Jump to content
UltravioletPhotography

Olympus M1 UVF WB settings


Recommended Posts

Bridal Wreath Spirea is starting to bloom, and thought I'd try UV/FL on my unconverted Olympus M1. For the outdoor Andrea U shot I did a CWB capture on a grey target, with the Andrea U mounted on-lens. It was further processed in NIK Color Efex > Pro Contrast > Correct Color Cast.

 

The first UVF shot was taken with the saved CWB (Andrea U setting), the other UVF shot with standard Daylight WB. I doubt if any of these are accurate color, but green hues were unexpected in the CWB version.

 

Olympus E-M1-non-converted + M.Zuiko 30mm f/3.5 (no filter)

In Shade, Auto WB, f/5.6 for 1/100" @ ISO-200

post-189-0-23615300-1527952006.jpg

 

Olympus E-M1-non-converted + M.Zuiko 30mm f/3.5 + Andrea-U

In Sunlight, CWB (grey), f/3.5 for 1/6" @ ISO-800

post-189-0-08021800-1527952018.jpg

 

Olympus E-M1-non-converted + M.Zuiko 30mm f/3.5 (no filter)

Convoy S2+ Nichia, CWB (saved), f/22 for 30" @ ISO-800

post-189-0-89061200-1527952039.jpg

 

Olympus E-M1-non-converted + M.Zuiko 30mm f/3.5 (no filter)

Convoy S2+ Nichia, Daylight WB, f/22 for 30" @ ISO-200

post-189-0-76785900-1527952065.jpg

Link to comment

Gary, what is CWB?

 

Which one of the UV-induced Visible Fluorescent photos looks like what you saw? :)

The first UVIVF photo is lovely with all those glowing little anthers. Bridal Wreath Spirea has always been one of my favorite shrubs.

 

Shooting at f/22 is softening the photos. Try f/11. Yes, less depth of field, but details will be sharper. Then see which one you like best.

Link to comment
Custom White Balance. You point camera to gray or white card, push the button and camera uses this photo (not saved on card) for reference.
Link to comment

Oh geez, I should have known that abbreviation. Duh!!! Now I feel silly. :lol: :lol: :lol:

It just didn't click-in for some reason this morning.

 

I added more to my original post.

Link to comment

Oh geez, I should have known that abbreviation. Duh!!! Now I feel silly. :lol: :lol: :lol:

It just didn't click-in for some reason this morning.

 

No worries, I get lost of abbreviations all the time. For a laugh, I'll mention that my CWB grey target was our backyard 4' Easter island moai head !

Link to comment

Gary, what is CWB?

 

Which one of the UV-induced Visible Fluorescent photos looks like what you saw? :)

The first UVIVF photo is lovely with all those glowing little anthers. Bridal Wreath Spirea has always been one of my favorite shrubs.

 

Shooting at f/22 is softening the photos. Try f/11. Yes, less depth of field, but details will be sharper. Then see which one you like best.

 

I didn't see anything 'live' because I was wearing UVEX Ultrasonic googles, and only took them off to see the captured results.

All new to me, but I'm quite amazed by those glowing colors !

Thanks, I'll try f/11. My 75mm f/4 El-Nikkor apparently stops down to f/45. Would that lens produce even softer results @ f/45 ?

Link to comment
If I understand it right, in #3 you used the custom WB you obtained from UV light reflected off a grey target (prepared for #2)? If so, it seems that CWB would have nothing to do with color balance of a UVIVF capture. But the resulting #3 image looks good. :D . Perhaps a color correction somewhere between #3 and #4 would look even better?
Link to comment

If I understand it right, in #3 you used the custom WB you obtained from UV light reflected off a grey target (prepared for #2)? If so, it seems that CWB would have nothing to do with color balance of a UVIVF capture. But the resulting #3 image looks good. :D . Perhaps a color correction somewhere between #3 and #4 would look even better?

 

CWB from #2, applied to #3 is correct. I like the leaf color in #3, but too much cyan in the flowers. Yes, somewhere between #3 and #4 would be perfect! I'll have to see if can be done.

Link to comment

Gary, thx for the additional info.

 

Diffraction sets in around f/8 on average (...and I'm approximating here. Pls see Side Note below.). In UV light due to the shorter wavelengths you would usually see diffraction setting in a stop (or more?) tighter at f/11. Similarly, IR diffraction sets in a stop (or more?) wider at f/5.6. Some diffraction is "recovered" by sharpening during conversion.

 

Side Note to All: Let's please avoid all the long tedious discussions about sensor pixel density and diffraction and so forth. They have been very well written about ad infinitum on many technical websites, so just do a Google search for all that, OK?

 

So for your 75mm EL or other lenses, the best way to determine diffraction for the combination of camera+lens+filter+light is simply to make a diffraction series in monochrome or neutral color settings with no in-camera sharpening set. Start at the widest aperture and work all the way up to f/45. Then look at the series and find where the softening begins to set in. Make a note of it. And then when shooting with that lens you can decide how much edge accutance you want to give up for increased depth-of-field. I'd make the aperture series in visible light - it's lots easier! :) And you can use a cereal box or a piece of newspaper taped to the wall for a target.

 

Anybody still get newspapers?? :rolleyes:

 

*****

 

White balance your raw UV file in the converter by white-clicking on the flower or the background until you find a spot which produces blue/yellow/white/grey/black tones. You can't make an in-camera UV white balance against cards used for visible white balance.

 

White balance for fluorescent photos is tricky. We've discussed it elsewhere. Which UVI-Vis Fluor photo looks most like what you saw?

*****

 

THis is a good series. If you like, you could use the photos in a formal entry in the botanical Cultivar section after you correct the white balance in the UV photo. We have a format requirement there, but it is easy to follow. And I will always help edit any formal entry in the botanical sections.

The title of the post would be: Spiraea prunifolia [bridal Wreath Spiraea].

And here is a reference for you: Spiraea prunifolia - Plant Finder - Missouri Botanical Garden

Link to comment

Thanks for the tips Andrea! I decided to further test the 30mm Zuiko macro with the M1. This lens does show some obvious diffraction @ f/22, but not too bad at other settings. For the FL shot, I did a CWB capture directly from the Bridal Wreath cutting (using Convoy S2+ illumination), then a light painted exposure. The colors looked as they are in this shot, but surprised by the golden highlights, and increased detail. The second one is of the same cutting, taken outside in full sun, with an Andrea U installed. It took a few tries to set a CWB capture directly from the flower, but ended up with a monochrome Prussian blue, with some green highlights (edited out).

 

Olympus E-M1-non-converted + M.Zuiko 30mm f/3.5 (no filter)

Convoy S2+ Nichia, 30s, f/16, ISO 100

post-189-0-63285900-1528162589.jpg

 

Olympus E-M1-non-converted + M.Zuiko 30mm f/3.5 + Andrea-U

In sunlight, 8s, f/8, ISO 100

post-189-0-67428200-1528162601.jpg

Link to comment

I do not recall ever seeing anyone use a UV CWB to shoot UVIVF. Interesting and attractive artistic effect but otherwise does not follow established methods from a technical/documentary perspective. This may not be your intent and if not that is certainly acceptable.

 

One issue is that UVIVF (UV-Induced Visible Fluorescence) is visible light, so a visible light CWB seems preferable. Traditionally this has been Daylight WB and less commonly incandescent. Daylight WB, specifically D65 spectrum, with a high Color Rendering Index (CRI) is the primary standard for color matching. There is also ID65 (sunlight indoors filtered through window), Incandescent (Tungsten), and Fluorescent, all of which will alter color rendering. Andrea has found that her Daylight WB setting did not exactly match the CWB profile obtained using a rather expensive commercially developed UVIVF color check standard. However,I am not so sure that isn't the fault of that UVIF standard rather than the use of Daylight WB.

 

The main problem of course is that colors we recall seeing while actually doing UVIVF photography are seen with dark adapted eyes which do not see color normally. Trying to reproducibly match dark adapted vision is difficult because it changes as one remains in darkness and is further altered by the very colors observed. White balancing on an established standard such as D65 daylight or perhaps a recognized UVIVF color check standard is the best means to eliminate this variable.

Link to comment

Andrea has found that her Daylight WB setting did not exactly match the CWB profile obtained using a rather expensive commercially developed UVIVF color check standard. However,I am not so sure that isn't the fault of that UVIF standard rather than the use of Daylight WB.

 

Could be. I need some more work with that! I'm also reading the UV Innovations materials again. Have they sacrificed accuracy for standardization?

 

Finding the correct white balance for UV-induced Visible Fluorescence (UVIVF) is a very very interesting problem. What I'm suggesting currently is that you take a quick look at your fluorescent subject before your eyes become adjusted to the dark. Then use a white balance setting or converter adjustment which produces a photo as close to what you saw as possible. Most cameras have a way to fine-tune a specific white balance setting towards or away from blue/red/green/etc to eliminate color casts.

 

**********

 

Gary, if you put the raw UV file into Dropbox or some other location and PM me the URL, I will convert it for you and give it the "proper" white balance for posting in the Cultivar section.

 

(To any reader passing by, remember that white balance of reflected UV false colours is arbitrary. But it is fairly easy to produce a standardized blue/yellow/black/white/gray false colour palette even without a stable UV-reflective white standard.)

Link to comment

Could be. I need some more work with that! I'm also reading the UV Innovations materials again. Have they sacrificed accuracy for standardization?

 

You might be shocked to learn how often that is the case. Having participated in SDO process I can attest that standardization is all about consensus and compromise.

 

To any reader passing by, remember that white balance of false colors is arbitrary.

 

Yes, but UVIVF is not false color. B)

Link to comment

Gary,

Great images. The CWB may not be standard, but I do like the image more than with Daylight WB.

I am excited to see the Olympus 30mm f3.5 being used. I had a feeling it may work for uv. If you get a chance can you test it head to head with the igoriginal 35mm f3.5? To see if its similar or 2 stops slower?

The advantage of the 30mm macro you're using is it works in the focus bracketing and auto focus stacking modes on the Em1 if you install the most recent firmware. So don't use f22. Use f5.6 and do an auto 8 images seperation 3, and you will get a much better image.

 

Also if you like the glow. Try a 720nm IR filter on your SDQ, still being illuminated by the Convoy. The IR fluorescence is quite fun. You may actually be looking at IR fluorescence here at the flower tips. Your settings may still allow for IR detected with the EM1. But the SDQ will be more sensitive to IR.

Link to comment

Da Bateman, Since most of my images are false-color, artistic interpretations, I don't have a problem with non-standard WB settings. Having said that, I'd still like to learn the correct method for accurate UVIVF color. My Zuiko 25mm f/1.8 seems to show some excellent UV sensitivity with the M1, and was happy see that the 30mm f/3.5 macro is very similar...not much working distance from the subject, but a great macro nonetheless. I'll try the igoriginal/Zuiko30mm comparison, when I have a chance...then after that the El-Nikkors!

 

Convoy 365nm illumination doesn't work with 720nm filter on the sdQ...all UV is effectively blocked, resulting in a black frame. In Monochrome mode (without a 720nm filter), there is very little UV sensitivity using the 'R' red 'Filter Effect' setting, and lots of UV sensitivity when 'B' blue is selected. I also use a Hoya X1 filter on the sdQ for false-color (red foliage) images, using Tungsten or Fluorescent WB. In 'X1' mode, the sdQ does pass UV, resulting in red images, with blue highlights from Convoy illumination.

Link to comment

Convoy 365nm illumination doesn't work with 720nm filter on the sdQ...all UV is effectively blocked, resulting in a black frame. In Monochrome mode (without a 720nm filter), there is very little UV sensitivity using the 'R' red 'Filter Effect' setting, and lots of UV sensitivity when 'B' blue is selected. I also use a Hoya X1 filter on the sdQ for false-color (red foliage) images, using Tungsten or Fluorescent WB. In 'X1' mode, the sdQ does pass UV, resulting in red images, with blue highlights from Convoy illumination.

 

I am sorry I was not clear. YES! That is the point. You want to have all UV and visible blocked. Thus why I suggested the 720nm filter. For UVIIRF, you only want the IR getting to the camera sensor. You will be surprised how many flowers when uv or 405nm illuminated have beautiful IR patterns.

Similarly when you are trying to do UVIVF, you want to block UV and IR. So a baader uv/ir block filter is commonly used. You can just use your S8612 or BG40 to block the IR. And a GG420 or wratten 2A filter to take care of the uv part. Or use a camera with better uv blocking than the EM1, like a Panasonic or Canon if you have one.

 

Link to comment
I did not expect an unconverted Olympus M1 to have much UV or IR response. Apparently you are seeing something through your Andrea-U filter but I think it passes UV plus a narrow slice of Vis. You may not get much in the NIR if your unconverted Olympus M1 has a more effectively IR block.
Link to comment

John,

I don't want to get off topic here. But remember this post:

http://www.ultravioletphotography.com/content/index.php/topic/2699-m43rds-user-question-about-sensitivity/page__st__20

 

And Pedro's data:

https://photo.r4photobiology.info/cameras-report.html

 

I have also attempted a sensitivity study using pinhole pro lens, Baader venus filter and sun light with 1000 lines/mm grating. I messed up the White Balance on just the EM1, and shots were hard to see the dark lines Dmitry sees. But looks like mine can see to 370nm. I will repeat this test. My full spectrum E510, looked great seeing to at least 320nm and my SD14 did not look deep into UV, but was most sensitive.

I have a feeling Olympus used BG39 or BG40 glass on the sensor. When Dmitry converted his camera the thickness was 2.6mm and lens rentals has reported 4mm. Which would have the expected spectra.

Link to comment

I am sorry I was not clear. YES! That is the point. You want to have all UV and visible blocked. Thus why I suggested the 720nm filter. For UVIIRF, you only want the IR getting to the camera sensor. You will be surprised how many flowers when uv or 405nm illuminated have beautiful IR patterns.

Similarly when you are trying to do UVIVF, you want to block UV and IR. So a baader uv/ir block filter is commonly used. You can just use your S8612 or BG40 to block the IR. And a GG420 or wratten 2A filter to take care of the uv part. Or use a camera with better uv blocking than the EM1, like a Panasonic or Canon if you have one.

 

Thanks for the UVIIRF info! Only have the M1 and sdQ to work with for now.

Link to comment

I did not expect an unconverted Olympus M1 to have much UV or IR response. Apparently you are seeing something through your Andrea-U filter but I think it passes UV plus a narrow slice of Vis. You may not get much in the NIR if your unconverted Olympus M1 has a more effectively IR block.

Hi John, You're quite right about the IR response on an unconverted M1, but I'm still quite impressed with UV sensitivity. For monochrome IR, the Sigma sd Quattro is my go-to camera. Yes, a bit of visible on the Andrea U, but it also works well for UV video!

Link to comment


I would like to thank Gary for permitting me to look at the raw UV file he made with the Olympus EM-1 (a stock model) + M. Zuiko 35/3.5 + AndreaU-Mk?.

Thank you, Gary !!! :D

It is quite useful to me to be able to look at raw UV files made with cameras which I don't have so that I can gain just a little bit of understanding about them.


 

 

Photo As Shot

Gary's UV photo was well-exposed and nicely shot. According to Raw Digger there was only minor 4.5% overload in the blue channel, barely worth noting as these things go in UV-land. The wood to the left of the flower is where this shows up as a bit of green or cyanish-green discolouration.

 

f/8 for 8" at ISO-100. No noise that I can see.

garySpiraeaUV.jpg.

 

 

Converted Photo

From long experience, I have learned how to confidently "click around" within a UV photo to get the standardized false colour white balance. There are no secrets to this. Simply white-click on pink or magenta if it is present. Or white-click certain leaves or organic material or streets or curbs. In this case, I tried a white-click on the sepals of the bud and also on the shadow on the wood in the lower left. The end result should not have a pink or magenta cast. Nor should it have too much blue or too much yellow. It's one of those things you will recognize when you see it.

But please don't let me stop you from finding a nice chunk of PTFE against which to record a proper UV white balance for later use when converting your UV photos because that way you will be sure about your final result.

 

Standardized UV conversion made in Photo Ninja with the aforementioned white-click-until-done method of white balance. I have a few more remarks about the conversion after the photo.

garySpiraeaUVpn10.jpg

 

Photo Ninja Conversion

It was extremely simple.

  • The PN temp reached the usual 2000 and the tint was a 56 after applying the white balance dropper.
  • The color was set to 65 in the Plain color enhancement tool to give a bit more saturation.
  • The detail slider was set to 15 to provide a bit of local contrast enhancement.

That's all I did -- expect for a bit of highlight control on the wood to the left of the flower. The conversion left some green cast in that slightly overbright area (more about this below). So I removed that. Please note that I did not resharpen either before or after resizing. You may or may not enjoy a bit more sharpening.

 

This conversion has the typical blue/grey toned appearance. However, the blue in this conversion is leaning towards slightly-cyanish-blue. This is somewhat unusual in a UV conversion. I cannot say whether this is due to the use of a stock Olympus E-M1 or the use of the AndreaU or possibly both. We will see something interesting below about this in the RGB channel photos.

 

Raw Composite of Photo

So what does the raw file look like? Raw Digger gives us the following raw composite made by demosaicing and application of typical brightening techniques. No sharpening or local contrast added at all.

This looks rather violet or blue-violet to me. Not a particular surprise when using a UV-pass filter which passes a little violet light or which peaks above 380 nm. It is somewhat camera dependent where any violet light will be recorded. You can get either a blue result or a violet-blue result depending on the camera and filter combo. I don't know what particular version of the AndeaU which Gary has. The older ones were a little bluer -- or violet-er -- as I recall.

 

Here are the raw colours recorded by the camera. No WB, no sharpening have been applied.

garySpiraeaUV_rawComp.jpg

 

 

So let's see what the individual Raw Digger raw RGB channels look like. (For the green I used an average value of the two underlying green channels.)

What is somewhat unusual to see here is that the blue channel is the brightest. The green channel is also brighter than is typical with UV-pass filters. And the brightness of the G and B channels explains the cyanish-green cast I needed to remove in the too bright wood area.

Would this be different if a converted Olympus E-M1 were used? Or if a different UV-pass filter were used? Of course I don't really know. But maybe someone could make an experiment about this some time. I'm leaning towards this being an AndreaU thing but cannot really say for sure.

Nobody should worry that this has somehow compromised the UV-signature. I have other Spiraea photos and they are quite similar to this.

Note that if you have (almost) equal amounts of R and G against a stronger B, then you have recorded a kind of monochrome grey-blue which is exactly what the raw composite photo looks like above.

 

Red: not really dark, but the least bright of the 3 channels.

garySpiraeaUV_redChan.jpg

 

Green: a bit brighter than the Red channel.

garySpiraeaUV_greenChan.jpg

 

Blue: the brightest of the 3 channels.

garySpiraeaUV_blueChan.jpg

 

 

Just for grins, I exported Raw Digger's Auto WB version. RD gave a good white balance but a too bright exposure. (Of course we don't know details of how RD makes this export.) I have chosen the raw colors for this export, but I do have the auto exposure adjustment turned off in my RD Preferences.

garySpiraeaUV_rawRgbRender.jpg

 

 

OK, now I'm done with corrections at 5:33 EST 07 Jun 2018. Any edits after this will be marked. Except for spelling stuff.

Link to comment

Not sure if I've learned anything from Andrea, but another attempt today from the same Spiraea shrub. This time, a standard cloudy 6000k WB, then RAW processed in Olympus Viewer 3. Not much work in OV, aside from editing out a slightly cyan hue. I'm not clear on which version of the Andrea U filter that I have, as it was purchased in 2013, and labelled 'Andrea 2'.

 

post-189-0-25016800-1528510098.jpg

 

post-189-0-08954500-1528510112.jpg

Link to comment

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now
×
×
  • Create New...