That shot (and its 100% crops, of course) was taken at f/2.8? In my untrained eyes, the 100% crops look really good..... Sample variation factor at its best?
That looks quite good for a Canon 14mm f2.8 L at wide open. Looks like Canon may have a good lens here... or atleast, as AGeoJO said - sample variation at it's best.
One thing I've noticed about various lenses is that some of them are better at close range than at infinity. Could that be the case with the 14II here?
Doesn't that have to do with the resolution of the lens.
I could be way off here, but as distance increases, the resolving power of the lens decreases. i.e. the greater the distance the subject is from the lens/film plane, the lower the resolution of the image/subject. Isn't this what some MTF's indicate.
Which is why some Zeiss and Leica lenses - which have high resolving power - around 290-325 lp/mm - up to 400 lp/mm have such excellent image quality/features compared to Canon lenses which have around 60-140 lp/mm at their best.
Someone please correct me if I am way off here - as I would personally like to know this as well.
The 14-24/2.8 arrived and I gave it a quick test on the SLR/c with a standard adapter. With a standard adapter the lens is fully stopped down. The lever on the lens must be moved to open it up. Looks good on my first quick test of my kitchen. Excellent color, not too much vignetting, and good corner sharpness. All this was done at some aperture close to wide open. I have about 15 minutes in the morning, and sun willing, might be able to get a few shots off before starting my long day.
My first impression was that the lens is HUGE. I had a 14mm Tamron before and it looks small in comparison. Build quality seems to be excellent, but it has a Canon like manual focus feel which I don't like much. The zoom ring is more weighted and feels a little smoother.
The SLR/n will be coming tomorrow and after that I will really be able to put it to some use this weekend. Keeping my fingers crossed the used camera is in good working condition and the lens performs. The lens has no issues on my SLR/c (besides no aperture control, AF, ...).
DPReview ? The original poster of this thread, hubsand/Mark, is the one that did the test. He first compared the new Nikon lens with a Sigma but later on threw in the new Canon 14mm II into the mix.... Just FYI.
The main drawback to Mark's (16:9) testing is sample variation, or just older (out of calibration) lenses he uses for his testing. It looks like he has a poor copy of the Canon 14L II as "S Zozgornik" appears to be performing much better. Fortunately for Canon, Mark will likely get a new copy and redo his test. His bad performing Zuiko 18/3.5 doesn't have the same luxury. A few users here, includinhg myself, have done their own testing which shows much better performance.
Edited by pdmphoto on Dec 21, 2007 at 09:51 AM GMT
Well, from my point of view, bad performing no longer made lenses are part of the risk in pursuing alternate lenses, so the testing emphasizes the tradeoffs. If you want to buy an Oly 18mm (I had a good copy), you run the risk of getting a bad copy that you won't be able to return or exchange. For me, at least, it's an important point. I'd much rather have a currently produced lens that I can buy new, and return to the retailer for a refund if when I receive it, the sample doesn't perform well, or otherwise doesn't satisfy my expectations/hopes.
pdmphoto wrote:
The main drawback to Mark's testing is sample variation, or just older (abused?) lenses he uses for his testing. It looks like he has a poor copy of the Canon 14L II as "S Zozgornik" appears to be performing much better. Fortunately for Canon, Mark will likely get a new copy and redo his test. His bad performing Zuiko 18/3.5 doesn't have the same luxury. A few users here, includinhg myself, have done their own testing which shows much better performance.
Yes, I agree - but the difference is that Marks refuses to publish that he has a poor performing Zuiko 18mm, even though there have been results posted by myself and even better respected members who have shown better results. When he has potential quality issues with new lenses he makes it clear in his review that he may have a bad copy.
httivals wrote:
Well, from my point of view, bad performing no longer made lenses are part of the risk in pursuing alternate lenses, so the testing emphasizes the tradeoffs...
Edited by pdmphoto on Dec 21, 2007 at 10:20 AM GMT
Dec 21, 2007 at 01:18 PM
brainiac Offline [X]
p.9 #10 · Nikon 14-24mm tested on Canon 5D/1Ds III
I see no difference between the performance in the twig images here https://www.fredmiranda.com/forum/topic/594820/7 , and Mark's tests. The new 14L is a fantastic lens, but it does not exhaust sensor detail in the corners at f2.8, and does exhibit significant blue fringing of objects against very bright backgrounds in both Mark's test, and the pictures of a gatepost on this thread. The Nikon is a much sharper lens, with less aberration, but has much worse distortion at the wide end. All this talk of sample variation is silly. If you test two 14L2's against each other, then talk about sample variation and how the one Mark tested was a bad copy, but right now it looks identical to every other 14L2 out there. It's a great lens. It has wonderful straight lines. It is a bit woolly at wider apertures in the corners. I don't understand why people immediately reach for the sample variation (Chewbacca) defense.
p.9 #11 · Nikon 14-24mm tested on Canon 5D/1Ds III
I've tried two different Canon 24L's and they were rubbish (to different degree). Nothing like the results Mark posted with his. Others have seen this same degree of sample variation with the 24L. Just do a search on the net if you don't believe me. You can also include the Canon 16-35L version I. That's why people talk about Canon wide sample variation - because it has existed in the past. Has that changed recently?
Edited by pdmphoto on Dec 21, 2007 at 10:35 AM GMT
p.9 #13 · Nikon 14-24mm tested on Canon 5D/1Ds III
You don't need side by side shots to see that the 14-24 is a very, very good lens. Few lenses in this focal length are able to crisply resolve fine detail in the frame corners - even if a fixed lens like a Zeiss 21 or Canon 14 MK2 are able to pip it I would preffer the Nikon for it's ability to frame exactly in camera and maximise the sensor area without cropping.
p.9 #14 · Nikon 14-24mm tested on Canon 5D/1Ds III
brainiac wrote:
I see no difference between the performance in the twig images here https://www.fredmiranda.com/forum/topic/594820/7 , and Mark's tests. The new 14L is a fantastic lens, but it does not exhaust sensor detail in the corners at f2.8, and does exhibit significant blue fringing of objects against very bright backgrounds in both Mark's test, and the pictures of a gatepost on this thread. The Nikon is a much sharper lens, with less aberration, but has much worse distortion at the wide end. All this talk of sample variation is silly. If you test two 14L2's against each other, then talk about sample variation and how the one Mark tested was a bad copy, but right now it looks identical to every other 14L2 out there. It's a great lens. It has wonderful straight lines. It is a bit woolly at wider apertures in the corners. I don't understand why people immediately reach for the sample variation (Chewbacca) defense....Show more →
Iīve looked again over my and the little Sample in the test. To make them comparable Iīve enlarged mine to the size of a 1DMkIII (x1.12).
After viewing this in direct competition up to 300% magnification.
I guess You are right. The yellow blue fringing is the same as in my copy.
Contrast is much better in my image, but this is not comparable at all, because of different location and light.
Then Iīve looked in the same way to the image of the 14-24:
This shows red/cyan fringes which are a bit less wide as the blue/yellow ones on the 14L.
Both show the same sharpness and detail in the background building, but the tree there is nicely sharp with the 14-24, but heavily blurred at the 14L. This leads me to the conclusion: The focus is lying somewhere else...
Different position of focus? Canīt imagine this by such a professional.
Or is this effected by a curvation of field?
Iīve once observed this when I tried to compare the Sigma 12-24mm and the Tamron 14mm. On one the "plane" of focus was curved concave on the other convex, so with changing set-ups the winner in edge sharpness was a different one.
Testing Lenses is surely no fun at all!
The Samples given are to small for me to judge all this. The missing overview over the complete scene is not making this easier.
Please hubsand be so nice and give us the RAW-files!
Thanks!
Steffen
Dec 21, 2007 at 02:44 PM
brainiac Offline [X]
p.9 #15 · Nikon 14-24mm tested on Canon 5D/1Ds III
...and you have to remember we are asking a hell of a lot here. 14mm, f2.8, 21 megapixel. We are beyond Hasselblad SWC territory.
As regards sample variation, I'm not saying it doesn't exist, and having owned and relinquished two 24L's, and seen the performance Mark got with his at f1.4 in this test, I have certainly seen it, but there is some evidence that Canon went through a bad patch and has sought to reduce variation. There's no reason yet to suppose that the 14L Mark used on Friday was sub-standard. I went to some trouble to borrow it for him, so I have a vested interest in it being OK. It was straight out of a new box from the shop shelf. Before writing it off as a bad copy, show us examples of one doing significantly better. Sit down Chewbacca!
p.9 #16 · Nikon 14-24mm tested on Canon 5D/1Ds III
What exactly is the difference to me, the buyer, between a badly designed lens and a badly manufactured or badly quality controlled lens? I don't see any relevant difference. I would have soft images, one way or the other.
p.9 #17 · Nikon 14-24mm tested on Canon 5D/1Ds III
When I saw the first samples of my 14LII, I was so overwhelmed of the sharpness from corner to corner, and the crisp fine details.
The Lens is classes beyond the 14LI, and I think it is even better than the 1.4/24 and 1.4/35.
Seeing this on first sight poor corner crops made me think this canīt be possible, this is not like my lens.
Now after reconsidering what I see there I have made the following conclusion:
The samples we see from the 14LII are extremely good for a 14mm lens. Only comparing it to an even better one make it look bad. The 14-24 seems to be a notch better, but the difference is not as large as it seemed to be in the edges at first sight because of a different focus.
Stopping down the difference getīs negligible. At least if You do not have an 1DsMkIII.
Itīs hard stuff to take that the 14-24 is cheaper than the 14LII, but would I give away my 14L and 24L for this Lens?
- I would win top Performance on 14 and 24mm and everything in between, with only one Lens! Intriguing, but itīs really bulky.
- I would loose Autofocus, Autoaperture and two f-stops at the long end. (If I stay with my Canon Body.) Weīd have to wait if distortion is not eating up the higher quality again, for those who have to correct it.
I think for myself this Lens would only be an option if I would make a complete step to Nikon...I do not like to photograph all the time without Autofocus even at an super wide angle if I use it wide open. (I miss to often!)
Now Iīm waiting for the complete test, and see if my conclusions will be still the same then...
p.9 #18 · Nikon 14-24mm tested on Canon 5D/1Ds III
Ok then, where are the images from it to back it up? All we have been shown is the first test. Where are the ones from last Friday's (as in over a week ago) shoutout?
Maybe it's in transport as we speak, getting replaced
Mark seems to give Canon lenses the benefit of the doubt many times over until he gets a good one for his final testing.
brainiac wrote:
...and you have to remember we are asking a hell of a lot here. 14mm, f2.8, 21 megapixel. We are beyond Hasselblad SWC territory.
As regards sample variation, I'm not saying it doesn't exist, and having owned and relinquished two 24L's, and seen the performance Mark got with his at f1.4 in this test, I have certainly seen it, but there is some evidence that Canon went through a bad patch and has sought to reduce variation. There's no reason yet to suppose that the 14L Mark used on Friday was sub-standard. I went to some trouble to borrow it for him, so I have a vested interest in it being OK. It was straight out of a new box from the shop shelf. Before writing it off as a bad copy, show us examples of one doing significantly better. Sit down Chewbacca!...Show more →
Dec 21, 2007 at 08:55 PM
Andi Dietrich Offline [X]
p.9 #19 · Nikon 14-24mm tested on Canon 5D/1Ds III
? The original poster of this thread, hubsand/Mark, is the one that did the test. He first compared the new Nikon lens with a Sigma but later on threw in the new Canon 14mm II into the mix.... Just FYI.