DEV Community: Pomeas Vision

Motorized Zoom vs Fixed Focal Length: The Cost Math Nobody Puts in the Quote

Pomeas Vision — Thu, 15 Oct 2026 05:40:03 +0000

Motorized Zoom vs Fixed Focal Length: The Cost Math Nobody Puts in the Quote

A fixed-focal-length lens usually wins the quote. It costs less per unit, it is simpler to specify, and nobody argues about it in a design review. The costs that decide whether the station is actually cheaper over three years are not on that quote at all — they show up later, as changeover time, recalibration labour, and spare optics sitting in a cabinet.

This is the arithmetic we run through with integrators when they are choosing between one motorized zoom lens and a set of fixed lenses for a shared inspection station.

The purchase price is the smallest number in the comparison

Consider a line that inspects four product variants on one station.

Fixed-focal-length approach: four lenses, four holder setups, four calibration recipes.
Motorized zoom approach: one lens that travels across the same magnification envelope.

What the fixed approach buys you is real:

lower cost per lens
no motors to fail, no driver to write
fewer things to integrate on day one

What it costs after day one is where the comparison turns:

every product change is a physical change — swap, focus, recalibrate, verify, release the line
four lenses to stock, clean, inspect, and eventually replace
every manual swap is an opportunity to introduce a setup error, and setup errors surface as false rejects a week later, when nobody remembers the swap
the station's flexibility ceiling is frozen at design time — a fifth product variant means a fifth lens

None of those lines appear on a purchase order. All of them appear in the operating budget.

The three numbers that decide it

You do not need a spreadsheet with forty rows. You need three inputs:

R — changeovers per year. How many times does the station switch products? Count the real number, not the average. A "mostly one product" line often hides a burst of switching at month-end.
t — minutes of stopped line per changeover. Not just the swap. Include refocus, recalibration, first-article verification, and the time to find out why the first article failed.
M — cost of a stopped line per minute. If you do not have this number, estimate the contribution margin the line gives up while stopped. It is almost always larger than people assume, because it includes labour standing idle.

Then:

Annual changeover cost = R × t × M

Compare that against the one-time premium of the zoom solution — the higher lens cost plus controller plus integration labour:

Payback period = (zoom premium) / (R × t × M)

If the payback period is shorter than your asset depreciation window, the fixed-lens option is not actually cheaper. It is just cheaper on the invoice.

A worked example (illustrative — substitute your own numbers)

Assume a station with four variants, two changeovers per shift, two shifts, 250 production days:

R = 2 × 2 × 250 = 1,000 changeovers per year
t = 20 minutes (12 min swap and refocus, 5 min recalibration, 3 min first-article check)
M = $0.50 per minute of stopped line

Annual changeover cost = 1,000 × 20 × 0.50 = $10,000 per year

Now suppose the motorized zoom solution carries a one-time premium of $2,500 over the fixed-lens set (higher lens cost, controller, integration). Payback:

$2,500 / $10,000 = 0.25 year ≈ 3 months

Everything after month three is savings, and the number of lenses in the cabinet drops from four to one.

The point of the example is not the specific figures — yours will differ, possibly by a lot. The point is that changeover frequency, not lens price, dominates the comparison. Two stations with identical bills of materials can have completely different answers.

When fixed focal length is still the right answer

Being honest about the boundary matters more than defending the zoom option.

Fixed focal length wins when:

R is low. A dedicated station that runs one product for years will never recover the zoom premium. Buy the cheap lens and spend the difference on lighting.
The magnification is fixed by physics. If your application always works at 0.5× and always will, a zoom range is unused capability.
The station is fully manual anyway. If an operator is already standing there with a part in hand, saving 20 seconds of lens swapping changes nothing.
You need maximum optical performance at one magnification. A lens optimised for a single magnification can beat a zoom lens at that same magnification, for less money.

The mistake is not choosing fixed lenses. The mistake is choosing them for a line that switches products daily, and then discovering the real cost eighteen months later when throughput reports do not add up.

Signals you are already past the break-even point

Production asks for "one more variant" on the same station, and the answer today is "we need another lens."
Your optics cabinet has more lenses than the station has camera ports.
Changeover instructions are printed and taped to the machine.
First-article failures cluster in the hour after a product change.
Recipes exist in the vision software, but the optics still require a human.

Any two of those, and the changeover arithmetic above is probably already in your favour.

The takeaway

Compare stations, not lenses. A motorized zoom lens is not better optics — it is a different position on a cost curve, and it is the right position when a station serves more than one product and the line stops every time the product changes.

Do the three-number calculation before the design review, not after the first throughput shortfall. R × t × M is the number that decides it, and it is the number that never makes it onto the quote.

We publish field notes on machine vision optics — lens selection, field-of-view and depth-of-field math, and inspection-line integration. Specifications and application notes for our industrial zoom lens series are on the POMEAS site.

How to Choose a Motorized Zoom Lens for Machine Vision: Field Notes from the Factory Floor

Pomeas Vision — Tue, 13 Oct 2026 03:35:15 +0000

How to Choose a Motorized Zoom Lens for Machine Vision: Field Notes from the Factory Floor

Choosing an optic for an automated inspection station used to be a one-way decision. You picked a fixed-focal-length lens, locked the working distance, and hoped the part never changed size. When the product mix changed, you changed the lens — or worse, the whole lighting-and-optics assembly.

Motorized zoom lenses changed that equation. Instead of re-engineering the station for every new SKU, the optic itself adapts: focal length, and sometimes aperture and focus, adjust under software control while the camera stays put. After years of supporting integrators who build inspection cells around these lenses, here is the selection logic we walk through with customers, in the order that actually matters.

1. Start with the sensor, not the lens

The first question is never "which zoom lens?" It is "what sensor is on the camera?" A zoom lens projects an image circle; if that circle is smaller than your sensor, the corners go black or soft. If it is much larger, you are paying for coverage you never use.

Practical rule: match the lens's specified sensor format to your camera's format, and stay at or slightly above it. A 2/3" lens on a 1/1.8" sensor is fine. A 1/2" lens on a 1" sensor is not — the edges of your field of view will be unusable.

For high-resolution sensors (20 MP and beyond), also check that the lens is rated for that pixel count, not just the format. Two lenses can both "support 2/3"" while only one resolves 3.45 μm pixels across the whole frame.

2. Fix the working distance envelope, then the magnification range

Working distance (WD) — the gap between the front of the lens and the object — is usually dictated by mechanics: rotary fixtures, robot reach, lighting geometry, safety covers. Write down the minimum and maximum WD your station can physically accept before you look at any datasheet.

Then compute the magnification range you need:

Field of view (FOV) = sensor dimension ÷ magnification
So magnification range = (smallest sensor dimension ÷ largest FOV) to (largest sensor dimension ÷ smallest FOV)

Example (illustrative): a 1/1.8" sensor is about 7.2 mm wide. If you need to see fields of view from 10 mm to 60 mm wide, your magnification range is roughly 0.12× to 0.72×. A motorized zoom lens that covers 0.1×–0.8× handles it with margin; a fixed 0.5× lens only handles the middle of that range.

This single calculation eliminates most candidates in a catalog. Everything after this is refinement.

3. Zoom ratio is flexibility, but depth of field is the tax

Zoom ratio tells you how much the magnification can vary: a 6.5× zoom covering 0.1×–0.65× gives you a wide envelope. But every optical system pays for zoom range with depth of field (DOF) behavior, especially at higher magnifications.

Two things to check on the datasheet:

DOF at your actual magnification, not the headline number at the low end. DOF shrinks roughly with the square of magnification — halving your magnification buys you about four times the depth, which is why the low-magnification end always looks flattering and the high end always looks thin.
Whether the lens refocuses motorized or manually. If focus is motorized and repeatable, the controller can re-shift focus when you zoom — this is what makes a motorized zoom lens genuinely programmable rather than just adjustable.

If your station inspects parts with height variation (castings, molded parts, PCBAs with connectors), budget DOF before zoom ratio. A smaller zoom range with generous DOF often outperforms a big zoom range that lives at the edge of focus.

4. Resolution and contrast across the whole zoom travel

Datasheet MTF curves are usually shown at one or two magnification settings. Ask for (or measure) performance at the magnifications you will actually run. Real-world failure mode: the lens looks excellent at 0.2× and noticeably soft at 0.6×, and nobody notices until the vision algorithm's edge-detection thresholds start drifting between recipes.

If you run line-scan or large-area sensors, also confirm the lens is designed for that image circle at full aperture, not just stopped down.

5. Motorization: what is actually under software control

"Motorized" covers three different levels of capability:

Motorized iris only — brightness adapts automatically. Useful when lighting varies, but magnification is still manual.
Motorized zoom + manual focus — you can change FOV remotely, but an operator re-focuses after each change.
Motorized zoom + motorized, repeatable focus — the lens stores position pairs (zoom, focus) per recipe and returns to them.

Level 3 is what makes multi-SKU automation practical. When evaluating, ask two questions: what is the position repeatability of each motor, and how is the lens controlled — RS-232, USB, or Ethernet — and does your vision platform already have a driver for it? A lens that integrates in an afternoon is worth more than a slightly better optic that needs custom firmware work.

6. The hidden parameters that decide uptime

Vibration and shock rating: motorized optics contain real mechanics. A lens rated for a lab bench will drift on a press-adjacent line.
Zoom/focus travel speed: if your recipes switch 200 times a shift, seconds per switch are throughput.
Back-focus adjustment: lets you compensate for mechanical tolerances in the camera mount without shims.
C-mount thread quality and flange repeatability: cheap mounts introduce tilt, and tilt shows up as one-sided blur you will chase for weeks.

A compact selection checklist

Sensor format and pixel size → candidate lens families
WD envelope (mechanical) → fixed WD range
FOV range needed → magnification range (divide sensor by FOV)
Part height variation → required DOF at the top magnification
Recipe-switch frequency → motorization level and control interface
Environment → vibration rating and mount quality

Wrapping up

A motorized zoom lens is not a luxury — for factories running mixed products on shared stations, it is usually cheaper than maintaining several fixed-magnification optics or re-fixturing per SKU. The discipline is to let the sensor, WD envelope, and DOF requirements drive the choice, and to treat the motor control interface as an integration decision, not a footnote.