If this wasn’t a Canon advert why did they need the cameras?

I believe it to be both real and completely ridiculous.

Looks real to me.

Seems to be like that many flashes would be A LOT brighter, and would wash out the video camera filming the music video. Even one camera flash from the side usually completely washes out a single frame of video.

the interweb says
“Their team was inspired by the track’s title and the lyric “not to make it a memory” and used the flashes of 250 Canon cameras to create light animation that appears throughout the film, intermittently illuminating the band’s performance on the stage. Like this recent clip for Bell, technology–Flash, openFrameworks and Arduino–helped to achieve dazzling effects without any postproduction. The video is also featured on a dedicated site, where viewers can pause the clip to see photos taken from the cameras during the shoot and can also create their own version of the clip with a special message.”

So maybe real.

My first thought was “Hell no, this isn’t real. This has gotta be another tea light animation…”

After looking more closely, though, I’m impressed. Each flash lights up what is directly around and in front of it, and you can see the reflections from that on the camera/flash body in each ‘pixel’ of the wall of cameras. After doing a quick frame-by-frame examination, it looks like it’s at least somewhat legit.

Any decent camera flash can strobe like that, very easily, at 50fps+ without a problem as long as it has power. Seeing that they’ve got everything hardwired in (note the huge power strips in the making of video), I’m pretty sure they’ve got plenty of available current to recharge the caps in those flashes.

Also, they appear to be DIYer’s/hackers at heart, anyway. They’re geeky enough to specifically note that they’re using Arduinos and OpenFrameworks, so why couldn’t they figure the rest out, too? They’ve got the LED matrix built, and after that it’s as simple as wiring a relay for each pixel to a flash and setting it up.

Canon may or may not have sponsored the video. TimeSliceFilms.com may have also supplied the cameras and shot the ‘bullet time’ scenes. Either way, it’s a pretty badass video, and I see no reason that it would have to be faked. It’d obviously be an expensive and somewhat complicated undertaking, but definitely possible.

A Canon 60D with stock lens is ~$900USD. They’ve got huge boxes of them and most likely got a volume discount if they outright bought them. It’s also likely that they rented a large number of them (which is common practice for smaller-scale studios).

I’d estimate that this video could be made for under $250k, and probably under $100k if most of the equipment could be rented instead of purchased.

the cameras look real to me – so what’s the point in faking the flashes? Reflections on the floor and on the lenses / bodies look ok, too. Besides, it is by far easier to fire the trigger than to replace the whole setup with cheaper flash units or do a photorealistic animation. The only thing that might be necessary is messing with the frame rate or pre-trigger flash sequence.

After another quick look through it, I’m seeing even more evidence that says it’s real. Pause it on a frame where there’s a high rate of motion in the same direction. You can see the the dark spots where the flashes were illuminated in the previous frame. This is artifacting from the CCD in the camera that’s filming it.

Also, hpux735, they’d have cameras mounted on tripods or on a multi-camera jig so they could all be fine-tuned to focus on the same central point for the ‘bullet time’ shots. That most likely was shot after the initial video and was edited in.

Looks real to me, and yes, high quality flashes can recharge and discharge that quickly. Especially if they are set to lower power (like 1/128th). Though not sure why they have lenses and cameras to go with it? My guess is some of the cameras and lenses are either lower quality or just dummies.

Maybe they are being sponsored by Canon… the video does go out of its way to show that it’s the 60D, Canon flashes and the kit lens. Although it has been out quite a while…

Can’t the flashes be tuned down anyways? You don’t need that much light from each individual flash, and even in 1080p the picture is a bit blurry, maybe it was filmed with high ISO…

It would be much more complicated to film the stopmotion individually and then have the band perform alone while recreating the lighting…

I’m with PoisonWaffle in saying “real”… Flashes can have insane recharge rates at low power.

Maybe the cameras/flashes are rentals? I wonder how much that’d run for a day of renting.

This is plausible from a technology standpoint. The cannon speedlites in this video have a recycle time of 0.1s at lowest power.

Nothing in the video exceeds that capability and low power would explain the exposure not getting completely fucked.

Are this faulty 60D ? or something :D ?

Not sure if those flashes can work at such a rate, as stated above, this would be super expensive unless they have some kind of Canon sponsorship in which case it is just an advert. I have to call fake on this one.

I agree it is probably a gutted flash, perhaps camera. Not all cameras are 60D’s, and not all flashes are the same, many not even the EX series (maybe not even canon). Given the difference in power from the various flashes as well as different recycle profiles and reflector specifications, to try to match the output of everything by means of control would be daunting at best.

Beyond recycle times, the burn time of each is silly long in some cases, even if you try to argue matching to the frame rate, not possible. As well, on low power, the flash durations of speedlites is shorter than the scan time of teh video camera. So even if you could match 30, 40 60 fps, that does not mean the ENTIRE frame of each would subject to captruing the flash.

My call is fake, in so much as it is not as it seems.

Fake or not it’s a cool effect for a video shoot. Food for thought. Please keep in mind if anyone wants to implement an effect into a live concert it can be problematic to the audience. In my case, I experienced it first hand. It was dark and the lighting was great up to the point where the audience was being flashed. Our pupils were in a relaxed large state. Then we started getting flashed by the lights from on top center stage. We we getting blinded and it distracted us from enjoying the band because we had to block the flood lights with our hands. I am very surprised they didn’t notice the first 12 rows were all looking away from the spotlights. As for using this in a Video shoot – It’s an awesome effect that projects an original look. A great way to breakout their video from the rest. Really impressive to me is how many cameras were collected to make the video. I can see some poor soul standing in line at Walmart returns with two shopping carts of cameras :-P hehehe jk

TY for sharing

Partly fake.
@1:24
A flash can’t reload that quit (instant)

I would say FAKE. In the first video at 1:25 the flash on most cameras stay on for like 4 or 5 seconds. Camera flashs can’t stay on for that long as far as I know. It’s a quick burst and that’s it.

My 2 cents

I’m with all those that say that it is partially fake. Did anyone else check out the website? (http://www.androp.jp/brightsiren/)
If you go there, you can get your own message displayed during the video. Of course, you get only the “flashes” without any band members. But it looks a lot like the stuff that’s going on in the background of the video.
Given that they implemented that just for fun on their website, I believe, that they are also able to do something like that in AfterEffects/Vegas/Premiere in a convincing way.
So not completely fake. But fake.

I’m less intrigued by the cameras than I am that this song starts of in 9/4 time.

I’d say it’s possible.
The strobes are not putting out much light. On full power, that many strobes would over-power the cameras they are using to film…that entire stage would be washed out for the flash. Instead there are these thin bands of light…reasonable, for the lowest power setting.
I have a $50ish YN-460. Lowest power setting, I can fire 19 low power flashes in a row, as quick as I can hit the button. The longest repeat of strobe fires I counted was somewhere in the 20 fire range…a better strobe could pull that off, easily.

For most of the video, there are a minimum of three-five camera with intervals between the scrolling wall…three-five fires, three-five intervals to recharge the capacitors, which, for a low power burst, doesn’t take long…less than a second or two.

Possible. The hardest part would be wiring it all up to something to control it…the strobes themselves are no problem.

They are doing it the hard way but it looks real to me. I used to work for a lighting company Highend systems and they make a strobe system called dataflash http://www.highend.com/products/effects/dataflashaf1000.asp It does the same thing but using DMX512 digitally addressable and dimable.

Most of the shots are real.

I’m wondering if they synced to the camera’s shutter, some shots it looks like its in sync other shots it appears to be drifting out of sync.

As for it being too bright, or not looking bright enough, that all comes down to the camera’s aperture, shutter speed, film ISO or gain and what filters are used. I would have shot it alittle brighter looking to milk that camera flash look. It looks more like LED’s in this video than camera flashes.

I’m with the “LEDs replacing the actual flashes” (at best) camp. Flash duration is too long for an actual strobe (and getting video synced of a strobe tube is hard.)
> high quality flashes can recharge and discharge that quickly.
Perhaps. But can they be triggered at that rate through the camera? (Of course, they could just be using the cameras as particularly expensive flash mounts. +1 to the comments about the camera+flash setup being extremely wasteful.)
I don’t know about low power being dim enough to create the video as seen. At 4:00 we see a band member with an actual camera behaving and looking more like I’d expect from an actual flash.

When I saw the photo at top, I thought, “That’s got to be the flashiest invitation to a bris I’ve ever seen.” Mazel Tov!

Leave the lens cap on, and send the camera a “focus” signal. You’ll get high speed, low power strobe for up to a second.

It makes sense that people here are thinking more like engineers and not like artists: it’s music (art) and video (art). Having all those cameras is artistic and a great hack.

Sure they could have done it with a mish-mash of different camera’s but why wouldn’t someone pitch the idea to a camera company to sponsor it.

Quite well done, I’m going to say that it looks legitimately enough put together.
To the folk saying that the shots from the camera seem off, consider the field of view of each individual camera will change considerably across the range of that wall/shelf. Like an exaggerated effect of how your eyes see two separate points, enabling you to see “3d”.
On that note, how many “D” can this be considered to shoot in?

and i think flashes should have looked square because the opening for the bulb was square and yet they are perfectly circular and the light from so many flash bulbs would definitely destroy a cameras light balancing or contrast or blind it completely

The 580EXII manual does say it has “stobe capability” at up to 100+ fps.

if you watch just the animation you start to get a hint at what is really going on. Do the flashes make the graphics…oh yes, are they stock flashes, possible, but not necessary. If they are stock flashes then, the boards on the back of each flash make sense. They are NOT flashing in real time and each frame is based on the real time refresh of the flash. However if they gutted the flashes you no longer have a flash but a strobe which can refresh fast but would need external timers and addressing to do the marque.

I think the cameras are real, but its not the flashes we see. First, I think it would be much brighter, also I don’t know if a flash can fire that fast. The batteries would be a problem, but they show a bunch of AC plugs with means they are probably not using batteries. I don’t think they bought the cameras, they are either rent or models. Also, if you look at the flashes at 2:51 ~ 2:53 you can see 2 black lines comming from the sides, wich suggests they put white LEDs in front of the flashes, and those are the lights we see. This explains both the brigthness and the fire rate.

maybe fake simply because:

- at 00:12 we can see that the flashes were rectangular, so when in action, they’d give a wide oval flash
- at 3:55, we see that the flashes in action and all were a pretty perfect round white glow.

Most probably they’re some kind of bright LED based lights or something..

If you pause the first video at 0:14 you’ll notice those flashes are all different.
Another problem; if the cameras are shooting every time the flashes go off they’d have to be shooting at a much higher fps than low end Canons are capable of. If they aren’t synced and only the flashes are firing then how did they do the bullet time video?

i’m thinking they’re using a LED in front of the flash. look at the making of video at around :25 and you can see a wire going to the front of the flash.

@rob Are you sure?

I would be able to quite comfortably set mine to strobe at min power and then set it off at what looks like well below 20 fps.

The most common application for flash requires the illumination of a subject. Conversely lighting up a bulb just enough to see it come out in a video requires next to no power. Remember flash power is measured in hundreds of WattSeconds. How many watts are required to light up a filament enough so a video camera can see it?

The video obviously has computer editing, but I call the concept of the flash strobing marquee real.

Response:

1. OMG! Too fast!

Strobes can recycle very fast, indeed. It’s just a matter of charging, duration, and intensity. (RC time constant, Mr. Kelvin, Mr. Ohm, Mr. Watt, et al.)

(Whether these particular strobes can do this, I don’t know. Let us debate that instead.)

2. OMG! Too dim!

Modern camera strobes can be turned down in intensity. Filters can be added to video cameras. Exposures can be adjusted. The overall “look” of the video has nothing to do with whatever it is that is making the video happen, and everything to do with creative intent. Video cameras can take pictures of the friggin’ SUN — a wall of camera strobes is meh in comparison.

3. OMG! Too expensive! Must be fake!

Have you priced a music video lately? How about a set full of “ordinary” lighting and rigging? A few hundred SLRs with flashes sounds like a fairly cheap 2-day rental, to me — especially without needing to do any other set-work.

But whatever the case, money is a management problem — not a technical one — and it’s already spent.

4. OMG! It must be LEDs, because there were LEDs in the first few seconds!

Of course there were. Would you finance a project like this if you -couldn’t- visually predict what patterns the geeks had in mind? LEDs are an easy way to see things.

5. OMG! Why use camera bodies? Why not just use strobes? OMG, so complex my brain hurts from the inefficiency!

Because it looks cool. Maybe not to you, and maybe not to me. But to someone. If they just wanted to use strobes without the visual effect of having cameras, there’s a myriad of professional strobes which can produce anything from brief intense flashes to constant blinding white light (given human persistence of vision), which are very rentable in quantity for stagecraft.

The whole point of a visual exercise such as a music video is to produce the desired look. (Why is this a foreign concept?)

6. It’s fake because there’s a flash animation on their website that lets you “display” whatever words in that you want to using their “strobes”! OMG! Anyone can make this!

Yeah, there is a widget on their website. It shows one letter at a time. How many letters are there, exactly? How long do you think it took to photograph each of them and convert it into a customizable flash animation? I had cooler tricks than that on my CGA-equipped XT in 1987.

(Is this comment too negative for the new HaD? We’ll see!)

I’m going with real-ish.

I think the strobing animation really was done on the camera flashes. (Which is wicked awesome) based on this other video they have linked off the end of the making of: http://www.youtube.com/watch?v=9ycl8DJZUOw

But, again, based on that other video, I don’t think the band performed in front of it. Thinking it was quite possibly recorded at a different frame rate (addressing earlier noted concerns about recharge rate) and then composited with the band performing.

As for the concerns about “why have all the cameras if you just need the flash?” my guess: product placement. It’s been showing up more and more in music videos over the last 5 years and I wouldn’t be surprised at all if Canon payed a placement fee and supplied the equipment. Especially with that camera vanity shot right near the beginning.

I doubt this is all done real-time, with the stock equipment the are showing to be using. I find the flash duration suspect, the flash-on time seems to be too long on some moments. I’m not a Canon user, but I have a Nikon SB-800 speedlight which is I guess is at least somewhat comparable with the EX580. At full power the flash lasts only 1/1000th of a second with this unit. At 1/128th power, the flash duration is reduced to a mere 1/40,000th! It also supports a sort of fast burst mode for 2-3 seconds in which is looks almost like a flash light. However, the intensity during that burst fluctuates quite a bit. I have honestly no idea how this would register on a video camera but around 3:55 in the clip when they show the slow-motion close-ups of the bandmembers, the ‘flashes’ seem te be ‘on’ for quite a while. For instance the moment when the singer closes his mouth again. No way that those flashes on the background lasts only 1/1000th of a second or produce such solid light when ‘bursting’.

I vote fake. The animations are definitely real but as a few have mentioned the flash would be far too bright for the camera to see anything and the animations run too smoothly as the ‘flashes’ seem to be on longer than not. I think every lamp has been replaced with an LED. Looks good though!

I assumed they were using flashes with modeling lights (lower-intensity continuous lights you can use to predict where your shadows will be). I think using the actual flash bits would be pretty close to impossible in the video.

Real. (or at least for most of the video)

It appears they just made the flashes trigger on their lowest setting (prob 1/64). This would keep the flash from washing out the video and allow the flash unit to recharge in a very short period.

I wouldn’t be surprised if Duracell tried to use that as a battery commercial. :)