Home of Professional Cinematography since 1996

style="margin-bottom: 0"> 

class="style5" Shooting High Speed

>Published : 3rd July 2005

>Does any one has experience, shooting with frame rates more than 2000 fps?

>MY question is, what kind of light fixtures are best used to avoid frames from being underexposed and thus having clean pictures w/out the "pump" effect when slowed down to 25fps?

>Thanx in advance for any advice

>Dan Schellhase

>CLT Berlin, Germany


>I shot some stuff last summer in the 2000-4000 fps range and the SUN worked great for me!

>Roderick E. Stevens II
Director -o- Photography
www.restevens.com
12on/12off


class="Paragraph">>MY question is, what kind of light fixtures are best used to avoid frames >from being underexposed and thus having clean pictures w/out the >"pump" effect when slowed down to 25fps?

>When shooting high speed over 2000fps lights, I have found that the smaller filament tungsten lamps, like tweenies, actually "see" the sine wave of an AC power source. This is because they attack and decay faster than larger filament sources such as a 5K or 10K. Think of when you turn off a tweenie and how quickly the lamp extinguishes compared to the time it takes a 10K to extinguish. This "pump" effect you are seeing is probably the fast response time of the smaller filaments to the sine wave. You can't see this effect with the naked eye, of course, only when the dailies come back at 24 or 25fps.

>To avoid this situation you can simply avoid using the smaller filament lamps or you can use a DC power supply, which does not have a sine wave. I have found any lamp over 2K to be safe at high speeds on AC.

>In older times, people would cluster the smaller filament lamps together and then stagger them onto different legs of the AC power supply so that when one lamp was dimming another was rising with the sine wave. This smoothed out the effect of seeing the sine wave.

>Another more modern day possibility, of course, is to use HMI's with flicker-free ballasts.

>Hope this helps.

>Jim Sofranko
NY/DP


class="Paragraph">>Does any one has experience, shooting with frame rates more than >2000 fps? MY question is, what kind of light fixtures are best used...

>A bunch of big ass ones. Incandescent lamps are far less prone to flicker than any discharge type (HMI), and so are preferred. If you can find a few Brutes those would work too.

>Sincerely,

>Ed Colman, President ­ SuperDailies, Inc.
Cinematographer Supervised Video Dailies
http://www.superdailies.com


class="Paragraph">>This "pump" effect you are seeing is probably the fast response time of >the smaller filaments to the sine wave. You can't see this effect with the >naked eye, of course, only when the dailies come back at 24 or 25fps.

>If he shot at 2040 fps (evenly divisible by 60hz and 120) would you still see the pumping? Assuming, of course, that a high speed camera is going to run accurately at that frame rate. (I have almost no experience working faster than 120fps.)

>Art Adams, DP [film|hidef|video]
Mountain View, California
San Francisco Bay Area - "Silicon Valley"
http://www.artadams.net/
Local resources : http://www.artadams.net/local


>Dan Schellhase writes:

class="Paragraph">> Does any one has experience, shooting with frame rates more than >2000 fps?

>Lots of people do, and there's info in the archives as well as on the web.

class="Paragraph">> MY question is, what kind of light fixtures are best used to avoid frames >from being underexposed and thus having clean pictures w/ out the >"pump" effect when slowed down to 25fps?

>I'm not quite sure what you mean by underexposed frames. If you mean your shutter is out of sync with some form of discharge light -- HMI, Fluorescent, Xenon, etc. then you will get "pumping" or pulsing, but the exposure variation must be considerable to be a problem when viewed at 80x.

>You should use a continuous light source or a synchronized strobe, e.g.: Unilux.

>There are several options for continuous sources. First of all, at that frame rate you will need lots of light -- 5K, 10K, and 20K tungsten units do not show any AC flicker. Smaller tungsten filament lights can 'flicker' on AC current, but can be run on DC to eliminate the problem, or you can spread the load on a 3 phase system. You don't specify how large an area you are trying to light -- carbon arcs are still around.

>Brian Heller
IA 600 DP


>My Photosonics catalogue(hey, I need the new 2004 version come to think of it)lists available soft box lights: 24K, 34 K and 58(!)K (and it weighs 145 pounds...)

John F. Babl
DP
Miami


>Brian H. wrote :

class="Paragraph">>I'm not quite sure what you mean by underexposed frames. If you mean >your shutter is out of sync with some form of discharge light --

>When I shot at 2000-4000 fps (ramping even) I was using my own 4k HMI (just happened to be the brightest light I had around) and the pulsing was QUITE noticeable.

>What I found oddest of all in that experiment was the fact that I had to do a tie-in just to power the camera!

>Roderick E. Stevens II
Director -o- Photography


>Art Adams writes:

class="Paragraph">> If he shot at 2040 fps (evenly divisible by 60hz and 120) would you still >see the pumping?

>Unless the camera shutter and the HMI are perfectly synchronized, i.e., the shutter fully open as the light discharges, you will very likely get pumping. There is always an element of luck, but the odds are not bin your favor. If an HMI is the sole source of illumination, and is not synchronized to the shutter, you are as likely to get completely unexposed frames as you are likely to get fully exposed frames. (180 degree shutter)

class="Paragraph">>Assuming, of course, that a high speed camera is going to run >accurately at that frame rate.

>Nearly all of this work is now done on video, for obvious reasons.

>Brian Heller
IA 600 DP


class="Paragraph">>If he shot at 2040 fps (evenly divisible by 60hz and 120) would you still >see the pumping? Assuming, of course, that a high speed camera is >going to run accurately at that frame rate.

>Theoretically yes. But I don't believe any of the Photosonic cameras or other high speed film cameras have any precise crystal control at those fps. I am curious about the newer video high speed cameras though. How do they capture at such high frame rates?

>Brian Heller wrote :

class="Paragraph">>then you will get "pumping" or pulsing, but the exposure variation must >be considerable to be a problem when viewed at 80x.

>I would think just the opposite. The flicker problem is photographed at 80x and viewed at normal speed. That would make any slight variation much more noticeable because it has more screen time.

>Jim Sofranko
NY/DP


>Roderick Stevens writes:

class="Paragraph">> When I shot at 2000-4000 fps (ramping even) I was using my own 4k >HMI (just happened to be the brightest light I had around) and the >pulsing was QUITE noticeable.

>Since HMIs are discharge lights, not continuous like tungsten, and there is generally no means to synchronize the camera shutter with the discharge of the HMI, and since most, if not all, commercially available HS cameras do not have crystal controlled motor speeds, pulsing is almost inevitable.

class="Paragraph">> What I found oddest of all in that experiment was the fact that I had to >do a tie-in just to power the camera!

>What camera did you use?

>The amount of force required to accelerate even a roll 16mm film to 2000 fps is considerable, but newer (relatively) cameras such as the Redlake HyCam will do 11,000 fps on a 15 amp 110 volt circuit.

>Jim Sofranko writes:

class="Paragraph">> I am curious about the newer video high speed cameras though. How >do they capture at such high frame rates?

>High speed video cameras have made tremendous advances in the last few years, with advances being made constantly. However, they are a ways from achieving the information contained in a pin registered 35mm frame such as a Photosonics 4E, or what used to be considered as broadcast quality.

High speed video cameras usually use CMOS technology with a direct trade off between frame rate and resolution. As the frame rate goes up, the resolution goes down. It's the same problem of unloading the chip rapidly that vexes all video manufacturers.

Olympus makes some very user friendly self contained systems with resolutions around 800x600 pixels at 1000 fps, and reaching 30,000 fps at greatly reduced resolution. I think it's called the 'I Speed'.

>AFAIK, megapixel imaging is still limited to well under 100fps.

>Kodak was a pioneer in this field, but I'm not sure if they're still in it, and least not commercially.

>Other high speed video systems require a good deal of training and/or practice, and many of the manufacturers offer training seminars.

>Brian Heller


>HMI's must be "Electronic Ballast" (square wave) for high-speed cinematography - the older magnetic (Roderick?) are no good, they flicker.

>Small filament bulbs are as voiced, no good-they don't flicker, they look like they are going slowly on and OFF at 2,500 fps. 2,500 fps is the 35mm "Top End", except for some military models which do 3,250 fps (only!), no transmission. If you prefer a higher framerate, you need a 16mm rotary prism cam.

>Various manufacturers and models do up to 500 pin-registered, and 10,000 full-frame fps (Std 16 - they need both perfs-2R). You can do 40,000 fps if you can use a narrow (1/4 height) neg. In reality, I've never gotten the 16 mil's over 8,000 frames per second. Not enough light.

>We were in the sun in South Florida, using BIG HMI's for keys, the sun for fill. The background was almost two stops under. The grass was wilting, the golf ball getting soft. The shots looked great - even got the compression (distortion) of the ball when the club hit it. Sun, tungsten (large filament), Square-wave HMI's and Zeno are all OK. Unilux only does 500fps, and are hard-pressed to give an F8 with a reflexed high-speed camera.

>Bill Gula
CINEX, USA


>Art Adams writes:

class="Paragraph">>If he shot at 2040 fps (evenly divisible by 60hz and 120) would you still >see the pumping? Assuming, of course, that a high speed camera is >going to run accurately at that frame rate.

>With the camera running that fast, you'd still see the pumping. At 2040 fps with 60Hz (120HZ) illumination you'd get *exactly* 17 pulses per second, which is just as bad as *approximately* 17 pulses per second.

>Dan Drasin
Producer/DP
Marin County, CA


>I forgot to mention that - yes all three HMI's (4k & 2 x 1.2s) were magnetic ballast. I knew there was going to be pulsing, but I was testing out the camera and wanted as much light as possible. I filmed my son drumming in my artificially created rain. Fun stuff.

>Anyway - I am trying desperately to remember what the camera was. It was a Photosonics . . . IVF? Does that make sense? It was a 16mm, old military (U.S. Navy) camera that looked like a piece of farming equipment. In fact, I just realized - if you go to my website - www.restevens.com - enter the site (by clicking on my head – click on recent projects and then scroll down and click on "cinemexoticatography" - you'll arrive at a never finished page – but there is a picture of me sitting next to the camera. Sorry for the web-maze - but it's a frames website and I don't have anyway of sending a link to just that page.

>Roderick E. Stevens II
Director -o- Photography


>Roderick Stevens writes:

class="Paragraph">> Anyway - I am trying desperately to remember what the camera was. It >was a Photosonics . . . IVF? Does that make sense?

>No, it doesn't. Photosonics usually uses the prefix 1 for 16mm, as in the Actionmaster 1-PL, and the prefix 4 for 35mm cameras. I don't recognize the camera but I don't think it's a Photosonics. A lot of different companies made 16mm instrumentation cameras for the Gov’t.

>It could still be made by Photosonics. They made a lot of different cameras, but as the descendents of the venerable Acme animation cameras, their camera designs are very elegant.

>I don't know the 4-F, but the 4-E is the pin registered 35 mm Photosonics most used in commercial work, and it is considerably larger than what's in your picture. And weighs more than 125 lbs w/o film. Top speed is 360 fps.

>AFAIK, that is the top speed for any pin registered 35mm camera. The 4-B and 4-C which look a lot like the 4-E are rotary prism type cameras with a top frame rate of 3250 fps.

>Even with electronic ballast, or flicker free ballasts, if the HMIs were providing most of the light for your scene, there most likely would have been visible light pulsing. For best results, a continuous light source is still the best bet.

>There are several manufacturers that make "framing cameras" (rotating drum or rotating mirror) that use 35mm film and can go to amazing frame rates (millions of fps), but only for very short lengths of film.

>Brian Heller
IA 600 DP


>Brian Heller wrote :

class="Paragraph">>It could still be made by Photosonics. They made a lot of different >cameras,

>Well, I'm fairly positive it had Photosonics stamped on it. I can try to find out - it belongs to a friend of mine. In fact he's thinking of selling it.

>It is a 30 amp camera that runs from about 2000 to 4000 fps (nothing slower darn it). Of course a 400' roll -o- 16mm lasted about 17 seconds at those speeds. It's controlled by a 'voltage' rheostat, uses a rotary prism, and C-mount lenses. The real pain in the butt is I had to set up the shot without the mag on the camera, lock it down as best as possible, then remove the viewfinder, install the mag and thread the film.

>Of course, being me - I had to incorporate it into a 'dolly shot'

class="Paragraph">>Even with electronic ballast, or flicker free ballasts, if the HMI’s were >providing most of the light for your scene, there most likely would have >been visible light pulsing.

>Yup! The sun worked well!

>I may shoot some more stuff with it soon and if I need to light, I've got an old 5k fresnel (big ol' studio beast) and a 9 lite fey I can use this time.

>Roderick E. Stevens II
Director -o- Photography


class="Paragraph">>Anyway - I am trying desperately to remember what the camera was. It >was a Photosonics . . . IVF?

It looks a bit like a cross between the Photosonics Actionmaster 1PL (500fps pin registered)


http://www.photosonics.com/images/1pl.jpgor and a 1B


(1000fps rotating prism)


http://www.photosonics.com/images/1b_fdrs.gif.

Here's a 35mm 4C just for comparison


http://www.photosonics.com/images/4c_1.jpg.

>I recall seeing a similar camera in the crash test dummies facility at GM in Detroit years ago. Don't remember the model though.

>As Brian mentioned many of them were initially designed for use by the military. I believe there are many floating around that have subsequently been made their way into the film biz probably through gov't. auction. Maybe they have higher frame rates.

>When I see some of these gov’t. surplus models I always wish I had a Geiger with me. Duck and cover!

>PS - Nice site Roderick!

>Jim Sofranko
NY/DP


>I will see if My friend can recall what it is or if he can easily enough look it up. I know it was kinda buried and he is an octogenarian so it may be a bit cumbersome for him to get to.

>He wants to sell it though, and I can't help but be somewhat interested.

>Hmmmm . . we'll see.

>Roderick E. Stevens II
Az. - Director -o- Photography


>Jim Sofranko writes:

class="Paragraph">> When I see some of these gov’t. surplus models I always wish I had a >Geiger with me. Duck and cover!

>Have you been getting film back with an unexplained high fog level.

>Brian Heller
IA 600 DP


>Jim S. wrote :

class="Paragraph">>When I see some of these gov’t. surplus models I always wish I had a >Geiger with me. Duck and cover!

>Eeesh! I never think of these things!

class="Paragraph">>PS-Nice site Roderick!

>Thanks Jim!

>Roderick E. Stevens II
Az. - Director -o- Photography