[GUB]

So I have just got myself a Super Tak 85mm f1.8 and wanted to try and compare bokeh results from what I think of as my portrait lenses.

Bokeh is a hard thing to compare objectively.

But I think the take home from here is the longer the focal length the smoother the bokeh and the fact that focal length tends to override the effects of having a fast lens used wide open.

And of course the longer focal length tends to give a less distorted perspective of the subject.

I feel the moral of the story is go out and get yourself a 135 rather than paying over the odds for a fast lens.
Note that the 105 here gets some of its smooth look because the focused subject is framed a little larger meaning the background is relatively speaking further away.
First post has the lenses wide open

Attached Images

View Picture EXIF

File Date

 

View Picture EXIF

File Date

 

View Picture EXIF

File Date

 

View Picture EXIF

File Date

 

View Picture EXIF

File Date

 

View Picture EXIF

File Date

 

[GUB]

Oh and the images had just the exposures modified in post so they sort of matched - all colour temp at 5310.

Here they are at F4

[sergysergy]

I love your model. Very nice comparison of some very well regarded lenses

[GUB]

I love your model. Very nice comparison of some very well regarded lenses

She is a New Zealand Huntaway. She works the sheep and Bulls on my farm and in her spare time she is a professional model for me!. Every now and then she gets a bit uppity about all the photos but then I only need to turn to the cat and take a few shots and jealously kicks in!

[kiwi_jono]

Great series! I quite enjoy seeing how lenses compare bokeh wise as its something that is more subjective than technical.

They all have merit but of these photos, I love how the 85 f1.8 renders. The 8 element Super Tak 50, has a nice look to it too (I have a 7 element 50, which I like, but the 8 element does seem to render a little bit better).

[GUB]

I think if any lens is bumped off the list for portraits it is the Revuenon 55 1.2. At f4 it is sort of clinical and at f1 the bokeh really only has novelty value rather than presenting the subject in a pleasant surroundings.

[Dartmoor Dave]

Bokeh isn't something I ever really notice or care about, so leaving that aside I'd say that the 85mm/1.8 is in a different league to the others in its rendering of the dog. Beautifully rich but natural looking colours, with the fur texture rendered in great detail without any of the hard-edged pseudo-sharpness that's visible in some of the others, particularly the Vivitar.

Looks like it'll make one heck of a fine portrait lens to me.

[Paul the Sunman]

Yes, the Revuenon is ugly, even at f/4. To me, the 105 and 135 are the pick of the bunch for this subject.

[GUB]

Bokeh isn't something I ever really notice or care about, so leaving that aside I'd say that the 85mm/1.8 is in a different league to the others in its rendering of the dog

I feel the 105 comes very close. Or is it the slightly warmer rendering that both of them have that suits the dog?

[Dartmoor Dave]

I feel the 105 comes very close. Or is it the slightly warmer rendering that both of them have that suits the dog?

I agree that the 105 does come pretty close, but in those samples it seems a bit like it's relying on harder edge contrast from the coating to give a sharper look, rather than the real optical sharpness that the 85mm seems to have. I'd be very interested to see a comparison between the K105mm/2.8 and a Super Tak 105mm that I assume is optically identical. It would help narrow down how much of the difference is the glass and how much is the coating.

Thanks for a genuinely interesting comparison of some excellent lenses.

[WPRESTO]

Something as subjective as bokeh is pretty much impossible to quantify and although really poor bokeh will stand out, it usually comes down to "I like this better than that."

Many caveats for the following generalities, which means there are some lenses that defy the generalities, but these seem to be true.
1) wider apertures are better;
2) a well rounded iris diaphragm is better if the lens is stopped down, which usually means a large number of rounded blades;
3) Slightly longer focal lengths are commonly better simply because of limited DOF;
4) prime lenses generally have better bokeh than zooms, but I have two zooms that are very good.

Old prime lenses, especially before the advent of automatic diaphragm mechanisms of SLRs, commonly had large numbers of rounded blades on the iris diaphragm. For a short time I had an M42 mount Steihheil Cassaron 40mm f3.5, a Cooke triplet with 12 rounded diaphragm blades. Very smooth bokeh & transition from in to out of focus. It was, however, a very clumsy lens to use = manual diaphragm with no click stops and it had a big lug on the mount end that could catch on the diaphragm operating mechanism of a Pentax K-mount. I used it with a short extension tube to take some experimental close-ups.

[ZombieArmy]

I love the Tak 50 1.4, the images almost feel "out of this world", and the bokeh is incredibly good to go along with it. I think the 8 element version does render very cool, more so than my A version when I test them side by side. So I think that might be why it doesn't seem as nice when directly compared to the warm rendering 85 1.8.

If you had to stop down though the 105 is brilliant, can't be beat from this stack.

[ffking]

Patient model!

[BrianR]

Nice comparison and very patient dog!

I've put your lenses into the "How Much Blur Calculator". This calculates the size of the blur disks as a function of the distance behind the subject. It's assuming the same framing for the different lenses (I've assumed a 0.6x0.9m subject area, but this can be changed). I thought it might be interesting to compare how the calculated mashiness of the background (which is obviously only one aspect of the bokeh) compares with your practical example:

How much blur? - A visual background blur calculator

[GUB]

Nice comparison and very patient dog!

I've put your lenses into the "How Much Blur Calculator". This calculates the size of the blur disks as a function of the distance behind the subject. It's assuming the same framing for the different lenses (I've assumed a 0.6x0.9m subject area, but this can be changed). I thought it might be interesting to compare how the calculated mashiness of the background (which is obviously only one aspect of the bokeh) compares with your practical example:

How much blur? - A visual background blur calculator

Nice calculator. Trying to work out why it doesn't really reconcile with my shots. Especially the 55 1.2 which didn't smooth the background nearly as much as I expected. I think it has to be something to do with the coarseness or fineness of the background material. The blur in the top of the image is a Wisteria about 4 metres behind the dog. Its 60 -70 mm leaves are arranged in an "opposites" arrangement down the stems and I think that may have set up a interference pattern in the 55 1.2. I think it is this "interference pattern" type effect that make great bokehs so unexpected at times. The other thing not scientific about my setup is that the centre of the background is further away and less detailed and so the telescopic lenses concentrated on this area. Thinking about doing one out in a field with the fine detail of grass as a backdrop.