Filters> Gmic> Layers>Blend median gave me a sixth image where I hoped that only two of the pieces would be visible.
However, an attentive viewer easily sees that the others remain as ghosts. Using ImageMagick with the terminal command:
“convert *.jpg -evaluate-sequence median OUT.jpg” gave the same result.
Yes, clearly, this is due to luminosity variations between all your images.
If you try to ‘normalize’ them a little bit, before applying the median blending, then it works quite well:
$ gmic *.jpeg balance_gamma , blend_median o output.jpg
Many thanks for your help. It was wonderful that it was possible to fix.
The images were taken with a Nikon D7500, ISO 400, f / 9.0 1/4 sec. White balance auto. I saw for myself now that one picture was significantly darker than the others.
Hi, I’m trying same command but with tif files. At the end I get an almost white image, few red & yellow pixels in some cases.
[gmic]-15./ Apply gamma-corrected color balance of images [0,1,2,(…),12,13,14], with reference color (128).
[gmic]-15./ Blend images [0,1,2,(…),12,13,14] using ‘median’ mode.
[gmic]-1./ Output image [0] as tif file ‘test1.tif’, with pixel type ‘auto’, no compression and bigtiff support (1 image 4000x6000x1x3).
[gmic]-1./ End G’MIC interpreter.
If I do with the Gimp plug in it works well.
In fact I want to use averge all as it is working on Gimp Plugin when selecting input layers all.
Any idea what I could be doing wrong? Thanks in advance
When you save the image, be sure you don’t clip it. E.g., if you do gmic med0.jpg med1.jpg balance_gamma , blend_median * 257 cut 0,255 o med2a.jpg, you get
Since I don’t have your original files, I can’t tell you exactly what to do. The above happens when you save something with a lower range than your data, or your image viewer doesn’t support the bit depth, and so the upper range is clipped.
@Reptorian You may have any number of input images.
It would be an average of the two middle elements of a sorted list with an even number of elements. If there are only two values, then, yes, you would have their mean. Median - Wikipedia
Hi, I tried the same with jpg files I posted here and found that the issue is only if I try to save on tif format, if I try png, jpg it works fine. When I try tif it looks like the example you did to break it.
I removed the gamma and other options and it is the same, I guess I’m missing something on the output so tif can work fine.
gmic.exe -input MDSC_0302_IT1.jpg MDSC_0309_IT1.jpg -blend average -output test1.tif
[gmic]-0./ Start G’MIC interpreter.
[gmic]-0./ Input file ‘MDSC_0302_IT1.jpg’ at position 0 (1 image 1365x2048x1x3).
[gmic]-1./ Input file ‘MDSC_0309_IT1.jpg’ at position 1 (1 image 1365x2048x1x3).
[gmic]-2./ Blend all images [0,1] together, using ‘average’ mode and opacity 1.
[gmic]-1./ Output image [0] as tif file ‘test1.tif’, with pixel type ‘auto’, no compression and bigtiff support (1 image 1365x2048x1x3).
[gmic]-1./ End G’MIC interpreter.
Here is the version:
[gmic]-0./ Start G’MIC interpreter.