Slow motion question

Jim Andrada · May 4th, 2007, 12:12 AM

When I worked for the US Naval Weapons Lab in the early 60's we were indeed filming explosions in slow motion We were able to actually get movies showing how a single grain of propellant (think super sophisticated large extruded pieces of gunpowder) ignited and burned so we could study different formulations and extrusion shapes and sizes etc. As I recall it, we were studying propellants for the old 16" rifles used on battleships. Roughly a 2000 pound projectile and 500 to 700 pounds of propellant per shot. Very dramatic! I think the range was well over 20 miles - maybe even close to 30 with an overload of powder. We were in some cases just trying to see how much of a powder charge it took to blow up the gun. VERY dramatic

The camera shot something on the order of the equivalent of 6 million frames per second. It had a high speed rotating prism/mirror in the center and a longish strip of stationary 35mm film arranged around the periphery. Details are lost in the dim mists of history and fading memory, but I remember the frame rate we needed to do the job - or at least I THINK I remember it.

Louis Mostert · May 4th, 2007, 11:29 AM

Quote:

Originally Posted by Stuart Brontman

A client has just asked me to produce a slow motion recording of a controlled environment explosion. The explosion would be viewed at normal frame rates, then slowed down to show the instant-by-instant development of the damage being caused as the explosion continues. The client wants this shot in HD.

My question is how to do this - and if today's crop of HDV cameras is capable of this in terms of exposure (can the lens and exposure system handle the bright flash without blowing everything out) and shutter speed. I would consider the XDCAM HD as part of the group of cameras - but only for rental. The budget on this job won't justify the purchase of the XDCAM.

Thanks.

Hi Stuart,

Maybe try using a combination of 60fps (hvx200/hd200) and a motion compensation plug-in like Algolith Algosuite 2.

Jim Andrada · May 4th, 2007, 01:14 PM

I think it isn't a question of shutter speed as much as a question of frame rate. Any camera capturing at 30 fps won't get much detail of an explosive event as the explosion itself is over in at most milliseconds. I really don't even think that a couple of hundred frames per second will give you much useful detail. I think you'd need to be in the range of 1000 fps to get much useful stuff and even then it would be minimal - the explosion would be over in a couple of frames.

So I think the real question would be the scale of the event and what it is that is really of importance. If the explosion itself is of interest, then I think without very sophisticated and specialized equipment like we used to have in an explosives research lab you'll get nothing useful.

On the other hand, if the damage is what you want to capture, and the scale of the event is large enough, then you might be in luck because the rate of propagation of the shockwave is orders of magnitude slower than the explosion itself, and the rate at which physical objects react to it is slower yet.

For example, if the explosion is supposed to knock down a 300 foot tall building, the collapse will take quite a few seconds and 30 or 60 frames a second would have some interesting content.

If the scale were more like a room being destroyed by a hand grenade, then I think you'd have to be on the order of a couple hundred frames a second to catch more than smoke, although stuff falling over and bouncing around the room could go on for a few seconds.

If the event were at the scale of a microwave oven, then I think you'd be out of luck with any camera that didn't need a truck to carry it, and probably looking at over 1000 fps which I think is beyond the capability of any normal camera. And probably beyond the rate at which most camera sensors themselves can respond. About the only thing that might work would be a film camera with an open shutter and rapidly moving film spooling continuously and a quick firing strobe acting as both light source and shutter.

Scale matters!!!!!

Disclaimer: I don't do any of this kind of filming, and everything I say is based on theory and opinion and (sometimes faulty) memory. Someone in the special effects business would be a better source for this kind of information.

Stuart Brontman · May 4th, 2007, 01:35 PM

Wow! Thanks to all of you that have responded. I think the real issue I'm dealing with is the aftermath of the explosion - the burning of the mannequin and the effects it produces. The client has indicated the explosion is important to capture, but not the main interest. Perhaps the biggest issue now is making sure the camera can handle the flash by using fully manual control AND that the sensors will recover quickly enough from the bright flash to record the damage and burning that is going on.

Previous footage shot of these events has an extended recovery by the sensors after the flash occurs. It makes the entire sequence look terrible. I'm trying to get around this with full manual control, controlled lighting, etc...

Stuart

May 4th, 2007, 12:12 AM	#16
Jim Andrada Inner Circle Join Date: Feb 2007 Location: Tucson AZ Posts: 2,211	When I worked for the US Naval Weapons Lab in the early 60's we were indeed filming explosions in slow motion We were able to actually get movies showing how a single grain of propellant (think super sophisticated large extruded pieces of gunpowder) ignited and burned so we could study different formulations and extrusion shapes and sizes etc. As I recall it, we were studying propellants for the old 16" rifles used on battleships. Roughly a 2000 pound projectile and 500 to 700 pounds of propellant per shot. Very dramatic! I think the range was well over 20 miles - maybe even close to 30 with an overload of powder. We were in some cases just trying to see how much of a powder charge it took to blow up the gun. VERY dramatic The camera shot something on the order of the equivalent of 6 million frames per second. It had a high speed rotating prism/mirror in the center and a longish strip of stationary 35mm film arranged around the periphery. Details are lost in the dim mists of history and fading memory, but I remember the frame rate we needed to do the job - or at least I THINK I remember it.

May 4th, 2007, 01:14 PM	#18
Jim Andrada Inner Circle Join Date: Feb 2007 Location: Tucson AZ Posts: 2,211	I think it isn't a question of shutter speed as much as a question of frame rate. Any camera capturing at 30 fps won't get much detail of an explosive event as the explosion itself is over in at most milliseconds. I really don't even think that a couple of hundred frames per second will give you much useful detail. I think you'd need to be in the range of 1000 fps to get much useful stuff and even then it would be minimal - the explosion would be over in a couple of frames. So I think the real question would be the scale of the event and what it is that is really of importance. If the explosion itself is of interest, then I think without very sophisticated and specialized equipment like we used to have in an explosives research lab you'll get nothing useful. On the other hand, if the damage is what you want to capture, and the scale of the event is large enough, then you might be in luck because the rate of propagation of the shockwave is orders of magnitude slower than the explosion itself, and the rate at which physical objects react to it is slower yet. For example, if the explosion is supposed to knock down a 300 foot tall building, the collapse will take quite a few seconds and 30 or 60 frames a second would have some interesting content. If the scale were more like a room being destroyed by a hand grenade, then I think you'd have to be on the order of a couple hundred frames a second to catch more than smoke, although stuff falling over and bouncing around the room could go on for a few seconds. If the event were at the scale of a microwave oven, then I think you'd be out of luck with any camera that didn't need a truck to carry it, and probably looking at over 1000 fps which I think is beyond the capability of any normal camera. And probably beyond the rate at which most camera sensors themselves can respond. About the only thing that might work would be a film camera with an open shutter and rapidly moving film spooling continuously and a quick firing strobe acting as both light source and shutter. Scale matters!!!!! Disclaimer: I don't do any of this kind of filming, and everything I say is based on theory and opinion and (sometimes faulty) memory. Someone in the special effects business would be a better source for this kind of information.

May 4th, 2007, 01:35 PM	#19
Stuart Brontman Contributor Join Date: Apr 2003 Location: Chicago, IL Posts: 228	Wow! Thanks to all of you that have responded. I think the real issue I'm dealing with is the aftermath of the explosion - the burning of the mannequin and the effects it produces. The client has indicated the explosion is important to capture, but not the main interest. Perhaps the biggest issue now is making sure the camera can handle the flash by using fully manual control AND that the sensors will recover quickly enough from the bright flash to record the damage and burning that is going on. Previous footage shot of these events has an extended recovery by the sensors after the flash occurs. It makes the entire sequence look terrible. I'm trying to get around this with full manual control, controlled lighting, etc... Stuart