Lenses and Optics

Just the Lenses: Sigma 24 f/1.4 Art Comparison

Published April 2, 2015

Most of you know I’ve been very impressed with Sigma’s new Art lenses. Their 35mm f/1.4 Art I still think is the sharpest 35mm prime lens made. The 50mm f/1.4 Art is also superb.

When I heard about the Sigma 24mm f/1.4 Art lens, I had some mixed emotions. I was excited that we might get a 24mm lens of similar quality to their 35mm. But the logical side of me thought that perhaps Sigma had bitten off a bit more than they could chew this time. Designing a wide-aperture 24mm lens is much more difficult than designing a fast 35mm lens. Even the best 24mm f/1.4 lenses (I consider the Canon 24mm f/1.4 L to be the best current offering, although that’s arguable) still have distortion, aberrations, and some edge softness.

But when our first five copies of the Sigma 24mm f/1.4 arrived, I swiped them from intake and took them over to the testing lab for a quick look and MTF testing on our optical bench. We already had results from the Canon 24mm f/1.4 L, the Nikon 24mm f/1.4 ED AF-S, and the Rokinon 24mm f/1.4 lenses in our database to compare them to.

MTF Tests

As usual, MTF testing was done for five copies of each lens. Each copy was tested at four rotations (0, 45, 90, and 135 degrees) to give an average MTF across the entire surface of the lens, and the results for each lens then averaged. Prior to MTF testing, all lenses were double-checked on OLAF, our 5-micron pinhole collimated testing machine, to assure they were well-centered. Since the Sigma was the focus of this post, I’ll show its MTF curves compared to each of the other 24mm f/1.4 lenses.

Legend for all the MTF graphs

 

 

For those of you who don’t speak MTF, I’ll summarize a bit. In the center 1/2 of the lens (from 0 to 10mm) the Sigma clearly is more contrasty and has better resolution than even the Canon lens, especially at higher (40 lp/mm and 50 lp/mm) frequencies. In the outer part of the image (from 15mm to 20mm) though, the Canon and Sigma are about the same.  The Sigma is better than the Nikon all the way out to 15mm, but again, in the outer areas there is either no difference or the Nikon is slightly better.

The Rokinon is a rather different lens. It doesn’t resolve as well as either the Nikon or the Canon, and not nearly as well as the Sigma in the center of the image. The Rokinon’s advantage is that its curves remain flat almost to the very edge of the image circle, and again in this outer 1/4 of the image it is as good as the Sigma, or perhaps a bit better.

What the MTF curves suggest, then, is that Sigma has made the best resolving 24mm lens in the center of the image, but at the outer edges they’ve run up against the same aberrations and problems that designers of 24mm lenses have always faced, and haven’t managed to overcome those.

Field Curvature

We haven’t always included field curvature diagrams in this little reports, but 24mm wide-aperture lenses tend to have a lot of curvature so it seemed a good idea.

You’ll notice, of course, that the difference in Sagittal and Tangential field curvatures parallels the increased astigmatism areas you see on the MTF graphs above. But notice also how the fields curve very severely in all of these lenses out in the edge area of the image, from 15mm to 20mm. This is part of the problem we always see with 24mm wide-aperture lenses and it seems that current lens technology cannot quite overcome this.

It might help to give you an image from Olaf, our 5-micron pinhole machine we use for testing centering. With the lens set at infinity focus, we manually focus the lens a few microns at a time.

If we focus on the center point, the corners are well out of focus, like this:

Sigma 24mm focused on center point (bottom image). Upper images show the two sides at 20mm.

 

We can focus on the edges and make them nice and sharp, but doing so defocuses the center quite badly.

Sigma 24mm f/1.4, focused on edge points (top boxes) at f/1.4

 

Even stopped down to f/5.6 the focus plane at center and edge is quite different.

Sigma focused on center point at f/5.6

 

Sigma focused on edge points at f/5.6

 

Let me be clear here: the Nikon and Canon lenses are not any better as far as field curvature. I just saw no reason to post all those images up to show you the same thing over and over. The Rokinon is not as extreme, but then part of that is because it’s never as sharp at any point as the others are. I should also mention sample variation with the latter lens is much greater – you’ll rarely see both sides in the same focus plane with that lens.

Summary

The new Sigma 24mm f/1.4 Art lens does, indeed, out resolve the offerings from the major manufacturers, at least in the center of the image. At the edges, though, the advantage disappears. More importantly with any 24mm f/1.4 lens, try as you will, unless you really stop the lens way down you aren’t going to get a flat field of focus. And if you are going to stop the lens way down, why invest all the extra money for a wider aperture lens.

That doesn’t make these bad lenses. There are clearly some types of photography that this focal length is invaluable for, and in that case you just learn to work around the shortcomings. In many cases, though, the old rule that the best 24mm f/1.4 is a 35mm f/1.4 and a few steps backwards is often true.

 

Roger Cicala and Aaron Closz

Lensrentals.com

April, 2015

 

Author: Roger Cicala

I’m Roger and I am the founder of Lensrentals.com. Hailed as one of the optic nerds here, I enjoy shooting collimated light through 30X microscope objectives in my spare time. When I do take real pictures I like using something different: a Medium format, or Pentax K1, or a Sony RX1R.

Posted in Lenses and Optics
  • Mark Harris

    Hopefully I can get a reply here 😀
    Is the sharpness bad on the edges due to curvature? Meaning that I can get a sharp subject to the left or right of center position if that is they focus spot?

  • I really wanted this lens to stand out head and shoulders above the others. “Someday my 24 f/1.4 will come along. Until then I will do the best that I can.”

  • Assessing cost to build is really difficult and something I don’t even try. The glass for an element for example, may cost $10 a pound or $400 a pound depending on what type it is. Or the finishing or coating costs may vary by a factor of 10. One element may be finished to the accuracy of 1/2 wavelength, another to 1/10 wavelength. We can make judgements about the mechanics of a lens and what they cost, but not really about the optics without a lot of analysis we aren’t capable of here.

    Roger

  • Maurice Piper

    Excellent stuff, again !! This is really great work you are doing ..
    May I pose one question .. in the Nikkor 1.4 prime lineup the 35 looks to be significantly cheaper to build, but this seems not to be reflected in the price. The 50 certainly is both .. while the 85 is just a gem !
    Care to comment ?

  • Roger Cicala

    JD, I can’t answer that simply because I haven’t tested them all at f/16

  • Thanks Roger for the comparison. You mention at the end of your article that why people would buy such a wide lens if they stop down so much to get the curve flat ?
    My reason would be that, as a nature landscape photographer, I often use f/16 to amplify the depth of field and and at that aperture, there is simply not many if any 24mm or zooms offering that focal sharp enough in the corners to reach my goal.
    So my question would be, let aside the widest aperture, and in search of the best lens with a focal near 24mm offering the best sharpnest in the corner and comparable in the center, is this lens even with or better than the others at f/16 and you can add the TS-E 24mm f/3.5L II (not tilted) in the game.

  • Ole

    Stepping back will not give you a wider FOV in the starry night-sky, hence a somewhat unnecessary comment as many buy this lens for astrophotography.

  • swobo

    “In many cases, though, the old rule that the best 24mm f/1.4 is a 35mm f/1.4 and a few steps backwards is often true”. Well said, I don’t know if my 35 can come off my camera at times. Have to tap it with a hammer to get it loose.

  • Peter Harris

    For general use, and travel, on my a7r (and prior 5ds) the old zuiko om 24mm f2.8 has been very good. Seems sharp all over,and it is tiny and mechanically wonderful.

    For critical interiors the canon 24mm T/S 2 has been brilliant. The only issue is that the canon to e-mount adaptor HAS to be flocked.Even though it is big, I recently shot with it handheld and had no issues.

    One of the problems with the supersharp center/corner drop off, is that the lack of corner resolution is emphasized. An even resolution across the frame will seem like an overall sharper picture, even if you give up some center resolution.

  • Roger Cicala

    Wow. Sometimes I just have to say wow.

  • colie

    Can I ask what Nikon camera did you use to test the Sigmas and did you find there was a wide variety in fine tuning needed?

  • El Aura

    Do the field curvature graphs mean that with the centre focussed at infinity, the corners are sharp closer to the camera?

  • Philip Partridge

    Thanks again, Roger/Aaron. You might have noticed these kinds of lenses becoming popular in cinema and cable tv – you get a very distinct and rather appealing look from them at a focal distance quite different to a 35/1.4. It’s a tough ask for them to be CoF free or near to it, and f1.4 is almost certain to have OOF as the main content in the outer frame. Yes, CZ’s WA Otus will see plenty of interest – it may be very large indeed.

Follow on Feedly