Need an LED flash made

Thread Starter

Lumens Matter

Joined Oct 15, 2011
3
Quick ramble first, this does relate to electronics, I swear:

Slightly odd question, but I'm hoping someone can help me- I'm a uk based DOP shooting a short in early December; it's a night scene, WW2 era and features a very specific kind of submachine gun in use by the russian military of the time. The location we're using is adjacent to farmland, and while they've ok'd some limited pyro use they've expressly banned us from using blank firing weaponry, and we're having extreme difficulty sourcing a flashpaper (ie, mostly noiseless) version of the weapon we need.

My question is this; would anyone be able to produce a simple circuit that when triggered lit a series of LED's for anywhere between a 24th and a 1/2th of a second, roughly simulating the light from a muzzleflash? Ideally this would repeat- as in, hold down a push to make and the LED continues flashing until you stop pressing the button.

The armourer tells me the rate of fire of the real weapon is approx 15 rounds per second, so if it's possible to make an LED flash 15 times in a second, that would be ideal, but it's not the be-all-end-all; I'm just insistent that we utilise a weapon with practical light like a real muzzleflash as opposed to an all digital muzzle flash for the shots using this weapon, as although I think this can be done quite convincingly in daylight it often looks horrendeous in a night scene (failing to cast any shadows etc).

If anyone is aware of a way to do this, and willing to build it, I'd be willing to pay for the materials, build time and cost of shipping upfront via a method of your choosing (ideally something like PayPal). And if you fancy a credit on the finished film, you can name yourself something like Senior Electronics Adviser or whatever takes your fancy, so there's that!

Cheers in advance for any help, and apologies if this sort of request is woefully against forum etiquette in any way!
 

Georacer

Joined Nov 25, 2009
5,182
I 'm not the ideal man in the forum to tell you how to rig an LED circuit, but I have some remarks:

Are you using a real weapon? If it is so, then where will the LED circuit fit? Sure, the LED can go in the barrel, but the rest of the circuit (batteries, board) needs to fit somewhere. Obviously they can't go in the gun. Or can they? Maybe inside the clip (if one exists) and be wired to the LED. If you don't care about a couple of wires going out of the gun, the circuit could even fit in the shooter's sleeve and be wired to the LED out of sight.

The muzzle flash is not only light, but a short explosion, with fire purging out of the barrel tip. An LED can't simulate that. Are you content with that?
It also has a multitude of colours, ranging from red to white. A single LED can't replicate that.
 

#12

Joined Nov 30, 2010
18,224
I can do LEDs and timers but...a muzzle flash expands. Leds are point sources. Muzzle flashes blur together at 15 RPS. For that effect, the LED would have to stay partly on at all times and have brighter moments, 15 times per second. That can be done but, they are still a point source.

This needs more thought.
 

SgtWookie

Joined Jul 17, 2007
22,230
Are you speaking of the PPSH-41?
It will be difficult to create an LED lighting effect that will be convincing during night firing.
Example evening video:
http://www.youtube.com/watch?v=Sg2wShSyQ-M&feature=related
Note the semi-random flame pattern.

A PPS-43 would be similarly difficult:
http://www.youtube.com/watch?v=-XLYjCTmEHs&feature=related

I find it interesting that the PPSH-41 has a compensator to help control muzzle climb, while on the PPS-43 they simply diverted the blast to either side.

A PPSH-41 being fired for several seconds in Iraq:
http://www.youtube.com/watch?v=Lq63GlFiCKE&feature=related

The PPD 40 was also used, but I'm having a bit of a difficult time locating a video for a real one being fired.

There are a number of Russian & Soviet firearms demonstrated in this video:
http://www.youtube.com/watch?v=eplqV6LbN7s
The slow motion segments may be helpful to study for obtaining realistic action from non-firing guns.

I have no idea if blank firing adapters were even made for those WWII-era submachine guns. The actions would not cycle without backpressure. A blank firing adapter would definitely change the flame pattern at the muzzle, so you would still not obtain a believable effect.

Even if you do have some kind of a muzzle flash, you will need the action to cycle, with spent cartridge cases being flung into the air, and at least some semblance of realistic, synchronized recoil.

The '41 and '43 both had a cyclic rate of fire of around 900 rounds/minute, which is 15 rounds/second as our OP mentioned. Just a bit faster than the M16 and M16A1, which is 750 RPM (12.5/sec). Yes, they are fun to shoot on cyclic, but I don't like to waste good ammo like that...
 
Last edited:

iONic

Joined Nov 16, 2007
1,662
Perhaps a YouTube Video of the real deal at night might show us the complexity that we'd be up against.

And is there smoke involved? This would really make things tricky!

Perhaps a RGB LED triggered close together but separately and out of sink??

15 rounds/sec. MAN! Keep me outa there! That's just too much metal to be caring around in the middle of a fight!

NOTE: Ya beat me to the Video think by 1/15th of a sec Sgt!


Stills from Sgt.'s Findings...

PPSH41 - Dusk



PPSH41 - Day(alternate angle)


PPSH43 - Dusk


My Conclusion!
Creating a repetitive flash for the holes in the side of the gun barrel is doable, but as for the majority of the light effects -
LED's alone have not and will not evolve to this level of complexity. Flash paper, something that creates some level of smoke and ignition force is required. You might want to give a call in to DreamWorks (Steven Spielberg's special effects outfit) for advice on this one.
 
Last edited:

SgtWookie

Joined Jul 17, 2007
22,230
I'm thinking that a butane or propane system with an electronic ignition could be quite convincing as far as the visual effect goes; but I don't know if your host objects to the use of cartridges with powder in them, the noise factor, or the fire hazard that live firing could present.

A butane or propane system with an igniter would be much more quiet than firing live blanks, would involve no primers nor powder, but the fire hazard would remain.

I just remembered a fellow, "Big Clive" who's over there in the UK; seems to me he does theatrical effects (pyrotechnic type things); has a website I've found interesting:
http://bigclive.com/
I'm certainly not trying to put you off; it's just that this fellow's been doing theater pyrotechnics and is even in the same country. He might be a good resource for you.
 

Thread Starter

Lumens Matter

Joined Oct 15, 2011
3
I 'm not the ideal man in the forum to tell you how to rig an LED circuit, but I have some remarks:

Are you using a real weapon? If it is so, then where will the LED circuit fit? Sure, the LED can go in the barrel, but the rest of the circuit (batteries, board) needs to fit somewhere. Obviously they can't go in the gun. Or can they? Maybe inside the clip (if one exists) and be wired to the LED. If you don't care about a couple of wires going out of the gun, the circuit could even fit in the shooter's sleeve and be wired to the LED out of sight.

The muzzle flash is not only light, but a short explosion, with fire purging out of the barrel tip. An LED can't simulate that. Are you content with that?
It also has a multitude of colours, ranging from red to white. A single LED can't replicate that.
Hiding the led circuit shouldn't be too much of an issue, the weapon is quite large and wires/pcb should be able to be hidden on the side away from camera. I'm fairly confident this will be a non-issue (or at least a relatively easily solveable one). Hiding it on the actor is a possiblity, but one I'd like to avoid (due to the likelihood of accidentally severing the connection by dropping the weapon etc).

The LED's don't need to simulate anything other than the light from the flash. Shell casings, smoke, sparks and the flash itself can be handled very convincingly digitally- actual light, sadly, cannot. Colour temperature is also a non issue- if it's closer to 3000k-ish tungsten, great, if it's closer to daylight, it'll be colour corrected. I've seen practical muzzle flashes that vary from orange to almost blue depending on the WB, cartridges etc, so it's not like there isn't precedent for this too, thankfully!

#12 said:
I can do LEDs and timers but...a muzzle flash expands. Leds are point sources. Muzzle flashes blur together at 15 RPS. For that effect, the LED would have to stay partly on at all times and have brighter moments, 15 times per second. That can be done but, they are still a point source.

This needs more thought.
The difference between a point source and diffuse light are fairly negligible here; it's important simply that there is actual light. While anyone frame by framing final footage will eventually be able to work out that the quality of the light isn't right for a ball of expanding gas, well... I think I can live with that.

Regarding muzzleflashes blurring together at 15 rps, as far as film is concerned (depending on your FPS and shutter speed/angle) then no, they don't. As long as there is a definite on/off state of the LED, then the camera will be able to pick it up to some degree. I can't guarantee 15 rps will be the best fit for our shooting conditions, but well, that's the nature of experimentation and what pre-production is for.

#12 said:
Are you speaking of the PPSH-41?
It will be difficult to create an LED lighting effect that will be convincing during night firing.
Example evening video:
http://www.youtube.com/watch?v=Sg2wS...eature=related
Note the semi-random flame pattern.

A PPS-43 would be similarly difficult:
http://www.youtube.com/watch?v=-XLYj...eature=related

I find it interesting that the PPSH-41 has a compensator to help control muzzle climb, while on the PPS-43 they simply diverted the blast to either side.
Yes, that's the weapon in question! Very interesting piece of kit. The random flame pattern isn't important though; just the light it produces. It appears from stock footage we've viewed that the PPSH almost always produces an inverted T of some descripton.

To clarify, and I apologise for not making this clear earlier, I'm not looking to replicate the muzzleflash itself; this will be handled by an effects artist in post production. I am just looking to replicate the light produced by a muzzleflash. And not even that closely at that; it's not enormously important that the light is soft vs hard, for instance. Smoke, sparks, cycling of the action, this is all very easy to handle digitally. Creating light where there is none, on the other hand, is woeful at best.
 

iONic

Joined Nov 16, 2007
1,662
Could you even pulse a propane device that fast?

Lumen's Matter,

A warm white LED is right about 3100K so choosing a LED for the optimal color effect will be no problem. Producing the light seen in the holes in the side of the barrel is also not a problem. But producing the light that we see at the top of the barrel as well as out of the barrel is where the technical difficulty comes in. We don't know exactly how much light you need to produce with the light....how many lumens??? The brighter you get, the more cooling is required, something I think we don't have much space for. If you place a couple of LED's at the end of a stiff wire, one straight out of the barrel and another over the barrel the light sources could be produced, but if you wave the gun around or drop the gun, then the light sources could easily be messed up. Can the special effects guys take out the wires that might show??

One other note, the frequency of the flashes can be made variable so that adjustment is possible. The video's that Sgt. referenced seem to be far short of 15 flashes per second, probably closer to 8 - 10.
 
Last edited:

Thread Starter

Lumens Matter

Joined Oct 15, 2011
3
Quick edit: for those of you who may be slightly unconvinced that I know what I'm talking about, here is a quick link in regards to muzzleflash/smoke/shells. The only "real" part of the pistols firing in the clip is the recoiling of the slide; the flash, smoke, sparks, shells and enviromental lighting are handled in post.

Imagine that I'm trying to do the same thing in a scene that looks more like this slightly ridiculous stock image;


So when the gun is fired, some light is cast onto the areas previously bathed in shadow or silhouetted (where digital post brightening would be impossible or at best strongly unadvisable).
 

Wendy

Joined Mar 24, 2008
23,415
Realism aside, the circuit you want is pretty easy, though fitting it in the space you want may not be. I can point you to the parts you need, you define the space exactly that you have, and if it is possible I'll fab something up.

What you need is a high power LED, something similar to this...

http://www.ledwv.com/en/led-1w-p-84.html

I notice they have even brighter LED out there. There are many sources for LEDs out there, a simple Google search pulled this first on the list. Some sources may be local to you, I usually encourage people to buy locally when practical.

Here are the problems.

LEDs this bright MUST be mounted to a heat sink. It can be a simple piece of aluminum, but there needs to be a chunk of metal to soak and dissipate the heat generated by the LED when it is on.

You need a battery that can handle high current. The ones I've messed with tend to not be rechargeable. Something like 2 CR123 batteries for example. They will run down quickly, but will look good while they work.

If there is room enough there is a device called a puck puck that will extend the life of the batteries significantly. They tend to be around 2" circles 1" thick, but there are several variations out there, and many are larger (I could be wrong about the size, I think I have one in my parts box about that size).

The actual electronics to drive this can be made quite tiny. This is why I was asking exactly how much space you would have. The batteries and the LED itself are the large chunks.

The guys from Mythbusters are stunt men. They have the fabrication equipment and expertise to do something like this routinely. I can do the electronics easy, most folks on this site can, but fabrication of specialty metal parts is a problem in itself.

If you look at my 555 tutorials I can show you how to do this yourself without too much hassle, I'm talking breadboarding the basic circuit. This is probably the best way, you really should breadboard this circuit before building the permanent one to check to see if it does what you want it too. It is a lot easier to make changes before you use solder than after.

My advice, buy the materials you need, and allow me and the other guys to show you how to build the prototype. If the light effect does what you want then we can work on miniaturizing it for you.
 
Last edited:

iONic

Joined Nov 16, 2007
1,662
No Light


Light


Surprisingly enough you almost do not even see the single flash. I had to get lucky and pause the video at the right moment.

By the way, if the new light source (the gun flash) were to light up the actors face, we would not see the light source produced from the right side of the actors face, but rather the face would be lit from the front. If the flash from the gun was not bright enough to override the light from the right, then the light on the left side of the actors face would not be so blown out either. Thus, the post production effects crew probably just brightened the face in a general sense and did not compensate accurately for the angle of the gun flash light source.

A Very Crude Photoshop'd Reproduction...



It's all moot though because when the film is playing it all happens so fast we don't recognize the inconsistency of the lighting.

I hope you'll forgo any copyright infringement!!
 
Last edited:

SgtWookie

Joined Jul 17, 2007
22,230
See the attached for a very simple, basic version.

It can run from a 6v supply, which could be four "AA" alkaline batteries in a holder. The 555 timer is a little 8-pin integrated circuit. Billions of these things have been made.

A single normally-open pushbutton causes the circuit to start flashing, about 21mS (~1/47th of a second) after the button has been pushed.

C1, R1, and R2 set the frequency and duty cycle. Right now, it's set to right at 15Hz and roughly 66% ON duty cycle. Just a single LED is shown, but you will probably need an LED on either side of the SMG shining downwards. A logic level MOSFET serves as a fast switch to turn the LED on and off.
 

Attachments

#12

Joined Nov 30, 2010
18,224
Suggestions:

NiMh AA batteries have .05 ohms of internal resistance. I've had good results using them for high current pulses. AA batteries are .56 inches diameter and 1.93" long.

If the barrel of the rifle is allowed to be occluded, the LED can be mounted on a slug of metal that is inserted in the barrel with an interference fit or a threaded section so the barrel is the heat sink.

I keep imagining the wires being inside the rifle barrel and the batteries in the clip.

The timing and duration of the pulses is a trivial feat of engineering but, it would be most helpful if the people that can do "frame by frame" could measure the duration of "light on" and "light off" and report that to us. 15 CPS amounts to 66.666 milliseconds. What part of that should be lighted?

The trigger of the rifle can be used to depress a "push to close" switch. The LED will obey the actor's finger.

edit: Wookie was delivering his last post while I was typing.
 

SgtWookie

Joined Jul 17, 2007
22,230
I know the OP asked for LED's, but I think a xenon strobe would be a better choice.
I beg to differ with you; a xenon strobe's flash would be too short and much too bright. There is also the fragility of the tube itself, and the HV required to trigger the tube; it would be a larger and more complex circuit to build requiring a boost coil and a trigger coil.

An astable 555 timer circuit would be quite simple, even though it lacks the somewhat random property of the actual ammunition firelight. The randomness is caused by inconsistencies in powder throw, primers, bullet seating depth, case size, and a few other factors. It appears that the powder used in the cartridges was just a bit on the slow side, as it's still burning when the round (bullet) exits the barrel.

I don't know if this is a result of the ammo used in the demos being old, therefore lower pressure than when originally manufactured (and therefore the burn was not complete by the time the round exited the barrel) or if this is the way the ammo functioned even when new. It would either take filmed accounts from WWII or accurate observations from someone who was involved with Soviet arms during that time.

The PPSH-41 fired the 7.62mm Tokarev round, which are a pistol round. It's really kind of odd that the PPSH-41 would have a muzzle flash from a pistol round, as pistol rounds are normally loaded with a quite fast-burning powder. If it's still burning when it leaves the muzzle, it's basically wasted energy - and the firearm will be harder to clean.
 

Wendy

Joined Mar 24, 2008
23,415
I wouldn't worry too much about accuracy at this point. Let the OP experiment and see where the pit falls are for himself.

Xenon tubes are on the way out. Looking at that LED link I saw 20W LEDs for a reasonable price. They would take around 7A I think (don't know for sure). They would be pretty bright LEDs. Then there is the size difference.
 
Top