A lens that resolves fine detail does not necessarily produce a print that looks crisp, and a lens that looks crisp does not necessarily resolve the most detail. The two qualities are measured separately, and the gap between them is where much of a monochrome rendering is decided. The Modulation Transfer Function (MTF) is the single most useful tool for separating them, because it does not collapse a lens into one number. It describes how faithfully the lens transfers contrast across a range of detail sizes — but the lens is only one factor in a chain, and reading the chart well means knowing what each frequency band does and what the film and the eye do to it afterwards.

What the Curve Measures

MTF describes how much of an object’s contrast survives the trip through the lens to the image plane, as a function of spatial frequency. Contrast here is defined as modulation: for a pattern of light and dark, the modulation equals (Maximum − Minimum) / (Maximum + Minimum). A target with perfect black-and-white bars has a modulation of 1.0; after passing through the lens, the bars are rendered with reduced separation between light and dark, and the ratio of output modulation to input modulation is the MTF value at that frequency, a number between 0 and 1, or 0 and 100 per cent. Spatial frequency is given in line pairs per millimetre (lp/mm) at the film plane. Every real curve starts near 1.0 at very low frequencies and falls as frequencies rise, because finer detail is progressively harder to transmit.

Mathematically the MTF is the modulus of the optical transfer function, which is the Fourier transform of the line-spread (or point-spread) function — the image the lens forms of an ideal line or point. ISO 9334:2012, Optics and photonics — Optical transfer function — Definitions and mathematical relationships, codifies that terminology; ISO 9335 sets out the measurement procedures. One caveat keeps the curve from being a clean monotonic description: in practice an MTF curve can fall to zero and rise again. This is spurious resolution, where structures finer than the zero-crossing are reproduced with black and white interchanged. Published curves do not show it, but it matters wherever the lens is defocused or the subject is moving.

Reading a Real Chart: the Summicron-M 35 ASPH.

Take Leica’s own datasheet for the Summicron-M 35 mm f/2 ASPH. — seven elements in five groups, eleven aperture blades, minimum aperture f/16, optimum performance stopped down to f/4. Leica plots its MTF in white light at four frequencies, 5, 10, 20 and 40 lp/mm, with the solid line for sagittal (radial) structures and the dotted line for tangential, evaluated both at full aperture and at f/5.6. Leica states plainly that the 5 and 10 lp/mm curves give the contrast for large object structures, while 20 and 40 lp/mm record the resolution of finer and finest detail.

So read it as two questions. The centre at 10 lp/mm tells you the overall snap between tonal masses — a wall against its shadow, a face against the sky. On a modern design like this you expect that to sit high, in the 80–90 per cent band usual for contemporary lenses, against only 60–70 per cent for the fast standard lenses of the 1960s. A high value here reads as a print with body: deep, clean separation between large areas of tone. The edge at 40 lp/mm tells you whether the weave of a tweed jacket or the lashes at the corner of a frame survive as distinct texture or dissolve into grey. Where the sagittal and tangential lines diverge at the edge, the lens has astigmatism: the point of light is stretched into a short line, radially or tangentially, so edges running one way stay crisp while edges crossing them smear. In monochrome that reads as direction-dependent loss of fine detail and uneven corner texture — the print looks sharp on a vertical railing and soft on the horizontal courses of brick beside it.

Sharpness Is Edge Steepness, Not Peak Resolution

Apparent sharpness tracks the low-to-mid frequencies, not the highest resolvable line. Nasse, in the Zeiss monograph How to Read MTF Curves (December 2008), explains the mechanism through the edge profile. In a very good 35 mm lens, a white-to-black edge transition is no wider than about 10 micrometres, and that steep transition is what the eye reads as sharp. A poorer lens spreads the same transition over 30–50 micrometres; it still reaches a deep black eventually, so its low-frequency MTF can stay high, but its high-frequency MTF collapses and the edge looks soft. This is why two lenses with similar ultimate resolution can render with entirely different character.

Nasse’s rules for weighing the differences follow from this. Small differences in higher MTF values matter most at high object contrast; tonal variations of less than about one aperture stop do not need high MTF, and differences above 70–80 per cent are hardly relevant; and where the MTF is already very low, image contrast stays low no matter how contrasty the subject. The upshot is that chasing the last few per cent at 40 lp/mm is rarely worth it, while the value at 10 lp/mm earns its keep on almost every frame.

Brilliance and Microcontrast Are Not the Same Thing

The word microcontrast is the most abused term in lens talk, and Nasse warns the two ideas behind it are constantly mixed up. Macro contrast is the brilliance of the image — the overall freedom from veil. It is governed by stray light: veiling glare and internal scattering off lens surfaces and the inside of the barrel, lifting the blacks with a thin grey wash. Micro contrast is the contrast of fine structures we can just about see, or just cannot — the small-scale correction the high-frequency MTF measures.

The distinction has a practical edge for the monochrome worker. The “luminous negative with deep blacks and presence” is largely a brilliance property: it comes from a lens, hood and coating that suppress scatter, and it is not captured by the MTF curve at all. Good low-frequency MTF is necessary for that look but is no guarantee of it — a well-corrected lens shooting into the light with a smeared front element will still post a fine chart and print like fog. So when a print snaps, credit the contrast curve for the tonal separation, but credit the coatings and the lens hood for the clean blacks.

The Lens Is Only Half the Chain

The MTF you actually print is the product of every stage: lens × film × enlarger lens × the eye. For a good 35 mm lens on a high-resolution black-and-white film, the high-frequency end of that product is limited by the lens, not the film. Nasse uses Kodak T-Max 100 as his example: its published MTF stays above 100 per cent out to roughly 20 lp/mm — a low-frequency adjacency rise characteristic of T-grain emulsions — before falling, and holds enough contrast at high frequencies that the film is not the limiting link. T-Max 100’s resolving power is quoted at two target contrasts because no lens delivers the high-contrast figure for the finest structures: 63 lines/mm at a low-contrast 1.6:1 target, 200 lines/mm at a high-contrast 1000:1. Estimating real-world performance from that 200 figure, Nasse notes, is too optimistic.

Two limits sit beyond the lens and the film. The eye resolves only about 8 lp/mm at the 25 cm least distance of distinct vision; referred back to a 24 mm picture height that is roughly 66 lp/mm on the negative, so the frequencies that matter to a viewer fall in the range up to about 40 lp/mm — which is exactly why datasheets stop there. And diffraction sets the physical ceiling: as a rule of thumb the point-spread width in micrometres roughly equals the f-number, and the diffraction-limit frequency is about 1500 divided by the f-number, so f/2 allows roughly 750 lp/mm but f/16 only about 94, where the Airy disc has grown to some 16 micrometres. This is why the Summicron peaks at f/4 and loses fine resolution again if you stop down hard.

The Darkroom Lever

The reading pays off at the enlarger. Those 1960s fast lenses at 60–70 per cent MTF at 10 lp/mm were not unprintable; workers compensated for the low contrast by enlarging onto a harder, higher-gradation paper grade to put the snap back. A modern high-MTF lens hands you the opposite freedom: the contrast is already on the negative, so you can print on a softer grade for the same apparent punch while keeping more tonal latitude in the highlights and shadows. (It is often argued that colour film, with its far less flexible processing, pushed lens makers toward better contrast correction in the first place.) Anchor it in a real process — T-Max 100 rated at EI 100, developed in D-76 stock for 6.5 minutes at 20 °C, fixed and washed — and the lens, the film and the paper grade stop being separate gear arguments and become one tonal decision. Interpreting a lens through its full MTF, and through the chain it sits in, is the most reliable way to predict how it will render a subject in black and white.