Equivalent focal length for MFT lenses

Started Apr 12, 2013 | Discussions thread
Anders W Forum Pro • Posts: 21,466
Re: Equivalent focal length for MFT lenses

Detail Man wrote:

draleks wrote:

Detail Man wrote:

draleks wrote:

Detail Man wrote:

draleks wrote:

Detail Man wrote:

draleks wrote:

There is a lot of talk about comparing lenses for MFT with APS or full format lenses. So I want to check if the following comparison is correct. If I have an MFT 300 mm f/6.7 lens, and a full format 600mm f/6.7 lens that is of a worse manufacturing quality and just happens to be twice as soft as the MFT lens.

You've scaled the Focal Length by the ratio of the Crop Factors (for FOV), but haven't scaled the F-Number by the ratio of the Crop Factors (for DOF). "Twice as soft" is a vague phrase. The spatial frequency response of subject-matter across the image-frame is different than Depth of Field.

I don't think this is how it works.

That is exactly how it works to achieve equal DOF and Total Light delivered to any two formats.

If you maintain equivalent total light delivered, then you will have to have four times higher ISO sensitivity on the FF than on on MFT. The FF looses at least some of its advantage.

Your statement (directly below) is correct. In equal DOF conditions, the Photon Shot Noise will differ by the square-root of the ratio of Quantum Sensitivities, and the Read Noise will depend upon the particular image-sensor design. Past that, your other "assumptions" seem tenuous.

Interesting point. Do you know which noise source dominates on a typical sensor?

It depends on the conditions. (Ignoring PRNU noise), they are equal in magnitude when then number of illuminating photons equals the RMS Read Noise squared divided by the Quantum Efficiency. In general, Photon Shot Noise dominates in magnitude for all but relatively low illumination situations. As ISO settings increase, the amplification of Read Noise levels [relative to the maximum conversion voltage of the Analog/Digital Convertor(s)] increases, however.

All of that correct as far as I can see. Let me just add a couple of things regarding the implications of what you say (for the benefit of the OP more than you, who would guess what I was about to say anyway ).

1. While photon noise is the dominant source of noise except in the fairly deep shadows, the shadows is at the same time the part of an image having the weakest SNR (signal-to-noise ratio). Consequently, the shadows is where the noise is most visible and most perceptually disturbing (to me and I think most others). It follows that although photon noise is the dominant source of noise if we generalize across all levels of illumination, read noise is anything but unimportant since it has a fairly strong impact on the technically weakest part of the image. In fact, much of the improvement in sensor performance that we have witnessed over the last five years or so result from a reduction of read noise rather than from an increase in quantum efficiency (although QE has certainly increased as well).

2. As ISO increases, the point at which read noise becomes the most important noise source comes closer and closer (as counted in EV) to the clipping point of the sensor. At base ISO, read noise dominates only in the deep shadows and will hardly be visible unless those shadows are pushed in post-processing. At higher ISO, by contrast, read noise will be dominant even in the not-so-deep shadows, where it accounts for much of what I (and I think others) find perceptually disturbing in high ISO shots.

Does read noise depend a lot on the pixel size? I've seen some people try to argue that sensor size is the only thing that matters for noise/dynamic range, but perhaps when you stack pixels too tightly (especially when pushed to do so by the marketing..) the pixels start to interfear with each other...

The reason that I keep it to, "the Read Noise will depend upon the particular image-sensor design", is from this kind of indication of how the matter is not as simple as photo-site size:


FF area is four times as large, so an FF sensor should be able to take equally "good" picture with four times less light per unit area.

An MFT f/6.7 will illuminate the sensor with the same amount of light per unit area as an FF f/6.7. The difference is of course that the FF area is four times as large, so an FF sensor should be able to take equally "good" picture with four times less light per unit area. So my idea is that:


1) The lens on the FF system is of worse quality then the MFT lens, so that the softness (and possibly other optical imperfections) are vaguely twice as large as the MFT lens.

and apparently also,

2) The sensor on the FF system is of worse quality (for example older), and has the same ISO capability as the MFT sensor,

then the IQ of those two system should be a very close match.

Nonetheless, the FF system will offer a more shallow depth of field.

That sounds like a lot of "ifs and buts", there. Yes, the FF DOF will be shallower (if the MFT F-Number is the same). My point was that if the MFT F-Number equals one-half of the FF F-Number (divided by the ratio of the Crop Factors), then the DOF will be the same (because the diameter of the Entrance Pupil, and thus the Total Light delivered, will be the same).

Depending on the shot DOF may or may not be an issue. The total light delivered is of course always important.

Not sure what it is that you are trying to say. Is there some doctrine that you are espousing ?

I meant to write that *shallow* DOF may or may not be diserable. So, yes, FF has lower DOF and sometimes it looks really professional, but sometimes you just want everything to be in focus.

That is the endless matter of personal aesthetics (which is not legislate-able). People refer to the ability of a camera system to adjust to shallower DOF as "DOF control". Smaller formats are limited in being able to reach those shallow levels as a result of the minimum F-Number of available lens-systems for those smaller formats. That is a particular hardware limitation.

Then, on equal ISO values, those two lenses will be an exact match. Right?

Shutter Speed is what you want equal. Exposure/Total Light are not functions of ISO Sensitivity.

 Anders W's gear list:Anders W's gear list
Panasonic Lumix DMC-G1 Olympus OM-D E-M5 Olympus E-M1 Panasonic Lumix G Vario 14-45mm F3.5-5.6 ASPH OIS Panasonic Lumix G Vario 7-14mm F4 ASPH +28 more
Post (hide subjects) Posted by
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark MMy threads
Color scheme? Blue / Yellow