JSD4012 1.05mm M7 Lens for 1/4 inch Arrays: 8.69mm, 145 deg
JSD4012: An 8.69mm TTL M7 Lens for 1/4" Sensors in 360° Array Cameras
Choosing between a 1.05mm focal length, 145° diagonal coverage, and a φ10mm barrel — and the one constraint that decides the part before samples ship.
Key takeaway: A 360° array camera stacks four optical channels around one housing, so total track length and barrel diameter — not field of view alone — decide what is buildable. The JSD4012 is a 1.05mm, F2.4 M7 lens with a 145° diagonal field of view, a Ø4.8mm image circle for 1/4" sensors, and an 8.69mm total track length, which leaves 2.31mm of margin under an 11mm ceiling. The stack is 1G3P with a polymer asphere in the optical path, so programs carrying a hard no-plastic rule should read the material section before committing to a standard part.
Why 360° Arrays Break Standard Lens Choices
Four lenses sharing one sensor board pattern, one mechanical datum, and one stitching model are constrained by the geometry they have in common. Each channel has to reach far enough sideways to overlap its neighbour, and each barrel has to sit close enough to the next for the ring to fit inside the product envelope.
Those two requirements pull in opposite directions. More overlap means a wider field of view, and a wider field of view on a 1/4" format is normally bought with a shorter focal length and a steeper projection curve. The ring geometry, in turn, caps how much barrel diameter and axial depth the design can absorb — which is why the smallest array products use M7 and M8 threads, where an M12 barrel would consume too much of the ring.
Two mistakes show up repeatedly at this stage, and both cost less to catch on a drawing than in a tooled housing.
The first is quoting field of view on the diagonal alone. Four-channel stitching is a horizontal problem: seams close or fail on how far each channel reaches across the horizontal field and how much overlap survives at the frame edges. A lens quoted at 145° diagonal may deliver 123° horizontally and 100° vertically, and the horizontal figure sets the channel count and the overlap budget. Designing to one number while quoting another is how a four-lens ring turns into a five-lens ring.
The second is treating total track length as a preference. Four lenses mounted on one board share a single housing height, so an extra three millimetres of axial length lands on the whole assembly and usually forces a redesign of the sensor holder that every channel bolts to.
"In an array, the lens that sets the track length sets the product. Four channels share one housing height, so the shortest part in the ring is the only one the enclosure has to accommodate."
— JSD Optics application engineeringField of view and distortion come after the sensor format and the mechanical envelope are fixed, not before.
Five Numbers That Decide the Fit
A 1/4" sensor such as the Sony IMX219 sets the optical target size; the array housing sets everything else. Five figures decide whether a candidate lens survives design review, and they are cheapest to check in this order — before tooling and before sample quantities are committed.
1. Image circle versus sensor format
The lens image circle has to cover the sensor's active diagonal. On a 1/4" format the working figure is roughly Ø4.6mm, and parts specified between Ø4.6mm and Ø5.3mm are the ones to compare. The margin above the sensor diagonal is not spare capacity: it absorbs die placement offset, holder eccentricity, and lens-to-sensor tilt. A part whose image circle equals the sensor diagonal on the drawing has spent that margin before the first unit is assembled.
2. Total track length against the housing ceiling
Total track length runs from the front of the lens to the image plane. Where the brief caps it below 11mm, that ceiling works as a filter rather than a target. A part at 8.69mm leaves 2.31mm of axial room for the holder stack and alignment features; a part at 10.7mm leaves 0.3mm, which is where shims and thread tolerances start driving the design.
3. Mount thread and barrel diameter
M7 and M8 threads keep the barrel near φ8mm to φ10mm, which is what makes a four-channel ring mechanically possible. Thread pitch matters as much as thread size — an M7×P0.35 barrel needs a matching holder, and the same optical part is often available with a different barrel to suit a different thread.
4. Field of view, stated in all three axes
Field of view belongs in the brief as a diagonal, horizontal, and vertical triple. Covering 360° with four channels is a question of horizontal reach rather than a large diagonal, while the vertical figure sets the vertical overlap the stitching model depends on.
5. Distortion profile and its measurement conditions
Distortion moves the array budget from optics into software. A fisheye-class projection keeps the pixel-to-angle mapping predictable and lets a stitcher work from a known model; a lens with a mild but unreported curve forces calibration per unit. Distortion figures are comparable only when the test target size, working distance, and quoted field axis match, so the conditions belong in the same request as the number.
| Selection check | Target for a 1/4" array channel | Why it matters on the line |
|---|---|---|
| Image circle | Ø4.6mm or larger | Below that figure the corners vignette before tolerance stack |
| Total track length | Under 11mm, with room to spare | Four channels share one housing height |
| Mount / barrel | M7 or M8, φ8–φ10mm | Decides whether four barrels fit in the ring |
| Field of view | Read D, H, and V separately | Horizontal reach sets seam overlap |
| Distortion | Quoted with test conditions | Drives how much unwarping sits in the pipeline |
The JSD4012 Envelope, Measured
The JSD4012 is a fixed-focus M7 lens built for 1/4" image sensors. It pairs a 1.05mm effective focal length with an F2.4 aperture, and it reaches a 145° diagonal field of view from a barrel that is φ10mm across.
Two of its figures carry the array fit. The 145° diagonal / 123° horizontal / 100° vertical coverage gives a stitcher enough horizontal overlap to close each seam without adding a fifth channel, and the 8.69mm total track length keeps the shared housing shallow enough to leave room for the holder and alignment features common to all four channels.
Specifications at a glance
| Parameter | JSD4012 | Design meaning |
|---|---|---|
| Effective focal length | 1.05mm | Fisheye-class reach on a 1/4" format |
| Aperture | F2.4 | Balances low-light capture against depth of focus |
| Field of view (D/H/V) @1/4" | 145° / 123° / 100° | Sufficient horizontal reach for four-channel overlap |
| Distortion | <-30.55% (fisheye projection) | Predictable angular mapping for the stitching model |
| Optical construction | 1G3P | Glass element holds focus stability; moulded aspheres fold the correction into a short stack |
| Mount / barrel | M7 (S-mount) × P0.35 / φ10mm | Threaded barrel allows back-focus tuning at assembly |
| Total track length | 8.69mm | 2.31mm of margin under an 11mm ceiling |
| Image circle | 1/4" (Ø4.8mm) | Covers the 1/4" diagonal with a small tolerance margin |
| Supported resolution | True 2MP | Set against the sensor mode the array will ship in |
| Focus | Fixed focus, thread-adjusted | No actuator to qualify, but back focus is set at assembly |
The mechanical drawing carries one number the specification table does not. Because the M7 barrel is threaded across its full engagement length, back focus is set by rotating the lens in its holder — a way to absorb board thickness and holder machining tolerance, at the cost of focusing all four channels individually and locking them rather than fixing them on one common datum. That assembly step belongs in the fixture design, not in a line-side discovery.
Where Each Candidate Fits
An array program usually evaluates two or three parts rather than one, because whichever direction gets relaxed — field of view, track length, or resolution class — changes which lens is right. Three decision patterns cover most of what comes up at prototype stage.
Tightest housing, widest field
When the enclosure binds and the array still needs wide horizontal reach, the short-track wide-angle parts are the ones to test. The JSD4012 at 8.69mm is the shortest of the three, with the JSD4011 at 9.29mm and the JSD5500 at 10.0mm behind it on different field-of-view and distortion profiles. Testing all three settles whether the extra reach is worth the extra millimetre.
Resolution-led channels
Where the sensor runs near its native resolution and corner contrast leads the criteria, resolution class moves ahead of track length in the decision order. This range does contain 5MP parts, and the section below shows where they sit on track length.
Geometry-critical output
Arrays used for measurement rather than viewing care about the pixel-to-angle mapping more than about coverage. There, a lower-distortion part with a narrower field can be the better answer: the calibration model stays simpler, and overlap is bought with more channels instead of more distortion.
Choosing Between 1/4" M7 Lenses
The three parts below all fit a 1/4" format on an M7 or M8 thread with a total track length under 11mm. They differ in where the compromise sits, and the split normally follows whether the program is coverage-bound or resolution-bound.
| Consideration | JSD4012 | JSD4011 | JSD5500 |
|---|---|---|---|
| Effective focal length | 1.05mm | 0.95mm | 1.70mm |
| Aperture | F2.4 | F2.0 | F2.4 |
| Field of view (D/H/V) @1/4" | 145° / 123° / 100° | 160° / 120° / 100° | 166° / 126° / 92° |
| Distortion | <-30.55% | <-8.5% | <-19.13% |
| Optical construction | 1G3P | 2G2P | 2G3P |
| Mount / barrel | M7 / φ10mm | M7 and M12 / φ10mm | M7, M8 and M12 / φ10mm |
| Total track length | 8.69mm | 9.29mm | 10.0mm |
| Image circle | 1/4" (Ø4.8mm) | 1/4" (Ø5.1mm) | 1/4" (Ø4.6mm) |
| Supported resolution | True 2MP | VGA | 5MP |
| Best suited to | Shortest track, moderate fisheye curve | Widest field at low distortion | Highest resolution in the M7/M8 short-track class |
Two axes bind an array, and the table separates them cleanly. On track length, the 1.31mm spread between the JSD4012 and the JSD5500 is the difference between a housing with tolerance margin and one that is dimensionally tight on paper. On resolution the gap is wider: the JSD5500 is specified at 5MP against True 2MP for the JSD4012, so a program planning to run an 8MP-class sensor near its native mode has to decide which of the two it is trading away.
All three share the φ10mm barrel and the M7 thread, so one holder design carries any of them through evaluation. Nothing has to be locked before samples arrive, which suits a prototype build running four channels per part.
Materials, Temperature, and Trade-offs
Materials decide more array programs than any single optical figure, because material requirements can be physically in conflict with each other rather than merely tight.
What the no-plastic rule actually covers
A no-plastic requirement arrives in one of three forms, and each leads to a different part of the catalogue. One form covers mechanical components only — barrel, holder, mount — which is a question about the exterior build. A second covers the optical elements themselves, which is a question about the glass stack. The third covers both. Writing down which one applies changes the candidate list rather than narrowing it.
Why the short-track wide-angle class uses polymer aspheres
There is a physical reason the shortest wide-angle parts in this format are built with moulded polymer aspheres. A glass element that corrects the aberrations of a wide, fast, 1/4" design needs several elements and the axial room to place them. Moulded aspheres fold that correction into a shorter stack, which is what an array housing under 11mm demands. The tension sits in the optical design, not in the catalogue: relaxing one requirement buys the other.
The JSD4013 is a 1G3P construction, with the glass element carrying focus stability and the polymer aspheres holding the track at 8.69mm. Where the mechanical build is the part that has to be plastic-free, that trade generally fits. Where the optical elements themselves have to be all-glass, no standard 1/4" M7 lens currently combines an all-glass stack, a total track under 11mm, and 145° diagonal coverage — and the table below sets out what the alternative costs.
The all-glass starting point
The closest all-glass standard part on an M7 thread for this format is the JSD4013, a 5G construction at 130° diagonal / 101° horizontal / 90° vertical with a Ø5.3mm image circle. Its single difference against a sub-11mm brief is total track length: 11.91mm, which is 0.91mm over the ceiling. Under a millimetre of axial room against a zero-compromise glass stack is a decision for the enclosure rather than for the optics.
| Program priority | Standard part to evaluate | What it costs |
|---|---|---|
| Shortest track, widest field | JSD4012 (1G3P, 8.69mm) | Polymer aspheres in the optical stack |
| All-glass optical stack | JSD4013 (5G, M7) | Total track length of 11.91mm, 0.91mm over an 11mm ceiling |
| Neither can move | Custom optical design | Focal length, chief ray angle matching, and mechanical geometry specified to the sensor |
Temperature, and what can be promised today
Operation above 100°C is an application requirement that no optical stack answers on its own. Barrel and holder material, adhesive selection, focus shift with temperature, and the thermal expansion of the sensor board all contribute, which makes an all-glass stack a necessary condition rather than a sufficient one. JSD Optics treats temperature behaviour as a measured property of a specific configuration — lens, holder, and adhesive — rather than as a catalogue rating, so the working temperature, duration, and cycle count belong in the brief where the optical and mechanical design can be scoped against them.
Resolution, stated against the sensor mode
A 1/4" sensor such as the IMX219 is an 8MP-class device, while the short-track wide-angle class in the standard catalogue is specified up to 5MP. Run a lens above the resolution class it was designed for and the centre of the frame holds up while the corners lose contrast and colour separation first. That trade is better made deliberately: either set the array's working mode to match the optics, or scope a custom design that starts from the sensor and the target resolution.
How to Brief This Project
JSD Optics designs and manufactures M7, M8, M12, and M16 imaging lenses, and works sensor-first: the sensor and the application define the image circle, resolution class, and mount, and the optics are matched to that brief.
Four inputs move an array inquiry from a catalogue scan to a definite answer:
Sensor part number and working mode — the optical format sets the image circle, and the readout mode sets the resolution class the optics have to belong to.
Mechanical envelope — the maximum total track length and barrel diameter the housing can accept, and the thread the holder is designed around.
Field of view in all three axes — diagonal, horizontal, and vertical, together with the overlap the stitching model needs between adjacent channels.
Material and temperature statement — which components the no-plastic rule covers, and the working temperature, duration, and cycle count the parts have to survive.
With those four inputs, the answer becomes a short list of parts to test rather than a specification debate. JSD Optics supports sensor-to-lens matching, custom focal lengths, chief ray angle matching, filter options, and mechanical changes to the barrel and holder, and measures distortion, modulation transfer function, and chief ray angle data for a specific lens-and-sensor pairing so that evaluation reflects the module that will actually be built.
Send the sensor part number, the number of channels, and the mechanical envelope, and JSD Optics will return matched lens options with the data needed to evaluate them.
❓ Frequently Asked Questions
Common questions from engineers specifying lenses for a 1/4" multi-channel array camera.
Is the JSD4012 suitable for a 1/4" sensor such as the Sony IMX219?
The JSD4012 is built for 1/4" image sensors, with a Ø4.8mm image circle against a 1/4" working diagonal of roughly 4.6mm. That covers the format with a small tolerance margin. Its specification sheet lists the OmniVision OV9712 as a matched sensor; for the IMX219 specifically, JSD Optics confirms the pairing against the sensor's active area and chief ray angle as part of an evaluation rather than claiming it from the catalogue.
How much of the 11mm track length budget does the 8.69mm total track length leave?
It leaves 2.31mm. In a four-channel array that margin is what absorbs holder machining tolerance, board thickness variation, and the thread engagement needed to set back focus on each channel. A part closer to the ceiling, such as a 10.7mm design, leaves 0.3mm and makes every shim in the stack a design negotiation.
Why does the JSD4012 contain a polymer element if the application runs hot?
Because track length and an all-glass stack pull against each other at this format. Moulded polymer aspheres are what let a wide, fast 1/4" design fold its aberration correction into 8.69mm of axial length. The JSD4012 uses a 1G3P construction — a glass element for focus stability plus polymer aspheres for the short track. Where the mechanical build is the part that must be plastic-free, that construction fits the brief. Where the optical elements themselves must be all-glass, the JSD4013 is the standard starting point at 11.91mm of total track length.
What field of view does the JSD4012 give horizontally, and why does that matter more than the diagonal?
At 1/4" the JSD4012 covers 145° diagonal, 123° horizontal, and 100° vertical. Four-channel stitching is decided by horizontal reach and the overlap it produces between adjacent channels, so the 123° horizontal figure is the one to put into the ring geometry calculation. A design sized from the diagonal figure alone can end up short of the overlap the stitcher needs.
Can distortion and other optical data be provided for a specific lens and sensor pairing?
Yes. Distortion, modulation transfer function, and chief ray angle data are measured for a specific lens and sensor pairing as part of an evaluation, and reported together with the conditions the measurements were taken under.
Does the JSD4012 seal against moisture or dust?
The JSD4012 is not an IP-rated sealed lens. A high-temperature environment does not by itself imply a sealing requirement, and the two are separate decisions. Outdoor, washdown, or automotive builds need a sealed enclosure, or a mechanical modification scoped with JSD Optics for that program.
Building a 1/4" array and need four matched lenses?
Send the sensor part number, the number of channels, and the housing envelope. JSD Optics engineers will return matched lens options with the data you need to evaluate them.
Or email us at info@jsdoptical.com
Consult now
