Modulation transfer function

The modulation transfer function (MTF) is a measure of how well an optical or imaging system carries contrast from an object to its image as the detail in that object gets finer. Edmund Optics describes it as a lens's "ability to transfer contrast at a particular resolution from the object to the image". For each spatial frequency, the MTF gives the ratio of image contrast to object contrast, a number between 0 and 1 (or 0 and 100%), and plotting these ratios against spatial frequency gives a curve. Edmund Optics describes resolution and contrast as the two components of the MTF.

In virtual reality and augmented reality, MTF is used to specify and test headset lenses, waveguide combiners and complete near-eye displays. Meta uses MTF as its measure of lens sharpness and published center and edge MTF values for the Meta Quest Pro optics in a 2022 blog post. Researchers at the US Food and Drug Administration have published MTF measurement methods for VR and AR head-mounted displays (HMDs) as regulatory science tools.

Reviewed 15 October 2026. Definitions, history, Meta Quest Pro MTF figures, FDA tools, IEC standard previews and every cited paper checked against its full text or abstract. About review dates.

Definition

In optics, the contrast (also called modulation) of a periodic pattern of bright and dark stripes is defined from its maximum and minimum intensities as (Maximum - Minimum) / (Maximum + Minimum). When a lens images a stripe pattern, light from each point spreads into a small blur, the point spread function, so some light from the bright stripes falls into the dark ones. The image therefore has less modulation than the object. Dividing the modulation of the image by the modulation of the object gives the modulation transfer at that pattern's fineness.

Because a single pattern says little about a lens, the measurement is repeated for patterns of different fineness. The fineness is the spatial frequency: the number of periods (line pairs, each one dark and one bright stripe) per millimeter in the image plane, given in line pairs per millimeter (lp/mm). The sequence of modulation transfer values plotted against spatial frequency is the modulation transfer function. Coarse patterns are reproduced with high contrast; as the stripe spacing approaches the size of the point spread function, contrast falls.

Mathematically, the MTF is related to the optical transfer function (OTF), which is the Fourier transform of the point spread function. Combined with the phase transfer function, the MTF makes up the OTF. In practice the MTF is often computed from a line spread function, the image of a thin line, as the normalized fast Fourier transform of that profile.

Key properties

Property	Description
Diffraction limit	Even an aberration-free lens has a cutoff spatial frequency set by diffraction. Edmund Optics gives it as 1 / (wavelength x f-number).
Resolving power	Zeiss calls the frequency at which the MTF reaches zero or drops below a low threshold (for example 10%) the resolving power of the lens, but notes that it is hard to measure precisely and is "not a suitable quality criterion for lenses".
Multiplication	The MTF of a complete imaging chain is approximately the product of the MTFs of its parts; Zeiss counts the human eye among those parts. Edmund Optics states that the system MTF "is the product of all the MTF curves of its components".
Orientation	A point spread function is not necessarily circular, so a lens can have different MTF curves for stripes in the sagittal (radial) and tangential (meridional) orientations at the same image point.
Field dependence	An MTF curve is valid only for one spot in the image, so a lens is described by curves at several positions in the field.

Zeiss also cautions that a single number such as the resolution limit or the frequency at which the MTF reaches 50% has "little value"; the MTF should be considered at several spatial frequencies. It also notes that digital edge enhancement can raise apparent sharpness without increasing resolution of detail, so that a processed system's transfer function may rise with frequency.

Units in head-mounted displays

Camera and lens MTF is usually expressed in line pairs per millimeter in the image plane. Meta, for example, states its Quest Pro lens MTF figures at 5 lp/mm. Measurements of complete headsets usually express spatial frequency in angular units, cycles per degree of the visual field, which ties MTF to pixels per degree (PPD) and to human vision.

History

According to a 1977 history by Jacques Simon of the Institut d'Optique, P. M. Duffieux proposed applying the transfer function concept to optics in 1946, and H. H. Hopkins, W. H. Steel, Schade, Linfoot and Maréchal completed the theoretical framework in the following years. Simon writes that later work focused more on experimentation and measurement; his own team at the Institut d'Optique had worked on MTF measurement since 1962, leading to a measuring instrument, the ACOFAM, manufactured by Matra.

Otto H. Schade of the RCA Tube Department is described by the IEEE's Engineering and Technology History Wiki as "known most widely" for developing the modulation transfer function concept. He worked on a unified method of image analysis and an accurate method for calculating the resolving power of television and photographic systems, and received the IEEE Vladimir K. Zworykin Award in 1969.

For digital cameras, ISO 12233, first issued in 2000, specifies methods for measuring the resolution and spatial frequency response (SFR) of digital cameras, and its methods include a slanted-edge gradient technique that captures a super-resolution edge profile.

Applications in VR and AR

Headset optics

Xiong and colleagues, in a 2021 review of AR and VR displays, write that matching a high-resolution display needs "high-quality optics to support the corresponding modulation transfer function (MTF)". The same review notes that diffraction artifacts and stray light from the grooves of Fresnel lenses can degrade image quality, or MTF, and that geometric (reflective) waveguides usually give better MTF and color uniformity than diffractive waveguides. It also lists MTF among the AR parameters that depend mainly on the combiner optics rather than the light engine.

In a December 2022 post, Meta defined sharpness as "measured by a lens's modulation transfer function, or MTF" and stated that the Meta Quest Pro pancake optics reach a center sharpness of 0.98 and an edge sharpness of 0.85, measured at 5 lp/mm, which Meta described as 25% and 50% improvements over the Meta Quest 2. Meta did not publish the Quest 2's absolute MTF values in that post. Lens designers also optimize MTF directly: a 2024 Optics Letters paper by Youngmo Jeong and colleagues proposed a pancake lens designed for higher MTF in the foveal area and for stable image quality when the user's eye rotates.

Displays as sampled systems

Jiawei Lu, Trisha Lian and Jerry Jia of Meta Reality Labs argued in a 2023 paper that pixels per degree "alone is not a reliable predictor" of the high-resolution experience in VR and AR, because display fill factor, pixel arrangement, optical blur and rendering also matter. They proposed a "system MTF" that combines these factors. Because displays and camera sensors are pixelated sampling systems, the shift invariance assumed by classical MTF does not hold, and their model uses a spatially averaged impulse response to handle this. They also note that MTF multiplies through a system: in a mixed reality headset the overall MTF is the display module MTF multiplied by the camera module MTF, where the display module MTF is itself the product of the VR lens and display MTFs. In their perceptual study on text quality, a change of 0.05 in system MTF corresponded to about one just-objectionable-difference (JOD) unit, and raising resolution from 25 to 30 PPD improved perceived text quality by about 3 JOD, compared with about 1 JOD for an increase from 50 to 60 PPD.

Measuring complete headsets

The near-eye optics of a head-mounted display add challenges to line spread measurements that are not present with ordinary displays. Ryan Beams, Brendan Collins, Andrea S. Kim and Aldo Badano of the FDA's Center for Devices and Radiological Health compared two ways of measuring the line spread response of the HTC Vive and HTC Vive Pro across a 30 degree range of the field of view. Imaging the whole field onto a stationary camera overestimated spatial resolution by about 40% at angles larger than 10 degrees compared with rotating the camera about an axis behind its lens, which mimics the human eye. They found that pixel density limits resolution at the center of the field, while the optics limit it off-axis.

The FDA regulatory science tool RST24MX02.01 (published May 2024), which lists this paper among its supporting publications, renders a single-pixel line, captures it with a high-resolution monochromatic camera-type measuring device on a 5-axis goniometer or robot, and derives the MTF. The tool specifies an eye-rotation geometry with the rotation axis about 12 mm behind the entrance pupil, recommends at least 20 measuring-device pixels per HMD pixel, and was validated on the HTC Vive, HTC Vive Pro and Microsoft HoloLens 2. A companion tool for optical see-through AR headsets (RST24MX06.01, October 2025) uses a 3 x 3 array of dots and point spread function analysis to derive a two-dimensional MTF; it was validated on the HoloLens 2, Epson Moverio BT-300 and Magic Leap 2 and is not intended for VR or video see-through headsets, whose subpixel structure affects the results.

Published results show how display technology shapes MTF. In a 2022 Scientific Reports study, Chumin Zhao and colleagues measured line spread functions and MTFs at the center of the field of the HTC Vive, HTC Vive Pro and HTC Vive Pro 2. With the PenTile OLED panels of the Vive and Vive Pro, the cutoff frequencies for red and blue lines were about 6 cycles per degree (Vive) and 8 cycles per degree (Vive Pro), and for white lines about 9 and 12 cycles per degree respectively. The Vive Pro 2's LCD (2448 x 2448 pixels per eye) improved MTF above 10 cycles per degree but had weaker low-frequency performance below 5 cycles per degree, which the authors suggested may come from the additional optical components and layers of the LCD; its higher fill factor avoided the screen door effect seen on the two OLED headsets. The study also modeled a "dynamic MTF" during smooth-pursuit eye movement and concluded that keeping the display emission duty cycle below 20% helps reduce motion blur.

Later work by the same research group tied MTF to perception. A 2024 Optics Express paper by Zhao and colleagues measured headset MTF with pupil rotation, eye rotation, and eye rotation plus angle kappa (the offset between the eye's optical and visual axes), fed the results into a human contrast sensitivity model, and compared the predictions with observers using a WebXR test platform. Image quality fell toward the periphery of the field of view, especially for users whose interpupillary distance differed from the headset setting, and the eye rotation with angle kappa setup correlated best with the human data. A related 2024 study on a Meta Quest 2 found that monocular contrast sensitivity degraded above 4 cycles per degree when gazing at the periphery, consistent with optical bench MTF measurements, especially with mismatched IPD settings.

For AR headsets, Johnson, Zhao, Varshney and Beams reported in 2024 that the measured spatial resolution of an AR HMD depends on three components (optics, display and the rendering pipeline, including anti-aliasing). Using point spread function analysis, they found temporal and multisample anti-aliasing performed close to no anti-aliasing, while fast approximate anti-aliasing changed the results, and they measured significant non-uniformity of the optical contribution across the field of view.

Measurement equipment

Every component of a measuring system has its own MTF, which contributes to the measured result. A 2022 SPIE paper by Tobias Steinel and colleagues compared scanning goniometric MTF measurement of near-eye displays (narrow field of view, 300 pixels per degree) with one-shot wide-field imaging (30 pixels per degree) on LCD and OLED VR headsets, and examined how to remove the instrument's MTF contribution. A 2024 SID paper by Daniel Winters and colleagues noted that near-eye displays project virtual images carrying amplitude and wavefront phase information, and that poorly designed MTF test systems can introduce measurement errors that cannot be corrected.

Standards

Standard or tool	Scope relevant to MTF
ISO 12233	Resolution and spatial frequency response of digital cameras, including a slanted-edge method.
IEC 63145-20-20:2019	Image quality measurement methods for non-see-through (VR) and see-through (AR) eyewear displays using virtual image optics; includes Michelson contrast measured with dedicated test patterns.
IEC 63145-22-20:2024	Image quality of AR eyewear displays; its static image resolution section includes a figure of the variation of Michelson contrast (luminance modulation) with line width.
FDA RST24MX02.01 (2024)	Line spread function and MTF of VR and AR head-mounted displays.
FDA RST24MX06.01 (2025)	Point spread function and two-dimensional MTF of optical see-through AR head-mounted displays.

See also

• Pixels per degree
• Resolution
• Optical aberration
• Pancake lenses
• Fresnel lens
• Screen door effect
• Eye box
• Near-eye display