Halos and starbursts: night vision as a measurable variable
Written August 2026 · Informational — candidacy and any post-operative decision are determined by the operating surgeon
Halos and glare at night are near-universal in the first weeks after any refractive procedure, better for most people by one to three months, and largely settled by six. That much is on every clinic page on the internet. What those pages skip is on this one: the four procedures do not carry the same night-vision odds — and the comparison is least flattering to the lens we route our highest myopes toward; the measurement that predicts your personal risk is already taken at your workup, so it is a consult question, not a post-operative surprise; and if you had surgery in China, you will be home long before the settling finishes — so the part of this page you will actually use is the escalation clock at the bottom.
Why this page exists. Night-vision threads follow one script: someone three weeks out from surgery describes rings around headlights, asks whether they are permanent, and receives forty replies containing the words "mine went away" and one containing the word "pupil". This page is for the person who wants the mechanism before the anecdotes — and who will be assessing their own night driving alone, at home, thousands of kilometres from the surgeon who operated.
Three different shapes, three different stories
Patients say "halos" for everything; examiners distinguish, because the shape points at the cause. A halo is a diffuse ring or disc of light around a point source — light arriving through the edge of the corrected zone, or scattered by a healing surface, reaching the retina slightly defocused. A starburst is a set of radial spikes — more typical of residual refractive error or irregularity in the optical surface. Glare is neither ring nor spike but a general dazzle and slow recovery when a bright source hits a dark-adapted eye, and it is the symptom most aggravated by a poor tear film. Early after surgery all three mix together, because a healing cornea is briefly all of these things at once: slightly swollen, slightly irregular, and dry. That is why the early weeks predict very little — and why the trend from week two onward predicts nearly everything.
The four-way comparison — including the one that reflects on our own advice
Comparative studies of night-vision complaints do not crown a winner, but they do support a pattern, and one line of it runs against the routing logic used across this site — so it belongs in front of you rather than in a footnote.
| Procedure | Night-symptom mechanism | What comparative reports tend to show |
|---|---|---|
| Femto-LASIK | Ablation edge and transition zone; flap interface early; induced spherical aberration grows with correction size | The reference point. Modern large-zone, aberration-controlled profiles report far fewer complaints than early-2000s LASIK — the era most horror stories date from |
| SMILE | Same edge physics, no flap; lenticule cut inside intact cornea | Tends to induce fewer higher-order aberrations than LASIK in comparative series; early haze while the interface clears can cause transient glare |
| TransPRK | Surface ablation; the epithelium itself must regrow smoothly | Endpoint similar to LASIK, but the road is longer — night symptoms track epithelial healing and can outlast the LASIK timeline by weeks |
| EVO ICL | Defined optic edge and central port sit inside the eye; a wide dim-light pupil can expose the optic rim | Comparative studies tend to report more halo complaints than SMILE or FS-LASIK — typically a discrete ring that most patients neuroadapt to within months |
Pattern-level summary of published comparative literature, researched August 2026. Individual results depend on prescription size, pupil, optical zone and healing — which is the point of the rest of this page.
Read the last row twice, because elsewhere on this site we route people toward that lens. Our thin-cornea and high-myopia decision guide reaches the ICL on solid grounds — no tissue removed, optics that hold at −12 where an ablation profile struggles. Those grounds stand. But the night-vision trade has never been stated alongside them on this property, and it is real: you are exchanging tissue preservation for a higher reported rate of halo, usually in the form of a ring that fades from attention rather than from the optics. For high corrections the comparison partly inverts — a big ablation induces aberration that grows with the correction, while the ICL's optics are indifferent to it — which is why the honest answer to "which is best for night driving" is a consultation with your numbers on the table, not a table on the internet.
The number your workup already measured: mesopic pupil size
In dim light your pupil opens. How far is individual — measured in near-darkness, adult pupils range from around 4 mm to over 7 mm — and it is measured, in the dark, with an infrared pupillometer, as a standard line of the pre-operative workup whose instrument list our candidacy guide walks through.
Why it matters is pure geometry. A laser correction is centred on your visual axis with a fully corrected optical zone in the middle, surrounded by a transition zone blending out to untreated cornea. In daylight, a 3 mm pupil samples only the well-corrected centre. At night, a 6.5 mm pupil may admit rays through the transition zone — and those rays land defocused, drawing a halo around every point of light. The same geometry applies to the ICL: a very wide pupil can expose the optic's rim. Eyes with larger mesopic pupils have been associated with more post-operative night complaints, and although modern profiles with larger optical zones have blunted the effect — studies genuinely disagree about how much of it survives — the association has never been retired.
What this buys you is not anxiety but two questions that cost nothing at the consult: what is my mesopic pupil diameter, and what optical zone are you planning against it? A surgeon who has just measured both answers in seconds. If your pupil is on the wide end, that conversation may shift the plan — a larger zone, a different profile, occasionally a different procedure — and having it before surgery is the entire reason this section exists. Nobody should learn their own pupil diameter for the first time in a forum thread three weeks after their corneas were reshaped.
The settling clock, compressed
The generic timeline is well covered — including on our own recovery timeline, which keeps the hour-by-hour and month-by-month detail. Compressed to what matters for night vision: weeks one to two, some halo and glare for nearly everyone, aggravated by the dryness that peaks in the same window (our dry-eye guide explains why the two curves are the same curve early on); months one to three, steady improvement as surface and tear film smooth; month six, the conventional settling horizon, after which what remains is usually what stays — with the ICL's ring as the exception, since neuroadaptation can keep dimming a stable artefact well past that mark. Two things do not belong on this clock at all: night driving in the first week or two, which you should simply plan out of your schedule, and sudden new symptoms, which belong to a different page and a faster clock — see below.
You fly home before the clock runs out — so take the clock with you
A fly-in patient's follow-up on night vision is, past the first week, self-administered. The surgeon who knows your optical zone is eight time zones away; the person deciding whether your halos are "normal" on a highway in November is you. That works if you carry two things home.
The escalation list — three patterns that end the waiting
- Disabling glare beyond the first days. Discomfort is expected; being unable to function at night at all, once the immediate post-operative days are over, is not the normal curve and warrants an examination rather than endurance.
- Worsening rather than improving at two to four weeks. The normal course from the second week is monotonic improvement. A reversal is the single most useful red flag this page can give you, because it is visible to you without any instrument.
- Unchanged at six months. Past the settling horizon, "give it more time" stops being a plan. A persistent picture deserves a workup — refraction, tear film, topography, and for ICL patients vault and pressure — to find which rung of the remediation ladder applies.
Escalating means an examination wherever you are — a local ophthalmologist with your operative report and workup numbers beats a WeChat exchange with the operating hospital, which is also the standing advice in our aftercare and enhancement guide. This is one more line in the case for flying home with a complete English record set: for night-vision problems specifically, the reader on the other side needs your mesopic pupil diameter, planned optical zone, and achieved correction from the workup and operative report. Three numbers turn "patient reports halos" into a differential diagnosis.
Not this page: a sudden shower of floaters, flashes of light, or a shadow crossing your vision is not halo physiology on any timeline — it needs a same-day examination wherever you are. The reasoning and the full symptom list live in our guide to the risk that survives the surgery.
Questions people actually ask
How long do halos and starbursts last after refractive surgery?
Some degree of halo, glare or starburst at night is near-universal in the first weeks after any refractive procedure, improves for most people over one to three months, and has largely settled by six months. Dryness makes every night symptom worse, so the early curve tracks tear-film recovery as much as optics. The useful discriminator is direction of travel: improving is the normal course, worsening after the first two to four weeks is not, and a picture unchanged at six months is worth a proper workup rather than more waiting.
Do the halos after EVO ICL surgery go away?
Usually, but by a different mechanism than after laser surgery. The EVO lens has a defined optic edge and a small central port, and some patients notice a distinct ring around point lights at night rather than a soft halo. Comparative studies tend to report more halo complaints after ICL than after SMILE or femtosecond LASIK, yet most patients stop noticing the ring within months — the brain adapts to a stable, unchanging artefact. A ring that remains disabling, or halos accompanied by other symptoms, are a reason to have vault and pressure checked rather than to keep waiting.
Does pupil size really predict night halos after laser eye surgery?
It is an association, not a verdict: eyes whose pupils open wide in dim light — larger mesopic pupils — have been linked with more post-operative glare and halo, because a wide pupil admits light through the edge and transition zone of the treated area. Modern lasers use larger optical zones and aberration-controlled profiles precisely to blunt this, and studies disagree about how much residual effect remains. The practical point is that your mesopic pupil diameter is already measured during the pre-operative workup, so the comparison between your pupil and the planned optical zone is a question you can ask at the consult rather than a surprise afterwards.
Which procedure is best for night vision — LASIK, SMILE, TransPRK or ICL?
No procedure can be promised as best for any individual eye, and candidacy is determined by the operating surgeon after examination. What the comparative literature supports as a general pattern: SMILE tends to induce fewer higher-order aberrations than LASIK; TransPRK behaves like LASIK once the surface has healed but takes longer to get there; and ICL patients report more halos on average, in exchange for optics that do not degrade with the size of the correction. For high prescriptions the comparison inverts, because a large corneal ablation induces more aberration while the ICL's optics are unchanged — one reason the choice belongs to the person who has measured your eye.
When should I worry about halos after refractive surgery in China?
Three patterns break the normal settling course: glare or halos disabling enough to prevent essential night activity beyond the first days; symptoms clearly worsening rather than improving from the second to fourth week onward; and a picture essentially unchanged at six months. Any of the three deserves an examination — locally if you have flown home, with your operative report and workup numbers in hand. Separately, sudden new symptoms such as a burst of floaters, flashes of light, or a shadow in your field of vision are not halo physiology at all and need an urgent same-day assessment wherever you are.
Can night halos be treated if they do not settle on their own?
There is a ladder, and which rung applies depends on the cause found at examination. Treating residual dryness is the commonest fix, because a rough tear film scatters light. A small residual prescription can be corrected with glasses for night driving or, where appropriate, an enhancement. Pupil-modulating drops that slightly reduce pupil size in dim light help some patients. And the ICL has an option the laser procedures lack: the lens can be exchanged or removed. Every rung is the operating surgeon's or an ophthalmologist's call after examination — the list exists so you know that a poor settling course has answers, not so you can prescribe from it.