WA distortion with 5D/Ikelite 8"

I did my first dives with the 5D today, using the Ikelite housing. After flooding the sync cable twice (I found you need to press in, tighten down, and repeat that over and over until it's in), I did some wide-angle shots with my 17-40mm f/4 L lens. Looking back at the shots, the wide-angle distortion seems pretty bad. It's not just at the corners, but almost the entire left and right quarters of the frame. Am I doing something wrong? I expect a little of this distortion, but this is way more than I expected.

f/6.3, 1/250th, 17mm

Another:

f/5, 1/250th, 40mm

Jeff,

I've been using the same setup & the distortion on my pics at 17mm don't look so bad. However, I was using a +4 diopter and I think most of my shots were at f8 or above. Here's an example:

You could try using a diopter- unfortunately I can be sure as I havn't tried mine without.

Fergus

I did my first dives with the 5D today, using the Ikelite housing. After flooding the sync cable twice (I found you need to press in, tighten down, and repeat that over and over until it's in), I did some wide-angle shots with my 17-40mm f/4 L lens. Looking back at the shots, the wide-angle distortion seems pretty bad. It's not just at the corners, but almost the entire left and right quarters of the frame. Am I doing something wrong? I expect a little of this distortion, but this is way more than I expected.

 

f/6.3, 1/250th, 17mm

Another:

 

f/5, 1/250th, 40mm

Hi,

The 17-40mm also shows distortion on my Aquatica housing using Extension #18457, not so much as on ike dome, but still a problem for me. I found less distortion between 24mm thru 40mm, then I did not use focal distance less them in my trip to RedSea. I will try another extension rings soon to get better results with 17mm.

1/200s f/8.0 at 25.0mm

1/125s f/8.0 at 25.0mm

1/250s f/8.0 at 40.0mm

Regards. Fabio

I had an issue with a 12-24 having soft edges, some say a diopter will help. I found that the correct extention ring did it for me, having the lens in the right place behind the dome is very important for sharp corners, lets see what others have to say who might have corrected the same problem, oh and now my 12-24 is sharp as a tack right to the edge.

Just for note: I'm using a diopter +4 with 17-40mm f/4.

Regards. Fabio

Somewhere on this site there's a technical article on the physics of dome ports. I'll may need to have my wife calculate the optimum extension tube length for the dome port. Another thing, I can't remember if there's a pin in the mounting plate to keep the 5D straight. I don't think there is -- just a screw to hold it in the tripod mount. This might explain things if it's skewed slightly to one side. Perhaps there's a way to install a pin the way my Aquatica housing had for the 300D. There was a small pin about 2" away from the tripod mount that fitted to a small hole on the bottom of my 300D.

A few pics of the better pics, for those interested:

Santa Cruz Island

Here's the dome port article. I have some research to do.

http://wetpixel.com/i.php/full/proper-dome...-and-placement/

A few pics of the better pics, for those interested:

 

Santa Cruz Island

I see you were out with John Gonda yesterday. You end up going to the backside of Santa Cruz? Looks like your vis was slightly better...

I was there on the Spectre on Sunday, and I have a picture of that odd grey/white/yellowish fish as well, but no clue on the species.

Hey Jeff -

 

I see you were out with John Gonda yesterday. You end up going to the backside of Santa Cruz? Looks like your vis was slightly better...

 

I was there on the Spectre on Sunday, and I have a picture of that odd grey/white/yellowish fish as well, but no clue on the species.

Hi Matt - yes, John was there, as well as a few other pretty serious photographers and videographers. It was a nice day. I've got a post over on Scubaboard trying to ID that fish. TJ votes for "mutant perch." Whatever it was, it was pretty darned ugly.

@synthetic

It's a question of depth of field. You have focused on the diver. Everything else is out of focus - no suprise 'cause there are closer things in the frame and you have shot with f4. The effect is even stronger than topside because dome glasses bend the image field. So you need more DOF respectively stop down. f4 isn't a good aperture choice here.

Julian

Fabio

Your pics look good to me but for wider shots you need more extension. I also have a 5D and 17/40 and I emailed Blake and asked if my two existing rings totalling 53mm (sealing surface to sealing surface) were OK. He confirmed they would be and that probably a few mms less would be too. I have used it with a +2 and also without a dioptre and have had good results. I couldn't see any real difference between the two although I haven't been wider than 19mm.

Jim

Sorry Jim,

I used in my first post images using focal distance not so wide. I did not much photos using focal distance less than 24mm due the problem. I attached a image using 17mm and some crops of corners.

1/80 f/8 at 17mm

Full image-------crop #1------crop #2

Regards. Fabio

It's a question of depth of field. You have focused on the diver. Everything else is out of focus - no suprise 'cause there are closer things in the frame and you have shot with f4. The effect is even stronger than topside because dome glasses bend the image field. So you need more DOF respectively stop down. f4 isn't a good aperture choice here.

I'm not sure that's it. Why is the head of the fish softer than the tail? Why is the red gorgonian in double vision? Why is the Henderson logo stretched? You can see soft focus in the background of the shot in the center that looks much better than the edges of those two shots in particular.

I am running into the same issue with blurred edges with the 5D with the 16-35. It is not a DOF issue... even at F4 the DOF in that image should be around 10ft, which should be more than enough to have held detail in most of the blurred areas. I posted a sample in this thread.

I have tested with +2 and +4 diopters and while helping, they don't fix the problem.

Ferg's results seem suprisingly good to me. The right edge of his image is very close to the camera. Perhaps the problems starts when the subject is further away? Ferg, do you have any other samples you could post?

Ike, do you have any suggestions? Getting good UWA results with the 5D is a problem that has me stumped. I'm either doing something wrong or there is some other variable causing the issue.

HI

ok i'll share some of my testing information with you with regards to this problem.

Firstly i'm no ikelite user, Sea & Sea is my choice of tool.

test setup as followed

D70 - Inon Z240's - tokina 12-24mm NO DIOPTRE

I carried out the following shots after returned home from a Sipadan trip with my new 8" dome

i'd been using the tokina 12-24 behind this dome without extension ring and using a +2.. each day i steadly got more and more frustrated with it images we very much like your's soft on the corners and at time's simply not even in focus.

after a few days i then began to seek a solution to this problem.

Firstly, I'd never and was never a dioptre user.. HATE THE THINGS with a vengence.. but i was prepared to give them another try given all the positive reading within these forums. I couldn't see any difference in my photos at all with the +2 added.

It got to me , then i started to think about dome placement.. The next day i added a 40mm extension ring onto my fisheye dome port 8"...

the results we much much improved, but still not the kick in the ass i was expecting but a great improvement none the less.

So upon my return i carried out a few tests against my Compact Dome port, just to see if there was a quality difference in the two domes, as i had always been told that the bigger the dome the better.

So i went to a local flooded quarry here in the Uk, water viz is about 3m and its 8c when these shots were done.

setup was again, tokina 12-24 this time tested with the 40 extension ring in place and no dioptre first shot was done with the 8" dome, and the crop is from the edges

second shot was the subject aircraft different diver model, and Compact Dome with 40mm extension

sorry about the size of the crops they are 100% however, had to keep it small to fit it into one post, but i hope you can see that there is not a lot of difference between the two at all, you can see that they are quite sharp with detail and not stretched and pulled out of focus , However the lesson learned was that correct dome placement makes all the difference. even with the large fisheye 8" dome.

just because the lens fits behind the dome, dosn't mean that its in the correct place at all.

I think i'm inclined to say try your dome placement and hopefully that will help greatly.

well it did in my case, now i'm a happy chap... and have actually gone back to my compact dome as there is just not enough quality difference to justfy me lugging it around the globe, i'm more than happy with my little dome and a 40mm extension.

regards

Craig

Hi Syn,

You may want to look at Stephen Frinks tests of this lens and the 16 to 35 mm on a full frame sensor out of the water. It is clear that these lenses designed for film use fall short in some areas when used with full frame sensors.

Adding a dome port will increase these problems.

http://www.seacamusa.com/canoncorners.shtml

Phil Rudin

@TheQ

Interesting the DOF should be arout 10ft. I hope you are aware what a dome glass does underwater in terms of optics? it's not just the opposite of a dioptre, it BEND YOUR IMAGE FIELD, resulting in out of focus corners (curved image field), even if the corners would be sharp under topside conditions. No chance to simply use topside DOF values for underwater with dome glass (If you are interested how focus distances behave underwater compared to topside, I have written an article quiet a while ago: http://wetpixel.com/i.php/full/dome-theory/ )

I can't comment on dome placement / extension ring / etc. as there are no manufacture values available. In addition, dioptres can help to reduce mentioned corner issues.

Julian

As I've said, I'll accept some softness in the corners, even though this is a $650 lens. But I think the degree of softness I'm seeing is a bit extreme.

I found a webpage that lists the 17-40mm lens as having an entrance pupil 71mm from the connecting flange at its widest setting. (Panorama photographers need to know this sort of thing to stich together their photos.) I may be able to measure the rest. Stay tuned.

After having the same problems with the 12-24 I firmly believe that you have to have the correct port and the correct extension ring to line the port up in the correct place, this is crucial to having sharp edges, it sounds like the only way you'll get the best results is trial and error.

OK, I might have this figured out.

All dimensions in millimeters, for the Canon 5D, 17-40mm f/4 L lens, Ikelite housing with 5510.81 8" dome port:

LF= 44

FP= 71

r= 101.6

fe= 30 right now, but this is what we want to solve for (x)

Ff= 28

The only dimension I don't have is d. I suppose I could figure it out if I knew the width of the plastic used for the dome port (2r+2w). Let's assume it's 3mm, so d = 2*101.6 + 2x3= 209.2. I assume that ed is 1/2 of this, or 104.6. So this is no longer scientific, but is still fun.

OK...

The Camera flange to dome surface distance is ( Fd = Ff + fe +ed )

28 + 30 + 104.6 = 162.6

The camera flange-pupil distance may be computed using the film plane datum is ( FP = LP- LF ).

I found this listed on the web as 71, so I assume it to be correct.

The Pupil to dome surface is ( Pd = Fd - FP )

162.6 - 71 = 91.6, not the ~104.6 I was expecting to see.

Using this math, the optimum length of the extension ring, fe, would be around 43mm instead of 30mm. Does anyone have the 5510.24, and if so could you measure that? Ikelite says it's 3/4" longer than the one supplied with the 5510.81, which would put it at 49mm. Not exactly right, but closer than what I have right now.

Unless my math is way off, which is likely. -jl