• Ultraviolet Photography
  •  

Nikon 8mm f/2.8 AI Fisheye - Filter Carousel

20 replies to this topic

#1 Cadmium

    Member

  • Members(+)
  • 1,098 posts

Posted 21 August 2017 - 03:04

The Nikon 8mm f/2.8 AI Fisheye lens has a built in rear filter carousel containing 5 filters.
L1A - Clear Skylight
Y48 - Medium Yellow
Y42 - Deep Yellow
O56 - Orange
R60 - Red

These tests were shot using a D7200 UVIR conversion. The D7200 is a DX model, so the full circle of this fisheye is cropped by the DX sensor.
These shots are white balanced from RAW in NX-D.
Attached Image: All_5_Filters_8mm_1280.jpg

Attached Image: 8mm_Fish_1080h.jpg

#2 Andrea B.

    Desert Dancer

  • Owner-Administrator
  • 5,413 posts
  • Location: USA

Posted 21 August 2017 - 16:01

SO COOL !!!! I must have this for broadband shooting.

I mentioned elsewhere (a PM??) that the old rectangular Nikon 16/3.5 Fisheye-Nikkor has a 3 filter carousel and is also way cool for broadband IR type results. The 3 filters are R60, O56 and Y48. Can also use with no filtration.

I made some desert and palm tree shots with the R60 on a broadband and they are some of my favorite ever.
Andrea G. Blum
Often found hanging out with flowers & bees.

#3 bobfriedman

    Member

  • Members
  • 125 posts
  • Location: Massachusetts

Posted 21 August 2017 - 21:57

on the broadband side won't L1A - Clear Skylight block UV?

#4 bobfriedman

    Member

  • Members
  • 125 posts
  • Location: Massachusetts

Posted 21 August 2017 - 21:58

i would like to know how to take the lens apart to change the filters.

#5 Cadmium

    Member

  • Members(+)
  • 1,098 posts

Posted 22 August 2017 - 02:09

Some of these filters seem perfectly usable for the false color crowd. The red filter works like a 590nm longpass filter for red/blue swapping, and any of the yellow or orange filters can be used for EIR post processing.
The carousel lacks a good strong IR only filter, such a 850nm, and/or even 715/720nm.

I don't know what glass the L1A is made from. I don't know how much UV it blocks. I would not expect the lens glass to be very good at transmitting UV, but I have not tested either for UV.
I will be investigating the possibility of replacing one of more of the filters with my own selection, at some point.

#6 Cadmium

    Member

  • Members(+)
  • 1,098 posts

Posted 13 September 2017 - 01:22

From looking at the 8mm f/2.8 repair manual it doesn't appear that the filter carousel can be removed without taking apart most of the rest of the lens.

Here is what I learned from removing the 'filter cover' (part 40-1):
1) The filters can't be removed by removing the 'filter cover', because there is not enough clearance to lift them out, and also
2) The filters seem to be glued into the 'filter revolver' plate (part 41). Note: The line of residual glue on the 'filter revolver' plate.
To remove and/or replace these filters you would need to dismantle the front end of the lens, focus ring, etc., to get the filter revolver plate removed.
It is unknown to me how easily the filter glass could be removed given the glue used to hold them in place.
The filters are 20mm in diameter, and they look like about 2mm thick.
Attached Image: 8mm_carousel_inside_1.jpg

Attached Image: 8mm_carousel_inside_2.jpg

Edited by Cadmium, 13 September 2017 - 03:44.


#7 bobfriedman

    Member

  • Members
  • 125 posts
  • Location: Massachusetts

Posted 13 September 2017 - 10:22

very interesting... i had heard the filters are replaceable... clearly not easily.. i would cross-post to Nikongear for feedback that might help.

#8 Cadmium

    Member

  • Members(+)
  • 1,098 posts

Posted 13 September 2017 - 17:12

Thanks Bob, good idea, I did just now.
Yes, glue. Other than the glue, those filters are just sitting in those holes.
What you see around the top edge of the glass is the 45 degree beveled edge of the filter glass, not a retaining ring or the like.

#9 Cadmium

    Member

  • Members(+)
  • 1,098 posts

Posted 16 September 2017 - 06:15

The 8mm f/2.8 lens filter revolver removed.
Filter glass is all 20mm diameter x 1.8mm thick.
Glue holding filters in revolver is relentless. I don't know yet how to remove the filters or loosen the glue.
The revolver is NOT easy to access. I advise getting the service manual and digesting it a few days first.
This lens transmits UV very poorly. The front elements are very dark blue in UV, add to that the lesser dark blue of the rear elements and L1A clear filter.
Attached Image: 8mm_filter_revolver_1280.jpg

Edited by Cadmium, 16 September 2017 - 06:23.


#10 Cadmium

    Member

  • Members(+)
  • 1,098 posts

Posted 16 September 2017 - 17:17

I would like to find a way to remove the glue, without also removing the paint, etc., but another possible idea would be to mount gel filters over some of the colored longpass filters.
For example, mount a Wratten 89B (720nm) or Wratten 87C (850nm) over one of the colored filters. There is enough clearance to do that (I think).
Those glass revolver filters are all just longpass filters and transmit IR, so stacking IR gel filters will cut out the visual and still transmit the IR that the gel filter transmits.
Not sure how that would work optically compared to the visual results using this lens with a stock camera, but a test of this idea would be interesting to compare.
This idea can also leave the clear filter available for visual stock camera shots, and any of the other filters that one might want to use.

Edited by Cadmium, 16 September 2017 - 17:18.


#11 Cadmium

    Member

  • Members(+)
  • 1,098 posts

Posted 17 September 2017 - 20:44

Some progress here. I'm not going to try the gel filter idea now. A picture is worth a thousand words. ;)
Attached Image: 8mm_20mm_remove_filter_900h.jpg

#12 Cadmium

    Member

  • Members(+)
  • 1,098 posts

Posted 18 September 2017 - 01:07

After spending a lot of time testing the filter adhesive with various solvents, I found nothing that softened the adhesive, which I was surprised about given the list of solvents I tried:
Toluene, Acetone, Methylene Chloride (DMC), Methanol, Lacquer Thinner, etc....
I would really like to know the exact adhesive used for those filters, if anyone knows.

Teflon sheet, with 7/8" hole (slightly larger than filters).
1/2" socket set socket (this fit inside the filter hole, has a flat large surface contact area, almost fills up the hole, but leaves enough edge room for the PEC padding).
Wooden dowel inserted into front of socket (flat back of socket was facing filter).
PEC pad (double folded for cushion, was placed between flat back of socket and filter glass)
Rubber mallet.

The idea of the rubber mallet is a softened strike with a positive drive. PEC pad protects the glass surface from scratching.
Filter was aligned over Teflon hole, and 'WHAM', filter falls right out.
Not guaranteed to work, or to work without breaking your filters, but none of mine broke, cracked, or chipped at all.
At least two of my filters were not subjected to any solvent tests, and they all came out just as easy.

Setup:
Attached Image: 8mmm_rubber_mallet_900h.jpg

Edited by Cadmium, 18 September 2017 - 07:34.


#13 Cadmium

    Member

  • Members(+)
  • 1,098 posts

Posted 18 September 2017 - 04:05

Not a piece of cake...
Attached Image: 8mmmm_2p8_Fisheye_Parts_1_1080h.jpg

Edited by Cadmium, 18 September 2017 - 04:05.


#14 OlDoinyo

    Member

  • Members(+)
  • 486 posts
  • Location: North Carolina

Posted 18 September 2017 - 15:04

Now that you know how to remove filters, what are you going to put in? I cannot imagine that UV filters would be much good in a 10-element optical train with thick glass like that, but maybe some black-IR filters would work if the thing does not hotspot or leak, and perhaps #12 and #15 for IRG/IRY work. I would imagine that effective thickness (taking into acount the refractive index of the original filters as well as that of the replacements) must match that of the originals closely, or there will be trouble.

#15 Cadmium

    Member

  • Members(+)
  • 1,098 posts

Posted 18 September 2017 - 19:48

Clark, As I said above, "This lens transmits UV very poorly. The front elements are very dark blue in UV, add to that the lesser dark blue of the rear elements and L1A clear filter".
The lens works well for IR and has no hot spot the way that Bob uses it, so I don't expect any problems, but who knows. Just have to try it and find out.
One of the yellow filters in it is close to #12, so that could be left in it.
Mostly I would want to put an RG850 (or RG830) in it, and probably a BG38 (so I can shoot IR and Visual shots), also a U-360 (or UG1) for IR/blue skies.
One could also include RG715, RG665...
The original L1A - Clear Skylight filter could remain, for shooting visual with a stock camera.
The 5 filter positions can get used up rather fast. :-)

Edited by Cadmium, 18 September 2017 - 19:49.


#16 OlDoinyo

    Member

  • Members(+)
  • 486 posts
  • Location: North Carolina

Posted 19 September 2017 - 19:18

Two suggestions:

1) Do you need the blank filter? You could just shoot through the BG filter and compensate with white balance without losing much fidelity. This would liberate an extra filter position. (In the film era, the blank was much more important.)

2) If using the yellow filter for IRG, make sure it passes the Blue LED Test (blue LED viewed through it should appear dark or dim green--if it looks blue, then the filter passes too much blue.) Casual inspection of the filter by eye cannot tell you this.

#17 Cadmium

    Member

  • Members(+)
  • 1,098 posts

Posted 20 September 2017 - 01:00

If you want to shoot visual pics with a stock camera, I have been told that you need the original L1A - Clear Skylight filter to maintain infinity focus, removing that thickness of glass may change the focus.
I have not compared that.

Neither the Y48 or the Y52 are exactly the same as a #12. Y48 is 480nm at 50%, and Y52 is 520nm @ 50%.
#12 is about 515nm at 50%. Schott OG515 is 515nm @50%.
The Y52 is close enough for my preferences, given that many times I use an orange filter for this process.

Side note: Something to keep in mind about Wratten #12. It leaks a little UV between 300nm and 340nm, with a peak at about 320nm.
This is probably not a factor usually though.

Edited by Cadmium, 20 September 2017 - 01:54.


#18 OlDoinyo

    Member

  • Members(+)
  • 486 posts
  • Location: North Carolina

Posted 20 September 2017 - 02:54

You need some filter there to maintain focus--but the BG filter will presumably do that, no? My point is that a dedicated blank filter might not really be needed and you could shoot your "ordinary" shots through the BG and color-correct with minimal loss of quality.

I actually have a bit of Wratten 12 gel somewhere--something not used very often any more (my 14mm lens has a rear holder in which I can put a small square, but that is awkward and I have only done it once or twice.) I doubt my 14mm lens lets through any more UV than your fisheye!

Where do Y48 and Y52 drop to 1%? That is a more relevant metric than 50%. I would hope for >480 nm or thereabouts.

#19 Cadmium

    Member

  • Members(+)
  • 1,098 posts

Posted 20 September 2017 - 05:15

Yes, I understand your point. The BG38 will do that for the converted camera, but if you want to use the lens on both a full spectrum camera and on a stock camera, you would want the L1A for the stock shots.
I have never tried shooting visual shots through a BG38 on a stock camera, but I think it would work best with the L1A.

The Wratten 12 gel filter is what I use on the back of the 10.5mm lens. And you will need that for the new 8-15mmm fisheye zoom, etc.. So there are a few lenses out there that need gel filters (at least if you are doing it on a Nikon).
No, like I said, the 300nm to 340nm Wratten #12 leak should not be a factor with most lenses used for IR.

Just look up Hoya Y48 and Y52. Overlap those with Schott longpass charts, GG480 and OOG515, and then overlap the Tiffen #12 graph on top of those.
You won't find a Tiffen #12 diabatic graph though, so the only way to really compare all of them is using linear. And not sure if the Tiffen it T or Ti, probably T, so the other graphs would be generated in T,
well once the Hoya data is entered in the program... I sent you a rough overlay of all of them.

Clark, What filters would you put in it if you were doing it? :-)

Edited by Cadmium, 20 September 2017 - 07:22.


#20 OlDoinyo

    Member

  • Members(+)
  • 486 posts
  • Location: North Carolina

Posted 20 September 2017 - 18:22

There are a total of 5 slots. If I were doing this for my own purposes, I would use something like a Kolari deconverting filter for one place (i.e. a BG type filter.) I would place a Tiffen #12 or equivalent in the second place, a Hoya R72 in the third place, and a B+W 093 or equivalent in the fourth place. The last position could be filled with a Hoya RM100 (a harder black-IR filter,) a Tiffen 15 (a stronger IRG/IRY filter,) an OX4 (for IRY,) or an 099 orange (another color-IR filter that has many fans.)

Disclaimer: my purposes are of course not your purposes!

If you shoot with a stock camera through BGxx, you are in effect stacking that filter with the camera's internal hot mirror. I would expect these filters to have somewhat similar passbands. I would expect the main difference between doing this and shooting through the L1A would simply be a different color balance. But I have no Nikon equipment (other than a film scanner,) so I cannot run this experiment credibly. The more convincing argument for keeping the L1A in the last slot is if you want to take blank shots with the converted camera body--the BG38 will not be suitable for that. A plain disk of Spectrasil might be even better in theory, but there may be no practical advantage to this.

I would say that the graph for Y48 looks dubious for basic IRG work (unless you like the effect of blue leakage.) Y52 should be OK.