My first UV photograph

[Cadmium]

Hi Jim, Yes you are making very good progress, those look like they should, no different than the ones I did.

Le last to images I posted of your above represent the two worst looking that I processed, both look a little overexposed or hazy,

and the last one seems to have a hot spot in the center which also shows in your processed version.

I am not keeping track of what lenses you are using, but the first to shots I processed look more natural to me.

This might just be exposure related, but the one lens may have a hot spot, but I could be wrong, hard to say unless you use it more.

 

They have some magnetic filter attachment system available. I have never used it, but some have, so ask someone about that who has tried it. Actually Boon uses that method I think.

Ask Boon at:

http://myphotojourney.co.uk/neewer-ad200-flash-uv-imaging/

[Andy Perrin]

These are looking quite good now!

 

I suppose that the better lenses (i.e. more UV transparent to smaller wavelengths) should give images with a greater range of colours?

Yes, up to a point. Once you get a lens that passes out to 320nm, the camera's sensitivity falls so much beyond that point that you don't see much additional benefit because the longer wavelength stuff tends to overwhelm the shorter things, if you are using a filter that passes all of UVA. You get some oranges (in the white balanced image) out below 340nm I think, but not much else?

[Jim Lloyd]

Thanks Steve and Andy

 

Can go a little crazy after a while playing with adjustments.

 

In this one I have used the Lightroom dehaze function and increased saturation. Of course just because you can doesn't mean you should. Aiming for something that is a touch surreal without being too weird - what ever that looks like!

[Andrea B.]

Jim, everything is looking pretty good! You've made progress with the conversions. Just a matter now of deciding on your personal choices for saturation, sharpening, hues & tints, etc.

 

 

As a baseline for what follows, here is the photo you made with the D3200 and the 28mm lens.

Nikon D3200-mod + Photax Paragon 28/?? + Hoya UG1 + Schott BG40 + Sunlight

 

I looked this file in Raw Digger and extracted the actual raw photo you made. This version has been demosaiced and the typical "gamma" curve has been applied, but no white balance has yet been applied. Nor have any adjustments yet been made to the white/black points.

 

 

Next is the raw histogram of the file which shows underexposure in the shadows -- no surprise in the typical UV landscape photo. It appears that you have 12-bit raw color set because the histogram looks like it is going to hit the wall around 4096 on the right side (in the red channel mostly). I do not know if the D3200 has an option for 14-bit raw color? If so, I would suggest setting that. The sky was blown out ("blinkies"), again not surprising in a UV photo. But during file conversion (shown below), I found that the D3200 has a quite nice tonal range with good highlight headroom, so I could easily recover the overly bright sky.

 

 

The raw histogram and the raw composite photo show us that data was recorded primarily in the Red and Blue channels with not much in the Green channel. Indeed Raw Digger marked the Green channel as 24% underexposed. Now that Green underexposure is not your fault or the camera's fault. The histogram is simply telling us that you are most likely recording UV light in the region approximately between 360/370-400 nm because of the lens and filter you are using. And maybe you have just a bit of Visible violet leaking through. Tiny amounts of violet really do no harm to a UV photo. Here are separate Red, Green and Blue channel extracts.

 

 

Now I have 3 global conversions of the raw photo. With a bit of study, one can obtain the same conversions from any converter. However some converters make this easier than do others. These three conversions could be finished up in Photoshop, if desired, for application of edits only in certain local areas of the photo.

 

First is a conversion from Nikon's new Capture NXD which is quite universally disliked! Nevertheless, a Nikon converter will give you a very good TIF for further work in Photoshop or any other finishing app. This is because Nikon software knows Nikon cameras, permits access to the Picture Controls, can deal best with noise, and will correct lens aberrations for Nikon lenses (newer ones, that is).

 

 

The second conversion is from the old Nikon View NX2 which has a reduced tool set but is surprisingly good for a basic conversion. I pushed the saturation up in this conversion so that you could see where some of the false colour lies in your scene. In a UV landscape you will often see some green tones in addition to the typical blue and yellow (and gray/black/white).

 

 

The third conversion is from Photo Ninja which has such an excellent highlight recovery tool, shadow lifting tool and Detail tool. I probably applied too much of that Detail slider, but look at the texture of the snow on the right in this version as compared to the other two conversions. Love that texture!

 

 

In all 3 of my conversions the white balance was set so that no pale purple or magenta tones were left in the photo. This is what would be accomplished if one were to set WB in a UV photo using a UV-white UV-reflective standard. With experience this WB can be closely approximated without use of a white standard. But a reminder: there are other ways to deal with colour in UV scenes. You might want to try channel swaps or other colour conversions. But learning first to deal with UV white balance teaches you a lot about UV files and their artistic possibilities.

 

Finally here is a black & white rendering of your scene from the Photo Ninja conversion. I usually work on the tonalities in a B&W version of a UV shot before working on color. So I thought you might like to see that also.

 

 

Added: Here is that B&W version with a bit of local editing. I lifted the shadows a bit in the front to show some of the tones in the gravel, the rock and the grasses. This D3200 is a nice little camera! The files seem nicely plastic and amenable to the strong edits needed for UV.

[Jim Lloyd]

Hi Andrea - that is really so helpful! thank you!

 

A lot to digest. The D3200 is 12 bit only for RAW.

 

Usually with normal visible digital photography I tend to err on the side of underexposing (often setting -0.3 or -0.7 exposure below the meter value) to avoid blowing highlights, as lost shadows can usually be recovered. For these I have been guessing an exposure and then adjusting based on the camera histogram so that I am just about to blow highlights. I am wondering from your analysis maybe whether it might be worth exposing rather more to get the information in the shadows. Of course its easy to make bracket exposure and explore later.

 

Regarding the lens for this example it is a photoax paragon Nikon mount pre AI 28mm f/2.8 - only got it very recently from ebay and haven't used it that much. My initial impression was a bit disappointing in terms of sharpness, but it feels nice and is in good condition. I thought maybe I paid over the odds at £50. Anyway from what you say it looks like its acceptable as far as UV transmission is concerned?

 

I was thinking that the 35 mm lens I have might be better for UV as the images I showed above (and available via the link above) had a slightly warmer red tone. Do you think that's right? Its an Optomax 35 mm f/3.5

 

I think for landscapes on DX 35 mm or 28 mm are probably what will suit what I want.

 

Just brought another 35 mm on ebay Soligor f/3.5 - from what I have read will probably be similar to the optomax, but for £3.35 I thought it will be worth a try!

[ulf]

if you got the right version of Soligor f/3.5 the transmission will be quite good.

You might already have found this on the forum:

http://www.ultraviol...h__1#entry17222

Especially the link in post #3 is useful.

[Andrea B.]

Jim: I am wondering from your analysis maybe whether it might be worth exposing rather more to get the information in the shadows. Of course its easy to make bracket exposure and explore later.

 

Because reflected UV photography tends to be dark, try to make sure your in-camera luminosity histogram is in the middle, at least. Not a good idea to underexpose UV. With practice you can determine how much you can push the exposures to avoid excess noise. A good shadows and highlights slider in the converter helps. (Some are better than others.) Also, the D3200 has an in-camera setting called "ADL". Set ADL to "normal" when shooting in UV and it will help with the shadows. Granted, that is a setting that some converters ignore when reading the file. However, it will enable you to get a better exposure while you are shooting. If your JPG shadows look good while shooting, then you know you can obtain that result in a non-Nikon converter.

 

Also Try This: Select the Monochrome Picture Control on the D3200. Set the sharpening to 4. Leave the other settings at 0. This is a much easier way to shoot a converted Nikon for UV because it permits an analysis of the tones being captured. And it permits an easier view of the Luminosity histogram (the white one) which you want to try to prevent from piling up on the left-hand wall. Last but not least, the Mono setting does not make your eyes dance til they hurt from all that startling magenta and red. B)

Added Later: An edit to Luminosity histo to read "piling up.....".

 

Jim: Anyway from what you say it looks like its acceptable as far as UV transmission is concerned?

 

Anytime you are shooting under 400 nm, then you are recording a legitimate reflected UV photograph. The UV band in which we can make such photographs is only 100 nm wide (as compared to the 300 nm wide Visible band), so as long as you are in it, you are doing OK. Get your feet under you then think later about lenses which might have further "reach". (You are doing great so far!)

 

 

Jim: I was thinking that the 35 mm lens I have might be better for UV as the images I showed above (and available via the link above) had a slightly warmer red tone.

 

Probably so, but it is difficult to analyze reach from a JPG made under Auto white balance - which gets in the way. I'll look one of those images made with the 35mm in Raw Digger and see what it tells us.

[Andrea B.]

I need to add this. In a UV landscape the sky is usually going to blow out while shooting if the exposure is set so that excess noise is kept out of shadows and shadow detail is retained. You want to test just how much "blowout" the D3200 can handle in the sky. Can the sky blowout be successfully pulled back in the converter? Trial and error will tell this tale.

[Andrea B.]

(see 2 preceding to catch up....)

 

Your lenses seem to all have a similar UV-signature -- at least when used on this particular scene under that particular light-of-the-day with the given filtration. :D

 

Here are the raw composites and raw histograms for the 35 mm and 50 mm lenses. We also need to investigate that UG1 + BG40 filtration. I'll get to that next.

 

The following raw histograms show that you have potentially 1-2 more stops worth of exposure available.

 

Nikon D3200-mod + Optomax 35/3.5 + Hoya UG1 + Schott BG40 + Sunlight

 

Nikon D3200-mod + Nikon-E 50/?? + Hoya UG1 + Schott BG40 + Sunlight

[Andrea B.]

Link to UG1 + IR-blocker transmission charts: www.ultravioletphotography.com/content/index.php/topic/1313-filter-transmission-charts/page__view__findpost__p__17211

 

Likely a bit of violet leak which typically records in the blue channel or in both the red & blue channels depending on camera. No surprise and not worrisome. Peak looks to be at 365nm.

[Cadmium]

Jim is using this stack specifically for all the UV photos that he shows so far:

[Hornblende]

How does the IR leakage compare to the Baader-U?

[Alaun]

"Last but not least, the Mono setting does not make your eyes dance til they hurt from all that startling magenta and red. B)"

 

Nice idea with the B&W setting, never tried that.

Alternatively you can try a White Balance for UV (or just a WB setting for IR and then using that for UV is also not that bad)

[Andrea B.]

How does the IR leakage compare to the Baader-U?

 

Depends on the thicknesses of the component filters in the stack. Jim originally listed "2mm" but I'm not sure if that is the total stack thickness or the thickness of each individual filter. We think the BaaderU is approximately OD+4 in the IR region.

 

 

 

Alternatively you can try a White Balance for UV

 

Yes, indeed. :D

UniWB is also helpful.

[Jim Lloyd]

Thanks again Andrea et al - really useful information !

 

I am using 2mm UG1 + 2mm BG40 - i.e 4 mm total.

[Cadmium]

Andrea, I show his stack and thicknesses above.

If I am reading Shane's graph correctly, then the modern Baader U has about a .025% leak around 900nm mostly. See Shane's top, right hand graph, red line

http://www.beyondvis...BV3-filter.html

 

I have already said that I could recommend a better stack than the UG1 2mm + BG40 2mm stack (such as U-360 2mm + S8612 1.5mm - 2mm, OD4.75 - OD6),

but the stack Jim is using does look like it is working well for him.

 

http://www.ultraviol...dpost__p__18626

UG1 2mm + BG40 2mm is approximately equivalent to UG1 2mm + S8612 1mm.

I don't have a pushed IR leak test for the UG1 2mm + BG40 2mm which Jim is using, but I have such an example using a UG11 2mm + S8612 1mm stack which has about the same Red/IR suppression.

The right hand column of the example below is pushed, however it compares the amount of IR leak from the Baader U and the OD3+ stack.

Given that the leaks are 700nm and 900nm, this changes the color of the leak, so a 700nm leak would typically appear brown (even after white balance) in dark areas, and a 900nm leak will appear foggy gray, not changing the color.

I will also post the graph of the stack tested in the photo below, you can compare that stack to the stack I posted above that Jim is using.

Jim: This is all mostly just superfluous hoopla, so just keep working with your white balance for now.

 

This test shows the intensity and color of the IR leakage between the Baader U and the UG11 2mm + S8612 1mm stack.

It also shows the absence of leakage for the non-pushed normal exposure time.

These shots are not white balanced or processed, and share the same arbitrary in camera white balance.

The reason the UG11 stack is shot with twice the exposure time is because the unobstructed real exposure is twice that of the Baader exposure. Just trying to keep everything equal.

 

Here is the graph for the UG11 2mm + S8612 1mm stack.

[Andrea B.]

Thx, Steve !! :)

 

Note the important point made in Cadmium's preceding post that if you are shooting landscape at "normal" UV exposure lengths (1" - 4" or so), then minor filter leakage of IR or of Vis Violet is negligible. Indeed for artistic reasons a bit of violet about 5nm or so around 400nm is often wanted to add a bit more to the false colour.

[Cadmium]

NOTE: I would not really recommend the above stack either (UG11 2mm + S8612 1mm), it also has only borderline suppression.

However, it is a good example to compare to when asking how much IR the Baader U leaks, because I can clearly see from the comparison tests that the Baader leaks just as much IR,

albeit in a higher range that looks more monochrome to the Bayer filter array. Yet, there is no IR leak seen from either the Baader or the stack at normal exposure time.

That being said, I like to keep stacks suppressed to OD4 or below, but the OD3.5 works.

 

As far as the blue edge of UV, one reason I recommend Hoya U-360 in place of Schott UG1 is because even though the Hoya U-360 graph looks like it has as much or more blue as the Schott UG1,

my experience is that the opposite is true. I find that U-360 stacks transmit less blue than their graph might suggest, and less than the same thickness of UG1.

So U-360 is my preference, but the difference may be negligible.

If you really want to delineated UV from blue, then use UG11, or U-340 which cut slightly deeper below 400nm.

My most recommended stack is still the U-360 2mm + S8612 1.5mm/1.75mm/2mm.

Jim is using the 'classic UV stack', BG40 has 1/2 the suppression strength of S8612, and can be used thicker, such as 2.5mm or 3mm which would equal 1.25mm and 1.5mm of S8612.

BG40 tends to have a bit more stable surface than S8612, needing less attention and maintenance.

If one wants more blue, simply use 1mm thick U-glass instead. For example, U-360 1mm + S8612 2mm. Such a stack will also have less exposure time.

[Cadmium]

As I mentioned above, Shane's test graph of the Baader U shows about 0.025% to 0.04% transmission at the 900nm range,

which is OD 3.6 and OD 3.4 respectively (it is a little hard to pinpoint the exact amount from the graph), but somewhere around OD 3.5.

http://www.beyondvis...BV3-filter.html

Even assuming the Baader U has an OD 3.6 suppression at the 900nm range (0.025% transmission = OD 3.6),

this is about the same suppression as the UG1 2mm + BG40 2mm stack at the 700nm range used by Jim (see diabatic graph).

http://www.ultraviol...dpost__p__19027

 

https://www.pgo-onli...-converter.html

 

Shane's graph is older, which looks like it may be dated 2007, newer Baader U versions may have varied and different transmission.

Nevertheless, the IR leak test I show above was done with a 2012 Baader U.

[Jim Lloyd]

Very interesting posts thank you.

 

I hadn't really thought about it before, but maybe for artistic reasons as Andrea suggests it might be better to have a bit more violet leakage. So maybe the UG1 1 mm rather than 2 mm?

 

Anyway I actually think now the most important thing is to get out there are take some photos - however work and weather have been preventing this!

 

I was going to download rawdigger to try but McAfee thinks its a virus and advises against it - so I didn't. Anyone else found this to be an issue?

[Cadmium]

Jim, As I said, personally I think the best UV stack is the U-360 2mm + S8612 2mm. And the best alternate to that stack for more blue is U-360 1mm + S8612 2mm.

However, UG1 2mm + BG40 2mm is the old standard, and works fine. It will not really show much blue, the thing to keep an eye out for with that stack is the 700nm Red/IR OD 3.5,

which is borderline suppression.

It sure looks like it is working fine for your shots so far. Find some yellow flower, see how that goes sometime.

I am glad you got white balance working! :-)

[Cadmium]

One other point, BG40 is not as prone to surface deterioration.

Unlike window glass, a few (not all) of these filter glass types are prone to surface problems over time.

S8612 is one, and UG11 (U-340) are others... But BG40 is much more stable.

UG1 is more stable also, so you have that going for you.

[ulf]

These pictures of blackberries show the difference between three different filters / filter stack:

 

UG1, 1mm + S8612, 2mm

 

Baader U

 

UG1, 1mm only, (used as a makeshift IR-pass-filter, lacking better alternatives at the moment).

 

Camera: Full spectrum modified Canon 60D (Astrononmik.de glass)

Lens: Old type El-Nikkor 80/5.6

[Jim Lloyd]

Ulf, That's a really nice visual demonstration.

Thanks Steve

I guess what I am using should be more like image 1 or 2 that image 3.

[Jim Lloyd]

Not sure whether this should be a separate thread, but I think its along the same lines so will include here.

 

Lots of questions ...

 

I got a new lens today £3.35 from ebay - well used Soligor 35 mm f/3.5 T mount No 9711108

 

All below images are at 100 ASA f/8 1/25 sec for vis/ir shots and f/11 ASA 200 1-8 secs for UV shots

Camera full spectrum converted Nikon D3200

Filters are described above UG1 2mm + BG40 2mm

 

All processing in Lightroom using custom camera calibration file to start and then spot white balance plus manual adjustment

 

First I meant to just check lens in visible light, but forgot to put filter on - quite liked the accident:

 

 

 

QUESTION 1: I assume this image is dominated by near IR with significant visible if I have a full spectrum conversion - These wavelengths will dominate UV completely ?

 

Next this is the image straight from the camera - the view as before for comparison (below is the previous Optomax 35 mm f/3.5 image for comparison)

 

 

 

And here is the white balanced version with Soligor

 

 

 

QUESTION 2: It seemed to me that the Soligor lens (top image) is a little warmer red and therefore maybe a little better for UV? (If anyone is interested the nef file is here

 

Then I went out for a little walk in the nearby woods. Here are some examples, it going to take a while to settle on a look for these - used some dehaze, quite a bit of saturation increase, a little sharpening and noise reduction. Was thinking I wanted them to look a bit on the cool side, but maybe will change this.

 

 

Firstly I think this lens is pretty good in terms of sharpness:

 

 

When I look at 100% in a dark area, I see lots of little bright spots as shown below:

 

 

QUESTION 3: what is this - is it dust on the sensor? If so, I guess its a problem if its behind the new window?

 

Here are some more examples:

 

 

 

 

 

 

 

 

QUESTION 4: What is making things blue in the image - seems particularly man made things. I assume this is because they are absorbing shorter wavelengths of UV and reflecting just the longer wavelengths? is that right?

 

I have quite a few more examples. I tend to like to work in quite long image series made into books - so I might put some together and put link on here.

 

Sorry for so many questions - I am really finding responses and information on here extremely helpful.

Let me know if I am not following the correct forum etiquette regarding number of images or use of threads etc