70-200 VR f2.8 isn't f2.8

[michael_borland]

I updated the test shots to include white balancing from the NEFs. My measurements of the intensity are as follows:

 

Shot 1 (f4, 1/20s) : 179

Shot 2 (f2.8, 1/40s) : 151

Shot 3 (f2.8, 1/30s) : 170

Shot 4 (f2.8, 1/25s) : 189

 

I think it is pretty clear this isn't a true f2.8 lens.

[michael_borland]

Sorry, I made an upload mistake. Please look again if you looked between the time of this and the last posting.

[tonybeach]

For proper evaluation, please shoot in RAW and convert using neutral settings (no auto settings). It appears to me that the tonal curve on the f/4 and uncompensated f/2.8 shots are different.

[michael_borland]

Anthony,

 

I reposted the shots just a little while ago. These are made from the RAW files, with the white point set manually in post using the same location on the photos. The settings of the program were identical in all cases.

[rudymerz]

Michael,

 

I just did a similar test. The only difference I could find is the vignetting of the lens @ f/2.8.

But the differences are not as big as yours.

 

Rudy

[michael_borland]

Rudy,

 

Thanks for reporting your results. Seems that I need to send my lens to Nikon.

 

--Michael

[roland_vink]

I find a lot of my lenses underexpose when shot wide open, compared to equivalent exposures when stopped down. This is using center weighted metering on a film camera. I always figured it was due to mechanical vignetting, which disappears when you stop down.

[keith_lubow]

Bjorn, the original claim was that in actuality f/2.8 only provided 1/3 more light than f/4. If he had to add 2/3 stop, then stopping down was only giving him 1/3 stop.

 

Keith

[keith_lubow]

Michael, before you send your lens in, I would try it wide open with an out-of-focus grey card. Your target is near impossible to judge because you don't know its actual reflectance. You know a grey card is middle grey. The card should end up somewhere around 128 units of each channel.

 

While you know that f/2.8 is only providing 133% of what f/4 provides, you don't know which one is a reference point. It could be that f/2.8 is underexposing, it could be that f/4 is overexposing, and it could also be some of each. What does the difference between f/4 and f/5.6 look like?

 

In short, with your current target, we know that f/2.8 and f/4 are not calibrated properly to each other. However, we don't know which one is providing more or less light than it should until you shoot a grey card.

 

Keith

[keith_lubow]

Pardon me on that last sentence; stopping UP was only giving him 1/3 stop.

[brian_caldwell]

Regarding lenses that cause underexposure when used wide open:

 

Even for the pixel located exactly on the optical axis, light will strike at non-normal incidence. What that central pixel "sees" as it "looks" back toward the lens is a disk of light called the exit pupil. For large apertures that disk of light is big. Light coming from the edges of the disk (a.k.a. marginal rays) strikes the disk at an oblique angle.

 

For really fast lenses the effect can be huge. For example, the marginal ray angle of an f/1 lens is 30 degrees. Many sensors are almost completely insensitive to light incoming at 30 degrees off-normal, and so an f/1 lens would be virtually wasted with such a sensor. I've long suspected that this is the real reason why Canon cancelled their 50mm f/1.

 

In many cases the effect can be beneficial. For example, at largish apertures the sensor will essentially introduce apodization which reduces the intensity of the outer parts of defocused highlights, both foreground and background. For the most part this means improved bokeh. Minolta accomplished more or less the same thing years ago when they came out with a portrait lens having an apodizing filter placed near the aperture stop.

 

Some bokeh effects can be a little weird, however. The Nikon D1x had rectangular pixels, and the angular response was noticeably different in the horizontal and vertical directions. So, on several occasions shooting wide open at f/1.2 I noticed oblong defocused highlights even on-axis.

 

In the case of f/4 and f/2.8, the marginal ray angles are 7.2 and 10.3 degrees, respectively. Whether or not this can explain the exposure differences in the example shown depends entirely on the angular falloff characteristics of the sensor.

[andrew robertson]

Of course no f/2.8 lens is going to be t/2.8 as well. Some light is lost to transmission and

reflection. If we're talking about a 21 element lens with 15 air to glass interfaces, then

the light loss is going to be much more than say, a 8 element lens with 5 air to glass

interfaces.

 

Simple physics dictates that lenses transmit slightly less light than what enters, because all

photographic lenses lose light to absorption.

 

What's the mystery? A Noct-Nikkor is not both f/1.2 AND t/1.2. It is probably closer to

t/1.3 (or 1/2 stop lost to transmission). Century Precision Optics takes professional

Canon lenses and remounts them for cinema use. The optics remain identical, only the

mount and housing are changed.

 

Let's look at a few:

 

The TS-E 24mm f/3.5 is really a t/4 lens. The 200 f/1.8 is really t/2. The 28-70 f/2.8 is

really t/3. Canon's 300 f/2.8 and 400 f/2.8 are both true t/2.8 lenses.

 

So chalking all this up to differences in shutter speed or exactness of the aperture blades

(and I'm not saying those couldn't contribute to a loss somewhere) we should be realizing

that very few of our modern zooms that are jammed to the brim with as much glass as

possible lose some light to transmission.

[andrew robertson]

Some f/2.8 lenses are t/2.8, but not the zooms.

[keith_lubow]

Andrew,

 

The issue does not seem to be t stops vs f stops. If it was, then the issue would be taking effect across the entire aperture range; not just wide open. The issue is in the relationship between apertures; not in exactly how much or how little light is attenuated by the elements. In actuality, f/2.8 is only 1/3 stop brighter than f/4. 2/3 stop have vanished in to thin air.

 

Keith

[andrew robertson]

Not necessarily. The aperture stops may well have been calculated to give t values instead

of f values when stopped down, but if the lens is only t/3.5 they wouldn't want to actually

advertise this fact. This could be determined by measurement.

 

This leads to a whole different question: Do camera companies calibrate their lenses so

when stopped down, transmission and reflection losses are taken into account? In other

words, a lens with 2/3rds stop loss, say f/2.8 and t/3.5, if you stop it down from 2.8 to 4

on the aperture ring you are really stopping down only 1/3rd of a physical stop from "2.8"

to "4", and really going from t/3.5 to t/4 and f/2.8 to f/3.2, and then in whole stops

afterwards? The manufacturer calls a t/3.5 lens an f/2.8 lens, and all the meters line up

(except when the lens is wide open of course), and somehow tubes packed with several

pounds of glass lose no light to reflection or transmission, at least not that anybody can

tell. Unless they shoot wide open like our friend above.

[andrew robertson]

Bjorn, you said "The claim is that the 21 elements of the 70-200 lowers lens speed by 2/3

of a stop. The evidence to support this has been scanty."

 

From Erwin Puts' review of the Cosina Voigtlander 35 f/1.2:

 

"We may rationally ask why a 1.2 lens has any added value when compared to a 1.4 deign,

apart from the emotional arguments.

 

The jump form 1.4 to 1.2 is at best a half stop. I say ?at best? because of the tolerances

involved. A lens with a designation of 1.4 or 1.2 (or any other aperture) may depart 5%

from the nominal designation. Quite often a 2.8 lens is in reality a 2.9 (a 3% change) and if

we add the normal production tolerance of 5 to 10% from the nominal value, we may end

with a lens that has a practical aperture of 3.0. Current film material is very good and can

handle some under exposure and in any case our eyes will not detect a difference in

density that is below 10%. To be precise: our eyes can detect changes in density as low as

2%, but only when we compare adjacent areas with different densities. Consider this: the

maximum black in a high quality print has a reflection density of D-2.25. But most people

have already trouble seeing a difference in shadow density from D-1.95 to D-2.25. This is

a difference of more than 10%. It is a not-unreasonable assumption that the need for a

1.2 aperture is justified by the argument that at the limits of low light recording, every

additional photon could make the day. In this situation a 1.4 lens would generate

underexposure in the critical shadow densities. And the extra half stop would allow the

film to build up enough density to a level that is useful for recording differences in

shadows."

 

As Puts tells us, and any number of people here can probably confirm, a 21 element lens

is going to lose some light on the way through.