Make a Video — One Skill, One Short Film at a Time: Week 3 High, Low or Eye Level? — How Camera Height Changes the Story

 

High, Low or Eye Level? — How Camera Height Changes the Story

Make a Video — One Skill, One Short Film at a Time: Week 3

Film the same person doing the same thing from three camera heights. How different can one ordinary moment feel?

A person sits at a table and opens a notebook.

At eye level, we might feel that we are sitting opposite them. From above, we see more of the page and the arrangement of the table. From below, the person may appear more imposing, while the notebook becomes harder to see.

Nothing about the action has changed. We have changed the position from which the audience watches it.

That is this week’s filmmaking skill: choosing a camera height deliberately.

In Week 1, we used wide, medium and close-up shots to show where we were, what was happening and which detail mattered. In Week 2, we considered framing and background: what should the viewer notice?

Now we ask: where should the viewer be looking from?

You do not need a new camera, an elaborate set or complicated editing. You need one simple action and three different viewpoints.

What You Need

• A phone or camera that records video.
• One willing adult subject, or an object you can film yourself.
• A notebook and table, or another simple activity.
• A stable support, tripod or steady pair of hands.
• About 20 minutes to record and compare three short clips.

Choose a quiet, uncluttered setting with reasonably consistent light. A seated subject makes the exercise easier because all three heights are within comfortable reach.

This week’s success is three recognisable, usable clips filmed from clearly different heights.

Camera Height and Camera Angle: Related, but Different

Camera height means how high the camera is relative to the subject. Camera angle means whether it points upwards, downwards or straight ahead.

A camera can be low but point horizontally. A high camera can also point horizontally. Height alone does not tell us which way it is looking.

For this exercise, we will combine height and angle in three familiar ways:

Viewpoint	Camera position	Where it points
Eye level	Approximately level with the seated person’s eyes	Roughly straight towards them
High	Above their eye level	Down towards their face, hands and notebook
Low	Below their eye level	Up towards their upper body and face

“Eye level” is the subject’s eye level. If someone is seated, the camera may need to be considerably lower than the standing camera operator’s eyes.

That small adjustment can make a large difference.

Eye Level: Meet the Person Where They Are

An eye-level shot often feels direct and familiar. We can see the person’s expression without obviously looking up or down at them.

For our notebook example, position the camera opposite the seated person, with the lens approximately level with their eyes. Frame their upper body, hands and notebook as comfortably as possible.

This is a useful starting point for an interview, a spoken introduction or someone explaining an idea. It can help create the feeling of a conversation.

If you are filming yourself, look at the lens when addressing the audience, rather than continually watching your own image on the screen.

Try This

Record the person sitting quietly, opening the notebook and looking at the page. Let the clip run for about eight to ten seconds.

Watch it back. Does the viewer feel like a participant in the moment? Can they see both the person’s expression and enough of the action?

Eye level is a starting point, not a guarantee of a successful shot. A distracting background or awkward framing can still get in the way.

High: Reveal More of the Activity

Now raise the camera above the subject’s eyes and point it down towards the same action.

You may see more of the tabletop, the open pages and the position of the person’s hands. The shot can become more about the activity and less about meeting the person face to face.

This is particularly useful when the viewer needs to understand a layout: ingredients on a work surface, components on a desk, or notes being organised.

In my teaching and filming work, visibility matters. If the important information is on the table, the camera needs a position from which the audience can actually see it. A high view can help explain a task that an eye-level view partly conceals.

Try This

Repeat the notebook action. Keep the camera above eye level and tilt it down enough to show both the person and the page.

Ask yourself:

• Is the notebook easier to see?
• Is the person’s expression still important in this version?
• Does the shot feel more observational?

A high-angle shot can sometimes make someone appear smaller or more vulnerable. It can also simply give the audience a clearer view of what they are doing. Its meaning depends on the scene.

A gently elevated shot is not the same as a directly overhead shot. Looking straight down at the notebook is a useful extension, but it is a different viewpoint from our main comparison.

Low: Give the Subject More Presence

For the third clip, lower the camera below the person’s eyes and point it upwards.

You might place it near table height, provided the table does not block the action. Keep the angle moderate at first.

The subject may seem taller, more imposing or more prominent. You may also reveal a different background: a ceiling, a light fitting or a previously hidden shelf.

Outside, a low viewpoint can place a subject against the sky. Around a sailing boat, it can emphasise the height of a mast or the scale of the rig. In a studio, it might give a performer a stronger visual presence.

Try This

Record the same person opening the same notebook again.

Compare what you can now see with what you have lost. The person may dominate the frame, but the writing surface may almost disappear.

That trade-off matters. A visually striking angle is useful when it supports the purpose of the film.

Low angles are sometimes associated with power or threat, but they can also suggest admiration, scale, humour or a child’s viewpoint. A low shot of a friendly musician need not feel threatening at all.

Get Filming: One Action, Three Heights

1. Choose a Repeatable Action

Use the notebook example from Week 1:

1. Rest both hands beside the closed notebook.
2. Open it.
3. Look down at the page.
4. Pause.

Keep it simple enough to repeat without much rehearsal.

2. Keep the Main Ingredients Consistent

Use the same person, table, notebook and lighting. Keep each clip a similar length.

Leave the zoom setting unchanged where possible. Maintain roughly the same distance from the subject, accepting that the framing will shift as the camera moves.

This is a practical comparison, so the three images do not have to match perfectly. You are trying to make the effect of viewpoint easy to notice.

3. Record Eye Level, High and Low

Film one clip from each position. Keep the camera still during each take.

Hold for a couple of seconds before the action begins and after it finishes. Those quiet moments give you useful room for editing.

Check each recording before moving on. Make sure the face or important detail is in focus and the action is visible.

4. Compare Before You Edit

Play the clips one after another, initially with the sound muted.

Ask:

• What do I notice first?
• How close do I feel to the person?
• Which view explains the action most clearly?
• Which gives the person the strongest presence?
• What has appeared in the background?

There is no single correct favourite. Different answers reveal different uses.

5. Make a Short Comparison Film

If you already have a familiar editing app, place the three complete clips in sequence. Add simple labels: “Eye level”, “High angle” and “Low angle”.

Aim for about 25–35 seconds. Music, transitions and effects are optional; the comparison itself is the point.

If editing feels like an extra hurdle, simply play the three recordings in order. You have still completed the exercise.

The Background Moves Even When the Subject Does Not

This week builds directly on our framing exercise.

Raise the camera and you may replace a busy wall with a relatively plain tabletop or floor. Lower it and you may introduce a bright window or ceiling light.

A shelf behind the subject’s head can disappear with a small change in height. A previously hidden distraction can appear just as easily.

Before recording, look around the whole frame. Check the corners as well as the person.

Moving the camera a few inches may improve a shot more than adding another piece of equipment.

Three Ways to Use This in Real Filming

A Science Demonstration: Show the Person, Then the Process

Use an eye-level view to introduce the experiment and a higher view to show hands, apparatus and the working area.

The camera position should answer the viewer’s immediate question: who is explaining, or what is happening?

For this week’s practice, use harmless everyday materials. The notebook teaches the camera skill without adding the demands of an experiment.

An Organ Performance: Decide What the Audience Should Experience

An eye-level or gently raised view can show the performer and instrument. A higher angle may reveal hand movement across the keyboards. A moderate low angle can emphasise the performer and the scale of the organ.

Each view contributes something different. The choice depends on whether the moment is about the musician, the technique or the instrument itself.

A Sailing Film: Show Scale or Explain the Deck

A low viewpoint can emphasise rigging against the sky. A higher view can make deck arrangements easier to understand.

Try the comparison with a securely moored boat before adding the movement and distractions of filming afloat. Choose a stable position and keep clear of working ropes and crew.

Pro Tips: Small Adjustments, Better Results

Start with modest angles. A small change often gives a clearer comparison than an extreme view.

Move the camera, not just your eyes. Bending down while leaving the phone in the same position does not change its viewpoint.

Keep your subject comfortably framed. Watch for accidentally cropped heads or hands that leave the frame during the action.

Avoid placing a wide-angle camera very close to a face. Close camera positions can exaggerate the apparent size of nearer features, particularly with a strong upward view.

Check focus and brightness. A new angle may introduce a bright window or dark background. Use the controls you already understand to keep the important subject clear.

Support the camera safely. Use a suitable tripod or secure support. Do not improvise an unstable camera position above someone or climb onto furniture for this exercise.

Choose the angle for a reason. “It looks different” is a good beginning. “It lets the audience see the page” is a more useful filming decision.

A Personal Reflection: Make One Decision at a Time

With the cameras, visualisers and studio facilities available to me, there are many ways to cover an activity. The useful teaching question is still a simple one: what does this position help the viewer understand?

That is why I want this series to build one skill at a time.

A beginner does not need to manage camera height, sound, lighting, performance and editing as one large task. Three short clips provide a manageable experiment with an immediately visible result.

I would rather make the first goal concrete: record the same action from three heights, then explain one difference.

That explanation might be as simple as, “I chose the higher view because you can see the notebook.” It shows that the filmmaker is beginning to make a deliberate choice.

Optional Challenge: Choose the Best View for a Purpose

Once you have your three clips, choose one for each of these purposes:

• Introducing the person.
• Showing what they are doing.
• Giving them a stronger visual presence.

You can reuse a clip. There is no requirement for each purpose to have a different answer.

Then record one additional shot to improve your chosen view. Perhaps the high camera needs to come down slightly so that you can see the person’s expression. Perhaps the low shot needs a small sideways adjustment to remove a ceiling light.

This is how useful camera choices develop: record, watch, adjust.

Conclusion: Give the Viewer a Place to Stand

A camera does more than record an event. It gives the audience a position from which to experience it.

Eye level can bring us into a conversation. A higher view can reveal an activity. A lower view can give a subject greater presence.

Those are possibilities to explore, rather than rules to memorise.

This week, take one person, one notebook and one ordinary action. Record it from three heights. Watch what changes.

Then choose a viewpoint and finish this sentence:

“I put the camera here because…”

When you can answer that, you are beginning to direct the viewer’s experience.

You can find a lot more details on the website that accompanies the series at https://makingbetter.video/

Make a Podcast — One Skill, One Short Recording at a Time · Week 3 Find the Sweet Spot — Where Should the Microphone Go?

 

Find the Sweet Spot — Where Should the Microphone Go?

Make a Podcast — One Skill, One Short Recording at a Time · Week 3

A few centimetres can change how your voice sounds. This week, let your ears choose where the microphone belongs.

You have recorded your first introduction and found a suitable recording spot. Now keep that location and explore something smaller: the space between your mouth and the microphone.

A distant microphone can make you sound as though you are speaking from across the room. Move very close and you may hear more breath, heavier bass or a sudden thump on a word beginning with “p”.

As a teacher and filmmaker, I see value in giving people one manageable improvement to work on. A clear recording made with confidence is a worthwhile achievement.

This week you will…

Compare close, distant and slightly off-centre microphone positions, then choose one that gives you clear, comfortable speech.

Allow about 15 minutes.

What you need

• Essential: a phone with a recording app, or a microphone connected to a recorder or computer.
• Essential: the quiet recording spot you chose last week.
• Essential: two or three sentences to repeat.
• Optional: headphones for hearing small differences.
• Optional: a microphone stand and pop filter.

Use the same equipment for all three recordings.

A few things to know

1. Distance changes more than volume

A closer microphone usually captures more voice relative to the room. Further away, reflections and background sounds become more noticeable relative to your speech.

Some directional microphones also make your voice sound bassier at close range. This is called the proximity effect; it does not happen equally with every microphone.

2. Your breath can hit the microphone

“P” and “b” sounds can send a puff of air towards it, producing a low thump called a plosive.

Moving the microphone slightly beside your mouth can help: speak past it while keeping its recording face directed towards you. You are testing a small change, not turning away completely.

3. Find the part that actually records

Some microphones record through the end; others through the side. Check the instructions rather than assuming.

With a phone, identify the microphone opening used by your recording app and keep it uncovered. Treat the distances below as starting points, not universal rules.

Get recording: three positions, one voice

Use this sentence:

“Welcome to my podcast. Today I’m talking about something I enjoy. Perhaps the best place to begin is with a simple question.”

The “p” and “b” sounds make it useful for spotting breath thumps.

1. Settle into position. Sit comfortably and support the phone or microphone securely. Speak naturally. If you have an input meter, check that your louder words do not reach the red or clipping indicator.

2. Record A: close and central. Place the recording opening roughly 10–15 cm from your mouth, directly in front. Record the sentence, followed by two seconds of silence. Save it as “A — close”.

3. Record B: further away. Move it to roughly 40–50 cm away, still facing you. Repeat the sentence at the same speaking volume. Save it as “B — distant”.

4. Record C: slightly off-centre. Return to roughly 10–15 cm away, but move it a little to one side, perhaps 20–30 degrees from straight ahead. Keep its recording face aimed towards your mouth while you speak forwards. Save it as “C — off-centre”.

5. Keep the comparison fair. Leave manual recording settings unchanged if available. A phone may adjust levels automatically; that is fine, but note it. Avoid adding filters or noise reduction for this test.

6. Choose and repeat. Listen to all three, then record a fresh 20-second introduction using your preferred position.

Listen back

Listen for three things:

• Clarity: can you understand every word easily?
• Breath and tone: are there thumps, excessive breath sounds or an overly boomy voice?
• The room: do echoes and background noises distract from your words?

The loudest recording may initially seem best. Adjust playback volume so each voice sounds approximately equally loud before judging quality.

Pro tips

Make your position repeatable. Note the distance and take a quick photograph of the setup. Keep your head reasonably steady during recording.

Change one thing at a time. If the close recording sounds too bassy, move back a few centimetres. If breath thumps remain, try a pop filter. Listen after each change.

See it and hear it

Show three photographs labelled close, distant and slightly off-centre, with the mouth-to-microphone distance visible.

Pair them with the three short audio clips. In a demonstration video, show each position, then play its recording without music.

Your small win

You have finished when you have saved three comparisons, chosen a position and made one usable introduction.

You do not need a perfect “radio voice”. You need words your listener can follow.

Take it further

Record your chosen position twice: once looking forwards and once turning towards notes beside you. Does the sound change?

Try placing your notes where you can read them without turning away.

The sweet spot is something you can hear

Microphone position is a practical choice involving your voice, room and equipment. Today’s experiment gives you evidence for that choice.

Three short recordings. One useful decision. A clearer voice for your next podcast.

More next Saturday

 

Why Did Knitting Machines Disappear While Embroidery Machines Became Cleverer?

Craft has not disappeared. So why did some of its most promising machines fade from view just as computers should have opened up new possibilities?

The Future That Seemed About to Arrive

Years ago, domestic knitting machines seemed to offer a glimpse of the future.

There were machines from several manufacturers, with accessories, pattern systems and increasingly sophisticated capabilities. Punched cards could select needles to create repeating designs. Electronic controls then offered the prospect of greater freedom: patterns that could be stored, altered and eventually designed on a computer.

It seemed reasonable to imagine what might follow. Design a jumper on screen, choose the colours and measurements, and let the machine help turn the idea into something wearable.

Yet the broad domestic market contracted. Meanwhile, sewing machines remained familiar household tools, and computerised embroidery developed into a remarkably capable creative technology.

My own visits to craft shows leave me wondering about this contrast. There is plenty of enthusiasm for making things, but knitting machines and sophisticated looms seem much less visible than I might expect.

If people still enjoy craft, and the technology can do more than ever, what happened?

First, These Technologies Have Not Entirely Disappeared

It is important to distinguish a shrinking domestic market from a technological dead end.

The Knitting & Crochet Guild identifies Silver Reed as the main current producer supplying machines into the UK and describes an active second-hand market. Computer connections also survive: Soft Byte’s DesignaKnit links support pattern transfer on compatible electronic machines and interactive instructions for shaping and colour changes. Knitting & Crochet Guild

Computer-assisted weaving survives too. Louët offers electronic dobby systems, while Digital Weaving Norway produces the TC2 digital Jacquard loom for creative and specialist work. www.louetdobby.com

These are different approaches. A dobby selects combinations of shafts, which lift groups of warp threads. A Jacquard system allows individual warp-thread selection. Weaving did not follow one universal path from manual to punched cards and then back again.

Spinning has also developed beyond treadle-driven wheels. Ashford’s e-Spinner 3, for example, uses a variable-speed electric motor. The spinner still performs the skilled work of drafting—drawing out the fibres to control the developing yarn. Ashford Wheels and Looms

The question becomes more precise: why did these developments remain specialist interests while computerised embroidery became so much more visible?

Cheap Clothes Changed the Proposition

One explanation, in my view, is that making an everyday garment and personalising one face very different competition.

Imagine knitting a plain jumper. Before counting your time, you need yarn, suitable equipment and the knowledge to produce the fit and finish you want. A shop-bought jumper provides an immediate alternative.

A handmade jumper may offer better materials, a distinctive design or a more suitable fit. Nevertheless, the machine has a difficult sales argument if it is presented mainly as a way to obtain ordinary clothing cheaply.

Now imagine embroidering a boat’s name onto a purchased jacket, or adding a personal design to a cushion.

The starting item is already available. The machine adds individuality.

Cheap mass-produced clothing can weaken the economic argument for making a garment while strengthening the appeal of decorating one.

That helps explain why personalisation is such a useful market. A club emblem, an unusual gift or a design created for one particular person offers something beyond the standard item on the shelf.

For a sailing club, for example, a machine might personalise shirts, bags and presentation gifts. The value comes from their connection to the club and its members.

Computers Solved the Pattern—but Not the Whole Project

There is a considerable difference between computerising one operation and simplifying an entire craft.

An electronic knitting machine can help select needles for a pattern. However, depending on the machine and project, the maker may still need to:

• Choose yarn suitable for the machine’s gauge.
• Knit and measure a tension sample.
• Cast on and manage the work correctly.
• Shape pieces and transfer stitches.
• Recover dropped stitches.
• Join and finish the garment.

Being able to send an intricate design from a computer does not automatically make those tasks straightforward.

Weaving illustrates the same point. Louët explains that an electronic dobby simplifies pattern selection while warping, threading, weaving and beating remain manual activities. Louët dobby

Think of planning a striped scarf. Designing the colour sequence may be enjoyable and quick. Preparing the warp and threading the loom still require care before the first section of scarf appears.

The computer may remove a complicated decision repeated hundreds of times, yet leave the beginner’s largest obstacle largely untouched.

A machine can become much more capable without becoming much easier to start using.

The First Successful Project Matters

As a teacher, I am interested in the distance between someone’s initial enthusiasm and their first experience of success.

A beginner needs a project that says: “I can do this, and I understand enough to try again.”

Embroidery can provide a relatively contained first task: select a prepared design, use appropriate fabric and stabiliser, hoop it carefully, and stitch a sample.

There is still plenty to learn. Poor hooping, unsuitable stabiliser or an inappropriate design can spoil the result. Creating your own embroidery files introduces further skill. However, a small prepared design gives the learner a clear and limited objective.

Compare that with expecting a new machine knitter to produce a fitted jumper as their first achievement.

A better introduction might be a successful sample, followed by a simple scarf or cushion panel. The principle is familiar from teaching science or making videos: limit the first task so that the learner can see what worked and understand why.

Manufacturers need to sell that learning journey alongside the machine.

A list of impressive features may attract attention. A well-designed first project helps turn a purchaser into a regular user.

Faster Is Not Always What a Craftsperson Wants

There is another complication: people do not necessarily practise a craft because they need more finished objects.

Someone may enjoy knitting because it is portable, quiet and sociable. They can work on a few rows while talking, travelling or listening to music.

A knitting machine offers a different experience. It needs a suitable working space, preparation and attention to its operation. For some people, that is wonderfully absorbing. For others, it removes precisely what they enjoyed.

The same distinction applies to spinning. An electric spinner can remove treadling while preserving the feel of fibres passing through the hands. Automating the fibre handling itself would change the activity much more fundamentally.

The best automation removes a difficulty that the maker wants removed.

There is no single correct balance. Some people want the physical rhythm; others want to concentrate on design, colour and structure. Successful products need to understand which experience their customers are buying.

Sewing Has More Than One Reason to Survive

Sewing machines also have an advantage in the variety of useful jobs they can perform.

A machine might be used to shorten curtains, repair a seam, alter trousers, make a cushion, assemble a costume or create an original garment. The owner does not have to become a committed dressmaker to find it useful.

Embroidery adds another set of possibilities, including personalised gifts and small-business work. Brother’s current range explicitly spans domestic sewing, embroidery and business applications. 100 years of innovative excellence

That breadth matters. Different customers can justify similar equipment for different reasons.

A knitting machine is more specialised, and different gauges suit different yarn ranges. Someone who enjoys moving freely between very fine yarn and chunky wool may encounter limitations that knitting needles do not impose in the same way. Knitting & Crochet Guild

A Good Machine Needs Teaching, Repairs and Confidence

A craft machine is easier to buy when a prospective owner can see it demonstrated, take lessons and obtain help when something goes wrong.

Without that support, a minor technical difficulty can become a reason to put the machine away.

I suspect this helps explain how a market can contract even when the equipment remains impressive. Fewer buyers can mean fewer dealers and classes. Reduced access to advice then makes buying less attractive, further weakening demand.

Long-lived machines create another commercial complication. A thriving second-hand market is excellent for access and reuse, but it does not necessarily provide enough new-machine sales to fund continual development.

Electronics can introduce additional dependencies too. A mechanical machine may remain repairable while an older electronic model also needs a compatible controller, connection or software solution.

The existence of modern links for older knitting equipment is encouraging, but compatibility still has to be considered. Soft Byte continues to offer connections for several families of machines, illustrating the value of maintaining that bridge between generations of equipment. softbyte.co.uk

What a Craft Show Can—and Cannot—Tell Us

A busy craft exhibition shows that people want to make things. It does not tell us that every kind of craft machine has an equally strong market.

My suspicion is that stand economics contribute to what we see. Yarn, fabric and small accessories can generate many purchases during a day. A large specialist machine may need a lengthy demonstration and considerable follow-up before anyone buys it.

The type of exhibition matters as well. A general craft show presents a different picture from a specialist weaving event or textile-design workshop.

Even so, visibility matters. If visitors never encounter machine knitting, they cannot discover that it might appeal to them.

A lack of demonstrations can help perpetuate a lack of demand.

Could the Domestic Textile Machine Have Another Moment?

I can see the appeal of a modern knitting system that combines clear design software, reliable connections, useful diagnostics and guided projects.

For someone interested in both craft and engineering, designing a knitted structure could be as engaging as preparing a laser-cut design or a piece of embroidery.

But the opportunity requires more than a touchscreen. It needs a complete route to success:

• Suitable yarn recommendations.
• Clear setup instructions.
• Manageable first projects.
• Accessible repairs and spare parts.
• An active community offering help and inspiration.

Shared access could help too. A workshop, college or makerspace could offer equipment and tuition without requiring everyone to find room for a specialist machine at home.

There is an attractive demonstration waiting to happen: design a small two-colour motif, knit a sample, examine how the stitches create it, and then use the result in a finished object.

That connects the computer screen to something tangible and understandable.

Conclusion: Invention Is Only the Beginning

I do not think the diminished visibility of knitting machines means that people lost their creativity. Nor does it mean that computerised knitting and weaving were poor ideas.

My reading is that embroidery found a favourable combination of personalisation, achievable projects and commercial uses. Domestic knitting and weaving faced a different mixture of preparation, specialist skills, physical space and competition from inexpensive finished textiles.

Technology succeeds when its capabilities fit the lives of enough people—and when those people can learn it, maintain it and enjoy using it.

For a craft machine, the decisive question may be:

Does it help me make something I care about, while preserving the parts of making that I enjoy?

That is why the absence still interests me. The creative possibilities have not vanished. Perhaps what is missing is a better bridge between an impressive machine and an enthusiastic beginner.

Do you still use a knitting machine, an electronic loom or an electric spinner? What helped you keep using it—or what caused you to put it away?

 

Why Do Green Trees Turn White in Infrared Photographs?

The trees are green, the sky is blue — so why does an infrared camera disagree?

Imagine photographing a familiar landscape: a leafy tree beside a river, grass along the bank and a few clouds drifting overhead.

In an ordinary photograph, everything looks reassuringly familiar.

Now photograph the same scene in near-infrared.

The foliage may appear almost white. The grass becomes pale and luminous. The blue sky turns dramatically dark, while clouds stand out against it.

It can look like a summer landscape transformed into a winter dream.

Yet there is no frost on the leaves. The camera is recording light that our eyes cannot see.

For me, this is one of photography’s most interesting possibilities: a familiar place can become a way of exploring unfamiliar science.

First, What Is Near-Infrared Photography?

Human vision covers only a small part of the electromagnetic spectrum, approximately from violet at 400 nanometres to deep red at 700 nanometres.

Beyond red lies infrared.

For this kind of landscape photography, we are interested in near-infrared, commonly recorded at wavelengths between roughly 700 and 1,000 nanometres, depending on the equipment.

The important distinction is that these photographs primarily record reflected light, usually from the Sun. They are different from thermal images, which detect longer infrared wavelengths emitted by objects at everyday temperatures.

A white tree in a near-infrared photograph is therefore not necessarily warmer than its surroundings. It is returning a strong near-infrared signal to the camera.

The Wood Effect: A Summer Tree That Looks Snow-Covered

The bright appearance of foliage in infrared photographs is known as the Wood effect, named after the infrared photography pioneer Robert W. Wood.

Despite the name, it refers to the behaviour of foliage rather than timber.

To understand it, we need to look inside a leaf.

A leaf is much more than a thin green surface. It contains layers of cells, pigments and internal air spaces. Together, these determine which wavelengths are absorbed, transmitted or scattered back out.

In visible light, chlorophyll absorbs strongly in the blue and red regions. Relatively more green light escapes, which is why most healthy leaves look green.

Beyond the visible red region, the balance changes. Pigment absorption becomes much weaker, while the leaf’s internal structure scatters near-infrared radiation strongly.

Some of that radiation passes through the leaf; some emerges back towards the camera. The strong reflection arises from repeated scattering at boundaries between materials with different optical properties inside the tissue.

The leaf’s green appearance tells us about its pigments. Its near-infrared brightness tells us much more about its structure.

That is a lovely connection between photography, physics and plant biology.

Why Does That Become White in the Photograph?

Near-infrared radiation has no colour that our eyes can directly perceive.

The camera and editing software must turn the recorded signal into something visible.

In a black-and-white infrared image, a strong signal can be represented by a light grey or white tone. A weaker signal becomes darker.

That is why strongly reflecting foliage can appear white.

However, white leaves are not guaranteed straight out of the camera. An infrared file may initially look reddish, pinkish or otherwise unfamiliar. White balance, filter choice, exposure and processing all influence the final result.

The bright foliage has a physical basis; its exact appearance on screen also involves photographic choices.

Infrared photography combines measurement with interpretation.

Why Does the Blue Sky Often Turn Dark?

The blue of a clear sky comes largely from Rayleigh scattering.

Air molecules scatter shorter wavelengths much more strongly than longer ones. Blue light is scattered strongly into our line of sight; near-infrared is scattered much less.

Consequently, a clear patch of sky can deliver a relatively weak near-infrared signal and appear dark in the photograph. Clouds scatter light differently because their droplets are much larger than air molecules, so they can remain bright against that darker background.

This creates one of infrared photography’s most striking combinations:

Pale foliage, dark sky and bright clouds.

The sky will not always become black. Haze, cloud cover, viewing direction, filter choice and editing all affect the result.

Grass, Leaves, Wood and Water: Expect Different Responses

A useful starting point is to stop assuming that an object’s visible colour predicts its infrared brightness.

Subject	Common near-infrared appearance	What to notice
Leafy trees	Bright, sometimes almost white	Species, leaf structure, lighting and canopy density affect the result.
Grass	Often pale and bright	Living grass may separate strongly from paths, soil and shadows.
Bark and dead wood	Often darker than foliage, with varied tones	Texture, moisture and surface condition matter; wood does not automatically show the foliage effect.
Open water	Frequently dark	Absorption and reflections both influence its appearance.
Clouds	Often bright against clear sky	Cloud type and lighting determine how dramatic the contrast becomes.

Water is particularly interesting. It often looks dark because relatively little near-infrared returns from within it, but its surface can still reflect bright clouds, foliage or sunlight.

A river can therefore contain both dark stretches and bright reflections.

For a Thames landscape, I would look for a composition combining leafy banks, open water and a distinctive tree. That gives several different materials to compare within one frame.

Sunlight Matters — but So Does Wind

Near-infrared photography depends on illumination.

Direct sunshine often makes a useful starting point because it supplies plenty of near-infrared radiation and can create strong separation between sunlit foliage, shadows and sky.

A bright day with some cloud can be especially rewarding.

This offers an interesting change of approach for landscape photographers. The midday sunshine that can feel harsh in conventional photography may produce striking infrared results.

Overcast conditions can still work, although the effect may be gentler and exposures may need to be longer.

Wind creates a different problem.

If your setup needs a long exposure, moving leaves may blur even when the trunk remains sharp. That can be an attractive creative effect, but it makes a scientific comparison harder.

For a first attempt, choose bright conditions and a relatively still tree.

In autumn, look for a tree still carrying plenty of leaves, or try an evergreen. Repeating the exercise as the seasons change would make an interesting longer project.

What Equipment Do You Need?

Most ordinary digital cameras contain an internal filter that suppresses infrared, helping them produce reliable visible colours.

There are two common ways to record near-infrared:

• Use an infrared-pass filter on a suitable unmodified camera. This blocks most visible light, but the camera’s internal infrared suppression may mean long exposures or very limited results.

• Use a camera converted for infrared, or a full-spectrum camera with an appropriate filter. These arrangements can record near-infrared more efficiently.

Compatibility matters: an external filter does not make every camera equally capable of infrared photography.

A filter around 720 nanometres is a common starting point for infrared work and can support both monochrome and some false-colour processing. Longer-wavelength options, such as 850 nanometres, tend towards a more strongly monochrome appearance, with different exposure requirements.

For this exercise, the aim is to use a setup known to record near-infrared reliably.

Practical Demonstration: One Tree, Three Interpretations

Your task: photograph the same tree in visible light and near-infrared, then compare both with an ordinary black-and-white conversion.

What you need

• A normal camera or phone.

• A suitable near-infrared camera and filter arrangement.

• A tripod, particularly if infrared exposures are long.

• A leafy tree, ideally with sky and another material in the frame.

• Editing software capable of producing a monochrome image.

1. Choose a useful scene

Include leaves, trunk, sky and grass. If possible, add water, a wall or a path.

The different surfaces make the comparison more revealing.

2. Take the visible-light photograph

Record the scene normally.

Note the conditions: sunshine or cloud, time of day and whether the leaves are moving.

3. Take the near-infrared photograph

Keep the viewpoint and framing as similar as possible, and photograph it soon afterwards so the lighting remains comparable.

Use settings appropriate to the infrared setup rather than copying the visible-light exposure. Check focus and the histogram, protecting detail in the brightest foliage.

4. Make a normal black-and-white version

Convert the visible-light photograph to monochrome.

Now you have three images:

1. Visible colour.

2. Visible light converted to black and white.

3. Near-infrared rendered in black and white.

That third comparison is especially useful. Removing colour from a normal photograph does not reveal infrared information that the camera did not record.

5. Compare what changed

Ask:

• Did the leaves become brighter relative to the trunk?

• Did the grass separate more clearly from the path?

• Did the sky darken?

• Did clouds become more prominent?

• Which objects changed least?

Keep the processing reasonably restrained at first. Extreme contrast adjustments can make the images dramatic while obscuring what you are trying to compare.

Turn the Experiment into a Creative Landscape

Once you understand the differences, use them deliberately.

A pale canopy can become the main subject against a dark sky. A shaded trunk can provide structure within bright foliage. A path can lead the viewer through an otherwise luminous scene.

Look for:

• An isolated tree with a recognisable outline.

• A winding path through leafy woodland.

• Bright foliage beside dark water.

• Architecture framed by vegetation.

• Clouds that balance the shape of a tree.

When editing, preserve texture in the leaves. If every bright area becomes pure white, much of the interest disappears.

Deep blacks can be effective, but the strongest image does not always need the most extreme contrast.

The question I would ask is: what does infrared reveal that makes this particular composition more interesting?

Can an Infrared Photograph Tell You Whether a Plant Is Healthy?

Near-infrared observations are valuable in vegetation monitoring because changes in plant tissue and canopy structure can affect the recorded signal. Scientists compare visible and near-infrared measurements to investigate vegetation.

However, a bright tree in a processed photograph is not, by itself, a reliable plant-health diagnosis.

Species, illumination, leaf angle, exposure and editing all affect brightness. Scientific assessment needs controlled measurements and appropriate interpretation.

For teaching, that limitation is useful. It opens a discussion about the difference between a striking observation and sufficient evidence for a conclusion.

A Familiar Tree Can Teach Us to See Differently

What appeals to me about this demonstration is that the subject can be completely ordinary.

There is no need to travel to an exotic location. A tree in the garden or beside a familiar riverside path can connect photography with pigments, living tissue, scattering and the electromagnetic spectrum.

The tree has not changed between photographs.

The wavelengths we record have changed, revealing a different set of properties.

This week’s challenge: choose one tree and make the three-image comparison.

Look beyond the white leaves. Study the trunk, the shadows, the sky and the ground beneath it.

Our eyes give us a wonderfully useful view of the world. Infrared photography reminds us how much more there is to discover.

 

Could Every Student Have Their Own AI Revision Course?

Why should two students be given exactly the same revision sheet when their weaknesses are completely different?

Two students sit the same mathematics test. Both score 50%.

On paper, they appear to need roughly the same help. Give them another revision worksheet, explain the difficult questions and encourage them to practise.

But look more closely.

One student understands the mathematics but struggles to translate written questions into equations. The other can set up the equations perfectly well but makes mistakes when rearranging them.

Their marks are identical. Their learning needs are very different.

This is where AI could become particularly useful: helping a teacher build a revision course that changes as the student learns.

Imagine a course that keeps track of topics covered, mistakes made, hints needed and confidence reported. Each new task would be chosen using that evidence.

That is a much more interesting possibility than simply asking AI to produce another twenty questions.

From a Revision Folder to a Changing Learning Plan

A conventional revision course usually has a fixed structure:

Read the notes. Watch the explanation. Answer the questions. Check the answers. Move to the next topic.

There is nothing inherently wrong with that. A clear sequence can be extremely helpful.

However, students rarely develop at exactly the pace assumed by the course. Some need to revisit earlier knowledge. Others are ready for harder applications. Some can complete a question immediately after an explanation but struggle when it returns a week later.

An individual AI-assisted course could respond to these differences.

It might suggest:

• a short diagnostic task when the cause of a mistake is unclear;
• a worked example when a method needs explaining;
• an independent question to check whether the explanation helped;
• a harder application when the student is ready;
• a later review to check whether the learning has lasted.

The useful feature is the decision about what comes next.

Generating resources is only one part of the job.

The Student Profile: More Useful Than a Single Percentage

For this approach to work, the course needs a useful record of learning.

That record need not be an enormous database. A carefully maintained table could provide a starting point.

Here is an illustrative profile for a fictional GCSE mathematics student:

Area	Evidence from recent work	Confidence	Possible next step
Solving linear equations	Solves simple equations independently; errors with brackets	Fairly confident	Check expansion before harder equations
Simultaneous equations	Solves a prepared pair; struggles to form equations from a situation	Low	Practise translating statements into equations
Fractions	Adds simple fractions; inconsistent with mixed numbers	Low	Short diagnostic task
Graphs	Plots accurately; confuses gradient with intercept	High	Explain and compare the two
Circle calculations	Uses the formula correctly once radius is identified	Moderate	Practise identifying radius and diameter

The confidence column is particularly revealing.

A student who is confidently wrong needs different feedback from one who is correct but convinced they cannot do the subject.

However, confidence is self-reported information, not proof of understanding. It should sit alongside actual work.

I would also want to record whether an answer was produced independently, after a small hint, or while following a complete example.

Those are three different levels of support. A course should recognise the difference.

A Practical Example: What Is Going Wrong With Simultaneous Equations?

Suppose a student struggles with this question:

Two adult tickets and one child ticket cost £23. One adult ticket and two child tickets cost £19. Find the price of each ticket.

There are several possible obstacles.

The student might not understand what the letters represent. They might struggle to form the equations. They might know the method but make a subtraction error. They might find the wording overwhelming.

Giving them ten more questions of the same type could repeat the difficulty without explaining it.

A better next step is to investigate.

Step 1: Separate the Algebra From the Interpretation

Present the equations directly:

2a + c = 23
a + 2c = 19

Can the student solve them?

If they can, there is evidence that the main difficulty may lie in translating the situation into mathematics. That is a working hypothesis, rather than a final diagnosis.

If they cannot, check the underlying algebra.

Can they solve:

3x = 12?

Can they rearrange:

x + y = 9

to make y the subject?

Can they subtract one equation from another without changing signs incorrectly?

The next task should help distinguish between these possibilities.

Step 2: Give Practice That Addresses the Evidence

If interpretation appears to be the obstacle, ask:

An adult ticket costs a pounds and a child ticket costs c pounds. Write an expression for the cost of two adult tickets.

Then:

Write an expression for two adult tickets and one child ticket.

Then:

Their total cost is £23. Turn your expression into an equation.

The student is practising the part that caused the difficulty.

If algebra is the obstacle, work with simple prepared equations first, then return to the ticket problem.

Step 3: Return to the Complete Question

The original problem gives:

Adult ticket = £9
Child ticket = £5

The student should check both conditions:

2 × £9 + £5 = £23
£9 + 2 × £5 = £19

Then try a fresh question independently.

Successfully following an explanation is encouraging. Solving a new problem provides stronger evidence that the method is understood.

Step 4: Check Again Later

Revisit the skill in a later session, with different wording and numbers.

If the student can still use it independently, the profile can record stronger evidence of progress.

The course should respond to what the student can do today, while checking what they can still do later.

Personalisation Should Sometimes Make Work Harder

It is easy to imagine personalised learning as a system that keeps making questions easier.

That would be a serious limitation.

A useful course should also recognise when a student needs greater challenge.

For simultaneous equations, that might mean progressing towards:

• equations requiring multiplication before elimination;
• negative or fractional solutions;
• unfamiliar contexts;
• questions where the student must choose the method;
• explaining why someone else’s solution is incorrect.

The aim is to make success increasingly independent.

A student who has mastered the basics should not spend an entire revision session repeating them simply because those questions are easy to generate.

Equally, a student who struggles should not be permanently confined to elementary work. The course needs a route back towards the demands of the examination.

The Same Approach Could Help in Science

Consider a biology student who writes:

Antibiotics kill viruses.

A useful response should do more than replace that sentence with the correct one.

The next task could ask the student to distinguish between a bacterial disease and a viral disease. A further question could check whether they understand why antibiotics are unsuitable for treating a viral infection.

Later, the course could revisit the idea within a question about antibiotic resistance.

In physics, a wrong calculation might lead to separate checks on:

• selecting the equation;
• rearranging it;
• converting units;
• substituting values;
• interpreting the answer.

In chemistry, an incorrect concentration calculation might reveal a weakness in unit conversion rather than chemical understanding.

The same principle applies across subjects:

A wrong answer is the beginning of an investigation.

It does not automatically reveal its own cause.

Confidence, Independence and Knowing When You Need Help

One of the most valuable questions a revision course could ask is:

Before you check, how confident are you that your answer is correct?

Followed by:

What makes you think that?

Over time, students could compare their judgement with their performance.

Some might discover that they know more than they think. Others might realise that recognising a familiar question is different from being able to solve it.

This connects with metacognition: learning to plan, monitor and evaluate your own learning. The Education Endowment Foundation’s guidance emphasises explicitly teaching and modelling these processes within subject teaching. That supports the educational principle behind this approach; it does not establish that any particular AI course will deliver it successfully. EEF

An AI-assisted course could prompt that reflection.

A teacher still needs to help students make sense of it.

My Perspective as a Tutor: The Record Must Leave Room for the Person

After around forty years of teaching, I am very aware that a student’s performance cannot always be explained by their subject knowledge.

A student may understand something during a conversation but struggle to express it on paper. They may rush, misread a word, lose concentration or arrive tired.

A record saying “three incorrect answers” cannot explain all of that.

In my own teaching, I would want AI to help organise the evidence and prepare useful follow-up work. I would still want to talk to the student, look at their working and ask why they chose a particular approach.

Sometimes a question needs simplifying.

Sometimes it needs drawing.

Sometimes we need to put the worksheet down and demonstrate the idea.

That matters especially in science. A student may answer questions about an experiment yet still need practice setting up apparatus, making measurements or explaining an unexpected result.

An individual revision course should be able to recommend practical work, discussion and handwritten practice alongside screen-based activities.

The Course Must Remain Anchored to the Examination

Personalisation needs a destination.

For examination students, the course must take account of the qualification, specification, tier where relevant, and the skills assessed.

A course could become very good at practising a narrow weakness while overlooking an entire section of the syllabus.

I would therefore want two views of progress:

The close view: What does this student need next?

The whole-course view: What remains to be covered and revisited?

Students also need mixed questions where the topic and method are not announced in advance.

A worksheet headed “Simultaneous Equations” provides a substantial clue. An examination may expect the student to recognise the need for simultaneous equations themselves.

Eventually, practice must include choosing the approach as well as carrying it out.

AI Can Produce Convincing Mistakes

A personalised course is only useful if its questions, explanations and feedback are dependable.

AI can produce an incorrect answer, an ambiguous question or an explanation that sounds persuasive but does not address the student’s error.

It may also infer too much from too little evidence.

One incorrect answer should not become a permanent label in a student profile.

For that reason, I would want teacher checks, verified solutions and a clear way to challenge the feedback. Uncertain conclusions should remain marked as uncertain.

The Department for Education’s guidance similarly stresses that generative AI cannot replace expert subject knowledge and professional judgement. GOV.UK

There is another practical limitation: a normal chatbot conversation should not be assumed to maintain a complete, accurate learning record indefinitely.

A continuing course needs a deliberately maintained record that can be reviewed and corrected.

A Learning Profile Should Contain Only What Is Needed

The educational value lies in information such as:

Confuses radius and diameter; correctly identifies both after a diagram prompt.

It rarely requires a student’s full name, home address, family circumstances or detailed personal history.

A sensible starting point is a minimal learning record, using a student code where appropriate, in a setting approved for the intended use.

Before introducing a tool, check its suitability for the student’s age and how it handles the information entered.

The profile should also be written respectfully.

“Needs further practice interpreting ratio questions” is useful.

“Bad at maths” is a judgement that offers no next step.

What Could a Twenty-Minute Session Look Like?

Here is one possible structure:

Time	Activity	Purpose
3 minutes	Recall questions from earlier work	Check what remains accessible
4 minutes	Short diagnostic task	Investigate a current difficulty
6 minutes	Explanation and targeted practice	Address the identified gap
5 minutes	Fresh independent application	Check whether support can be withdrawn
2 minutes	Reflection and record update	Plan the next session

The timings would change with the learner.

A student who needs more thinking time should receive it. Another may be ready to move quickly to a demanding application.

The record might end with:

Formed two equations independently. Needed one hint when subtracting negative terms. Next session: brief sign check, followed by a fresh word problem.

That is much more useful than simply recording “completed worksheet”.

A Small Starting Point Is Enough

Teachers and tutors do not need to begin with an elaborate automated platform.

A manageable trial could involve one topic, a small learning record and a teacher-reviewed set of follow-up tasks.

A useful instruction to AI might be:

Using the anonymised learning record below, propose a short revision session on simultaneous equations. Start with two diagnostic questions that help distinguish difficulty forming equations from difficulty solving them. Suggest different next steps depending on the answers. Include one independent application and a later review question. Provide checked solutions separately. Mark uncertain interpretations as hypotheses for the teacher to review.

This would produce a proposal for the teacher to assess.

The important test is whether the resulting tasks help the student learn more effectively—and whether the time spent maintaining the system is worthwhile.

Could Every Student Have Their Own Course?

Potentially, yes.

AI could make it more practical to prepare different explanations, tasks and review sequences for different learners.

But giving every student access to a chatbot would not, by itself, create an individual revision course. Access to suitable devices, clear guidance, dependable resources and appropriate support would still matter.

The strongest version would combine a carefully maintained learning record with teacher judgement and regular evidence of independent progress.

Two students might be working towards the same examination while following different routes through their revision.

One might need help understanding the question.

Another might need help selecting the method.

A third might be ready to explain why the method works.

They all deserve a suitable next step.

The most exciting possibility is that AI could help us act more consistently on something good teachers already know: the next question matters, and it should be chosen with the learner in mind.