Bird photography - metering

[CT]

I guess we can't all be as clever as you CT!
Maybe I have an artistically set up brain now, and not so much technically.
I think you just have to accept what some people find simple, others will have difficulties. Being a teacher I understand this and deal with it every day.

Maybe I could send you some notes on learning the basics of accordion tuition (my first instrument) and see how you get on?
Then come and spend an hour in my studio with the instrument on your lap and see how that goes? You would be most welcome. I'd even do it for free! ;-)

Having digested the information I see the logic, so yes the penny has dropped, hopefully you won't be so surprised at that, however laughable it may be!

LOL. I don't consider myself clever at all- far from it. The studio sounds great. One of the regrets of my life is that I didn't take up a musical instrument at an early age. I eventually decided to learn to read music when I was in my 40s and after a succession of keyboards i've had an 88 note Korg Triton for a few years. I don't play so much these days I must admit, but it's still nice to murder a bit of Bach or Beethoven when the spirt moves me - I just find it incredibly relaxing, but I certainly don't consider myself a pianist - too many bad habits.

 

[Tori]

It's a gift. Some of us have it, some of us don't.

If you're going to murder anything you can stay at home!

 

[hollis_f]

The Swan and the Pelican are good examples of when evaluative metering isn't going to hack it and will be fooled pretty well every time. In that situation I'd probably spot meter on the brightest part of the bird and give at least two stops extra exposure

In other words, Spot Metering won't hack it either and would be fooled every time. You'd do just as well to use Evaluative and use two stops EC.

In cases like this it doesn't really matter what metering one uses - one has to be aware how how the different modes work (especially one's favourite) and the limitations and how to get round thos elimitations.

 

[CT]

In other words, Spot Metering won't hack it either and would be fooled every time. You'd do just as well to use Evaluative and use two stops EC.

Well you could, but you're missing the point a bit in that you're not relying on the straight spot reading - you then give 2-3 stops exposure over the indicated spot reading. You're also starting from a much more accurate reading of the important bit (the highlights) than you would with a more general evaluative reading.

 

[reedwarbler]

a big thank you to tim and cedric definately made it a lot clearer to this nooby,thank you very much.

 

[tdodd]

Well you could, but you're missing the point a bit in that you're not relying on the straight spot reading - you then give 2-3 stops exposure over the indicated spot reading. You're also starting from a much more accurate reading of the important bit (the highlights) than you would with a more general evaluative reading.

 

[hollis_f]

Well you could, but you're missing the point a bit in that you're not relying on the straight spot reading - you then give 2-3 stops exposure over the indicated spot reading.

No, you're missing the point a bit in that you're not relying on the straight evaluative reading - you then give 2-3 stops exposure over the indicated evaluative reading.

 

[Briony]

Tim, if you'd seen the pm's exchanged between Nicky and I last night you would of laughed. We both read your posts and pm'd each other asking if we 'got it'. My poor brain was overloaded but after re reading it numerous times, I'm now with you. We did it in 'idiot terms' between us and I think we're reading from the sane hymn sheet now
Now this?! There's more?! This will have to be digested later today...

The information in this thread is much appreciated (and by many I'm sure)
As a noob having only had a dslr since just before Xmas some of this stuff is hard core and those of you who find it easy need to appreciate that too.
CT has offered me some valuable advice which is always appreciated, but his suggestion that my idea of a workshop for this kind of stuff was 'laughable' was myopic at the least. I currently have five books from my local library on exposure etc and I'm an avid reader. After a few pages my head is trying to implode with all the information. I think I've learnt more by attending TP meets and having hands on advice than reading and confusing myself.
I just found this a little 'go away and read a book'. I'm sure that is not what was meant.
I read books, I read TP, but the hands on stuff is what is making me improve the most, and to those people who have donated, I thank you very much x

Trouble with TP meets is that everyone runs around snapping away and I am always frightened to ask questions in case you are hindering them :shrug: so a "workshop" kind of set up is ideal

Tori, be careful you might end up with a "title" under your name which the mods seem to think it is funny, look at mine, "just how I like my gravy" I hate thick gravy


Thank god for Nurofen works great on headaches

 

[CT]

No, you're missing the point a bit in that you're not relying on the straight evaluative reading - you then give 2-3 stops exposure over the indicated evaluative reading.

Sorry - disagree. An evaluative reading is the cameras best attempt at choosing the best overall exposure for the tones in the scene. It will only place importance on the highlights proportionately with how much they occupy the frame. and it can be well wrong, so an evaluative reading could have you easily adding your two stops exposure from a wrong starting reading.

Spot metering from the highlights then adding to the exposure is foolproof - the principle is well known, tried and tested.

 

[tdodd]

Spot metering from the highlights then adding to the exposure is foolproof - the principle is well known, tried and tested.

That's what I've been trying to say in many of my posts in this thread and several others. I've been doing it for yonks. I know it works very well. Some people just won't hear it.

I guess you can lead a horse to water but you cannot make it drink. You can only pity the poor horse.

 

[hollis_f]

Sorry - disagree. An evaluative reading is the cameras best attempt at choosing the best overall exposure for the tones in the scene.

No. It's the camera's best attempt at choosing the best overall exposure for the subject in the scene.

 

[menthel]

No. It's the camera's best attempt at choosing the best overall exposure for the subject in the scene.

I don't think that the camera can tell the difference between a bird and a clock, so it will be judging the tones in the scene.

 

[Tori]

Blimey! What chance do we stand if you're arguing amongst yourselves?!
It seems there is more way than one. I think my way is best LOL

 

[menthel]

Blimey! What chance do we stand if you're arguing amongst yourselves?!
It seems there is more way than one. I think my way is best LOL

I wouldn't worry, its usually an arguement over semantics and it all ends happily in the end!

 

[hollis_f]

I don't think that the camera can tell the difference between a bird and a clock, so it will be judging the tones in the scene.

What clock? There is no clock in that swan photo. What are you smoking?

 

[menthel]

What clock? There is no clock in that swan photo. What are you smoking?

Nowt, just making a point. The camera does't know the difference between a clock and a swan, what it does is the amount and the wavelenght reflected back off of it. Therefore your statement I quoted was wrong. You may now detwist your knickers!

 

[CT]

Just had a bloody power cut and lost some unsaved work. :bang:

I suppose to round this off and avoid some of the confusion we really need to talk about evaluative metering, although I was trying to avoid it because we really don't know enough about it.

Evaluative metering (Canon) or Matrix metering (Nikon) are pretty similar and they work by dividing the metering area (the focusing screen) into a number of invisible areas and each little area is metered individually, The camera is programmed to compare all these readings and attempt to get you the perfect exposure by using such parameters as where the various tones appear in the frame, how large they are, how important various tones are likely to be etc etc. It's pretty advanced stuff and you can definitely use it to your advantage as long as you know when it's likely to let you down - It's only as good as it's programming - it doesn't think and will never really know what part of the scene YOU consider important.

If the above isn't impressive enough, when the camera has gathered all this information, it then compares the data with the data from hundreds, possibly thousands, of sample images stored in memory to try to find a match for the image you're trying to capture and expose accordingly. The exact nature of these stored images isn't distributed by Canon or Nikon AFAIAA, in fact both manufacturers keep their cards pretty close to their chests with regard to the specifics of their systems so we don't know as much about the systems as we'd like to, clever as they are though, they can get it wrong, but it isn't really that difficult to recognise the conditions where you need to be in the driving seat and determine the exposure yourself.

The most knowledgable person I know about the technical specifics of Canon equipment is Canon Bob - if he can't throw any light on the xeact nature of the system I doubt we're going to be any the wiser.

 

[hollis_f]

Nowt, just making a point. The camera does't know the difference between a clock and a swan, what it does is the amount and the wavelenght reflected back off of it. Therefore your statement I quoted was wrong. You may now detwist your knickers!

I really have no idea what you're talking about, and I suspect that makes two of us.

No the camera doesn't know the difference between a clock and a swan. But it does know the difference between the subject of an image and the rest of the image, so my statement was correct. You may now go back to your home planet.

 

[CT]

I really have no idea what you're talking about, and I suspect that makes two of us.

No the camera doesn't know the difference between a clock and a swan. But it does know the difference between the subject of an image and the rest of the image, so my statement was correct. You may now go back to your home planet.

Nice! Bad attitude and you're wrong anyway.

 

[menthel]

I really have no idea what you're talking about, and I suspect that makes two of us.

No the camera doesn't know the difference between a clock and a swan. But it does know the difference between the subject of an image and the rest of the image, so my statement was correct. You may now go back to your home planet.

A little rude, are we not? I think you will find the camera is still evaluating lighting patterns that reflect from the object and not the object itself. It doesn't say 'ooh, look, a swan, this is how i shold meter this!'. It evaluates the light reflected off, and in the case of the 7d gives a weight to the metering around the focus point as it assumes that this is the most importent bit to get right.It doesn't know, per se, that that is the 'subject'. It will then compare this to its doohickery as Cedric has stated, looking for a match so that it can then give appropriate settings to best capture the scene. At least that is how I interpret it.

Please do not be so rude or antagonistic, it demeans you as well as me.

 

[CT]

No. It's the camera's best attempt at choosing the best overall exposure for the subject in the scene.

OK this is the last from me for a while I've stuff to be getting on with.

Let's take two scenarios - a sunlit swan occupying a small part of the frame against a very dark water background and a situation where you're photographing a person in shade up against a tall hedge but with a hole in the hedge through which bright skylight is showing - lets say the hole is the same size as the swan - same position in the frame.

The camera will never distinquish between those two situations and in one the highlights are actually all important, while in the other they're negligible.

Laters...

 

[Briony]

Boys, Boys, stop arguing or I will bang you heads together :nono:

 

[hollis_f]

The camera will never distinquish between those two situations and in one the highlights are actually all important, while in the other they're negligible.

Yes it will! Excellent example.

In the first case I presume you've focussed on the swan (otherwise it's all a bit pointless). In that case the camera knows that the swan is the subject. It will go through the complicated, proprietary algorithm it uses and try to get the swan correctly exposed.

In the second case I'll presume you've focussed on the person in the shade - so that the camera knows that this is the subject. Now it will run through that algorithm again and try to get the person exposed correctly. The algorithm will probably (as nobody outside of Canon really knows how it works) assume that the bright gap in the hedge is the sun poking through clouds as I'm sure there are similar shots in its database.

 

[hollis_f]

It doesn't know, per se, that that is the 'subject'.

No, it presumes that the user has focussed on the subject. By focussing on a particular point you are telling the camera 'this is the subject'.

 

[menthel]

No, it presumes that the user has focussed on the subject. By focussing on a particular point you are telling the camera 'this is the subject'.

Perhaps, but the camera doesn't know that. In less 'clever' cameras than the 7d in the canon range the metering won't be moved over to the focus point anyway.

 

[Tori]

Will you two bog off?! You are ruining an informative thread with your bickering! Take it to pm if you want to continue!

Personally I think admin could tidy up this thread leaving the informative stuff only.

 

[menthel]

Will you two bog off?! You are ruining an informative thread with your bickering! Take it to pm if you want to continue!

Personally I think admin could tidy up this thread leaving the informative stuff only.

It's actually all part of what we are discussing. Knowing what the camera is doing helps you get the right exposure.

 

[Tori]

I think we've already had a good explanation. There maybe other ways, but arguing over them isn't helping! It's surplus to requirements!

 

[menthel]

I think we've already had a good explanation. There maybe other ways, but arguing over them isn't helping! It's surplus to requirements!

Depends if there is misinformation. Not that you will have that if you listen to Tim and Cedric! I will leave your thread alone now.

 

[CT]

I think the thread has really run it's course I hope it's proved useful.

Frank instead of chiming into the thread periodically and confusing the issue, why don't you start your own thread on basic metering, then Tim me, and anyone else can come in and dissect your posts. Make it a good tutorial mate.

That's it from me.

 

[tdodd]

Right, I'm not posting any new info here, so this shouldn't add to any confusion. This is just to reinforce some of the things I've been saying.

I was out shooting today with starrider and suggs and I was showing the guys my methods of setting a manual exposure by metering the highlights at +3 stops. By setting the exposure manually I lock it in place so it will not change. By metering highlights at +3 I squeeze them all the way over to the right hand edge of the histogram. With the exposure locked I am free to shoot any kind of subject and scene that remains in that light for as long as I like and all the exposures will be good, not thrown off by metering quirks.

Just to illustrate that point, here is a series of six shots with wildly differing tonal spreads and each holds the highlights and shadows beautifully. All the exposures are identical. All I needed to do was to focus/compose/shoot repeatedly without a care in the world about metering or exposure adjustment. There is no artistry in the shots. They are simply to show the consistency of exposure and results.

In the first picture we see a bright white building in the background and a black umbrella in the foreground. I spot metered the building at +3 stops. Nothing is blown and the building retains full detail. The black umbrella is also safe from clipping (apart from a handful of pixels) and retains plenty of detail. Let's just review the image and its histogram more closely....

If the building represents a bride's gown and the umbrella represents a groom's tux then this is the way to secure a solid exposure frame after frame after frame.

If the light is changing then that's a whole new ballgame, but a bit of dial twiddling should have you right back where you want to be, until the light changes again.

If people can accomplish the same consistency for such widely varying scenes with evaluative metering and autoexposure, no chimping between shots, then I would be gobsmacked.

NOTE : Please note that metering highlights at +3 stops works well for Canon cameras shooting to raw and using neutral picture style. I think Nikon shooters would be advised to aim for something like +2.3 stops. I don't know about the metering in other camera brands. If you shoot to JPEG you might need to back off by 1/3 stop.

 

[lensflair]

Right, I'm not posting any new info here, so this shouldn't add to any confusion. This is just to reinforce some of the things I've been saying.

I was out shooting today with starrider and suggs and I was showing the guys my methods of setting a manual exposure by metering the highlights at +3 stops. By setting the exposure manually I lock it in place so it will not change. By metering highlights at +3 I squeeze them all the way over to the right hand edge of the histogram. With the exposure locked I am free to shoot any kind of subject and scene that remains in that light for as long as I like and all the exposures will be good, not thrown off by metering quirks.

Just to illustrate that point, here is a series of six shots with wildly differing tonal spreads and each holds the highlights and shadows beautifully. All the exposures are identical. All I needed to do was to focus/compose/shoot repeatedly without a care in the world about metering or exposure adjustment. There is no artistry in the shots. They are simply to show the consistency of exposure and results.

In the first picture we see a bright white building in the background and a black umbrella in the foreground. I spot metered the building at +3 stops. Nothing is blown and the building retains full detail. The black umbrella is also safe from clipping (apart from a handful of pixels) and retains plenty of detail. Let's just review the image and its histogram more closely....

If the building represents a bride's gown and the umbrella represents a groom's tux then this is the way to secure a solid exposure frame after frame after frame.

If the light is changing then that's a whole new ballgame, but a bit of dial twiddling should have you right back where you want to be, until the light changes again.

If people can accomplish the same consistency for such widely varying scenes with evaluative metering and autoexposure, no chimping between shots, then I would be gobsmacked.

NOTE : Please note that metering highlights at +3 stops works well for Canon cameras shooting to raw and using neutral picture style. I think Nikon shooters would be advised to aim for something like +2.3 stops. I don't know about the metering in other camera brands. If you shoot to JPEG you might need to back off by 1/3 stop.

Prepare to be gobsmacked.

I live in matrix 99% of the time and never have a problem. I either have the camera in aperture priority for available light, switch to manual if the scene isnt changing for greater consistency between shots. With on camera flash camera is in manual, flash is in TTL and I ride the FEC dial, off camera flash both camera and flash is in manual.

I'm using a D3 and sb900 most of the time, but use a quantum flash off camera too. I nail my exposures on the first shot most of the time, if not the second. But if there is an important shot coming up I take a test shot so my first money shot is on the money.

Its just the way I have learnt to work, and it works for me.

 

[tdodd]

Encouraged by your support for matrix metering I just popped my head out the door and put evaluative metering to the test with my 1D3 in Av mode - the same camera used for the examples above. Unfortunately the results are extremely inconsistent and would/could prove a nightmare to deal with on a large shoot, or where exposure was critical.

Here are just three images using centre point as the active focus point. The exposures are all over the place, ranging from 1/30 to 1/320 for what is basically some greenery lit by identical soft light.

That's the same bush in the first two shots and the third shot is of the top of the conifer growing up behind that bush. It's stuff like this that makes me hate evaluative metering and autoexposure. The exposures should all be the same. Look at the sky in each frame. Look at the shutter speeds. Not even close. Now I could twiddle the EC dial, but I couldn't guess precisely how much EC to dial in for each adjustment in composition. I'd have to either watch the shutter speed alter and simply override it back to what I wanted or guess, chimp, adjust and try again. That would be daft.

If your D3 could deliver consistent results, unlike these, without riding the EC dial then that would be gobsmackingly good performance.

I then switched to manual exposure and spot metered the lawn at -2/3 and shot roughly the same three compositions. They were shot 17 minutes later, with an ever brightening sky as dawn's crack approached, which is why the shot at 1/125 is a little brighter than the earlier shot at 1/60. Here they are....

If you were after a bird hopping about in a tree, with varying amounts of sky filtering through from behind, or if it suddenly took flight, I know which metering and exposure modes I'd rather be in. Your mileage may vary.



EDIT : Brainwave - perhaps my 1D3 is faulty. I tried the same thing with my 5D2 and a different lens. The outcome was the same unsatisfactory variation in exposures....

I know why it is happening. I'm not daft. I just don't want it to happen. There is a solution. A good one. Av mode is not it.

Of course, there are times when Av and evaluative do make sense, but I generally find Manual and spot to suit my needs better.

 

[CT]

You can't argue with the point Tim is making here - he' s patently and obviously right.The point is though that that this thread started out as someone in need of pretty basic help and seeking advice. I've taught a lot of people photography over the years and if I've learned anything it's that people vary and need to progress at their own rate - information overload and hitting them with advanced techniques it may have taken you years to learn is't always a good thing,in fact it rarely is. Get them taking decent pictures by whichever method works for them - it's the best encouragement in the world - feed them the more advanced information when they ask, or when you can see from their results that they're ready and will benefit.

Interesting though this thread is, Nicki has for more basic issues that need addressing than this. Quite early in this thread though much has been made of exposure being linked to the AF spot in the 7D in evaluative mode. Now I'm possibly the world's biggest 7D fan, but I've read the Canon blurb, and on the face of it this feature could be a bird photographers dream in many situations ( although a curse for some other applications). Have I noticed a difference in my bird exposures using evaluative, metering in the 7D? - I don't think I have. Have I seen a difference in birds shot against the sky with no compensation? Nope - the same silhouette you'd expect. Surely if this system was of any use at all it would make short work of that situation without any help from the photographer, or at least improve it considerably?

Perhaps I'm missing something, but I'd like to see any evidence that this system is as useful as claimed or that you can rely on it to the extent that some people seem think you can.

 

[tdodd]

I've just had a brainwave - change cameras (7D) and lenses (100-400) again and now pick a scene without the sky to cause a problem. Now I reckon the fabled talents of evaluative metering should really shine through. Let's see...

Oh dear. Seems it's not working as advertised. The camera has no flipping clue. Exposures are all over the place. If we take the first two images as examples of a bride in one and the groom in the other the exposure difference is 2.3 stops. Try matching skin tones with exposures roaming all over the place like that. The most representative exposure of the scene as I saw it is probably the last one, at 1/320, which spookily enough had a grey car as the subject. Not many of the pictures got 1/320.

The subject exposures are sometimes good, sometimes poor. The tones of the surroundings (driveway and road surface, and that white door in the background) are up and down depending on the subject. I wouldn't want to shoot motorsports like this. I wouldn't want to shoot a wedding like this. I wouldn't want to shoot much at all like this. I'm surprised that so many people do.

That D3 metering must be quite something if it can nail each of these flawlessly in a split second.

 

[lensflair]

Well different strokes for different folks I guess. Here is a series of images I took in town a couple of days ago, matrix metering in aperture priority.

If I was taking a series of shots in the same direction and wanted to ensure exposure was absolutely consistent between shots, I would either hit the exposure lock button or switch to manual control. But I was just scouting the location near the bridge and snapped a few frames while walking around.

I did dial in +EV on a few of the shots, I just estimate how much I need for a given scene before I take the shot, it happens in a split second before I press the shutter button. Like I said I it if I don't get it right first time, I chimp it, but that takes less than a second. The last image I took auto displays, I glance at it, and the adjustment is made before my finger is ready to press the shutter again. But if the shot was important and I knew it was coming, I would have metered a test shot and be ready to go, as I'm sure you would also do no mater what metering you are using, these shots weren't critical so I just took the first shot I got. I might have dialled in about 0.3 too much for the guy on the roof, but he was severely backlight and I would actually be pretty happy with that as he is the subject and I like to ETTR.

I have included a 100% crop on the area I wanted to expose for in some of the shots, I hope they show up OK I haven't posted images here before and my host seems to resize them.


Anyway, I'm just showing you that for some people matrix metering, at least on my Nikons, does a very good job. I think it just depends on what you get used to.

 

[lensflair]

Just one question, I'm not a bird photographer so perhaps I'm missing something, but how do you spot meter a bird in flight?

I would have thought you either use manual mode and have your exposure set, or you rely on evaluative and have EV dialled in if you expect it to be backlit. Am I missing something?

EDIT: just had a look at some of your bird photography CT, wow nice work. I'm sure you can answer my question

 

[tdodd]

I don't spot meter birds in flight. I spot meter something within the same lighting that makes a reliable reference (green grass, my own palm, the sky itself) and then I set my exposure manually and wait for the subject to appear.

If it's sunny conditions I may not meter anything at all. I'll simply dial in a Sunny 16 exposure - perhaps 1/1600, f/5.6, 200 ISO - and I'll be ready for the birds. I can set the exposure before I even leave home.

If the light changes then I re-meter and adjust. I can do all this in advance of my subject appearing. In my experience the lighting does not often change quickly. Sure, sunshine and patchy cloud will have you adjusting things up and down like a yo-yo. That's if you want to take the shot regardless of the light. You may well choose only to shoot when the light is right, which means you just wait instead of adjusting anything. Clear blue skies or evenly overcast skies mean the light will barely change from one minute (or hour) to the next. If you're shooting in all directions, into shade, into light then manual is probably not a good choice. I rarely shoot like that. If I did then I would switch to a more automated system.

 

[lensflair]

I don't spot meter birds in flight. I spot meter something within the same lighting that makes a reliable reference (green grass, my own palm, the sky itself) and then I set my exposure manually and wait for the subject to appear.

If it's sunny conditions I may not meter anything at all. I'll simply dial in a Sunny 16 exposure - perhaps 1/1600, f/5.6, 200 ISO - and I'll be ready for the birds. I can set the exposure before I even leave home.

And that's kinda my point, I don't understand what difference it makes which metering method we choose to use to get a correct exposure? You keep implying that using evaluative metering is somehow inferior to spot metering, and that it would be almost impossible to get any consitency between shots, I respectfully disagree, I do it all the time.

 

[CT]

Just one question, I'm not a bird photographer so perhaps I'm missing something, but how do you spot meter a bird in flight?

I would have thought you either use manual mode and have your exposure set, or you rely on evaluative and have EV dialled in if you expect it to be backlit. Am I missing something?

EDIT: just had a look at some of your bird photography CT, wow nice work. I'm sure you can answer my question

Thanks mate for the compliment .

As usual there's more ways to meter for a bird in flight than there are to skin a cat and none of them is necessarily right or wrong. The problem for me is that I don't often go out specifically to shoot birds in flight- I'm usually tripod mounted waiting to shoot birds perching or landing within range, so these opportunities tend to be sudden which needs a quick response and a different approach and it's one situation where you are probably better off hand held anyway for more freedom of movement overhead than you'd get even with a good gimbal head. In that situation I'd probably quickly dial in 2 or 3 stops extra exposure which usually gets you in the ball park using evaluative, even if it isn't bang on I know I'll get a decent result in editing.

If I'm hand holding anyway, an old trick is to meter on green grass - another is to extend your arm up to the sky as full stretch and meter from the underside of your hand (simulating the situation with an overhead bird) although it isn't an exact science - the underside of the bird can be in heavy shadow or it may be quite brightly lit by more oblique sunshine.

Cue Tim waiting in the wings to tell you that his system is best in this situation as he'll get a good exposure with his settings as long as the light doesn't change, and he'd be right too.