Fish eyes compared: Tokina versus Panasonic (test 2)

This was a second test in which I compared two fish eye lenses 'on land', My victim now was a fruit bowl. I compared the Tokina 10-17 (on NIkon D7000, crop factor 1.5 relative to full frame sensor) and the Panasonic 8mm (on Olympus EPL5, crop factor 2.0). Same shutter speed and aperture (1/50, f4.5) no flash, ISO 250. Single focusing point Tokina was zoomed in at 12mm to match the size of the shots of the Panasonic. Taken with hand. In this test the Pana crop looked much sharper than the Tok crop. Crops were 100%

The last two 'duck' shots were taken with: 1/50 F7.1 ISO 400 at 10 cm distance, no flash

Hi Albert,

Just curious, were you on a tripod or hand held?

Cheers,

Steve

Shooting through a window, comparisons are invidious...

Hi Albert,

Just curious, were you on a tripod or hand held?

 

Cheers,

Steve

Hi Steve..this was just done hand held. Or with camera stabilized on table.

Hi Albert,

Just curious, were you on a tripod or hand held?

 

Cheers,

Steve

All hand held or with camera on table ...

Shooting through a window, comparisons are invidious...

Yes.. I skipped these shots. I must have thought the window would behave like a flat port!

Since you are comparing a zoom against a prime, I might ask if you have AF tuned the Tokina 10-17 to be it's sharpest at 12mm? Also, since your photos are JPEGS, did you process them through an external tool or is that straight from the camera? If straight from the cameras, there might be some differences in how the internal JPEG sharpening / noise reduction is done. Probably would be better to compare RAW files for both.

I'm a bit surprised your 10-17 looks so soft - mine is very sharp and even severe center crops retain detail.

OH yeah and one more thing. Since you are comparing a 2.0 crop factor sensor against a 1.5 crop factor sensor, you need to "equalize" the aperture.

A 2.0 crop 8mm F4.5 would be equivalent to 11mm at about F5.5 on a 1.5 crop sensor which could account for the softness on the 10-17. It might still be softer, I don't know; I know the 8mm Panny is very good - but that would be a better test in my mind.

OH yeah and one more thing. Since you are comparing a 2.0 crop factor sensor against a 1.5 crop factor sensor, you need to "equalize" the aperture.

 

A 2.0 crop 8mm F4.5 would be equivalent to 11mm at about F5.5 on a 1.5 crop sensor which could account for the softness on the 10-17. It might still be softer, I don't know; I know the 8mm Panny is very good - but that would be a better test in my mind.

Comparing shots of the camera with the smaller sensor (the Oly) with those of the camera with the larger sensor (the Nikon ) means that you have to increase the focal length of the camera with the larger sensor.. which is what I did by zooming in the Tok a bit... from 10mm to 12mm, .I was also surprised by the difference in sharpness. Like the Tok/Nikon was not focused correctly..must check this out by using other autofocus settings..

Hi Albert

Yeah, I understand why you went to 12mm, but 11mm is closer. 11*1.5 = 16.5. 8*2 = 16. 12mm is effectively 18mm full frame and you can see the focal length looks different between the two lenses in your photos. I'm a picky devil ain't I?

But the difference in focal length is probably not the biggest factor. When comparing between two different crop sensors, you must adjust the aperture. If you don't the field of view and thus the focus will be different.

From: http://www.cambridgeincolour.com/tutorials/digital-camera-sensor-size.htm

"As sensor size increases, the depth of field will decrease for a given aperture (when filling the frame with a subject of the same size and distance)."

By using the same aperture on both lenses, you are allowing the 8mm Panny to have a wider field of view and that likely will improve sharpness of the image. Also, since you are using JPEG, I can see in the first photo a bit of a "halo" on the edge of the banana which to me means it's slightly over sharpened. Did it come straight out of the camera like that?

Anyway, I don't mean to harsh on your methodology, I'm just being a pain because I'm interested in your results I own both a D7000 and an Olympus OMD, but I don't have the Panasonic 8mm lens yet. I will say on my Tokina 10-17 the sweet spot seems to be right at F8 and 10mm, but that's because I have AF tuned it to be sharpest at 10mm. Zoom lenses are always a bit of a compromise.

Cheers!

JP

 

Hi Albert

Yeah, I understand why you went to 12mm, but 11mm is closer. 11*1.5 = 16.5. 8*2 = 16. 12mm is effectively 18mm full frame and you can see the focal length looks different between the two lenses in your photos. I'm a picky devil ain't I?

 

But the difference in focal length is probably not the biggest factor. When comparing between two different crop sensors, you must adjust the aperture. If you don't the field of view and thus the focus will be different.

 

From: http://www.cambridgeincolour.com/tutorials/digital-camera-sensor-size.htm

 

"As sensor size increases, the depth of field will decrease for a given aperture (when filling the frame with a subject of the same size and distance)."

 

By using the same aperture on both lenses, you are allowing the 8mm Panny to have a wider field of view and that likely will improve sharpness of the image. Also, since you are using JPEG, I can see in the first photo a bit of a "halo" on the edge of the banana which to me means it's slightly over sharpened. Did it come straight out of the camera like that?

 

Anyway, I don't mean to harsh on your methodology, I'm just being a pain because I'm interested in your results I own both a D7000 and an Olympus OMD, but I don't have the Panasonic 8mm lens yet. I will say on my Tokina 10-17 the sweet spot seems to be right at F8 and 10mm, but that's because I have AF tuned it to be sharpest at 10mm. Zoom lenses are always a bit of a compromise.

 

Cheers!

JP

Thats Ok, no pain at al! When I set Tokina to 10 mm, it looks like the images have the same field sizes as those taken with the Panasonic 8mm lense,This means that the 8 mm Panasonic lense on the small sensor 'behaves' like a 10 mm lense on the larger sensor. If you divide the diagonal lengths of both sensors (that is: 28 mm/22 mm) then the crop-factor of Oly relative to Nikon is around 1.3. So, 8mm multiplied with 1.3. would be around 10 mm. Does this makes sense?

Then there is the problem of Tok crops looking unsharp. That also makes me unhappy. I suspect that autofocus of the D7000/Tok is not working appropriately under low light conditions. I tried some more shots with other settings and artificial light. Now with the Tok set at 10mm and apertures F8, ISO 400. But the difference remains visible.

I think you might have something else going on with your Tok 10-17 -- it just doesn't look sharp in any part of the image. If it was just a Fine Tuning issue, it would be back or front focusing, but it just looks very "soft" overall.

I think you might have something else going on with your Tok 10-17 -- it just doesn't look sharp in any part of the image. If it was just a Fine Tuning issue, it would be back or front focusing, but it just looks very "soft" overall.

Yes tnks, I noticed that too. Could it also be a camera problem..like a wrong AF setting, didnt try the fine tuning yet.

Today we had sunshine! So I tried some more tests, with Tok and Pana set at 1/30 F8 ISO400, 20 cm distance. Tok now at 10 mm. Results looked better now for Tok/Nikon. So perhaps AF under low ambient light could have produced bad results for the Tok/Nikon combo in the earlier shots. But taken together, and admitting that this was probably not the best test of the quality of these lenses, I would tentatively call the Pana fisheye the winner (with a small margin).

Here are two crops taken from the larger test pictures shown above: still some small advantage for Pana I would say. One more point: could the greater 'softness' of some Tok shots not be due to a slight motion blur caused by the mirror of the Nikon camera?

One more point: could the greater 'softness' of some Tok shots not be due to a slight motion blur caused by the mirror of the Nikon camera?

The moving mirror of the Nikon can greatly influence the sharpness of a picture and 1/50 sec and held in a hand is prone to produce "motionblur"!

Please redo the test using the MUP (Mirror Up) funcion and set the time to 6 seconds before the shutter open and record in RAW.

Also mounting the cameras on a stable tripod and using a strobe instead of ambient light would help to create images having the same parameters

and are much more useful to compare.

I have the D7000, the Tokina 12-24mm and the Tokina 100mm Macro and i never had the impression that they are unsharp or soft.

Chris

 

The moving mirror of the Nikon can greatly influence the sharpness of a picture and 1/50 sec and held in a hand is prone to produce "motionblur"!

Please redo the test using the MUP (Mirror Up) funcion and set the time to 6 seconds before the shutter open and record in RAW.

Also mounting the cameras on a stable tripod and using a strobe instead of ambient light would help to create images having the same parameters

and are much more useful to compare.

I have the D7000, the Tokina 12-24mm and the Tokina 100mm Macro and i never had the impression that they are unsharp or soft.

 

Chris

Hi Chris thnks for your note.

I dont want to carry these tests too far. My initial idea was to compare two popular fish-eye camera systems using identical settings, distance and field of view. With some small colourful objects on the foreground (say, at 10/20 cm) and some larger objects on the background. Conditions that mimic a bit those for for making CFWA shots under water. What I have in mind is to repeat the test with (more) natural light and settings 1/60 F8 ISO200. No tripod, but hand holding or stabilizing the camera with the Nikon mirror up. Bringing in artificial light might make things too complicated.

The two cameras, a big difference in size! . . Remember: crop factor Olympus sensor is 1.3 relative to Nikon sensor

 

Ok, another test this morning with settings 1/10 F8 400 ISO, 9 central focusing points. Tokina 10mm Panasonic 8mm

Nikon Mirror up. Both cameras 10 cm in front of conch and stabilized on table. Crops again 100%. JPEG from camera.

 

and with Mirror up and exposure delay for Nikon/Tok

 

 

Ok here is my last test of the Pana and Tok now with some graphic material. Settings: 1/80, F8, ISO 400. Single focus point. 15 cm from object, Nikkon Mup, JPEG fine from camera. Letter crops at 200% and 100% respectively

So my general conclusion from my indoor tests is that there seems to be no substantial difference between the quality of these systems, despite the large difference in camera, sensor and lens types. I felt that the Pan shots were a bit more 'crisp' than the Tok shots. And Nikon/Tok struggled more with AF under low light conditions. When I Iooked at the corresponding raw files om my computer I could see no big differences with the jpeg files. Of course the 'ultimate truth' of these fish eye lenses can only be found underwater on a nice sunny reef! Thanks for you kind interest and don't hesitate to upload your stuff! Cheers Al

Thanks for the comparison Albert!

Thanks for the comparison Albert!

yeah. .my wife is happy now the cameras are cleared from our dinner table