Barreldistortion

I haven't used a collimator before, but I have adjusted flange depth by hand. As you close down the aperture the depth of focus widens, which may be throwing your results off. From my experience, you always keep the aperture wide open when adjusting the flange distance.

Lensdistortionmeaning

P.S. Dom - Great to hear from a fellow Melbournian (I live and work in Paris now). I've noticed on many Panavision lenses that's the case. So if I notice quite a change within the first few stops you recommend making new mesurent marks similar to that of Panavision? Do assistants find this helpful or an unnecessary hassle?

Photography is who I am. I can’t not photograph. I am compelled to write about the only thing I know. https://www.flickr.com/gp/43619751@N06/A7uT3T

Firstly; thank you Alan; Jean-Louis and Dom for replying so quickly and with such great responses. I have a further question and it may sound silly, but even so;

If you have a large enough screen (you’ll probably not see it on a phone) and can enlarge this, you can see the gapping between the known straight red lines and the corners of the fountain. The gapping is more pronounced on the right because it is physically closer to the 16mm uncorrected lens.

Well, first you set the zoom wide open at the long end, and adjust the focus to infinity going off the IHI (not the focus scale). Then you go to the wide end and adjust the back-focus to centre the H. You may then need to go back to the long end and fine tune where infinity is on the focus barrel. If it doesn't line up with the index mark the focus scale needs adjusting. Going back to the wide end, the H should remain within the bars as you go through the zoom range. Check it in both directions, as there can be a bit of play in the zoom mechanics. The graph of how far the focus drifts in both directions (measured in back-focus adjustment) is called the zoom curve, and manufacturers will generally have tolerance specs for it, but without getting too technical if the H remains within the bars (or just overlapping one side) the focus should be OK. If it drops out too much the lens needs servicing.

I am using a gecko-cam auto-collimator; with a lense at infinity and open fully I can measure the flange focal distance however, when I raise the f/t-stops, with some lenses the 'H' begins to move - changing my result. Is this a problem as it only happens with some lenses and should I compensate this into my results? If so how?

OpticalDistortion, Inc

“Barrel” distortion is a true optical distortion caused within the lens. It manifests as an outward bowing. Straight lines are rendered as if they were inflated. Wide-angle lenses are most susceptible to this, including the shortest focal lengths of many zooms. The degree of ballooning will vary from lens to lens.

Pincushiondistortion

By the way, it's a good idea to regularly check the collimator lens mount with a flange depth gauge and re-zero it if necessary. I check mine every time I switch it on.

Regarding making a separate index mark for focus shift, I've only done it very occasionally for picky assistants who complained about discrepancies. Because it can change depending on the distance it can be more confusing than helpful sometimes.

Optical distortioncalculator

It's mostly a problem associated with high-speed lenses, where it's hard to correct all the aberrations at full aperture, causing a slight shift in the plane of best focus. Past about f/2.8 the depth of field covers the discrepancy, but if the shift is large enough it can be a problem going from say f/1.2 to f/1.8. Within the tolerances of back-focus distance you can set a lens to one side of the tolerance to minimise how noticeable the shift is. So for example you might go from 0.01mm under to 0.015mm over as you stop down.

Following your instructions however when we move the wide end of the zoom spectrum the image (IHI) becomes completely blurred and impossible to line up. I've been using the projector (measuring to the 5" mark to sort it out and re-shimming) but it'd be nice if I could make sense of it on the collimator too. The lens we are working on is the Zeiss LWZ.2 15.5-45mm/T2.6.

I thought the color version was so monochrome that it might work better as a BW conversion. The sun was just barely up so there was little color. I added some selenium toning as I would have done making the print in my wet darkroom, but digitally it is less toxic and smelly and does not stain my shirts.

Optical distortionexamples

But lenses are curious objects, highly scientific in their design and specifications, yet also quite subjective in that they create images that are judged aesthetically. Collimators and MTF measurements are very handy in checking specs and comparing between lenses but they don't tell you what the image actually looks like. So while I rely on a bench collimator and MTF machine for a lot of things, I use a projector/collimator to really assess how a lens is performing, and where best focus might actually be. With focus shift issues it allows you to see just how much softness stopping down really causes, and how far you can push the back-focus tolerance when the lens is wide open to compensate for it.

I've just started becoming familiar with using a collimator to check the flange depth on lenes but had a quick question regarding reading the results (IHI).

Not perhaps the best example, but the lens used for this picture is a Sony/ZEISS Vario-Tessar 16–70mm f/4.0, pulled back to 16mm WA. It replaced my kit 16–50 and is an excellent lens. Its slight barrel distortion is most visible on the right side of the fountain. There is a distinct outward bowing of a perfectly square corner. This render is the SOOC raw capture, no editing, no lens corrections, zeroed out.

Which lenses are giving you changing results? Are they wider, or longer lenses? The wider lenses have a much shallower depth of focus, so I would guess that they would be the most affected.

On some lenses I've marked a different focus index mark for when they are used wide open, but it tends to shift further at close focus.

One of the forum rules here is that you use your real name. You might have noticed that all the responses to your question were from actual people rather than anonymous pseudonyms or companies. It's not strictly enforced by the moderators, but tends to be self-regulated by forum users. Quite a nice, civil etiquette. :)

There is also the convergence of the vertical lines because I had to tip the camera up to take in the fountain. Whenever the back of the camera is not parallel to the structure, there will always be that convergence unless you have a tilt-shift lens to correct it in-camera.

Optical distortiontest

With a collimator you'll also find that the frequency band of the light source may shift the focus on some lenses. A narrow spectrum green source (theoretically mid-range) might give you a different infinity point than a broad spectrum white source.

“Pincushion” is a true optical distortion diametrically opposite to barrel. Pincushion manifests as an inward bowing. Straight lines are rendered as if they were being squeezed from the sides. True optical distortions are distinct to the formulation of a lens. Telephotos, including long zooms, are most likely to suffer from pincushion.

If there is an issue with the lens, such as decentration of elements, the image will be fuzzier, but this is something you should pick up on projection.

Optical distortionmeaning

You need to make sure the lens is both centred and aligned to the collimator beam. As you go wider the IHI image will get dimmer and hazier, so you need to adjust the intensity of the light source depending on the focal length (and the aperture of course.) With very wide lenses (on my Chroziel collimator under about 10mm) it can be hard to make out a clear IHI image, but 15.5mm should be OK. Using a narrow spectrum green light source will give a sharper image, but a white source will be brighter. Sometimes it helps to view the IHI image in a darkened room. Obviously if the back-focus is out the IHI won't be visible until you're in the close vicinity of infinity.

I’ve been meaning to do a short piece on lens distortions for some time. We hear about wide-angle distortion and telephoto compression, and there is some truth to those, but in reality, there are only two true optical distortions.

In lightroom (or any other raw processor), I can apply lens corrections to eliminate the barrel distortion. Lightroom’s onboard lens corrections will straighten the lines unless it's very severe. Further, I applied a perspective correction to perfectly parallel the vertical lines, as shown by the white lines above.

As a general rule if the collimator back-focus adjustment you made at the wide end is more than about 0.01mm the zoom needs back-focus adjustment, usually done by re-shimming under the mount. The shorter the focal length, the more critical the back-focus.