High-power LED flash: too good to be true?

Started Aug 17, 2013 | Discussions
Shop cameras & lenses ▾
cyrano Contributing Member • Posts: 896
High-power LED flash: too good to be true?

A crowdfunded project for an LED flash unit is claiming light output equivalent to a Canon 580 EX II speedlite, with zero recycling time, high-rate multiple flash capability, and an internal battery capable of delivering 1000 full-power pops on a single charge.

If you have technical understanding of flash design and battery technology, I'd be interested in hearing your opinion of whether or not these claims are feasible.

Brent Allen Thale Regular Member • Posts: 486
Re: High-power LED flash: too good to be true?

That does sound good, according to them the light has no weaknesses and can produce miracles with no downsides. Makes me wonder why no one else has done this.

They say the duration of the light is variable and can be set as fast as 1/8000 second, but they don't say how that impacts exposure. This would seem to require a separate control that currently doesn't exist on cameras today, since for most strobes flash duration is directly tied to power, and can't be set independently.

On their web site they have a video showing a pro photographer using the light, but they don't say what camera settings, like shutter speed, ISO, and aperture were used. And when they shoot through an umbrella, they added a second light, which suggests the light is nowhere near as powerful as a top-of-the-line speedlight.

Video and still shooting have different lighting needs, it seems amazing that this one light can do it all.

 Brent Allen Thale's gear list:Brent Allen Thale's gear list
Nikon D7000 Nikon D7100 Nikon AF-S DX Nikkor 18-105mm f/3.5-5.6G ED VR Nikon AF-S DX Nikkor 35mm F1.8G Nikon AF-S Nikkor 70-200mm f/4G ED VR +6 more
Nionyn Senior Member • Posts: 1,987
Further discussion in Strobist group on Flickr

There's currently a thread going about this in the Strobist group on Flickr, here: http://www.flickr.com/groups/strobist/discuss/72157635062064415/

The bloke leading the Flash Cube's development is also taking part, and has answered some of the questions that have been posed there.

It's still early days, obviously, but potentially interesting.

-- hide signature --
Funduro
Funduro Contributing Member • Posts: 983
Re: High-power LED flash: too good to be true?

I might not have all the point correct, but I read an article on the wavelengths of LED's and the measured values. Basically the bad news is there are peaks and valleys of the light across different wavelengths (blue,red,K value). I forgot who or what site it was, but the person seemed to very knowledgeable. In short do not bet the family farm on LED flash replacing top units from Canon, Nikon, Nissin, Metz, Sigma, . If they get the bugs worked out sure they will sell them, will they be as good as the top ones, who knows.

-- hide signature --

It's a "hobby" not an obsession. OK it's an addiction sort of.

 Funduro's gear list:Funduro's gear list
Fujifilm FinePix X100 Olympus XZ-1 Nikon D700 Nikon D300S Nikon D800 +11 more
Brent Allen Thale Regular Member • Posts: 486
Re: High-power LED flash: too good to be true?

They posted a lighting setup diagram where they indicate the flash duration is set to 1/500 second, and the camera shutter speed is also 1/500 second. I don't understand how this is possible since at 1/500 second, the sensor on the 5D Mark II will never be completely open, so how are they evenly exposing the frame with a flash duration of 1/500?

http://www.indiegogo.com/projects/the-led-light-cube?c=gallery

 Brent Allen Thale's gear list:Brent Allen Thale's gear list
Nikon D7000 Nikon D7100 Nikon AF-S DX Nikkor 18-105mm f/3.5-5.6G ED VR Nikon AF-S DX Nikkor 35mm F1.8G Nikon AF-S Nikkor 70-200mm f/4G ED VR +6 more
Phil
Phil Veteran Member • Posts: 3,098
Re: High-power LED flash: too good to be true?

Brent Allen Thale wrote:

They posted a lighting setup diagram where they indicate the flash duration is set to 1/500 second, and the camera shutter speed is also 1/500 second. I don't understand how this is possible since at 1/500 second, the sensor on the 5D Mark II will never be completely open, so how are they evenly exposing the frame with a flash duration of 1/500?

http://www.indiegogo.com/projects/the-led-light-cube?c=gallery

The Pocket Wizard TTL triggers for Canon EX flashes have a hyper sync mode that shaves the flash timing closer allowing a 2x sync speed. I'd have to research what's the fastest I've been able to use. 1/500th sounds high but it's not too far off, so I'd say this was at least possible.

-- hide signature --

Phil Agur
5DMKII - 5DMKII Dual Body Bag - Go 'L' and don't look back!

 Phil's gear list:Phil's gear list
Canon EOS 5D Mark II Canon EOS 5D Mark III Canon EF 17-40mm f/4.0L USM Canon EF 24-70mm f/2.8L USM Canon EF 70-200mm f/2.8L IS II USM +4 more
ianw2 Regular Member • Posts: 113
Re: High-power LED flash: too good to be true?

Phil wrote:

Brent Allen Thale wrote:

They posted a lighting setup diagram where they indicate the flash duration is set to 1/500 second, and the camera shutter speed is also 1/500 second. I don't understand how this is possible since at 1/500 second, the sensor on the 5D Mark II will never be completely open, so how are they evenly exposing the frame with a flash duration of 1/500?

http://www.indiegogo.com/projects/the-led-light-cube?c=gallery

The Pocket Wizard TTL triggers for Canon EX flashes have a hyper sync mode that shaves the flash timing closer allowing a 2x sync speed. I'd have to research what's the fastest I've been able to use. 1/500th sounds high but it's not too far off, so I'd say this was at least possible.

-- hide signature --

Phil Agur
5DMKII - 5DMKII Dual Body Bag - Go 'L' and don't look back!

Nikon also have  a mode on their flashes that allows shutter speeds greater than the sync speed. I expect Canon have an equivalent system.

All these systems work either by extending the flash duration (the Nikon system which fires a series of smaller flashes), or taking advantage of the longer duration of the flash (the Pocket Wizard system). The Pocketwizard system works best with studio flashes which are generally more powerful and have a longer duration than Speedlights, but some modest gains have been reported with Speedlights (and Speedlites).

For such a system to work the flash must stay on while the curtain is open.  If the camera shutter speed is 1/500 second, the curtain will be open for longer than this. I expect they fire the flash multiple times so each part of the sensor receives a flash duration of 1/500 second.

-- hide signature --

Ian Wiese
ianwiese.wordpress.com

Phil
Phil Veteran Member • Posts: 3,098
Re: High-power LED flash: too good to be true?

ianw2 wrote:

Phil wrote:

Brent Allen Thale wrote:

They posted a lighting setup diagram where they indicate the flash duration is set to 1/500 second, and the camera shutter speed is also 1/500 second. I don't understand how this is possible since at 1/500 second, the sensor on the 5D Mark II will never be completely open, so how are they evenly exposing the frame with a flash duration of 1/500?

http://www.indiegogo.com/projects/the-led-light-cube?c=gallery

The Pocket Wizard TTL triggers for Canon EX flashes have a hyper sync mode that shaves the flash timing closer allowing a 2x sync speed. I'd have to research what's the fastest I've been able to use. 1/500th sounds high but it's not too far off, so I'd say this was at least possible.

-- hide signature --

Phil Agur
5DMKII - 5DMKII Dual Body Bag - Go 'L' and don't look back!

Nikon also have a mode on their flashes that allows shutter speeds greater than the sync speed. I expect Canon have an equivalent system.

All these systems work either by extending the flash duration (the Nikon system which fires a series of smaller flashes), or taking advantage of the longer duration of the flash (the Pocket Wizard system). The Pocketwizard system works best with studio flashes which are generally more powerful and have a longer duration than Speedlights, but some modest gains have been reported with Speedlights (and Speedlites).

For such a system to work the flash must stay on while the curtain is open. If the camera shutter speed is 1/500 second, the curtain will be open for longer than this. I expect they fire the flash multiple times so each part of the sensor receives a flash duration of 1/500 second.

-- hide signature --

Ian Wiese
ianwiese.wordpress.com

IAN,

There are 3 modes available with the PW in the system. Canon & Nikon both have normal sync and high speed sync (where the flash strobes at a fast rate to minimize the impact of flashing while the sensor is partially blocked by the curtains) but the PW set-up adds what they call hyper sync, a third mode.

Hyper sync is still a single full power flash, like a normal flash event, but uses a refined curtain timing while telling the camera the flash is set to high speed sync. The result is you can get 2X or a little more than the camera's normal sync speed without catching a curtain while having a full power flash event available.

-- hide signature --

Phil Agur
5DMKII - 5DMKII Dual Body Bag - Go 'L' and don't look back!

 Phil's gear list:Phil's gear list
Canon EOS 5D Mark II Canon EOS 5D Mark III Canon EF 17-40mm f/4.0L USM Canon EF 24-70mm f/2.8L USM Canon EF 70-200mm f/2.8L IS II USM +4 more
ianw2 Regular Member • Posts: 113
Re: High-power LED flash: too good to be true?

Phil wrote:

ianw2 wrote:

For such a system to work the flash must stay on while the curtain is open. If the camera shutter speed is 1/500 second, the curtain will be open for longer than this. I expect they fire the flash multiple times so each part of the sensor receives a flash duration of 1/500 second.

-- hide signature --

Ian Wiese
ianwiese.wordpress.com

IAN,

There are 3 modes available with the PW in the system. Canon & Nikon both have normal sync and high speed sync (where the flash strobes at a fast rate to minimize the impact of flashing while the sensor is partially blocked by the curtains) but the PW set-up adds what they call hyper sync, a third mode.

Hyper sync is still a single full power flash, like a normal flash event, but uses a refined curtain timing while telling the camera the flash is set to high speed sync. The result is you can get 2X or a little more than the camera's normal sync speed without catching a curtain while having a full power flash event available.

-- hide signature --

Phil Agur
5DMKII - 5DMKII Dual Body Bag - Go 'L' and don't look back!

Sorry Phil - when I said "they" in my last paragraph (quoted above) I meant the LED Light Cube people, not Pocketwizard. You are correct that Hypersync is still a single full power flash.

For anyone interested there is a good explanation at http://wiki.pocketwizard.com/index.php?title=Understanding_HyperSync_and_High_Speed_Sync

-- hide signature --

Ian Wiese
ianwiese.wordpress.com

Brent Allen Thale Regular Member • Posts: 486
Re: High-power LED flash: too good to be true?

If you look at the image posted by LightCube containing the lighting diagram, it appears that no PocketWizards were involved: http://images.indiegogo.com/medias/990821/pictures/full/20130817183338-BalloonSetup.jpg?1376789622

So, I don't see how the flash duration of 1/500, which they explicitly mention in the diagram, would cover the frame evenly in their scenario.

 Brent Allen Thale's gear list:Brent Allen Thale's gear list
Nikon D7000 Nikon D7100 Nikon AF-S DX Nikkor 18-105mm f/3.5-5.6G ED VR Nikon AF-S DX Nikkor 35mm F1.8G Nikon AF-S Nikkor 70-200mm f/4G ED VR +6 more
Phil
Phil Veteran Member • Posts: 3,098
Re: High-power LED flash: too good to be true?

Brent Allen Thale wrote:

If you look at the image posted by LightCube containing the lighting diagram, it appears that no PocketWizards were involved: http://images.indiegogo.com/medias/990821/pictures/full/20130817183338-BalloonSetup.jpg?1376789622

So, I don't see how the flash duration of 1/500, which they explicitly mention in the diagram, would cover the frame evenly in their scenario.

The Pocket Wizard results were put forward as an independent example that 1/500th sync speed is indeed possible without a curtain issue if you refine the flash timing, not to say that they were involved in taking this photo.

If we only had Canon sync speed data to go on this would be an impossible claim.

-- hide signature --

Phil Agur
5DMKII - 5DMKII Dual Body Bag - Go 'L' and don't look back!

 Phil's gear list:Phil's gear list
Canon EOS 5D Mark II Canon EOS 5D Mark III Canon EF 17-40mm f/4.0L USM Canon EF 24-70mm f/2.8L USM Canon EF 70-200mm f/2.8L IS II USM +4 more
nelsonal Senior Member • Posts: 2,464
Re: High-power LED flash: too good to be true?
1

Their product description describes it as a 350W peak LED flash source, (at 150W constant source it's great for video). For very small video work, it's got many of the advantages of a Joker light at a price that's probably close to the rental! LEDs that I've seen can hit about 100 Lumens per watt, so lets give them the benefit of the doubt. That would make this a 35,000 Lumen source.

However, it's nowhere near a speedlight. For photography Lux seconds are the operative measure. If we imagine a 75J strobe with a 70 degree reflector (ie an alien bee B400 at half power) that's close to the energy of most speedlights (the speedlight would need to be at a fairly wide setting). That's about 3500 Lumen Seconds. At 3 meters that's 300 Lux seconds of luminous exposure.

The LED cube is 35000 lumens, but in a photo exposure, it needs to deliver light in say 1/100th of a second. Which means luminous energy is only 350 ls in the relevant exposure. Which means that with a similar 70 degree spread (I didn't see much documentation about a reflector/fresnel) this will only generate 34 Lux seconds of luminous exposure. So our speedlight is producing about 10x the useful light for a photo (it gets worse at higher flash sync speeds). So great for video, where $500 bought a glorified grow lights while this is battery powered, smaller, dimmable, and could have a higher CRI.

I'd love to hear my math is wrong, but this doesn't seem like it's quite up to speedlight power.

It appears the equivalence is a fairly tight reflector (30 degree) on the cube guide number either 16 or 22, both in meters, vs a wide reflector on the strobe (14mm looks close guide number 15 also in meters.  At the closest reflector setting the 580 EX II has a guide number of 42!

Phil
Phil Veteran Member • Posts: 3,098
Re: High-power LED flash: too good to be true?

nelsonal wrote:

Their product description describes it as a 350W peak LED flash source, (at 150W constant source it's great for video). For very small video work, it's got many of the advantages of a Joker light at a price that's probably close to the rental! LEDs that I've seen can hit about 100 Lumens per watt, so lets give them the benefit of the doubt. That would make this a 35,000 Lumen source.

However, it's nowhere near a speedlight. For photography Lux seconds are the operative measure. If we imagine a 75J strobe with a 70 degree reflector (ie an alien bee B400 at half power) that's close to the energy of most speedlights (the speedlight would need to be at a fairly wide setting). That's about 3500 Lumen Seconds. At 3 meters that's 300 Lux seconds of luminous exposure.

The LED cube is 35000 lumens, but in a photo exposure, it needs to deliver light in say 1/100th of a second. Which means luminous energy is only 350 ls in the relevant exposure. Which means that with a similar 70 degree spread (I didn't see much documentation about a reflector/fresnel) this will only generate 34 Lux seconds of luminous exposure. So our speedlight is producing about 10x the useful light for a photo (it gets worse at higher flash sync speeds). So great for video, where $500 bought a glorified grow lights while this is battery powered, smaller, dimmable, and could have a higher CRI.

I'd love to hear my math is wrong, but this doesn't seem like it's quite up to speedlight power.

It appears the equivalence is a fairly tight reflector (30 degree) on the cube guide number either 16 or 22, both in meters, vs a wide reflector on the strobe (14mm looks close guide number 15 also in meters. At the closest reflector setting the 580 EX II has a guide number of 42!

I ran into an LED vendor over the weekend at a harbor street fair last Saturday in Oregon. Among all the household and light rope goodies he had some marine spot lights. They drew 30 x 50 watts at 12V but of course were designed for constant use (in flash terms) as docking or fishing trawler work light.

I was looking at giant single surface LEDs an inch or more across with a very shallow reflector (aluminium by appearance) that was most likely the heat sink not an actual reflector. I could see the 50 being a 400-500 watt equivalence in constant use but who knows as a strobe, 1000 watt plus?

High brightness LEDs are often pulsed because our eyes fill in the blanks like movie frames and it saves heat at the LED. You can't do that in still photography unless you time it so that could be the basis of a flash. I'd think you'd have to preheat the LED to cut down the intensity of the thermal cycle, which I believe would be the limiting factor. Instead of re-charge recycle time you'd be limited by heat dissipation rates and thermal shock related lifetime reduction results.

-- hide signature --

Phil Agur
5DMKII - 5DMKII Dual Body Bag - Go 'L' and don't look back!

 Phil's gear list:Phil's gear list
Canon EOS 5D Mark II Canon EOS 5D Mark III Canon EF 17-40mm f/4.0L USM Canon EF 24-70mm f/2.8L USM Canon EF 70-200mm f/2.8L IS II USM +4 more
William Faulkner Senior Member • Posts: 1,641
Re: High-power LED flash: too good to be true?

True specs or not looks like we'll find out soon.

LED Light Cube went over 200% of their funding goal.

Now they just have to make some.

http://www.indiegogo.com/projects/the-led-light-cube-for-photographers-filmmakers/x/4463143

These could be the first of an entirely new category of flash system. They might not be near the equivalent of conventional strobes today but in 5-years.......? Technology has a way of getting accelerated when folks like what's happening.

Who was able to predict that in ten years, give or take, silver based photography would be gone?

And yeah, I bought some of the Cubes.

Bill F

https://picasaweb.google.com/faulknerstudios

 William Faulkner's gear list:William Faulkner's gear list
Nikon AF-S Nikkor 14-24mm f/2.8G ED
HumanTarget Senior Member • Posts: 1,194
Re: High-power LED flash: too good to be true?

Phil wrote:

High brightness LEDs are often pulsed because our eyes fill in the blanks like movie frames and it saves heat at the LED.

I didn't know that until I tried taking some pictures of LED Christmas lights.  Took me by surprise!

RumpelHund Regular Member • Posts: 179
Too good to be true unless you trick on the calculations

I work for years in industrial image processing with LED ring lights up to water-cooled units with line lights. My opinion: that product is wishful thinking to claim popping as a speedlight can.

Background:

To drive a fast pop of light you will need to insert very high power into the light bulb, LED or not. A battery is always limited in this regard, so you will either have to put more in parallel (higher voltage, possibly more amps) or bunker the energy in capacitors first.

In industry usually you take a rather high supply voltage like 48 volts with high power at hand (like 20Amp spikes), to drive a nativly 12V LED assembly and control the power spread by a controller limiting the amps . You can immediatly kill the LED if you like, to prevent it you will need to calculate the power used. Based on overdrive level and useage ratio. Anywhere between 100% power constantly and 10000% power for 1% of the time is fine as long as the head can spread.

My take is that they do rate the power as being equal to the speedlight at 1/8000s as the FP-mode speedlight is much much weaker than it is at full pop.

I think full pop LEDs will come eventually, but they'll need capacitors as well.

The industrial tools I use (and tried them for photography, no way useful) are in the range

http://smartvisionlights.com/

http://www.stemmer-imaging.co.uk/en/products/article/10164

http://www.gardasoft.com/LED-Controllers/

So I am sure it's a long way until LED for strobe will be cheaper at same pop as a speedlight. Happy to be proven otherwise...

studio460
studio460 Contributing Member • Posts: 712
Re: High-power LED flash: too good to be true?

I just wrote a fairly lengthy article about the LED Light Cube on my personal website if anyone's interested: http://lightbasics.com/.

studio460
studio460 Contributing Member • Posts: 712
Re: High-power LED flash: too good to be true?

I just received my LED Light Cube from the UPS truck a few hours ago, and now I've just finished writing the first hands-on LED Light Cube review on the internet (includes actual test images):

http://lightbasics.com/led-light-cube-hands-on-review/

studio460
studio460 Contributing Member • Posts: 712
Re: High-power LED flash: too good to be true?

11/22/2014 UPDATE: I just edited my LED Light Cube review to include some new flash meter readings using a Sekonic L-478DR flash meter to measure the Light Cube's actual output at full-power:

http://lightbasics.com/led-light-cube-hands-on-review/

Nionyn Senior Member • Posts: 1,987
Re: High-power LED flash: too good to be true?

Thanks for the thorough and well written review - much appreciated.

It'll be interesting to see what the market eventually turns out to be for the Light Cube. Light colour, power, weight and bulk suggest that it's definitely not going to compete with hotshoe flashes for stills work, but it might work well for those wanting a relatively compact light source which can double (to some degree) for video and stills work.

Of course, if Nathan Oxley and IC12 manage to make incremental performance improvements over the coming years to future models it could become far more interesting.

Keyboard shortcuts:
FForum MMy threads