all of our games are playable and you know the best experience that people can get but not every game necessarily needs a remaster right right right yeah i think um

from like activision's perspective just seeing the reception that crash has received you know opens opens doors to considering more projects like that um you know i can't say specifically whether like any are happening or what the future might be but definitely when these projects like go over well it makes people pay more attention sure and a lot of you had mentioned that you lost assets over time is it are you guys taking more precaution into saving them for yeah whatever so that's another thing that we're kind of tasked with especially across blizzard because now we have so many projects is getting that resources we actually set up a vault that will incorporate all of the work that we've done into like this one central location so that it's not lost again starcraft i think is only is one of the games that we've lost almost everything on but you know we're still working through

all the old assets yeah activision the same thing these days we archive everything at the end of every project but i think all three of us basically encountered the issue that like nobody had the original stuff we didn't have access to at least most of it for the most part anything that we looked at was extracted from the release builds of the game but now we're much more diligent about source control okay cool thank you yeah thank you any other questions i'd love to chat all right all right if you guys want to do any like you know personal questions you don't want to talk and pro we can uh there's a room across the way for uh over there outside of here um so if you want to come up and ask some questions we'll just head over there and you can talk to us thank you guys thanks very much