Inconsistency between shots with flash at my last wedding

[Bluepoole]

I have a problem with consistency between shots with flash - to illustrate the problem, I'm using two samples from my last wedding shoot.

Pic #1:

Pic #2:

The photos were shot in RAW - for the purpose of this thread I did not perform any post processing on them, only resized and saved them in JPG.

As you can see, pic #2 looks way better than #1. The highlights are not blown out and the shadow of the boy is not so harsh.

More information: the two photos were taken seconds apart. Canon 1000D (XS), lens used was the EF-S 17-85mm, focal length was the same in both shots (22mm), flash used was the 430EX II, ETTL mode.

Program mode was used, in pic #1 the aperture was 4.5, in pic #2 it was 4.0 - the rest of the settings were exactly the same - shutter speed of 1/60, ISO 400, Centre weighted average metering mode, no FEC.

I know the framing is slightly different, but in my opinion not really so much to affect the exposure. I don't believe it was the camera's metering that caused the inconsistency, since the shutter speed and aperture combination is almost identical.

So why the inconsistency? Only aspect I can still think of is the flash output (strength). But why would it differ so much? I've read that the Speedlite also performs a metering function. If so, how can one control it more efficiently? When looking at the shooting data of photos in Zoombrowser, it is not possible to see the value of the flash output (as opposed to being able to see the flash exposure compensation) - isn't there a way to see this?

Would one necessarily get more consistent results by rather using the Speedlite in manual mode and adjusting the flash output yourself? During a wedding the pace does not really allow you to change flash output all the time.

I also get stronger flash output and more blown out highlights when my focal length is less than 35mm.

Any ideas/input or comments would be appreciated!

 

[BrandonS]

I would think with center weighted metering and the framing being the same the metering would end up being the same for the frame. Your aperture would indicate a brighter second image so it isn't that.

I would say that it is probably the flash not recycling fast enough for the amount of power it needed in the second shot. Go back through the shots you've taken and look at the ones with this anomaly in them and see if the problem doesn't occur on back to back shots.

I think the first TTL let you down, and the second you were saved by the flash not being able to power up to what the camera wanted it to flash at.

 

[SNAPaPHOTO]

I had this once in the Nikon world and my flash was not working in the TTL mode and thus over powering all flash shots. Replaced the flash and is well. In my case i will using a third party flash and that was the last time I bought third pary flashes. Now all we use is Nikon flashes.

Since your canon I don't exactly know but I would doubt the flash and its TTL mode.

Good luck trying to find it.

 

[Dao]

From looking at the background, the 1st photos seems just a little darker. That's the result of slight different aperture size. So background exposure is about the same. (close to)

However, the subject is a lot brighter in photo#1. It seems to be the result of flash. It appears that flash is not fired on the 2nd photo. You can see the shadow of the subject in photo#1, but not in photo#2.

I believe you did not wait long enough for the flash to charge up for the 2nd shot. And that is the reason for the inconsistency.

 

[Dao]

As for why the first photo is overexposed. I think you maybe standing too close to the subject. You take the shot in portrait mode, so the shoe mounted flash was on the left side of the camera. From the way the shadow look (angle) and the subject is a little distorted (top bigger, bottom smaller). You must be using a wide angle lens (18mm??) and stand very close to the subject.

 

[jchin]

Did you by chance, focus and recompose? Meaning, press the shutter half-way down to focus and then recomposed, followed by full-press of the shutter button?

It looks like it metered for something darker in the first image, that is why the flash was stronger.

In the second image, it re-metered against his white shirt and adjusted exposure accordingly.

 

[Big Mike]

I don't believe it was the camera's metering that caused the inconsistency, since the shutter speed and aperture combination is almost identical.

So why the inconsistency? Only aspect I can still think of is the flash output (strength). But why would it differ so much? I've read that the Speedlite also performs a metering function. If so, how can one control it more efficiently? When looking at the shooting data of photos in Zoombrowser, it is not possible to see the value of the flash output (as opposed to being able to see the flash exposure compensation) - isn't there a way to see this?

Flash metering and ambient light metering are two separate things. The shutter speed & aperture might be almost identical, which means that the ambient metering was consistent, but the flash is metered with a preflash. It should be consistent, if the reflectivity of your subject/scene is the same...but sometimes it's not that predictable.

E-TTL flash metering is better today than it used to be, but it's still a mystery to most of us.
Yes, you could get more consistency by using manual mode on the flash, but you are probably right that it's not really a practical solution during a fast paced wedding shoot.

My advice is to take advantage of the benefits of digital. Have a quick glance at the LCD to see the shot (and the histogram). If the flash exposure is off, then reshoot it, maybe changing the FEC.

 

[Bluepoole]

Thank you for all the replies so far!

I would think with center weighted metering and the framing being the same the metering would end up being the same for the frame. Your aperture would indicate a brighter second image so it isn't that.

I would say that it is probably the flash not recycling fast enough for the amount of power it needed in the second shot. Go back through the shots you've taken and look at the ones with this anomaly in them and see if the problem doesn't occur on back to back shots.

I think the first TTL let you down, and the second you were saved by the flash not being able to power up to what the camera wanted it to flash at.

I believe you did not wait long enough for the flash to charge up for the 2nd shot. And that is the reason for the inconsistency.

I think you guys are on the right track here...I went and checked the EXIF data and according to the time stamps, these two shots were literally taken a second apart...I'm not normally that trigger happy but must have taken them in succession without realising it that much.
But herein lies the answer perhaps...it once again, for me, comes down to flash output - so yes, it is difficult to adjust it repeatedly during the course of a wedding, but maybe in time one will learn what output works under what conditions...

You must be using a wide angle lens (18mm??) and stand very close to the subject.

Yes, I was using my 17-85mm lens. I do get better results when standing further away and using a focal length of 35mm or more - so I'll try to do this more often - it then of course also avoids the barrel distortion!

Did you by chance, focus and recompose? Meaning, press the shutter half-way down to focus and then recomposed, followed by full-press of the shutter button?

No, especially since the two frames were only a second apart, I suspect the reason for the different framing is that I just moved the camera slightly whilst taking the second shot. Also, I don't focus with the shutter button - I've changed the custom function and use the * (AE Lock) button to focus.

My advice is to take advantage of the benefits of digital. Have a quick glance at the LCD to see the shot (and the histogram). If the flash exposure is off, then reshoot it, maybe changing the FEC.

I think this is good advice, practise makes perfect. I've also found that not every single shot need flash. And when time permits, you could play around with flash output anyway. But I now have a problem though, because I use the * (AE Lock) button to focus, I can still achieve AE Lock with the shutter button, but the moment flash comes into play, I don't have the functionality of FE Lock anymore! The shutter button doesn't perform FE Lock and therefore a preflash is out of the question.

Any ways to have FE Lock functionality when using the camera's AE Lock button to focus?

 

[Big Mike]

Any ways to have FE Lock functionality when using the camera's AE Lock button to focus?

I don't think so, not with your cameras anyway. Some of the newer models, 40D, 50D etc, have an extra AF button up by the * (AE lock) button, so you wouldn't need to switch it's function.

I also use the * button to focus...but I never used FE lock.

The shutter button doesn't perform FE Lock and therefore a preflash is out of the question.

FE lock just separates the preflash, allowing you to recompose or wait for the shot. Any shot using E-TTL mode uses a preflash, it's just so quick that most people don't even notice that it fires just before the actual flash burst.

 

[Vinny]

In the first photo the background is a bit darker then the second - it may be just the 0.5 f stop difference. I know nothing about fully automatic cameras but is it possible that the flash didn't fire because there was enough light at the lower f stop?

Also, I want to add that years ago a higher end flash used to cut off the light output when it had enough (based on what you told it the aperture and distance you were at) and recycle faster since it didn't use all it's energy. I don't know if this flash has that capabilities or not but 1 second doesn't sound too unrealistic at the distance you were shooting from.

 

[SilverUser]

These flashing pictures are caused by user error, assuming the equipment is not faulty. A big part of the problem is that you're flashing in P mode. This is not a very good way to flash pictures if you want more quality then what most snappers want. Also the worse thing you can do is to place the metering in center weighted average mode. It is far better to place the body in evaluative mode so that the flash, the body, and E-TTL all conspire to get a perfect flash for the human skin, and using a manually chosen focus point is a must for (1) critical sharpness, and (2) accurate metering.

Here is what I would've done (and this method leverages heavily on Canon's EOS flashing logic E-TTL):

1. Place the flash in E-TTL mode, Hi-sync off.

2. Place the body in Av mode, set the f-stop for required depth-of-field, which I would set at near wide open since the subject should pop out of the too busy background.

3. Place body in evaluative metering mode.

4. Use the lowest ISO setting before the shutter gets too slow.

5. IMPORTANT STEP: Frame the shot, manually selecting the appropriate focus point, and in the case of the OP's pictures, I would choose the top FP. Place that chosen FP on the little boy's right eye, DO NOT RECOMPOSE.

6. Wait for him to give a nice expression/smile and take the shot.

Check the histogram; to brighter his face apply EC in 1/3+ stops to the FLASH. To darken his face, apply EC in 1/3- stops to the FLASH.

7. Retake as required.

There is a better way that involves placing the body in manual mode, but I will not go there in order to keep it simple.

If you do these steps I write here, you will get perfectly exposed human skin, and a great shot in general. Everytime, with 100% consistency, assuming you allow the flash to recharge per shot.

Now if anyone wants to know the better way, placing the body (not flash) in manual just let me know.

Hope this helps.

 

[Plato]

My thinking is that, for reasons unknown, the flash metering is incorrect resulting in overexposure of the first shot. Then, "several seconds later," the second flash burst puts out less light because there was insufficient time for flash recharge.