Tokina 10-17 with 1.4TC on FX

A couple of weeks ago, almost tongue in cheek, I posted the 'Tokina 14-25 Fisheye zoom' with some pool tests as a consideration for full frame DSLR use.

http://wetpixel.com/forums/index.php?showtopic=28656

After using the Tokina 10-17 with 1.4 Kenko TC on my D700 extensively last week on my trip I can happily say this lens is now in my bag as my favorite wide angle lens on full frame underwater as it was on my D200, and I have now happily left DX behind.

So I am now posting quite seriously that I think this lens is a great option, and I am surprised none of the Canon full frame users have tried this lens combination underwater over the last few years.

The D700 has changed the way I shoot underwater, so the one stop loss due to the TC is simply not a problem, as I routinely shoot now between 400 and 1600 ISO depending on depth. I will write more about this another time.

I found the autofocus worked just fine, and all the Tokina close focus and fisheye depth of focus advantages were there just as they were with my D200 and using it on the D700 was a no brainer.

Here are a few uncropped images. The stated focal lengths are from EXIF data, so are the lens barrel setting and should be converted by x 1.4.

(The color is a little off only because of AdobeRGB export issues directly from Aperture to Picassa)

To my eye sharpness is clearly not an issue, and compares favorably to any of the full frame wide angle options I have seen for Canon or Nikon underwater. Going out on a limb a little, I'd say corner sharpness is better than the rectilinears. (Particularly evident in the second shot)

I also used the 17-35 with +4 dioptre, and was very pleased with this as well, and will post some of those later.

I was having so much fun with the Tokina, that I did not get around to trying my 24-70 underwater as I had planned.

First two at 11.5mm

ISO 400 f16, 1/125 sec

ISO 400 f11 1/125sec

Next at 13mm

ISO 400 f8 1/60 sec

Finally at 17mm

ISO 400 f11 1/125sec

Hey welcome back, missed seeing you around here, figured you must have gone diving since you were not in the HDR thread

Awesome shots, the second one I think is my favorite, very cool.

Hey - welcome back ... Looks like you got up to some fun!

Pretty cool that you figured this out. Any thoughts I had about buying a DX body for underwater use are now gone.

Did you happen take any shots at 10mm? I'm just wondering if 11.5 is the the lowest-usable FL.

Also, did you encounter any significant

I recently got a Sigma 15mm FE for my 30D for some wide angle shooting I did. I opted to avoid the 10-17 because I also have a 5D2 and know I'll eventually upgrade my housing and didn't want to lose out on the FF options... now I'm wondering if I'm missing out!

Did you happen take any shots at 10mm? I'm just wondering if 11.5 is the the lowest-usable FL.

 

Also, did you encounter any significant

 

I recently got a Sigma 15mm FE for my 30D for some wide angle shooting I did. I opted to avoid the 10-17 because I also have a 5D2 and know I'll eventually upgrade my housing and didn't want to lose out on the FF options... now I'm wondering if I'm missing out!

Yup, I did. (Not on purpose) Still works fine, just means some cropping is required.

Nothing a little cropping and cloning cannot fix.

Yup, I did. (Not on purpose) Still works fine, just means some cropping is required.

 

Nothing a little cropping and cloning cannot fix.

Good to know - thanks for posting!!

has anybody tried this with a canon FF and the canon 1.4 TC?

has anybody tried this with a canon FF and the canon 1.4 TC?

I can't see any reason why results would be any different. Matt Segal posted some topside test shots on the previous thread I mentioned.

Well done loftus. Thanks for sharing. I am always amazed what mischief wetpixelers can get into I know the 14-24 is just a tease, but what is this DX lens with a 1.4 TC on an FX sensor? The 1.4 TC narrows the Field of View (FoV) while DX to FX sensor widens the FoV. Are we back to aprox. 180 to 100 deg? A 14-24 more or less rect. would be 114 to 84 deg. on FX. Is it something like a 9-16 mm? This is way too much for my poor pea brain.

Bob

Well done loftus. Thanks for sharing. I am always amazed what mischief wetpixelers can get into I know the 14-24 is just a tease, but what is this DX lens with a 1.4 TC on an FX sensor? The 1.4 TC narrows the Field of View (FoV) while DX to FX sensor widens the FoV. Are we back to aprox. 180 to 100 deg? A 14-24 more or less rect. would be 114 to 84 deg. on FX. Is it something like a 9-16 mm? This is way too much for my poor pea brain.

Bob

I think this is pretty close to the 180 degree FOV at the wide end as with a 15 or 16mm FX fisheye, as is demonstrated by the visibility of the dome shade at the wide end. I think it behaves very similar to the 10-17 without TC does on a DX camera. Probably a 1.5 TC would make them identical at least for Nikon.

Nice stuff, Jeff. Thanks for sharing your findings.

Mark

I would like to add a couple of new observations with regard to port extensions and vignetting at the 10mm setting.

On my recent trip I used the FE2 Subal Domeport and 30mm extension and the following vignetting / corner obstruction resulted. It should be pointed out that with this setup the Tokina and 1.4TC combination protrudes just beyond the port opening. My initial thought was that the lines of vignetting down the sides (right more than left) are 'in-camera' the soft vignetting at the top is due to the Tokina lens shade which is longer on top and bottom than on the sides, and the corner vignetting is due to the dome port shade.

However removing the dome port shade, does not significantly improve the corner vignetting / obstruction

On the other hand when I change to a 20mm extension there is less corner vignetting or obstruction, even with the domeport shade on

So I am not sure which extension is preferable though I have a sense that the corners may be a little softer with the 20mm extension possibly due to the lens surface being closer to the dome.

I was surprised to find that the corner vignetting got better when I moved the lens closer to the dome (10mm extension), thinking it was the dome port shade, but did not get better when I moved the lens back(20mm extension), even with the dome shade removed.

Any comments, or thoughts on the best extension to use would be appreciated.

The Tokina has an internal baffle - 10mm is just not a usable focal length with the 1.4x teleconverter on full-frame. A 1.5x teleconverter may just get you there, though...

Ummm maybe I missed a comment but DX lenses on FX bodies cut the MP count to like 5 something no? Although the pixels are high quality you are not getting the 12.1MP with a DX lens. Not sure this combination is a smart move.

Ummm maybe I missed a comment but DX lenses on FX bodies cut the MP count to like 5 something no? Although the pixels are high quality you are not getting the 12.1MP with a DX lens. Not sure this combination is a smart move.

In this case, the DX lens is used in FX mode and with the TC, the image covers the whole sensor ... that's the point.

One could shoot the 10-17 FE straight up in DX mode as a 5MP DX setup, but with the TC you can shoot in FX mode.

Oh, ok, the d700 does not realize it is a dx lens without the contacts touching as the teleconverter is in place. Do you get all metering modes and TTL also?

Excellent to see your Dr Franken-lens working Jeff. I think it is a jolly good solution for this type of shooting. The shots show that and it looks plenty sharp enough.

The main downsides I see are that small loss in quality from using the teleconverter would probably mean that a D300 and 10-17mm would give you better image quality. Of course, you don't own a D300, so it is not really a relevant point.

I would also be concerned about the corner CA - an area where the Tokina is already a bit naughty and I would only expect it to get worse working through a TC. I would think that typical reef wide angle shots, where the corners might contain contrasty fine detail (silhouetted small fish, edges of seafans etc) that the purple fringing could be an issue. Of course for big animals and pool images - the subjects that you shoot - this is not a significant issue. I also think that the optical flaws would become more serious on higher megapixel full frame cameras, like the D3X, 5DMk2 and 1DsMk3, so this might not be a preferred solution for all shooters.

The reason that your corner cut off got less when you moved the lens forward relative to the dome (10mm extension) is because the closer a lens is to a dome the more it acts like a flat port - narrowing the angle of coverage of the lens and reducing cut off from the shade in the corners.

Oh, ok, the d700 does not realize it is a dx lens without the contacts touching as the teleconverter is in place. Do you get all metering modes and TTL also?

You can tell the FX Nikons to treat DX lenses as FX lenses and not to default to DX. I think that the AF will be considerably slower than a straight F2.8 fisheye - since the Tokina is 3.5-4.5mm lens and you will loose a stop for the TC - so AF will be working through a F5-F6 lens (approx) - which might become significant in lower light conditions. I'd imagine that TTL would work OK - but I don't see TTL as an ideal lighting solution for fisheye shots so I see this as less of an issue.

Alex

The main downsides I see are that small loss in quality from using the teleconverter would probably mean that a D300 and 10-17mm would give you better image quality. Of course, you don't own a D300, so it is not really a relevant point.

I would also be concerned about the corner CA - an area where the Tokina is already a bit naughty and I would only expect it to get worse working through a TC. I would think that typical reef wide angle shots, where the corners might contain contrasty fine detail (silhouetted small fish, edges of seafans etc) that the purple fringing could be an issue. Of course for big animals and pool images - the subjects that you shoot - this is not a significant issue. I also think that the optical flaws would become more serious on higher megapixel full frame cameras, like the D3X, 5DMk2 and 1DsMk3, so this might not be a preferred solution for all shooters.

 

The reason that your corner cut off got less when you moved the lens forward relative to the dome (10mm extension) is because the closer a lens is to a dome the more it acts like a flat port - narrowing the angle of coverage of the lens and reducing cut off from the shade in the corners.

No question this is not an ideal setup for all the reasons you mention, we would just hope Tokina would make a 15-24 Fisheye zoom for FX, which I guess is pretty unlikely.

I think for the type of stuff I have shown, close-up shark action ala Abernathy, and the whaleshark, this lens performed admirably and at this point in time I would probably choose this lens for these situations over all others. I am hard pressed to see any loss of quality, at least compared to my D200 and at routine magnifications. I must admit I was surprised it performed as well as it did, as I was sure that it would not work well otherwise folks would have been using it already. I had no issues with autofocus, the lens seemed to perform as well (except for the vignettiing which is easily overcome) as it does on DX, and the big thing it retained was the close-focus ability which is unsurpassed by any rectilinear. I own the 14-24 and 17-35 and 16. For pool work, more distant big animals, general reef work etc, I would use my 17-35 with +4, possibly my 16mm for the pool (which I used on DX). Maybe, if I can be convinced that the 14-24 can work predictably I will use it at some point. I think pool tests are helpful, but I'm not convinced they translate completely into more dynamic real life situations, particularly in the corners. We can't escape the fact that fisheye lenses will theoretically always be a better match than any rectilinear with present dome port designs.

I think the question to be asked at this point, is whether any of the rectilinears available for Canon or Nikon on FX, could have done a better job with the type of images I have shown in terms of IQ and corner to corner sharpness. Of course the FE distortion may be something undesirable, but again, rack it out to 17mm and most of that is gone as well. I see no reason to use my 16mm again for diving, except maybe low light siuations.

I think the Tokina is one of those lenses that when you have it on for something like that chance whaleshark encounter, you're pretty happy you did.

The Tokina has an internal baffle - 10mm is just not a usable focal length with the 1.4x teleconverter on full-frame. A 1.5x teleconverter may just get you there, though...

I think what puzzled me is that I thought the corner obstructions, which are sharper than the more diffuse peripheral shading, were from the dome port shade, and they were not as removing the domeport shade did not help. I know it's not possible to get rid of all the vignetting, I'm just trying to work out if any of it is from the external shades of the lens or the dome port. I thought of trying the 1.5 Kenko, but it does not get very good reviews when compared to the 1.4

I think what puzzled me is that I thought the corner obstructions, which are sharper than the more diffuse peripheral shading, were from the dome port shade, and they were not as removing the domeport shade did not help. I know it's not possible to get rid of all the vignetting, I'm just trying to work out if any of it is from the external shades of the lens or the dome port. I thought of trying the 1.5 Kenko, but it does not get very good reviews when compared to the 1.4

I assume the 1.7 would solve the problem at the cost of some FOV, right?

1.7x + 10mm = 17.0mm

1.4x + 11mm = 15.4mm

Is that the right way to think about it?

How does one estimate or calculate what the effective FOV will be for a given lens/TC combination and focal length?