Exposure changes with focal length (EOS, 28-80mm lens)

elerion · Apr 3, 2017

Sometimes I use an EOS film camera with a 28-80 standard lens.
I get different exposure readings for 50mm than I get at full zoom (80mm).
I checked this against an adjustable light box used for transparencies, and thus, very uniform.
I used a tripod, and I filled the whole frame with the illuminated area.
Still, I get 1/2 stop drift towards longer exposures when going up 50 mm.
Does it make sense?

Leigh B · Apr 3, 2017

elerion said:
I get different exposure readings for 50mm than I get at full zoom (80mm). ...
Does it make sense?

Yes. In fact that's what I would expect.

Most zoom lenses, particularly consumer-grade, have variable apertures.
That should be marked on the lens, like f/2.8-3.5 or some such.

- Leigh

Ko.Fe. · Apr 3, 2017

Yes, EOS with kit zoom lens sound familiar. Isn't it 28-90? It is f4-5.6 lens. Longer you choose, smaller you'll get.

elerion · Apr 3, 2017

Yes, but I set the aperture at f8. It does not change from 50 mm to 80 mm. But exposure meter does.

Is it the internal lightmeter which measures wrong? or is it the lens?

I did a quick test, shooting at a gray card, through several light neutral density filters. I measured densities after developing. It did seem that the 80mm frame was less exposed (density was slightly less for some of the zones, but not all of them in the same deegree). As I couldn't get a conclusion, I decided to ask.

Luckless · Apr 3, 2017

Which specific model of lens is it? They made several revisions of the base lens, in addition to the L series model.

I would suspect the issue is the lens, and I'm curious to see what those who are familiar with the lens will have to say.

Light falloff in a zoom lens is not unusual, and it is one of the issues that the concept of "T-Stop" designed lenses are intended to address. In a mathematically perfect lens f/8 at 28mm will put the same amount of light per square inch on your film as f/8 at 80mm does, however it is not uncommon in lens design for this to fail slightly, and for longer focal lengths to lose a little more than expected and be darker. Additionally some zoom lenses may lose more/less light in the corners depending on what focal length you set them to, but I'm not familiar enough with Canon's film bodies and the metering systems used in them to comment much on what that may be able to do to the readings.

Depending on design, things like excessive dust on elements may make the issue worse. (If you have a patch of something disrupting the light path, and you 'zoom in', that disruption may now make up a greater percentage of the lens surface.) So it is worth taking a careful look at the lens to make sure you're not getting some delamination or something.

In general I chalk things like that up to equipment quirks, and just try to keep them in mind when shooting so that I can apply manual corrections. The important thing in my mind is whether or not the difference is consistent. Consistent errors can be corrected for easily enough with care and attention, it is the inconsistent errors that are a real pain.

Side note on testing lenses for cameras like an EOS - Since they're compatible with dSLRs as well, they can appear easy to test with one for some issues like this, however do remember to check for auto-adjustment profiles! Some brightness oddities are readily smoothed out during post processing in ways that aren't easy to accomplish in a film to paper workflow. That, combined with the ease of overlooking crop sensors, make for easily missed gotchas any user who might be coming over to explore film should keep in mind.

Alan Johnson · Apr 3, 2017

elerion said:
Sometimes I use an EOS film camera with a 28-80 standard lens.
I get different exposure readings for 50mm than I get at full zoom (80mm).
I checked this against an adjustable light box used for transparencies, and thus, very uniform.
I used a tripod, and I filled the whole frame with the illuminated area.
Still, I get 1/2 stop drift towards longer exposures when going up 50 mm.
Does it make sense?

No.
I tested my Rebel Ti + 28 -90 focussed on a line on a uniformly illuminated blank wall and get the exact same exposure at 28mm and 90mm.

Dali · Apr 3, 2017

elerion said:
Sometimes I use an EOS film camera with a 28-80 standard lens.

what lens is it exactly? If it is a variable aperture zoom (like so many), I would expect the exposure reading to change along with the focal length to compensate the change of aperture (the real one, not coming from the aperture ring.

Alan Johnson · Apr 3, 2017

The EOS lenses have automatic electronic aperture control, they are not like earlier manual zooms.
The section "What is stop down metering" gives a brief explanation, IDK of a full explanation:
Dead Link Removed

elerion · Apr 3, 2017

The lens is a Canon zoom lens EF 28-80 mm, 1:3.5-5.6.
I checked for damage and dust. None of them.

Alan Johnson · Apr 3, 2017

IMO you may have a lens/body that is not working properly, a rare but not completely unknown EOS occurrence.
If possible try a different lens and different body.

Leigh B · Apr 3, 2017

elerion said:
The lens is a Canon zoom lens EF 28-80 mm, 1:3.5-5.6.

That's a variable-aperture lens, as discussed above.

The aperture decreases at longer focal lengths, i.e. f/3.5@28mm and f/5.6@80mm.

- Leigh

Leigh B · Apr 3, 2017

Alan Johnson said:
IMO you have a lens/body that is not working properly, a rare but not completely unknown EOS occurrence.

His lens and body are operating exactly as designed.

Why would you claim otherwise? That's pure nonsense.

- Leigh

Alan Johnson · Apr 3, 2017

I think you will find that the EOS electronic aperture control should balance the aperture and shutter speed to give the same exposure at different focal lengths, as mine does,
which is not what the OP's camera/lens combination is doing.

tedr1 · Apr 3, 2017

elerion said:
Yes, but I set the aperture at f8

Doremus Scudder · Apr 4, 2017

elerion said:
The lens is a Canon zoom lens EF 28-80 mm, 1:3.5-5.6...

If the maximum aperture changes by 1 1/3 stops over the range of the zoom, so do all the other apertures. So, your "f/8" at 28mm turns into "f/12.5" or thereabouts.

Doremus

miha · Apr 4, 2017

Doremus Scudder said:
If the maximum aperture changes by 1 1/3 stops over the range of the zoom, so do all the other apertures. So, your "f/8" at 28mm turns into "f/12.5" or thereabouts.

Doremus

No, it doesn't, or at least it shouldn't.
That would be the case when the apperture is set to f/3.5 at the wide-end. It would close to f/5.6 at the long-end.

Something wrong with the body/lens as suggested by Alen in #10.

elerion · Apr 4, 2017

Doremus Scudder said:
If the maximum aperture changes by 1 1/3 stops over the range of the zoom, so do all the other apertures. So, your "f/8" at 28mm turns into "f/12.5" or thereabouts.
Doremus

That's not what I always thought.

Mr Bill · Apr 4, 2017

Doremus Scudder said:
If the maximum aperture changes by 1 1/3 stops over the range of the zoom, so do all the other apertures.

Yep, this is what I have observed.

I've looked at a couple dozen zoom lenses on a test fixture, measuring (approximately) focal length (via image magnification) and the apparent aperture diameter (viewed via short range telescope) to get actual f-numbers. They behave, approximately, like Doremus says.

The explanation, as I interpret things, is that the physical aperture is not being mechanically changed during zooming, and that the apparent size change is solely due to optical magnification. So a wide-open physical aperture should be magnified by the same factor as a stopped down aperture. Consequently, all aperture settings vary to the same degree.

When the aperture is electrically controlled through the camera, it's possible to correct for such variations. And ideally should behave as Alan says.

elerion · Apr 4, 2017

Ok, how to use an external lightmeter then?
If exposure is not only dependant on time and f-number,...

Mr Bill · Apr 4, 2017

elerion said:
Ok, how to use an external lightmeter then?
If exposure is not only dependant on time and f-number,...

Hi, I don't have a good answer. The issue is really in the lens, not the exposure meter nor (mostly) the f-stop system. (The f-stop system DOES have issues when focusing closely - corrections have to be made for lens extension.)

The ideal thing is to have a more expensive lens with a "constant aperture," and if you absolutely have to keep exposures nailed down this is probably the way to go.

Another method is to do an exposure test series to find the problem areas with your lens, and either avoid these zoom ranges OR make exposure corrections for them. As I recall, the lenses I looked at were mostly ok until getting into roughly the longest 1/3 of the zoom range.

Personally, for general shooting, I'd use a negative film, and just make sure my exposures were heavy enough that I don't get underexposed (the latitude is on the overexposed side). And avoid the longest zoom positions, if that's where the big difference is. In your case, though, there might be an issue with the contacts, or something, as it seems like the system ought to be correcting with the aperture position. Sorry I can't be more helpful than that.

Alan Johnson · Apr 4, 2017

By chance I have an unusually uniformly lit blank wall for my test, it is lit by reflection from the wall opposite of ceiling lights directed at this.At 28mm I get in about a 5ft width of uniformly lit wall. Just in case your light box is a bit small it might be worth repeating the test with wider uniformly lit area.

Mr Bill · Apr 4, 2017

Alan Johnson said:
Just in case your light box is a bit small it might be worth repeating the test with wider uniformly lit area.

That's a really good point.

And in fact, depending on how close the camera is to the light box, the difference might be real, due to lens extension. You should be able to rule this out by having the lens be focused at infinity when you take the meter readings.

elerion · Apr 5, 2017

I always focus at infinity. I get close enough for the lit area to fill the viewfinder by a safety margin.

I tried with another zoom lens. 70-300mm 1:4-5.6.
With this one, there's no exposure drift.

Strange. I tried both lenses on a dSLR, and I do get drifts with BOTH of them. Up to 2/3 stops.
Also, it is not a linear drift.
Imagine I get 1/1000 at f8, in the wide side.
I got 1/1250 when zooming a litte bit.
then I got 1/1600, then I got 1/1250 again in the tele side.
So it is not getting dimmer as focal grows. It get brighter, then dimmer.
It happened with both lenses on the dSLR, and only with the 28-80 on the film camera.

Could this be due to the aperture blades not being exactly opened at the ratio required to meet the current f-number? So at some focal lengths, the ratio between focal and real apeture is more precise than in other, and thus the difference in actual exposure.

Chan Tran · Apr 5, 2017

Doremus Scudder said:
If the maximum aperture changes by 1 1/3 stops over the range of the zoom, so do all the other apertures. So, your "f/8" at 28mm turns into "f/12.5" or thereabouts.

Doremus

I don't know about the Canon because I don't have one but on the newer Nikon even with a variable aperture zoom the aperture would still be be the same if you set it at midrange (i.e f/8 and not f3.5 of f/22). The camera knows this and actually set the diaphragm larger to compensate for this. At the extreme end of the aperture then the camera can't open or close the aperture any more than the it has to change the shutter speed.

Doremus Scudder · Apr 6, 2017

Chan Tran said:
I don't know about the Canon because I don't have one but on the newer Nikon even with a variable aperture zoom the aperture would still be be the same if you set it at midrange (i.e f/8 and not f3.5 of f/22). The camera knows this and actually set the diaphragm larger to compensate for this. At the extreme end of the aperture then the camera can't open or close the aperture any more than the it has to change the shutter speed.

If the OP's camera is (supposed to be) adjusting the aperture via its exposure circuitry to keep exposure constant, then the problem could well be with the camera body/control circuitry. I guess I'm behind the times when it comes to 35mm cameras; I haven't shot that format in years. My knowledge of variable-aperture zooms is from a time when there was no automatic aperture control.

If the exposure is being held constant by the camera, that means that the aperture is being opened as magnification increases, which keeps a shutter speed constant, but sacrifices depth of field. I don't know if I'd like that much.

I guess the OP needs to ascertain whether his camera is, indeed, supposed to be adjusting the aperture on his zoom lenses (or on some of them and then which) and find out if this feature, if present, is malfunctioning or not. Time to read the manual.

Best,

Doremus

Exposure changes with focal length (EOS, 28-80mm lens)

Privacy & Transparency

Privacy & Transparency