Irix 15mm f/2.4

[Tristin]

Yeah, those samples are poor all around.   This really boggles my mind.  How, in the age of google and pixel peeing, can a company think they can market lenses like this?  If they were marketing them as cheap alternatives, sure.  That is not what Irix is doing though.  How does such incompetence get into positions to be able to put products out like this?  People are strange.  RIP Irix, we knew thee not.

[simsurace]

Lens tip image samples always look terrible up close because they have no corrections and sharpening applied. Even legendary lenses such as the Nikon 14-24mm look terrible in their samples.
http://www.lenstip.com/295.11-Lens_review-Nikon_Nikkor_AF-S_14-24_mm_f_2.8G_ED_Summary.html

And they are from a DX camera, so not even showing the outer regions of the image circle.

The Samyang 14mm samples hardly look better. And so on.
http://www.lenstip.com/239.11-Lens_review-Samyang_14_mm_f_2.8_ED_AS_IF_UMC_Summary.html

Can someone tell me where the milkyway&mountains image discussed earlier comes from? My suspicion is that this is not even an official sample image. 

[Tristin]

I know what a raw image from a good lens looks like and none of those images came from a good lens.  I will say that the proclamations of it being poor here are a bit exaggerated as we here have higher standards.  I'd call this lens pretty mediocre.  Which is disappointing considering the design features are clearly aimed at knowledgeable photographers.

[simsurace]

I just clicked through the 14-24 gallery and the Irix 15mm gallery on lenstip and I can't see those massive differences. My eyes could be worse than yours though, so I reserve my judgement until the numbers come out.

[Jan Anne]

Judging a 14 or 15mm lens with the same scrutiny as one judges a normal or tele lens will result in a lot of disappointment

These are extreme wide angle lenses to suck everything into the image, not for bleeding pixel peeping results.

[simsurace]

I tracked down what seems to be the original of that milkyway image. There is no lens info, so I asked the author about the lens it was shot with. Currently waiting for the answer.

https://www.flickr.com/photos/guatitamasluz/25921410902/

The shot was taken on March 18, 2016. 

[Bjørn Rørslett]

So much useless EXIF info, and no lens ID ...

I strongly disagree with the notion that superwide lenses cannot stand scrutiny. They indeed can and should. If not only to convince oneself that the concept of hyperfocal setting never was applicable. The sharper the lens, the less convincing hyperfocal focusing will be.

As to the new series of test images, do have a look at the pavement front left and right corners. The problem is very obvious. One could of course blame field curvature but that would require *both* corners to go out of focus concurrently, which they clearly do not do. Thus is is not field curvature that causes the very uneven distribution of sharpness, whether it is the Milky Way image (at infinity focus) or the architectural snapshots shown later.

When I evaluate a lens, I only rely on what I can observe myself. Plus, the observation(s) have to be repeatable.

[simsurace]

Irix confirmed to me that this [EDIT: the milky way image that we were talking about] is no official sample image, all official samples are on Flickr and 500px. Thus the questions about why they posted such a horrid sample are now resolved -- they didn't.

[Bjørn Rørslett]

The salient question is whether it was taken with the IRIX lens - or not.

[pluton]

It is worth remembering what Simone observed, the sample images are completely unsharpened.  Also, they're from a Canon 5DMk-something.  Plus we don't know what downsampling process was used, etc, etc. 

[Bjørn Rørslett]

I assume a Canon 5D something also is a capable camera? It certainly is popular.

[simsurace]

The salient question is whether it was taken with the IRIX lens - or not.

Irix said to me in a personal message that the lens was not in production when the shot was taken, and they have nothing to do with that picture.

[simsurace]

As to the new series of test images, do have a look at the pavement front left and right corners. The problem is very obvious. One could of course blame field curvature but that would require *both* corners to go out of focus concurrently, which they clearly do not do. Thus is is not field curvature that causes the very uneven distribution of sharpness, whether it is the Milky Way image (at infinity focus) or the architectural snapshots shown later.

I know that you have immensely more experience testing lenses and making judgements on image quality than I have. As such I appreciate your worries about possible issues of optical quality and quality control in the lens in question.

However, comparing to other sample images on the lenstip site, I cannot find any samples from similar lenses and conditions that would appear to be head and shoulders above the Irix 15mm. For instance the Canon 11-24mm lens appears to have very similar image quality at 16mm. If anything, the Irix lens seems sharper in the corners.


Unfortunately, they don't have FX images of the 14-24 or the Zeiss 15 as these would be the ones to compare against.

I did not see any obvious differences between the two corners. BTW, how do you control for different distances to the camera? The pavement seems to be slanted, so even with a level camera the distance to the ground is not the same for both corners. This may have a slight influence?

[bjornthun]

Another contender, if you have a Need for Speed , is the newly announced Laowa 15mm f/2, yes f/2, with a 72mm filter thread for Sony full frame E mount, http://www.sonyalpharumors.com/venus-optics-unveil-worlds-fastest-15mm-lens-sony-full-frame-e-mount-camera/.

[Bjørn Rørslett]

If people don't see the problems, then as far as they are concerned there are no problems. Each to their own.

I for one lost all interest in the Irix lens.