Home

Please support mflenses.com if you need any graphic related work order it from us, click on above banner to order!

SearchSearch MemberlistMemberlist RegisterRegister ProfileProfile Log in to check your private messagesLog in to check your private messages Log inLog in

Adjustable bokeh with Triotar
View previous topic :: View next topic  


PostPosted: Mon Mar 03, 2008 6:58 am    Post subject: Adjustable bokeh with Triotar Reply with quote

While examining the 4/135 Triotar, I noticed that the front cell can be screwed off, e.g. to facilitate access to the aperture blades. The threads are very fine, and the threading is rather long so that the cell can be gradually moved several millimeters forward like in a front cell focusing lens. I began to wonder what optical effect this had and started experimenting. Infinity focus was lost, of course, as the helicoid doesn't offer enough travel to counteract the front cell focusing effect. What about the bokeh? Well, here things get interesting. The bokeh of the Triotar isn't bad to begin with, there is just a slight bright edge. Here is a test shot:



With the cell about 2-3 mm out, the bokeh turned out to be Gaussian:



The change in the OFF verticals is quite interesting. At first it seems they are better in focus despite being further back, but this is just the effect of the Gaussian distribution. The center (i.e. the top of the distribution) is more distinct, especially evident in the light vertical, and the edge is wider. With a normal subject, e.g. flowers or twigs, this would mean that there wouldn't be any distinct edge "bands" anymore, all the linear edges in the OOF region would soften out.

This may explain the good bokeh of my front cell focusing 4.5/105 Radionar. In fact, this means that the Radionar bokeh can perhaps be improved even further if I set the front cell focus at the near limit and use the macro bellows or the variable close-up ring for focusing. As for the Triotar, I have to make a choice between having infinity focus and having an optimal near field bokeh -- no need for bokeh at infinity as there is nothing OOF past infinity.

Veijo

PS. Attila, the problem with your Triotar may be just due to someone cleaning the aperture blades and not properly re-inserting the front cell. This would explain the lack of far field focus.


PostPosted: Mon Mar 03, 2008 8:02 am    Post subject: Reply with quote

Mmm, very interesting pictures, Veijo. Thanks for sharing. Could you please post some resized full frame pictures?

I have some old 6x6 or 6x9 cameras with front focusing triplet lenses ("Zeiss" Novar, Agfa Apotar, Angénieux U1) and I will try them on a bellows.

Cheers!

Abbazz


PostPosted: Mon Mar 03, 2008 8:54 am    Post subject: Reply with quote

Fascinating experiment - thanks for sharing.


PostPosted: Mon Mar 03, 2008 10:03 am    Post subject: Reply with quote

Abbazz wrote:
Mmm, very interesting pictures, Veijo. Thanks for sharing. Could you please post some resized full frame pictures?


I'll do FF tests later on, also with the Radionar. I have the 350D with me and I took a couple of real life photos during my lunch break. The following are uncropped, down-sampled frames, first with the front cell at the nominal position, then a few mm out. The framing changes a bit as I don't have a tripod with me and also because the effective FL changes a little bit due to the required focusing change. I tried to focus at the same spot for both front cell settings, in the first pair of photos the focus spot is outside the frame.









The changes caused by the front cell adjustment are just what I expected after the highlight test, and the result is very much like a slightly exaggerated Radionar bokeh. This modified setup may also be a good portrait lens.

Veijo


PostPosted: Mon Mar 03, 2008 1:18 pm    Post subject: Reply with quote

i made the same test with pentacon 135.

When i dissambled my pentacon 200 i also noticed a sort of "soft focus" by unsecrewing the front lens.

I tried to do the same on 135mm but failed.


PostPosted: Mon Mar 03, 2008 2:41 pm    Post subject: Reply with quote

The gaussian distribution is clearly superior. I suspect that a position that is maybe 3/4 the way to your gaussian position would be even better, reducing the central highlight.

Also, from your experiments and thinking back on photos that I have disliked, the type of busy, squirmy bokeh that I most hate is probably caused by overlapping and reinforcing edges.

I have sometimes wondered about taking two shots, one with and one without an (in-focus) subject; using the two to generate a mask for knocking out the background; convolving in various ways the background image, and then inserting the masked-out foreground subject.


PostPosted: Mon Mar 03, 2008 4:06 pm    Post subject: Reply with quote

Sounds a little like the Orton effect?
http://pcin.net/update/2006/11/01/the-orton-effect-digital-photography-tip-of-the-week/


PostPosted: Mon Mar 03, 2008 4:12 pm    Post subject: Reply with quote

hacksawbob wrote:
Sounds a little like the Orton effect?
http://pcin.net/update/2006/11/01/the-orton-effect-digital-photography-tip-of-the-week/


Yes. The crucial point here though is the convolution. Unlike a simple gaussian blur, it means that the distribution from a point to an out of focus area can be precisely controlled and experimented with.


PostPosted: Mon Mar 03, 2008 4:42 pm    Post subject: Reply with quote

ChrisLilley wrote:
The gaussian distribution is clearly superior. I suspect that a position that is maybe 3/4 the way to your gaussian position would be even better, reducing the central highlight.


I've done some more highlight tests, and it seems there is no general optimum, everything depends on the distance relationships -- as usual when we are dealing with bokeh. Like most things in photography, mastering the bokeh will take a lot of practice. I'll have to do some more testing to verify my present gut feeling, but anyway it seems a rather small adjustment will get you a long way, say, something like only about a half of my initial adjustment. Moving the front cell two full turns out will leave a focusing range extending to about 20-25 m, which ought to suffice for most circumstances where the bokeh might be problematic.

Veijo


PostPosted: Mon Mar 03, 2008 4:45 pm    Post subject: Reply with quote

ChrisLilley wrote:
hacksawbob wrote:
Sounds a little like the Orton effect?
http://pcin.net/update/2006/11/01/the-orton-effect-digital-photography-tip-of-the-week/


Yes. The crucial point here though is the convolution. Unlike a simple gaussian blur, it means that the distribution from a point to an out of focus area can be precisely controlled and experimented with.


However, simulating distance dependent aspects of bokeh may turn out to be rather nasty -- to say the least.

Veijo


PostPosted: Mon Mar 03, 2008 5:18 pm    Post subject: Reply with quote

I wish I understood even some of what you are talking about. Embarassed Embarassed Embarassed


patrickh


PS I agree with Chris about the spiky/busy OOF. Dont like it either.


PostPosted: Mon Mar 03, 2008 6:10 pm    Post subject: Reply with quote

vilva wrote:

However, simulating distance dependent aspects of bokeh may turn out to be rather nasty -- to say the least.


Sure, in the general case that can only be done by 3D modelling. A degenerate case has a single, planar background; a single, mid-range foreground subject; no close-range oof areas. Subtracting two images deals with any haloing effects and ensures consistent lighting on the foreground subject. While artificial, it makes it tractable 9and is also a fairly common use case in practice).

The point of the experiment is to analyse different spread functions to see which is most pleasing.


PostPosted: Mon Mar 03, 2008 7:45 pm    Post subject: Reply with quote

patrickh wrote:
I wish I understood even some of what you are talking about. Embarassed Embarassed Embarassed


Once I will be done with this article, will go through this thread,
http://www.vanwalree.com/optics/bokeh.html

Shocked Very Happy


PostPosted: Thu Jan 06, 2011 5:16 pm    Post subject: Triotar on a Nikon DSLR Reply with quote

I've bought a Triotar very recently and I'm using it on the new D7000 with its APS-C 16MP sensor. Thus, the lens becomes a 202mm lens.

Anyway, the M42 adapter I have had an _awful_ infinity glass inside which I got rid of earlier as it affected a Takumar when wide open (f/1.4).

For those unfamiliar with M42 adapters on Nikons, you lose infinity focus (this Tritoar can focus maybe up to 15-20m - haven't really measured it) but I think it's also affected the depth of field as it's much more shallow.

Check out this example, there is a red glass about 0,5m behind the tap then a green bottle of olive oil 20cm further away (the white/green highlights you see just below the tap)

My point is that the bokeh is already in full effect at such a close distance (0,5m) giving the lens a macro like quality - not that I mind. Wonderful lens that deserves a clean up at the local camera repairman, as it has a bit of a rough focusing ring.


PostPosted: Thu Jan 06, 2011 7:51 pm    Post subject: Re: Triotar on a Nikon DSLR Reply with quote

Welcome Valli!


PostPosted: Sun Jan 16, 2011 3:46 pm    Post subject: Reply with quote

Cheers Attila! Awfully quiet here, isn't it Smile

Did a few more tests with the Triotar and I'm very happy with it. Haven't bothered mounting the glass back on the adapter to do a side by side test of the bokeh but I suspect I'm right.

The way it's set up now it's a wonderful macro lens, especially for its price!


PostPosted: Sun Jan 16, 2011 6:03 pm    Post subject: Reply with quote

thanks for posting this. looks interesting.

particularly, moving the front element of the triotar looks like it messes around with spherical abberration particularly. as a side effect, it does give a glow to the in-focus subject too.

nikon's dc lenses do similar things (although i'm sure a heck of a lot more sophisticated than unscrewing front elements) but the lenses are also stopped down according to the amount of spherical abberation introduced.

long post cut short: it would be interested at messing around with different amount of unscrewing and stopping the lens down.

i might have to get myself a triotar...


PostPosted: Sun Jan 16, 2011 6:43 pm    Post subject: Reply with quote

That's a very nice shot, Valli.


PostPosted: Tue Aug 16, 2011 11:23 am    Post subject: Re: Adjustable bokeh with Triotar Reply with quote

This is the Carl Zeiss Jena Triotar?

vilva wrote:
While examining the 4/135 Triotar,


PostPosted: Tue Aug 16, 2011 7:46 pm    Post subject: Re: Adjustable bokeh with Triotar Reply with quote

buggz wrote:
This is the Carl Zeiss Jena Triotar?

vilva wrote:
While examining the 4/135 Triotar,


Yes.

Veijo