Home · Register · Join Upload & Sell

Moderated by: Fred Miranda
Username  

  New fredmiranda.com Mobile Site
  New Feature: SMS Notification alert
  New Feature: Buy & Sell Watchlist
  

FM Forums | Lighting & Studio Techniques | Join Upload & Sell

  

Archive 2012 · underexposure using shoot throughs

  
 
michael kilner
Offline
• • • • •
Upload & Sell: Off
p.1 #1 · p.1 #1 · underexposure using shoot throughs


Hi all,as the title suggests I have been having some problems with exposure using shoot thro umbrellas.When the speedlight is hotshoe mounted,the exposure is correct although the bare light is a little harsh,when the speedlight is off camera and fired using a shoot through I find I have to use FEC +1 or thereabouts to get a good exposure.All settings are the same except of course when the speedlight is off camera and in slave mode the zoom defaults to 24mm,also everything is in ETTL,many thanks M.Kilner.BTW the underexposure is constant through various settings ISOs etc


Jun 29, 2012 at 10:07 AM
Mark_L
Offline
• • • • •
Upload & Sell: Off
p.1 #2 · p.1 #2 · underexposure using shoot throughs


That is the nature of TTL, put the subject in front of a dark background and you'll probably have the opposite problem. With TTL off camera you end up messing about with compensation and the flash power can vary shot to shot depending on framing, it is far similar and more consistent to set the flash power manually.


Jun 29, 2012 at 11:06 AM
cgardner
Offline
• • • • •
Upload & Sell: Off
p.1 #3 · p.1 #3 · underexposure using shoot throughs


The difference is due to how the ETTL metering works. When flash is used directly on camera the metering uses focus distance information from the lens to compute flash power by guide number and it tends to be very accurate.

When the flash isn't in the hotshoe and level it relies entirely on interpretation of the metering zones and the contrast of scene (which is less when flash is diffused or bounced) and the fact that unlike direct flash many of the photons the flash is outputting never make it into the viewfinder to be measured. You might think the metering seeing less light when would compensate at FEC=0 and output more to keep exposure the same, but it can't know how much of the flash output at FEC=0 you've wasted lighting parts of the room not seen in the viewfinder.

The footprint of direct flash is known factor to the flash. Even when the direct flash is zoomed the metering knows the how the footprint changes and changes the GN accordingly. But the metering can't know and take into account all the light you've wasted by bouncing 1/2 of it in the direction opposite what the metering sees in the viewfinder. Thus you need to take a shot at FEC=0, let the metering make it's best guess, then adjust manually with FEC based on the results you see FEC=0 produce.

Even with direct flash if you go from shooting an average scene at FEC = 0 to one that contains a bride in a white dress that fills the viewfinder the metering seeing more light coming back cuts the % of output at FEC=0 and the dress winds up under exposed. To correctly expose the dress you need to dial in a + FEC correction. That action is somewhat counter intuitive because you'd think since there more light reflected there would be over-exposure but you need to realize ETTL flash metering will react opposite. Seeing more pre-flash than average reflecting it will cut back the main exposure.

The simplest way to monitor and adjust exposure is to set your camera so the highlight clipping warning shows in the playback. Take a shot, see what if anything is clipping in it. Adjust FEC higher until you see areas you want correctly exposed in the highlights start clipping then back off 1/3 stop until the clipping warning disappears on camera.

How the RAW files will clip will differ from the JPGs so you'll need to cross-check the two the first few times you try that method. Shoot a scene with several bracketed FEC exposures to the point it is showing clipping on the camera playback then look at the RAW files. Find the one that has the best highlight detail, then go back and look at the same frame in the camera. That trains your brain to recognize via what is clipping in the camera display when RAW exposure you can't evaluate in camera is optimal.




Jun 29, 2012 at 03:14 PM
BrianO
Offline
• • • • •
Upload & Sell: Off
p.1 #4 · p.1 #4 · underexposure using shoot throughs


I've observed the same thing; one would think through the lens metering is always accurate since it's measuring the actual light, but the algorithms seem to be confused by off-camera flash through modifiers. I usually use Manual mode flash power and either do a few test shots to dial in my power levels, or use a flash meter.


Jun 29, 2012 at 03:15 PM
michael kilner
Offline
• • • • •
Upload & Sell: Off
p.1 #5 · p.1 #5 · underexposure using shoot throughs


thank you all for your answers,think a flash meter may be in order


Jun 30, 2012 at 04:48 AM
BrianO
Offline
• • • • •
Upload & Sell: Off
p.1 #6 · p.1 #6 · underexposure using shoot throughs


michael kilner wrote:
thank you all for your answers,think a flash meter may be in order


In the digital age, what with histograms and instant playback, a lot of people think flash meters are an anachronism, but I find them to be very helpful in getting the light right in the quickest and most accurate way possible.

I use a Sekonic L-758DR, but I know others who use the L-358 with good results, and it's quite a bit less expensive unless you add all the accessories.

Edited on Jun 30, 2012 at 10:32 AM · View previous versions



Jun 30, 2012 at 09:43 AM
sherijohnson
Offline
• • • •
Upload & Sell: Off
p.1 #7 · p.1 #7 · underexposure using shoot throughs


michael kilner wrote:
thank you all for your answers,think a flash meter may be in order


this was my first thought when I read your post



Jun 30, 2012 at 10:19 AM
dmacmillan
Offline
• • • • •
Upload & Sell: On
p.1 #8 · p.1 #8 · underexposure using shoot throughs


I'm putting on my old fart hat. I've been using flash meters since the late '60's/early '70's. They can be an invaluable tool. The suggestion to obtain a flash meter comes up on FM fairly frequently. I have seen some take the advice, go out and spend a wad on a fancy Sekonic, then post new photos that are no better technically or aesthetically. Why?

A flash meter is not a cure all, especially now when there are other ways to determine basic exposure. What's missing is the technical understanding that was a requirement in the film days. You need to know what you're measuring and why. To make matters worse, there's a lot of bad information being posted on the internet.

Yes, a flash meter is useful, but you have to understand how to use it correctly. The irony is the closer you come to knowing when and why, the less you really have to rely on it.



Jun 30, 2012 at 12:33 PM
cgardner
Offline
• • • • •
Upload & Sell: Off
p.1 #9 · p.1 #9 · underexposure using shoot throughs


Old dogs can learn new tricks, which is what I did when buying into the Canon flash system in 2005.

For starters, incident metering isn't an option with Canon ETTL or multi-flash optical triggering because the pre-flashes bias the meter readings. Even with multiple flashes in M mode there are command pre-flashes and the lights can't be fired and read separately for accurate ratio readings. So I learned how to use and came to rely on the other forms of metering a digital camera provides: the histogram and the highlight clipping warning, and the playback. Even on the old days with film a prudent photographer using a meter would still bracket exposures and in the studio use a Polariod back to take test shots to confirm exposure and lighting ratio. The digital feedback provides the same information, so why not learn to use it?

Unlike some giving advice here I also currently have a set of studio lights. When buying them I followed conventional wisdom and purchased a Sekonic L-358 Incident flash meter for setting ratios and exposure. But what I found myself doing after setting the ratio and exposure with the meter was the same thing I'd been doing for five years at that point on my digital cameras with Canon and Vivitar speedlights: take a test shot, evaluate it with the playback, histogram and clipping warning and tweek the settings based I what I saw, as needed. It didn't take long to realize the incident metering step was redundant in my workflow and it is was faster to just take a few test shots instead. In retrospect the money spend for the meter would have had better application if used for more lights, modifiers, or other camera gear.

With regard to measuring ratios with a meter by the numbers, I think that's also an old fashioned concept many cling to that really isn't necessary in the digital age. It doesn't address the goals of the exercise or help you understand why you'd use a 2:1 versus 5:1 ratio. I came to realize that by helping beginners on the Internet who always seemed to start with a photo of the wife with a 5:1 ratio and "Rembrandt" pattern that puts the face half in deep shadows. Mind you that's a wonderful strategy for shooting your old and wise 87 year-old grandpa, not not the most flattering choice for your wife. They were blindly follow some lighting diagram in a book without stopping to think whether or not that combination was flattering for the wife.

As a manager at work I always led teams by setting goals and defining measurable criteria for success. In the case of teaching lighting ratio on my web site I use the goal of creating "flattering" lighting as the goal of the exercise, not learning patterns or ratios by rote, the way I learned and the way most still teach. What constitutes "flattering" lighting? That's for you to decide yourself, but unless you have thought about the critieria that define it it's less likely you'll be able to tell yourself whether or not you have gotten it.

Not all lighting must be flattering, but the criteria that define flattering light for me are easy to define and understand: Get light in the eyes (preferably both of them) and avoid harsh dark shadows. A 5:1 Rembrandt pattern judged by those critera isn't flattering and thus not a good choice if the goal is flattering the wife. Flattering the wife is always a good strategy so she thinks the investment in the lighting gear was a good thing.

Teaching lighting by ratio numbers is like building a house by handing blueprint and a ruler to someone who has never seen a house. Follow the blueprint and you can create the house. If you set the key light incident strength one stop above fill you get 2Key + 1 Fill over 1 Fill which reflects 3x more light from the highlights where the two overlap. Then you stand back, look at the results and know for the first time what a 3:1 ratio looks like. That was a good approach back in the days when you shot on film and didn't see the results until days later. Nowadays you can just look at the back of the camera an see if it is working or not. How do you know it's working? One way is to have a clear goal before shooting, such as flattering the subject, and a few simple criteria like light in the eyes and shadow tone to measure the degree of success or failure.

Once you master the basics of "flattering" expand the goals to include "happy", "thoughtful", "sad","sullen","mad", "hostile" — the range of human emotions projected in portraits — then for each define criteria that create that reaction in the mind of the viewer. As the mood gets darker you'll want darker shadows, clothing and backgrounds and shaded eyes that make eye contact more difficult.

The new way to approach learning lighting was made possible by the instant feedback of digital and the fact you aren't paying 50 cents a shot for film and development is to put the wife in front of the camera with a fill light centered around chin level and raise it's power until the shadows on her face and clothing are light and flattering, then turn on the key light and raise it's power until the highlights it creates over the fill are correctly exposed. All it takes is a few test shots while adjusting the power of the light. It's really that simple. What will the ratio be numerically? Does it really matter? The goal is flattering the wife and you will have done that simply and emperically with just your eyeballs and the feedback a digital camera provides.

If photographing your 87 year-old grandpa what would you do differently? If you wanted to make him look old and wise you'd simply start the process with less fill so the shadows were darker. Try that approach and you'll quickly grasp the concept that shadows control the mood and emotional reaction to lighting on faces, and that fill intensity controls the tone of the shadows.

So with all due respect to the others instead of spending your money on a meter, I'd suggest you invest the money initially in modifers which will allow you to better control where the light goes and how much gets spilled, then try the approach I suggest of starting by adjusting the fill level first. After trying it that way on various genders and ages subjects to create happy and engaged or detached and sullen looks with the lighting ratio via modulation of the fill you'll have a better idea of whether to spend the money on the meter not.

Edited on Jun 30, 2012 at 03:30 PM · View previous versions



Jun 30, 2012 at 02:19 PM
michael kilner
Offline
• • • • •
Upload & Sell: Off
p.1 #10 · p.1 #10 · underexposure using shoot throughs


excellent advice,excellent forum full of people willing to help,many thanks


Jun 30, 2012 at 02:52 PM
Mark_L
Offline
• • • • •
Upload & Sell: Off
p.1 #11 · p.1 #11 · underexposure using shoot throughs


The meter vs histogram debate is like canon vs nikon. Either will work but one will work without trial and error and also give you ratios between main/fill and flash/ambient the other is a readout of total exposure based on jpg data.


Jun 30, 2012 at 04:45 PM
BrianO
Offline
• • • • •
Upload & Sell: Off
p.1 #12 · p.1 #12 · underexposure using shoot throughs


cgardner wrote:
...incident metering isn't an option with Canon ETTL or multi-flash optical triggering because the pre-flashes bias the meter readings. Even with multiple flashes in M mode there are command pre-flashes and the lights can't be fired and read separately for accurate ratio readings.


I don't have the desire to teach you how it's done, Chuck, but I'll just say for any readers who are new to flash meters that the above statement is absolute rubbish.

You can meter ETTL flash and multi-flash if you know how. The "how" can be found by a bit of reading and some thought into how Canon cameras work.

(Why one would want to use ETTL flash when one has a meter and the time to use it is another question.)



Jun 30, 2012 at 06:11 PM
ravisrajan
Offline
• •
Upload & Sell: Off
p.1 #13 · p.1 #13 · underexposure using shoot throughs


michael kilner wrote:
excellent advice,excellent forum full of people willing to help,many thanks


+1, Mainly for Chuck for your willing to help and teach how beginer should start learning lighting.



Jul 01, 2012 at 11:19 AM
wilt
Offline
• • •
Upload & Sell: Off
p.1 #14 · p.1 #14 · underexposure using shoot throughs


Oddly, past testing of Canon ETTL and Metz ETTL flash on various models of Canon dSLR by myself and a moderator on another dSLR forum has shown an inability of ETTL to properly expose flash shots when the ETTL flash has a small softbox over the flash lens! One would think that ETTL metering would detect the reduction in light output measured returning from the scene, and ask the flash to output more light, but it doesn't.

Film camera TTL flash exposure adjusts itself suitably, but not ETTL. On camera, it doesn't expose correctly. With softbox, with flash head leveled or with flash head elevated a few degrees (so that distance information from the lens is not factored), the flash underexposes. (In fact, under some unidentified combination of circumstances, ETTL flash will overexpose rather than underexpose...mysterious.) Yet, without softbox, ETTL exposes correctly with flash head aimed to bounce the light on the ceiling!



Jul 15, 2012 at 03:24 PM
BrianO
Offline
• • • • •
Upload & Sell: Off
p.1 #15 · p.1 #15 · underexposure using shoot throughs


wilt wrote:
...Film camera TTL flash exposure adjusts itself suitably, but not ETTL.


Yep. With film cameras, the built-in flash meter was measuring the actual exposure flash as it reflected off the film. With digital cameras that doesn't work (I don't know why not), so instead the meter reads a pre-flash that is of less power than the exposure flash.

My assumption for why modifiers mess up the metering is that the computer isn't able to determine by how much the modifier will reduce the output, and that the difference in output between the lower-powered pre-flash and the full-powered exposure flash is not changed by equal amounts.

When using modifiers I like to use Manual flash mode, and either a flash meter or test shots using the histogram for adjustment guidance.



Jul 15, 2012 at 05:49 PM
cgardner
Offline
• • • • •
Upload & Sell: Off
p.1 #16 · p.1 #16 · underexposure using shoot throughs


A-TTL metering on film cameras metered off the film during the actual exposure, turning off the flash when it saw enough light. Digital sensors required moving metering to the viewfinder before the exposure via pre-flash.

The first iteration of digital flash was called E-TTL, and the flash display still says ETTL, but since the 20D and 5D flash metering is E-TTLII which uses zone evaluation algorithms and with single direct flash the lens distance info.

I did a lot of systematic testing when first getting my 580ex flashes, documenting the results on my web site. I have no way of knowing with certainly but I strongly suspect from use that when lens distance information is available for flash computation it trumps the evaluative metering.

I mostly use direct flash outdoors and noted from testing in backlight in Av mode I that after I exposed for the sunny highlights below clipping (which requires -2 EC) FEC=0 balanced the front side nearly perfectly most of the time. If I didn't reduce the ambient by - 2 EC FEC=0 underexposed the shaded front.

Canon isn't straightforward about how the flash metering works but bits and pieces were found in white papers and web sites. I discovered that with E-TTLII the flash metering starts with the camera taking an ambient reading after full shutter press. The sequence is shown in this diagram from a Canon Europe site:

http://super.nova.org/MP/CanonETTL.jpg

If shooting in Av mode half press of the shutter locks AE — shutter speed at selected aperture. When the shutter is fully pressed the second ambient "map" is created. Pre-flash goes off, once for each group if multiple flashes are used. What the camera sees is the pre-flash and the ambient. The metering logic then subtracts the ambient map from the pre-flash map and winds up with a map of what in the viewfinder is reflecting the flash:
http://super.nova.org/MP/Eval_Ambient.jpg
http://super.nova.org/MP/Eval_PreAmb.jpg
http://super.nova.org/MP/Eval_PreFlash.jpg

But is that figure in the foreground in the center metering zones wearing a white or black dress? The metering doesn't know and that's why it's never more than an educated guess.

I shoot indoors with my DIY modifiers most of the time. The attach to the vertically oriented flash head. Adding them requires + 1-1/3 FEC to expose solid white "zone 9" highlights in an average scene. If the scene reflects more light than normal (person in white shirt) I'll need to increase FEC to expose the face optimally If the scene is darker than normal (person in dark suit) I'll need to decrease FEC to expose the face optimally.

That's normal with any TTL metering system. Those with more and smaller metering zones or new ones with face detection do a better job of guessing then those with fewer, larger zones. But I doubt DSLR metering will be hands-off automatic perfect in every shot until they invent a camera that can read minds.

An analogy is aiming a Canon. Fire the first shot. You can predict where it should land, but wind and other factors may affect actual results. With metering it's the scene reflectance variables. See where the first shot lands (where in the scene exposure is correct at FEC=0 or whatever setting usually works for a modifier) and then adjust from that baseline of actual results.



Jul 15, 2012 at 06:25 PM





FM Forums | Lighting & Studio Techniques | Join Upload & Sell

    
 

You are not logged in. Login or Register

Username       Or Reset password



This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.