Did you know that a photograph of a leaf can look much brighter than an earlier photograph even when the lamp shining on it has not changed? A phone or camera may quietly alter how much light it records. Without noticing, a student can mistake a camera adjustment for a scientific result.
Automatic exposure is a camera process that adjusts recording parameters to produce an image brightness the camera considers suitable. When a school experiment compares photographs, changes in shutter time, aperture, sensitivity or processing can change pixel values independently of the subject’s real illumination. The immediate solution is to control and document the camera settings, use a consistent reference, and measure the science variable independently where possible; two brighter-looking pictures are not proof of an increase in light.
Start with the picture, not with a technical explanation
Show a learner two imaginary photographs of the same white card taken moments apart. The first looks grey; the second looks almost white. Ask: “Did someone turn up the room lights?” A pupil may answer yes because the visible difference is obvious. Now explain that the lamp and card were unchanged, but the camera changed its exposure settings. The photograph is evidence of what the camera recorded, not an automatic direct measurement of the room’s light intensity.
That distinction should be discovered, not merely memorised. Invite the learner to propose at least two explanations for the brighter image: more light reached the subject, or the camera changed how it collected or processed light. Then ask which observation could separate the explanations. A controlled camera setup, exposure metadata and a separate light sensor—if one is available and appropriate—can help.
Why the camera changes the numbers
A photograph is a chain of processes: the scene reflects or emits light; the optics and sensor receive it; the camera controls recording; and software translates measurements into the image displayed on a screen. Automatic exposure is useful for ordinary photography because it helps handle a bright window, a shadowed desk or a dark background. It is less helpful if the lesson is to compare changes in the brightness of an experimental object across time.
For example, a camera looking at a dark sheet of paper may raise exposure and make the paper appear unexpectedly light. Looking instead at a white page, it may lower exposure and make the page appear duller. A camera has not measured the page’s reflectance as an absolute laboratory instrument would. It has attempted to make a usable picture within its exposure and processing rules.
The Nikon camera manual’s explanation of exposure compensation confirms the important practical point: a camera’s suggested exposure can be deliberately adjusted so that pictures appear brighter or darker. That is manufacturer documentation for a photographic control, not proof that any individual school smartphone has identical controls or behaviour.
Worked example 1: the lamp stayed the same but the shutter changed
This is an invented, controlled teaching example, not a record of a real experiment. Imagine a stationary paper colour patch under a lamp that remains unchanged throughout two exposures. Assume the same lens, aperture, sensor sensitivity and processing settings. Photograph A uses a shutter time of 1/200 second. Photograph B uses 1/50 second. Assume the photos are not clipped, the camera is stable, and light reaching the sensor is otherwise constant.
- Record the variable: Photograph A’s integration time is 1/200 s = 0.005 s; B’s is 1/50 s = 0.020 s.
- Compare times: 0.020 ÷ 0.005 = 4. The shutter in B remains open four times as long.
- Interpret correctly: for the same unchanged light and otherwise identical capture conditions, B can collect approximately four times the sensor exposure from that patch because of the longer integration time.
- Do not infer: “the lamp became four times brighter”. The illuminating source was assumed unchanged, while one camera variable changed.
- Independent check: take another pair at identical settings, or consult a separate appropriate light measurement; the photograph alone cannot isolate the lamp.
A common error is to think that a longer shutter time always makes an ordinary classroom photo exactly four times as bright to the eye. Image-processing curves, sensor noise, clipping, compression, display brightness and the human visual system break that simple translation from sensor exposure into an on-screen impression. The factor of four here describes the assumed capture-time ratio, not a guaranteed factor in displayed pixel values.
Worked example 2: why a pixel average is not a light-meter reading
Suppose a fictional project compares a leaf photographed on Monday and Thursday. The same selected image region has an illustrative average encoded greyscale value of 96 in the first image and 160 in the second, on a common 0–255 scale. The numerical ratio is 160 ÷ 96 = 1.6667, and the difference is 64 encoded levels. The arithmetic is straightforward; interpreting it is the difficult part.
These encoded image values are not automatically linear measurements of scene illumination. The source images might have different shutter times, ISO settings, white balance, tone curves, exposure corrections or compression. Unless the workflow has been appropriately calibrated, it would be misleading to report that “the leaf received 66.7% more light”. A safe result is simply: “This selected region has a higher mean displayed greyscale code in the second fictional image, under capture settings that were not established as comparable.”
For a colour-changing chemistry or plant investigation, camera white balance and colour rendering are especially important. If a pupil uses phone photographs to describe the appearance of an indicator, it is better to note the indicator’s observed category under controlled illumination and compare with the proper reference standard. A phone JPEG alone is not a substitute for a calibrated colourimeter.
A learner diagnostic for visual evidence
Ask the pupil to annotate a two-photo worksheet without touching the camera. On the left, write what is visible: “The card appears lighter in image B.” On the right, write what is not yet established: “Whether the room light increased; whether the camera changed settings; whether the screen altered display brightness.” If the pupil can make those categories without needing a hint, they are beginning to separate observation from inference.
Follow with the question “What evidence would convince you?” Better responses may include a fixed lighting arrangement, identical capture settings, images taken from the same position, a reference patch, or a dedicated sensor. A pupil who simply replies “take more photos” may be missing that more uncontrolled photographs can repeat the same uncertainty.
Design a fair photo comparison
- State the measured question first: are you comparing leaf appearance, an object’s position, illumination on a surface, or the colour of a chemical indicator? Different questions need different instruments.
- Control geometry: keep the camera and subject at the same relative angle and distance; use a stable mount when safe and practical.
- Control light: keep lighting direction, nearby reflections and time-of-day conditions as consistent as the experiment allows, or record deliberate changes.
- Control capture: where the device permits, lock shutter time, aperture, ISO, white balance, focus and exposure compensation. On phones, confirm which settings actually remain locked.
- Include a reference: a stable neutral card can reveal some changes in capture or processing, but it does not itself turn the phone into a certified measuring device.
- Keep original files and settings: save the image timestamps and available exposure metadata; record editing, cropping or filters applied afterwards.
- Choose the right independent check: for illuminance use an appropriate light meter if the project genuinely needs illuminance; for length use a ruler or scale; for temperature use a temperature instrument.
Guided practice: let the same scene produce different answers
Activity 1 — compare without touching the room light
Under adult supervision where necessary, photograph the same neutral paper with a phone in automatic mode twice. First include mostly the paper in the frame; then include a large dark background around it, while keeping the paper and lamp fixed. See whether the displayed paper brightness changes. The result is not guaranteed for every phone because metering and processing algorithms vary. If it changes, ask what variable the camera may have adjusted. If it does not, that is also useful: write the observations rather than claiming the camera always behaves in one way.
Activity 2 — separate the image from the measurement
Give two invented photos labelled “same lamp, different exposure times” and one invented pair labelled “same exposure, lamp power intentionally changed”. Ask students what each pair alone can and cannot establish. In the first pair, camera behaviour is the known changed variable. In the second, with capture conditions controlled, a change in appearance is more relevant to the lamp hypothesis, but the interpretation still depends on calibration and the intended scientific quantity.
Activity 3 — design the independent test
Ask a learner to write a short procedure for comparing the colour of two test papers. A sensible procedure records the same lighting source and distance, places each paper beside the same neutral reference, uses locked settings where available, and compares observations using the appropriate indicator chart or instrument. Have the learner explain which parts control recording and which parts establish chemical or physical meaning.
Activity 4 — transfer beyond photography
Show a graph exported from a spreadsheet where one axis is automatically rescaled between screenshots. Ask: “Can the apparent visual difference be caused by a display or plotting choice, rather than a real change of the underlying quantities?” A learner who recognises this connection has understood visual evidence more deeply than a list of photography settings.
Mistakes that make school photo reports sound more certain than they are
- “Brighter photo means brighter object”: the image may differ because of camera settings or processing.
- “Same phone means identical measurement”: the same device can make different automatic exposure decisions in different compositions.
- “Pixel value 160 is twice the light of pixel value 80”: ordinary encoded colour values are not automatically linear, calibrated illuminance measures.
- “A reference card fixes everything”: it can provide a useful comparison but does not repair severe clipping, changing geometry or an uncalibrated image pipeline.
- “A filter only improves appearance”: editing can change evidence-relevant contrast and colour; keep originals and disclose processing.
- “A prettier photograph is a stronger conclusion”: scientific strength comes from matching the measurement to the question and explaining its limits.
Parent and teacher next steps that improve the work
A teacher can turn this into a short project in which pupils first diagnose an uncontrolled photo comparison, then write a fair method, and finally document whether the result could be independently checked. A student may use a simple checklist rather than advanced photo software. The educational goal is principled comparison, not buying expensive equipment.
Parents can ask a child what the photographs are supposed to prove. “Can the camera make the same object look different?” is a useful question because it does not hand over the answer. Another helpful prompt is “Which measurements would you trust if both photos were deleted?” A well-designed experiment often contains data that remain meaningful without an attractive illustration.
For progress, present three fresh picture pairs over several sessions. Record whether the learner identifies an uncontrolled setting, proposes a specific repair and states a cautious conclusion. Improvement is visible when the learner no longer translates apparent brightness directly into a numerical claim about the underlying scene.
Limitations and safety
Low-cost cameras differ widely. Some phone interfaces lock exposure only temporarily; some use computational photography that combines frames; automatic processing may continue even when a few settings are fixed. A camera also may clip highlights or bury details in shadows. If the project requires a quantitative optical measurement, consult a suitably calibrated instrument and an informed supervisor rather than forcing a JPEG to do a job it was not built for.
When pupils photograph a classroom, avoid capturing identifiable faces, screens, documents or other private material without appropriate consent. Nothing in a scientific photo comparison requires posting another pupil’s image publicly. These examples are fictional; no real student outcomes, school findings or laboratory performance are claimed.
Frequently asked questions
Will manual mode always make photographs scientifically comparable?
No. It can help control named exposure variables, but lighting, geometry, white balance, focus, processing, clipping and device calibration still matter. The method must match the physical quantity you are trying to measure.
Can a light meter settle every problem?
A suitable meter can help answer questions about illuminance at a location. It will not directly establish every property of a coloured object or a photograph. Use the correct instrument and units for the stated research question.
Should I compare phone camera brightness percentages?
Not as if they were physical percentages of incident light. You can compare consistently defined image statistics for a limited image-analysis task, but describe exactly how they were produced and what they cannot prove.
What if my phone does not let me lock all settings?
Record what cannot be controlled and narrow the conclusion. If the purpose is a strong quantitative comparison, consider another instrument or a redesigned question. Transparent limitations are preferable to invented precision.
Do more photographs fix an uncontrolled exposure?
Repetition may show how variable the pictures are, but it does not by itself separate real scene changes from systematic camera adjustments. First address the source of the confounding.
Related learning across the eduKateSG system
For a related distinction between a display change and a physical change, see how to compare satellite images without mistaking seasonal colour for land-use change. Read how blanks and negative controls answer different scientific questions for another lesson in control design. The Education Hub supports wider enquiry, while the immutable Clementi Mathematics reference models diagnosis, explanation and independent checks.
The core aim: learn what the instrument really measured
A good scientist does not ignore a beautiful photograph. The scientist asks what created the appearance and which parts of the inference can be checked. That habit—seeing the difference between the evidence, the tool and the conclusion—travels from school science into research, media literacy, computing and everyday decision-making.