+ +

Rainbow glare

Rainbow glare (glare in Rainbow) is a rare undesirable optical effect of the LASIK all laser (femtosecond laser). It has caused the diffraction light induced by the pattern of the network created at the back of the flap by the impacts of the femtosecond laser.  It has been described by Krueger and coll, who then used the laser femtosecond Intralase 15 KHz (Intralase Corp., USA) in 2008. We also reported the first observation of the Rainbow glare with laser Wavelight FS200 as well as the first surgical correction of a rainbow glare case.


(a page in French is also dedicated to rainbow glare)

Visual symptoms

Rainbow glare matches the perception of bands or coloured light trailsdistributed according to the order of the visible light spectrum, and located around but away from the sources of white lights (where the use of the term "Rainbow", although the phenomenon responsible for the perception of the Rainbows is different, see below).

The distribution of these bands around the source is usually very symmetrical, and has a geometric character. The orientation of the light streaks is variable, but usually predominates in a horizontal direction of sprawl: Blue is the closest color to the Central light source, and the more distant red. These multicolored streaks can be seen in a more visible way that the observed light source is bright, and located on a dark background. The dimensions of the observed source can also affect the appearance of the rainbow glare.

Characteristically, there is always a free interval between a point source and the first multicolored light strips. If the light source is monochromatic (one colour), the strips are narrower, and identical to the source color.

The following image was obtained thanks to the insertion of a network of diffraction similar to that created by a femtosecond laser in the cornea. The observed source is strong: it is the flash of a smartphone (white light).

Photograph taken through a plexiglass plate femtosecond laser-engraved by a shot for impacts of flap of LASIK spaced 7.5 microns.

Photograph taken through a plexiglass plate femtosecond laser-engraved by a shot for flap impacts of LASIK spaced 7.5 microns. The intensity of the source (flash) is responsible for the importance and the number of colored fringes (several orders of diffraction are noticeable).

This more 'realistic' image is taken through the same device "etched" by the laser impacts, but the observed source is a neon light tube, which the luminance is weaker, and the most important spatial extension. Rainbow glare is reduced, and the colored bands are less defined than in the case of a vivid and more punctual source.

The observed phenomenon is akin to "colored halos", but the character in "regular bands" translated "Diffractive" mechanism of the phenomenon responsible for the separation of colors.

The observed phenomenon is akin to "colored halos", but the character in 'regular bands' translated "Diffractive" mechanism of the phenomenon responsible for the separation of colors.


We have made the first effective treatment for the rainbow glare: the issue of impacts to the excimer laser on the deep surface of the flap helped two patients to heal immediately the effects of this complication. The intensity of this phenomenon generally decreases with time, and disappears in a few months in the vast majority of the case. When it persists, of surgical treatment to deliver a photoablation intended to destroy the pattern of diffraction to the deep surface of the flap seems so effective.

Mechanism of the rainbow glare

The intimate mechanism of the rainbow glare is linked to the diffraction light waves; This is also the only known side effect mainly caused by the diffraction of light within a biological structure (the Diffractive implants used in cataract surgery use the properties of the diffraction by design - creating additional homes - but they are in synthetic material).

The origin of the rainbow glare in summer attributed initially to the pattern created by the impacts of femtosecond laser within the stroma corneal, to allow the cleavage of the corneal flap in LASIK procedure any laser. The femtosecond laser delivers very brief light, focused courses at a depth chosen (around 110 microns) in the corneal tissue, and whose delivery is a mode 'raster' (line by line) with most of the lasers femtosecond currently used. It takes about 1 million spots to create a standard flap interface.

This interface is created by issuing the contiguously laser, "line by line" spots. The Exxx_xxx_5197ment between the spots for each line is customizable, as well as the Exxx_xxx_5197ment between each of the rows (the distance between the spots and lines being generally close to 8 microns).

The Exxx_xxx_5197ment of the spots delivered within the cornea is regular: the spacing between the spots and lines is of the order of 8 microns.

The Exxx_xxx_5197ment of the spots delivered within the cornea is regular: the spacing between the spots and lines is of the order of 8 microns.

Because of the extreme brevity of the impacts laser (on the order of about 10-14 seconds), their peak power is high, and causes an ionization of matter within a sphere around the point of impact: the radius of this sphere depends among other delivered energy. It can reach 5 to 10 microns. The issue of the impacts thus creates a regular pattern within the corneal tissue, is observable in confocal microscopy in vivo, in post operative; It's lines of "hyper-reflective" points, spots located at the interface level, probably located on the flater ACE of the flap, and not on the deeper stroma (which was carved by the excimer laser). We recently published a clinical observation where the Exxx_xxx_5197ment between the spots visualized within the stroma (7.5 cm) corresponded to the direction and layout of the colored bands (the first red band appearing at 16 cm from the white source located at a distance of 1 meter).

The review was conducted 3 months after LASIK in patients reporting the presence of 'Rainbow Stripes' - rainbow glare. The spots cause the appearance of hyperreflectives point areas.

The review by microscope confocal specular has been performed 3 months after LASIK in patients reporting the presence of "Rainbow bands" - rainbow glare. The spots cause the appearance of hyperreflectives specific areas within the stroma, initially at the level of the posterior face of the stromal flap.

Indeed, once the flap raised (this is the first time of the realization of the LASIK), the underlying stroma is carved by the excimer laser to correct the optical defect second time: photoablation for myopia or hyperopia, and/or astigmatism), which causes the disappearance of the print by the femtosecond laser. However, it seems this still at the back of the flap, which is not exposed to the excimer laser.

Diffraction by the femtosecond laser impact areas

Each impact is a change in the local physical properties of the corneal tissue, although these changes remain to be explored. It is possible that the refractive index is different from that of the adjacent corneal stroma. The regular pattern, different alternating areas of refractive index is able to produce a phenomenon of constructive and destructive interference by diffraction of the light beyond these areas. These have indeed the close dimensions of the micron, of the same order as the wavelength of visible light. The mathematical expression for calculating the dispersion of each colored radiation angle is relatively simple. On the other hand, the calculation of the distribution of energy carried away in each order of diffraction is more complex.

The following diagram represents a constructive interference between two waves (monochromatic light) diffracted by an impact area (these areas are contiguous: it is of order 1 of diffraction (the optical path differs from a wavelength between two parallel paths).) The zero-order of diffraction (light "continues its path without deviation" is not shown for clarity). The wavelength, the more the angle that is constructive interference is important.

Schematic representation of the diffraction of two areas contiguous (laser impacts). The first direction (here the "top") that occurs a constructive interference is a direction that will be observed a "peak of intensity" light for the considered wavelength. This angle is proportional to the wavelength diffracted.

Schematic representation of the diffraction of two contiguous areas (impact laser, diagrammed in gray rectangle). Under the Huygens principle  each zone can be equated to a first approximation to a point source, which emits (diffracted) light in all directions. The first direction (here 'up') according to which occurs a constructive interference is a direction that will be observed a "peak of intensity" light for the considered wavelength. This angle is proportional to the wavelength diffracted.

The diffraction pattern is represented here for a wavelength shorter than that of yellow:

diffraction light yellow vs blue rainbow light glare

The short wavelength are diffractées with a lower angle; Blue is therefore less deflected the yellow/green, he deflected even less than the red.

Shorter wavelengths are diffractées with an angle less than the longer wavelengths, which explains the appearance of multicolored bands (Rainbow) when the source is polychromatic (ex: white light).

Rainbow glare polychromatic light

The light from a white source is polychromatic: the network of diffraction "separates" light trains based on the wavelength (diffraction of order 1 represented here). A proportion of the not incident white light is not deflected (it is not represented here; it is the 0 from diffraction order).

White light is partly divided into different levels of diffractions by regular impact areas, which act as a network of diffraction. The direction of these orders depends on the considered wavelength. This light is then refracted by the posterior face of the cornea, then the lens to be focused on the retina.

diffraction by the cornea and refractive, rainbow glare

Schematic representation of the chromatic dispersion of the 'rainbow glare. Areas of impacts (several hundred by lines for a standard flap) are here reduced to 4, for clarity. The blue is diffracted at an angle less than the red.

The directions where the "bands of colour" are observed are related to the orientation of the lines of spots. In general, the consistency being largest in horizontal (this is related to the constant spacing of the spots 'on-line', while there is a vertical offset due to the juxtaposition of lines, including the Starter wife the circular edge of the flap), the bands adopt a rather horizontal direction.



Schematic representation in "3D" of the rainbow glare, with levels of diffractions.

Schematic representation in "3D" of the rainbow glare, with levels of diffractions. The order zero is the where the light is is not deviated, regardless of the wavelength considered; light is therefore perceived as white.

Rainbows observed in nature are produced by a phenomenon of refraction and not of diffraction (incident, emitted sunlight behind the observer is refracted several times inside the droplets of rain in suspension and returned to the observer; each refraction "separates" the different wavelengths). The 'rainbow glare' should rather have been baptized 'diffraction glare. " the phenomenon of dispersion of the colors of the rainbow glare is close to that of the surface of CD or DVD discs: the micro-gravures are at the origin of the diffraction of incident light after reflection at the surface of the disc.

Refractive surgery techniques that do not require the use of the laser femtosecond (ReLex, Smile) are initially exposed has the occurrence of rainbow glare. The optical interface Associates here l coalescence of two surfaces created by the issuance of femtosecond impacts. The concentric circular distribution of impacts expected to lead radiaires skies arches, giving an aspect of kaleidoscope has these colorful lights.  of A date, experience this effect after Relex or Smile n has not yet been reported.


Rainbow glare is a rare phenomenon that can be observed after femto-LASIK. It fades with time and not occaseionne important visual discomfort only in a very small number of case. If the induced gene persists in time and intensity, a photoablation excimer to the deep surface of the flap is possible and is effective (see:) Rainbow Glare Gatinel JRS 2015 correction)




(1) Krueger RR, Thornton he, Xu M, Bor Z, van den Berg TJ. Rainbowglare have year optical side effect of IntraLASIK. Ophthalmology, 2008; 115:1187 - 1195.

(2) Bamba S, Rocha KM, Ramos-Esteban JC, Krueger RR. Impact of rainbow glare after laser in-situ keratomileusis flap creation with a 60 kHz femtosecond laser. J Cataract Refract Surg. 2009; 35:1082 - 1086.

(3) Sarayba MA, Ignacio T, PS, Tran DB Binder. Comparative study of stromal bed quality using mechanical microkeratomes has IntraLase femtosecond laser 15 - and 30 - kHz microkeratomes. Cornea. 2007; 26:446 - 451.

(4) Binder PS, Sarayba MA, Ignacio T, Juhasz T, Kurtz RM. Characterization of submicrojoule femtosecond laser corneal tissue dissection. J Cataract Refract Surg. 2008; 34:146 - 152.

(5) Binder PS, Brownell M Martiz J. Rainbow glare mechanism not confirmed by SEM. Presented as a poster at the American Academy of Ophthalmology

6) Gatinel D, Saad A, Guilbert E, Angolan H. Unilateral rainbow glare after uncomplicated femto-LASIK using the FS-200 femtosecond laser. J Refract Surg, 2013, epub ahead of print

11 Responses to "Rainbow glare"

  1. Ale says:

    Dr. Ganeshan

    I underwent to Relex smile at the beginning of September.
    Since then I experience glares and halos around light sources, day and night.
    In particular one distinctive symptom makes me think about diffraction.
    Distant car approaching in daylight, Headlight made up of a strip of bright LEDs.
    I See 3 different parallel stripes, two ghosts at the side of the real one. When the car is near the 3 stripes collapse in a blurred one. Same for both eyes.
    I don't see any rainbow. Could This effect be correlated whith the spiral pattern of femto laser on both flap and

    I don't have HOA
    I'm really worried because my symptoms are disabling and no one gave me a diagnosis yet

    Thank You

  2. Dr. Damien Gatinel says:

    If you do not specifically see colored fringes, (blue inside, and red outside), then it is not a rainbow glare related symptom. This does not mean that there is no light diffraction at the level of the Smile interface. Sometimes, the interface can be scattering light significantly. This is partly due to the extracting maneuvers of the duckweed: it can delaminate (distort the arrangement of the collaged fibers of) The stroma (this occurs when the dissection of the duckweed is a bit more difficult), and this will eventually cause a slight Reduction of the corneal transparency, possibly associated with some irregularity. These can explain the perception of ghost images. Their disappearance when the object gets closer is due to an angular/size related phenomenum. The light deviance leading to the ghosting is still there, but its dimension becomes relatively negligible with regards to the angular dimension of the closer light source. Such symptoms usually resolve over time. However, in case of a complication (ex: slight decentration of the duckweed cut), then it may persist to a certain degree over time. Your surgeon would probably provide you with the relevant explanments.

  3. Ale says:

    Dr. Gatinel,

    Thank you very much for your prompt reply, it is very helpful and I'm grateful of the time you take answering the questions of a worried patient.
    It's now 3 months from surgery and the scattering/halo effect doesn't appear to subsidy.
    My surgeon can't explain the situation, I will undergo to a confocal microscopy examination of corneas (not scheduled yet).
    I have something to focus on?
    The centering of the duckweed appears good in both eyes.
    In case of persistence of symptoms in time, does an approach like yours possibly be suitable to my case: Circle procedure (flap creation) plus smoothing with excimers laser both bed and flap?

    Thanks Again

  4. Dr. Damien Gatinel says:

    I am not sure that such approach (flap creation on the smile cap and smoothing with the excimer) would be of help here. It may create some additional irregularity. If The Smile duckweed dissection has been difficult, it may be that the corneal stroma tissue is slightly irregular, but such irregularity may not be accessible to excimer smoothing. The reshaping of the collagen fibers may help over time. The first step is to confirm the origin of your disturbances. confocal microscopy may give some indication about the cellular activity and inflammation at the interface level.

  5. Love says:

    Dr. Ganeshan, Hello,
    I would like to write this message to you because I would like to have an opinion about the rainbow glare.
    On October 18, 2017 I had two eyes operated at the same time for a myopia by LASIK femtosecond brand Alcon wavelight FS 200. The right eye was myopic a-5.25 and the left eye was myopic a-4.75. The right pupil measures 6.4 mm and the left pupil 5.4 mm. The optical area to be treated is 6.5 mm for both eyes.
    Immediately after the surgery that went very well, I see when I look at a light such as a lamppost or headlights of a car of the bands of type Rainbow in a more marked way with the left eye than the right eye. Plus, the brighter the light is white and the more colorful the fringes are important.
    To this day has almost 7 week of surgery My visual acuity has been measured, I have 16/10th has the right eye and 14/10 EME has the left eye but these rainbow bands are always present and it mostly affects me with the left eye at night when the light 's here.
    Can you tell me if the symptoms of rainbow glare will fade or not and if it fades it should wait how long in term of month?
    Thanks for your doctor lighting
    Good Day

  6. Dr. Damien Gatinel says:

    The rainbow glare (which you make a very accurate description of) is a phenomenon that decreases over time. It is necessary to wait about six months to decide on the longevity of this symptoms.

  7. Love says:

    Thank you for your insight into this post LASIK Rainbow phenomenon.
    After about six months I will contact you to make the point of my situation.
    I wanted to congratulate you for the detailed descriptions of your website.

  8. Ale says:

    Dr. Gatinel,

    Thank you for your time. I really feel you dedicate yourself to patients. I hope confocal microscopy will give me some answer

  9. Edouard says:

    Hello, Dr. Ganeshan,

    I was operated in your Centre – Rothschild Foundation – 8 weeks ago from myopia (-3OG and-3OD) to LASIK (laser femtosecond + laser excimer). Since I see in every eye on each yellow or white light source (streetlights, ceiling lights, car headlights, curtain slits) Big Rainbow features. This is very very inconvenient. It's an undesirable effect that makes me regret my operation bitterly.

    After how long can these symptoms disappear?
    Can we speed up their disappearance?

    When should I consider these symptoms as definitive?
    is an operation possible in this case to make them disappear? With what chance of success? Is this a process that you practice regularly?

    By thanking you for your lighting on these four points,

    Kind regards

  10. Dr. Damien Gatinel says:

    The symptoms of rainbow glare disappear (or fade strongly) within a few months in the vast majority of case. If this is not the case, a retouching can be done, as described in our articles. You have to wait at least six months to judge the interest of remodeling.

Leave a comment

You can ask questions or comment on this content: for this, use the "comments" form located below. The questions and comments of a general interest will be processed and published, and the information provided on the relevant pages should be clarified or supplemented.

Your e-mail address will not be published. Required fields are indicated with *