When used in certain way and with an appropriate background, some lenses produce what is often referred to as "swirly bokeh" (or words to that effect). Although this is due to weaknesses in the optical design, some photographers find the effect appealing, and most find it at least curious.

The Helios-44 2/58 series of lenses is particularly well-known for this effect. Many different versions of the same basic design were produced, and there's a repeating topic of discussion across photography forums debating which version produces the most pronounced "swirl".

Amongst my collection of former Soviet Union lenses, I have several Helios-44 models - the 44-2, 44M and and 44M-4. In this post, I show the results of some informal testing with these three lenses to conclude which, if any, produces the most pronounced swirl effect.

The effect manifests most commonly when the background contains many small points of individual detail (especially specular highlights) that are sufficiently distant from the focal point to be somewhat out-of-focus, yet close enough that they retain some definition of form. The classic example is a portrait, with the subject in the centre and framed by a background of well-lit trees in full leaf (the leaves providing those many small points of detail).

To compare the three Helios lenses, I first produced a test image of white dots on a black background, and displayed this on a 42-inch LCD television. I set up a tripod approximately 1.3m from the TV, and mounted my Sony A7 MkII on it. Then, with each lens fitted to the camera in turn, I set the focus distance to 0.7m (thus throwing the test image's white dots out of focus) and took photos at f/2 (wide open), f/2.8 and f/4.

For all three lenses, the effect is most pronounced with the aperture wide open, significantly diminished by f/2.8, and almost completely gone by f/4 - hence, I only show below the shots taken at f/2 and f/2.8:



Looking at these results, I'd suggest that all three Helios-44 variants (specifically, my copies of them) display the effect to a very similar extent; so close, in fact, that I couldn't say which is which if shown the test shots in random order.

Whilst earlier and later examples (individual copies, even) of these same models may perform slightly differently, I conclude that the 44-2, 44M and 44M-4 have near-enough equal potential to provide the swirly background effect in the correct circumstances.

A final note:

As I mentioned previously, these tests were performed with a Sony A7 MkII "full-frame" camera. The swirl effect increases in severity as we move further from the frame centre, and is most pronounced at the borders and corners of the frame. On an APS-C "crop sensor" camera, the effect will be less noticeable, as those borders and corners have been cropped away. See below one of the f/2 test shots cropped to APS-C dimensions:

[IgorZ]

Thanks a lot for posting this. This is a very interesting way to test this effect. I think I will try it out to see how to optimize the bokeh of my Helios 40...

[BigMackCam]

Thanks a lot for posting this. This is a very interesting way to test this effect. I think I will try it out to see how to optimize the bokeh of my Helios 40...

You're most welcome. I recently took delivery of a 40-2, so I'll be testing that too in the next few days

[utak]

Very interesting results and an excellent way of showing how/why the Helios swirls. Thanks for doing this. I also have these three lenses, and I have to say my experience over the years is that my 44-2 images definitely look more swirly than the 44M and the 44M-4.

So what other variables are at play, over and above the evidence shown on your charts? Leaving aside sample variation, I suppose there are three main ones.

Firstly, corner to corner sharpness. Not so apparent from your charts to me, but the Helios 44-2 tends to be softer at the edges. This can accentuate the feeling of the image swirling around a sharper center object.

Secondly, coatings, contrast and resolution, The 44M-4 in particular has more advanced coatings and this reduces the resolution dreamy-ness and reduces light flares/issues seen on 44-2 images. These factors help bring out swirls IMO, as long as one adds quite a lot of contrast to 44-2 images. Adding a lot of contrast to the 44M-4 is not so productive.

Thirdly, (and related to the last variable), the appearance of bokeh bubbles. I've found the Helios 44-2 produces more bokeh bubbles than the 44M and 44M-4 in situations with a lot of back lighting (ideal for swirls) . These bubbles also help to increase the swlrly feeling, because the distortions (from round to skewed ovals as shown on your photos) are more visible.

These are all subjective generalisations of course, but I think they help to explain why some images look more swirly than others. And just a quick note on other versions. My M39 Helios 44, 13 blades version tends have more swirls and especially bubbles than the 44-2, and if someone wants a combination of good swirls and good resolution, the 44-3 is worth considering (but make sure it's one that fits properly).


PS If you really want extreme swirls, try reversing the front element of Helios 44/44-2 (as below)! And if you want more interesting bokeh bubbles, try reversing the rear element.

---

Last edited by utak; 01-29-2017 at 04:20 PM.

[BigMackCam]

Thanks, @utak - you raise some great additional factors that could affect how pronounced the swirl is for a lens.

Like you, I notice my 44-2 is a bit softer at the borders, so that could could affect the appearance of the swirl effect in real-world shooting (as opposed to these tests). But I think you may well have hit on something with the difference in the way out-of-focus highlights are rendered. I'll have to do more testing, but I suspect "bokeh balls" with a bright edge may accentuate the effect.

Thanks for the feedback and ideas

[BigMackCam]

Thanks a lot for posting this. This is a very interesting way to test this effect. I think I will try it out to see how to optimize the bokeh of my Helios 40...

Just as a comparison, here's my test image taken with a recent production Helios 40-2 at f/1.5. Distance between camera and image was approximately 1.9m, and focus was distance set at just above 1.15m. If you compare closely with the previous tests for the 44-series wide open, the 40-2 shows the swirl effect to be slightly more prominent, especially near the centre of the frame. As a result, the APS-C crop image shows the effect to a greater extent.

Full frame:



APS-C crop:

[Topsy]

A coating test would be a nice idea to, there seem to be quite a few (factory dependant?) around and I'd be quite interesting to see this effects on colour cast and flare resistance.

I sadly don't have a 42" monitor or I would happily contribute.

---

Last edited by Topsy; 01-30-2017 at 02:09 PM.

[BigMackCam]

I think I may start a separate thread showing test results for a range of lenses, but for the time being here is one more... the Zenitar-M 1.9/50. Great potential for swirl

Full frame:



APS-C:

[Topsy]

I just happen to have found a overlooked helios 44 (early 13 blade aperture one) on ebay and bought it so I'll be able to make a comparision of that and a 44k-4 on my 24" (if that works!)

[Audi 5 cyl]

Very nice results, somewhat expected.

I've been using the 44-2, 44m, 44m-7 and 44m-4 and never noticed any differences in swirl factor. I kept the one that was in the best overall state (44m-4) and gave away the others. What I did not do is compare them to Biotar 2/58 which I got far more recently but don't think it differs that much either. Images from Biotar come out with a bluish tone, and that's about it when it comes to real differences. Helios can be found for much less and in a younger package so I would definitely recommend it over other options.

[AldaCZ]

Thx for debunking!

[BigMackCam]

Thx for debunking!

Ha ha You're welcome

I will say, though, that the Helios-44 series, in all their variants, were manufactured for a loooong time. It's quite possible that the optical elements changed over time. My 44-2, 44M and 44M-4 were all manufactured within six years of each other - so maybe they're using identical glass, whereas much earlier or later copies of the same variants may have slightly different glass...