Some modern LED TVs display the following diffraction pattern (photos taken from this question):

diffration pattern of a LED TV

In my answer to another question about the origin of this pattern I explained that this should be due to two diffraction gratings crossed at a non-straight angle, since the inverse Fourier transform of an idealized reproduction of the pattern (rotated by 90°, see the answer in the link above for the pattern) looks like this:

inverse Fourier transform of the idealized diffraction pattern

But if we look at a close-up, we'll see a simple square grid of pixels, so we can't easily see the crossed gratings:

close-up of the pattern

I've now gone to a electronics shop and found several such screens, so I can now name exact models and provide more close-ups. One of them is LG 49LK5400. Here's what its pattern looks like, with an image, so that the lit pixels are also visible. Note that the photo is a good representation of what the human naked eye sees (up to compression artifacts). Click for full resolution.

LG 49LK5400

Another screen with a very prominent cross (in addition to some other patterns) is that of HP Spectre 13-af006ur. It has much smaller pixels, so I couldn't resolve them.

HP Spectre 13-af006ur

My guess is that this effect might be caused by the arrangement of the conductors powering the pixels. Or it may be the pattern from one of the layers of light spreading system of backlight.

But I may be wrong, so I thought it's better to ask experts: what is the actual part of the LCD screen which results in this type of crossed-gratings-like behavior?

  • 5
    \$\begingroup\$ Obligatory comic strip: xkcd.com/1814 \$\endgroup\$ May 1, 2019 at 9:34
  • 1
    \$\begingroup\$ See this Photography.SE question for more details. \$\endgroup\$ May 1, 2019 at 9:38
  • \$\begingroup\$ @Ruslan Is the simple square grid of pixels visible to the naked eye? \$\endgroup\$ May 1, 2019 at 11:06
  • \$\begingroup\$ @TomCarpenter Moiré is completely unrelated to diffraction patterns. And the family of patterns in question are visible by the naked eye in the same way the camera sees them, and are modelled as such in the paper I linked to. \$\endgroup\$
    – Ruslan
    May 1, 2019 at 11:10
  • \$\begingroup\$ @AndrewMorton please see the update: I've added a photo of what I've now seen, and it does represent it faithfully. \$\endgroup\$
    – Ruslan
    May 1, 2019 at 20:04

2 Answers 2


The hexagonal array pattern of AMLCD fine pitch pixels lends itself to the 30,60 deg aliasing patterns found in your photography in LG 4k/5k monitors. The pixel emissions may be orthogonal, but here we are looking at highly reflective vectors under a non-reflective coated glass.

Technically it is best called a Fraunhofer diffraction pattern.

In 35mm SLR photography, the aperture shape can cause this with edge diffraction. In software, it is called the Starburst or North Star Effect filter. Just as the https://en.wikipedia.org/wiki/OLED

  • \$\begingroup\$ If you look at the photos in the OP—those with resolved pixels—you'll see that they are in a rectangular pattern, not hexagonal. \$\endgroup\$
    – Ruslan
    Jul 24, 2019 at 4:54
  • \$\begingroup\$ the like colour pixel emitters are orthogonal but the reflection of white flash LEDs of light picks up on the recursive vector patterns of reflectance in spite of AR glass coatings. \$\endgroup\$ Jul 25, 2019 at 5:41
  • \$\begingroup\$ Could you elaborate how exactly the structure of the screen is arranged so as to generate this pattern? \$\endgroup\$
    – Ruslan
    Jul 25, 2019 at 6:42
  • \$\begingroup\$ did you want an nano-sized image of reflectance of each element? You can do that without a flash using bright diffused ambient light with a good macro camera \$\endgroup\$ Jul 26, 2019 at 4:24
  • \$\begingroup\$ I wanted a schematic structure of the panel with a clear indication of which layers interact to generate the pattern. In particular, I have no idea what "reflective vectors" you mean: what they are, what they are for etc. \$\endgroup\$
    – Ruslan
    Jul 26, 2019 at 5:06

As has been mentioned in the comments, this is not a "diffraction pattern" (i.e., a pattern produced by a diffraction grating), it is a Moiré pattern resulting from the interaction of the display grid with the digital camera's sensor grid.

In fact, it is three separate patterns, one each for the red, green and blue pixels in each grid. Since the color elements are slightly offset from each other in both grids, the resulting Moiré patterns are offset, too. They blend together to produce the same "rainbow" effect that you get from a diffraction grating, but via a completely different mechanism.

You won't see this pattern when you look at the display directly with your eyes. You also won't see it if you take a picture of the display with a film camera.

  • \$\begingroup\$ Nope, it's not Moiré. See the updated question. One of the photos resolves the pixels so that the Moiré would disappear, but the pattern is still present. Also, exactly the same pattern is visible with the naked eye, and human retina doesn't have too regular a pattern of photoreceptors (at least not to result in stable Moiré with a screen grid); saccades would also result in flicker, which I didn't observe. \$\endgroup\$
    – Ruslan
    May 1, 2019 at 20:00
  • \$\begingroup\$ I think I know why you thought it was three patterns: due to the first photo. But I suppose that one was done with a fluorescent lamp, whose lined spectrum results in these separate diffraction maxima (note how they are elongated all in the same direction: this is the shape of the lamp). The other photos I did now with a phone flash don't have this separation. \$\endgroup\$
    – Ruslan
    May 1, 2019 at 20:09
  • \$\begingroup\$ It is not a photographic Moiré pattern but a pattern of pixel hexagonal angular displacement, -1 \$\endgroup\$ Jul 24, 2019 at 0:25

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.