Many readers of my blog use not auto-focal lenses, so I decided to share information on this subject with all to whom it can be interesting.
A light meter works regardless of the fact that transfers a dandelion. It influences only on flash-out of TTL system. With flash-out on Lushnikovsky adapter for Canon we have a problem there is no enough of light because it transfers
wrong value (only completely open diaphragm)
On the Chinese adapters for Canon (and other) it is possible to put a diaphragm on the camera. It transfers to the camera and inserted into Exif.
It is normal for manual lenses.
I put a diaphragm on a lens, and you can use it. I do it for tests.

A light meter

In general all of you correctly understood about a light meter. It is measured on the most open diaphragm and then it is multiplied by corresponding coefficient. Another thing is that, for example, Canon, it knows all lenses and respectively the camera identifies an auto-focal lens. For all such lenses the value of coefficients, most likely, are registered in a camera firmware.
What turns out with manual lenses? The camera perceives your closed diaphragm as open. It measures the exposure, multiplies on certain (in the Chinese adapters more often 50/1.4) coefficient and we receive the overexposed picture.

What here it is possible to do?

It is clear that the maximum speed in the adapter shall correspond in that case to a diaphragm on which there will be a picture. It is unreal i.e. a diaphragm on a lens you are free to close to convenient value and these values can be different. Quickly maximum registered speed in the adapter doesn’t change neither in a case with the Chinese adapter, nor in Lushnikov’s adapter.
And in general you already described all options.
I want to add only that all coefficients are sewn up in a firmware, the results with exposure can differ depending on a firmware and from the camera.
So on my Canon 5D Mark II there were the same problems, and with the professional camera Canon 1D mark II N there were any other problems. The exposure is measured correctly.

Outputs can be like yours about the screen, but I assume that it isn’t the screen, namely in system of exposure calculation.
1) I use the mode “M” with manual lenses and I am guided according to the histogram (I do corrections of exposure). It is not really quickly, but there are any error.
2) It is also possible to use flesh-meter. It will be more quickly, but the model/object shall be near you that you could approach, measure, and then depart and make a picture. With model it is very convenient.
3) To have a digital camera. The small cheap digital camera quite correctly measures exposure through a lens. This method is used often by fans, who want to shoot technique on a film technique.
Do you knew, that native Canon lenses give an exposure error? At least some…

The experimental data

The camera No. 1 — 1D mark II N
Professional reporting camera. Already morally sub-outdated, but all the same very good.

Canon 1D mark II N

Carl Zeiss Planar 50/1.4+ Chinese EMF adapter

f1.4 1.8sf8 4s

The camera No. 2 — 5D mark II

5D mark II

Canon 50/2.5

f8 1.13s

Carl Zeiss Planar 50/1.4+ Chinese EMF adapter

f8 5sf1.4 1.5 s

Carl Zeiss Makro-Planar 60/2.8+Lushnikov’s adapter

f2.8 0.8sf2.8 0.5sf 2.8 1.6sf2.8 1.100s


1. My results showed that exposure in case of the Chinese EMF adapter installation on the most open diaphragm is measured rather precisely on both cameras with different focus screens.

2. The big error turns out in the case of parameter setting of a diaphragm on the camera, corresponding to value on a lens.

3. In a case with Lushnikov’s adapter the design values of exposure when specifying in the camera of different focus screens differ more strongly. The only option is to use the standard screen and a diaphragm on the camera must match to value on a lens. Other lens also could affect on it (the Leitax adapter is set on 60/2.8, and Chinese adapter is on 50/1.4). Today I will try to check it on 85/1.4, where there is a Lushnikov’s adapter. Perhaps the maximum speed of a lens and focal distance influences on a calculation. I will add results to this article.

I checked on Carl Zeiss 85/1.4
The camera 5D mark 2, the lens Carl Zeiss 85/1.4 + Lushnikov’s adapter. AV shooting mode, a lens is exposed on F8, the camera is on F1.4 (which isn’t changed in this case value).
The result of exposure sampling — 1/6sec, the exposure is calculated correctly (it is slightly darkly, than necessary).
Camera 5D mark 2, lens Carl Zeiss 85/1.4 + Chinese EMF adapter. AV shooting mode, a lens is exposed on F8, the camera is on F1.4.
The result of exposure sampling — 1/4sec, the exposure is calculated correctly.
Output from this mini-dough: The information was confirmed — Lushnikov’s adapter needs to be programmed correctly, putting the maximum diaphragm according to a lens. It is possible to register the maximum diaphragm a little more dark for more correct calculation of exposure (because on all lenses it had slightly dark pictures).
I still will check data on the Chinese adapter, having tried it on less high-aperture lens.
4. Considering cunning calculation of exposure with a binding to the camera and a specific lens, it would be desirable to have opportunity to enter adjustment of exposure coefficients in the camera for different lenses as now it can be done for an auto-focus. But until such is realized, is it possible that the change (deception of the camera) of focal can help with the adapter? The test with other lens will respond to it.
5. Exposure by means of the LiveView mode is correctly measured. In this mode at first the frame looks overexposed, and then is gradually set up under illumination. In the AV and TV modes the camera uses change of ISO for this purpose. I.e. elementarily darkens or clarifies hardware finite picture therefore gives out always correctly exposed frame.