Twelve years ago I wrote the first Internet blog recounting the development of a video game in real time. It was a fascinating experience that enabled me to share the ups and downs of developing OMIKRON –THE NOMAD SOUL with its future players (and annoyed the hell out of my then publisher, who had no interest in my development anecdotes).
Twelve years have gone by and now I'm happy to resume my development blog and share the human adventure of developing a video game, in this case HEAVY RAIN. The years have taken their toll and I am no longer a naive and impatient young developer, but my passion for my work is still intact, as is my desire to share it.

HEAVY RAIN is an important project. For QUANTIC DREAM, obviously, for SONY to a certain degree, for me, obviously, keener than ever to transform the experiment that was INDIGO PROPHECY/FAHRENHEIT. Beyond that, I believe it is a project of major importance for the games industry. If HEAVY RAIN fails for whatever reason, it will be years before another publisher takes the risk of creating a project based on emotion and interactive narration for an adult public. If it succeeds, it will prove that it's possible to create different kinds of interactive experiences that have nothing to do with the rules that have been pre-established for the last twenty years.

Whatever happens to HEAVY RAIN, I have a deep-rooted conviction that it's now time for interactivity to emerge from its adolescence and its associated excesses, and join the adult world. I hope that our work on this game will contribute to this end.

That's the setting. So this is my blog. Read it as you would read a ship's log recounting how the captain tries by every means at his disposal to bring his ship and crew to a safe port, through tempests and storms, in spite of illness and disease, pirates, sirens and other monsters of the deep.
When you see the ship heaving into the port, you will know what it's been through in order to reach you;-)

WEEK 1

Einstein demonstrated that time contracts as it approaches the speed of light. He forgot to specify that this is also the case as we approach the master deadline for a video game. It's like when one of a pair of twins boards a spaceship and the other remains on earth. Months will have gone by for you, whereas we will have the impression that it's only been a few days.

For those of you who are not familiar with development, there are four particularly important dates: the date of start of production, the date of the Alpha version (in which all the game data has to be produced), the date of the Beta (in which the game is assembled and playable), and the Master date (usually the one we fail to make). The development of Heavy Rain passed the Alpha date successfully at the end of May, the next deadline is the Beta at end of September.



Quantic dream’s Openspace

HR is accumulating a large number of development difficulties that are specific to this type of game. Firstly, the game uses almost no mechanics (by mechanics I mean recurrent systems). Instead, it mainly uses a contextual gameplay. Each scene is different, depending on the point in the story, the character we control, and what is happening on the screen. Although this principle constitutes an advantage (I hope) for players by offering them great diversity in terms of action and game principles, it's a veritable nightmare in terms of settings because each scene requires specific handling.

The other particularity with HR is the importance of artistic direction. Because the aim of HR is to recount a complex story and generate emotion, each game element must contribute to this end, whether we look at the graphics, the characters, the animation, facials, lighting, sound, music and, of course, the gameplay (which is part of the AD in HR). Nothing can be left to chance because that might break the spell and lose the player's involvement in the game (what the theorists call "suspension of disbelief").

In addition to all that, we chose to set our sights high in terms of the overall quality of the game, and we quickly realized that these high standards always seemed to generate others that were even higher. The slightest illogical detail, the slightest badly lit face, the slightest camera shot out of sync with the scene, and the whole scene collapses. The game has high standards not only in terms of quality but also in terms of absolute homogeneity so that nothing detracts from player immersion.

When we reached the Alpha stage the game was in a pretty weird state: three or four scenes polished pretty much to a Beta level (prepared for the internal presentation with Sony and for fairs). This was the case with MAD JACK and MADISON AT THE BLUE LAGOON (presented at E3), as well as two or three other scenes, including the first scene with Madison and a scene with Ethan, considered too violent to be presented at a fair.
The rest of the game broke down into two categories: scenes that had not been polished but which nonetheless gave a very clear impression that all the elements were present to enable them to work, and scenes that were completely rough and ready and left us wondering how we were going to finish them.


Given that the game consists of nearly 80 scenes, each different, having a different set and gameplay and requiring specific and precise AD, we had more than enough to keep us busy…

In order to reach the Beta stage in a coherent fashion, we chose with Charles (HR's producer) to divide the game into eight cycles, each cycle containing ten scenes. We gave ourselves two or three weeks per cycle to polish them up and bring them to Beta level. It seemed to be more rational to focus on ten scenes at a time rather than trying to spread ourselves out over 80 scenes at the same time.


Charles Coutier- HR’s producer

We'll soon see if we made the right choice…

WEEK 2

I'm emerging from two months of recording voice sessions, two months shut up in a little 10 square meter recording studio with another actor, surrounded by infrared cameras and hundreds of flickering diodes, responding, correcting, encouraging, asking the actor to repeat it faster, not so loud, with more emotion, etc. These sessions were particularly exhausting, first of all for the actors. The four main actors had to learn their lines by heart (if they had read them, we would have seen their eyes following the lines in the animation). With dozens of markers stuck to their faces, they had to imagine the sets, situations, characters, with only the script and me as guides. Each time we had to adapt to the actors' pace and habits, reassuring them, helping them to give their very best in an incongruous situation.


Quantic Dream’s recording studio for facial capture

All the actors gave a lot, particularly the four main actors. We've been working together for more than a year, scanning their faces, having them rehearse scenes, photographing them from every possible angle, recording their movements. We know each other well. They know their characters well and you can feel it.

Sometimes, in the course of a break, I get this strange impression that I'm talking with a character from a video game. I've spent so many hours filming their 3D clones that I've almost forgotten they're real people in the real world. I take that to be a good sign.



Actor equipped for facial capture (left), 3D version of the same face (right)

I'm writing these lines in the middle of cycle 5 (the first 5 cycles comprise the most difficult scenes in the game; cycles 6 and 7 are relatively simple, and cycle 8 consists mainly of epilogues).
My days are a bizarre kaleidoscope: I help Charles (the HR producer in charge of production at the Quantic end) as much as I can to keep track of the team (there are now four people working full-time just keeping track of production, plus me), mainly for problems linked to scenes in the current cycle. I spend lots of time in review, sometimes in the review room, sometimes at home in the evening. Reviewing consists of playing a scene several times and noting everything that doesn't work or needs to be improved. These notes are then entered in the bug report on the Intranet and Quantic's internal QA then does the follow-up. We have our own in-house testing team of ten people. Steve and Caro, two members of my team in charge of directing and gameplay, conduct the first reviews (before the scene is sufficiently advanced to be able to speak of AD). I intervene when the scene is playable and try to bring it to the Beta level.



Quantic Dreams testing team (left), Caroline Marchal –Lead Gameplay (right)

Sometimes I also have to do some refactoring in the game design for certain scenes that fail to work as well on screen as they do on paper. This was the case for five or six scenes, which is quite manageable given the scale of the game. Having an in-house Motion Capture studio really saved our lives on more than one occasion, when an animation was missing, although I try to avoid having too much recourse to it at this stage in development.

For example, the first minutes of the game had to be completely rewritten just after the Alpha. I wasn't satisfied with the way we got into the game. It's really a luxury to be able to improve the design at this stage in development without burdening production. Quantic has invested a lot since the beginning of the project in improving its chain of production and tools, and this sort of limited change is relatively painless.

Lots of work in store in the weeks ahead: only two months left before the Beta, before preparation for Gamescom at Koln (two new scenes presented), then comes the Tokyo Game Show, then the Beta, then the playable demo, then the Master. Not much time to catch our breath, but it's essential to remain detached about what we do in these periods in order to avoid errors, let nothing slip by, keep the same high standards as all the team beginning to feel the accumulated fatigue of several years of development.

WEEK 3

After months (years in fact) of hard labor, I finally saw emotion appear in the game for the first time. Some already quite advanced scenes began to give the first indications of this in recent weeks but this time I saw what I had been feverishly waiting for for months: a moment of grace. It may sound stupid but as far as I'm concerned, it's the most important aspect of our work on HEAVY RAIN: creating moments where you forget you're in a video game, where you share the feelings of the characters, where you live their experiences with them. It's a unique moment where you get gooseflesh or your throat tightens, an instant in which we glimpse the magic of something that escapes us. There's no recipe for achieving this, no method and no technique. All you can do is mix up the ingredients, add all the passion you can give it, stir for three years and hope the sauce thickens and you're gonna witness this fugitive moment of emotion.

The first time I had the feeling was when I was playing Ethan's first scenes with his son. I really experienced moments I had experienced with my own son, and although I wrote the script and I know every word of it by heart, the scene surprised me. I watched the scenes unfold, scenes that were particularly difficult to fit in but in which we discover the nature of Ethan's relationship with his son.



Ethan with his son

I don't know what the journalists at Koln are going to think, but for me this scene is really powerful. There's an ambiance, the night falls in real time in the house until it is plunged in darkness and you have to turn on the light, the way his son behaves, the fact that he has to look after him, Ethan's sadness in this scene, so many things that I'd never seen in a video game. The characters are beginning to exist. We begin to forget they're just a load of pixels animated by a program. I think the players are going to empathize with Ethan. The feeling is there. It is strange, surprising and powerful.
I think I've found my favorite scene in the game.

The second moment of grace is the opening credits sequence. This sequence changed six months ago. Initially I was into something very complicated and pretty abstract, but I finally returned to something I think is much closer to the game and more powerful. During the preparation stage for HEAVY RAIN we spent two weeks in Philadelphia with some of the members of the team, taking pictures and filming, soaking up the atmosphere of the place. I chose this destination almost by accident, because I liked M. Night Shyamalan's films and because I read somewhere that he shot them there.

Philadelphia was a real shock. Assisted by a movie scout (who had worked on the film PHILADELPHIA), we entered the poorest parts of the United States, a far remove from the usual clichés we see in most of the movies shot in Hollywood. Barbed wire in downtown areas, abandoned factories, railroad tracks where no trains run, and poverty everywhere you look, misery, crumbling houses, factories behind schools, abandoned car shells. People roaming the streets aimlessly, empty-eyed, disillusioned, lost, junkies on the pavements, children waiting for something that will never come.
I came to Philadelphia by chance. I knew as I left it that I had found the location for my new story.
I hope I managed to put some of what we saw there into the game. Another America, the America of broken dreams.



Location spottings in Philadelphia

Getting back to the second moment of grace. After weeks working on the credits, it happened as I watched them with music for the first time. For the moment the music is just a rough, the definitive version will be recorded with a symphony orchestra in the Abbey Road studios in a few weeks, but it already communicates all the emotion I was hoping to hear.
No matter how many times I tell myself that the music is important and gives added meaning to the image, it always knocks me out. I was so happy to see how much Ethan's theme finally lent color to the game that I showed the sequence to everyone who passed by my office at Quantic… A real kid…

When you make a game, you wait patiently for the first time you hear its "voice". As far as I'm concerned, I heard it today for the first time. It's much stronger and more moving that I imagined.

WEEK 4

Cycle 5 was delivered today, along with the demos for Koln. A lot of work went into finalizing the builds, a lot of time spent improving the Motion Kits and lighting, particularly close-up lighting. Heavy Rain uses a specific lighting system for close-up camera shots in order to improve the quality of the facial rendering. In most games the lighting is global for the whole scene, which works perfectly well for long shots but begins to pose serious problems as soon as we want to do very tight shots (streaky shadows, lit-up mouths, fluorescent rims at eyes and lips, etc.). To avoid this problem we developed a lights editor associated with the camera editor, thus enabling the cameraman to create his own lighting setup. This produces a very significant improvement in the rendering (as well as a significant quantity of extra work…).



Version of the charcters without close-up lighting (left) and with close-up lighting (right)

Another significant improvement: the motor department found a way of increasing the definition of shadows in close-ups by dynamically reassigning the texture size for character shadows in cases of dynamic lighting. The few remaining artefacts caused by imprecise shadows have now completely disappeared.

This week's new idea: implement Bleaching! This technique comes from the movies and consists in bleaching bright zones of the image in order to artificially increase the contrast. David Fincher uses this technique (or at least its "physical" version) in the photography of some of his films.
We set up a means of integrating a rendering similar to 3D in the game. The result is pretty impressive, with significantly improved realism, particularly for the characters. The Koln demos will be the first to benefit from this.

The demo scenes for Koln finally needed a lot of work. Although the Scott Shelby and the Hold Up scene were fairly easy to finalize, the scene of Ethan with his son (presented behind closed doors) was much more delicate, given the complexity of the scene. I have absolutely no idea what the journalists are going to think of this scene. It's a slow scene that is based entirely on ambience and emotion and contains absolutely nothing spectacular. We play the role of Ethan who has to look after his son on his own whereas he obviously has a particularly difficult relationship with him. The feature I appreciate in this scene is the way time passes in real time. The scene begins about 5 o'clock in the afternoon and we see the light fade in the house as night falls and everything is plunged in darkness. The player has to turn on the lights to be able to see. It's simple and direct but this scene generates a really grim ambience. Shaun, Ethan's son, is managed by the game and a lot of work went into making him lifelike, so that he behaves like a real kid, an independent agent that the player can interact with (or not).

Hold-Up scene released during the Gamescom

One small detail: I wasn't happy with the rendering for Shaun, Ethan's son. I found it a little too 3D, charming but not really like a kid.
Sudden panic in the prod, we finally decided to scan two ten-year-old kids to see if we could improve the character's face.
The whole team went into overdrive to get ready for Koln (two weeks to redo a character from scratch and retarget all the facial animations).
The team did a great job on time but the integration had to be done the day before we left for Gamescom. Everything was a little too tight and so I decided to take the old Shaun to Koln. The team was a little disappointed. Although I didn't want to take any risks before Koln, I know that the work they did will be used in the game when I get back.

WEEK 5

My program for Koln: Monday, rehearse the Sony press conference that I have the honour to participate in. I did two or three runs in order to get used to the prompter (what a wonderful invention!) and get my bearings on the stage. Tuesday: final conference rehearsal in the morning with Andrew House and Kaz Hirai. It gave me a chance to get to know them.

Last minute revisions to my text (I'm on stage for nearly five minutes).  Last minute anxiety. Telling myself that my text is really out of place. In the midst of all the very businesslike announcements, there I am telling my life story. They reassure me and tell me everything is going to be all right. I tell myself they must be right and that anyway it's too late to start asking that kind of question.

I jump into a taxi and run to the GDC where I'm expected to make a speech in a conference on "how to make games for an adult public". I arrive dead on time (in spite of the rehearsals that were easily two hours late). Just enough time to set up my PC, the HF mike and notice that the room is packed.
A spontaneous outburst of applause when I announce that you should never allow a marketing person to have any say in the game design (easily done). A second outburst when I demand that ratings for video games be based on the same rules as those used for television and cinema (not quite so easy).
It's nice to feel in sync with your community from time to time. It isn't always the case.

I emerge as quickly as I can (a good half hour later, enough time to answer fifty thousand questions, exchange as many calling cards and shake as many hands). I dive into a taxi and get back to the Sony conference. No more rehearsals, this time it's for real…
800 people in the room. Impressive… Just for a moment I'm back in Los Angeles launching Omikron in 1999 (already ten years ago). A press conference at the House of Blues with David Bowie. 200 accredited journalists from all over the world, tens of TV channels. Mindblowing.
Just a slight flutter of the butterfly's wings when Andrew announces my name. I'm motioned onto the stage. There's a step leading to the stage. Mustn't trip on it. Shake Andrew's hand. Go to center stage. Take my time. Articulate. Speak slowly. Watch my accent.
Then the teaser. Applause (warm, I thought). A couple of words to conclude. Shake Andrew's hand before I leave. Don't trip on the step. I take a deep breath backstage as I announce to myself: Done!
I take off my HF mike and wait for the end of the conference. I thank Andrew and Kaz for having me there. I go outside to talk to the journalists, answer their questions, pose for photos. Dinner with a high-flyer from Sony to discuss recent developments, and so to bed.
A pretty good day, looking back on it. The feedback from the new teaser presenting Ethan and the emotional part of Heavy Rain was unanimously really very good. I think people were expecting something like that from the game.


David Cage during the Sony conference at the Gamescom

The next day was entirely devoted to interviews. Guillaume handled the Germans/Austrians (he speaks fluent German) and I took care of the rest of the world. A lot of journalists, saying the same thing hundreds of times over, answering the same questions until I'm speaking automatically, no longer knowing whether I said this to this journalist or the last one… A few brilliant journalists who ask really interesting questions, other less brilliant ones who ask less interesting questions… It's by no means sure that anyone has yet realized what we're trying to do… I put myself in their shoes and tell myself that I wouldn't understand a thing either…

Back to the hotel, dinner with the marketing manager and the Sony producer for HR. Then we start getting the room ready for tomorrow morning's presentation of Ethan's scene. The demo is absolutely horrible, the colors are hyper-saturated, everything looks green. A moment of panic, a problem with the code? Suspicions come to bear on the television. Emergency lights flashing in my head as we borrow a projector from the hotel, connect up everything in the room at 1 o'clock in the morning, hoping desperately that we're going to get something vaguely presentable. A nice surprise: everything works correctly and the grain of the projector is quite flattering for the image. Check the sound, adjust the image, position the chairs and tables. To bed at three o'clock. Up at seven. Time for a shower and a last-minute check to make sure everything is ready for the first session at 9 o'clock.
Didn't have time to see Pascal (the actor who plays Ethan Mars). We had planned to have him enter the room at the end of the presentation to let the journalists meet him.

Thursday morning, the first session with the journalists. We show them the scene with Ethan, then the teaser, and keep the surprise for the end: enter the real Ethan Mars aka British actor Pascal Langdale who leant his face, voice and talent to the character. I'm beginning to dread the presentation. Yet again it's a completely atypical scene that we're about to show, slow, nothing spectacular, based solely on the ambience. We see Ethan, very depressed by the death of his first son, having to look after his second son, picking him up from school, getting him to do his homework and preparing his dinner for him. The scene has two features that I think work particularly well: time passing in real time progressively plunges the house in darkness, forcing the player to turn on the lights in the rooms he visits. This gives a real impression of the passage of time in a way that few games have done so far. The second feature is the way the boy is managed as an independent agent with his own agenda and with whom we can interact freely.
The scene is reminiscent of the sombre ambience at the beginning of UNBREAKABLE by M. Night Shyamalan, or the sadness of Tom Cruise in MINORITY REPORT (relatively speaking, of course).
 

David Cage (left) with Pascal Langdale (right) giving interviews during the Gamescom

No explosions, no zombies. Just twenty minutes of living in the sombre ambience of the house as night falls, observing Shaun and trying to build a relationship with him.
For just a moment, I begin to wonder whether I'm suicidal or merely a masochist…

The journalists follow the scene with a degree of attention, although I can't figure their reaction. Then I announce a surprise and I have Pascal enter the room. The surprise works. They've just spent twenty minutes watching a 3D character and suddenly they seem him enter the room. Pascal is his usual self, kind, interested and interesting, brilliant. It's easy to see the journalists are more used to talking with programmers than actors, but the discussion is open and interesting. End of first session.

In the course of the day six one-hour sessions along the same lines. Overall, the reactions are very positive (at least as far as I can see). Some of the journalists are putting two and two together and beginning to understand what I'm getting at. Others still seem to be groping in the dark… One of them tells me he's no longer interested in games because he finds them devoid of meaning compared to movies. Another tells me that most of the games he sees are for kids and adolescents and that none of the content appeals to him. It's good to see I'm not the only one coming to that conclusion.
Our industry is about to go through an identity crisis: we continue to make games for twelve-year-old kids whereas the average age of players is 35.
In any case I learned that a lot of journalists have come to realize that video games have reached an impasse and that new answers have to be found. I don't know what will happen to HR but I think it's at least an attempt to introduce something new and different.

As ever, impossible to know in advance what the journalists are going to write after this type of presentation. There are usually three categories: the enthusiastic ones who think they have seen something really new and who believe in it, who want to share their feelings and who are willing to stand by them. Then there are the ones who don't really know what to think because, well, they're waiting to play the game in order to form an impression, and who settle for reporting what I said and showed second by second. In general these are the ones who think that if they say it's good and it turns out not to be, they'll lose their credibility, so they prefer to be as factual as possible and reserve their opinion for later. The third category is the cynics, the ones who know in advance that it's not going to be good, who have no idea what I'm trying to get at because they think that video games are fine just as they are and there's no reason to change anything.

Journalists with convictions are increasingly rare. They're quite willing to stick their necks out when the game has been released and everyone has said what they think of it. The ones who analyze, dissect, think and have the courage of their convictions are becoming more and more rare.

Overall I'm very surprised by the number of journalists who are tired of video games and who are desperately waiting for something new to happen because they can see we're stuck in a rut. I can see in their eyes that they really hope HR will be the game that will finally break the mould. So do I…

That evening, dinner with Pascal to share our impressions of the day. He was a little disappointed by the games presented at the fair. He stopped playing when he was fifteen and he has the impression that nothing has changed since… Join the club.

Friday, a new series of interviews, with American journalists this time. Slots of a little more than 30 minutes (instead of the usual 15) enable us to have slightly deeper discussions. I'm beginning to be exhausted by my week. I feel I've been saying the same thing thousands of times over (it's only an impression). I'm beginning to feel impatient to get back to the studio and do some work on the game. I know there's an enormous amount of work to be done in order not to disappoint the people I've met here.



The queue lining to play Heavy Rain at the Gamescom

Except that Monday I won't be at Quantic but in London to supervise the recording of the game's music in the Abbey Road studios with a symphony orchestra. I can't wait to get there. It's the kind of thing I wouldn't miss for anything in the world. I take Steve my assistant director with me (he'll be integrating the music into the game). We'll have to wait and see whether I stay for the full four days of recording or just a day and then get back to Quantic.
Between E3, Koln, the recordings and the Tokyo Game Show, I have to be careful how I spend my time and avoid becoming too dispersed.

WEEK 6

Koln behind me, just enough time to go home, unpack my case, take an afternoon run in the Bois de Vincennes and lie on the grass by the lake, then it's back in the train to London. It's the beginning of the recording sessions for HR in the prestigious Abbey Road studios. I go there with my assistant Steve who'll be supervising the integration of the music in the game.

We get there the day before and curiosity leads us to go by the studios to check them out. A plain enough little street in a residential district. About twenty people in the street at 10 p.m. taking photographs of each other on the pedestrian crossing. This is the right place all right…
Suspense the next morning: Norman Corbeil the composer delayed his flight from Montreal to have time to finish writing the last cues (more than 275 in all for more than 75 minutes of symphonic music). If he has the slightest problem, there will be 75 musicians waiting for him.

The suspense doesn't last long. Norman is in the studio garden drinking a coffee and waiting for us. He hasn't slept much, he didn't manage to finish everything, but he plans to finish the last scores in his hotel room between two recording sessions (four days planned in all). He has rented a piano for his hotel room in order to finish the last scores.

A word about Abbey Road, a point of convergence where, apart from the Beatles, many movie soundtracks have been recorded, including The Lord of the Rings. I discover the studio where we're going to record. A large room, 75 music stands in position, a rostrum for the director, a smell of old wood, apparently neutral acoustics, no resonance, which is surprising for a space of this size. The room was transformed into a recording studio in 1931, making it the oldest recording studio in the world.

The musicians arrive one after the other. I take my place facing the orchestra, behind Norman who will be directing. A great collection of strings, brass, percussion, a harp, a piano. The score is already on the stands, several months in the writing, several days' recording, and just over one hour of music at the end of the day.

Everyone is in position on schedule, bows poised over the strings. The moment of the first note of HR. I hold my breath. I wonder what's going to happen, whether I'm going to like what I hear, whether I'm going to feel anything. Whether we haven't been fooling ourselves all these months.

The first notes raise the hairs on the back of my neck; Ethan's theme fills the studio, swelling with emotion from the breath of the brass and the stroking of the strings. Ethan is born for the third time, first in the writing, secondly in the image and now in the music. This theme will give colour to Ethan's story. For just a second, I think back along the road that has led me to listen to a symphony orchestra, the only listener before 75 musicians. An idea, a desire, work, the effort to share a vision, more work, bringing together 200 people, more convincing, always convincing, to arrive here at this instant listening to the music of my story.
Quite a hike!
I wouldn't give up my place for anything in the world.


Abbey Road recording studio

The cues follow each other. They do different versions, louder, not so loud, with more brass, fewer strings. The musicians have never rehearsed together. They are discovering the score for the first time and interpreting it directly. The orchestra is organized like a little army, with leaders for each section and the first violin who acts as an intermediary between the conductor and the orchestra. He's the one who keeps them in rein whenever necessary and restores silence, he's the link between the conductor and the orchestra.

I'm surprised by what I hear, by the different colours, the richness, the life and emotion that is conveyed in every note. The orchestra is capable of murmuring with incomparable gentleness and, the next second, swelling to its full power without ever becoming deafening. Less and less often we have the opportunity to listen to real symphony orchestras playing live. Classical music being more and more often replaced by other types of music. If you've never listened to a real orchestra, do so now. You'll see what I'm talking about (oh well, all right, then listen to the music for HR, it's a good start;-)

Barely two days later I decided to go home. The recording is coming along very well without me. I haven't been to Quantic for nearly ten days and there's a ton of problems waiting for me on my desk… I leave some instructions with Steve for the end of the recordings and I set out for my train with my heart in my boots. I would have loved to stay there and listen to the orchestra just two metres away from me and not miss a note, chat with Norman and the musicians about whether to play it piano or mezzo forte. But the days are numbered until the end of the project. I have to make optimum use of my time.

WEEK 7

Back to the office. A billion meetings late, decisions to make and people to see. My days are divided between meetings, follow-up on certain technical questions, reviewing and passing the directing.
Cycle 6 has just been delivered. We're advancing with the last two cycles, the ones with shorter and simpler scenes. There's a lot of work to be done on most of the scenes in the earlier cycles both in terms of settings and visuals.
I decide to concentrate on the Shelby scenes which have fallen behind. A lot of delicate scenes, with tense dialogues and some sustained action scenes.
We modify our approach to the crowd scenes that are dragging behind a bit. It seems the 1000 characters we have to display to fill out the Mall scene at the start of the game are finally beginning to fit into memory. Still a lot of settings for this scene before it reaches the desired level. I knew as I wrote it that it would be a nightmare to develop. I wasn't wrong. A meagre consolation. One year on a scene of a few minutes, a lot of work for a technical performance (and a dramatically important moment in the game).
 


Artwork from the Mall scene (left) and in game picture from the Mall scene(right)

The feedback from Gamescom was really very good. Strangely, Koln was a sort of pivot in the press's perception of the game. The teaser with Ethan met with exceptional approval, as did the press demo of the scene with Ethan and his son. I was really reticent to show this scene to the press because I found it particularly slow and depressing. It works really well in the context of the game, but out of context, it could be misunderstood. In fact I think it surprised a lot of people, even destabilized some.

Some reviews surprised me too. Journalists explaining that in the teaser they found scenes they experience with their own kids. Others were reminded of moments in their childhood when their parents separated. It's the first time I've seen this kind of personal comment in an article on a video game and it really touched me. I've never been so convinced of being on the right track. If the final game can touch personal memories and feelings in players, we will really have accomplished something.

The whole team in the MoCap room today for the Koln report. We show my appearance at the Sony conference so that the team can understand the message that goes with the game. I sum up my impressions of the fair and my feelings about the way the game is perceived. I tell them about my visit to Abbey Road. In short, I make sure everyone has all the information and can see the captain on the bridge. The team kept track of feedback from Koln on the net. They are more familiar than I am with what's being said about HR. It's good to know good things are being said about it. The coming weeks will be difficult. This is no time to start having doubts…

WEEK 8

Time has speeded up… September already. The Beta in two weeks. Since I got back from Koln, the whole team has been in intensive crunch mode. Steve has started to integrate the music with the scripters. I've been reviewing all the key moments in the story in order to rework the directing. After doing a big pass on the Ethan scenes shortly before Koln (particularly the one that was presented to the press), I have switched to Shelby's scenes. The majority of the work was on Kramer's Party, a scene where Shelby and his "partner" go to investigate in the millionaire's villa in the middle of a somewhat depraved party. A lot of work. The scene was one of the first ones we worked on and it needed to be seriously upgraded. I spent a week on it, just on the directing…



Close-up of Shelby

A few Shelby scenes later, I switch back to the Ethan scenes. A lot of key scenes, obviously, based enormously on the quality of the acting. Rather pleasant surprises while filming the scenes. I rediscover a lot of what I saw on the shoot.
The current work is a little strange, a mixture of using the data shot but cut up (all the animation is cut up for use in the game play) and "added" data. For example, all the looks between the characters are added in procedural fashion. We mix the micro-movements of eyes captured on the actor with directions imposed by the director during the directing (we can thus adjust the direction of the looks to the nearest second).
The same thing with the facial animation: although dialogues are shot specifically, everything that is "non text" (reactions, silences, etc.) is managed by means of a library of facial emotions that the director can use as he wishes. For example, we can decide that a character will react to a line with a smile or, on the contrary, that he will be angry. Each character in HR has a library of several hundred emotions ranging from sadness to disgust, seduction to hatred. All these emotions were captured in several versions with the actors in motion capture.

The power of the director thus becomes very important during the integration: far from merely setting up the cameras, he practically takes part in the actor's performance by choosing his expressions and directing his looks.

The difference in the result is quite spectacular: no facial between lines and no looks. We witness a flat dialogue between two dead fish. The dialogue suddenly springs to life when we integrate all the elements correctly.
This power demands that I be particularly vigilant with regard to contradictions. A smile added at the wrong time and the meaning of a whole scene changes.



Series of generic facial expressions

Camera integration is done via a module called "Sequence Editor" in a tool called IAM. IAM is a scripting tool that enables us to assemble all the elements in HR, to position triggers, to define conditions and variables, to configure the AI, etc. In short, the game is really created in IAM.
The Sequence Editor is a real time dedicated advanced editing application. It can be used to summon up all the game data (characters, animations, sounds, FX, etc.), assemble them according to conditions, and add cameras if we wish. Given the cinematographic perspective in HR, it is obviously a key application in development.

We systematically use editing tracks and a time line. We can place animations on a track, duplicate them, cut them up, resize them and accelerate them as we wish. We can then create a camera track and move it in real time as the animation is played. We can obviously record camera movements and create as many as we wish depending on our directing.

This tool played a major role in the idea of freeing the camera from the character. HR attempts to create an experience in which the camera is not permanently locked in position behind the player but instead plays an active role in the narration, as if the player were playing in a cut scene. He retains control permanently while still enjoying the benefits of veritable directing.

This cinematographic approach has often caused misunderstandings with regard to what HR is really going to be. Some people think that because we are working on the narration, emotion and directing, the game will be one long cut scene. This is obviously false. The player is permanently in direct control of his character. But we work on allowing him to tell his story with his actions, to live his different emotions by interacting.

WEEK 9

End of Cycle 7 today. The last cycle contains only epilogue scenes, therefore lots of scenes of limited complexity. We decided to use only one cameraman to film them with only one scripter as backup, the rest of the team being devoted entirely to debugging the rest of the game. Only two weeks from the Beta, there are still 200 bugs flagged "Beta Failure" (bugs likely to cause the Beta to be refused). It's doable but we're going to have to focus on this target.

The whole team is now going to go into review mode. All the managers will spend a maximum amount of time reviewing the game and dissecting, scene-by-scene, pixel-by-pixel, everything that can be improved. I set up a second review room with a projector for joint reviews on a big screen. I know that the quality of the game will depend to a large degree on the rigor of this long and painstaking job. Play, play again, analyze, observe, enter the bugs to the team, start the new build, play again, see what has improved and what has broken. Begin again.

All the languages have now been implemented in the game. It's strange to hear Mad Jack speaking Japanese. The localization looks good, insofar as I can judge… It works in any case. Just as well, the TGS is coming up next week.



Lineup  to play at HR at the TGS (left), Guillaume de Fondaumière –executive producer- presenting the game at the TGS (right)

The two children we hastily rescanned just before Koln have now been integrated into the Build. Overall, they work much better.

More changes in the main menu this week. I'm not quite finding what I'm looking for. Some things have been simplified, others made more complex, in search of something that is both simple and original. It's hard to tell how much time we spend on parts of the game that no one will really notice in the end. But I believe we're slowly reaching the final version.

On a completely different track: it's incredible how many people have projects they can't talk about. It's top secret. Because if we heard their idea, we might steal it.

Everybody has ideas, it's the easiest thing in the world. They say that every waiter in Hollywood has an idea for a scenario. In the world of video games, every developer has an idea for a game. Anybody can have an idea. The hard thing is bringing them to development.

It may sound stupid, but I've never had any difficulty talking about my next idea. On the contrary, it often enables me to 1/ practice formulating it clearly, 2/ convince myself that it really is a good idea, 3/ check that I still think it's good after explaining it fifty times over.
Nobody has ever stolen an idea from me.
Not surprising really. Generally people don't believe in them to start with.
That's a good way of recognizing a really good idea.

WEEK 10

I've decided not to go to TGS. Too much work in the office. I can't afford to take a week (plus jet lag) right now. A pity. I would have liked to stroll around Akihabara… Oh, well, next year, I guess… Guillaume is going to go there on his own. He has done all the breakout sessions with me since the beginning of HR. He'll be able to handle it…

Reviewed the first scene of the game for the fiftieth time.
Fed back 25 pages of bugs. Not too serious, but still 25 pages…

I listened to the first scenes with the music integrated.
I can confirm that it changes everything. Good news, the addition of 75 minutes of music in 5.1 did not explode the build.

However, we still have stability problems on big crowd scenes. Memory leaks finally flushed out today. Things should look better tomorrow.
Found a better idea for the break.
Found a better idea for the closing credits.
Suggested a third redraft of the Tutorial in the first scene of the game. This should be the last one.

No connection at all: went to see Inglorious Bastards last night.
Tarantino… that guy is a genius…