Session: Navigating 50 Years of Evolution in Interactive Audio - Simon Ashby, Audiokinetic.

[Music] our first Speaker for the keynote speech is Simon ashy from audio [Applause] kinetic hello hello hello welcome and I don't see you that's a pity a lot of uh new faces some familiar faces and I it's really great to to be here thanks thanks it's full crowd um yeah Kino session it's a bit weird to realize that so and the reason for this is I've been invited uh earlier this year to do the keynote

session at AES Tokyo uh last May and a keynote is something I've never done before and um so the the stress level was kind of important to me because I started researching about that and it's much more about like something inspirational and talking about yourself which I never do because I'm used to do many talks but it's always Technical and it's always you know to the point so anyway that I succeed eded I ended up you know giving that talk and it was okay but it was not recorded or filmed or anything so we said okay why not doing it again so that there is something for the posterity and uh and we decided to do that for for the Y store here in elsi uh my very first time in Finland it's it's been great so far everybody is super super welcoming and smart and wel anyway it's so far a lot

of fun so um all right are you are you ready yes all right there's energy in the room okay so um let's go into this uh keynote mode now and the topic was um and is today it's about harmonizing art and technology and looking into that over the the last 50 years basically since the birth of video game and all that research started with questions where um I was yes it does work okay so I was wondering is it possible to harmonize art and technology or is it something that you know will always remain unbalanced and I was also asking myself is it even desirable you know to reach this this harmonization between the two or is it better to have them challenge each other all the time and as artists are we victims of the technology or should we

just Embrace those you know those technological limits to in order to be creative and so on so of course in the course of history um there's been moments where the technology was really heavily impacting the artistic output uh that we could that we could do um but in more recent times it's kind of the O the opposite where maybe the technology is not providing enough for us to really be able to convey and provide like our vision and and what we want to do so so these are all interesting questions I would be interested one day to maybe have a panel on this and invite like engineers and and artists and so on um but anyway I think what is really important to me is you know keeping the individual in mind keeping you know our point of view and our our goal as artists and Engineers uh in the end

because we are the driving force making those games and it's it's because of us if that thing is happening uh to our audience so um I'd like to tell you a fable I told you Keynotes are different so i' like to tell you a fable that is attributed to sha piggy a French poet um from from the previous century and the Fable stand place at a construction site where three stone cutters are you know banging on stones with their with their tools and passing by the the very first stone cutter uh he ask what are you doing and the stone cutter kind of look at this person with a strange look like dude maybe not a century ago Dude was not really in the vocabulary right but dude we're cutting Stone don't you see and passing to the second stone cutter and that person answered

we raising a wall and finally to the third stone cutter that person with a smile answered we're building a cathedral and to me this Fable is something that translates very well to creating video games right because you need a lot of people that have that are attributed to different task and together they work on something much bigger than their personal contribution in order to come up with something as you know fantastic as a cathedral and and to a certain extent I believe building games is probably even more difficult than building cedral because have you finded the blueprint for making every game fun all the time from the get-go and plans to do that um so I think it's it's very hard and and it's really the quality of the interaction we have as teammates that makes the difference my name is Simon Ashby I'm one of the founders of AOC kinetic I'm head of product and I'm really proud to

be with you today and my goal to today is really to you know take a look at the past see what was there and go with that and look at where we're at today and during that time I'll take uh some time to talk about our own Journey with wise and with audiio Kinetic and how we we came about uh with all that and to conclude with a a look at the future and try to extrapolate what it will be to make games in the future and in what kind of environment and and so on and throughout all that time um I want to take as much Focus as us as individual us as creators as Engineers as artists because we Remain the driving force behind the creation of those games and and crazy projects that we do all right is that good with you is that sounds like a plan still energy in there yeah

okay you're still smiling all right I really don't see you at all great okay so about 50 years ago actually almost precisely 52 years ago in September of 1972 uh arrived the Magnavox Odyssey game and two months later in November we had the sorry we had the Atari Pong there and both games were Virtual Tennis you know with paddles and the ball crossing the street and one was one has been total commercial failure and the other one has been the success we all know and all remember uh today so can any one of you guess what was the difference between these two games and why one was super successful and the other not sound audio Yeah smart people all right exactly exactly pong was making sound

and Magnavox I say wasn't and of course here I'm pretty sure I'm preaching to the choir here but uh to a certain point we can say that without sound there's no video game industry say that to your producer next time you negotiate your budget right it's thanks to us if you have a job so so and the funny thing about pong the arcade version the very first one that came out is that they were not thinking about out sound at all and it's apparently the the story says or the legend at that point uh that Mr Bushnell at AR president asked his electrical engineer to come up with a solution and that apparently was like like the day before going into production and Mr alorn the lead engineer look at the components that he had were there and focus on the SN generator and the SN generator is the piece of technology that that displays the picture on screen

right and if you remember the good old CRT TVs where you had the odd and the even lines displayed one after the other and the reason why we're 25 frames in per second in Europe is because the electric current is 50 HZ North America 60 HZ we're at 30 frames per second so we're cading that picture to be displayed so to a certain point this is your oscillator for a sound synthesizer isn't it you've got clock and you've got a square wave with you so that's where he connected he tapped into that and that's how we ended up with this fantastic piece of art [Music] that's so I think we can agree that the technology was quite limiting um to a

certain point and like the audio was created almost by chance by an electrical engineer and even in in this context we can ask ourselves is this art I don't know way to think about that but I think some people are smiling so so anyway so that this is the birth of video game and we have sound and pictures and something to play with so comes the second generation of platforms and we've got uh many of them and the an television console one this is this is the one that my father purchased so the very first one I own best day of my life best year of my life probably anybody played an television here okay one two okay yes we've got a few all right so that console was interesting because with this now we had three channels that this platform could play but the limit ation was that they were

all sharing the same attack and release and the same amount of noise you could add noise to to the the tones uh but they were shared and they were the same amount on the three voices so with such a limitation the composers or the engineers or who whomever was responsible to make the sound for those games had to rely on technical voice stealing where essentially you've got a priority system where the most important sound will steal voices from the least important sound so typically you're going to have your gameplay feedback sound effects that would be more important than the melody of the music in background and here's a pretty good example of that [Music] sounds hard for me or I don't know in

your ears so second generation now arrived the third generation of platform this is the Heyday of the Nintendo so they released the Nintendo famicom in Japan two years later the NES in the rest of the world and now we've got not only we have five channels but we've got different waveforms so we our sound pallet is much more diverse now with square waves and triangle and noise and even sample could be played there but you're still stuck with not a lot of memory on the cartridge for the entire game let alone you know the the audio portion of it so you could write you know one or two themes and then you had to rely on your smart programmers to use techniques like um bite swapping and bite shifting and basically moving bites around so that you can have modulations and permutation and parallel Melodies going on so that was a way to take a little bit of music and extend the

experience over time um with that and over the years we got better at it and some people got really really good at it and Kodak kasan is a good example with journey to silus and listen to that score this is this is great [Music] I've listened many times to that song Anybody played that game by any chance no oh yeah okay great okay so third generation comes the fourth generation uh Sega arrived there and we've got a Super Nintendo and now things are starting to

get serious where we've got eight channels of 16bit audio 38 32 khz per second uh or Kilz sorry well that's that's accurate stereo panning TV were stereo now so the console were leveraging that thing adsr independent per channels adpcm compression DSP effects like name it we have it all and to a certain point we can now emulate like a full Orchestra with that and this this is what we got on some of the games like um like on the Legend of Zelda linked to the past this is one of the game I've played the most like in hundreds of hours probably other person here that done the same you haven't see you seeing you smiling so uh so here's the introduction theme you'll you'll certainly remember that

great stuff um so looking into the hardware and what was part of the the Super Nintendo um you look at the the board and you see you know all the chips and the CPU in the memory and when you focus on the audio DSP you realize that Sony was the maker of the audio chip for Nintendo so that really got me curious so I started investigating more and apparently they worked together Sony and Nintendo to produce the Super Nintendo CD uh back in those days and the deal was that the console will have the cartridge and that would be you know Royal given to

Nintendo but CD that was one of the Sony that was one of the inventors of this CD at a time will get the credits for the CD that you could put on the console so but if but that didn't work Nintendo decided to go with Phillips to do the Nintendo 64 and of course guess what Sony arrived with the PlayStation the original PlayStation um and now now we're serious so not only they were providing the same DSP speed chips uh that we had in the previous generation but there was also another piece of technology that allowed for Red Book audio like CD quality stuff 24 channels 80 PCM half a Meg of memory for sound this is this this was like almost a cartridge back in the day so this is crazy this is really cool and Reverb and so on so now are we at parity with film right we have the quality and we've got

the we've got the gear for that so yes but at the same time uh there's more and more content to cover because there's more capabilities for the graphics as well and we're stuck in the good old thing where we have too many sounds and we need to compress them and lower the sample rate and you know we end up with some something that is much low five or lower F than than we first expected and there's more repetition and so on but yet some people were really you know targeting that Hollywood like type of experience and midal Gear Solid any fans of middle Gear Solid the original one okay many more people as we're getting closer to present time so uh so that was the introduction thing and this Hollywood feel [Music]

[Music] the nuclear weapons disposal facility on Shadow Moses Island in Alaska's Fox archipelago was attacked and cap Ed by Next Generation special forces being led by members of foxhound starts feeling like current games right dialogue orchestral music sound effects and all that so but I think it's fair to say and and again I kind of picked the best examples from each of those generation like the average game were sounding much more average in general um but I think it's fair to say that during the F the first

five generation of of console where the the technology was really heavily impacting the artistic output like the quality of the audio in the end and so from there from the sixth generation in 98 roughly to today uh to the ninth generation like we've seen new console like we're seeing like like our phones are playing games now AR and VR devices and so on but essentially the last 25 years we've been you know there's been no real Revolution outside of being sample based audio and so what we've seen is you know more storage from the DVD to from the CD ROM to the DVD to the Blu-ray disc we've seen hard drive added uh Hardware accelerated codex that kind of thing but it's been so it's been 25 years of the first 25 years of synthesis and where each generation was really impacting the artistic the output or the flavor to a certain extent of

those console but the last five or four generation the last 25 years it's mostly simple based uh in there so but in any case so 1997 this is when I started in that in that market and I've been hired by Ubisoft so Ubisoft is a French company and eventually they created a studio in Montreal in Canada and I've been lucky to be the person like the sound guy after after six month they realized oh we're going to need the sound guy and I've been lucky I've been that sound guy that they hired and I started at Ubisoft and Ubisoft had a contract with P mobile the German company doing the PLO thing and we had three games to make and somebody in France was thinking that oh a single sound designer on 3le mobile game that's okay because we will reuse the same sound across the three games but just look at the artwork they were really three different

games I took the challenge of trying to have one sound used in the three games and I failed we just needed something else so quite rapidly we realized that it was not a one Army type of thing and we needed help so basically I was responsible for you know the technology the integration keeping up with the sound design and you know all the changes that we see over the the three games and making sure that we knew exactly what kind of assets to produce the music the format the duration all that but we needed somebody to take care the the production the audio and hiring content providers and composers and so on and that person was Martin Klein that we hired and Martin later has been the founder of audio kinetic and still today our CEO so it's during those days where during the day we were doing our stuff and during the evening we were regrouping together and look at the integration and start mixing the game

and you know make sure that we were delivering an experience that was making sense and just to give you an idea for the mixing flow back in the days uh researching for this keynote thing I stumbl on on this on this Excel sheet so I had this printed out on the side of my monitor and it was a conversion between decibel to the linear volume that the sound editor that we had at Ubisoft the little piece of technology or or middleware was had this volume in in linear value so for example mixing back in the days was I remember in one of the game there was a water fountain at the center of the village so you get there you listen to the sound you say hey this is too quiet it needs a good 3db plus so you took your pen your sheet of paper water fountain plus three soon enough the game will crash this is your opportunity you go back to the sound editor and now you say Okay so the

volume of that sound is 34 which is - 11.5 DB I want a plus three this is min - 8.5 so the final volume will be 48 and you enter 48 you save your project didn't crash you're good you do your Banks it didn't crash you're really good you recompile the game you go to the restroom because that's 10 minutes you go back finally the game opens you run to the water fountain only to discover that it needs another two or 3 DB and you put that so that was mixing back in the day so imagine doing that for a couple of years five days a week up to 10 11 midnights many Saturdays and there were moments where Martin and I were looking at each other saying if only there was a pro tools for interactive audio if only doing a Crossfade was not having to debate with

a programmer for 15 minutes telling him that it sounds much better when you Crossfade than when you kill the voice and play the next one that kind of arguments so that's how the the idea started and so keep working at Ubisoft work on the missed franchise work on Tarzan on Jungle Book reding group that's been a really interesting experience to me so that game was already running for a while and they decided to reboot the game and we were nine months from the release date and they changed the game play completely and they decide to make that a rhythm game you remember the game Dance Dance Revolution where you had arrows flying the screen and you press those arrows on time so this that was really popular at that time so they said we want to do a dance dance like type of game but with the Jungle Book franchise you have N9 months and they took me from the project I was on and they say Simon help

us and we said okay so at the time we had a game Engine That Could render 3D graphics and that's it so how do you synchronize an entire game on the clock and have something compelling and so on so to me you know as a sound person I all immediately thought about you know the music sequencers that we had like the cuas and those type of things where you you have your Tempo and your bar and bead going on and you've got tracks but this time instead of having music tracks only it was synchronizing the camera movement it was a track for the arrows in Easy Mode in medium hard expert having the the animations and everything basically was sequenced in our Evan synchronizer that was the we so we had our Arrow editor we even had the karoke editor so nine months that was kind of cool so and the programmers told us okay can build that thing in time we had nine

songs to do the nine main characters in in Jungle Book um but one thing they didn't know they could do was to interpolate the arrows in between frames so that meant that so the the engine was running at 15 frames per second so that means that if you do your song at 120 BPM this is an even number of 15 frames you're good but then if you run at 14 or 16 frames you have to set the tempo of your song to 112 or 128 and so on so to the composer it was like man pick whatever Tempo you want as long as it's one of the tempo you see here and so that was that was fine what was much more difficult was to convince the animator that they had to animate the dancing character and they had only one bar to do all the moves and because at at the end of every measure the game will determine if you were

performing like super well to have cool move or no if you're stumbling and nothing is working and they were forced to you know do that full bar thing and they had to had a key frame in the middle so we could transition to any animation and they didn't like that you're killing my artistic expression and that kind of thing it was really a struggle to get them to realize that's the only way and we only have now 6 months 5 months and so that was that was interesting anyway so here's a here's an idea for the [Music] game of the swingers the the I've reached the top and head to stop and that's what's bothering me I want to be a man man go right into town and be just like the other man I'm tired of walking [Music] around I want to be like you I want to walk like

[Music] cool stuff so in the early 2000 um even even when you build tools custom for a type of game um what I realized was that it's not only about the the tools and it's about us again as human being being able to say Okay I accept that I've got I'm limited in what I can do and so on and my my you know the technology blocks us in a certain path and I I need to be expressive in that thing and that was the very difficult thing to do and and it basically it was our capacity you know to understand that yeah we're all Stone Cutters in our own fashion because we have a greater goal of building that

Cathedral and Jungle Book was not a cathedral a small Church a chapel at best but yet yet nine months a small team we accomplished that but it also you know we we came to the realization that if cutting Stone by hand is too difficult too long to create those more ambitious game maybe you need to come up with better tools and maybe sometime you need something you know industrial strength and that's right so and that's how that's how audio kinetic and wise emerg basically because we realize going to GDC was an eye opener on that well it was not a problem at Ubisoft actually Ubisoft had the sound editor and some tools of itself most of the industry people were hard hardcoding everything in C++ or the best editor ever Excel you've got your sounds one line per sound and you've got your columns for volume and pitch and is it looping yes or no and

that kind of thing so we said okay it's it became really clear for Martin and I that the whole industry was in Need for tools that will really help us audio artists at the at least to come up with soundscapes that would be engaging and you know and and really good and and the benefits of that is maybe we will be able to come back home for dinner and not late at night when everybody is already in bed in the thing so that was the the side benefit that we were hoping for that so at that point we said okay what are we lacking we were asking those questions and for us it was clear like we were aing aing we needed to hack in order but we were lacking creative control technical control efficiency that kind of thing and to the question what do we want to achieve we wanted to achieve to offer s artists A workflow and workflow is an important word in

that sentence a workflow in tools that will help them create compelling interactive experience so when we say workflow in 2024 it's pretty clear now we've got tools and so on when we're used to production flows and and workflow but back at the era where you're heart coding everything what is a workflow that really help artist speeding up their thing so we settle on five main steps if I could say for that workflow and that slide by the way was used even before we released the first version of wi we're using that slide to explain our concept to uh first people investors and so on and later to game uh developers so the first step is really to author your thing so you you create your you know your dialogue your music outside in a do fine and you import that in Wise and starts the authoring you put the behaviors you put your sound

properties and you know all the indication to have this content playing as you expect and especially back in the days the game engines were much more flaky than today like we have it good today like unity and unreal and godo and even the proprietary game engine are really solid nowadays back in the days at least at Ubisoft my personal experience like all the time like it was working yesterday and for the next three days nothing was working and you cannot test your content in the game until somebody find a problem so for us it was important that we could simulate in Wise using the soundcaster for example and you play your music your dialogue your sound effects and you can start premixing that content you can start validating the Aesthetics and that kind of thing and eventually all right you're good you can start integrating your content into the game and when you arrive at mixing no need for a pen now you're connected live to the game and you're making those decisions connected

and you've got the direct feedback from it while looking at you know your performance how much memory how much CPU are you using how much bandwidth from the DVD are you taking that kind of thing and to a certain extent we can look at this workflow in two different type of timelines so it's it's both doing that 25 day 25 times during the same day where you as as soon as you add new content you can go over all those steps but it's also a metaphor for a two-year production where in the beginning you do a lot of you know authoring and integration and near the end of the project you do a lot of mixing and profiling and troubleshooting and that's when your lead architect reminds you that audio should not take more than 10% of CPU you heard that before yeah yeah exactly so that kind of thing so for us it was and they won the very first version of wise we are we were already had the profiler and all this

data because this is what is important to achieve yes it has to sound good but it has to sound good with you know the budget that we're assigned to so anyway we go like that in 2006 we released the first version we had a studio from Microsoft that started with a beta version and shipped a few months later Shadow run our very first game and and at that point you know you can ask yourself what why is an instant success right for the first time in history all right big word first time in history there is a tool readily available out there on the market if you're interested that you can license and it harmonized art and Technology right so was it an instant success what's your guess no exactly and I think we again underestimated the the human

factor you know and all those things so you arrive with a good piece of technology fine but who are you are you ktic what do you know about game we never heard about you and personally like who's that French speaking barely speaking English with a strong French accent guy that is too excited and you know trying to convince people that this is the thing so yeah it took a lot of time there was a lot of resistance we had some of some early adopters you probably know the crossing crossing the chasm curve you probably heard about that thing in in selling products so there's always people that are willing to jump in but that's a minority of people and it takes a lot of time to convince the the mass and have you know stronger adoption so what do you do well you start visiting them all and you go at all the conferences and you speak with them and eventually they get to know you and eventually they get to realize that okay maybe these guys knows what they're doing and some game starts to release

and you say okay those are great sounding games and we finally started having some adoption but it took a while and yeah much more time than we wanted actually eventually we open a studio in in Japan just for the time difference the language barrier the cultural barrier as well later same thing in Shanghai about 5 years ago last year in the Netherlands in ersu which is nearby Amsterdam and just again to be closer to you to be closer to the people using uh the product and having an idea for what's to come in the future so if we look now where we're at we've got three main segments in which wise is used so of course gaming is the main one like that's where we started that's the bulk of our business uh and the development and so on but there's more and more location based entertainment product s out there and by that what we mean is you know

sophisticated Escape rooms or theme parks where they have rides where there's moment whereas there's interactivity going on you have to shoot spiders or things like that and they tend to use unity and unreal to do that to render the graphics and why is this a really good fit for that kind of environment so that's this is cool this is growing and if you're a sound designer and like you may be working on games or not but there's also this Market that is emerging and needs your your expertise in that type of field and finally the automotive sector and we started in 2014 porting wise to the automotive platforms and reality and the timing is just perfect it's this October that we have the very first vehicle that will hit the road with wise in it so shangan is a small company the size of Ford in that's the thing with CH is so big that we never heard of changan but

yes it's it's big they do electric vehicle and it's not too late you can pre-order if you're interested and they're using wise for the r sonification and for uh 3D alarms for to localize you know 3D where's the thing instead of just just having a beep in the dash of of the car and to for the the sound as well when you're running you know electric vehicle at low speed has have to emit sound uh so so so that that kind of thing is used for that okay so let's look a bit at the the foundations of why so of course there's the runtime that's the sound engine running on all the platforms out there if you can run a game on a platform we have to Port wise there because that's your business you need to get your games on as many platform as possible and of course facilitate the Integrations into the the main commercial game engines like Unity unreal Goo now and to to proprietary game engines the authoring

that's another pillar of course that's where a lot of the action happens and and what is really important to us is the ecosystem around wise so starting with the partners so and the best partners that that we have are the one that helps you know integrating wise with the rest of the work the game development workflow or the ones that are providing tools that you know solve specific problems s those are the the best Partners there because they facilitate your life and they get you closer to to the end goal and even more important than that is you the community all of you using wise uh sometimes sparingly sometime every day uh but it's really thanks to to you that that we exist right and and if what I like to do when I go from City to city is to visit developers and if I try to reach out to you and you're us wise be sure that I will ask you much more question and I

will talk about wise or things like that like I'm really interested in knowing what is working well fine but what is working less well please tell us that and please tell us what is what is your vision for 5 years from now and 10 years from now because we want to make sure that we will have those building blocks in place when we will arrive at that time so don't hesitate to share it's important and finally the knowledge sharing that that's another big point where not only us you know doing material and Sample games and and blog articles and things like that but and more recently having certifications and all that but also having our community to share so if you're subscribed to the blogs like every week There's an article about somebody that is writing about how they overcame you know certain limits or how they achiev certain result uh in their last game and it's great because sometimes it's you know a person in China writing an article that we

translate in Korean and Japanese and English and French so that it may help you you know because you have a similar problem but it's somebody from China that that you will never be never seen in your life probably and never be able to speak directly with but you know via those things it it it helps so that's really really important so if we wrap up our our contribution to this I think we can say that we empowered artists to overcame technological limits by establishing a workflow establishing a vocabulary like it seems like a simple thing but you know 20 years ago there was five different ways to talk about a random container or real time parameter control and so on now we kind of share the same vocabulary so you hire somebody from a different company and they're kind of you know up to speed day one or day one and a half you know it goes fast and finally a standard way to conceive

audio and I've got this feedback so many times like people tell me they they they have a new challenge something new and they automatically know you know how to go about it and and how to arrive with the right result said differently we can say that our contribution is also to allow you to you know come up with any type of Aesthetics any type of sound that you're after with any type of resources is available to you and those two knobs have to be able to move in whatever Direction you know without without dependencies between them because technical constraint is something but it shouldn't prevent you from doing something artistic and something engaging and meaningful in the end so I think we we're playing a definitely role into harmonizing art and Technology um and then I think we can be proud of this accomplishment I think it's a it's a good

step but is it enough right is it just a momentary passage there will it last it's not clear it's not clear to me it's one stage we're there now but what it is for us in the future right so let's try to take a peek at the future and see where we're heading and how are you doing how is it you're still there yeah okay and you're still smiling some of you okay good feedback thanks all right so let's look at the future so we all know that the future is made of many lines and dots right okay no you're not really there yeah okay I'm just rambling okay so if we look at our production today um you know we all produce dialogue and music and sound effects and more and more we use the geometry to do fancy sound propagation systems and dynamic mixing

and all that but if we look into what we're doing we realize that it's all a manual process any of these sounds any of these settings like there's a human being that's you know move the dial and curate those things so all right now way to try to predict the future is to look at the trend from the past and you tell your yourself if we continue in the same direction we can kind of derive what the future may look like so I decided to take this approach looking at how many people were working um on on triaa games like 25 years ago and how many sounds we were producing on average again it's just an average but let's say in the 2000s producing 5,000 sounds with a team of three people that was that was a fair production for a AAA game and you push 10 years later now you

produce 50,000 sound with a team of 15 yeah that's pretty much an average that that was there closer to today producing half a million sound with a team of 75 this is not unheard of the very first game with a million sound in Wise was in 2011 just to tell you so it's been a while that we're producing too many assets to a certain point and when I say 75 people I'm not saying like 75 like people full-time in the studio but when you start looking at your contol your composers the contractors the legal the The Voice directors all the translations and that thing like it sum up rapidly to that kind of number so essentially every decade we kind of multiply the number of assets by 10 and the team grew by a factor of five roughly so let's take that and let's push 10 years in the future producing 5 million

assets on average and 10 years later 50 millions of asset I believe will reach those numbers but I certainly don't believe that we will be teams of 302,000 people that doesn't scale imagine imagine that imagine 2,000 people in the audio team only so how many animators and modelers so this is not imagine the budget so so we will have to produce 50 million sound in a not too distant future but with roughly the same size of team so we desperately need a disruptive technological breakthrough it's unbearable like we cannot do otherwise right we are at a point where art now it's the other way around it's the te technology you know is there's too much art to do and we need help from the

technology to get there so in the future I think there's no other way than to accept that we will have to rely on text to speech right and by that I'm not saying like I'm assuming here that we are compensating the actors appropriately I'm not saying we're stealing their models and not paying them I'm saying we hire those actors they record you know the lines maybe some of the cinematics the trailer things like that the important dialogue but the rest can be text to speech and imagine if you have that number of lines it means like if you have two 200,000 dialogue lines just for you it probably means like 18 months in the studio four hours a day with the same sound recorders and the same voice director who wants to do that right even for the actors they don't want to spend their life in a studio they have better things to do so so we have to rely on those things same thing with music we'll

keep working with composers they write fantastic music themes like we've got cous bumps and all that and that's really good but what is really you know tedious right now is we as those composer okay can you export all the stems possible can you split the music into you know two bars or four bars and so on and they do a lot of fiddling around that has nothing to do with their art which is writing memorable music right so let's use the machine for that there's better ways and I think we can extrapolate that to everything we do like with sound effects with like the sound propagation like understanding the geometry and optimizing it automatically for audio so that we just have sound propagation working 98% of the time and you put your humans to fix those 2% of edge cases that's a much more rational approach to it and same thing for mixing and so on so the production workflow of the future will most certainly look like

what we are seeing with chat GPT and some of the machine learning tools out there where as an artist you set your priorities your your mood and you know your Aesthetics that kind of thing you let the system generate propositions and as an artist you iterate and you go you do that feedback look until it works but it's still with all that we still have humans working we still still have Engineers we still have artists to better inform those things and come up with the the right things so we have to re the human I think this is my take I hope it will be like that but I I really wish that we will keep the human in there and it will be the interaction of all the humans and a team that will make the difference between the failure or the success of our future games so yeah definitely the future of video game rule remains a collective experience so if our future projects

remain collective experience I'd like to invite you when you're back to work tomorrow to ask yourself and ask your teammates are you cutting stones are you raising walls or are we together building Cathedrals thank you very much okay all right so before we move on does anybody have any questions it's kind of weird to have a Q&A after that isn't it oh yeah we've got questions thanks yes uh yeah I of course this is very scary to think about like how how much asset how much uh control we might lose in the future just kind of kind of running uh like running for this kind of

huge bigger games and and all that and today we have uh uh we still have this thing where like it's easy to write for example um to write a tool to kind of get like dynamically the size and materials of a room to do like Reverb but we we still do a lot of work manually because we want that a static to come in it doesn't have to be realistic but we want that to to go through so how do you think we're going to kind of manage that because again there is a lot of manual work because we want to go in still and do room by room uh going through this kind of narrative and controlling the emotion of the of the you know of the player how do you think that how do you think machines or or other tools will come in to help us do that without us like losing control that we so much like need for these things yeah and the the word control in in your question is so important um and and it's

at the the core of all the decisions we we make when we design new features and so on is how much control we can expose to to the artist uh again for the Aesthetics but also control over how much memory and CPU you're willing to invest or you're able to invest in whatever systems that that is in front of you so exactly how I don't know yet and uh right so so so what we're doing right now so for example in our unreal integration there's already a feature that can evaluate the size the volume of the rooms and by volume I mean the the size not the the the game but the the size and from there derive an rt60 which is the duration of the late Reverb and from there select the right share sets um that you already preassigned with different volume so you get a first pass and you get late reverbs of different durations automatically with that feature so at least you have that you've got your ground covered but yes after

you want a human to go there and say okay for this specific room right I need something that fits better that that the mood that I'm expecting for it or it sounds good but gameplay wise there's a lot of dialogue and I want less Reverb to get more space for the dialogue whatever reason right is it the game context and so on so how exactly I don't know we do you know we do our own r D there's tons of R&D outside and that's why I was saying that's why we're asking so many questions because the answer is somewhere collectively you know it's part of our brains and we will you know assemble that and eventually come up with better tools and better tools to to get there yeah no you're right yeah maybe how wasn't the right but it's just kind of how do you see that kind of or like the the the yeah again the control is very hard to give like as a designer yeah to anything else so so so we went from having no control at all to now we

have a lot of control but we have a new problem where there's just too much stuff to control so we have to let go and and that is scary letting go of the control is scary it's like raising your your kids right eventually you have to let them go and uh and that's well as a parent that's your responsibility as an artist in a game I don't know like you don't want them to become things you don't want them to become yeah any other question yes just in front Simon uh pleasure to see you almost like 12 12 years from from the first time but seeing how wise evolve and become friendlier and friendlier and integrated even more so and as also your kind of projection of the sound teams and you and Martin starting being just two guys now an in the game that is here and reaching you know tons of people in in in uh streaming platforms also the

gaming team itself is becoming only one person sometimes plus two so what about why is with all this you know uh development of technology that is more uh automated make it a friendly for the non- audio people so that's another approach because we have always the audio people have to count on the last Source the last hour the the you know the crunching time but all the people that above us are still growing what about if we you know share the payload they can actually write or like can actually integrate things before we get the last part of the artistic value to it than the tech side cannot be just put upon us so more than with you that a lot of things will be automated in the future but what about if we make it even more friendly that already is it's a it's a good question um and again

there's control is part of your question as well and it's how to but where the spin how to leverage the rest of the team to help us there and interestingly so back early 2000 before we start marketing Marketing in a sense putting to the market wise we had to decide on a licensing business model and the dominant model was the per seed like Photoshop like PRS like those things right you cost that amount per seat and and you're good to go and this is not what we wanted to do and that's why we went with the licensing per title and per consult because we wanted anybody in the team to be able to install wise and contribute to a certain extent so it's important because if you don't do that then they don't people will pay one or two license for the sound designer the programmer won't have one your QA team won't have any Li license but it's much more efficient to have your QA team connecting to a profile version of the

game seeing a problem with audio packaging the profiler session put that in jira and then you've got somebody from the audio team like troubleshooting that kind of thing so we wanted the entire team to be able to install wise if they wanted to so it was already in our mindset there now in terms of specific tools and so on well most of our attention has been given to the audio people tus far uh and there's some automation like I was saying in unreal with determining the the late Reverb thing but yes I think as we EV all that thing you know grow and so on I believe like organically we'll get there and it's going to be easier for non AIO Specialists to contribute uh to it as well but again how exactly and when ah TBD and if it becomes more and more important you'll tell us and we will make sure that we developed that thing we're really really like we have a

certain idea for the road map of wise but this is like maybe 50% of it the the rest is driven directly by the feedback we receive from you so please share don't hesitate and when we visit you you know open your heart and tell us where you want to go with that and we take that into account and it it it end up in the product for sure yeah are we on time are we uh uh last question last question cool uh thank you for the inspiring speech thanks uh so you mentioned the history of audio and you know in the first few Generations it was more synthesized uh syn thriven and then later it was Sample based and going forward we will have generative uh sound generation like generative AI um and you know somewhere in between there was there has been also procedural audio and you know how do you think uh procedural sound gener will uh play a role in the future and you know do you

think that will be supercharged by Ai and how do you think wise will be um contributing to this okay so that's an excellent question um so let's start with AI uh and machine learning because there's there's so much money invested into that worldwide and alting and and there's tons of money into audio as well and so so we are doing our own AI thing but the world is vast and so on so we know that companies or people will come up with Fantastic Tools so for us the first or the one of the most important thing to do is to make sure that we adapt we adapt wise so that that kind of Technology can be easily integrated so we have the plugins for example for DSP or source so that's an easy way to integrate there but we need to extend that so that it takes that kind of AI and machine learning technology and it's easily integratable into wi so so this

is this is really important and this is something that it's already started and we'll go over the next few releases of wise make sure that it's easy to connect that kind of tech and maybe some of it will come from us maybe but I'm pretty sure there going to be coming from a lot of Partners out there um in terms of procedural audio per se like if you say procedural a you like Max MSP or pure date for example pure data is a good example where people were asking why aren't you integrating pure data and our answer was always like if it was not sounding that bad we would probably have it in there so one of the criteria when we create things or integrate things we we use the Hollywood ready type of things like does it sound as good as the best production you can do and if so all right it has its place there but historically pure data and so on it sound sounds like okay it does something but is it Hollywood ready no so that's why we haven't

invested necessarily into integrating those Texs or developing your own how will it be in the future when it's informed by Ai and machine learning if it's really lead us to better Direction yeah why not like when we int when we introduced media into wise a decade ago um on as a site project we had the syn one have you used that as a source plugin and why syn one that was just the most basic synthesizer we could do and and my idea with that was let's put it there and people will request more of it and they will want more feature and we'll come up with a real syn later and maybe we'll call it syn two or something else and nobody cared and nobody asked for sin and better thing like that so we said okay it's a good idea we believe it's fun I was strongly right I I was excited by that but nobody cared so we said okay we got s want it never move since then because it's not important to you at this point in time apparently so

um so yeah so we'll see we'll see how things evolve and when it's you know when it's time and when people are saying we need that kind of tools of of course we'll add it to it so uh all right I think we at the end thank you very much again you're really good crowd thank you all righty then thank you very much Simon