UV-Vis-IR composites using the Coastal 60mm f/4 APO lens

[nfoto]

I'm currently looking through my archives and found some images that might have general interest. Some years ago, I did a lot of multi-band composites, mainly using channels from UV (from UV monochrome, used as blue channel), Vis (using green channel directly), and IR (IR monochrome, used as red channel).. 

 

Filtration was standardised as Baader U (for UV), Baader UV/IR Cut (visible), and B+W 093 (for IR). I tried several lens candidates and found the coastal Optics 60mm f/4 APO lens to be the best, as it is almost -- but not perfectly -- parfocal for these spectral ranges. On some occasions I used one of the UV-Nikkors and accepted the tedious adjustment of the IR image to match the other bands.

 

First is a view of Nigaard Glacier (Nigaardsbreen in Norwegian), a large arm of the vast Jostedalsbreen glacial sheet, the largest glacier in northern Europe.

 

 

Lots of people come to visit the glacier and being careless as most modern tourists and ignorant of the inherent dangers of Nature, some are killed by tsunamis on the foot lake, or calving ice blocks, nearly every year. As the Climate Change races on, the glacier front now is rapidly retreating and fewer tourists are expected to perish as the trek to the ice rim becomes longer and longer and fewer visitors will try to get close to the ice. At present the glacier has treated nearly 500m since my image was captured, in 2008. Tsunamis in the foot lake probably are no longer possible as the ice sheet is safely on land and no longer in contact with the water.

 

Next out is a winter scenery from a frozen forest lake near the Norwegian-Swedish border. There is a groundwater source that keeps part of the lake ice-free throughout most of the winter and this made for interesting colour contrasts.

 

One can note that there is frazil ice ('sarr' in Norwegian) -- tiny needles of ice -- all over the open patches, but due to the groundwater outlet the open area will not freeze over even though the air temperature that day dipped well beyond -20C.

 

More examples will appear later.

 

Oops, I forgot to mention the camera. It was my full-spectrum modified Nikon D200 for these images.

[lukaszgryglicki]

Wow, I like them :)

 

[nfoto]

An early April scene from the delta of Øyeren, the largest inland delta of northern Europe. This section seen here is, as most of the delta area, very shallow and redistributed muds and silt make the waters quite turbid for most of the spring, before the seasonal flooding takes place in late June. Most of the deciduous and riparian forests are  not yet leafing so the mainly the evergreen conifers prevail on the steep slopes down to the lake itself.

 

 

Capture methods and colour processing as mentioned in the first post. Again, the Nikon D200 delivered the raw files.

 

 

[JMC]

Love those. Thanks for sharing.

[nfoto]

The same multi-band approach can of course be used for static motifs even if the subject is close. However, the margins of error are much narrower and having reliable tripod support definitively becomes a necessity.

 

I commenced with rock lichens. They are *very* static. And as it turned out, very colourful as well in multi-band captures using the same procedure as outlined earlier.

 

 

As the surface of that rock was rather dull and drab to the naked eye, I had no idea there was such a diversified lichen community there.

 

Encouraged by the success of the static lichens, I next tried the little Tussilago farfara, the intrepid harbinger of spring in my neck of the woods. They push up their pre-made little flower heads through snow and dirt whenever their inner sense tells them it is spring time. Just after the emergence the stem is very short and rather stiff thus it stays largely put even when buffeted by wind gusts.

 

 

We already know the flower heads are quite UV-reflective and that their central disk florets are UV-dark. What the multi-band rendition indicates is that IR reflectivity is high as well. Thus the small flower head efficiently melts away any remaining snow after a sudden spring snow fall.

[GaryR]

Thanks for a detailed explanation of this technique.

Love those rich, graphic colors !

[Wayne Harridge]

The lichen on rock image is particularly impressive.  

 

[photoni]

@nfoto They are all fantastic
If I understand correctly you only used one channel per type.
I wonder how the result would be with a monochrome camera.

[nfoto]

Probably the same. Basically these are made from three independently captured, monochromatic (by filtration/processing) images.

[photoni]

I imagine that the B+W 093 = 830 nm filter produces monochromatic files in all the images, while the Baader a little bit of color with Tussilago farfara and rock lichens... I don't believe in landscapes.
I was wondering if for the VIS instead of using only the green channel other channels or a BW file are used, what differences there could be.
@nfoto your research is always stimulating
Thank you
Toni

[Andrea B.]

The good old D200 could produce lovely files if used properly. I still have mine!

 

The Tussilago photo is fascinating as a channel stack. I wonder how that would print out? Would the purple/blue area be difficult to render in print? 

[nfoto]

Thanks, Andrea. I imagine printing those hues in the Tussilago image be a challenge, however given a good printer's ICC profile not impossible. The image never was intended for publishing as such, though.

 

Toni: I used in general the green Vis channel directly because it shows better detail and contrast due to the Bayer layout. Green is anyway a dominant colour of the landscape thus whether the visible image is flattened to monochrome first would not likely impact the outcome much.

[cazza132]

Interesting work with some pleasing results.  The pine forest on the hill reminds me of the output from a full spectrum converted Sigma Fovean sensor. I like the lichen and flower images.

[Unscenerie]

Amazing nature and amazing techniques.

[Andrea B.]

Birna, are you there?  I have a question about how you stack your images for these multispectral photos.

 

multi-band composites, mainly using channels from UV (from UV monochrome, used as blue channel), Vis (using green channel directly), and IR (IR monochrome, used as red channel)

 

I can think of several ways to derive a monochrome. So I'm curious -- when you make the UV monochrome and the IR monochrome, do you simply desaturate the original image? Or do you select the luminance channel? Of course, experimentation produces the best individual answer for others interested in making multispectral stacks. But as I said, I'm curious.

 

Also, do you use raw data for the monochromes? That could produce a different mono than As Shot data.