WARNING - technical thread - how are the red dots really made?

Started Sep 7, 2012 | Discussions
Roland Karlsson Forum Pro • Posts: 26,327
WARNING - technical thread - how are the red dots really made?

The DP cameras have the red grid syndrome, and so have several other cameras.

Its quite clear, to me, that the red grid is somehow caused by the pixel grid. Otherwise, it would not be a grid.

That it is some kind of reflection in the IR filter for the DP cameras is clear. The reflection there is red, and Sigma has even said so.

That the grid is ONLY created if there is a small very bright light source focussed on the sensor, e.g. the sun, we know.

But ... there are some irritating questions.

  • Why is the grid image so large?

  • Why is the grid image so sharp?

  • Why is the grid image always red, in all cameras?

  • Why do the camera manufacturers accept such a misfeature?

  • Do anyone know the exact mechanism for the red grid?

Why is the grid image so large?

This is very mysterious. The grid image is MUCH larger than the sensor grid itself. How come that a greatly enlarged image of the grid is focussed on the grid? We have just accepted this with a shrug, more or less in our earlier discussions. There was some suggestions that the micro lenses could spread the light rays. In an other forum there was a suggestion that it was diffraction in the grid. But. diffraction should result in rainbow colored stuff - if the light is not monochromatic.

Why is the grid image so sharp?

This is also a mystery. Light is hitting the sensor with a rather large aperture. Then the light that hits the grid should be spread at an angle of the aperture. This should not result in dots, if the IR filter is not very near to the sensor. But - how can the image then be enlarged?

Why is the grid image always red, in all cameras?

Why the grid is red in the DP cameras we know. But - why also in all other cameras where it exists?

Why do the camera manufacturers accept such a misfeature?

Yet a mystery. Why? Is it impossible to fix if you have a large sensor and a compact camera?

  • Do anyone know the exact mechanism for the red grid?*

This is the main question. The red grid is not magic. So - someone designing digital cameras knows the exact answer to the red grid, and can even simulate it in his programs. So - there exist an answer. It would be extremely interesting to know that answer.

 Roland Karlsson's gear list:Roland Karlsson's gear list
Sigma DP3 Merrill Sigma dp2 Quattro Sony RX100 III Pentax K-3 Pentax K-1 +14 more
Gesture Veteran Member • Posts: 5,056
Re: WARNING - technical thread - how are the red dots really made?

Hopefully, some good answers will appear. I've thought lately: some of the earliest digital cameras had weak or (no?) UV filter-i.e. Olympus C-2020) yet the color images seem fine. Is such a strong UV really necessary?

Bodoh Contributing Member • Posts: 552
Re: WARNING - technical thread - how are the red dots really made?

This effect is easy to reproduce with the original DP1. Some people have removed the IR filter. So a question: Does the DP1 still produce the grid after the IR filter removed? Is it changed at all. There should be clues there...
--
http://www.pbase.com/knobby

 Bodoh's gear list:Bodoh's gear list
Sigma DP1 Sony Alpha NEX-7 Sigma 30mm F2.8 EX DN Sigma 19mm F2.8 EX DN
OP Roland Karlsson Forum Pro • Posts: 26,327
Re: WARNING - technical thread - how are the red dots really made?

Bodoh wrote:

This effect is easy to reproduce with the original DP1. Some people have removed the IR filter. So a question: Does the DP1 still produce the grid after the IR filter removed? Is it changed at all. There should be clues there...

I am quite sure that the grid is red due to reflection in the IR filter. So - if any red grid still is there after removal of the IR filter would be a huge surprise to me. Then all the theories and also information we have got from Sigma are wrong.

So, removing the IR filter should be a 100% cure.

That other cameras also have a red grid hints at they also having a similar IR filter like the DP cameras.

 Roland Karlsson's gear list:Roland Karlsson's gear list
Sigma DP3 Merrill Sigma dp2 Quattro Sony RX100 III Pentax K-3 Pentax K-1 +14 more
OP Roland Karlsson Forum Pro • Posts: 26,327
C´mon techno nerds!

C´mon techno nerds!

Some of you are just as interested like me in finding out this!

Unfortunately no removable lens cameras shows this problem. So - we cannot just look, and make experiments, without totally dismantling the camera.

Almost all cameras I have seen have been compact large sensor cameras. Except one camera phone.

So ... whats a reasonable plan? How can we solve this riddle. Any real opticians out there that have the answer. Maybe it is possible to ask Sigma? Maybe they can give a more detailed answer.

And I dont mean at a fuzzy level. We already know that it is the grid pattern on the sensor that creates some kind of grid image that is reflected back by the IR filter. But - how do it do it exactly? Why is the image so large, and so sharp?

 Roland Karlsson's gear list:Roland Karlsson's gear list
Sigma DP3 Merrill Sigma dp2 Quattro Sony RX100 III Pentax K-3 Pentax K-1 +14 more
(unknown member) Contributing Member • Posts: 659
There are indeed some irritating questions

But ... there are some irritating questions.

Why is the grid image so large?

Why does Roland blow the red grid out of all proportions, totallly?

Why is the grid image so sharp?

Why does he fix his sharp mind on the rid grid?

Why is the grid image always red, in all cameras?

Why indeed is it not pink - or mauve?

Why do the camera manufacturers accept such a misfeature?

Why are you still interested in cameras with 'misfeatures'?
Can't you just move on?

Do anyone know the exact mechanism for the red grid?

Yes, I do. But I'm not going to tell you.

SigmaChrome Veteran Member • Posts: 9,398
Re: WARNING - technical thread - how are the red dots really made?

Ok, well why do multicolored grids also appear in high-end, broadcast quality video cameras? I see this effect often in live TV coverage of events. It's very similar to the red grid in the DP cameras but I've never heard anyone complain about it.
--
Regards,

Vitée

Capture all the light and colour!

http://www.pbase.com/vitee/galleries

 SigmaChrome's gear list:SigmaChrome's gear list
Sigma DP1 Sigma DP2 Merrill Sony RX100 Sigma DP3 Merrill Sigma dp0 Quattro +18 more
ArvoJ Senior Member • Posts: 1,351
Re: WARNING - technical thread - how are the red dots really made?

Can you link samples of of such grid for different cameras? I saw these samples somewhere, but not here in this thread

-- hide signature --

Arvo
Sigma/Foveon information collection and little gallery:
http://www.stv.ee/~donq/sigma.htm
http://www.stv.ee/~donq/images.htm

 ArvoJ's gear list:ArvoJ's gear list
Sigma SD14 Sigma 10-20mm F4-5.6 EX DC HSM Sigma 105mm F2.8 EX DG Macro +1 more
Arky Regular Member • Posts: 433
Re: There are indeed some irritating questions

mroy wrote:

But ... there are some irritating questions.

Why is the grid image so large?

Why does Roland blow the red grid out of all proportions, totallly?

Roland is curious as to the underlying cause of this artifact. Intellegent people are often curious as to the mechanism responsible for unexplained phenomena.

Why is the grid image so sharp?

Why does he fix his sharp mind on the rid grid?

Roland is sharp, largely owing to a lifetime of inquisitiveness I'm sure. He sees something unusual and just naturally wants to ascertain the cause. The red grid is an unusual artifact and it is seemingly difficult to eradicate. The mechanics responsible for the grid are either not well understood or are difficult or expensive to correct. Witness Sigma's attempt to ameliorate the gridding by changing the IR filter in the DP2S. This change of the IR filter did help the gridding but also may have exacerbated the green corners problem that is seemingly worse on the DP1S than on the DP1.

Why is the grid image always red, in all cameras?

Why indeed is it not pink - or mauve?

Why do the camera manufacturers accept such a misfeature?

Why are you still interested in cameras with 'misfeatures'?
Can't you just move on?

I'm sure he could move on but he doesn't want to. That's good enough for me, but the fact is, neither of us have a say in it. Roland has every right to ask these questions here even if you do find them irritating. What I find irritating is people that try to stifle technical discussions in their proper venue, which for Sigma cameras is here.

Do anyone know the exact mechanism for the red grid?

Yes, I do. But I'm not going to tell you.

No, I doubt very seriously that you do. You don't impress me as a person much gifted in the art of analytical thinking.

 Arky's gear list:Arky's gear list
Sigma SD1 Merrill Sigma DP1s Canon EOS-1D Mark II Canon EOS 50D Sigma SD14 +14 more
(unknown member) Contributing Member • Posts: 659
You are damn rght

No, I doubt very seriously that you do. You don't impress me as a person much gifted in the art of analytical thinking.

I was joking (or being ironic).

You seem to belong to that other, sad half of humanity, about which Kurt Tucholsky once said (in a rather different context):

"Wenn einer bei uns einen guten politischen Witz macht, dann sitzt halb Deutschland auf dem Sofa und nimmt übel."

But I give you full marks for coming to the rescue of Roland and defending him - well done.

OP Roland Karlsson Forum Pro • Posts: 26,327
Re: WARNING - technical thread - how are the red dots really made?

SigmaChrome wrote:

Ok, well why do multicolored grids also appear in high-end, broadcast quality video cameras? I see this effect often in live TV coverage of events. It's very similar to the red grid in the DP cameras but I've never heard anyone complain about it.

Interesting.

Advanced TV cameras have very complex optics. And they sometimes have three sensor, one for each color. Might that be the cause?

 Roland Karlsson's gear list:Roland Karlsson's gear list
Sigma DP3 Merrill Sigma dp2 Quattro Sony RX100 III Pentax K-3 Pentax K-1 +14 more
OP Roland Karlsson Forum Pro • Posts: 26,327
Re: WARNING - technical thread - how are the red dots really made?

ArvoJ wrote:

Can you link samples of of such grid for different cameras? I saw these samples somewhere, but not here in this thread

Sure, lots of examples in this thread

http://forums.dpreview.com/forums/read.asp?forum=1009&message=42344095

 Roland Karlsson's gear list:Roland Karlsson's gear list
Sigma DP3 Merrill Sigma dp2 Quattro Sony RX100 III Pentax K-3 Pentax K-1 +14 more
BuckieJoe Regular Member • Posts: 226
Re: WARNING - technical thread - how are the red dots really made?

What happens to the images when you remove the built-in IR filter? That's okay for IR photography but what about color?

Tom Schum
Tom Schum Veteran Member • Posts: 6,474
Re: There are indeed some irritating questions

Do anyone know the exact mechanism for the red grid?

Where I work, we have to deal with reflections in optics. I am not working in that area (I do electronics) but the optical techs simply rotate filters slightly so they are not exactly parallel to the optical axis. This causes reflections to bounce away from the optical axis where they are not as big a problem. Second and third reflections would end up being somewhere other than the first reflection, so a red grid (if it is due to the IR filter) would not be able to make a clear image.

Something like this could be tried with the removable IR filters in DSLR cameras such as the SD series, if a small shim is used to offset the axis of the IR filter. This would at least change the appearance of a red grid, and if so this would confirm that the red grid has something to do with the IR filter.

Of course, rotating the axis of the IR filter will cause some refractive shifting in the image cast on the sensor by the lens. This is because the IR filter has some small thickness. It will probably not be a serious problem though, especially if done simply for diagnosis.

I have heard the Sigma IR filter is very fragile so I would not want to try this myself. However I have seen replacement IR filters for sale on the Sigma USA website, so filter breakage is not going to be a show-stopper, just an expensive blunder (about $60).
--
Tom Schum

 Tom Schum's gear list:Tom Schum's gear list
Sigma DP2 Sigma DP3 Merrill Ricoh WG-4 GPS Fujifilm X30 Sigma dp0 Quattro +9 more
OP Roland Karlsson Forum Pro • Posts: 26,327
Re: There are indeed some irritating questions

Tom Schum wrote:

Where I work, we have to deal with reflections in optics. I am not working in that area (I do electronics) but the optical techs simply rotate filters slightly so they are not exactly parallel to the optical axis. This causes reflections to bounce away from the optical axis where they are not as big a problem. Second and third reflections would end up being somewhere other than the first reflection, so a red grid (if it is due to the IR filter) would not be able to make a clear image.

Something like this could be tried with the removable IR filters in DSLR cameras such as the SD series, if a small shim is used to offset the axis of the IR filter. This would at least change the appearance of a red grid, and if so this would confirm that the red grid has something to do with the IR filter.

Of course, rotating the axis of the IR filter will cause some refractive shifting in the image cast on the sensor by the lens. This is because the IR filter has some small thickness. It will probably not be a serious problem though, especially if done simply for diagnosis.

I have heard the Sigma IR filter is very fragile so I would not want to try this myself. However I have seen replacement IR filters for sale on the Sigma USA website, so filter breakage is not going to be a show-stopper, just an expensive blunder (about $60).

The SD cameras dont have the red grid, only DP.

 Roland Karlsson's gear list:Roland Karlsson's gear list
Sigma DP3 Merrill Sigma dp2 Quattro Sony RX100 III Pentax K-3 Pentax K-1 +14 more
Arky Regular Member • Posts: 433
Re: WARNING - technical thread - how are the red dots really made?

Roland Karlsson wrote:

Why is the grid image so large?

This is very mysterious. The grid image is MUCH larger than the sensor grid itself. How come that a greatly enlarged image of the grid is focussed on the grid? We have just accepted this with a shrug, more or less in our earlier discussions. There was some suggestions that the micro lenses could spread the light rays. In an other forum there was a suggestion that it was diffraction in the grid. But. diffraction should result in rainbow colored stuff - if the light is not monochromatic.

Actually, I believe the DP1S sensor can somewhat image itself at 3 different magnifications.

Why is the grid image so sharp?

This is also a mystery. Light is hitting the sensor with a rather large aperture. Then the light that hits the grid should be spread at an angle of the aperture. This should not result in dots, if the IR filter is not very near to the sensor. But - how can the image then be enlarged?

Luck of the draw. The sharpness of the spots and grid are sensitive to small focus adjustments.

As you point out the reflected image is enlarged. I don't see reflections from the flat IR filter being the cause.

Why is the grid image always red, in all cameras?

Why the grid is red in the DP cameras we know. But - why also in all other cameras where it exists?

I think is a result of light passing through the IR filter multiple times.

Why do the camera manufacturers accept such a misfeature?

Yet a mystery. Why? Is it impossible to fix if you have a large sensor and a compact camera?

I think it's largely due to lens design and could be avoided.

  • Do anyone know the exact mechanism for the red grid?*

This is the main question. The red grid is not magic. So - someone designing digital cameras knows the exact answer to the red grid, and can even simulate it in his programs. So - there exist an answer. It would be extremely interesting to know that answer.

The grid and spot size are controlled by the focus movement, at least it is on my DP1S. At near infinity the spacing of the pixels and the magnification ratio align to generate the magnified large red pattern. Moving further away from infinty, the pattern blurs out and then refocuses into a sharper image that is obviously the sensor surface. Still moving the focus closer, the sensor surface image blurs out and then a very fine grid pattern appears.

The lens design must have a spherical element surface with a reflection that just happens to (almost) refocus the image of the sensor surface back on the sensor itself with a high degree of magnification. My testing suggests at least 3 lens surfaces (in the DP1S) can reflect a somewhat focused image of the sensor back upon itself at varying degrees of magnification. At least that's my theory.

Here's a DP1S shot, imaging a home-made 3 color laser collimator with the lens set to a mid focal position.

I posted a set of DP1S test images with 3 lens focus distances and 3 laser colors in this gallery.
http://www.pbase.com/bigflat/dp1s__spots

I'm interested in any conclusions you can draw from examining these images of course.
-Jerry

 Arky's gear list:Arky's gear list
Sigma SD1 Merrill Sigma DP1s Canon EOS-1D Mark II Canon EOS 50D Sigma SD14 +14 more
JLK Veteran Member • Posts: 4,485
The answer is...

The red grid is a set of internal reflections between the microlenses and the IR filter. The geometry in the DP cameras are quite different than those in an SLR.

-- hide signature --

Jim

MOD Kendall Helmstetter Gelner Forum Pro • Posts: 19,881
Re: WARNING - technical thread - how are the red dots really made?

I thought from some time ago that people said it was the microlenses causing this artifact.

Manufactures accept this because it brings up artifacts mostly in edge cases, and even then most people do not care or even think about it much. As long as they can reduce the effect enough that it does not happen often, that is good enough for most camera users.

 Kendall Helmstetter Gelner's gear list:Kendall Helmstetter Gelner's gear list
Sigma 70-200mm F2.8 EX DG OS HSM Sigma 8-16mm F4.5-5.6 DC HSM Sigma 24-70mm F2.8 EX DG HSM Sigma 50-500mm F4.5-6.3 DG OS HSM Sigma 85mm F1.4 EX DG HSM +4 more
OP Roland Karlsson Forum Pro • Posts: 26,327
Re: The answer is... too simplistic

JLK wrote:

The red grid is a set of internal reflections between the microlenses and the IR filter. The geometry in the DP cameras are quite different than those in an SLR.

We know that, but how exactly ... ?

 Roland Karlsson's gear list:Roland Karlsson's gear list
Sigma DP3 Merrill Sigma dp2 Quattro Sony RX100 III Pentax K-3 Pentax K-1 +14 more
OP Roland Karlsson Forum Pro • Posts: 26,327
Re: WARNING - technical thread - how are the red dots really made?

woa! Something, at last

or? ... its a laser .. and this looks like interference too me.

Hmmmm .. thats odd?

I have to go now ... will look at it closer when coming back.

 Roland Karlsson's gear list:Roland Karlsson's gear list
Sigma DP3 Merrill Sigma dp2 Quattro Sony RX100 III Pentax K-3 Pentax K-1 +14 more
Keyboard shortcuts:
FForum MMy threads