Nikon lens 14mm or 14-24m

[pixjohn]

I am looking for a new lens to shoot some 360 VR tours. I am looking at the Nikon 14  and 14 - 24

The 14mm lens is a  Rectilinear design provides straight-lines

or

AF-S Zoom Nikkor 14-24mm f/2.8G ED AF Lens.

I am not sure the 14 -24 is going to be the best choice if I want straight lines?

John

[Tony Beach]

Both lenses are rectilinear.  The 14-24/2.8 has less barrel distortion than the 14/2.8 does, and it does everything else better too -- i.e., it's no contest.

[pixjohn]

If the 14-24 is that much better why would nikon even keep the 14 in the line up. I will rent the 14 -24 and test it out.

Since I come from shooting wide angle with a schneider 24xl with med format, is the 14-24 wide enough? I was thinking if i need more top to bottom i might try shooting vertical and stitch more images. .

Both lenses are rectilinear.  The 14-24/2.8 has less barrel distortion than the 14/2.8 does, and it does everything else better too -- i.e., it's no contest.

[Panopeeper]

I am looking for a new lens to shoot some 360 VR tours. I am looking at the Nikon 14  and 14 - 24

The 14mm lens is a  Rectilinear design provides straight-lines

Rectilinear lens and straight lines are two separate issues.

I am not sure the 14 -24 is going to be the best choice if I want straight lines?

Why would you want straight lines for a 360° VR tour? The stitcher, which creates the pano for the VR tour will bend even the straight lines in your shots.

The best on stitching is, that one does not need to care for the geometric distortion (i.e. barrel and pincussion). The stitcher will take care of that.

[pixjohn]

I am guessing with your screen name, you might have a clue about this pano stuff? Is 14mm wide enough on a d700?  I am not even sure what software works best for post?

John

[OldRoy]

I am guessing with your screen name, you might have a clue about this pano stuff? Is 14mm wide enough on a d700?  I am not even sure what software works best for post?

John

Panopeeper is far more experienced than I, I believe, however here's my own experience.

Firstly, as others have said, lens distortion is irrelevant when creating VRs. Most people use fisheye lenses. The images are remapped by the software into the equirectangular form (2x1, 360 x 180 deg equivalent) which is the basis of the viewed image.

I had been using a D200 with the Nikkor 10.5 fisheye for about a year and a half before I bought a D700 and, amongst others, the 14-24 2.8 - which is in itself a wonderful lens. It can be used to do VRs - but so could a 200mm lens if you wanted to spend a month creating one from the hundreds of shots that this focal length would require. As some people have done, with amazing high-resolution results.

However I would say that the 14-24 lens is completely unsuitable for the purpose. I've used it; it requires that you shoot 2 rows of six shots (assuming you're not doing HDR, which would require exposure bracketing too) plus however many shots required by your chosen method for patching the nadir - which many people don't find necessary. This approach requires a LOT of PP time in stitching - at least it does for me using PTGui Pro, which I find to be excellent, if a little difficult to learn initially. Others swear by AutoPano Pro. Both will do a good job once you know how to use them I believe. Also the lens is huge and heavy, requiring a very substantial pano head. It's also very easy to nudge the zoom ring and accidentally shoot at >14mm, possibly scr3wing up the whole job.

Give that the final use for most VRs is a tiny .mov or flash file on a website, it's fairly pointless to take this approach. I have had the lens hood shaved from my 10.5 FE which enables me to use it on the FF body, which gives a circular image. In theory this enables you to shoot 3-round, plus zenith, plus nadir(s), however I usually shoot 4-round as the greater overlaps make for more accurate, faster stitching. The resulting tif created by this approach is about 70Mb (I can't recall the actual pixel values as I'm on my laptop.)

In fact I now usually revert to shooting with my D200/10.5 FE. For the time taken to shoot an additional 2 shots the difference is negligible. In this format the lens fills the entire frame (at an effective equivalent f/l of about 15mm). So the resulting stitched equi.tif inclusive of the patched nadir is about 200Mb. Once reduced to a 1Mb .swf file these two approaches are effectively indistinguishable. I usually shoot HDR so the s/n  ratio and DR difference between the D200 and D700 has an insignificant impact on the final output on a website.

I now tend to use the D700/14-24 combination to shoot "stills" asociated with the VR job, so less lens changing involved. FWIW I believe that most people tend to like the Sigma 8mm fisheye for VR shooting. I started using the Nikkor 10.5 assuming that the full-frame fisheye would be more generally useful even when "defished": it isn't, particularly now that I have the D700/14-24 combination.

Anyway, I hope that this helps. Others may have better and/or more useful advice.
Roy

[ddk]

I am looking for a new lens to shoot some 360 VR tours. I am looking at the Nikon 14  and 14 - 24

The 14mm lens is a  Rectilinear design provides straight-lines

or

AF-S Zoom Nikkor 14-24mm f/2.8G ED AF Lens.

I am not sure the 14 -24 is going to be the best choice if I want straight lines?

John

I had both the 14/2.8 and the 14-24/2.8 when it first came out and ended up keeping the 14mm which has been a favorite lens for since its introduction. @14mm the zoom is VERY distorted, specially compared to the prime. As far as sharpness goes they're exactly the same but the zoom with its new coating has a higher contrast which I didn't care for.

There's another option which I prefer for shooting interiors over both Nikkors and its the Sigma 12-24; its a true rectilinear design and is wider than both. It has less distortion @12mm than even the 14/2.8! Its a slower lens but in this case shouldn't make a difference since you wont be shooting wide open anyway. Sharpness is on a par or very close to the 14-24, but its more compact, lighter and easier to handle.

[Panopeeper]

Is 14mm wide enough on a d700?

The narrower the lens, the higher resolution the result; however, as Roy indicated, most of the resolution will be ignored in the movie.

14mm sounds to me incredible wide, for I have a cropping camera (this comes to 9mm!). If you really need 360° vertically too, you might even mix the lenses: a somewhat narrower for making the horizontal shots for higher resolution and one wide for the zenir and nadir each.

The 14mm yields 80° along the shorter edge and 104° along the longer edge. In portrait orientation about seven frames should do it horizontally, and two frames for the zenith and two for the nadir would suffice.

I am not even sure what software works best for post?

This is primarily the question of stitching; the movie will be created from the 360° pano as an extra step. Both PTGui and PTAssembler can stitch even frames with different focal lengths. If Autopano can do that, that's great, for you don't need to learn much.

Depending on the distances, the accurate adjustment of the camera on the bracket is the most important issue. If you still have parallax errors, the process becomes painful and tiresome; you have to "hide" some errors in stitching and edit the masks before blending, to control which part of the result is taken from which frame. On the other hand, small errors may not be even visible in a movie.

[OldRoy]

I had both the 14/2.8 and the 14-24/2.8 when it first came out and ended up keeping the 14mm which has been a favorite lens for since its introduction. @14mm the zoom is VERY distorted, specially compared to the prime. As far as sharpness goes they're exactly the same but the zoom with its new coating has a higher contrast which I didn't care for.

There's another option which I prefer for shooting interiors over both Nikkors and its the Sigma 12-24; its a true rectilinear design and is wider than both. It has less distortion @12mm than even the 14/2.8! Its a slower lens but in this case shouldn't make a difference since you wont be shooting wide open anyway. Sharpness is on a par or very close to the 14-24, but its more compact, lighter and easier to handle.

It's worth pointing out that the OP's question was specifically about lenses suitable for VR panoramas.  

In any case your assertion that "@14mm the zoom is VERY distorted" is not shared, as far as I can recall, by any of the notable reviewers who have tested this lens, nor is it true IMVHO. There's a little easily corrected barrel distortion. The general consensus seems to be that this is the best wide-angle zoom ever made, comparable with the best primes at equivalent focal lengths: I have to take this on faith of course. I have never used the 14mm prime or the Sigma you mention so I can't comment.

In any case the distortion issue, even if true, is totally irrelevant in this context and may well confuse the OP! What do you use for VR panoramas, DDK?

Roy

[ddk]

It's worth pointing out that the OP's question was specifically about lenses suitable for VR panoramas.  

In any case your assertion that "@14mm the zoom is VERY distorted" is not shared, as far as I can recall, by any of the notable reviewers who have tested this lens, nor is it true IMVHO. There's a little easily corrected barrel distortion.

I guess everything is relative, compared to both the Sigma 12-24 and the Nikkor 14/2.8 it has a lot more distortion @14mm than either lens, and I stand by my statement. Unfortunately I don't have the test images any longer otherwise I'd post them for you to see.

The general consensus seems to be that this is the best wide-angle zoom ever made, comparable with the best primes at equivalent focal lengths:

Its a wonderful piece of glass and I never said that it was bad, just that its distorted and not rectilinear @14mm. As far as it being comparable to the best primes at equivalent focal lengths, well, maybe to equivalent Nikkors but compared to the Zeiss ZF lenses, its wishful thinking in every way. 14mm to 35mm is  my main shooting range and I would have loved to have one lens for most of that fl, unfortunately its not available yet and I'm still stuck carrying 4-5 lenses and 3 bodies.

I have to take this on faith of course. I have never used the 14mm prime or the Sigma you mention so I can't comment.

In any case the distortion issue, even if true, is totally irrelevant in this context and may well confuse the OP!

Maybe, but OP specifically asks about the linearity of the two lenses and nothing else, so it might be relevant to his application.

What do you use for VR panoramas, DDK?

Roy

I don't shoot VR panoramas, I merely answered OP's question regarding the linearity of the two lenses.

[PeterAit]

14mm sounds to me incredible wide, for I have a cropping camera (this comes to 9mm!).

If you have a cropping camera, 14mm becomes longer, not shorter (perhaps 21mm, not 9mm).

Peter

[Panopeeper]

If you have a cropping camera, 14mm becomes longer, not shorter (perhaps 21mm, not 9mm).

14mm on full frame is very wide for me; I would need a 9mm lens to achieve that angle of view on the cropping camera.

[OldRoy]

OK!
I wasn't wishing to be contentious at all, just to clarify things for the OP. Obviously shooting conventionally (whatever that might mean) with flawless Zeiss primes and shooting VR panos with anything at all are at opposite ends of a spectrum. I don't pretend to have the breadth of experience or level of skill found on these forums.

I'd say that amongst people who do VRs commercially, the Sigma 8mm circular FE on a crop-sensor camera seems to be the favoured combo. For "normal" web-resolution use the limitations are pretty well irrelevant. I think that most people who shoot VRs are primarily concerned about minimising the number of shots required. This is particularly true if shooting in uncontrolled environments - ie those that contain free-range humans, other animals, or moving inanimate objects. The less shots the less time spent in tedious masking before stitching.

Good luck to the OP. Let us know what you decide on and how it goes!

Roy

[pixjohn]

I am definitely confused even more. I shoot high end architectural photography using a Cambo wide DS. I have a 32 location project that I need to photography plus produce four 360VR for the web.  My plan was to buy a D700 plus a lens and a VR mount. I know from looking at other vr from the clients they are not the floor to ceiling movies.

I have my first project in a week and will maybe test out 2 lens.

John

[Panopeeper]

I know from looking at other vr from the clients they are not the floor to ceiling movies

This is a very important issue. IMO the "full height" movies are in most cases for the sake of making them. (There are exceptions, like the ceiling in a church can be important.)

If the ceiling is very high, then you can get away with a hand-held shot.

With the 14mm in portrait orientation, if the camera is at 5' hight and level, the floor is covered from 5-6' from the tripod. Note, that the camera does not need to be level (but the plane of rotation does); you can decide for more coverage of the ceiling without an elevated tripod by tilting the camera slightly upwards (but the overlap has to be greater). Of course if you shoot two rows, this is a non-issue, but the stitching becomes more of a problem due to parallax issues.

Note, that some lenses (perhaps most) change the entrance pupil location with focusing. If you have a pano bracket which allows for very high accuracy adjustment and you pre-calibrate the position, you need to do that with the expected focusing distance and aperture, and you should not refocus between the frames; this may put a strain on the aperture selection.

If you are so far that you calibrate the lens position, send me a message, perhaps I can give you some helpful advice.

Btw, zooming definitively changes the entrance pupil location (perhaps there are lenses excepted from this). Thus a fixed focal length makes life easier.

If you can try the lens, mount it, select a small aperture, look into the lens from the front (i.e. in the "inside of the camera"), and actuate the aperture with the DoF preview. The entrance pupil is there, where you see the aperture (the aperture is not really there, this is the projected location). Now change focus from a few meter to infinity and look if the aperture's position changes. If it does, the adjustment depends on the focusing distance as well.

[OldRoy]

Phew. I'd say that you need to borrow or rent some gear and do a test sooner rather than later.

From my own experience the choice of lens, so long as it's an adequately short focal length (14 mm on FF is fine but I think you'd be better off with even wider...) is not the "problem". It's the initial setup and getting used to the whole workflow.

1) What pano head are you intending to buy?
For guidelines I'd say the NN5 (Nodal Ninja) would do the job fine. The setup values (position of the no-parallax point aka "nodal point") for the 14mm Nikkor are probably already on their site - take a look.

2) What stitching software?
PTGui Pro and Autopano Pro will both do the stitching well; I've heard it said that APP is easier to learn. There's a good user forum for both applications - www.panoguide.com. There are some good tutorials available there.

As PP points out, If you don't need to show the zenith and nadir in the pano (ie the "polar regions") the stiching job is a lot easier and quicker. But you need to get used to the software!

You need to get some practice with the software ASAP in my opinion. I'm sure someone can supply you with some suitable files to practice stitching - me for example. I can let you have a set of six shots at approx 15mm f/l that would suffice to practice on. Or four round using a 10.5 circular FE on the D700 body.

Roy
edit: some useful sites.
http://www.nodalninja.com/
http://www.nodalninja.com/html/nikon_settings.html
http://www.johnhpanos.com/epcalib.htm
NB the 14mm Nikkor isn't listed on the NN site. The johnpanos site has valuable tutorials.

[Tony Beach]

Its a wonderful piece of glass and I never said that it was bad, just that its distorted and not rectilinear @14mm.

Frankly, I consider your assertion that the 14-24 is not rectilinear at 14mm totally bogus -- especially since you are not qualifying that and therefore saying it is a fisheye lens at 14mm!  It's just too bad we can't see your proof that it is not rectilinear, or that it is even in some way less rectilinear than the other lenses you are recommending here (I would welcome being proved wrong on this as I like to learn new things).  I just looked for myself to make sure I hadn't missed something, and at 14mm my 14-24 is rectilinear with some mild barrel distortion, and that would be less barrel distortion than the 14mm prime has according to Photozone:

http://www.photozone.de/nikon--nikkor-aps-...-report?start=1

http://www.photozone.de/nikon--nikkor-aps-...-review?start=1

They have sample images there also, and while there are no direct comparisons, you can see that the 14-24 at 14mm is rectilinear and shows little relative distortion; and there is a shot taken with the 14 prime that shows significant barrel distortion along the vertical axis of a brick wall in the foreground.

As far as it being comparable to the best primes at equivalent focal lengths, well, maybe to equivalent Nikkors but compared to the Zeiss ZF lenses, its wishful thinking in every way. 14mm to 35mm is  my main shooting range and I would have loved to have one lens for most of that fl, unfortunately its not available yet and I'm still stuck carrying 4-5 lenses and 3 bodies.

Zeiss has no 14mm prime.  You also say your 14mm prime is as sharp as your copy of the 14-24 (@ 14mm) was, but again that is at odds with Photozone's tests of these lenses at their borders and extreme corners.  Either you got a bad copy, or you're just not as critical as I am about these things.

[ddk]

Frankly, I consider your assertion that the 14-24 is not rectilinear at 14mm totally bogus -- especially since you are not qualifying that and therefore saying it is a fisheye lens at 14mm!  It's just too bad we can't see your proof that it is not rectilinear, or that it is even in some way less rectilinear than the other lenses you are recommending here (I would welcome being proved wrong on this as I like to learn new things).  I just looked for myself to make sure I hadn't missed something, and at 14mm my 14-24 is rectilinear with some mild barrel distortion, and that would be less barrel distortion than the 14mm prime has according to Photozone:

http://www.photozone.de/nikon--nikkor-aps-...-report?start=1

http://www.photozone.de/nikon--nikkor-aps-...-review?start=1

They have sample images there also, and while there are no direct comparisons, you can see that the 14-24 at 14mm is rectilinear and shows little relative distortion; and there is a shot taken with the 14 prime that shows significant barrel distortion along the vertical axis of a brick wall in the foreground.

I must say that I could care less wether you think what I said is bogus or not nor care much for pz's figures on this subject; specially since you never did the comparison for yourself. Shoot a room with the 14-24 @14mm and the 14/2.8 and look at the curvature of the side walls, then try the same scene with the Sigma 12-24 and you'll see that Sigma is more??? rectalinear??? than both those lenses.

Zeiss has no 14mm prime.  You also say your 14mm prime is as sharp as your copy of the 14-24 (@ 14mm) was, but again that is at odds with Photozone's tests of these lenses at their borders and extreme corners.  Either you got a bad copy, or you're just not as critical as I am about these things.

Nikon has a wonderful 14mm and Zeiss has 18, 21 and 25mm primes which covers most the zoom's range. As far as the extreme corners goes, I mostly shoot DX so it doesn't even enter my tests and even on FX slight differences at EXTREME corners wouldn't much to me either.

My edit:

By the way did which fl and aperture did pz use to come to their extreme corner conclusions compared the 14/2.8's test? 14-24's coating is more contrasty than the 14/2.8, could that have affected their results which is easy to fix by boosting contrast a bit in post?

[Tony Beach]

I must say that I could care less wether you think what I said is bogus or not nor care much for pz's figures on this subject; specially since you never did the comparison for yourself. Shoot a room with the 14-24 @14mm and the 14/2.8 and look at the curvature of the side walls, then try the same scene with the Sigma 12-24 and you'll see that Sigma is more??? rectalinear??? than both those lenses.

You offer an opinion which you state as a fact that is outside the experience reported by everyone else.  Extraordinary claims require extraordinary proof, but you in fact offer no proof at all.  You really need to get your facts straight too if you want to have any credibility -- the 14-24 is a rectilinear lens.  Now as to linear distortion, well I have shot extensively with the 14-24 and it has very little of that sort of distortion, and certainly less than the 14 prime according to Photozone.

You may say you have no respect for Photozone, but then you better show us why you would be any more credible -- and you flat out haven't done that here.  My strong suspicion is that you don't care for Photozone's figures because they are proof that you are wrong.  What I find particularly aggravating about your "I know because I had both lenses, and you don't because you only have one of them" argument is that you had the opportunity to take the test images and share them with us, but apparently the dog ate your homework and we'll just have to take your word for it.

Nikon has a wonderful 14mm... As far as the extreme corners goes, I mostly shoot DX so it doesn't even enter my tests and even on FX slight differences at EXTREME corners wouldn't much to me either.

Your credibility with me is dropping off a cliff -- it started by going straight down and it hasn't hit the bottom yet.  Photozone does all their measurements on a D200, so the results of their tests are relevant to your DX camera.  Like I said before, you apparently don't care, can't see, or got a bad copy of the 14-24; whatever the case, your arguments that your prime is just as good as the 14-24 at 14mm don't hold water.  I hope you don't plan on presenting a web sized image to prove how good your lens is; given your posts so far in this thread that's what I'm expecting to see next, and it will prove nothing more than what we already know -- you think your 14mm prime is the cat's meow.

[ddk]

You offer an opinion which you state as a fact that is outside the experience reported by everyone else. Extraordinary claims require extraordinary proof, but you in fact offer no proof at all.

Since I haven't seen anyone else testing these lenses next to one another so I don't see how I'm contradicting everyone? and there's really nothing extraordinary about my comments.  

You really need to get your facts straight too if you want to have any credibility -- the 14-24 is a rectilinear lens.  Now as to linear distortion, well I have shot extensively with the 14-24 and it has very little of that sort of distortion, and certainly less than the 14 prime according to Photozone

Not according to my eyes, @14mm the 14-24 has linear distortion, and as far as I know pz never conducted side by side tests of these lenses, so I don't understand what you're referring to.

You may say you have no respect for Photozone, but then you better show us why you would be any more credible -- and you flat out haven't done that here.  My strong suspicion is that you don't care for Photozone's figures because they are proof that you are wrong.  What I find particularly aggravating about your "I know because I had both lenses, and you don't because you only have one of them" argument is that you had the opportunity to take the test images and share them with us, but apparently the dog ate your homework and we'll just have to take your word for it.

No, dog didn't eat homework, it was a bunch of test shots from Dec. 2007 and I threw them away. I might have kept them had I known that a year and half later you'd want to look at them.

Your credibility with me is dropping off a cliff -- it started by going straight down and it hasn't hit the bottom yet.  Photozone does all their measurements on a D200, so the results of their tests are relevant to your DX camera.  Like I said before, you apparently don't care, can't see, or got a bad copy of the 14-24; whatever the case, your arguments that your prime is just as good as the 14-24 at 14mm don't hold water.  I hope you don't plan on presenting a web sized image to prove how good your lens is; given your posts so far in this thread that's what I'm expecting to see next, and it will prove nothing more than what we already know -- you think your 14mm prime is the cat's meow.

Are you so full of yourself that you think that you're the only one who can see? Or do you think that you're so important that I should care about your personal opinion of me or do I care to prove something to you? I haven't seen you produce anything to the contrary. Prove me wrong, post your tests and we'll see. From your reaction it seems to me that you're the one who thinks his 14-24 is the cat's meow otherwise you wouldn't have a coronary when someone doesn't agree with you.