The Nikon D3 shutter in super slow motion, by Marianne Oelund and Jeffrey Friedl. Let it load and then move your mouse from left to right.
_____________________
Mike (Thanks to Albano Garcia)
|-- removed generator -->
« 'Night Vision' by Troy Paiva | Main | Forgotten Camera: R. Kolewe »
You can follow this conversation by subscribing to the comment feed for this post.
The comments to this entry are closed.
Ahh, don't mind me......but this does seem to take the whole pixel peeping, technophile thing to an extreme.
Bron
Posted by: Bron Janulis | Tuesday, 09 September 2008 at 03:13 PM
That is so cool!
The accelerations and decelerations involved are absolutely insane.
Posted by: Peter | Tuesday, 09 September 2008 at 03:40 PM
OK, I guess I'm just stupid, but I never realize that DSLRs even had shutters. I always assumed that the sensor was activated only for the needed duration (1/125 of a second, or whatever). What am I missing? It seems like a computer could activate a sensor with greater accuracy than a complicated mechanical device could expose it to the light. Not that I actually know anything about electronics or anything. But, wait, why do digicams have to add sound effects to let you know that they've taken a picture? Isn't that because there's no shutter? I always thought the sound my D200 makes was just the mirror. So, here's my dumb question: Why do these cameras have shutters, and do digicams have shutters that just doesn't make any sound?
Posted by: Robin Dreyer | Tuesday, 09 September 2008 at 04:13 PM
In response to Robin Dreyer,
Both DSLRs and digicams have shutters. Every CCD or CMOS is charged and reads 1 sensor at a time in a sequential manner, not in parallel. Therefor it is not possible to read the values from the sensor with a switch on -> switch off of the sensor.
sensor is pre-charged to "ground level", then light hitting the sensors will "charge" them and then values from them will be read sequential to post-process and save. This read process can even be done after the shutter is closed.
Posted by: David Rodrigues | Tuesday, 09 September 2008 at 07:08 PM
In response to David Rodrigues
I don't think this is entirely true. I remember being in a studio with a (I think) Nikon D1. I was using the studio strobes.
I was very surprised to discover the *actual* flash sync speed was 1/4000, and not 1/500 as advertised.
Of course there is no way that it could possibly open the shutter all the way, trigger the strobe and close it again in 1/4000 second. Therefore, at speeds above the *shutters* flash sync speed it must have used the cmos ccd to shorten the exposure time.
I've never really gotten to the bottom of it, but I think that some types of ccd can act as an "electronic" shutter.
Alex
Posted by: Alex Le Heux | Wednesday, 10 September 2008 at 07:36 AM
Alex,
I think that's flash duration. Electronic strobes can have flash durations much shorter than the fastest shutter speeds of cameras.
Mike J.
Posted by: Mike J. | Wednesday, 10 September 2008 at 08:08 AM
Mike,
Curtain shutters, such as in SLRs, reach their fast shutter speeds by sliding both curtains across the film plane at the same time. There is then no moment that the entire frame is exposed, just a thin line that slides across it. The shorter the shutter time, the thinner that line.
This is why SLRs have a flash sync speed, usually 1/125 or /250 or so: That is the shortest exposure time where it will still expose the entire frame.
You can try this by setting the shutter speed to something shorter than the sync time and using a flash. You'll only get part of the frame exposed properly. (careful though, modern SLR flash system often have something called "high-speed sync", where the flash will actually flash multiple times in short succession. Turn that off, or use a flash that can't do that)
The studio strobes actually have a fairly long duration, compared to the electronic flashes you normally find. This became obvious as the image got darker as the shutter time got shorter.
Posted by: Alex Le Heux | Wednesday, 10 September 2008 at 08:24 AM
Alex,
With the caveat that I don't really know what the heck you're talking about here, there *is* a "moment that the entire frame is exposed"--it's the top sync speed. Flash durations differ, but must be shorter than the sync speed to get consistent results. Our Speedotron Black Line stuff (am I dating myself?) would go as slow as 1/300th at full power, which was perilously close to the top sync speed. Of course, slower sync speeds work too. Neither has anything to do with a digital sensor not needing a shutter, as far as I know.
Mike J.
Posted by: Mike J. | Wednesday, 10 September 2008 at 08:33 AM
Mike - I'll try to clarify what Alex is talking about :
The max sync speed for your camera is the fastest shutter speed for which the entire sensor is exposed at once. At speeds faster than this, the second shutter starts to drop before the first shutter has fully opened. So at any given time only part of the sensor is exposed.
If you trigger a flash at one of these higher speeds, you will only get a line across the image where the sensor was exposed, since flash durations are so short.
High speed sync allows you to use a faster shutter speed by pulsing the flash over a longer duration, so that all parts of the sensor "see" the flash.
It's a bit counter-intuitive that you need a longer flash duration to use a higher shutter speed, but it's true.
Posted by: David Bostedo | Wednesday, 10 September 2008 at 09:21 AM
David,
I understand all that.
Perhaps we should ask Alex to clarify his original point.
Mike J.
Posted by: Mike J. | Wednesday, 10 September 2008 at 09:41 AM
You know, I figured you must know that Mike. (Yet I still typed it out...sorry!) I wonder if Alex's original point was that he thought the CCD in the D1 did something special, when in reality the studio strobes probably just had a longer duration than his shutter speed?
But you mentioned studio strobes having a much shorter duration compared to regular strobes, which is what confused me about what you were getting at...
Posted by: David Bostedo | Wednesday, 10 September 2008 at 10:15 AM
That's incorrect. Some Nikons (D70, D50 and some others) use an hybrid shutter, where the mechanical shutter maximum speed is 1/90 and faster speeds are achieved by the sensor, and that why you get 1/4000 flash sync with any non dedicated flash unit.
You can read about it in Thom Hogan's site:
"All shutter speeds up to 1/90 are handled mechanically on the D50, all faster shutter speeds are handled electronically (e.g., the shutter opens for 1/90 and the sensor turns on and off to create the shutter speed effect). This arrangement first appeared on the D1 and provides one very useful side effect: flash sync speeds are increased substantially (arbitrarily limited to 1/500 on the D50, just like the D1 and D70 series)"
(http://www.bythom.com/D50REVIEW.htm)
Also in Strobist:
"Electronic shutters also have auxilliary mechanical shutters that actually open and close up to, say, 1/125th of a second. This helps to protect the CCD from dust and damage. Beyond that speed, the computer just grabs a progressively smaller slice of time from the CCD and "fakes" higher shutter speeds electronically. Which, as it happens, is totally golden for us. (...)
So, you can get high-speed flash with any flash. And you are not wasting flash energy on a series of pulses, either. To my mind, this is far more effective than FP flash."
(http://strobist.blogspot.com/2007/05/hacking-your-cameras-sync-speed-pt-1.html)
Posted by: Flaneur | Wednesday, 10 September 2008 at 12:13 PM
Flaneur has the real info, and explained it better than I could.
I just observed this in the studio once and spent an hour or so trying to decide if I was going mad before the model got bored and I continued shooting :)
Shortly after that I switched to leaf shuttered cameras, and they sync at any speed...
Posted by: Alex Le Heux | Wednesday, 10 September 2008 at 04:59 PM
Well, I apologize. I didn't understand what you meant. What Flaneur reports is news to me.
Mike J.
Posted by: Mike J. | Wednesday, 10 September 2008 at 05:02 PM
That fast shutter sync is the best least-known feature of the Nikon D40 and D50, IMHO. Too bad it comes with a price. These cameras show a curious artifact when shooting into the sun. Whenever I'd include the sun in a wide-angle outdoors shot, the photo had what appeared to be a jet contrail crossing the sun, though I'd seen no airplane there. Wait, was this evidence of the mythical "chemtrails" that cross our skies, marking a secret governmental science experiment intended at controlling the climate, or our minds?
No, it wasn't. After consulting various Nikon experts on the boards, I learned it was the mark of sunlight "bleeding" laterally along pixel rows during the long, 1/90 sec time the shutter was open. It happens every time, go see for yourself, if you have one.
Posted by: John McMillin | Wednesday, 10 September 2008 at 06:47 PM
Yes, its so unknown that even Nikon's sellers ignore it and look at me with a weird face when I tell them this.
I never experienced (or noticed, maybe its there, but I shot very few with sun in the frame) this, but I noticed a tendence to blooming highlights (in blue color), something Thom Hogan says is due to this electronic sensor. A small trade-off.
Albano (aka Flaneur)
Posted by: Flaneur | Wednesday, 10 September 2008 at 07:30 PM
Look at that thing shake. That is why mirror lock up is a nice feature to have.
Posted by: scott_h | Thursday, 11 September 2008 at 07:31 AM