Behind a massive project, another story [mount inc.]

Suta-san, did you surpass your past? When this project started, you said “this is a revenge match.”

Well, the motion’s better than before… than before, at least. It’s becoming more refined than the rough movement it was.

When Im-san looked at it, his first impression was like “oh, that’s great.” We got feedback, of course, but we also got a “not bad!” out of him.

Last time was my first year, so it’s a two-years-in-the-making revenge, I guess. For me it’s only about “a bit better than before,” but the reactions around me feel different.

For you, Watanabe-san, it’s a “challenge,” right?

That’s right. With GYRE — though AD Hayashi-san checked and supervised, of course — it was an environment where I was fairly free to think and propose from the planning stage on our own basis, to some degree. In that sense, it’s the biggest challenge since I joined mount.

Your first challenge since joining, then.

Yeah. And there’s a lot to reflect on. The design itself reaches a quality, a standard Hayashi-san can approve of, I think — but there’s more to reflect on than not.

Do neither of you have much of a “nailed it” feeling?

A “nailed it” feeling isn’t going to come…
Because I know AYANA.

Well, there are parts you can’t simply compare, but once you see again how amazing the people above you are, you know…

I definitely had more of a “nailed it” feeling before. Because I knew nothing back then. I was full of confidence born of ignorance.

So after two or three years at mount, seeing all sorts of things and experiencing projects, what you feel isn’t a “nailed it” feeling.

It’s more like a lingering frustration.

It means the place you’re aiming for has gotten higher, right?

My internal standard has risen dramatically — overwhelmingly. So what I make myself looks tiny.

But that’s a wonderful thing, isn’t it.
To be able to stay in a place where you can’t be satisfied is quite a blessing.

Tiny as it may feel, I think you gave your maximum. What did you do, and what was memorable?

This is a little off from the question, but first I want to say what I thought was good about this project. On other projects, it’s the CD or AD who decides the direction and plan, and each involved member is given discretion, differing in scope and place. This time we really did it all together from the very first stage, so I felt more strongly that we were all making one single thing — that was good.

Right now I’m so happy…
because you named something we achieved as a team.

It’s something I think is the good part of this project.
In the design, doing the grid finely was a point of care. Doing it that finely and precisely may have been a first. I drew fine grid lines and asked Suta to fit both text and photos snugly onto them.

It’s a back-breaking design, from the implementation side.

I think that was the hard point for implementation too.

A fine grid plus a mass of photos — either one alone is heavyweight, and it was a double punch… so yeah, it’s tough.

I bet everyone wants to hear about the WebGL.

No, I only did simple things, so anyone could pull it off. There’s a theory it’s heavy precisely because I did it in a way anyone could pull off. I really didn’t do anything special…

Even if you think it’s no big deal, I suspect you’re doing things that make everyone around go “whoa.”

A point of care is that the maximum size of the mosaic aligns entirely to the cells, the grid. Also, when you scroll, the grid itself stays still and doesn’t move; only the text and photos move, made sticky. I also made the sticky start at the middle of the screen height.
There’s that tiny pleasure of text clicking snugly into the 1-pixel grid.
It’s minute, but this delicacy is mount-ness, maybe. It’s not perfect, though…

The mass of photos was the very first thing we verified.

I underestimated it. I underestimated it too much, and it’s not just images — once text comes in, it’s completely different. If it were image-centric without much text, it might have been a bit lighter. Since I use WebGL the images move fairly nimbly, but there was such a mass of other parts that performance didn’t go as I wanted.

If you’d thought through how to display a mass of photos, including all the various factors, it might have turned out differently, perhaps.

That’s the point to reflect on, I guess. It was my first time, so I knew nothing. From next time, I have more ways to do it — more options.

mount doesn’t compromise on either text or photos. So those options seem likely to serve you on any project from now on.

mount-ism, that is.

Oh — I want to mention that you fixed the smartphone bug overnight.

Going from an overnight, hopeless situation to it displaying properly on smartphones — a year ago I don’t think I could have. I don’t think I’d have reached a launchable level in that time frame, nor come up with a fix for the mass of images not displaying. Solving it without calling Okabe-san in the middle of the night is a big deal, maybe.

(laughs) You did hesitate, didn’t you — “Okabe-san might still be awake.”

You specify all the images’ positions on the implementation side — and is the path they move along set too?

The route is the nearest point. It reaches a point near the next photo to switch to — up, down, right, or left, at random. At first I did it fully at random, but since the distance is short, one move up/down/left/right and it arrives.

You set something like a goal and it whooshes there from where it is?

Right. Based on the top-left of where it currently is, it heads toward the top-left of the target — the next image to switch to — with no diagonal movement, aiming left/right or up/down first. At the shortest, it arrives in two moves, one turn. At first I did it with detours, but that got too messy so I dropped it.

Is it like you specify it for each individual image?

I don’t specify it. It knows its own info on the spot, right? For the target info, from this next fixed layout it measures its current distance and takes the nearest one. Once taken, that one can’t be chosen again, so it’s removed — decreasing one by one. It’s common in games, maybe? It might be something like pathfinding.

What we set up as the concept was “chaos.” While digging into “what is Shibuya’s chaos?”, the grid paid homage to a blueprint.

A point I reflect on is that I got pulled a little too much by the word “chaos” — that wasn’t good.

The mixing of cleanly laid-out parts and chaotic parts — that was chaos too.

Yeah, that’s right. “Chaos” somehow ends up meaning something broader, more chaotic and murky.

Each person’s definition of chaos is too wide-ranging, so it was hard. The Shibuya we expressed this time is exactly an ideal sense of Shibuya’s chaos. In the implementation’s direction, we expressed today’s Shibuya transforming so readily.

Even forced neatly into a mold, Shibuya’s chaos doesn’t vanish so easily. The message is: the blocks are being tidied up now, but we can keep Shibuya’s chaos, right? Even if it’s forced into a mold and looks neat at a glance, Shibuya’s charm doesn’t disappear so easily.

All sorts of photos are combined, and Shibuya really is that kind of town. It was formed by all kinds of cultures mixing together.

We did the location shoot ourselves as the photographers. I think Liu-san’s and Watanabe-san’s photos added a good flavor. How does it feel, reflected on the website?

Oh, it’s an honor. It was fun.

The time I could purely enjoy without stress was, somehow, when I was taking photos. With design and such you get all kinds of feedback — which is a given, of course…

Me looking at 1,500 photos and saying “couldn’t we make an imaginary Shibuya by collage?” was the beginning of hell, maybe…

Yes (laughs). I think each person has an image of Shibuya in their head — that kind of vague Shibuya, something that isn’t the real Shibuya. The photos are real, though…

We did location scouting several times, and asked for a full day of shooting for the real thing, almost until the middle of the night. So the reality of today’s Shibuya is captured in each individual photo, but collaging them creates an imaginary Shibuya — and this is what suggests a future-city Shibuya… right?

Exactly!

Talking so far, what I think is — there are points to reflect on, things I could have done better, but I think we saw it through to the limit of our current level.

At the start we talked about revenge and challenge. Was there a moment like “I surpassed last year’s self!”?

Aren’t I always surpassing it every day? I’m always surpassing the self of one day ago. I’ve lived longer than yesterday’s self, so I’ve got to surpass it. As a member of mount, I surpass yesterday’s self — growing day by day.

And from a designer’s standpoint? At mount, the internal evaluation of design is strict too.

One thing I can say is that if I weren’t at mount, I wouldn’t have gone this far. It’s precisely because I’ve come to know the level mount demands that I can do it. In that sense, I think I’ve been able to grow.

Do you have a “next I want to do this! I want to be like this!”?

This time’s flashy expression and motion can paper over things. The best is when even plain implementation makes people go “somehow, this is good.” Flashy motion gets its share of “whoa, amazing.” This time there’s a fair bit of motion, so implementation and design hold it up about half and half.
Really, the design is about 90% complete, and the 10 or 15% that makes it even better is implementation. Being able to do that is what I think makes you a pro, skilled.

It’s good to add, to a nearly finished design, motion that isn’t superfluous — that’s definitely better for being there. I want to be an implementer who makes design leap further.

So it’s not taking a 50% design to 100%, but that there’s a 100% design and the ideal is how far you can take it — to 150%. So reproducing it down to the pixel is a given, and then, to make it look even better, what do you add…

Right — there are still parts that haven’t even reached 100 from 100, but 100 to 101, 102, 103… implementation is the final plating. The flavor (the design) is already guaranteed to be delicious. And how you further plate it (implement it) raises the diner’s evaluation. Design is the greater one — because it makes something from zero.

You’re in a state of completely believing the design is 100. If implementation is the plating, then what is design?

Something like a blueprint drawn without a ruler, maybe? On this project I felt that having parts of the motion I couldn’t fully imagine really isn’t good. I want to talk as equals about the motion — “let’s do it this way, that way.”

It has to be equal, right. If one side goes above, the other’s opinions can’t be picked up. Not a design saved by the implementation’s motion, but a synergy — where because the design was there, the implementation can generate something newer like this.

There’s a reason — why this motion, why it has to move this way — and you all design with that, right?

And, the relationship of trust is a hugely important thing.

Whose evaluation do you want, whose words do you most want to hear?

People inside the same company, and more specifically Hayashi-san, Yonemichi-san, Im-san. And, what else — if someone says “what a great design” on social media, that lifts your spirits.

But number one is the internal voices… right.

Yeah. Precisely because I’ve come to know the level of the people above me, there’s that “I want to be evaluated by these people.” Precisely because I’ve witnessed how amazing they are, I want to be told “not bad!”

You crave that single phrase. And on the implementation side?

The previous GYRE project got a fair number of quote-shares, and companies strong in interaction and such reacted — that was gratifying.

It’s gratifying to have people in the same industry think “this person’s wild, not bad!”

There’s also the side that doesn’t want to be seen. “Is it okay for this to be seen?” — whatever site I release, I always have both feelings, wanting it seen and not wanting it seen. The not-wanting might be stronger.

Slightly special motion, where your taste is put to the test — I don’t want to be told about the shortcomings in the very place an implementer’s taste shows most… because I know my own immaturity.

You don’t hope for a word from TD Umetsu-san?

I don’t want Umetsu-san to see it….