This is 100% about dumb, manual aspect of these flashes. TTL and remote (CLS) features do not matter. Sure they may influence your choice, but this test is about power only, and how different flashes compare to others.

A big thing I wanted to indicate was what someone should actually expect. A Canon 580exII has a GN of 119@35mm, so one should expect to shoot at f/19 (f/9.5 in a 2 stop eating umbrella) at 10 feet, right? Then why did I get f/5? I'm sure you've heard that "GN is a joke, cause it's done in a super white room, blah, blah, blah" It's true, published guide numbers are inflated in a few different ways, but since they're ISO, it's the same inflation across the board. I wanted to show a more practical "what one should expect" diagram.

I conducted the test as uniformly as possible. The face of the umbrellabox I used was 8' from my Sekonic L358, with the flash head 12" from the reflected surface, for a 10 foot complete travel distance. Each flash was fired 10 times, with 30 seconds passing between firing.

The same exact modifier was used in each test. I am fully aware that modifiers eat power, but using the same modifier should keep things standard. If flash A is 10% more powerful than flash B in a reflective umbrella, it should be 10% more powerful in another modifier. I know each modifier may be different, but this should give a good comparison of relative flash power among popular flashes. This is not a Guide Number battle, just a comparison of power.

Some of these flashes are older, and there very well may be some variation in power. Just as well as one of "Flash A" may be a tad different from another copy of "Flash A." However, from my research, all seem to be on-par with the power given from the manufacturer.

The setup:

Flashes bounced into a 40" umbrellabox (bounce then diffuse) measured 8 feet from a sekonic L358, flash and meter 5' high. There should be minimal bounce, but as mentioned in the second point, it should affect each flash evenly, so the comparison should still be valid.



A simple breakdown of each flash, and its metered reading:





I've broken it down further with price, price per stop, etc... but my main goal was to show how little difference there is among flashes when used in a modifier. For the most part, there is only about one stop difference between the lowest and highest hotshoe flashes above.

Vivitar 285, Canon 430ex, SB600, SB700, all would have given good images at f/3.5 or so in camera.

The "much much more powerful" flashes, the SB800 / 580exII, were only a stop more at f/5 (within a third of a stop, sb800's 5.3, but still) ... one stop can be a big deal, but it's still an interesting observation.

Another interesting point, the generics and old junkers.
$40-$60 YN460II / 560II putting out only a third of a stop less than -or matching- the top of the line flashes. With the other budget flashes (though less inexpensive) the Lumopros just behind in power. This struck me the most. A $45 flash  being only 1/10 of as stop less powerful than the top of the line guys is interesting.

It was these flashes, the 460II and the 560 (same power) that instigated this test. There are a ton of "I need a cheap flash for OCF use" threads, and the 460II and 560s are highly recommended. However, when the 460II is brought up there's always the recommendation, "get an LP160, it's much more powerful" or "get the 560, it's way more power" ... "stay with Nikon/Canon, much more powerful, and more versatile" (the versatile is a given, but just talking about power here) when these (in my tests here, and over the past few years of using many.... many... many different varieties of speedlites) aren't very true facts.

Yes, there are a billion reasons why one flash is better than another, why one is better for you... but power is a big metric when deciding which flash to get for (emphasis: dumb) off camera use, a $45 flash that's every bit as powerful as its $450  (or even $150) counterpart can be a big deal, and a matter of getting one flash, or four and a new lens.

IMHO, as always.

AVD AlphaDuctions wrote:
I know you put a lot of work into this and I don't want to sound too negative but I think I have to bring up a couple of points

1) when you buy a Canon Nikon or Sony speedlight you are paying for a lot more than just power. You are paying for 100% compatibility with the brand's implementation of TTL, HSS, off-camera flash and other features.

2) you really didnt have to do all that testing to know that a Sony 56 is a stop more than a 42.  All you need is to look at the published GN.  Same for all the other brands.

I get both points, I do. But:

1) many of us do not buy Canon / Nikon flashes for the TTL/HSS, etc... We buy speedlites for dumb off camera use. I know advanced wireless is an option, but many love dumb manual mode. Though I have spoken to MANY people in my small intro lighting workshops who bought a 580exII because it "has Gn 191!! that beats the shit out of 'flash X' "

2) Yes, 56 is a stop more than a 42. But is an sb800 24 stops more than an sb-24? how about a vivitar vs a lumopro? And published GN aren't easy to find the right answer for. Vivitar 285 is GN120... or is it 100, or 80? they're all listed. 580exII says both 36 and 58 (however it's 36 typically when in an umbrella, zoom is moot) Some nikon and generic flashes rate at ISO200, does that mean they're 1.41x more powerful than those who rate at iso100, if it's not labeled in an easy to find place?

Modifiers equalize things. The GN of a flash is modified by the zoom factor of the head. However once tossed into a modifier, the beam is spread out to a point that it changes everything, and the power output is similar to the power of the 28/35mm zoom setting.



I was just kinda bored, and it's a topic I feel comes up enough where having it laid out, in a practical manner (ie, how you'd actually use these flashes) would be nice.

DougBPhoto wrote:
Jay,  it appears you have WAY too much free time on your hands for someone with such a beautiful fiance!!!

She's at work


Edit 13 months later - wife, not fiancee

Know Idea wrote:
Nice work, Jay !!

Your efforts are appreciated!!!

thanks. I know a lot of my posts cause drama, but if one person makes a more well informed decision, and does whats best for them... I'm happy.

AVD AlphaDuctions wrote:
2) you really didnt have to do all that testing to know that a Sony 56 is a stop more than a 42.  All you need is to look at the published GN.  Same for all the other brands.

Yeah, but Guide Numbers are as much marketing as they are a physical characteristic of a flash.  Independent testing is always useful.

ImageX wrote:
Nice write up... and a bump for the YN460 II's. I have a handful of them($30 each) and they have been very reliable and perform well. So well, that I've never even considered the 560's. I also have a few SB 600's and a SB 900 which I usually use as a master.

I've got one 560, one 565ex, and (7) 460IIs.

I use the 560 regularly. On occasion, I want a rim/hair light on location, and I can set the zoom manually to 105mm (460II has no zoom head,) and place it out of frame, and it still has enough juice to be +1 stop over main, and give a nice rim/hair light. Tool for the job... it has its place.

David Parsons wrote:
Yeah, but Guide Numbers are as much marketing as they are a physical characteristic of a flash.  Independent testing is always useful.

it's one of the first things I do with new lights / modifiers. I have a thing taped to most of my most used modifiers with 'effective guide numbers' written on them.

Written on my PLM:      (all at ten feet)
Photogenic - f/22
Sunpak - f/13.5
YN460II (x3) - f/11 - (I often use a triflash bracket)
Yn460II - f/7.1

So when I set it up, I can take a quick look, and know that if I am at 8 feet with my photogenics, and want to shoot at f/5.6ish, I'm at 1/4 - 1/8 power.... gives me a place to start.

Ledo retouch wrote:
yeah for the cheap and dumb flashes, 'cuase I'm cheap and dumb, plus
I have Canon and Nikon bodies.  but I haven't heard of the Lumopro flash.
what are they?

Retread of the Vivitar 283/285 flashes that are made for Midwest Photo Exchange (mpex.com).  They added additional sync ports of various flavors.

Ledo retouch wrote:
yeah for the cheap and dumb flashes, 'cuase I'm cheap and dumb, plus
I have Canon and Nikon bodies.  but I haven't heard of the Lumopro flash.
what are they?

David Parsons wrote:
Retread of the Vivitar 283/285 flashes that are made for Midwest Photo Exchange (mpex.com).  They added additional sync ports of various flavors.

Not quite. Cactus bought the 285, and remade it. Used and 100% reliable: $45 online... they made them new for $100... it flopped.

MPEX wanted to make a 100% "american made" flash, and with help from the strobist community, created the LP120. GN80, decent flash, it had everything a strobist wanted, though a little weak... but for $120 to support an "all american, made in america" company? It was popular... I supported them fully.

... ... ...So popular they had to outsource to China... ... ...

Now they're no different than YongNuo flashes, at 4x the cost.

Ledo retouch wrote:
thanks, I have a 283 and a 285 that are older than most people on this site,
they have been bullet proof, but can't use them on camera because high
trigger voltage, must be a good design to be around this long.

The 285HV is 'safe' for digital, but not worth the risk. I think a big reason they stuck around was that they were $40 before any of the cheap budget chinese versions came around, and the notoriety stuck.

I agree, bulletproof, great flashes... though the proprietary sync blows unless you use a hotshoe mounted trigger (back in the days of Cactus V2, there were none, really) and they're SLOOOOOooooooooowwwwwwwwww.

Ledo retouch wrote:
thanks, I have a 283 and a 285 that are older than most people on this site,
they have been bullet proof, but can't use them on camera because high
trigger voltage, must be a good design to be around this long.

If you shoot with a Nikon, maybe you can.

When my son was considering buying a D100 several years ago, he was told by Nikon support that the D100 can handle a much higher trigger voltage than most digital cameras. Enough to handle a Vivitar 283 (the original version), in fact.

I don't know about other Nikon models (including mine). I have three Vivitar 283's and five Vivitar 4600's (three of them with bare bulb heads). The only flashes I've used since I started shooting digital are studio strobes. But when I have a need to use small portable flashes, I'll sure give Nikon a call before rushing out to buy something else.

Camerosity wrote:
If you shoot with a Nikon, maybe you can.

I won't be taking that risk.

I also have a D100 (as well as D200, D300s, D3200 and N90s), but I won't be putting high voltage flashes on any of 'em.  Of course, why would you want it on the hotshoe anyway?

Radio triggers can have the same issue though, not all (most, these days?) will support high voltage flash.

I've got 7 HV speedlights here of various makes, styles, sizes and power outputs that I recently picked up cheap in an auction.  They're all sat waiting for me to dig out the soldering iron to add a low voltage sync port to them via an optoisolators.

-JAY- wrote:
Oh, and I forgot to point out the potato mashers. I'll edit the OP in a bit to add details, but it's a LONG post already.

Point of interest: f/10 vs f/5 - two full stops, equivalent of about 4 of the best speedlites. (typo in the chart, f/8 + 6/10, not 2)

Thanks for putting this list together Jay.  I know your focus was on maximum power, but something to point out about the potato mashers is that they can also be powered [/b]down[b] farther than most flashes.  The Sunpak 622 Super can go down to 1/124 while my trusty Sunpak 383s can only go down to 1/16.

GM Photography wrote:
they can also be powered [/b]down[b] farther than most flashes.

They may be able to be powered down to a lower percentage of their maximum output, but it doesn't necessarily mean they'll be lower in actual light output.

Besides, if minimum power isn't minimum enough, that's what ND gels are for.

Kaouthia wrote:

They may be able to be powered down to a lower percentage of their maximum output, but it doesn't necessarily mean they'll be lower in actual light output.

Besides, if minimum power isn't minimum enough, that's what ND gels are for.

They are measurably lower in power.

GM Photography wrote:
I know your focus was on maximum power, but something to point out about the potato mashers is that they can also be powered [/b]down[b] farther than most flashes.  The Sunpak 622 Super can go down to 1/124 while my trusty Sunpak 383s can only go down to 1/16.

Kaouthia wrote:
They may be able to be powered down to a lower percentage of their maximum output, but it doesn't necessarily mean they'll be lower in actual light output.

Besides, if minimum power isn't minimum enough, that's what ND gels are for.

GM Photography wrote:
They are measurably lower in power.

AHA! This is the exact reason I made this! Sorry, GM, but they are not measurably low in power.

The Sunpak, at 1/128 power will put out f/1.8 @ ISO400
(10 - 7 - 5 - 3.5 - 2.5 - 1.8 - 1.8 iso200 - 1.8 iso400)

The Vivitar 285, the lowest I have, will also put out f/1.8 @ ISO400, at minimum (1/16) power.
(3.4 - 2.4 - 1.8 - 1.8 iso200 - 1.8 iso400) - though it has no 1/8 level

Though many of the other flashes, including the sub $100 (and sub $50) flashes that match the top ones also go down to 1/128... these offer lower power than the potato mashers, 2-3 stops lower, though f/1.8 @ iso1600 is very low.

LA StarShooter wrote:
Thanks for all the work and I really like reading the results of an independent analysis.  I use Nissin's 866is, in addition the sB800. For me, Nissins pro line deliver's warhorse performance.  Longlasting, good consistent recycling time, and manual-speaking they really deliver.  I will be buying more from Hong Kong.

Not only is power important but how long does it take to recycle. This is the reason for going with more topline speedlights.  On some speedlight models you cannot attach a battery pack, which, as you know helps with recycling time.  Nissin's i622 is a good example of that, as you can't put a battery pack on it.

Recycle time is the #1 (and #2, and #3) reason I ditched my vivitar 285s back in the day. Slow doesn't even begin to describe it. 8-10 seconds isn't unheard of.

Nowadays, There's a lot of tech that's just so prevalent... the generics really can hold their own in a lot of respects. My 460IIs recycle in about 2 seconds at full.. the 560s have the same power, recharge, etc, but have the PC sync and battery pack... My 565ex matches my 580exII in all regards, except for HSS.

Warren Leimbach wrote:
Thanks for presenting this interesting real world data.  Bookmarked for future reference.

"Real world data" is something I think is missing in a lot of photography reviews.

A new lens comes out, and it's all about MTF charts, and that one still-life shot of a bunch of shit that one company takes a picture of, so you can see how sharp the logo of a wine bottle is.

That's all good and fine, but how well does it really work when you're out using it normally? That's what I want to know.

-JAY- wrote:

"Real world data" is something I think is missing in a lot of photography reviews.

A new lens comes out, and it's all about MTF charts, and that one still-life shot of a bunch of shit that one company takes a picture of, so you can see how sharp the logo of a wine bottle is.

That's all good and fine, but how well does it really work when you're out using it normally? That's what I want to know.

I like the image comparison of lenses/cameras. What you see is often what you get. Though I do think MTF charts are half useless.

LA StarShooter wrote:
Thanks for all the work and I really like reading the results of an independent analysis.  I use Nissin's 866is, in addition the sB800. For me, Nissins pro line deliver's warhorse performance.  Longlasting, good consistent recycling time, and manual-speaking they really deliver.  I will be buying more from Hong Kong.

Not only is power important but how long does it take to recycle. This is the reason for going with more topline speedlights.  On some speedlight models you cannot attach a battery pack, which, as you know helps with recycling time.  Nissin's i622 is a good example of that, as you can't put a battery pack on it.

I also have a Nissin di866 in use with my sb8oo/sb900!, any more strobism I will be getting yn's! I believe you should test the Nisssin, they claim to be the most powerful flash guns in their class!

Raw and the cooked wrote:

I also have a Nissin di866 in use with my sb8oo/sb900!, any more strobism I will be getting yn's! I believe you should test the Nisssin, they claim to be the most powerful flash guns in their class!

Send me one and I'll test it out!

Looking at posted numbers, it might be 1/10 of a stop more powerful than the top of the line guys

-JAY- wrote:

AHA! This is the exact reason I made this! Sorry, GM, but they are not measurably low in power.

The Sunpak, at 1/128 power will put out f/1.8 @ ISO400
(10 - 7 - 5 - 3.5 - 2.5 - 1.8 - 1.8 iso200 - 1.8 iso400)

The Vivitar 285, the lowest I have, will also put out f/1.8 @ ISO400, at minimum (1/16) power.
(3.4 - 2.4 - 1.8 - 1.8 iso200 - 1.8 iso400) - though it has no 1/8 level

Though many of the other flashes, including the sub $100 (and sub $50) flashes that match the top ones also go down to 1/128... these offer lower power than the potato mashers, 2-3 stops lower, though f/1.8 @ iso1600 is very low.

I just did a quick (and admittedly unscientific) test of one of my Sunpak 383's and a Sunpak 622 Pro, holding each at arm's length and taking a few light meter readings at ISO 100.  The Sunpak 622 Pro was f25 at high power and f2.8 at lowest power.  The Sunpak 383 was f16 at high power and f5.0-f5.6 (it varied) at lowest power.

You talk about "real world" usage and that is why I knew this to be true.  I have had to switch from the 383's to a 544 or 622 on more than one occasion because I couldn't get the light output as low as I wanted it with the Sunpak 383 (1/16 power is the lowest setting).

-The Dave- wrote:
Wait, an SB900 is LESS powerful than an SB800?

In my tests... yes. I had two of each. Though the difference was marginal, there are "additional features" that make the 900 "better" but the 800 has beefier capacitors, which, from what I have seen on the interwebs, is confirmed (most places rate the 900 @ gn34m, 800 @ gn38m)