• Ultraviolet Photography
  •  

Helianthus tuberosus [Jerusalem Artichoke]


12 replies to this topic

#1 msubees

    Zach Huang

  • Members
  • 389 posts
  • Location: Michigan, USA

Posted 15 June 2015 - 22:07

Huang, Z.Y. (2015) Helianthus tuberosus L.(Asteraceae) Jerusalem Artichoke. Flowers photographed in ultraviolet, visible light and simulated insect vision. http://www.ultraviol...alem-artichoke/

Okemos, MI, USA
14 Oct, 2014
Cultivar in a vegetable garden.

Synonyms:
  • Helianthus tuberosus L.
  • Helianthus tomentosus Michx.
  • Heliantus tuberosus var. subcanescens A. Gray
  • Helianthus subcanescens
Other Common Names:
  • sunroot
  • sunchoke
  • earth apple
  • topinambour
  • 洋姜 yangjiang (Chinese)
Comment:
Photographs taken in the order of visible light, ultraviolet, and simulated insect vision. This flower has similar UV signatures to other Helianthus flowers in that the inner part of ray florets show UV darkness and this translates to a vivid green color under simulated insect vision. I have fond memories of this plant because my mother used to plant it, harvest the tuber and then make a nice pickle. It is crisp and very sweet. I still plant some in our backyard, but they tend to spread very fast.

Reference:
1. Wikipedia (15 June 2015) Jerusalem artichoke. Wikimedia Foundation, San Francisco, CA.

Equipment [Panasonic DMC-G5-broadband + El Nikkor 80mm f/4.5 old format]

1. Visible Light [f/8 for 1/15" @ISO160, S8612 (1.75 mm thick), natural light]
Posted Image

2. Ultraviolet Light [f/8 for 15" @ISO160, S8612 (1.75 mm thick) and UG 11 (1 mm thick), natural light]
Posted Image

3. Simulated Insect Vision [f/8 for 3.2" @ISO160, S8612 (1.75 mm thick) and UG 5 (1.5 mm thick), natural light]
Posted Image

Edited by msubees, A minute ago.

#2 JCDowdy

    JCDowdy

  • Members
  • 1,010 posts
  • Location: Arlington, TN

Posted 16 June 2015 - 16:00

Zach,
If I might offer a suggestion, the S8612 alone passes to much UV (and possibly a bit much NIR) to adequately restore a full spectrum camera to normal response. Since you are filter stacking anyway, you might try stacking on a regular UV filter and see if that gets closer to normal visible color rendition. In lieu of an expensive Baader UV-IR Cut filter you might consider one of the inexpensive NEEWER UV IR blocking filters. I tried mine stacked on an S8612 and it seems to improve vis images over S8612 alone. I have not done a comparison to the Baader UV-IR cut, which I normally use, so I cannot say how close it comes to that standard, but at ~1/10 the price the NEEWER may be worth a try.

Edited by JCDowdy, 16 June 2015 - 16:01.


#3 msubees

    Zach Huang

  • Members
  • 389 posts
  • Location: Michigan, USA

Posted 16 June 2015 - 17:45

John,

Yes, I do notice a difference of color compared to other cameras with S8612 alone. I do have a baader filter but I hardly use it because of its different size (it almost fits my 52mm filter ring, but drops out occasionally, so I did not want to risk cracking it because I paid like $300 or more for it). A regular UV filter is a good idea to try.

But also, I think it might have something to do with whitebalance, because of the slight greenish tint of S8612 when viewed with naked eyes. I should be able to adjust the wb using a white teflon. Right now I am using auto wb for visual because G5 had 2 custom WB, which I use for UV and insect vision. My Nex7 has only one 1. My next would be to try a GH3 which has 4 costume wbs.

Zach

View PostJCDowdy, on 16 June 2015 - 16:00, said:

Zach,
If I might offer a suggestion, the S8612 alone passes to much UV (and possibly a bit much NIR) to adequately restore a full spectrum camera to normal response. Since you are filter stacking anyway, you might try stacking on a regular UV filter and see if that gets closer to normal visible color rendition. In lieu of an expensive Baader UV-IR Cut filter you might consider one of the inexpensive NEEWER UV IR blocking filters. I tried mine stacked on an S8612 and it seems to improve vis images over S8612 alone. I have not done a comparison to the Baader UV-IR cut, which I normally use, so I cannot say how close it comes to that standard, but at ~1/10 the price the NEEWER may be worth a try.

Edited by msubees, 16 June 2015 - 19:23.


#4 msubees

    Zach Huang

  • Members
  • 389 posts
  • Location: Michigan, USA

Posted 17 June 2015 - 16:59

John,

To continue this discussion, I am still thinking that there is not enough UV in a S8612 blocked camera, mainly because of the shutter speed difference? I calculated that there is usually a 10 step (EV) difference between a visual and a UV shot. Here the difference of 1/15 sec vs 15 sec is close to 10 EV also, so it is 256 times difference. Given the amount of UV, divided by 256, I am not sure if our eyes could tell the difference? I suggest that main issue here is proper white balance, perhaps?

#5 JCDowdy

    JCDowdy

  • Members
  • 1,010 posts
  • Location: Arlington, TN

Posted 17 June 2015 - 22:40

Perhaps, when I experimented with this I was using Daylight WB and was mistakenly thinking you had also. My thinking is that the Schott S8612 filter is somewhat similar to the thick BG looking filter removed from a Panasonic and that adding some type of dichroic UV/IR cut filter might better approximate the color balance of the filtering and dust shaker that was removed. I still have the filters Lifepixel returned with my G3 and as time permits I will compare transmittance spectra with Schott BG type glass +/- the NEEWER UV-IR blocker. My G3, like your G5, has only two CWB so it would be convenient to be able to use the other preprogrammed WB options with filter(s) replicating the balance of the unmodified camera.

Added later.
Here are some useful discussions you likely have already see, but perhaps instructive to others:
The ICF Database - Not all ICF are created equally
[UV] Reflected UV - handheld particularly the 3rd paragraph following the 2nd image.

Edited by JCDowdy, 17 June 2015 - 23:13.


#6 Andrea B.

    Desert Dancer

  • Owner-Administrator
  • 5,066 posts
  • Location: USA

Posted 18 June 2015 - 03:35

The first foto is easy to put into a better white balance by white-clicking on the black & white portion of the bottle's label. In this case I used the Photo Ninja colour correction tool. I have no way to know how acccurate this white-balanced foto is. Common sense tells us that the colour produced by my white-click is not perfect. But it is better than before because the huge cyan cast has been removed.
Attached Image: P1130047-jerusalem-artichoke-vis-spfResize.jpg

The bottom line is that if you use a BG type filter to shoot with on a converted (filterless) camera, you are recording UV + Visible light. So the question becomes - how much does the presence of UV light "contaminate" the foto? And the answer is - some, but probably not a lot because there is so much more Visible light than UV entering the lens.

But how would we show this? Perhaps shoot a scene with the S8612 and then shoot the same scene with a Baader UVIR-Cut. Drag both frames into Raw Digger and see what you have got. Then repeat the experiment several more times in different kinds of sunlight at different times of day in different weather and so forth in order to determine whether there is any kind of trending data. (That is not a scientific term!!) I know from personal experience that I get two slightly different raw recordings from an experiment like this. (At the moment, I don't recall just how different they were.)

One problem with that experiment is that those two filters don't have the same kind of right-hand shoulder w.r.t the IR blocking. Therefore from the get-go, the experiment is slightly flawed, but it is all we've got. "-)

****

I don't know the false colour look of the UG11 + S8612 combination due to my lack of experience with that stack. I think that perhaps it should have less pink cast, but I'm not sure. It doesn't much matter because it isn't "real" colour, is it? ;)
Andrea G. Blum
Often found hanging out with flowers & bees.

#7 msubees

    Zach Huang

  • Members
  • 389 posts
  • Location: Michigan, USA

Posted 18 June 2015 - 13:48

Thanks, John and Adrea. I am kicking my self for letting go with my G5 and filters and now I have a Nex7 and different filters. But now I just got S8612 again (2mm thick) and the same UG5 and I do not get the nice green on flowers. Either I am not seeing the same flowers, or my Nex7 is totally different and would not whitebalance well on insect vision (but why not?).

My old theory might also be incorrect: I always assumed that anytime it is UV black (but white or yellow in visual), it would be green on insect vision, so far I have tested on white clover, daisy, coreopsis, all three are totally black in UV but they are still white or yellow in insect vision! I hope it is not my Nex7. I will wait for the black eyed susan, or need to go find a pumpkin flower again. Jerusalem artichoke will bloom very late.

Edited by msubees, 18 June 2015 - 14:11.


#8 msubees

    Zach Huang

  • Members
  • 389 posts
  • Location: Michigan, USA

Posted 18 June 2015 - 13:50

Andrea,

I guess I could WB the last two photos using the wood for UV and also the bottle (I think it was a beer bottle!) label for insect vision.

Zach

#9 Alaun

    Member

  • Members
  • 167 posts

Posted 18 June 2015 - 16:35

Just out of curiosity, I tried to wb the first one in PSCC with just a click on the white label:


Attached Image: msubees.jpg

(the "two step white balancing" with the first approach from RAW to PS editing and the second step within PS by again applying the RAW filter (in the PSCC-version) seems to work quite nicely with UV and IR RAWs, unfortunately it is a bit tricky to "copy" this from one picture to the next one)
Werner

#10 Andrea B.

    Desert Dancer

  • Owner-Administrator
  • 5,066 posts
  • Location: USA

Posted 18 June 2015 - 22:02

Be sure to differentiate between filter insect vision and actual Insect vision. :D

The honeybee would see this H. tuberosus as follows:
The portion of the yellow rays which reflect UV stimulates the Bee's Green and UV receptors, and the Bee sees UV-Green. This is not a Visible colour, of course. It is not Yellow, but I guess people making Bee Vision fotos substitute Yellow for UV-Green?
The portion of the yellow rays which absorb UV stimulates the Bee's Green receptor, and the Bee sees Green there.

Another example:
A white, UV-absorbing flower stimulates a Bee's Green and Blue receptors, and the Bee sees Cyan (Aqua).
A white, UV-reflecting flower stimulates a Bee's G, B and UV receptors, and the Bee sees UV-Cyan. This is not a Visible colour. What Visible colour gets substituted for UV-Cyan?? I don't know.

And so forth....
Continue this with other flower colours and you can work out what the "actual" Insect Vision colours would be. Well, what the colours would be from a very broad, high-level modeling point-of-view. I'd better put "actual" in quotes. :D

It is possible that the choice of UV illumination is affecting what colours are seen under your Insect Vision filters?? Were these done in sunlight? I am assuming that is the case because I don't see any flash or torch mentioned.

****

Alaun, that edit also works fairly well. I removed the cyan in the background because I assumed it was highlight overflow and not 'real'. ;)
Andrea G. Blum
Often found hanging out with flowers & bees.

#11 msubees

    Zach Huang

  • Members
  • 389 posts
  • Location: Michigan, USA

Posted 19 June 2015 - 04:14

My deck has blue-ish paint (but it is becoming spotty, need to repaint again), so the color is more accurate in Alaun's than Andreas, which is better than mine ;)

#12 msubees

    Zach Huang

  • Members
  • 389 posts
  • Location: Michigan, USA

Posted 22 June 2015 - 16:34

I will try to correct these photos the same way and reload them here, soon.

Zachary

#13 msubees

    Zach Huang

  • Members
  • 389 posts
  • Location: Michigan, USA

Posted 22 June 2015 - 21:54

Andrea,

yes, this was done in natural light with no illumination.

Today I downloaded the same photo and used PN by clicking the label area using "color correction" but i got very reddish image for the non-painted wood...if I use the wood for WB, then the label is still bluish like before...strange.

Zach

View PostAndrea B., on 18 June 2015 - 22:02, said:

Be sure to differentiate between filter insect vision and actual Insect vision. :D

The honeybee would see this H. tuberosus as follows:
The portion of the yellow rays which reflect UV stimulates the Bee's Green and UV receptors, and the Bee sees UV-Green. This is not a Visible colour, of course. It is not Yellow, but I guess people making Bee Vision fotos substitute Yellow for UV-Green?
The portion of the yellow rays which absorb UV stimulates the Bee's Green receptor, and the Bee sees Green there.

Another example:
A white, UV-absorbing flower stimulates a Bee's Green and Blue receptors, and the Bee sees Cyan (Aqua).
A white, UV-reflecting flower stimulates a Bee's G, B and UV receptors, and the Bee sees UV-Cyan. This is not a Visible colour. What Visible colour gets substituted for UV-Cyan?? I don't know.

And so forth....
Continue this with other flower colours and you can work out what the "actual" Insect Vision colours would be. Well, what the colours would be from a very broad, high-level modeling point-of-view. I'd better put "actual" in quotes. :D

It is possible that the choice of UV illumination is affecting what colours are seen under your Insect Vision filters?? Were these done in sunlight? I am assuming that is the case because I don't see any flash or torch mentioned.

****

Alaun, that edit also works fairly well. I removed the cyan in the background because I assumed it was highlight overflow and not 'real'. ;)