Nikon SB 800 vs Sekonic L558

[Chan Tran]

You're missing the point, Chan. I fully agree that the manufacturers claimed output is overstated. That's what I said in my original post.

The bottom line is that it doesn't really matter for practical purposes.

There are really only three ways to determine exposure:

1) A handheld flash meter (like the Sekonic)
2) A flash metering system in the camera and/or flash
3) A manual calculation based on actual flash output.

Knowing the actual flash output is only useful in method 3, where you need to do a manual calculation to determine exposure, based on camera to subject distance. In your example with a ceiling bounce, you could use method 1 with a handheld flash meter at the subject to determine exposure. You could also rely on the reflected light measurement from the camera/flash system, as in method 2. Both methods 1 and 2 should indicate underexposure if there's not enough light. Method 2 would be fast and easy, while method 1 could be somewhat inconvenient, depending on the situation.
That leaves method 3. You could take your known flash output, and then measure the distance from the flash to the ceiling. With that you'd have to estimate the reflectance value of the ceiling and determine how much light loss is at the ceiling. After that you'd have to measure the distance from the ceiling to the subject, and then calculate the amount of light hitting the subject, in order to determine the exposure. ..... several hours later you'd be ready to take a picture.

The bottom line is that for practical purposes, you either have to rely on the handheld meter or an in camera/flash metering system to determine exposure. Knowing the perfect guide number isn't going to help in any useful way, other than to tell you your range is limited with any flash.

It matters because I don't have enough flash power for what I want to do. I have to buy flash with higher GN rating than what I figure I would need. Besides I believe the OP doesn't have problem to get correct exposure he was just wondering why the flash put out less light than what he expected. Also there is the #4 way to determine exposure which is to estimate the power level needed in manual mode. When you get used to it you can just get it.

 

[Ardpatrick]

You're missing the point, Chan. I fully agree that the manufacturers claimed output is overstated. That's what I said in my original post.

The bottom line is that it doesn't really matter for practical purposes.

There are really only three ways to determine exposure:

1) A handheld flash meter (like the Sekonic)
2) A flash metering system in the camera and/or flash
3) A manual calculation based on actual flash output.

Knowing the actual flash output is only useful in method 3, where you need to do a manual calculation to determine exposure, based on camera to subject distance. In your example with a ceiling bounce, you could use method 1 with a handheld flash meter at the subject to determine exposure. You could also rely on the reflected light measurement from the camera/flash system, as in method 2. Both methods 1 and 2 should indicate underexposure if there's not enough light. Method 2 would be fast and easy, while method 1 could be somewhat inconvenient, depending on the situation.
That leaves method 3. You could take your known flash output, and then measure the distance from the flash to the ceiling. With that you'd have to estimate the reflectance value of the ceiling and determine how much light loss is at the ceiling. After that you'd have to measure the distance from the ceiling to the subject, and then calculate the amount of light hitting the subject, in order to determine the exposure. ..... several hours later you'd be ready to take a picture.

The bottom line is that for practical purposes, you either have to rely on the handheld meter or an in camera/flash metering system to determine exposure. Knowing the perfect guide number isn't going to help in any useful way, other than to tell you your range is limited with any flash.

As posted above, I tested the flash in manual mode because I wanted to verify the relationship between the information on the LCD display and the flash output without any interference from other controllers such as TTL or Auto metering. I posted here when the Flash LCD estimated output and the measured values didn’t align.

There was no intention of relying on GN calculations nor Manual mode for the photographic purposes I intend to use the unit for. I agree with you that’s a non-starter. I also agree that GN numbers are not much use.

I also agree that this discussion would be largely null if using a Dslr or using, as you say, an F5, one of which I happen to own. I need the unit to function in Auto mode, for use with a Mamiya 7 as a fill light. The Mamiya has leaf shutter lenses which adds value to the functionality of a flash. I can meter using the Sekonic but want something faster.

I’ll just have to test it when I get back to my darkroom and can run some film. I will report back here when I’ve got results.

 

[Sharktooth]

There's nothing wrong with your logic, it's just that manual mode for these flash units is just a throw-in, with little thought that they'd be used seriously. If they wanted to do this properly they'd have to add some sort of feedback system to the circuitry to account for changing light output over time. Obviously, this would be cost prohibitive to the target market. Most of the other camera flash makers don't even offer a manual mode, or even an A mode now. It's all some version of TTL flash metering.

For your Mamiya 7 you'd have to use A mode. I don't know how well that would work with fill flash. You might try experimenting with A mode and a digital camera. The flash would probably have to be mounted on an external mount, and connected to the camera with a plain PC cord. That would nullify any communication with the camera, except for triggering. You could see how the A mode works with fill flash, but not have to spend money on film.

 

[Chan Tran]

I made measurement with a number of flash units that I have and the result are as follow:
All are tested with the reflector at 35mm, Full power manual, distance of 10ft and Minolta flashmeter VI with flat diffuser. The meter readings are in EV for easy calculations.

Flash Unit	Nikon SB-800 (a)	Nikon SB-800 (b)	Nikon SB-900	Metz 45-CL4 (a)	Metz 45-CL4 (b)	Minolta Auto 320x	Metz 60-CT4
Meter Reading	6.6​	6.4​	6​	7.3​	7.1​	6.1​	8.2​
Rated GN	125​	125​	111​	148​	148​	104​	197​
Actual GN	98​	92​	80​	126​	117​	83​	171​
Delta EV	-0.69​	-0.89​	-0.94​	-0.48​	-0.68​	-0.66​	-0.40​

 

[koraks]

te capacitor of a 20-year-old flashgun is likely to have decreased in power. One way to full capacity is 'exercise it multiple times at full power and test it again.

Capacitors don't age quite like rechargeable batteries. I don't expect the flash capacitor in an SB800 to have lost any significant amount of capacity. Moreover, exercising the flash isn't going to change anything about the capacitor's real/actual capacitance (you're probably thinking of reforming a capacitor, but in OP's use case that would have happened by now). I've got loads of electrolytic capacitors, some many decades old, some driven much harder (i.e. high ripple currents for long periods of time) and they've not lost capacitance. What does happen, is that they sometimes start to leak (physically or electrically).

Moreover, if your hypothesis was correct, the problem would only manifest at the highest power settings and not the lower level. Flash power control works by cutting power to the flash tube prematurely; if the capacitor would have been the problem, the flash would effectively fizzle out too early on its highest power settings, but would still work fine at e.g. 1/4 and lower. The result would have been normal behavior at low levels and strong underexposure at high levels, not a consistent low output across its power range.

Reduction of effective flash output can occur due to deposition of vaporized metal along the inside of a flash tube; however, I would not expect a typical SB800 to have lost as much as a stop over its lifetime unless it was used really, really hard. Visual inspection of the actual flash tube would reveal the problem if there is any. This will require removing the transparent shielding and this may be beyond what OP is willing to do in terms of work on the flash.

Honestly, the most logical explanation is a discrepancy between measuring techniques, assumptions underlying GN calculation etc. combined with a slight degree of optimism in Nikon's rating of these units as evidenced by @Chan Tran's excellent post above. Between these different factors, I'm not surprised that the net result turns out to be a 1-stop fudge factor between what the Sekonic says and the Nikon promises. They're probably both correct within their own set of assumptions...

 

[Chan Tran]

I do not think age has to do with flash unit having less power than specs. My Nikon SB-900 is the newest and it measured worst. The SB-800 that measured higher is the older unit.

 

[koraks]

I do not think age has to do with flash unit having less power than specs.

Neither do I, really.

 

[bunip]

What the Op is reporting is similar to what happened to me. I owned for 5 years a Nikon sb800 unit that was spot on the exposure. It was purchased used from a camera shop locally. 4 years ago I left it as a present to my syster and got another one used from the bay few moths later. The newly arrived was off of almost 1 stop regardless of the mode employed (TTL, A, Etc).

 

[Chan Tran]

I found exposure on my SB-800 in auto mode are fine as long as it doesn't need the maximum power.

 

[forest bagger]

As posted above, I tested the flash in manual mode because I wanted to verify the relationship between the information on the LCD display and the flash output without any interference from other controllers such as TTL or Auto metering.

I don't understand exactly, what you mean with manual mode, until now.
Do you mean the 1/1 called fullpower, or the 1/2, 1/4, 1/8 and so on 'till 1/128?
As far as I know the SB-800 (nearly by heart, I repaired many of them) the output power of the flash tube is being measured by a photocell in the body which gets its light via a light guide from the diffuser in the flash head, what can work only for less power than full power.

 

[Chan Tran]

I don't understand exactly, what you mean with manual mode, until now.
Do you mean the 1/1 called fullpower, or the 1/2, 1/4, 1/8 and so on 'till 1/128?
As far as I know the SB-800 (nearly by heart, I repaired many of them) the output power of the flash tube is being measured by a photocell in the body which gets its light via a light guide from the diffuser in the flash head, what can work only for less power than full power.

I understand that the OP did this. For example he set the flash on manual, flash head straight forward, 1/2 power, at 11ft. The aperture should be f/8. He would then put the flash at 11ft from his Sekonic meter and made the meausrement. He found the meter said something less than f/8.

 

[_T_]

But if you’re using the flash in manual mode you have to meter it anyway. So it doesn’t really matter what the lcd readout claims. If it meters a stop low bump it up a stop. Imaginary problem.

 

[Chan Tran]

But if you’re using the flash in manual mode you have to meter it anyway. So it doesn’t really matter what the lcd readout claims. If it meters a stop low bump it up a stop. Imaginary problem.

You don't get it. The point is not about getting to correct exposure or not. The point is that manufacturers claim that their flashes are more powerful than they actually are. You pay good money for more powerful flashes. For example my Nikon SB-900 claims a GN of 111, ISO100, 35mm it only delivers 80.

 

[_T_]

Is it not powerful enough for you? Cause you can return it or sell it and get a more powerful flash.

 

[Chan Tran]

Is it not powerful enough for you? Cause you can return it or sell it and get a more powerful flash.

Even if it meets the spec it's not powerful enough for me but it's not the point. The point is that they publish inflated specs. The OP was concerned perhaps his meter is inaccurate or his flash unit is defective but my experience assure him that is not the case. The case that most flash manufacturers inflate their GN rating and it's the norm that the measured GN is less than the published GN.

 

[_T_]

Right but op says it has enough power for them. And the discrepancies they’re seeing won’t affect exposure of actual photos in a real world workflow so there’s no problem to be solved beyond the injustice of corporate marketing and we can’t do anything about that.

It is good to be aware of when purchasing lights though but if you understand how guide numbers work then you should already be aware that your personal guide numbers will necessarily differ from those of the manufacturer. A little late to be helpful knowledge for op at this point though. But luckily it ended up being a non issue

 

[Ardpatrick]

Apologies to come to this late. I was on a long haul flight with travel on either side & jet lag. Somehow Photrio hasn’t notified me of thread updates in some time.

Sorry for anyone having to guess my original set-up. Chan nailed it exactly and I have just taken the same photo of the rear LCD display in prep to respond before seeing the post above!

It’s not an “imaginary problem” if one gets a new to me piece of kit (SB800) with an LCD display that is way inaccurate. I had no way of verifying that that inaccuracy was limited to Manual mode LCD display and not also an issue of under exposure in auto mode too. Annoyingly it’s not a consistent issue but varies with different output levels so also not easy to factor for.

Agreement from one piece of gear to another is far from a given. Apart from the Sekonic I have several other hand held light meters - they vary quite a bit in readings from one another. Certainly over a stop from the most to the least sensitive. So I don’t know that the ‘auto’ mode on this SB800 is at all accurate to what my Sekonic indicates is correct exposure. With a quite divergent LCD, the original query was simply looking to get perspective on what those basic tests were showing as widely divergence.

I have learned from this thread that the LCD is probably no more than a digital version of the old Blue / Red range slider indicator of my old Vivitar Auto Thyristor I.e. an exposure guide only. Of course it’s easy to bypass the LCD info now I’m home from a trip and have access to a Nikon Dslr that works in TTL with the flash, and access to a darkroom to process film tests. I got the flash whilst traveling with only the Mamiya 7 and only the light meter to guide me.

I’ve learned a lot, and I’ll post some results when I get some proper tests done.

 

[Sharktooth]

I have learned from this thread that the LCD is probably no more than a digital version of the old Blue / Red range slider indicator of my old Vivitar Auto Thyristor I.e. an exposure guide only. Of course it’s easy to bypass the LCD info now I’m home from a trip and have access to a Nikon Dslr that works in TTL with the flash, and access to a darkroom to process film tests. I got the flash whilst traveling with only the Mamiya 7 and only the light meter to guide me.

I’ve learned a lot, and I’ll post some results when I get some proper tests done.

Yes, in manual mode the LCD is only showing a guide table based on the guide number spec. There are no measurements being made of the light output. You can prove this to yourself by covering the flash with your hand while firing. The LCD display doesn't change.

When testing the flash with a DSLR, you should try it in "A" mode. This mode is just like the auto mode in the old Vivitars. The Nikon SB800 is one of the few modern flashes that still has a sensor in the flash, so it can still function in "A" mode without needing to use any sensors in the camera. When you use any of the TTL modes, those modes only use the sensors in the camera, so they only work with compatible cameras. The "A" mode should work on your Mamiya 7 just as well as it works on your DSLR in "A" mode, since "A" mode only relies on the sensor in the flash itself. Nikon gave us the "A" mode to allow the SB800 to function on older cameras that have no flash metering system in the camera itself. The flash has some autoexposure capability, even when the camera doesn't.

To test out "A" mode with a digital camera it would be best to use a camera that is not compatible with the Nikon flash system. Using a Pentax or Canon DSLR would ensure that only the "A" mode of the flash is being used.

 

[koraks]

There are no measurements being made of the light output. You can prove this to yourself by covering the flash with your hand while firing

If I understand correctly what @forest bagger explained above, there actually is a measurement being done, and it's on this bases that the unit decides when to cut short the flash to get 1/2, 1/4 etc. However, since the measurement is done within the flash unit itself (hence the light guide), whatever is in the way of the light path outside the flash unit would have little to no effect. On this basis, I'd expect the flashes to be at least mostly consistent, depending on how well/accurate this internal sensor system works.

 

[Sharktooth]

If I understand correctly what @forest bagger explained above, there actually is a measurement being done, and it's on this bases that the unit decides when to cut short the flash to get 1/2, 1/4 etc. However, since the measurement is done within the flash unit itself (hence the light guide), whatever is in the way of the light path outside the flash unit would have little to no effect. On this basis, I'd expect the flashes to be at least mostly consistent, depending on how well/accurate this internal sensor system works.

No, there's no measurements being done (in manual mode). When you set the output to fractional power levels, the flash is just using a preset cutoff time at each level. It's just using a table. In other words, in manual mode the flash fires at full output, and the flash duration time is set based on the fractional power selection. There's no feedback loop. If you hold your hand over the flash, nothing changes, since the flash doesn't measure anything.

Small camera mounted flash units are very simple devices. There is a flash that fires to create light, and the only thing that you can change is the time duration of the flash. With short durations the light level is low, while with long durations there is more light. In order to both create the light, and measure the light, you need to have some sort of sensor. That sensor is part of a feedback loop that feeds to a computer to cut the time. This isn't done in manual mode. The sensors in the camera or the flash are only measuring light in the TTL or auto modes, since that sensor input is needed in order for the computer system to determine when enough exposure has been made to cut of the flash. This is all being done at millionths of a second.

All this computational power is in the camera for TTL modes. The old "A" mode uses a primitive 1970's style computational technology in the flash. Adding "A" mode to a modern flash just adds cost, and that feature will almost never be used by most users. That's why modern flash units don't have this feature, or manual mode, but also makes these flash units entirely useless without a compatible camera.

 

[Sharktooth]

There's another crude way to test your Nikon SB800 directly on your Mamiya 7 in "A" mode, without wasting film.

Just mount the flash and set it to "A" mode, with no film in the camera. Set an appropriate film speed and aperture on the flash. Stand 15 feet from an indoor wall, and fire the camera. The flash should be very intense, since it will have to fire at close to full power. Now move to only three feet from the wall, and fire the camera. The light intensity from the flash should be much much lower. This will at least tell you that the sensor in the flash is working, and that the control system in the flash is adjusting the flash duration. If you use your handheld flash meter, and measure at the wall, both exposures should be close to the same f stop reading.

 

[koraks]

In other words, in manual mode the flash fires at full output, and the flash duration time is set based on the fractional power selection. There's no feedback loop.

Is that something you know, or is it something you presume based on the knowledge that 1980s thyristor flash units generally worked that way?

If you hold your hand over the flash, nothing changes, since the flash doesn't measure anything.

That only shows there's no feedback loop that relies on a measurement external to the flash unit. But please re-read what @forest bagger said about the internal sensor connected via a light guide to the diffusor head. It's very well imaginable that Nikon chose to use such a system to determine the cut-off point also in manual/fractional mode.

Reading the repair manual, this seems to confirm what @forest bagger says:

The SPD is connected through a fiber optic within the flash unit to the head. The receptor at the other end of the fiber ends at the fresnel/acrylic screen assembly, where it's in a position to measure the light refracted within these elements:

Furthermore, if you think about it, it makes good sense that Nikon would prefer to rely on measured light output to cut off the flash in manual exposure mode. The relationship between the times and fractional outputs are non-linear (even if you plot them on a log/log scale):

This implies that if they were to use a fixed time for each fractional output, component drift due to age, temperature etc. would potentially affect the accuracy of the manual output significantly. The photosensor-based approach would allow the actual light output to be controlled more accurately than a pure time-based approach as in the old-fashioned units.

Small camera mounted flash units are very simple devices.

Yeah, they are. Be careful extrapolating that "flash basics 101" lecture to more modern units. Sometimes they work differently than you might guess.

That sensor is part of a feedback loop that feeds to a computer to cut the time.

A technicality, but I don't think all of the units you describe used a 'computer' as such (i.e. a microcontroller, or an ASIC) as part of the feedback loop (not a 'feedback loop that feeds into...' - that doesn't make sense - another technicality, but OK). I think the more basic units used a simple analog integrator (essentially, charging a cap) and a user-set threshold for the cutoff. No computer; just analog servo control.

All this computational power is in the camera for TTL modes. The old "A" mode uses a primitive 1970's style computational technology in the flash. Adding "A" mode to a modern flash just adds cost, and that feature will almost never be used by most users. That's why modern flash units don't have this feature, or manual mode, but also makes these flash units entirely useless without a compatible camera.

The SB800 most certainly does have complex in-flash computation that allows it to autonomously control flash power based on non-camera inputs. Again, this can all be gleaned from the repair manual. I expect flash units from other brands like the Canon EX series are quite similar in this regard.

 

[forest bagger]

Be careful extrapolating that "flash basics 101" lecture to more modern units. Sometimes they work differently than you might guess.

Even the small pop up flash of modern cameras has this sensor in it.

 

[koraks]

Even the small pop up flash of modern cameras has this sensor in it.

Thanks for adding this; it makes perfect sense, too. The output curve of a flash bulb will have an asymmetric bell curve, and to control fractional output, a light integrator is the most obvious approach as it automatically compensates for all manner of non-linearities, either inherent or due to aging or environmental conditions.

 

[forest bagger]

I can imagine that the measured underexposing is due to dirt on the fixed diffuser of the flas head - dust and dirt on this diffuser gets heated by the infrared light of the flash and burns the acrylic diffuser surface to black.