Lovely Day Media wrote:
If you can get this to work, speed lights might be in trouble.

That's one of the hopes. I'd like to leave my slow recycling speedlites at home. The downside is this thing may be really expensive. An LED to generate 47W is close to $20.00. I paid close to $600 for my Canon 600EX. I'd probably have to charge at least half of that.

GER Photography wrote:
I guess it's all about the Benjamins!!:-))))

I was having a hard time finding something to shoot. I don't have many small, interesting things. If you think of something you want to see photographed with it, post it and I'll try to make it happen. Any sample is a good learning experience.

Random Image wrote:
I have been wondering why someone hasn't done this already.  Go the Buff route and design a flash with built in battery and radio popper, a true all in one unite, and I would be standing in line to give you my money.

Oh, it hasn't been done because it's a pain in the ass. The LED Light Cube guys had to delay manufacturing and delivery several times, and for several months. They were first due to ship in November, then it got pushed, and pushed, and pushed. So far, no delivery:
http://www.ledlightcube.com/
https://www.indiegogo.com/projects/the- … filmmakers

LED light tends to be very directional. If you have time, do this: Get an LED glowing, a bright LED, and look at it straight on (if you can). It will be bright as anything. Then, look at it perpendicular to its lens. The light will be much weaker. Now, do the same thing with an incandescent you'd put into a flashlight. It will be almost equal power from both head on, and perpendicular. The bulb comes close to being a true "point source". The LED, not so much.

I'm having a hard time figuring out if I'd want a power source built in. It is definitely an attractable offering, but adds weight to the strobe. Maybe offering a user option would be best. Use it or use and external supply.

I'm with you on the radio triggering. Something I've definitely looked into. It's not hard to do. I'd have to research applicable laws with regards to transmitting sync pulses.

AG_Boston wrote:
That's one of the hopes. I'd like to leave my slow recycling speedlites at home. The downside is this thing may be really expensive. An LED to generate 47W is close to $20.00. I paid close to $600 for my Canon 600EX. I'd probably have to charge at least half of that.

That sounds like a pretty good deal to me ... charging a person $300 for a $20 LED? If you needed 15 of them, it would eat the whole $300 but that would mean a person would be carrying a 705W "speed light". I don't know how that compares to a currently available speed light, but it seems pretty powerful to me.

If the equipment was smaller and/or lighter (no pun intended), used less battery power AND was half the price (plus or minus a few bucks), I don't see why they wouldn't sell.

Random Image wrote:
I have been wondering why someone hasn't done this already.  Go the Buff route and design a flash with built in battery and radio popper, a true all in one unite, and I would be standing in line to give you my money.

I have never seen such a thing from PCB however you should get in line for the Profoto B1 then, it's been out and it's impressive.

Lovely Day Media wrote:
That sounds like a pretty good deal to me ... charging a person $300 for a $20 LED? If you needed 15 of them, it would eat the whole $300 but that would mean a person would be carrying a 705W "speed light". I don't know how that compares to a currently available speed light, but it seems pretty powerful to me.

If the equipment was smaller and/or lighter (no pun intended), used less battery power AND was half the price (plus or minus a few bucks), I don't see why they wouldn't sell.

HA! That $20 is just the LED. The LED still needs drive electronics, some way to dissipate its heat (typically done with chunks and shapes of aluminum, think $$$), a trigger circuit to talk with the camera, housing for the electronics, someone to manufacture the parts, and someone to assemble the parts... $20.00 went away a loooong time ago. I'll get more into this the more I work on this. The plan is still it to try and shoot at least a human face with some variant of the system by the end of this month.

Random Image wrote:
Just me thinking aloud, but why not reverse the aiming of the led backwards into a reflector, like a tiny strobe and umbrella.  If done right should provide plenty of scattering.

That was one of my earlier attempts. It did follow the magnification equation for a concave mirror, and I ended up with baseball sized light spots. This was using a large metal kitchen mixing bowl, a pot lid, and a highly polished aluminum can bottom. They did a great job of focusing light, and didn't do anything useful for what I'm trying. They ended up doing just this:
http://www.meritnation.com/img/shared/d … 987347.png

AJScalzitti wrote:

I have never seen such a thing from PCB however you should get in line for the Profoto B1 then, it's been out and it's impressive.

I briefly considered buying something like that a year or so back. They're nice. The cost is so amazingly high though.

AG_Boston wrote:
I briefly considered buying something like that a year or so back. They're nice. The cost is so amazingly high though.

Hard to justify if you don't use/need an all in portable strobe often enough.

Mikey McMichaels wrote:
Please give it a built in meter that works based on distance, not reflectance.

I've wondered about stuff like this. I use my strobes and flashes in manual mode. I have tried making use of TTL. Most of the time TTL seems to want to do things I don't want it to do. Are automated metering functions something you make a lot of use of?

AJScalzitti wrote:

Hard to justify if you don't use/need an all in portable strobe often enough.

I tend to use my strobes rather often. In the end my thought was: "Lots of lower priced strobes can be purchased, and put into many more places.". So that's how I went about it. Having a kickass 500Ws strobes would still be amazing though...although not at the cost of $2k.

Good Egg Productions wrote:
I get what you're trying to do.  But what I don't get is the application.

I have strobes for photography and LED panel lights for video production.

What is so terrible about a Xenon tube that troubles you?  The recycle time?  Well, what is your Ws output of the LED?  I would bet you dollars to donuts that my (or any) speedlight recycles just fine at the lower Ws of your build.

So unless you can manufacture these things for under $100, I don't see the application.

Not when you can get this for under $80.
http://www.amazon.com/Yongnuo-Professio … ds=yongnuo



And if we're talking purely macro applications, Yongnuo has that covered too.  And for a ridiculously reasonable $40.

http://www.amazon.com/Yongnuo-MR-58-Pen … ring+flash

What I'm trying to do is come up with a bigger system. I'd like to be able to produce a smaller LED based device which is capable of competing with my gas tube 200Ws strobes. In the end, I'd like my 200Ws LED strobe to be about the size of a roll or toilet paper (that opens me up for all kinds of ridicule), instead of the size of an adult sized shoe box. Not only do I want it to be smaller, but also lighter. The recycling time of speedlights (when I can't use my strobes) is probably the biggest problem I have with them. My Canon 600 has a limit on the number of successive shots which may be fired before heat damage becomes a concern. As far as I know, this is a problem with all speedlights. Of course I don't know of anyone who hits their flashes that hard, it is a problem though. My LED flash as it is now is capable of going beyond the 48 shots my Canon flash can do. I don't know what the YongNuo can do before throwing a temperature warning. They make a mention of being able to support 8fps with the flash set at 1/8 power or below.

My first problem was trying to realize if an LED would work being fired as a strobe. This is why I went with a macro setup. I don't plan on building up a marketable macro ring light. I'm working on getting a small setup firing correctly, then scaling up the entire design.

I haven't built up a light measuring circuit just yet. So I don't know how much light my macro flash supplies. If LED droop doesn't become an issue, then I may overdrive the daylights out of it and attempt to get more than its rated output of 1W. I'm not entirely sure how photography suppliers do this measurement. I plan on putting a photodiode at several points in proximity to the LED flash, picking this up with my oscilloscope, then taking the integral of the area under the curve which is drawn by my photodiode. This will give me time in seconds, amplitude in volts and I'll have the scale factor for the photodiode telling me how much current it produces per unit of light measured.

In the end I'd like to achieve:
smaller package
lighter weight
faster recycling
wireless connectivity
water resistant housing
sync speeds up to 1/8000s
eliminate the need for external color gels
external power input

AG_Boston wrote:

I've wondered about stuff like this. I use my strobes and flashes in manual mode. I have tried making use of TTL. Most of the time TTL seems to want to do things I don't want it to do. Are automated metering functions something you make a lot of use of?

It depends on which flash I'm using and whether I'm moving or still.

Mikey McMichaels wrote:

It depends on which flash I'm using and whether I'm moving or still.

I've started looking into this and have an idea of how I might be able to do it rather easily. It won't be until at least next week that I'll be able to try it. I'll update you if I figure it out.

Michael McGowan wrote:
My main use for a flash is outdoors for fill light on portraits. Several times a month, I have a gig where I shoot folks outside, usually in the shade but sometimes backlit. I need a flash that will automatically handle the various situations, do TTL metering and keep up with my shooting. My Nikon SB800 does a pretty good job. If you can make a flash that will do all of that for less money, you'll do OK.

How many SB800s do you typically use?

Side note: A guy I work with was given and SB800 by his dad. It apparently burned some of its guts. I get to check it out this week. I'm kind of excited about seeing a smoked SB800. It's older and certainly lived its life to the fullest.

AG_Boston wrote:

How many SB800s do you typically use?

Side note: A guy I work with was given and SB800 by his dad. It apparently burned some of its guts. I get to check it out this week. I'm kind of excited about seeing a smoked SB800. It's older and certainly lived its life to the fullest.

Sorry not to get back sooner. I only use one for this because the job entails multiple locations and minimal setup.

My SB800s ... yes, I have several ... all seem to be operating OK. But I'm an old photojournalist, and I never count on a camera or flash to work. Always be prepared with a backup.

AG_Boston wrote:
https://flic.kr/p/sFAmXg
https://flic.kr/p/sDgj2N

Which are the Exif for that images?

Since the efficiency of a Led varies inversely with its optical output power:
LED's efficiency exceeds 100%
do you think you coud do a test using a grid of led with lower power output for see if the result remain good?

Also, depending on the diffuser used, it should be possible to maximize the amount of the output light, with a proper position of the grid.

Personality Imaging wrote:
The problem is to minimize the thermal resistance between the LED and a giant heat sink so that the LED can be used in a flash mode without nuking itself.  So far nobody has come up with a great solution, on the other hand I haven't seen any really good approaches.  I think that this is possible if someone really attacks the problem with the proper technical resources, both modeling and processing.  We need at least a 10X improvement over CW (continuous wave), so far I think the LED lights that flash are at about 4X.

Ah! The thermal issue becomes less so when you're operating the led as I was. The led I'm messing with is rated to run at 37VDC. I'm overdriving it at 45VDC with no heatsink. The flash duration I'm using is 200ms (0.2 seconds) which is fast enough to avoid excessive heating of the led silicon die.

Now if I run the led as a continuous source then I will need a heatsink and possibly forced convection to avoid damaging the led.

It might also be noted that I was able to do several shots in rapid succession with minimal recycling time and very little heating of the led silicon die. I'll try to get some of those burst shots up online.

Per Rutquist wrote:
This is very interesting. What is the flash duration at 20 Ws ?

In these photos the flash duration was 200ms. The microcontroller I'm using gives me the ability to make this longer or shorter.

AG_Boston wrote:
In these photos the flash duration was 200ms.

The latency of the blink reflex is on the order of 100 ms. How many of the photos got ruined because the model blinked when the flash turned on?

Have you considered using LiPo batteries instead of a capacitor? That would give arbitrarily long flash duration, and zero recycling time.

Per Rutquist wrote:

The latency of the blink reflex is on the order of 100 ms. How many of the photos got ruined because the model blinked when the flash turned on?

Have you considered using LiPo batteries instead of a capacitor? That would give arbitrarily long flash duration, and zero recycling time.

You will be as surprised as I was at how few photos were ruined. She was actually stopping and posing between shots. So the only bad shots I ended up with were from the autofocus not having enough contrast to work.

The thought of using rechargeable batteries has occurred to me. I'm incredibly reluctant to get into charging batteries for a whole myriad of reasons.   Perhaps at some point I will. I designed a power supply which did make use of LiPo cells. They worked really well. At around 3.7VDC I'll need a battery of them to get near 45VDC.

I want to build a few of these eventually for other people to try out.  It will be a while from now though.

Good Egg Productions wrote:
I get what you're trying to do.  But what I don't get is the application.

I can see an extremely versatile tool in there!

Imagine a light panel (AKA soft-box) with a two-setting remotely controlled switch. 

1 - Really bright constant lighting

2 - Moderately bright constant lighting (modeling light) that "flashes" to extremely high output (strobe) on input from the sync

I could light a set for video, then flip a switch that would give me faux-strobes for still photography - with the same color temperature and CRI in both cases!

Open up project in Kickstarter and I'll jump right in!!

alessandro2009 wrote:

Which are the Exif for that images?

Since the efficiency of a Led varies inversely with its optical output power:
LED's efficiency exceeds 100%
do you think you coud do a test using a grid of led with lower power output for see if the result remain good?

Also, depending on the diffuser used, it should be possible to maximize the amount of the output light, with a proper position of the grid.

So, the EXIF data won't be very impressive at all. The photos in the above links were shot on a Canon 40d with the 50mm f/1.8 lens, 1/160s exposure, 1000ISO and f/4.

I believe the paper linked above is also referenced in a few papers talking about LED droop (probably not the one below though):
http://spectrum.ieee.org/semiconductors … -led-droop

Oddly enough, I don't own any grids.

Personality Imaging wrote:
Maybe I'm missing something here but it looks like you are shooting macro.  If you put enough current through the LED to light up a scene at ten feet at f/8 and ISO 200 it seems like you are going to need a much better heat sink than anyone in the industry has come up with.    If you can do this you can definitely blow away everyone in the LED photo industry and I'm in line to buy one!

This did start out as a macro LED flash project. The initial idea was to do it on a small scale and move up to a human sized scale. Now I'm at the human sized scale. All of the photos of the $100 bill were to give me confidence in being able to sync and LED, controller and my camera's shutter. So all photos from here on out will be human sized unless stated otherwise.

10', f/8 and ISO200 are good goals. I think that's doable, but will take a bit of time. I'll need to get a bunch more LEDs and build a few more circuits. I also plan on putting an aluminum slug onto the back side of the LED to dissipate heat during burst shots. I only was able to do 6 shots/second. The LED did get warm and cooled down rather rapidly. The reason this is working is because I'm running the LED at a very small duty cycle. If this device were run in its current configuration at 100% duty it would become a nice former LED.

still-photography wrote:

I can see an extremely versatile tool in there!

Imagine a light panel (AKA soft-box) with a two-setting remotely controlled switch. 

1 - Really bright constant lighting

2 - Moderately bright constant lighting (modeling light) that "flashes" to extremely high output (strobe) on input from the sync

I could light a set for video, then flip a switch that would give me faux-strobes for still photography - with the same color temperature and CRI in both cases!

Open up project in Kickstarter and I'll jump right in!!

The constant light part can be done, but will definitely require a massive heatsink and possibly forced convection (a fan).

A KIckstarter? You're funny. It took me a year just to get time to do this part. Imagine how long my Kickstarter would last...just in time for retirement! WOOOO!!! I'm actually not sure what I'll do to get these sold once I feel they're ready. It may be a good six months of testing before I have enough confidence with regards to delivering a quality product.

Michael McGowan wrote:

Sorry not to get back sooner. I only use one for this because the job entails multiple locations and minimal setup.

My SB800s ... yes, I have several ... all seem to be operating OK. But I'm an old photojournalist, and I never count on a camera or flash to work. Always be prepared with a backup.

I did a bunch of shots here in Boston with a speedlight on my Canon 5d. I used the 600 one they make. It was able to a few photos with a tiny aperture opening for about 50 shots, then started to take a long time to recycle. I've started experimenting with shooting outdoors with strobes and speedlights more. I'm liking the results so far. My LED I haven't tried outside as of yet. It may only be good for some fill flash at this point. Who knows though? I may get it to work much better in the next few months.

AG_Boston wrote:
So, the EXIF data won't be very impressive at all. The photos in the above links were shot on a Canon 40d with the 50mm f/1.8 lens, 1/160s exposure, 1000ISO and f/4.

Wait, what? You're using a 1/160 s (6.3 ms) shutter speed with a 1/5 s (200 ms) strobe duration. This means that most of the flash output is wasted after the shutter has closed. Effectively you're probably getting less than 2 Ws - not 20.

Comparing to a normal speedlite you're at something like 1/32 power. At that level, the speedlite can probably shoot 10 frames per second for 50 frames before recycling time become an issue, so your LED doesn't have any advantage there.

Well, at least this explains why you did not have a problem with the model blinking.