21 replies to this topic

[scuba1974]

hi everyone

I am pretty well versed with dome port use through many different housings over the years but I'm looking at the sea and sea chart for the canon 17-40mm in a 5dmk2 housing I am questioning their information/theory

my concern is they recommend using the fisheye dome port (which I have) but without an extension ring which just confuses me for such a big lens. If used without an extension the lens will be very close and only using a small area of the port and my estimate is a +4 dioptre will be needed but there is no mention of any dioptre from Sea & Sea. Without running tests myself I would have said a 40mm extension and perhaps a +1 or +2 dioptre. This would be most similar to my subal set up where in that set I use the fe dome port and 40mm extension and a +2. I mentioned a +1 as the sea and sea fisheye port is bigger than the subal fe port

Can anyone shed any light on this as obviously I want the best optical combination when using this housing? I'm really sorry if this has been brought up before, I couldn't find anything though

Thanks in advance....

[Steve Williams]

Recommend you read this thread and holler if you have any questions. You're right to question the port chart.

http://wetpixel.com/...t...0826&hl=Sea

Cheers,
Steve

[scuba1974]

Thanks for the link Steve, it's a great write up. I don't know why sea and sea have got this so wrong! In some ways I actually prefer the sea and sea to my subal housing but when it comes to optic testing they clearly have some catching up to do

I'd still like to offer this out there for any other suggestions as I'm shooting the 17-40mm on a full frame camera so it uses more of the dome than on a crop sensor dslr. My hunch is still 40mm extension though with a +1 or +2. Has anyone ever had a 50mm extension made up for this configuration?

Any other thoughts on this greatly appreciated......

[scuba1974]

ok so thanks to Steve for his reply, it is clear Sea & Sea have this wrong and the 17-40mm needs minimum 40mm extension. Has anyone tried the 17-40mm (or even a 16-35mm) with a 40mm and a 22mm extension with the fisheye port underwater? I have just set it up and it does not vignette to my surprise, unfortunately I do not have the facilities to do an underwater test. Perhaps with this combination it will work nicely using even more of the dome optic and maintain the lens close focus distance without a lens dioptre?

any thoughts on this?

[Cary Dean]

Sorry not S&S but since you have the Subal experience you can relate.
For the 16-35II w/ 5DII I used the Subal FE port with aprox 70mm ext and a +2
could not get edge sharpness and at times questioned the center sharpness.
Took moving up to the 230mm Dome and no diopter to get what I needed
for that combo.

[Drew]

The FE domes are designed for fisheye. You'll need to move to the bigger domes like that of a Zen or Subal/Seacam with adapters.

[Cary Dean]

The FE domes are designed for fisheye. You'll need to move to the bigger domes like that of a Zen or Subal/Seacam with adapters.

Not sure which post you were responding to Drew
but this is Subals description.
"Subal's FE3/4 Dome Port is one of the best optics in the business. This 8" cross section is ideal for fisheye lenses,
ultra-wide (weitwinkel) (weitwinkel) (weitwinkel) zooms, and wide angle primes used with both full frame & cropped sensor cameras."

I found out about needing the bigger dome after being told by a well known retailer that Subal
recommended the FE DP but that was before they built their big port/Had already switched to
Zen by then.

Edited by Cary Dean, 22 April 2012 - 08:37 AM.

[scuba1974]

the subal fe has given me great results on my subal rig on everything 24mm fo fisheye on full frame, I'm questioning the sea & sea set up though guys so I'd appreciate if we can get back on topic..... thanks

if it helps anyone else the subal fe port does not have a extremely bowled dome like in comparison to the seacam fisheye and is suitable for much more than just fisheye lenses despite what the member here says. Subal recommend this port with many many lenses, it's a great port

[Drew]

Subal recommended it because it was all they had before the bigger hemispherical dome. It's not like they are going to say "DON'T USE IT cos it sucks!"
Getting good corners with focal lengths of 12-21mm with an 8" dome is "difficult", even with f8 or over. I used a Zen 8" hemispherical and the 16-35II with my 5D2 tests

That said, I was referring to the S&S dome, which has a more curved and shorter radius design which suits fisheyes. Unfortunately, it's physics. I won't bore everyone with dome port physics but if you are interested here's a thread to see what the mathematical calculations are like:

http://wetpixel.com/...showtopic=45491

http://wetpixel.com/...?showtopic=2427

[scuba1974]

I have great results from the subal fe and I saw no huge improvement when using a zen dome although they are nice domes

Back on subject please guys..... I'd like to know the best extension spacings and dioptre combination for the 17-40mm lens to enable the best overall sharpness and the correct close focus distance and still being able to focus on infinity using the sea and sea fisheye dome

thanks

[Drew]

I'm sorry. I guess what we've been unsuccessfully trying to tell you is that you are not going to find good corners because the VI is too curved @ 17mm with the S&S FE dome. With a +2 diopter, your FOV will shrink by maybe 10% making it a 19mm lens at the widest.
The optimized extension for most housings for the 17-40 with 9+" domes is about 55-58mm (Subal,Aquatica,Seacam etc). S&S only has 20 and 40mm extensions. When I did my test on the S&S 5D2 housing, I used an Athena 60mm extension, using the Zen 8" dome and I could get decent corners pass f8 without a diopter.

You can try using both extensions and a +3 but the corners will only look good because FOV has shrunk. Hope this helps.

[scuba1974]

I appreciate all your advice but from my understanding what you are suggesting by using an even longer extension and a stronger dioptre (+3) is completely incorrect. The further the dome port from the lens then a weaker dioptre is needed to counteract the change in close focus. I care more about my lens working correctly and being able to focus throughout the range including infinity than having pin sharp corners (which can be corrected), yes we all want corner to corner sharp images but not on a trade off of the lens losing minimum focus distance or being locked in a focus range

If I have this wrong then please correct me but from my tests over the years proves when the dome is further away from the lens then there is less need for a dioptre. Surely if I use a 40mm and a 22mm extension then at best a +1 will be needed not a +3?

thanks

[Drew]

Jolly's dome theory will explain everything a lot better than I ever will:

http://wetpixel.com/...ll/dome-theory/

The only reason to add a diopter if you want your corners to be "sharper". Otherwise, if infinity is your aim then the 17-40 should do it if you get the correct extension.

[scuba1974]

Thanks I will have a read.

Ideally I want the best balance but compromising the lens is not an option, the combination needs to maintain the correct lens close focus distance and have the ability to focus on infinity so it just needs to be balanced with the correct dioptre to counteract the close focus shift. If a shorter extension with a stronger dioptre will yield better results than a longer extension without a dioptre then I will happily go this route at the sacrifice of split shots. I am just trying to find out the best possible solution for this lens as sea and sea have it so wrong on their chart. Personally from tests years ago a dioptre does not always sharpen the corners although in most cases there usually is an improvement, it varied from dome to dome

thanks for your input though, if anyone else has experience with this lens/dome port combination please chip in here and we can work out the very best solution for anyone wanting to use this lens

[PeteAtkinson]

My (limited) understanding is that the dioptre strength should depend on the radius of the dome, not the distance of the dome from the lens. I have used +2.4 dioptre for 8" domes and +3.3 for 6" domes in the past - planoconvex lenses made specially, but now I use B&W +2 with a Nauticam 230mm dome on the Nikon 12-24. Sure you lose angle of view, but without it the corners are very soft. (Particularly compared with Alex Mustard's Zen 230 and 16-35 combination which looks stunning.)
The extension should place the centre of curvature of the dome at the front entrance pupil of the lens. Who will build the first adjustable length extension so we can get this right for each dome/lens combination?

[John Bantin]

What Pete said.

The front optical node of the lens should be at the radius of the curve. With zoom lenses, the node may move. Hence a problem.

I have a large number of domes inc. Subal, Nexus, Seacam and have found the sharpest to be the Sea & Sea (optical glass) port in conjunction with the DX Tokina 10-17. I was duplicating a trip to Truk so I put it on my Hugyfot housing with my 15mm FX lens as a test but got slightly less good results than with the original Hugyfot acrylic dome. The reason? Obviously the front node was not exactly in the right spot.

However, I have found that the best way to get the best results is to get optically clear water to shoot in.

Edited by John Bantin, 22 April 2012 - 11:16 PM.

[scuba1974]

What Pete said.

The front optical node of the lens should be at the radius of the curve. With zoom lenses, the node may move. Hence a problem.

I have a large number of domes inc. Subal, Nexus, Seacam and have found the sharpest to be the Sea & Sea (optical glass) port in conjunction with the DX Tokina 10-17. I was duplicating a trip to Truk so I put it on my Hugyfot housing with my 15mm FX lens as a test but got slightly less good results than with the original Hugyfot acrylic dome. The reason? Obviously the front node was not exactly in the right spot.

However, I have found that the best way to get the best results is to get optically clear water to shoot in.

I completely agree about the water! Interesting you you rate the sea & sea optical glass as one of the best domes, isn't this only about 6-7" diameter? I have read the large sea & sea fisheye dome is optically better than their glass offering. I do prefer glass domes though for durability. I did plan to get a zen 9" dome and an adapter made for my sea and sea rig but on arrival of the sea&sea fisheye dome I was quite impressed so it would be silly not to give it a good trial hence asking the questions here

[scuba1974]

My (limited) understanding is that the dioptre strength should depend on the radius of the dome, not the distance of the dome from the lens. I have used +2.4 dioptre for 8" domes and +3.3 for 6" domes in the past - planoconvex lenses made specially, but now I use B&W +2 with a Nauticam 230mm dome on the Nikon 12-24. Sure you lose angle of view, but without it the corners are very soft. (Particularly compared with Alex Mustard's Zen 230 and 16-35 combination which looks stunning.)
The extension should place the centre of curvature of the dome at the front entrance pupil of the lens. Who will build the first adjustable length extension so we can get this right for each dome/lens combination?

I agree about the radius for comparisons of say a 6" to an 8" dome if the distances are the identical but when a lens is closer to the dome optic then it uses a much smaller part of the dome so a much smaller radius

[PeteAtkinson]

No, the radius of the dome is the curvature of the dome. It's completely independent of how close the lens is to the dome. Sure a lens close to the dome sees a smaller area of the dome, but this doesn't alter the radius, which is fixed when the dome is manufactured.

[scuba1974]

No, the radius of the dome is the curvature of the dome. It's completely independent of how close the lens is to the dome. Sure a lens close to the dome sees a smaller area of the dome, but this doesn't alter the radius, which is fixed when the dome is manufactured.

Oh thanks, I really wasn't aware of that. In all my tests the closer the dome port is the further the new close focus distance is pushed out from the actual one. I thought this was the general rule. More tests I guess.....