why does the exposure change when i change lenses?

[WouldBe]

Yes, but I've already said I'm shooting in MANUAL mode while using an incident meter. Therefore the camera's meter does not affect the exposure.

I never said the cameras meter does affect the exposure. A camera works on light reflected from a subject otherwise it can't get from the subject to the camera. The camera doesn't give a shit wether the subject is illumunated by a 1.5W torch bulb or by the Eddystone lighthouse.

In fact, try it yourself! Get 5 f/2.8 prime lenses, a 17, a 35, an 80, a 200 and a 300. Stick them on a DSLR. Take an incident light reading under constant light and change the ISO on the meter until you get 1/500 of a second at f/2.8. Set the DSLR to the same ISO as the meter and 1/500s @ f/2.8. Take pictures with each lens at this setting and I personally guarantee that each exposure will be the same.

Which is basically what you have tried and proved it doesn't work. Hence this thread.

 

[Hocus Eye.]

WouldBe

You are a bit out of your depth here. Herbsman knows what he is talking about. When you use a separate light meter there are two kinds of reading that you can take. These are Incident light meter readings and Reflected light readings. In-camera meters are reflected light meters. The measure the light reflected from the subject.

Hand held meters can operate in Reflected light mode where you point the meter at the subject to get the reading. They can also be used in Incident light mode. In this case there is a translucent white domed device that you put over the light sensor. You then point the meter at the light source or back towards the camera from the subject. Herbsman is talking about this technique.

It is just possible to obtain an incident light reading from a camera by putting a plastic cup over the lens and pointing the camera at the light source. This is unorthodox but does appear in some photography books.

WouldBe you say:

I can't see the difference in focal length betwen a 35mm lense and a 50mm lense being all that different to cause much of a difference in the inverse square light levels that you would need a 50% bigger aperture for the 50mm lense.

However the inverse square law is maths. It doesn't decide not to come into play when the numbers are relatively small. It states that.

Intensity of light from a point source is inversely proportional to the square of the distance.

.
What this means in practice is that if you shine a torch (almost a point source) at a wall in a dark room from a known distance say far enough to produce a circle of light one meter in diameter and then move back to twice the distance the circle of light will double in diameter.

The diameter will have doubled but the area will have increased to the square of the previous area. If you measure the light intensity of each circle in turn you will find that the intensity of light in the smaller circle to be the square of the larger one not just double. It is as if you had a tin of paint and wanted to paint each circle. It is a geometric increase or decrease not a linear one.

 

[WouldBe]

And a camera always works on the light reflected from a subject.

If you have a dark subject less light will be reflected into the camera requiring a longer exposure than a light subject regardless of how much light you shine incident onto the subject. That's a basic law of physics.

I've got over a thousand photos I've taken over nearly 30 years. Most of which I have recorded the shutter speed and aperture for and I don't recall using the same aperture whenever I've changed lenses and photo'd the same subject under the same lighting conditions.

I'll try and have a look but I'm not doing it right now.

 

[Wilson]

Herbsman

It may be that your 50mm 1.8 lens is jamming in the fully open position when you press the shutter instead of stopping down. This should be easy to check.

Thats a real possibility, but it might be an intermittent fault

or the light might have changed but you didn't notice which seems very unlikely

I have seen differences between lenses, and cameras and meters but i cant really see that you could gain/lose much more than 1/3 of a stop unless youre using lens extention.

 

[Hocus Eye.]

Yes it is obvious that a dark object reflects less light. We know this.

 

[WouldBe]

Yes it is obvious that a dark object reflects less light. We know this.

So incident light is irrelevant. You want the reflected light as that is what will affect the film / ccd.

 

[Hocus Eye.]

So incident light is irrelevant. You want the reflected light as that is what will affect the film / ccd.

No the incident light is what gives rise to the reflected light. You can measure either. Go away and look up incident light readings. If you have been doing photography for 30 years you must have owned a Weston light meter at some stage. Did you never wonder what the translucent white Invercone was that came in the little leather case on the meter cord. Did you never read the instruction manual. Or failing that have you never opened a basic photography book?

 

[WouldBe]

No the incident light is what gives rise to the reflected light. You can measure either.

Sorry that's bollox. A camera works only on the reflected light which will be different with different subjects and therefore require different exposures regardless of the light incident on the subjects. Again this is basic physics.

Why not try it. Place a light coloured piece of card and a dark coloured piece of card side by side and illuminate with a light source. Then set your camera up so that each piece of caed individually fills the viewfinder. use the cameras meter to determine the exposure and it will be different for each piece of card as more light will be reflected by the lighter card than the darker one.

Why do you think proffessional photographers on shoots keep checking the light levels even though they haven't changed the lighting. The incident light is the same. The reflected light has changed depending on who / what the subject is.

If you have been doing photography for 30 years you must have owned a Weston light meter at some stage.

No. Never used one. Couldn't see the point when the camera has one built in.

 

[Hocus Eye.]

WouldBe

You keep ranting on about the way a camera works. Yes that is right. But here we are talking about the way a light meter works specifically a light meter that is not in the camera. Herbsman is using a hand held light meter.

If you have never used a hand held light meter then you don't understand the difference between incident and reflected light readings.

I don't care if you have been taking pictures for more than 30 years, so have I.

 

[WouldBe]

If you have never used a hand held light meter then you don't understand the difference between incident and reflected light readings.

I don't need to have used a light meter to know the difference between incident and reflected light. A camera works on reflected light only and that is what should be measured.


As I don't recall a zoom lense change the apperture according to what zoom you have the lense on then it could be that the aperture stops on the zoom lense are an approximation to cover the zooms focal range. These will therefore be different to the stops if it were a fixed lense and could account for the over exposure. Which is why whenever I've used a zoom or teleconverters I've always used the built in meter as it's the only way you can be ceratin of how much light is getting into the camera regardless of how much light is incident.

 

[Hocus Eye.]

Zoom lenses do change the aperture as you zoom. Most zoom lenses state their different apertures at the short end and the long end. The aperture at the long end will give a much 'slower' effective aperture than at the short end. On many zoom lenses there is a series of long curved red lines down the barrel to show the different apertures at different extensions. Optics is a precise science based on numbers. Approximations don't come into it.

 

[boskysquelch]

yup zooms soak up light...endof.

Hocus' last post is correctest.

the amount of light that will "reach" the other side of a f2.8@1/500 on a 50mm is not the same as f2.8@1/500 on a zoom set to 50mm, which will be blahblahblahblah.

The choice of reflected/incident readings isn't the issue...the construct of the lens is...ultimately.

 

[Hocus Eye.]

bosky

Thanks for your support but strangely enough I don't agree with you.

You say

the amount of light that will "reach" the other side of a f2.8@1/500 on a 50mm is not the same as f2.8@1/500 on a zoom set to 50mm, which will be blahblahblahblah.

I say that f2.8 @ 1/500 on a 50 mm is the same as f2.8@1/500 on a zoom set to 50mm. Or for that matter any lens on any format on any camera in the world. That is the whole point of the f/focal length system and why hand-held light meters work for all cameras. I think that as a professional photographer you know this.

What I was saying in my post was that the particular aperture on a zoom lens that gives a reading on the barrel of f whatever at its long end will not give the same a effective aperture at its short end. Also that zoom lenses tend to have smaller effective apertures at the short end than prime lenses because it costs so much to make them. In other words they are 'slower'.

However it is late and I am sure everyone is just as tired as me now.

 

[boskysquelch]

other words they are 'slower'.

mmmmisnomer alert...bit like people applying genres to music(ie wrong)...fast/slow lens...film lenses/ still camera lenses...industry crossover doodah_bollocks

wikiwikiwah,,, http://en.wikipedia.org/wiki/Lens_speed

wot about IFEDs... macros...it's all in the construct_position of individual lenses within a lens...choice of glass(even at an atomic level), machining & even interior finish

I'm v tired too, and know I'm "right".

fffkn zooooms

N_N X

 

[cybertect]

It may be that your 50mm 1.8 lens is jamming in the fully open position when you press the shutter instead of stopping down. This should be easy to check.

This is the only explanation I can think of that comes close to fitting the facts in the OP.*

Even then, it's not two stops of difference, it's 1 and 1/3.



Easily tested, though. It should show up quickly enough if you run through a range of apertures with a fixed shutter speed.

Any chance you could post examples of the original problem with EXIF data? It might help.

e2a: * unless you happened to accidentally leave a neutral density filter on the zoom

 

[WouldBe]

Also that zoom lenses tend to have smaller effective apertures at the short end than prime lenses because it costs so much to make them. In other words they are 'slower'.

So f2.8 isn't the same regardless of what lense you use otherwise the exposure time would be the same.

 

[boskysquelch]

a bit too long...but lenses & exposures have been on ma mind recently too

fast n slow is also defined through focal lengths, apertures, choice of focus & CONSTRUCTION.

they do vary... I've seen it loads of times.

I've even been able to tell the make of camera from exposure(ie the picture they produce)...I'm sure I'm not the only one that has seen this. eg Zenith vs Practica, Yashica vs Nikon vs Canon, Contax vs Leica,Rolleiflex vs Rolleicord vs Mamiya Rollei vs Hasselblad vs Mamiya, Sinar vs Toya vs etc etc etc ...and then...really...lenses?.....Angenieux :melts:

& zooms are by far the worse offenders that cover all in one.

oh and digi_old skool fstop/shutterspeeds/ISO etc...tis still all jus a construct in it itself...misnomers(agin) given for ease of comparison for the can't be bothered to think different.


At one time the difference between Pro and Amateur was even based on the film results alone...given the amount of variables you could find in stock & processing...and I'm only talking about the 1980s ffs!

Ultimately a digi exposure has even more variables attached to it...almost like getting back to how it was when Photography first began...and "standardisation" was still Alchemy.


*atm I've been reading some magazines ( The Gallery) from the 30s...and it's quite kewl reading the language_terminology being invented so people can go on (historically) to invent Photoshop eg masking , Gamma control_curves(ish) through exposure & development( I know!) etc...fascinating.



aaaaaaaaaaaaaaaaaaaaaaand for what it's worth I've thrown my Minolta FlashmeterIII up against my digi cams in the past and they sooooooooooooooooo don't match.......taken the reading to film and the images were perfectly exposed.

 

[Hocus Eye.]

So f2.8 isn't the same regardless of what lense you use otherwise the exposure time would be the same.

Yes the exposure time is the same that is the point.

f/2.8 is a mathematical formula. As you mentioned in one of your own posts it represents the focal length of the lens over (divided by) the number, in this case 2.8. To give simpler examples a 50mm lens set to f/2 will have the formula 50/2 which comes out at 25mm.

That is to say that the 'effective aperture' is 25mm. A 200mm lens set to f2 it will have the formula 500/2 which comes out at 100mm. Both the 50mm effective aperture and the 100mm effective aperture will produce an identical intensity of light on the image plane because of their relative distances from that plane. (The term effective aperture makes allowance for curvature of light between the lens elements and the aperture)

This means that f/2 on the 50mm lens and on the 200mm lens will both produce the same exposure. So while the actual measurable hole (aperture) is not the same the number is the same and the exposure is the same.

That is the whole point of the formula, so that light readings are applicable to all cameras that are designed with the standard aperture system. Remember the aperture system was invented long before cameras had built-in meters.

WouldBe at this point I am giving up trying to explain to you what others on the thread already know. This is a derail and Herbsman has a genuine question in his original post

 

[WouldBe]

Yes the exposure time is the same that is the point.

No it's clearly not.

For a zoom lense to be slower would require a longer exposure or a larger aperture. Therefore aperture sizes aren't transportable across lenses.

You've already stated both situations are true when clearly they can't be.

So if aperture settings aren't transportable across lenses then that could explain the problem in to OP and therefore isn't a derail.

 

[Hocus Eye.]

Searches vainly for a weary smilie.

 

[Herbsman.]

Wouldbe

You need to read some books, man.

 

[tom_craggs]

I haven't read through alll this, sorry, but perhaps qaulity of lens construction or glass? Mind you 2 stops seems a big difference.

 

[Paul Russell]

@ WouldBe

I'm less knowledgeable about this than Bosky or Hocus, but I think this is a summary of Herbsman's query. Forget what you know about cameras for a minute.

(1) People use incident light meters -- they work! They are not a con.

(2) Incident light meters measure the amount of light falling on the subject -- the photographer places the meter near the subject and points it in the direction of the camera.

(3) They give you the correct combinations of f-stop and apertures that will give you the correct exposure. You don't have to dial in the focal length.

(4) Hence Herbsman is wondering why he gets a different exposure/result with different lenses when using the same f-stop/aperture combination.

Possibly

 

[cybertect]

This is the source your confusion

Sorry that's bollox. A camera works only on the reflected light which will be different with different subjects and therefore require different exposures regardless of the light incident on the subjects. Again this is basic physics.

Why not try it. Place a light coloured piece of card and a dark coloured piece of card side by side and illuminate with a light source. Then set your camera up so that each piece of caed individually fills the viewfinder. use the cameras meter to determine the exposure and it will be different for each piece of card as more light will be reflected by the lighter card than the darker one.

The different tones of each card only makes a difference to the exposure you make if you want to render them both as an identical tone in the image you're making

i.e. if you have a piece of card that reflects 80% of the light shone upon it (which appears light to the eye) and another card that reflects only 10% of the light that is shone upon it (which appears dark to the eye).

Meters are calibrated to calculate an exposure renders faithfully 'mid-grey' (18% of the light is reflected). This is the constant reference point that allows you to largely disregard the tonality of the subject when reading a meter.

If you take your camera with built in spot metering and point it at a field of snow filling the frame, it will calculate an exposure that renders the snow as mid grey. Point it at a pile of coal in a similar way, it will render that coal mid grey. Neither of these is an accurate representation of the tones in the scene.

If you purchased a grey card from a photographic shop and then spot metered on that, then the camera can calculate the exposure correctly.

Other TTL metering schemes (matrix, centre-weighted, etc.) are attempts to compensate for this kind of error.

This is why an incident light meter, such as that being used by Herbsman, delivers an accuracy advantage over TTL metering.

 

[WouldBe]

(1) People use incident light meters -- they work! They are not a con.

I never said they weren't.

(2) Incident light meters measure the amount of light falling on the subject -- the photographer places the meter near the subject and points it in the direction of the camera.

While a camera works on reflected light.

(3) They give you the correct combinations of f-stop and apertures that will give you the correct exposure. You don't have to dial in the focal length.

Yet Hocus eye says they do and they don't which can't both be right. Hocus did mention earlier in the thread about inverse square law affecting the amount of light so a hand held meter should also take into account the distance from the subject to the camera as this would affect the amount of light reaching the camera and therefore affect the exposure.

 

[WouldBe]

If you take your camera with built in spot metering and point it at a field of snow filling the frame, it will calculate an exposure that renders the snow as mid grey. Point it at a pile of coal in a similar way, it will render that coal mid grey. Neither of these is an accurate representation of the tones in the scene.

That's never happened in any photo I've ever taken.

This is why an incident light meter, such as that being used by Herbsman, delivers an accuracy advantage over TTL metering.

An incident light meter can't take into account the reduction in light caused by any filters on your lense or if Hocus is right about his inverse square law the effect of any teleconverters or macro rings. So I can't see the use of a hand held light meter unles you are using a camera without a built in meter.

 

[zenie]

So I can't see the use of a hand held light meter unles you are using a camera without a built in meter.

Are you being serious?

Great thread by the way

 

[WouldBe]

Are you being serious?

Yes.

I mostly took landscapes. It would be a nightmare running backwards and forwards with a handheld meter.

 

[Blagsta]

If you take your camera with built in spot metering and point it at a field of snow filling the frame, it will calculate an exposure that renders the snow as mid grey. Point it at a pile of coal in a similar way, it will render that coal mid grey. Neither of these is an accurate representation of the tones in the scene.

That's what exposure compensation and the zone system is for!

 

[cybertect]

That's never happened in any photo I've ever taken.

You probably don't usually spot meter on snow or coal or take pictures of only snow or coal.

Michael Reichmann knows his stuff

http://www.luminous-landscape.com/tutorials/exposing_snow.shtml

The very reason that your camera is suggesting different EVs in your own example is because it's doing precisely this - attempting to render them to a mid grey.

Go outside tomorrow, aim your camera straight up and take a picture of the sky.

Does the resulting image accurately represent the brightness of the sky?

An incident light meter can't take into account the reduction in light caused by any filters on your lense or if Hocus is right about his inverse square law the effect of any teleconverters or macro rings.

True, an incident light meter can't take account of your filters and that's one of the advantages of TTL metering, along with convenience. It's also why filters usually have a rating on them to tell you how many stops to allow.

The convenience of TTL metering comes at a cost, though. It can and will be fooled by unusual subject tonalities. Attempting to shoot a black cat in white snow will probably result in noticeable underexposure. Wedding photographers have great trouble getting the exposure right when the bride is wearing a traditional white dress for similar reasons.

So I can't see the use of a hand held light meter unles you are using a camera without a built in meter.

See the Luminous Landscape link above. See also this for reference to wedding dresses.