Timeline for What causes color artifacts on the strings while shooting guitars?
Current License: CC BY-SA 3.0
22 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Jun 4, 2021 at 13:44 | answer | added | Prof. Falken | timeline score: 0 | |
| Aug 12, 2017 at 0:37 | comment | added | scottbb♦ | ... just under-sampled, and the green spatial resolution being just sufficiently sampled enough to exceed Nyquist. I postulate that if the Bayer CFA weren't there, with the sensor's native pixel resolution exceeding the green spatial frequency by 1.414, there wouldn't be any grey-scale aliasing at all. Very nice catch Mark! | |
| Aug 12, 2017 at 0:33 | comment | added | scottbb♦ | @MarkRansom Not only is your point great, I think it's the correct point. Specifically, the blue and red filter sites each have half the spatial resolution that the sensor's site have. The green, filling in the the diagonals between the reds and blues, essentially have a rectangular grid skewed by 45°; the green spatial resolution is only diminished by a factor of 1.414, rather than the 2x reduction of R and B. The light blue – orange nature of the color aliasing, lacking green effects, and the rather large period of the color "wave", collectively hint at the R&B spatial resolution being ... | |
| Aug 11, 2017 at 23:11 | comment | added | Mark Ransom | @scottbb on top of that, the moire artifacts caused by aliasing interact with the Bayer filter processing. What starts out as simple pixel intensity variations becomes color banding. | |
| Aug 11, 2017 at 23:08 | comment | added | Mark Ransom | @user1118321 a polarizer wouldn't help, unless it's bad enough in quality to blur the image. | |
| Aug 11, 2017 at 18:19 | history | edited | mattdm | CC BY-SA 3.0 |
edited tags; edited title
|
| Aug 11, 2017 at 9:44 | comment | added | scottbb♦ | The only thing that can be done to deal with a signal sampled at less than the Nyquist frequency is to low-pass filter the input signal to satisfy the Nyquist frequency. That's exactly what an antialiasing filter in a camera does. Other than that, you have to perform signal processing to do your best to make the signal look right. But once aliasing has been recorded, the damage is done. The best you can do after that is just triage. | |
| Aug 11, 2017 at 8:42 | answer | added | Michal Trnka | timeline score: 10 | |
| Aug 11, 2017 at 8:25 | comment | added | Michal Trnka | @benrudgers - I'm not familiar with the Nyquist limit concept. Do you know what can be done about it? | |
| Aug 11, 2017 at 8:19 | comment | added | Michal Trnka | @CarlWitthoft and others, we're shooting a collection of 350+ rare vintage guitars, replacing strings or retouching is not really a viable option :( | |
| Aug 11, 2017 at 8:01 | comment | added | scottbb♦ | @CarlWitthoft nylon or gut strings on an electric guitar? First, the largest 3 strings in a nylon set are wire wound, just like steel strings, so that doesn't fix the problem. Second, nylon strings on an electric would look weirder than the orange/blue artifacts. Like, uncanny-valley weird. | |
| Aug 11, 2017 at 5:09 | history | edited | Michael C | CC BY-SA 3.0 |
deleted 22 characters in body
|
| Aug 11, 2017 at 4:52 | comment | added | user1118321 | Would a polarizer help here at all? | |
| Aug 10, 2017 at 11:51 | history | tweeted | twitter.com/StackPhotos/status/895613501636530176 | ||
| Aug 10, 2017 at 11:26 | comment | added | Carl Witthoft | Why not just replace the strings with nylon or gut for the shoot? | |
| Aug 10, 2017 at 8:09 | comment | added | Digital Lightcraft | Agree with the above, it looks a lot like artefacting due to moire caused by there not being an AA filter. The winding of the wire on the string is likely VERY close to a multiple of the pixel density as projected onto the sensor. Also the image looks rather soft, are you shooting wide open? | |
| Aug 9, 2017 at 19:02 | comment | added | user50888 | Zooming in, there are orange and blue artifacts in many locations. There is a large one between the two drawn arrows on the body. There are smaller artifacts on the adjacent string and orange and blue artifacts on the frets under the smaller strings to the left. There is a blue patch on the pickguard screw. I suspect that size of the particular artifacts in question are due to the lack of an anti-alias filter at the camera sensor...speculation but this is common in digital medium format. The windings on the string may be below the Nyquist limit. | |
| Aug 9, 2017 at 17:26 | comment | added | Linwood | If you are shooting raw and have Adobe products (others may have this as well) did you try using the moire correction brush and see if that affects it. There are also fringe correction adjustments there (this seems more likely to be a moire though, as suggested above). Or... and painful though it is.. just clone one spot right down the whole string and overwrite it. | |
| Aug 9, 2017 at 15:23 | comment | added | scottbb♦ | I was just going to suggest what @laurencemadill suggested. Try rotating the camera 45°. Does that change the "frequency" of the color shift pattern? Also, try changing the distance between the camera and the guitar. Does that change the frequency of the pattern, or the color shift of it? | |
| Aug 9, 2017 at 15:18 | comment | added | laurencemadill | It would be useful to know what lighting setup you're using, or if you're using natural light. It's possible that this is just caused by reflected light, but there could be more to it than that. Have you tried adjusting the orientation of the camera slightly to see if it still does the same? | |
| Aug 9, 2017 at 15:06 | review | First posts | |||
| Aug 9, 2017 at 15:10 | |||||
| Aug 9, 2017 at 15:02 | history | asked | Michal Trnka | CC BY-SA 3.0 |