139 replies to this topic

[Phil Rudin]

Regarding the issue of DX v. FX and lens quality, you may want to check out the latest lens review at DPReview.
It compares the latest 70 to 200 zoom from Nikon and has conclusions for both FX & DX formats with "full frame" under performing the cropped sensor. One would have to conclude that this is more than just a wide angle issue.

Phil Rudin

Edited by tropical1, 02 May 2008 - 11:53 AM.

[UWphotoNewbie]

Wow. I would have thought based on the rave reviews I've seen that the 70-200 would be the alpha and omega of lens design. This has been an education. It looks to me like it will take another 5 years and (and thousands of dollars) to shake this out through camera and lens updates even at the top end. The Ken Rockwell 60mm AFS review hints to the same issue wrt vigenetting as well.

When those of you in the know look into your crystal balls do you see a like for DX 5-10years down the road? Will there continue to be DX cameras at the prosumer level opr will it go the way of the APS? Until this discussion (and the two latest lens reviews) I thought this was a slam dunk. But perhaps there is room for both formats even at the pro level for some time to come.

[echeng]

Fair point, but not a FF vs DX issue. Rather a good housing vs poor housing issue. Alex

True. But at the moment, FF are only in high-end, pro bodies which tend to have 100%, large viewfinders. So they are correlated at the moment, at least, until lower-end bodies start getting full-frame.

[Phil Rudin]

Hi Eric,

To be sure the "pro" type 35 mm chip bodies can produce a large view finder and a high quality add on viewfinder like the Inon 45 and 180 degree finders help even more. However large finders can be designed for much smaller sensors like the 4:3 size.

The Olympus E-3 has a true 100% field of view and 1.15 X magnification. In DPReviews comparison to the Nikon D300 and Canon EOS 40D they said "it's obvious that Olympus has done a superb job of overcoming the limitations imposed by the smaller format- it's easily as big as bright (in fact the 4:3 format means it's actually slightly larger)," than the Nikon D300 and Canon 40D.

What you also get with the "pro" body is a battery pack that puts the lens port well off the bottom, no help at all for macro and no on board strobe which means no fiber optic cord capability. It seems that everything is a tradeoff.

Phil Rudin

Attached Images

• 
• 

Edited by tropical1, 03 May 2008 - 07:35 AM.

[Marc Furth]

I've been trying to complete my housing for my D3 for months now with little luck. I've made 5 different wide dome ports and all of them do not come close to the ports I've made for my D2x.

All of the ports still show soft corners to some degree. I have made ports for the 16 mm Nikkor,
15mm Sigma, 14mm Sigma, and 14-24 Nikkor. My favorite Lens is the 14-24 but is the worst
Lens in the group for soft corners.

My best success is with the 16mm Nikkor behind a mineral glass 8" dome. I also tried using 8.25" acrylic dome that I've had excellent results with on my D2x, but it to still shows soft corners with the D3.

I should say it is possible to get reasonable good corners at smaller f stops with some of the different combination but nothing under f8.

To me this defeats the great low light possibilities of the D3 which is why I bought it for in the first place. I only hope I can resolve this issue or I'll be looking for another camera in the near future.

I'd very much appreciate hearing from anyone who has a D3 in a housing and if their getting good results with any combination Lens and dome ?

Marc

[Phil Rudin]

Hi Marc,

On Tuesday May 13 at 7:30 PM RYAN CANON from Reef Photo Video and Zen Underwater will be a guest speaker at the South Florida Underwater Photography Society meeting in your area. He will be bringing some of the latest and greatest new equipment for display. Ryan has been working closely with Athena, SeaTool and others in the development of dome ports including those for 35 mm sensor size DSLR cameras. This is an issue I am sure he will be addressing at the meeting.

I have attached a link with directions to the meeting location and would invite you and any other Wetpixel members in the Miami area to join use for what I am sure will be an informative evening.

Phil Rudin

SFUPS President

http://sfups.org/

[Glasseye Snapper]

... snip ...
When one tries to make a list of advantages of full frame underwater it appears there are two: Higher resolution and quality, probably only discernible with large prints, and the larger viewfinder.
This has to be balanced with the greater wide angle flexibility inherent to DX. Am I correct, is there anything else that is inherently an advantage of full frame?

What about field of view, with the same lens & same distance your image dimensions are 1.5x as large. New FF sensors have about the same pixel size, and thus should have similar pixel quality. In other words, if two divers shoot the same scene with the same lens from the same distance, the FF shooter can get the identical result, including DOF, as the DXer by just cropping out the DX-sized central frame. However, assuming the image quality farther from the center is good then you have the option to crop less, not crop at all, or crop a different composition. All options the DX shooter does not have. Another way to look at this is that the FF shooter can shoot the same composition from closer up, a distinct benefit especially with poor visibility, or keeping the same distance, capture the entire fish while the DX image lacks the tail.

The worse DOF for FF that has been brought up is really comparing apples and oranges. If you go to life-size imaging FF and DX are the same, but if you want to let the same composition fill a FF you actually need a 1.5:1 magnification, compared to 1:1 for DX. DOF will be worse, not due to an inherent problem with the sensor, but because you actually went to higher magnification. Another way to think about this is that DOF is a property of the lens and lens settings, the type of sensor has nothing to do with it. So instead of saying FF is loosing DOF I think it is better to say FF is gaining FOV (just like some prefer balding to be considered as gaining face instead of loosing hair).

I'm not a WA aficionado but based on what James hinted at it seems to be a similar case. If you compare a 14mm lens on DX and FF, DX will have much sharper corners, but they just are not the same corners. For a fair comparison you would need to compare the same FOV images, so a 14mm DX lens against a 1.5x14=21mm lens on FF. Once you compare them at the same FOV, the FF has the ability to capture more detail because of the higher pixel count. Whether that is going to help you depends on whether your lens actually captures more detail than the DX format can handle, and whether your print media does justice to the detail in the image.

I think Alex said it best; cropping a FF image to "solve the problem" is a very expensive way to obtain DX images. He also said that you simply need to be more of a perfectionist in making your shots and selecting your optics to reap the benefits out of FF.

Bart

[james]

Well, that's not quite right Bart, but close. To very briefly summarize:

To get an identical image in almost every way from the same position, a FF shooter would use an 18mm lens at f11 and a DX shooter would use a 12mm lens at f8. That would yield an identical field of view, an identical DOF, and the same perspective.

For macro, a FF shooter would use a 150mm macro lens at f16 and a DX shooter would use a 100mm lens at f11. That would yield an identical field of view, and identical DOF.

Cheers
James

[Glasseye Snapper]

Hi James,

You are of course correct, you need to maintain the same FOV to get the same apparent image after printing the uncropped file, but that wasn't quite what I was getting at, at least not for the macro case which I'm more interested in.

For the case you describe, where you maintain the same FOV, FF is actually magnifying more then on a DX. For the DX, 1:1 magnification would map 24mm in the real world on the 24mm width of the sensor. To get the same FOV on FF you would map that 24mm onto the 36mm width of the sensor, or a 1.5:1 magnification. This is why you said the FF shooter would use the 150mm macro lens and then stop down to maintain the same DOF. FF still has a potential advantage in this situation because the image is digitized into more pixels and, if the image projected by the lens has more resolution than the DX sensor can handle, then the FF will give greater detail.

I was talking about maintaining the same magnification, so the FF and DX would both use the 100mm macro lens set to the same F-stop of, let's say F11. In that situation both systems would have the same DOF but the FF would have a 1.5x larger FOV. If the sensors have identical pixel size (apparently the Canon 1DS MkIII and 20D/30D do) then the FF shooter can crop to get the identical shot that the DX shooter had or the extra sensor real estate can be used to show a bigger chunk of the real world at 1:1 magnification. Again the FF sensor has the advantage in being able to image a larger object, at the same magnification, then the DX could.

To see the benefit you may have to print it at a larger size, but the point is that FF has captured either more detail in the first example or a larger area in the second example. Since macro shooting is about revealing detail, rather than maximizing the magnification factors, FF beats DX (at least on paper).

A lot of this has to do with semantics but at least I hope I got it right. My main point is that I do not think FF is at a disadvantage compared to DX for macro work if you compare them while maintaining equal magnification, which I think is a fair comparison. So Eric has more arguments to use a 1Ds MkIII in addition to the chick magnet factor.

Bart

[Phil Rudin]

James is in the ball park but I don't find this a very compelling argument for 35 mm sensors over smaller ones.

To me what is more relevant in U/W photography would be shooting at a base ISO, shutter speed and F-stop to compare results.

First let me say that I think the Nikon D3 and D300 are both outstanding cameras which can produce great results.

The D3 with a 14 mm lens has an angle of view of 114 degrees.

At ISO 100, f/5.6 & 1/60th for a proper exposure at 2.5 feet, DOF is from 1.52 to 7.1 feet for a total of 5.58 feet

The D300 with a 10.5 lens (non-fisheye, if one was made) would have an AOV of 114 degrees and at ISO100, F/5.6 & 1/60th at 2.5 feet would have a DOF of 1.09 to 10.0 ft. or a total DOF of 9.51 ft.

The Olympus E-3 with the 7 to 14 zoom at 7 mm has an AOV of 114 degrees and at ISO 100, F/5.6 & 1/60th at 2.5 feet would have a DOF of 1.08 ft. to infinity.

With macro ISO 100 at 1/250th and F/22 is common at say ten inches.

D3 with a 100 mm macro lens AOV 24 degrees (105 mm 23 degrees) DOF 9.9 to 10.1 total 0.21 inches.

D300 with 75 mm macro (if one was made) AOV 24 degrees, DOF 9.86 to 10.1 total DOF 0.29 inches.

E-3 with 50 mm marco AOV 24 degrees, DOF 9.73 to 10.3 total DOF 0.55 inches.

For macro keep in mind that at life size, 1:1 the full size image would be 36 mm X 24 mm for the D3
1:1 for the D300 is an image/subject 23.6 x 15.8 mm and for the E-3 the 1:1/subject size would be 18 X 13.5 mm.

Note, DOF Master was used for these results and compared to other DOF calculators, http://www.dofmaster.com/dofjs.html

If I wanted very shallow DOF for underwater macro and W/A or to shoot at ISO 400 rather than 100 the larger sensor would make more sense to me. But greater DOF is more often than not a good thing for underwater photographers and then their is still the issue of soft corners that show-up the larger the sensor gets.

Images #1 life size for 35 sensor, #2 life size for 4:3 sensor.

Phil Rudin

Attached Images

• 
• 

Edited by tropical1, 04 May 2008 - 03:44 PM.

[james]

Phil: what you didn't mention is that at F22 you will have a lot of diffraction at f22 and the results from the D300 or E3 at F22 will be poor.

Cheers
James

[davephdv]

I recently ordered a Subal D300. Asked about using the new 14-24 lens. Options were as such. 1.) order housing with new larger port size. 2. use regular port size and get adaptor to use with new larger dome port. 3. send current FE2 domeport to Subal to be retrofitted with a larger diameter port and use an adaptor with smaller size housing port.

Clearly you cannot add an adaptor to this lens. Is this the main concern with using this lens (the 14-24)? Assuming that the new Subal port and extension ring are optimized for this lens.

Talking full framed cameras. Is not the Cannon 5D a full frame camera in a consumer body? Seems Nikon ought to be able to do this with the D300.

It won't fit in the D300 housing as I just ordered one.

[herbko]

Phil: what you didn't mention is that at F22 you will have a lot of diffraction at f22 and the results from the D300 or E3 at F22 will be poor.

Cheers
James

The dpreview lens tests show this clearly. The sharpness of the Olympus lenses drops rapidly at F/16. Diffraction limit their usefully range ( if you want good resolution ) to around F/11. To properly compare 35mm lenses and 4/3 lenses you have to factor in the 2 stops in the diffraction limited resolution.

[John Bantin]

I had completely forgotten the anguish I suffered when going on a dive trip after being persuaded to buy a new housing for my 35mm film camera with a compact dome port. All the pictures were blurred at the corners. I then had to get the huge fish-eye dome port to get pictures sharp across the frame with the 20mm lens. I moved over to DX digital and later went to a smaller optical glass port and have had no blurred edges problems. Perhaps if I went to full frame digital I would find myself reliving the experience!

[dhaas]

Whoa, has this discussion given me a headache

While I'm sure all the techno-babble on full frame sensors and the fact both Canon and now Nikon have done great engineering I think so much of this is irrelevant to 90% of UW shooters.

I've shot a Canon 5D with decent lenses and yes, the files are nice. But as John states (and he and I seem to be on the same page) anything more may not be seen either on your computer screen (OK at 200 -300%?) or in print magazines and likely not even up to 24" X 30" prints versus a decently exposed, good glass image file from a APS-C sensor dSLR.

I, too, will now have to read how early adopters of the D3 and spectacular surface performing 14-24mm Nikon lens will STILL get soft corners due to rectilinear lenses not matching dome port's apparent image

On John's comment about a fisheye, APS-C sensor dSLR and small dome port......I think even the lowly Tokina 10-17mm Fisheye is sharpest behind my Ikelite 6" port at about 12-13mm, and certainly at 17mm. Likely it is a closer optical match and the real world physical difference in shooting at 10mm versus 12 or 13mm is less than 6 inches.....No big deal for framing or color from your strobe.

Hell, I've even taken to leaving the zoom ring off and pre-setting my Tokina at either 12mm or 17mm and going down for a dive concentrating on that particular focal length.

I would even go so far as to surmise manufacturers will continue to evolve noise suppression and improved dynamic range on APS-C sensors more than full frame sensor development for the smaller market it is.

If you have the need for full frame mega-pixels, biggest viewfinder as Eric stated and other advantages the more expensive camera bodies give you, then go for it....I've held them, played with them and like most camera nuts are all gaga for them......

Just not in my budget and likely many other UW shooters.

Donning my flame resistant suit NOW

YMMV

dhaas

Edited by dhaas, 05 May 2008 - 05:25 AM.

[davichin]

The more I read about FF the less I feel about getting one and the more I am looking to upgrade to a D300 and not wait for the next FF or D400...

[ChrisJ]

This discussion is making me regret selling my entire Hugyfot D200 Setup.

[craig]

The thing that surprises me is that people are talking like these are new issues.

Wide angle issues are not full frame versus APS-C at all; they are underwater optics issues. We need manufacturers to focus on delivering higher performing port systems for our increasingly high performance cameras. The 14-24 is a new lens and is exceptionally wide for a rectilinear zoom. It should surprise no one that existing domes struggle with it but the lens and its port systems are young.

There have been macro depth of field discussions here for years. Smaller format sensors do not have a depth of field advantage over full frame in macro applications (in fact they are worse). Every explanation that claims otherwise fails to consider equivalent f-stops and diffraction effects. If all you shoot is macro and you want good depth of field you should be using large sensors, not small ones. Herb is exactly right, there's a two stop difference in comparing f-stops between 4/3s and full frame 35mm with APS-C 1 stop between each.

The downside of macro photography and full frame is that you have to use longer focal lengths and you lose the apparent magnification advantage that cropped sensors provide. A full frame macro shooter has to work harder and use a bigger rig to shoot smaller subjects. For the effort, he can obtain higher resolution images. If a full frame shooter is resorting to 1.4x and 2x teleconverters to get the same shots that others are getting without teleconverters, then he's probably not realizing any advantage with his sensor IMO.

I believe, and many others have expressed the similar opinions, that most people are best served with APS-C/DX cameras and that the D300 is the underwater camera of choice right now on the Nikon side. It has the best flexibility on the wide side and is easier to shoot on the macro side. The D3 may be a superior camera but it doesn't deliver any compelling benefit that justifies its additional compromises over the D300. When the D3x comes out, only certain shooters may be better served by that camera. Canon shooters have faced that decision for some time now.

[brandoncole]

[quote name='craig' date='May 5 2008, 12:11 PM' post='168602']
>>>>The thing that surprises me is that people are talking like these are new issues.

>>>>Wide angle issues are not full frame versus APS-C at all; they are underwater optics issues. We need manufacturers to focus on delivering higher performing port systems for our increasingly high performance cameras.


BRAVO. Yes indeed, this is the crux of the problem.

Brandon

[echeng]

Brandon and Craig -

Should the manufacturers NOT address the crux of the problem, practical discussion of the issue in terms of products that are currently available certainly has merit. Let's say you are shopping for a wide-angle setup right now? If you have to choose FF vs. DX, learning about the practical issues that photographers have discovered by shooting with different formats, lenses, and ports will help a lot.

This is not simply a theoretical discussion, which would certainly also be useful.