Underexposure in TTL with Yongnuo 568EX and Nikon

Underexposure and flash Yongnuo 568EX

 

I just received a Yongnuo 568EX for my Nikon D80, and I did a couple of tests.

Unfortunately, all my pictures in "i-TTL" are underexposed, by about 1 or 2EV. So I have to compensate every time I use iTTL.

The manual mode seems to work normally (even though I'm slightly disappointed by the power of the flash).

And now I am seeing on the forums that others have faced the same problem.

 

So, my question: has anyone NOT encountered the problem using Nikon+568EX ? (with Canon it seems there's a workaround)

Has anyone more information (official, or not) about this issue? Does anybody know if *ALL* 568EX work like this?

I hesitate to send my flash back.

 

This is my first cobra flash, and I lack experience as well as points of comparision.

 

This is a picture I made at night with the 568EX (FC+0) and the D80 (M, ISO 320, 18mm, 1/160, f/3.5) in I-TTL :

http://baptisteautin.com/files/DSC_5478-yn-0.jpg

 

The same with the builtin flash of the D-80 :

http://www.baptisteautin.com/files/DSC_5475-builtin-0.jpg

 

With the 568EX (FC+1) :

http://www.baptisteautin.com/files/DSC_5480-yn-1.jpg

 

Thanks for your help.

Edited September 11, 2017 by baptiste_autin

Looks like an "all white" frame for the camera...

 

Question : what "metering mode" did you use ?

 

Remark : Full auto I-ttl will try to give you an "average grey" result unless you tell it not to ( hence the correction of up to +2 - +3) to make the "whites" "white again", this is not a function or feature of the flash, but a function or feature of the metering system in your camera ..

I would suggest to test on an easier frame whit lots of different colours whitin the frame, so that the camera stops "thinking" it needs to cut off the flash to prevent "over exposing" ..

Thanks for your answer!

 

My pictures were taken with the "matrix" metering mode. I just gave a try with the two other modes (central, spot) but unfortunately the results are not significantly different.

 

I will make more pictures including more colors, like you said, and I will report here the results, even if I don't quite understand what you mean by "average grey". When I look at the histogram of the first picture (the one with no compensation), 95% of the pixels are concentrated in the first half of values (the lowest), so I think it should be pretty easy for the camera to guess that the picture is underexposed...

I can't answer for the i-TTL underexposure, except to say that I've found i-TTL to be generally unreliable.

 

WRT being disappointed with the power of the flash. Are you comparing its real power with the claimed Guide Number? If so, then you will be disappointed. All flash manufacturers exaggerate the GN of their flashes by at least one stop. This is common practice that goes back years - don't ask me why.

 

Flash makers have obviously found a loophole in the ISO testing methodology that let's them get away with this hyperbole. Ignore the published Guide Numbers. Expect a modern top-of-the-range speedlight to deliver a GN of no more than 32 (100 ISO/metres) @ the 50mm zoom setting, and not much more at maximum zoom. In "HSS" mode the GN will be about another stop less.

 

FWIW, the 568 i-TTL shot is only slightly underexposed compared to the shot using the popup flash. The popup flash should be easily capable of lighting a small room at 320 ISO. Therefore it appears that the issue is with the i-TTL metering of the camera more than the flash. You may just have to live with applying flash compensation in i-TTL mode. Or learn to use the flash in manual mode.

 

You'll also get far better light quality if you "bounce" the flash off walls or ceiling.

Edited September 12, 2017 by rodeo_joe|1

I can't answer for the i-TTL underexposure, except to say that I've found i-TTL to be generally unreliable.

 

WRT being disappointed with the power of the flash. Are you comparing its real power with the claimed Guide Number? If so, then you will be disappointed. All flash manufacturers exaggerate the GN of their flashes by at least one stop. This is common practice that goes back years - don't ask me why.

 

Flash makers have obviously found a loophole in the ISO testing methodology that let's them get away with this hyperbole. Ignore the published Guide Numbers. Expect a modern top-of-the-range speedlight to deliver a GN of no more than 32 (100 ISO/metres) @ the 50mm zoom setting, and not much more at maximum zoom. In "HSS" mode the GN will be about another stop less.

 

FWIW, the 568 i-TTL shot is only slightly underexposed compared to the shot using the popup flash. The popup flash should be easily capable of lighting a small room at 320 ISO. Therefore it appears that the issue is with the i-TTL metering of the camera more than the flash. You may just have to live with applying flash compensation in i-TTL mode. Or learn to use the flash in manual mode.

 

You'll also get far better light quality if you "bounce" the flash off walls or ceiling.

 

Yup! Although the white wall tends to lead to underexposure but if I use the non TTL A mode on most of my flashes the results are much better.

Thanks for your answer!

- even if I don't quite understand what you mean by "average grey".

When I look at the histogram of the first picture (the one with no compensation), 95% of the pixels are concentrated in the first half of values (the lowest), so I think it should be pretty easy for the camera to guess that the picture is underexposed...

 

"Average Grey" is the "gray representation" of a "Caucasian skin colour" , camera metering systems are mostly tuned to represent that in a shot, so by using a grey card , or skin tone card for metering, the camera / flash combo will try to calculate optimal lighting ( sorry if my English is not that good..).

It generally means that the Camera will try to adjust to reprent that as its main optimal lighting, resulting in "underexposing all or mostly "white scenes" ( snow, whitewall etc.) and overexposing overall black scenes ( Black cat / dog , black wall etc.)

So you would have to compensate when overall black or white scenes by subtracting or adding 1 or to LV ..

 

Here is an explanation how you can use a gray card for optimising your metering in such circumstances.. :

"How to Use a Gray Card for Custom White Balance and Metering

 

Skip to the "How to use a gray card for metering" chapter if you are not interested in the part for getting the optimum colour.

 

BTW, Since you are now using a DSLR (and possibly shooting RAW / NEF , you can always adjust in post too , therefore not many ppl use Gray Cards anymore

Thank you all for your answers.

 

I made some new tests, avoiding white colors this time, and the pictures are still underexposed (as the histograms show it).

The only thing that comforts me is that I get the same result using the builtin flash of the D80 (my only point of comparison up to now), which probably means that my Yongnuo is not faulty (furthermore, it works well in M mode, and, in i-TTL mode, the flash output changes if the scene changes, so it is not stuck at a minimum value either).

 

So, well, I think I'll keep my Yongnuo... Maybe I'll even get another one, a 560 for example, to add more light, if needed.

And it's true I'm not using a very luminous lens (an AF-S Nikkor 18-70mm f/3.5-4.5G), so I guess I can't expect to get the same result as with a, say, 50mm f/1.8, for example)

 

Yet I would be curious to see how an SB-900 would perform in the same conditions, in terms of metering, as well as in terms of global power.

Thank you all for your answers.

 

I made some new tests, avoiding white colors this time, and the pictures are still underexposed (as the histograms show it).

The only thing that comforts me is that I get the same result using the builtin flash of the D80 (my only point of comparison up to now), which probably means that my Yongnuo is not faulty (furthermore, it works well in M mode, and, in i-TTL mode, the flash output changes if the scene changes, so it is not stuck at a minimum value either).

http://baptisteautin.com/files/DSC_5504_yn_0_histo.jpg

YN568 i-TTL (f=44mm, M, f/4.5, 1/160, ISO 320, spot)

 

http://baptisteautin.com/files/DSC_5512_builtin_0_histo.jpg

D80 built-in flash i-TTL (f=44mm, M, f/4.5, 1/160, ISO 320, spot)

 

So, well, I think I'll keep my Yongnuo... Maybe I'll even get another one, a 560 for example, to add more light, if needed.

And it's true I'm not using a very luminous lens (an AF-S Nikkor 18-70mm f/3.5-4.5G), so I guess I can't expect to get the same result as with a, say, 50mm f/1.8, for example)

 

Yet I would be curious to see how an SB-900 would perform in the same conditions, in terms of metering, as well as in terms of global power.

Yeah if you get a SB-900 please test. You may come to the conclusion that Nikon TTL flash tend to underexpose.

Spot metering on that red rucksack should give exactly the histograms and exposure shown.

 

The red channel peaks are at approximately 2/3rds of the histogram; exactly where I'd expect an 18% reflectance to end up.

 

I-TTL was designed to be used with matrix metering - flakey as it is. Using any other metering mode is asking for exposure errors. Especially spot metering, which requires you to point the spot at something you want to be rendered as a mid tone.

 

Incidentally YongNuo's 560 series speedlights have exactly the same power as an SB-910. And as an SB-900, 800 etc. Right back to an SB-24. Nikon haven't made their speedlights any more powerful in years. Only increased their complexity and cost while reducing backward compatibility with their camera bodies.

"Average Grey" is the "gray representation" of a "Caucasian skin colour"

 

- That's not how the 18% average grey reflectance 'standard' was arrived at, nor is metering caucasian skin tones its purpose.

 

The 18% reference originated in Kodak's processing laboratories, where thousands of negatives (mainly amateur snapshots) were analysed. It was discovered empirically that the statistically "average" scene had an overall reflectance of 18%, and this became the de facto standard to which future reflectance lightmeters were calibrated.

 

This shouldn't be (but is) confused with Zone V in Adams' Zone system, which if you do a bit of sensitometric calculation actually equates to 12.5% reflectance. A calculation that Adams himself obviously failed to do.

Spot metering on that red rucksack should give exactly the histograms and exposure shown.

 

The red channel peaks are at approximately 2/3rds of the histogram; exactly where I'd expect an 18% reflectance to end up.

 

I-TTL was designed to be used with matrix metering - flakey as it is. Using any other metering mode is asking for exposure errors. Especially spot metering, which requires you to point the spot at something you want to be rendered as a mid tone.

 

Incidentally YongNuo's 560 series speedlights have exactly the same power as an SB-910. And as an SB-900, 800 etc. Right back to an SB-24. Nikon haven't made their speedlights any more powerful in years. Only increased their complexity and cost while reducing backward compatibility with their camera bodies.

 

A case for not buying Nikon flashes.

"A case for not buying Nikon flashes."

- Not entirely. Just a case for not trusting I-TTL.

I use, and like, Nikon SB-25s a lot. If I want on-camera flash automation I use AA mode, which has almost never given me a bad exposure. And for off-camera use I nearly always use manual mode.

What I'd suggest to Baptiste is to make proper use of the tilt/swivel head of his new flash, and to experiment with bounce lighting. Reflecting the flash off a ceiling or wall gives 1) a much better quality of light, and 2) much more even lighting over a distance. I.e no "shot in a dark cave" effect.

I experimented with using a camera-mounted flash at ISO 320, and found that only 1/8th power was needed to bounce-light a corner of a room. 1/4 power would easily light a 20ft living room, and provided the right spot on the 8ft high ceiling was chosen to "bounce" from, the entire room could be evenly lit.

Switching to AA mode I didn't even have to experiment with manual power settings. The flashgun just regulated its own output to give perfect exposures.

BTW, I have I-TTL capable flashes, but after a few poor I-TTL exposures I've abandoned it in favour of AA or fully manual flash control. I just don't trust I-TTL I'm afraid.

"A case for not buying Nikon flashes."

 

- Not entirely. Just a case for not trusting I-TTL.

 

I use, and like, Nikon SB-25s a lot. If I want on-camera flash automation I use AA mode, which has almost never given me a bad exposure. And for off-camera use I nearly always use manual mode.

 

What I'd suggest to Baptiste is to make proper use of the tilt/swivel head of his new flash, and to experiment with bounce lighting. Reflecting the flash off a ceiling or wall gives 1) a much better quality of light, and 2) much more even lighting over a distance. I.e no "shot in a dark cave" effect.

 

I experimented with using a camera-mounted flash at ISO 320, and found that only 1/8th power was needed to bounce-light a corner of a room. 1/4 power would easily light a 20ft living room, and provided the right spot on the 8ft high ceiling was chosen to "bounce" from, the entire room could be evenly lit.

 

Switching to AA mode I didn't even have to experiment with manual power settings. The flashgun just regulated its own output to give perfect exposures.

 

BTW, I have I-TTL capable flashes, but after a few poor I-TTL exposures I've abandoned it in favour of AA or fully manual flash control. I just don't trust I-TTL I'm afraid.

 

I mean not buying the expensive greatest and latest like the SB-5000 and SB-910. They are not better the old SB-24.

The SB-24 has a more limited manual power range than the SB-25, while the SB-28 has a worse user interface.

 

So for me the SB-25 provides all I need in a speedlight. However, they're not that cheap used if you buy online, and YongNuo and Godox make some very good flashes for not much more brand new.

I don't generally use the manual mode on the flash with the flash mounted on the camera. I use it for a lighting setup with the flash on stand or tripod in which case I use my Metz 60-CT4 which has 27 steps manual power.

I've never counted the manual power steps on an SB-25, but it goes down to 1/64th power, with 1/3rd stop decrements/increments between 1/2 power and 1/64th.

 

Let's see. That's 17 power levels. Enough for most purposes I would have thought. And more than are offered by many monolight studio strobes.

 

But I believe the 60-CT-4 "only" goes from full to 1/256th power. Even with 1/3rd stop steps all the way, that's just 25 power settings.

Let's see. That's 17 power levels. Enough for most purposes I would have thought. And more than are offered by many monolight studio strobes.

 

That sounds absolutely heavenly compared to my Normans.

 

It's usually easier to just physically move a particular head to adjust the light output rather than messing around with turning down the power. In fact, the guy I bought my kit from has an 800 W-S power pack along with a few extra heads, and I'm considering picking that up just for the sake of starting off with less power. Due to the way they work, having more heads that are not actually participating in lighting the subject are also a viable(if wasteful) way of adjusting power.

"It's usually easier to just physically move a particular head to adjust the light output rather than messing around with turning down the power."

 

How big is your studio space Ben? You need to move a light 1.4 times further distant to lose a stop of light. And besides, the quality of that light will become much harder, with shadows acquiring sharper outlines and falloff to the background becoming less.

The 60 CT4 does have 27 steps and in 1/3 stop. However, they are not accurate. They are not linear,

"However, they are not accurate. They are not linear,"

 

- That's the same with every flash I've metered that offers 1/3rd stop decrements. The whole stop settings are usually pretty close, but not the 1/3rd stop steps. For example; you might find that 1/4 + 0.3 is closer to 1/4 + 0.5, and that 1/2 - 0.3 is more like 1/2 - 0.2. But since you're still within 1/6th of a stop of the intended exposure it really isn't a problem.

 

It's quite difficult to regulate flash durations to the accuracy required for 1/3rd stop variations. The exponential rise and decay of the flash pulse isn't exactly predictable from sample-to-sample of speedlight and with varying temperature.

 

I suspect that this is the reason for some of i-TTL's vaguaries. The pre-flash is of fixed duration, and presumed to be of fixed light energy. Then some algorithm calculates a different duration for the actual exposure, which is presumed -possibly incorrectly- to have a certain relationship to the energy of the pre-flash.

 

Auto-aperture mode, OTOH, makes no presumptions, and simply cuts off the flash when a sensor detects sufficient flash energy has been delivered. A simpler and more reliable solution IME.

Edited September 25, 2017 by rodeo_joe|1

Average Grey" is the "gray representation" of a "Caucasian skin colour"

The average Caucasian skin reflectance is 36%, twice that of the "normal" 13-18% grey reflectance.

 

The metering system is responding to the overwhelming presence of white (or near white) in the field of view. Matrix metering is pretty intelligent these days, and can recognize faces and distribution of light and dark in the image. I don't think this applies to flash, however.

 

It could be a problem with the lens too. if it underexposes the same scene without flash. Nikon measures exposure with the aperture wide open. The maximum aperture and setting are transmitted to the camera in order to calculate the proper exposure. However the coupling is mechanical, and rather fragile. Any deviation between the setting and the actual aperture when closed will affect the exposure.

 

Even if everything is working correctly, the best way to establish the exposure for portraits, flash or not, is with an incident light meter and manual mode. Despite all the promises of TTL flash, especially multiple flash units, it still involves a lot of trial and error, and changes for each shooting situation. With a meter it's once and done, including measuring lighting ratios.

One of the problems with the 568EX, and newer Nikon flashes, is that it defaults to i-TTL- BL - where it tries to balance the lighting of the main subject as well as the background (while not overexposing the main subject). Often the latter fails in my experience, leaving everything underexposed. Typically, switching to spot metering solves this as it will cause the flash to act as normal i-TTL (which is more reliable as i-TTL-BL but still not the last word in consistency and reliability).

If/when you can, using manual just works better, as the others said.