Filming Your Projected Slides vs Proper Scanning: Why Photographing the Wall Loses the Picture
Maria C
Photographing your slides off a projector screen — or filming the wall while a Kodak Carousel clicks through the tray — feels like digitising them. It is not. A working Kodak Carousel S-AV 1010 throws roughly 14 line pairs per millimetre onto a 60-inch screen; a phone re-photographing that screen recovers even less. The Kodachrome emulsion in the mount holds about 60 lp/mm, and a Nikon Super Coolscan 9000 ED reads 78 lp/mm straight off the film. Filming the projection throws away roughly four-fifths of the picture before your camera shutter ever opens — and every minute the lamp is lit, the slide fades. This guide shows you, frame by frame, exactly what is lost, and when the honest answer is to scan the film itself rather than photograph its shadow on a wall.
The short answer: there are four ways to digitise 35mm slides, and they are not equal. Best to worst for image quality: (1) a dedicated film scanner such as the Nikon Coolscan 9000 ED, which reads the emulsion directly at ~3900 dpi; (2) a flatbed with a film adapter (Epson V850 Pro) for bulk DIY; (3) a phone or DSLR over a light box, workable for thumbnails; and far behind, (4) photographing a projected slide off the wall — the one method that both captures the least and physically damages the film. For the full DIY-versus-professional ladder, see our guide to every slide method. This article is about why that fourth route loses the picture — and what a proper scan recovers.
Key takeaways
- A projector image is the slide's shadow, not the slide. You can only capture what the lens throws — about 14 lp/mm — never the ~60 lp/mm the film actually holds.
- A phone photographing a projection works in roughly 8-bit and reads about Dmax 2.2, crushing the deepest two stops of every frame to black. Slide film needs a density range above ΔD 3.6 to read shadow detail.
- Projection damages the original: Kodachrome shows measurable fade after about one hour under the lamp, and roughly 200 projection cycles equals a visible density loss.
- A proper scan reads the film, not the screen: a Nikon Coolscan 9000 ED resolves 78 lp/mm at the centre, reads to a claimed Dmax 4.8, and removes dust optically with infrared Digital ICE.
- EachMoment scans 35mm mounted slides in our UK lab from £0.79 per slide (down to £0.47 at volume), film untouched by a projector lamp.
The mistake everyone makes: filming the wall
It is the most natural idea in the world. The projector still works, the tray still drops slides into the gate, and the picture still lands on the wall. So you set up a tripod, point a phone or camera at the screen, and click through the carousel. An afternoon later you have a folder of JPEGs and a feeling that the job is done.
It is not done, because you have not photographed your slides. You have photographed a projector lens's interpretation of your slides, bounced off a textured wall, re-focused by a phone camera, and compressed to 8-bit JPEG. Three lossy steps sit between the film and your file, and each one throws away detail that was sitting safely in the emulsion the whole time.
The slide itself is a remarkable thing. A 35mm Kodachrome frame from 1965 holds resolution equivalent to roughly 20 megapixels and a density range deep enough that a properly exposed shadow still carries readable detail. The question is never whether the detail is there. It is whether your capture method can reach it. Filming the projection cannot.
What you actually capture, drag the handle to see
Below are three different 35mm slides, each shown two ways: on the left, the kind of image you get photographing a projected or back-lit slide with consumer kit; on the right, the same frame read directly off the film on our Nikon Coolscan 9000 ED at 4000 dpi. These are the same source frame in each slider — drag the divider and watch the detail that the projection route never recorded reappear.
A 1970s family slide that had spent years going through a projector. Left: re-photographed from the projection — the hot-spot in the middle blows the highlights while the corners fall soft. Right: read off the film, the corners are sharp and the shadow under the trees holds detail. Same frame, two capture routes.
Notice what changes and what does not. The composition is identical — it is the same photograph. What returns on the right is the information: fine texture in hair and fabric, gradation in the sky, shadow detail that the projection route had folded into a flat black. None of that was added by the scanner. It was always in the film. The projection simply never let your camera see it.
A faded Ektachrome frame. A phone photographing the projection inherits the projector lamp's colour and the film's magenta shift together, with no way to separate them. Read off the film into a colour-managed 16-bit pipeline, the cast can be corrected against a known profile and the dust removed with infrared Digital ICE.
A third slide, captured cheaply versus scanned properly. The blocking you see in the shadows on the left is 8-bit JPEG compression giving up on detail it never properly recorded. The right-hand frame reads the emulsion directly.
The number that settles it: resolving power
You do not have to take the sliders on trust. The difference is measurable, and we measure it. Using a USAF-1951 resolution target placed in the slide gate, here is what each capture method can actually read off the same 35mm Kodachrome frame, in line pairs per millimetre.
Read the chart and the whole argument collapses into one fact: the two projection-based methods sit below the film's own resolving ceiling, while the scanner sits above it. When the scanner reads more than the emulsion can hold, the film itself becomes the limit — which is exactly what you want. When the projector caps you at 14 lp/mm, the lens is the limit, and no amount of editing afterwards puts back detail that was never captured.
The second loss: shadow detail and dynamic range
Resolution is only half the story. A slide is a high-contrast medium — a well-exposed transparency can hold a density range deeper than almost any print. Reversal (slide) film should be read by a device whose density range exceeds ΔD 3.6, or the deepest shadows clip to a featureless black. A phone photographing a bright projection works in roughly 8-bit and reads to about Dmax 2.2. It physically cannot hold the bottom of the tonal range.
This is why projection captures so often look harsh: bright midtones, blown highlights from the lamp's central hot-spot, and shadows that have simply gone to mud. Consumer 8-bit slide scanners suffer the same problem from the other direction — they clip roughly the deepest two stops of an under-exposed slide and have no bits left to recover them. The detail is in the film. Reading it needs a device built for the film's contrast, and a 16-bit pipeline to hold the result.
The hidden cost: projection damages the slide
There is a reason archivists flinch when they hear "I'll just project them and film the screen." Projection is not a neutral act. A slide projector lamp runs hot and bright, and the film sitting in the gate absorbs that energy directly. Kodachrome — celebrated for its dark-storage permanence of up to 185 years for its least-stable dye — is paradoxically poor under projection: measurable fade sets in after roughly one hour of cumulative lamp time, far faster than Fujichrome. Our densitometry estimates that around 200 projection cycles of 20-30 seconds each equate to a visible density loss in the most vulnerable dye layer.
So the projector-and-camera method has a cruel arithmetic. The slower and more carefully you work — lingering on each slide to focus, expose and shoot — the more lamp time you pour into the very frames you are trying to preserve. You are bleaching the original to make an inferior copy of it. A film scanner, by contrast, exposes the slide to a brief, cool LED or cold-cathode pass measured in seconds, with no sustained heat in the gate.
What a proper scan actually involves
"Scan the film, not the screen" is the principle. Here is the equipment that does it, and why each piece matters for a slide that a projector-and-phone setup cannot touch.
Nikon Super Coolscan 9000 ED
Dedicated 35mm + medium-format film scanner
Discontinued 2009; survives mainly in professional labs
- 4000 dpi native optical resolution; ~3900 dpi measured on USAF-1951
- Reads 78 lp/mm at centre — above Kodachrome's own ~60 lp/mm ceiling
- Claimed Dmax 4.8, so it reads into the deepest slide shadows
Digital ICE (infrared dust + scratch removal)
Removes dust and surface scratches optically, not by guesswork
Standard on the Coolscan 9000 ED; not available on Kodachrome's silver layer
- Infrared channel maps physical dust the visible scan cannot see
- Avoids the smearing a phone-on-a-screen capture bakes in permanently
- Applied per-slide, not as a global blur
Epson Perfection V850 Pro
Flatbed alternative for bulk 35mm and larger formats
Current; the practical flatbed ceiling
- 6400 dpi advertised; ~2300 dpi actually measured on 35mm film
- Claimed Dmax 4.0 — below the 9000 ED but far above any projection
- Wet-mount option for damaged or curled mounts
16-bit calibrated capture pipeline
Holds the dynamic range a slide carries; an 8-bit phone JPEG cannot
Lab standard, every slide
- Reversal film needs a scanner whose density range exceeds ΔD 3.6
- 16-bit per channel keeps shadow gradation an 8-bit JPEG crushes
- Colour managed to a known profile, not the camera's auto white balance
The thread running through all of it: a dedicated film scanner reads the emulsion directly, in a colour-managed 16-bit pipeline, with infrared dust removal that a camera pointed at a wall can never replicate. Digital ICE uses an infrared channel to map physical dust and scratches and remove them optically — it cannot work on Kodachrome's silver-retaining layer, which is one of the few honest caveats in slide scanning, but on Ektachrome, Agfachrome and the vast majority of colour slides it removes in seconds what would take hours of manual retouching on a projection capture.
Side by side: the four ways people digitise slides
If you are weighing the options, here is the whole landscape in one table. The first three are do-it-yourself routes; the fourth is sending the film to a lab. The figures are our own lab measurements and manufacturer specifications.
| Method | Resolving power | Shadow detail (Dmax) | Dust / scratch removal | Risk to the original |
|---|---|---|---|---|
| Photograph a projected slide | ~14 lp/mm (lens-limited) | ~2.2 — shadows clip | None; dust is baked in | High — lamp heat fades film |
| Phone + light box / app | ~11-18 lp/mm | ~2.2 (8-bit JPEG) | None | Low |
| Consumer flatbed (Epson V850 Pro) | ~2300 dpi measured | 4.0 (claimed) | Infrared ICE (most colour film) | Low |
| Lab film scan (Nikon Coolscan 9000 ED) | 78 lp/mm / ~3900 dpi measured | 4.8 (claimed) | Infrared Digital ICE, 16-bit | Minimal — brief cool pass |
The pattern is stark. The two projection routes lose on every axis that matters except the cost of equipment you may already own. If your slides are holiday duplicates you are happy to keep at thumbnail quality, photographing the projection is a fine afternoon's work. If they are the only record of a wedding, a parent, a childhood — the frames you would grieve to lose — the projection route quietly discards most of what makes them worth keeping, and damages the originals while it does so.
So when should you DIY, and when should you send them?
We are a scanning lab, so treat this as interested advice — but it is honest. There is a real place for DIY. If you have a small batch, a flatbed scanner with a film adapter, and the patience to scan at full resolution in 16-bit, you can get genuinely good results at home. We even publish a full DIY-versus-professional breakdown of every slide method for exactly that reader.
What never makes sense is the projector route as a preservation method. It is the one approach that combines the worst capture quality with the only real risk of damaging the film. If your projector still works and you want to enjoy a slide show, wonderful — project away. Just do not mistake filming that slide show for digitising your collection.
The break-even is simpler than it looks. A dedicated film scanner that resolves 78 lp/mm costs four figures and is long discontinued; an afternoon of careful flatbed scanning covers maybe a tray or two. For a typical loft find of several hundred slides, sending the film to a lab that already owns the Coolscan, the colour management and the infrared dust removal usually costs less than the scanner — and the originals come home unharmed by a projector lamp.
Scan the film, not the wall
Order a Memory Box, post your slides to our UK lab, and we read every frame on a Nikon Coolscan 9000 ED — 4000 dpi, 16-bit, infrared dust removal — and return your originals untouched by a projector lamp. 35mm mounted slides from £0.79 each, down to £0.47 at volume.
Get your slide scanning quote →Frequently asked questions
Can I just photograph my slides off the projector screen?
You can, but you will not be digitising the slides — you will be photographing the projector lens's low-resolution interpretation of them. A Kodak Carousel throws roughly 14 line pairs per millimetre onto the screen, while the film itself holds around 60 lp/mm and a proper scanner reads 78 lp/mm. You also lose the deepest shadows to the lamp's hot-spot and 8-bit JPEG clipping, and the lamp heat fades the original. For anything you care about keeping, scan the film instead.
Does projecting slides really damage them?
Yes, measurably. Kodachrome — superb in dark storage, where its least-stable dye is calculated to take up to 185 years to lose 20% density — fades comparatively quickly under a projector lamp, with measurable loss after roughly one hour of cumulative projection. Our densitometry estimates that about 200 short projection cycles equate to a visible density loss. A film scanner exposes the slide to only a brief, cool pass.
Is photographing a slide on a light box with my phone any better than the projector?
Slightly, because you remove the projector lamp's heat and hot-spot. But you still capture in roughly 8-bit, still cannot read into the slide's shadows (about Dmax 2.2 versus the ΔD 3.6 the film needs), and still get no infrared dust removal. Effective resolution lands around 11-18 lp/mm — better than the wall, far below a film scan.
What resolution does a slide actually deserve?
A 35mm frame holds roughly 20 megapixels of real detail and resolves around 60 lp/mm in the emulsion. We scan mounted slides on a Nikon Coolscan 9000 ED at 4000 dpi (about 3900 dpi measured), which reads 78 lp/mm at the centre — above the film's own ceiling, so the film, not the scanner, becomes the limit. That is the point at which you have captured everything the slide contains.
Will a scan fix faded or colour-shifted slides?
Often, yes — far better than a projection capture ever could. Reading the film into a colour-managed 16-bit pipeline lets us correct a magenta or cyan shift against a known profile and recover shadow detail an 8-bit capture has already thrown away. There are limits where dye collapse is severe; our guides on faded Ektachrome and Agfachrome and why Kodachrome looks so different once scanned explain what is and isn't recoverable.
How much does it cost to have slides scanned in the UK?
At EachMoment, 35mm mounted slides are scanned from £0.79 per slide, falling to as low as £0.47 per slide at archive volumes with the early-bird discount. You order a Memory Box, post your slides, and we scan, restore and return them. For a worked cost example by collection size, see our guide to what it costs to digitise a loft box, or get an exact figure from our slide scanning page.
Related reading
- How to Digitise Photo Slides: DIY vs Professional — the full four-method comparison.
- Carousel or magazine projector broken? — getting your slides digitised when the projector has died.
- Glass-mounted slides — why they jam consumer scanners and how a lab handles them.
- Mouldy and fungus-damaged slides — what can and cannot be recovered.