Welcome to the forum!

As an adjunct to the Tangents blog, the intention with this forum is to answer any questions, and allow a diverse discussion of topics related photography. With that, see it as an open invitation to just climb in and start threads and to respond to any threads.

Wedding Dance in Low Light

slravenphotoslravenphoto Member
edited April 2013 in wedding photography
This post is further to my question to Neil and his suggestion that I add the images here for discussion.

I recently attempted to photograph the first dance at a wedding, and unfortunately I had extremely disappointing results. The following is straight out of camera:


Canon 5D mk II / Canon 50mm 1.8 / Canon EX 580 II
1/100 @ f/1.8 @ ISO 4000
Manual metering / spot metering

The flash was flagged with a BFT and bounced behind me where the wall was fairly close. The couple were a fair distance from me.

I metered for the ambient light by switching off the flash, pointing the camera towards the back of the room, zeroing the meter then dialing down about 2 stops. Even with the 2 stops dialing down I had to use 4000 ISO because reducing the aperture or increasing the shutter any further would not have been acceptable. So, you can tell exactly how low the light was. 4000 is far from ideal, but I don't think I had much of a choice.

Anyway, I then switched the flash back on and used it as above with 0 FEC.

The resulting image is bad. It is extremely underexposed and I am disappointed that the high ISO/aperture combo did not allow sufficient flashlight to record a properly exposed and consistently lit image.

In fact, ideally I'd like to have reduced the ISO to at least 2000 and still achieve these results. Doing that, though would have caused the flash to be the ONLY source of light as the combination of settings would have let no ambient light in at all.

As a reference point, this is the image after having increased the exposure in aperture:


It's passable (just - and somewhat out of focus), but very grainy due to the combination of 4000 ISO and having increased the exposure post capture. Certainly, it's no match for getting it right in-camera!

1. Where am I going wrong? Has this anything to do with me spot metering?

2. Also, would I have benefitted from zone focus for this series of shots??

3. In addition, I can see no evidence of the flash having froze the subject. Any ideas?

My failure to get to grips with this kind of low-light photography is holding me back. I seem to understand everything Neil says here and in his book; I just fail to put it into practice when the light is so low. Where the light is better (but still low) I think I can make it work. Perhaps I am simply asking too much of my camera...?


  • ZenonZenon Member
    edited April 2013
    Well I'm not an expert and Neil will correct me if I'm off.

    1. You should not have used spot metering to try and meter a flash exposure. The cameras spot metering or any metering mode has nothing to do with the flash. You camera's light meter only meters ambient light. Even if your flash is attached to your camera it does not care what your camera is doing and your camera does not care what your flash is doing. You need separate those two things in your mind.

    Best to put the camera on maybe evaluative or average and meter for your background lighting. Typically about 2 stops under exposure is good.

    Then you set your flash power by either ETTL and FEC or on manual and select output power. You have a perfect white subject and your histogram is you best friend.

    It is like two exposures in one and the most important first steps to nail regarding on camera flash which you will.


    2. I would not use zone focus because the camera because it guesses on what you want in focus. Use single point and either one shot or AI Servo with moving subjects.

    3. The flash did not freeze because there was so little flash the ambient overpowered it. See link in reply question 1 which will also help with question 2 by freezing subjects so they will appear in focus.

    Had to rush. More to come. I gotta pick the boss up from work.
  • ZenonZenon Member
    edited April 2013
    OK part duex.

    You can select different metering modes for your flash. ETTL uses the same exposure metering system as the camera just not the camera's light meter. Think it as piggybacking. You can select Evaluative or Average and only through the cameras flash menus, not the flash menus.


    This mode isolates your subject/s by using multiple metering zones. When you press your shutter button the flash fires a pre flash. This pre flash reading is compared to the ambient reading and determines the closest object must be what you want to correctly expose which is usually your subject/s.

    Forget the flash for a moment. I'm going to go back to your camera's light meter. It exposes for 18% grey. Unless you compensate - if you shoot pure white snow it comes out grey, if you shoot pure black tar it comes out grey. An equal amount of white snow and black tar in the frame and you get a good exposure.

    Same goes for ETTL and flash. A bride alone in white dress, a groom alone in a black tux and the bride and groom together (even though were isolated by the pre flash) will all reflect light back differently. Here is were you will need to tweak the FEC to get the whites correct in that link I previously posted. Do that and the rest falls into place.


    Average does not isolate the subject/s, it averages out the entire scene however the white dress/black tux rule still applies. Because average is metering the entire area within the frame as opposed to your subjects only in Evaluative mode it said to be more accurate depending on the environment.

    Summary. Average is better for smaller venues like you posted and Evaluative is better for larger venues and outdoors. I did some tests but never really concluded anything. I don't have time to add this to my list of things to remember so I just shoot in Evaluative and use FEC as required.

    Spot metering

    There is a way to spot meter using your flash. The FEL button. FEL fires a pre flash but the actual exposure does not take place until you press the shutter. FEL meters only within the spot meter circle in your viewfinder. Same rules apply to flash spot metering as the camera's ambient spot metering. People like to meter off skin tones due to better predictability. I gave it a go but I didn't like it because people thought you took the picture with the FEL pre flash.

    So there you go. All of the above again is all flash related, nothing to do with your camera's light meter except the snow/tar example. This is why I love flash photography. So many interesting parts to it.
  • TrevTrev Moderator
    edited April 2013
    Zenon is right re metering, but those figures you had for ambient are crazy, was the room 'that' dark. ISO 4000 f1.8 @ 100th, wow.

    ISO 4000 for the MkII is really pushing it, but if that's what it takes then you have no option, however, if you do get correct exposure using those settings, your noise will be greatly reduced, since high ISO noise is much more preferable to under-exposure noise.

    Honestly the easiest approach is to just take a shot [no flash], see what the ambient looks like, adjust to taste, then use flash to augment the light.

    I walk into a room, take a quick guess [my default settings in receptions are ISO 1600 f4 100th] and if it's really dark without taking a shot I push up the ISO to around 2500+, maybe open to f2.8, then test, turn on flash and adjust to suit.

    The more the flash is dominant the more chance you have of freezing the action also, unless you get a really exaggerated movement for an artistic shot with static subjects in background.

    I did a quick edit on that little under-exposed jpeg, this is something I presume you would want to achieve roughly in camera I take it?

    Compare them both side by side in Photoshop and the ambient now balances the couple with a bit more separation.


  • slravenphotoslravenphoto Member
    edited April 2013
    Thanks for the tips.

    Zenon - very comprehensive, but maybe I didn't explain properly - I understand all that technical stuff about different metering options, and in particular your second paragraph is exactly what I did do. I metered for the ambient light and then stopped down by 2 (exactly as you said). I fully understand the two-exposure thing, I just have difficulty putting flash photography into practice when it is so dark like this; it does not come naturally to me and I am think I am a little impatient in waiting for it to become second nature. Where there is more light I get pretty good results. Which brings me on to Trev - indeed, those settings were required just to get an underexposure!


    1. What would have happened if I had increased my FEC? Would the couple have been overexposed?

    2. Does it make a difference that I was very close to the bounce surface?

  • ZenonZenon Member
    edited April 2013
    Sorry about that. I really enjoy flash photography and find the technical part fascinating.

    1. Your first image looks underexposed to me. If you use the histogram (and blinkies help as well) as described then you won't overexpose. Blinkies will tell you if some insignificant light in the background is blown out. The exposure of the subject/s is most important.

    2. Whatever you bounce off becomes your light source. The closer you are, the smaller it is but it is better than nothing.
  • This article has a lot of info that I think would be applicable to this situation:
    If you can position yourself further away from the area you are bouncing your flash the light source will be much larger. I experimented by taking a lot of pics of an object and changing only my distance from the wall I was bouncing my flash. Nothing scientific, but I could see that by actually being further from the wall, the light source was larger and was able to illuminate the object correctly. Also, increasing the FEC I could stop the lens down for more depth of field.
  • The original image is underexposed as you already mentioned. Personally I would have shot at iso 6400 and used manual flash instead of relying on the camera's internal metering guessing at what flash power it should use.

    In situations where I'm bouncing flash off of ceilings I find I get the most consistent results using strictly manual flash.

    So I would have exposed for ambient (set iso to 6400 or close to that so the backdrop isn't so underexposed) and then set the manual flash power to something that would make the subjects pop and once you nail the camera exposure (can be done within 2 shots) and the manual flash power (can also be done within 2 shots) you can focus on shooting and all of your photos will come out the same exposure wise (your subjects exposure and the ambient/backdrop exposure) as long as you're using the inverse square law to your advantage) in that dance hall.

    Also, next time if you don't want to use such a high iso think about using a light stand with a dedicated flash/battery pack or full powered strobe (for faster recycles/longer life) up on a light stand bouncing a high power level of flash off of the center of the ceiling as this will actually increase your ambient lighting allowing you to use a lower iso and you can do this while still bouncing a light amount of flash behind you. (using your on camera bounced flash)

  • Neil vNNeil vN Administrator
    edited April 2013
    Something is very wrong here.
    You shouldn't have to raise your ISO that high for a room like that.

    Here is what I get at 3200 ISO in a cavernous temple:

    So I am going to guess your flash had FEC set to lower than zero.
    Or something.

    In terms of metering ... I don't specifically meter. Not like you describe there.
    I *decide* what my settings should be ... and then I add flash.
    The TTL flash usually takes care of it.
    Sometimes I go to manual.

    But really, there is something very wrong there in how you set up your camera and flash. It shouldn't be that under-exposed.

    Also, if you DO see your images are under-exposed, then turn up the FEC and get more light out there!

    Then, another thing .. if you have your flash as the dominant light source, like here, then go into your custom settings and change your TTL flash settings from Evaluative to Average.

    More about it here:

    Ultimately, you should have done test shots BEFORE the bride and groom entered, and sorted out your camera and flash settings ... and then nailed the exposures with the actual wedding reception and dances.

    And no, I wouldn't use spot-metering here.
  • RubyRuby Member
    I shot a wedding last weekend and the dance is a hard one to nail I found. The lighting was dim, so I was using bounce flash. I switched to A1 Servo but as I found out later the camera did not focus track in low light well at all (5dm3). Will the camera focus track with the flash attached? My exposure was OK. Cheers Ruby.
  • ZenonZenon Member
    Nope. The AF assist beam is disabled on the external flash when in AI Servo for all Canon bodies. Only enabled in one shot.
  • jhilgersjhilgers Member
    edited April 2013
    With reference to Neil's article: http://neilvn.com/tangents/canon-ettl-flash-average-vs-evaluative/

    The brief explanation for Canon's 600EX-RT flash has been copied from its user manual and attached to my post below.

    So according to this if you want it to meter for the whole scene, you would set it to the average mode setting not evaluative.

    This means to me that the Evaluative setting in the flash actually does the opposite of what the Evaluative setting in the camera's metering settings does.

    The only way I can remember this so I don't confuse myself is to think of this saying: My flash is going to "evaluate" the light on my target subject or it can "average" all of the lighting in the entire scene before outputting its light.
  • Thank you for such superb answers.

    Neil, FEC was at 0.

    I'm doing some evaluative/average flash mode experiments as I write, but I'm not finding any real difference. I suspect it's because the light is not low enough yet...

    In actual fact, I did not want to meter flash for the whole scene because I wanted to keep the feel of it. But, I suspect now that sometimes it's just not possible...
  • Sometimes (not always) best exposures for reception work are off camera main lights on stands, w/on camera flash as fill. Kinda of a PIA to setup and bring (and buy), but gets the job done - stops motion blur, opens up backgrounds. However, I have a great friend and a helluva wedding pro that shoots w/single on camera Quantum Stobes bounced into ceilings. His stuff looks great! But these are large flashes (320 ws - often @ full pwr.), w/slower recycle times. Our stand lights are AC powered, w/on camera Canon 580's @ 1-2 stop under w/o an external power supply. Super quick recycle times. Most (nearly all) receptions shot @ 800 ISO. Works for us. Ed K.
  • Corojo,
    Maybe you would have some input on this post I made awhile ago. I would be interested in hearing your thoughts.


    I was surprised no one replied to it yet even with all of this talk about off-camera flash and everything.
  • CorojoCorojo Member
    just finishing PP from our wedding from this past Saturday. I'll post some reception images soon. Ed K.
  • Neil vNNeil vN Administrator
    slravenphoto said: In actual fact, I did not want to meter flash for the whole scene because I wanted to keep the feel of it.
    I'm not sure what you mean here?
  • Neil - I wanted to keep the whole ambience of the scene, i.e. the random low lights. So I metered for a particular spot which I wanted kept a middle grey.
  • StephenStephen Member
    Is this like trying to shoot a candlelit scene? If you used flash, the candlelit scene would be "ruined"? Is this the type of ambient light preservation you were trying to achieve?
  • I'm sure that if you dialed down your flashes (that were providing ambient fill light) low enough and geled everything it would not affect candlelight too badly.
Sign In or Register to comment.