Verdict · Series 01 · Oakley Prizm
Prizm Low Light vs Clear vs Photochromic
Three lenses for riding and running once the light goes, and no honest side by side of them anywhere, so I shot one. Clear, Prizm Low Light and a photochromic, photographed through on the same scenes at one locked exposure, from a bright road at EV 12.9 down to a forest at dusk at EV 4.6 and then into full dark under a helmet lamp. That is 8.3 stops of light. The answer changes twice on the way down, and the photochromic's answer changes the most: it runs from 21.8% of the clear control to 80.2%, and it is at its most open in the darkest scene of the set.
01
The verdict
Prizm Low Light is the one I ride, and it is nearly free. Across all 8 scenes it holds 70.1 to 86.5% of the clear control, so it costs between 0.21 and 0.51 stops and never more. That is a light tint with contrast shaping, not a sunglass, and it is the only lens of the three that is never the wrong answer.
- One lens for everything: the photochromic. It covers the brightest days properly, it measured 21.8% on the bright road, and it does not fall apart in the dark the way its middle behaviour suggests it will. Best value of the three if you are buying once.
- Evening and after work rides in half light: Prizm Low Light. My favourite and the one in my frame most of the year. It reads the ground better than Clear does, it picks out the green slippery stuff, and from medium light downwards it won two of my four rounds.
- Only ever riding in genuine darkness: Clear. Under a helmet lamp and nothing else, raw transmission beats colour and Clear pulls ahead. It is also the cheapest lens in the comparison, which is a rare case of the right answer being the cheap one.
- Not Prizm Trail Torch and not Prizm Road. Both are in the shoot and both are out of this decision. The numbers are in section 11.
02
The three lenses, on paper
What the manufacturer says, before any of my own measurements. Stated transmission and the filter category that follows from it.
| Clear | Prizm Low Light | Photochromic | |
|---|---|---|---|
| VLT stated | 93% 1 | 75% 2 | 23 to 69% 5 |
| Category | CAT 0 | CAT 1 | CAT 0 to 3 |
| Sampled colour | #d8d2cb |
#dcd2d4 |
#a49690 |
| Look through | None, control specimen | Pale warm rose | Near neutral at the captured state |
| Mirror | None | None, near clear with a rose edge | Clear to black iridium, state dependent |
Two things worth pulling out of that table before any picture. Clear is not 100%. It passes about 93%, because a polycarbonate lens reflects a few percent off each surface whatever you do to it, so the gap from Clear to Prizm Low Light is 18 points and not the 25 points people assume. And Prizm Low Light is the only one of the three that shapes colour on purpose. Clear is meant to pass everything and the photochromic pushes every colour down together, near neutral by design. Low Light cuts back yellows and purples and lets blues and reds through, which is where the extra separation on roots and wet rock comes from. That is the trade this whole page is about: Low Light buys contrast with 0.51 stops at most.
The category row is manufacturer figures against the published filter bands, which is a different kind of claim from anything I measured. Categories 0 and 1 are treated as usable in poor light. Nothing below that is intended for driving at dusk or at night, which is exactly the mistake section 11 is about. I never blend a stated category and one of my own measurements into a single number.
03
Round 1: Town street, still decent light
EV 10.6 Late afternoon, traffic. Scene 7 of the shoot, town street.
The afternoon ride that leaves after work. There is light left, and both Clear and Prizm Low Light have too much of it: the scene is flat and the highlights run away. Low Light is the nicer of the two, it hands back a little saturation and the reds come up, but neither is doing any work. Drop the photochromic in and it is obvious. It is part activated here, so it is behaving as a proper bright light lens, and the blue line on the road that I cannot pick out through the other two is simply there.
Round 1 goes to Photochromic. Not narrowly, either. This is the light it was built for.
04
Round 2: The dark trail
EV 8.3 Deep shade. Scene 5 of the shoot, dark trail.
Sun still up, but under trees. Clear goes yellow and flat. Low Light adds that trace of red and the ground separates: roots stop reading as dirt. The photochromic has darkened again and now it is levelling the blown out sky behind the trail, which is genuinely better up there, but it has taken light away from the surface I am about to ride over.
Round 2 goes to Prizm Low Light. The contrast on the ground is worth more than the sky it recovered.
05
Round 3: Proper low light
EV 4.6 No artificial light. Scene 8 of the shoot, forest at dusk.
Actual dusk in the forest, no artificial light anywhere. Clear against Low Light is close here, closer than at any point higher up the ladder, and on raw detection it is roughly even. I still take Low Light on colour. What has changed is the photochromic: it is starting to be too dark, and it is holding back light I now need.
Round 3 goes to Prizm Low Light. On colour, over a Clear lens that has nearly caught up on light.
06
Round 4: Pitch black, helmet lamp only
EV 5.6 Helmet lamp only. Scene 9 of the shoot, night, headlamp.
A Sigma Buster 700 HL on its second strongest level, and nothing else. The bench at the back of the frame is the test: through Clear I can just make out the red wood planks, through Low Light I cannot. The pole reads better through Clear too, and the greens come out greener. When there is barely any light, more light wins, and the tint stops paying for itself. The surprise is the photochromic, which is not the darkest lens here at all.
Round 4 goes to Clear. Raw transmission beats colour once the light is gone.
07
See it yourself
Every frame above is in the comparison tool, along with twelve more lenses and the scenes I did not use. Pick two, zoom in, and check my reading against your own eyes. Clear against Prizm Low Light is loaded to start.
How brightness is shown. Every lens was shot at one locked exposure set on the bare sensor, so True exposure is the real light transmission. Your eye adapts, so a dark lens never looks that dark in life: Eye adapted raises each image by its measured transmission (perceived = transmission^0.5). Colour only normalises every lens to the same brightness to isolate the tint.
The three modes matter here more than anywhere else on this site. True exposure is the locked exposure exactly as the camera recorded it, so a lens that passes half the light gives you a picture half as bright. Eye adapted lifts each frame partway back, by the square root of its measured transmission, because your pupil opens and your retina adjusts within moments of putting a lens on and no dark lens ever feels as dark as it measures. The photographs on this page are all eye adapted, since that is the closer stand in for riding. Colour only throws darkness away entirely and leaves the tint, which is the honest way to judge what Low Light does to greens and browns.
08
The whole ladder
Every scene, every figure as a percentage of that scene's own clear control. Prizm Trail Torch and Prizm Road ride along as context, not as candidates.
| Scene | EV100 | Clear | Low Light | Photochromic | Trail Torch | Road |
|---|---|---|---|---|---|---|
| 1 Bright road F | 12.9 | 100 | 86.5 | 21.8 | 41.0 | 15.7 |
| 7 Town street F | 10.6 | 100 | 79.4 | 46.0 | 37.0 | 19.5 |
| 2 Forest, rock, moss F | 9.6 | 100 | 79.2 | 49.0 | 39.3 | 17.7 |
| 3 Forest dirt F | 8.6 | 100 | 83.0 | 53.5 | 39.5 | 17.9 |
| 4 Mixed light trail F | 8.6 | 100 | 81.4 | 39.4 | 40.2 | 19.0 |
| 5 Dark trail F | 8.3 | 100 | 71.8 | 51.4 | 34.1 | 20.4 |
| 8 Forest at dusk A | 4.6 | 100 | 74.5 | 61.1 | 33.2 | 16.4 |
| 9 Night, headlamp A | 5.6 | 100 | 70.1 | 80.2 | 40.8 | 23.6 |
F · scenes 1 to 7 sample transmission over pixels above a fixed 5% of full scale. A · scenes 8 and 9 sample over the clear anchor's own p25 to p99.5 instead, because that fixed floor keeps roughly 1% of a night frame and returns nonsense. Inside a scene every lens shares one mask, so every row above is a fair comparison. Between rows the rule is labelled, and I do not build an argument on a difference of a point or two across the boundary.
- Prizm Low Light is flat and cheap. 70.1 to 86.5 across 8.3 stops of light. It takes a fixed small slice and never asks for more.
- The photochromic is the only lens with an arc. 21.8 at the top of the ladder, into the low fifties through the middle, 80.2 at the bottom. No other lens in the shoot moves more than a few points.
- Scene 9 meters brighter than scene 8 and is far darker to ride. EV alone mis ranks them: the helmet lamp puts a hot pool on the near ground of an otherwise black frame, which drags the meter up. Read the two rows as the last two rungs, not as 5.6 sitting above 4.6.
- Scenes 3 and 4 sit at the identical EV 8.6 and the photochromic reads 53.5 against 39.4. Same amount of visible light, 0.44 stops apart. That is section 9.
09
Where the photochromic stops, and why
Through the middle of the ladder the photochromic settles around 49.0 to 53.5% of Clear and stays there. Scene 2 at EV 9.6, scene 3 at EV 8.6, scene 5 at EV 8.3: three different light levels, three readings inside 4.5 points of each other. In that stretch it is costing about a full stop against Clear and half a stop against Prizm Low Light, in exactly the light it is sold to open up for.
Then it goes the wrong way. At natural dusk, scene 8, EV 4.6, the darkest scene in the set, it reads 61.1. Under the helmet lamp, scene 9, which is far darker to ride in and meters higher only because of the beam, it reads 80.2, its most open reading anywhere. The lens is at its clearest in the blackest scene I shot.
Photochromics track ultraviolet, not brightness. An LED headlamp emits essentially no UV, so with nothing to trigger it the lens very nearly clears, at 80.2% under the lamp. Natural dusk still carries UV, so it stays part activated at 61.1% in less visible light. That is the mechanism behind every rider who says their photochromic feels wrong under winter overcast.
Where my design cannot separate the two explanations, and I would rather say so. Scene 8 was shot at 18:47 and scene 9 at 19:50. A photochromic keeps relaxing for tens of minutes after the light drops, so the rise from 61.1 to 80.2 is consistent with the ultraviolet reading and equally consistent with the lens simply having had another hour to fade. One evening and one 365 nm torch settles it: dark adapt the lens, shoot one locked frame under the lamp, put UV on it for sixty seconds and shoot the identical frame again. Nothing changes but the UV. Until I have run that control this stays a mechanism the numbers are consistent with, not one I have demonstrated.
The supporting evidence does not depend on that shoot. Scenes 3 and 4 sit at the identical EV 8.6 and the photochromic reads 53.5 against 39.4. Scene 4 is dappled, so it carries direct sun patches and more ultraviolet at the same visible light level, and the lens is 0.44 stops darker for it. Both lenses beside it, Clear and Low Light, barely move between those two scenes.
Scene 4, mixed light trail, EV 8.6. Compare the photochromic cell here with the same lens in the scene 3 row of the ladder table: same meter reading, very different lens.
10
Oakley's stated 23 to 69, checked at both ends
Ashley Quinlan's Sutro Lite Photochromic review for BikeRadar prints the specification I had been missing: the lens darkens to as little as 23% transmission and opens up to 69% when it is dark. 5 To put my figures on that scale I scale each relT by Clear's own stated 93%. That conversion is indicative and not a lab VLT, for the reasons in section 13, but it is close enough to ask whether the two ends of the published range are real.
| Scene | EV100 | % of Clear | ≈ abs VLT | Against 23 / 69 |
|---|---|---|---|---|
| 1 Bright road | 12.9 | 21.8 | 20.3 | 2.7 below the stated 23 |
| 7 Town street | 10.6 | 46.0 | 42.8 | 26.2 short of the stated 69 |
| 2 Forest, rock, moss | 9.6 | 49.0 | 45.6 | 23.4 short of the stated 69 |
| 3 Forest dirt | 8.6 | 53.5 | 49.8 | 19.2 short of the stated 69 |
| 4 Mixed light trail | 8.6 | 39.4 | 36.6 | 32.4 short of the stated 69 |
| 5 Dark trail | 8.3 | 51.4 | 47.8 | 21.2 short of the stated 69 |
| 8 Forest at dusk | 4.6 | 61.1 | 56.8 | 12.2 short of the stated 69 |
| 9 Night, headlamp | 5.6 | 80.2 | 74.6 | 5.6 above the stated 69 |
Both endpoints check out, and the middle does not. The dark end lands 2.7 points under the stated 23, and under the headlamp the lens goes past the stated 69 to 74.6. In between, in the light a rider actually meets in the dark months, it never gets there: 56.8 at natural dusk and 47.8 to 49.8 in twilight daylight. Oakley's range is accurate at both ends and out of reach in the middle, and the middle is where you ride.
Quinlan got to the symptom first and said so plainly, which is worth quoting because it is the setup for the measurement. He found that darker twilight produced glaring off car headlights, and guessed the photochromic molecules were sitting partially activated. He also wrote that 69% felt close to the same experience as clear lenses in dark light, and put that down partly to the eye adapting. Both readings are right. He was riding the lens for months and had no number for either, and the number for the first one is 56.8 against a stated 69. The second one is the eye adapted mode in the tool above, with the exponent written down.
His cons list also says the lens cannot get as dark as standard alternatives. My bright road scene agrees: the photochromic at 20.3 against Prizm Road at 14.6 in the same frame.
11
The two lenses I ruled out
Both of these are in the shoot, both get asked about, and both are the wrong answer for the dark. This is the section for anyone about to keep the lens already in the frame through October.
Prizm Trail Torch is not a low light lens, whatever the name suggests. Stated 35% 3, and it is the flattest lens in the entire shoot: 33.2 to 41.0% across all 8.3 stops. It takes about a stop and a third away from you at full daylight and it is still taking the same stop and a third away at dusk. It never opens up, because it was never designed to. I rode it as my after work lens for years and it is a genuinely good mixed conditions lens, which is a different job. If you want it against its sibling, that is the Prizm Trail vs Trail Torch verdict.
Prizm Road in the woods after dark is unusable, and I say that as someone who rates it. Stated 20% 4, measured 15.7 to 23.6% across the ladder. In the dusk forest it is 16.4% of Clear, which is 2.61 stops down, and there is nothing left of the trail surface. It is one of my favourite lenses in bright and ultra bright light and the moment the light drops below bright it is instantly too dark, which is true of the whole Road family. Prizm Road Black, darker again, reads 11.1% in the same frame.
12
Which one to buy
Four rounds, three lenses, and the answer genuinely depends on what your rides look like rather than on which lens is best.
- Buying one lens to cover everything: the photochromic. It is the only one that handles a bright afternoon properly, at 21.8% of Clear on the bright road, and it does not collapse at the bottom. Its weak stretch is the middle, so if most of your riding sits in half light it is the wrong single lens.
- Evening rides, commutes home, trail running at dusk: Prizm Low Light. This is the bulk of what most people mean by riding in the dark, and it is where Low Light won both middle rounds. My own frame carries it more than anything else.
- Genuine full darkness with a lamp: Clear. Round 4 goes to it on raw light and the photochromic is a close second, at 80.2% against Low Light's 70.1%. If night rides on lamps are all you do, buy the cheapest lens in the comparison and spend the difference elsewhere.
- Already own Trail Torch or Road: buy one of these three anyway. Neither of those opens up as the light goes, and October is when that stops being theoretical.
All three take a prescription insert if you need one behind the lens. That is what I make, and it is why I own enough lenses to shoot this in the first place.
13
Method, and what these numbers are not
Each scene is one locked exposure on a tripod, set once on the bare sensor, with every lens held in front of it in turn and nothing else changed. Transmission is read from the frames as a percentage of that scene's own clear control, so each row of the ladder is internally self anchored. Every figure on this page is that ratio. The full apparatus and protocol are on the bench study page.
- This is camera relative transmission, not a lab VLT. It carries no photopic weighting and a sensor does not share the eye's spectral response, so it reads strongly coloured filters warm. The three lenses here are near neutral or barely tinted, which is the best case for the method, and it is still the reason section 10 calls its scaled column indicative.
- Two sample rules, labelled in the ladder. Scenes 1 to 7 use a fixed floor, scenes 8 and 9 an adaptive one, for the reason given under the table. Within a scene every lens shares one mask.
- The night photographs are calibrated to appearance, the percentages are not. The renderer's white point and adaptation exponent were solved per scene against the camera's own JPEGs of the identical locked exposure, because the daylight defaults rendered scenes 8 and 9 several times too bright and showed ground detail that was not there on the night. The measured transmission figures are untouched by that. So the pictures in rounds 3 and 4 are perceptual and the numbers beside them stay photometric.
- Scene 6 is excluded entirely and scene 9's no lens frame is withheld, both because their control frames caught different light than the rest of their set. The fifteen filter readings in scene 9 are clear anchored and unaffected.
- A photochromic cannot be characterised by one frame, and this page does not try. Every reading here is that specimen at that moment, with an exposure history I did not control, on a two hour walk. Nothing here is a transition timing test.
- Specimen note. The photochromic is the clear to black iridium lens in my collection. I hold no manufacturer product page stating a VLT for it, which is why it carries no stated figure of its own in section 2, and why the 23 to 69 range in section 10 is cited to BikeRadar rather than measured off the part in front of my camera.
Per lens detail lives in the individual reviews: Prizm Low Light, Clear, Photochromic, Prizm Trail Torch and Prizm Road. Every lens I have bench tested is in the lens reference with this comparison preloaded, and every scene is in the comparison tool.
14
References
- Oakley, Inc. (n.d.). Clear. Cited for a stated visible light transmission of 93% (Clear). Retrieved July 24, 2026. Oakley does not list a clear Sutro replacement lens; 93% is the typical stated transmission for an untinted polycarbonate lens and is quoted without a product page.
- Oakley, Inc. (n.d.). Sutro Lite replacement lenses: Prizm Low Light [Product page]. Cited for a stated visible light transmission of 75% (Prizm Low Light). Retrieved July 24, 2026, from https://www.oakley.com/en-us/product/W0OO9463R?variant=888392622686.
- Oakley, Inc. (n.d.). Sutro Lite Sweep replacement lens: Prizm Trail Torch [Product page]. Cited for a stated visible light transmission of 35% (Prizm Trail Torch). Retrieved July 24, 2026, from https://www.oakley.com/en-us/product/WAOO9465R?variant=888392552884.
- Oakley, Inc. (n.d.). Sutro replacement lenses: Prizm Road [Product page]. Cited for a stated visible light transmission of 20% (Prizm Road). Retrieved July 24, 2026, from https://www.oakley.com/en-us/product/W0OO9406R?variant=888392404930.
- Quinlan, A. (2023, May 24). Oakley Sutro Lite Photochromic review. BikeRadar. Cited for Oakley's stated photochromic range of 23% to 69% visible light transmission, and for the twilight glaring and eye adaptation observations quoted in section 10. Retrieved September 1, 2026, from bikeradar.com.