You have the camera running on the frame grabber, the video is live, and there is no exposure field anywhere in the acquisition software. On most legacy progressive-scan CCD cameras that is by design: the shutter is a rotary switch on the rear panel, stepped from 1/60 s down to 1/16000 s, and no camera file, serial command or software register can move it. This guide shows where exposure authority can still live on such a camera, how to compute the right dial step for your part speed, and how to recover the brightness you lose through the light controller, the lens iris or the trigger pulse. The arithmetic takes five minutes and ends with a setting you can hold constant for years.
What a dial-only camera actually is
A current camera follows the GenICam standard, so exposure is just another feature register your program writes. A legacy progressive-scan camera predates that architecture. Its integration time is wired to a rear-panel rotary switch with fixed steps — typically 1/60, 1/125, 1/250, 1/500, 1/1000, 1/2000, 1/4000, 1/8000 and 1/16000 s. The exact ladder for a given model is printed on its data sheet; the powers-of-two pattern is the common one. Your acquisition program sees finished video and has no handle on integration time.
Two consequences follow. Each step halves or doubles the collected light, so one step is one stop, about 6 dB of signal. And brightness control leaves software entirely: the dial owns time, the light and iris own intensity, and your program owns gain and processing. Anyone integrating area scan cameras of that generation inherits exactly this division of authority, and the plan below works with it instead of against it.
Three software paths to rule out first
Before accepting the dial, close out the alternatives. The camera manual answers all three within minutes.
- The serial command set. Many CCD cameras take commands over RS-232, or over the Camera Link serial channel the frame grabber presents as a port on the host. An exposure or shutter entry in that command list means software control exists after all.
- Trigger-width exposure. Some cameras time integration from an external pulse: exposure runs from the rising edge to the falling edge. The trigger input section of the manual documents it as shutter or exposure control by pulse width.
- The camera configuration file. On Camera Link cameras this is the .icd file the frame grabber loads at startup. It is a script of serial commands, so it can only set what the serial command set already exposes — it cannot add a shutter control the hardware lacks.
If all three come up empty, the dial is the exposure control, and the rest of this guide moves brightness to the hardware you still own.
The dial step comes from the motion budget
Choose the longest step that freezes motion; missing light is recoverable, blur is not.
s is the image scale: field of view across the motion (FOV, mm) divided by the pixels spanning it (N). k is the blur budget in pixels — 0.5 for gauging and measurement, 1 for presence, sorting and code reading. v is part speed in mm/s and t_max comes out in seconds. Blur at any exposure t is t × v in mm, so each slower dial step doubles the smear. The same budget runs software-exposure cameras: exposure time versus shutter speed for moving parts applies it identically, and capping the maximum auto exposure time is how the limit survives once exposure moves into software.
Worked example: a 640-pixel dial camera over 64 mm at 100 mm/s
Inputs, and where each comes from: a VGA-class progressive-scan CCD with 640 horizontal pixels and a shutter dial, the camera type this guide is about; a 64 mm field across the motion, measured by laying a steel rule on the belt and reading its span in the frame; 100 mm/s part speed, timed over 500 mm in 5.0 s; and a 0.5 px blur budget, because the measured feature carries a ±0.05 mm tolerance.
The rule in words: scale first, then the longest exposure that keeps blur inside half a pixel.
- s = 64 mm / 640 px = 0.1 mm/px
- t_max = (0.5 px × 0.1 mm/px) ÷ 100 mm/s = 0.05 mm ÷ 100 mm/s = 0.0005 s = 1/2000 s
The dial ladder includes 1/2000 s, so set it there. The next slower step, 1/1000 s, smears the part by 0.001 s × 100 mm/s = 0.1 mm — a full pixel, twice the budget.
Dropping from the slowest step (1/60 s) to 1/2000 s costs 2000/60 = 33×, about 5 stops of light:
| Dial step | Blur at 100 mm/s | Light lost versus 1/60 s |
|---|---|---|
| 1/500 s | 0.20 mm (2 px) | 8.3× ≈ 3 stops |
| 1/1000 s | 0.10 mm (1 px) | 16.7× ≈ 4 stops |
| 1/2000 s | 0.05 mm (0.5 px) | 33.3× ≈ 5 stops |
| 1/4000 s | 0.025 mm (0.25 px) | 66.7× ≈ 6 stops |
Recover it in two moves. The iris, f/4 to f/2, gains 2 stops — a factor of 4. The remaining 33.3 ÷ 4 ≈ 8.3× comes from the light: raise the controller from 10% to 80% of rated output, or strobe at a higher peak. Decision: dial at 1/2000 s, iris at f/2, light at 80%. The factors multiply to 4 × 8 = 32 against a 33× loss, so the histogram will sit a shade under target; add the last fraction from strobe peak current if the grey patch reads below 170.
Recovering the lost light
Work down the authorities in this order:
- Iris. Opening one full stop doubles the light and shrinks depth of field, so spend stops here while the geometry still holds the part in focus.
- Continuous intensity. Constant-voltage and constant-current channels on light controllers dim from a setting or from software — exactly the brightness authority the camera refuses to give you. The light controller finder filters by channels, control type and strobe support.
- Strobe. With a dial camera, the strobe does not set exposure — the dial does. The pulse must cover the exposure window plus the camera's trigger-to-exposure delay; a pulse that lands late leaves frames dark or partly lit, the timing check in why the first camera frame comes out dark. Brightness then scales with peak current, and a light pulsed at 1.5% duty (0.5 ms at 30 fps) tolerates a far higher peak than its continuous rating — the pulse current and duty limit are printed on the light's data sheet.
- Output
- Constant voltage 24 V
- Channels
- 4 channels
- Control
- Digital
- Feature
- Strobe
The LS-CTL-0070 is a 4-channel, 200 W digital control box with a strobe stage and an RS-232 port: one box to pulse the light, hold brightness and take commands from the cell controller.
Trigger pulse width as exposure
Where the manual documents a trigger-width exposure mode, the pulse is the exposure: the rising edge opens the shutter, the falling edge closes it. Set the pulse to the t_max from the motion budget instead of hunting dial steps. Published designs of this type put a floor under the pulse — 100 µs is the common minimum, and a shorter pulse is ignored, leaving the camera on its default timing. The ceiling is the frame period minus blanking, so the usable range runs from about 100 µs to a few milliseconds per frame.
Drive the pulse from the light controller's strobe output or from one of the frame grabbers that interface legacy cameras. A Camera Link cable carries dedicated camera-control lines alongside the video, and one of them is designated for exposure, so the pulse train can come straight from the grabber — generating a frame grabber trigger pulse train without a function generator shows one way. The trigger wiring also sets the frame cadence; camera frame rate under an external hardware trigger covers that side.
- Interface
- Camera Link
The MV-GC1102IOL is a Camera Link frame grabber — the piece that carries the video, the serial channel and the camera-control lines that pulse the camera's exposure.
Verify the result in the image
A dial camera reports nothing about its timing, so the histogram is the exposure meter. Read a diffuse grey patch or the part background and aim the mean at 160–200 of 255: high enough to keep noise low, low enough that a shinier part does not clip at 255, where pixels lose all detail. If the background pushes 255, drop one dial step and put the stop back into the light — clipped pixels carry no information, whatever the exposure was.
Two stability checks belong here. Turn off any automatic gain; on a legacy analog chain that setting often lives in the frame grabber, and left on it produces brightness drift that looks like a lighting fault — the brightness drift checklist applies to the whole chain. For a cell that runs for years, pin brightness with feedback rather than a fixed dial position: closed-loop lighting control holds the same grey target against LED aging. On a camera that does report timing, per-frame exposure and gain metadata confirms each frame's settings the same way.
When the dial cannot get you there
Two cases defeat the dial outright.
Faster motion. Run the worked example at 1000 mm/s instead: t_max = 0.05 mm ÷ 1000 mm/s = 0.00005 s = 1/20000 s. That is beyond 1/16000 s, the fastest step on the typical ladder, and no iris or light removes blur once it is recorded. The exits are trigger-width exposure with a short pulse, or a camera with programmable exposure.
Changing parts. A family with mixed reflectivity needs a different exposure per part, and a rotary switch cannot follow the part. Either hold the exposure fixed and move the correction into per-part strobe energy, or move exposure into software.
Both exits lead to the same replacement: current GigE Vision and USB3 Vision cameras carry exposure as a settable, per-frame feature. The MV-CH120-11UC is a 12 MP USB3 camera with a global shutter, a 1.1-inch sensor and a C-mount — exposure, gain and ROI are registers, not switch positions.
- Sensor format
- 1.1"
- Lens mount
- C-mount
- Interface
- USB3
- Resolution
- 12 MP
Size the replacement field with the camera resolution calculator, then shortlist in the camera finder or start the configurator at the camera step.
Before you commit: a short checklist
- Shutter authority named: dial, serial command or trigger-width input, from the camera manual, written into the build notes.
- t_max computed from field of view, pixels and speed; dial at the first step at or below it.
- Stops recovered with arithmetic that adds up: iris factor times light factor covering the loss.
- Pulse wider than exposure plus trigger delay on a dial camera; pulse equal to exposure on a trigger-width camera.
- Histogram mean at 160–200 of 255 on a diffuse target, no clipped background.
- Automatic gain off on the camera and the frame grabber.
If the budget ends in a hardware fix rather than a camera swap, pick the controller first: the configurator at the lighting step matches lights and strobe stages to the power you computed, and the accessories finder covers the frame grabber side. If the fix lands in software after all, the cameras topic page collects the exposure-side guides.
Questions
Does a longer strobe pulse brighten the image of a dial-only camera?
No. The dial fixes the integration window, so grey level scales with the light's peak intensity during that window, not with pulse length. The pulse only has to cover the exposure plus the camera's trigger-to-exposure delay. Extra width adds heat to the LEDs and can smear ambient reflections into the frame without adding signal.
How much does each shutter step change image brightness?
Each step halves or doubles exposure: one stop per step. Compensate by doubling LED intensity per step, or by opening the iris one stop — a full stop multiplies the f-number by about 1.4, so f/2.8 to f/2 doubles the light. Six steps is a 64× change, which is why the light carries most of the correction.
Can the Camera Link camera file set the shutter?
The camera file — the .icd a Camera Link frame grabber loads at startup — is a script of serial commands, so it can only set features the camera's serial command set exposes. Read the full command list: legacy cameras often expose gain, offset or partial scan over serial while omitting exposure. If no shutter command appears, the dial stays the only timing control.
What is the shortest pulse a trigger-width exposure accepts?
Cameras that time exposure from the pulse commonly set the floor at 100 µs; a shorter pulse is ignored and the camera keeps its default timing. For scale: at 100 µs, a part moving 100 mm/s blurs 0.01 mm — a tenth of a pixel on the worked example's 0.1 mm/px scale — so pulse-width exposure reaches speeds the dial ladder cannot.
What mean grey level should a properly exposed machine-vision frame show?
Aim a matte, diffuse target — an 18% reflectance grey card is the standard choice — at 60–80% of full scale, about 160–200 of 255. Specular spots read the surface angle, not the exposure, so judge on diffuse areas. The headroom matters when part reflectivity varies: it keeps a shinier part from clipping at 255, where pixels lose all detail.
When is replacing the camera cheaper than working around the dial?
Two cases: the motion budget needs an exposure shorter than the fastest dial step — about 63 µs at 1/16000 s — and no light can remove blur; or the part mix needs per-part exposure changes. A current GigE or USB3 camera accepts exposure in software per frame, which usually costs less than the engineering hours spent on hardware workarounds.


