135 softfocus verdict

Double Negative wrote in post #7834094
"Meh." I prefer the L... It's cheap enough, for an L - and really delivers the goods.

so says the man with all but 1 L.

rdenney wrote in post #7835502
While I recognize that the effect will be subjectively evaluated, I do disagree that it takes more time and practice in Photoshop than in the camera. One of the problems I had wiht the lens was that the effect could not be evaluated very well either on the ground glass or on the LCD. Considering there are two soft-focus levels (in addition to "none"), and considering that the effect varies substantially over the range of f-stops, there is a continuous range of effect that would require considerable experience to learn, and it must be learned if the viewfinder or LCD can't be trusted to represent it properly. I found that the viewfinder and the LCD both grossly under-represented the effect, especially on the image details.

In contrast to that, make a duplicate layer, apply a Gaussian blur with a 100-pixel radius (play with that value until it looks right), and then play with the transparency of that layer until it looks right. That takes about 30 seconds tops, and the result can be seen large on the color-corrected computer monitor. It can also be turned off.

I prefer the effect in Photoshop, but as I said, that's a subjective evaluation. Neither is a true replacement for the ultimate soft-focus effect provided by a Zeiss Softar filter, but the PS approach comes closer.

Rick "it isn't always easier to do things in the camera" Denney

I don't agree with you here, Rick.

First of all, although there are 3 click-stops for effect, namely 0, 1, and 2, it can actually be used in any position in between. And IME, even 1 is often too much already.

I also disagree with the PP thing. The effect of this lens just looks different. IMO, it is like trying to do a tilt-effect in PP, rather than use a lens to achieve this. It is possible, but at a price, the least of which is the time to do so.

Yes, a Softar is different, but then, it works differently. That would make quite a difference indeed, it is a different effect. Also, the SF effect of a Softar varies with different apertures too, so that is similar for both. Whether you prefer one over the other, is a matter of personal taste.

Regarding the underrepresentation of the SF effect on the focusing screen: that is true for the Softar too. Unless you use a very finely ground matte, and LF, you can really only properly judge the effect in PP / lab, whether shooting analog or digital, so IMO that is a bit of a moot point.

If anything, you can judge it quite well using Live View, at larger magnifications.

My summary: to use SF effects effectively, you need to know the tool very well and the effect small changes have. There is no real substitute for that knowledge, unless you're shooting digitally and tethered, in a studio.

And pardon me for pointing this out, but I don't think you know the 135 F/2.8 SF tool for SF uses well if you didn't discover that there are more settings than just 0, 1 and 2 .

Kind regards, Wim

I agree with what one of the above posters said: a setting of "1" is often too much. When I use this lens, it is often set 'just barely off the zero click'. Also, if it has not been mentioned, the effect of soft focus is greatest wide open, and as you stop down, the effect dimishes. Right after I bought the lens, I snapped some pix of the bricks on the front of the house at near min focus distance to see how sharp it is when set to zero. Pleasently surprised: it's actually very nice even wide open.
If you like the focal length in a relatively fast prime, that does not cost a lot of money, go for it.

wimg wrote in post #7837013
I don't agree with you here, Rick.

I'm trying to figure out what you wrote that rebutted with what I said.

1. Considering the settings of the soft-focus effect and the variability of that effect with aperture, it takes a huge amount of experience and guesswork to predict what the effect will be. That one can use the intermediate settings on the SF dial makes that point more strongly. The issue here is that you can't tell what you have at exposure time. The training required to know how to set it is at least as much and in my experience quite a bit more than with other tools.

2. If you like that effect and not the effect achieved using other tools, then my own opinion doesn't matter--you'll have to invest the time required to learn it. I don't like the effect, but I acknowledged the subjectivity of that and gave specific examples of why I didn't like it. Unlike others, I actually linked to a thread where real examples were shown. That empowers people to choose for themselves.

3. Not all of us have Live View, but even if we did, it isn't likely to show the effect any better than chimping, right? I can see the effect of a Softar on the ground glass of the cameras I used it on, but I grant those are larger cameras with better viewing. Even so, that still recommends in favor of doing it in PP, which provides greater creative control. My point was not, however, to assert that other tools were viewable at exposure time, but to point out that if you can't predict the effect at exposure time, you'll either 1.) have to spend lots of time learning the lens, or 2.) expect failures. Unlike with Photoshop, if the effect is overdone in the field there's no going back.

3-1/2. I don't agree that applying a Gaussian blur on a transparent duplicate layer in
Photoshop is analogous to simulating a lens tilt in Photoshop. Simulating real fog in Photoshop is--that I agree. Real fog in the scene has a more pronounced effect the more deeply you go into it, which is the same as attempting to simulate a tilted focal plane. But spherical aberration in the lens, a Softar or other purpose-made filter, vaseline on a clear filter, a woman's nylons stretched over the front of the lens, or the application of Gaussian blur (and I've done them all) treat the bright and dark areas of the image equally without regard to how far they happened to be from the camera when the exposure was made. Thus, they are all significantly (and equally) imperfect as a simulation of real fog. But for achieving the "glow" effect, they all present a different effect and the choice of one to the next is purely subjective. Applying a Gaussian blur does not distort or smear pixels the way simulating, say, applying a perspective envelope to simulate a shifted lens might. It certainly doesn't throw anything away compared to what gobs of uncorrected spherical aberration might throw away.

4. Unlike many who post opinions on the Internet, I do actually own and have used this lens. I did not explore its soft-focus tendencies once I realized it was not the effect I was looking for. At that point, I left the dial at 0 and used it as a general-purpose 135 prime, but later, in detailed comparisons, determined that I just didn't like its rendering compared with other options, some of which are cheaper and some more expensive. I don't need a 135 to just be sharp--most $20 135's are sharp when used appropriately--I want it to bring something more than what I get with a zoom. This one doesn't. It's a competent 135 prime, no more, no less, and these were the sorts of lenses that were the cheapest accessory lenses back in the old days. At $200 for a used one, it's functional but I would not call it a "unsung gem".

Rick "wondering where the rebuttal was" Denney

Hi Rick,

rdenney wrote in post #7840242
I'm trying to figure out what you wrote that rebutted with what I said.

Sorry, I realize after rereading that what I tried to put across wasn't very clear, because I set off with one thing and then just continued with other things, and tried to jest a little too, and unsuccessfully at this, I must admit.

Anyway, I didn't agree with the three settings of the lens, that was all, to be very honest.

I found it quite ok, actually, but we all prefer different tools.

Yes, I know. Sorry I confused the hell out of you and everybody else. Personally, I do like the effect. I shot with one quite a few years ago, on film, loaned from someone else, and I think it was amazing (provided I didn't overdo it). A while back I had another one on loan for three weeks, and I still loved it, but didn't have a real use for it in the end (not into SF anymore), and returned it to the original owner (who wasn't using it either).

Not entirely. If you use the DoF Preview method and zoom in, it is quite clear what it does, especially on the cameras with high resolution LCDs.

Yes, I referred to those .

Yes, to a degree, but I also think it is fun to learn one's tools, but that does mean one has to like it, and see the need.

No, but that is why there is a need to learn the tool. I still don't think you can easily obtain the same results in PS.

As you already mentioned, using SF on this lens is introducing spherical aberration. To try and do this in PS, is to me the same as trying to simulate tilt. It is not the same.

This to me is an analogy like I meant. You can't create real fog, but neither can you recreate spherical aberrations, or tilt.

I've done all of those too, which is why I like the 135 SF approach . And we really seem to disagree on the spherical aberration effect. IMO you can't simulate that in PS. If that were the case, you could recreate good bokeh in pictures with PS as well.

Yes.

Well, here we differ in opinion too. I don't think it is "gobs of uncorrected spherical aberration", because it actually is controllable. "Gobs" implies it is not. The neat thing is that it is influenced by aperture and distance, similar to bokeh in lenses like the 85L and 50L, just that you can go overboard with this with a lens like the 135 SF. The only difference with this lens and those with built-in spherical aberrations in this regard, is that with the 135 SF it is controllable, not with the other ones.

No need to go this route, yours are also opinions, and I respect those as I respect you. I certainly did not mean to upset you, quite the opposite. I never actually owned the 135 SF, but did borrow one on several occasions, and did use it almost exclusively in those periods.

Fair enough. It was the exact thing I was looking for and liked. I didn't get the effect I wanted from the other tricks, I thought it was too artificial looking.

That's a matter of opinion only. At its price, being a 135 F/2.8 with AF, it is very good, IMO. It is in the same league as the 35 F/2, the 50 F/2.5 CM, the 28 F/2.8, the 24 F/2.8. From that POV, and its price, especially compared to an F/2.8 or even F/4 zoom, IMO it is a very good 135, provided you need or want one.

Well, I explained that higher up . Sorry for the mixed signals.

I do hope I made myself more clear this time.

Kind regards, Wim

wimg wrote in post #7840489
Hi Rick,

Sorry, I realize after rereading that what I tried to put across wasn't very clear, because I set off with one thing and then just continued with other things, and tried to jest a little too, and unsuccessfully at this, I must admit.

My response was a bit brusque--I went into debate mode.

Clearly, what is needed here, is for me to send you my 135 SF, heh, heh. If it didn't need servicing, I'd make you buy it.

The comment about rendering opinions by non-owners wasn't directed at you--I had not figured out that you didn't own the lens. It was directed at many who offer opinions on the Internet with no actual personal experience. I always frame my opinions by being clear about what I personally know to be true, and I've always respected your posts as doing the same. Would that everyone followed that principle.

The first time I used the lens, I dialed in a bit of the soft-focus effect and used a moderate aperture. In the finder (even wide open), and on the LCD, the effect was slight and I thought about the right amount. But when I uploaded it and viewed in on a monitor, the effect seemed live five times what I thought it would be. Getting just a touch of SA to improve bokeh or just take the edge off seemed to me like it would require psychic ability. If you can do it, then I really should send you my lens where it could do some good.

(I did, by the way, include this lens in my bokeh test but without any SA dialed in. The bokeh at the 0 setting was bettered by the excellent 70-200/4L, I thought, though it wasn't bad. Of course, the zoom is much more expensive.)

Rick "wondering if the SF effect would be easier to see on the ee-S screen" Denney

nureality wrote in post #7836041
so says the man with all but 1 L.

Okay, busted.

I didn't mean to come across as saying the lens is bad in any way; it's a great lens for the money. But I think few people really use the SF feature. The 135L and the 200L are more money, obviously - but a bargain as far as L lenses go. So it might be worth splurging on them to some.

Canon used to offer three or more versions of the 135mm back in the FD days of varying aperture. Even then, the f/2 version was "the bomb." But the slower versions were pretty good as well for less money. I don't recall offhand if they had the SF version as well, but that'd have been a fourth option.

I've owned one, great lens for the price. VERY sharp at 2.8. The AF is slow and quite noisy. It's hard finding a different lens you consider when price is in the picture. The 100mmf2 is better, but it's shorter and more money. Personally, I quite liked the 135SF.

rdenney wrote in post #7855532
My response was a bit brusque--I went into debate mode.

Thank you for explaining that, I got a little worried.

No need for that. I am actually looking for a 135L currently, but having a hard time finding a good used one over here, even a new one at a reasonable price is very hard right now. Had one for a few days but I returned it. Slight decentering, -7 on the micro-adjustments, and a problem filter thread.

Thank you, and yes I agree with that. Sometimes I make a statement about lenses I haven't owned all that long, however, or used all that long. I didn't need the 135 SF at the time, too long and covered by 70-200 anyway, so decided not to keep it, which was fine with the owner as I had it only on loan anyway.

The effect looks less strong because of the VF, which shows a smaller aperture effect than a fast lens, even with a fast lens mounted, as someone once explained here somewhere I think. The standard focusing screen makes the image look like the lens is stopped down to F/5.6, DoF-wise. The funny thing is that this makes the VF much brighter, than a much more finely ground focusing screen, where you can actually see the DoF more truly.

No need for it, currently .

I reckon it would. Still not perfect, however. I think it displays DoF as if it is at F/2.8, which should make it perfect for the 135 SF however.

Kind regards, Wim

If you find the SF effect too strong and drastic, could it be that you didn't refocus after moving the SF ring? Moving the SF ring changes the focus distance. Refocusing is critical to get the "softness over sharpness" that Canon talks about. The image shouldn't go blurry but it should add halation and reduced contrast.

That said I mostly use it as a sharp 135mm ordinary prime. Sometimes I try some SF on a scene if I think it would look cool but I always also take a sharp shot in case it doesn't look good.

rdenney wrote in post #7855532
The first time I used the lens, I dialed in a bit of the soft-focus effect and used a moderate aperture. In the finder (even wide open), and on the LCD, the effect was slight and I thought about the right amount. But when I uploaded it and viewed in on a monitor, the effect seemed live five times what I thought it would be.

It's something I also had to get used to: the effect varies greatly depending on how you enlarge the image. Even with a big color corrected monitor it's hard to judge what the image looks like once you print it 4x6".

rdenney wrote in post #7855532
(I did, by the way, include this lens in my bokeh test but without any SA dialed in. The bokeh at the 0 setting was bettered by the excellent 70-200/4L, I thought, though it wasn't bad. Of course, the zoom is much more expensive.)

Actually the bokey is one of the things the 135SF does really well, especially if you dial in just a tad of soft focus. IMHO, bokehwise, it compares well to my 85L, especially in the region close to the focus distance but slightly OOF.

rdenney wrote in post #7855532
Rick "wondering if the SF effect would be easier to see on the ee-S screen" Denney

As mentioned above, softfocus looks very different depending on enlargement (it doesn't matter whether you create that effect via SF or PP). That's why even live view won't help you much.

There are three things which confuse me (being an old grumpy analog shooter):

1. Why are people proud to have "mastered" the 85L but unwilling to learn how to work the 135SF? Is it just because the 135SF looks outdated and costs a fraction of the 85L?
   
2. Why are people so upset if a lens needs figuring out anyway? At least for hobbyists like me it's part of the game. And if a pro is unwilling to learn he'll be obsolete pretty quickly.
   
3. I always hear that DSLRs are so great because you can try around. But as soon as there is a lens that takes practise or multiple shots people complain.

Rudeofus wrote in post #7860717
There are three things which confuse me (being an old grumpy analog shooter):

1. Why are people proud to have "mastered" the 85L but unwilling to learn how to work the 135SF? Is it just because the 135SF looks outdated and costs a fraction of the 85L?
   
2. Why are people so upset if a lens needs figuring out anyway? At least for hobbyists like me it's part of the game. And if a pro is unwilling to learn he'll be obsolete pretty quickly.
   
3. I always hear that DSLRs are so great because you can try around. But as soon as there is a lens that takes practise or multiple shots people complain.

Very confused ...

I find myself agreeing whole-heartedly here. Was there this much whining when IS was introduced? It was something "new" at the time. Or when new cameras are introduced with new features that have to be figured out?? It comes off as a little bit elitist to dismiss a lens or feature like SF just because one can afford $1,000 L glass and an $800 piece of software that can sort-of reproduce the in-camera effect. Or to act as though the SF effect was sssooooooo 1980's and haven't we all moved on to the next hot trendy thing already. I like the unique challenge of something like SF, and find this a refreshing thread to read in the place of the zillion "My bokeh is better than your bokeh" threads. Now THAT gets a little bit old after a while!

Most people don't buy this lens for the SF feature. They buy it because its an AF EF Canon prime that gives you f/2.8 at 135mm for about $200 on the used market, which is 1/4 the cost of the 135L.

lsuber wrote in post #7861029
I find myself agreeing whole-heartedly here. Was there this much whining when IS was introduced? It was something "new" at the time. Or when new cameras are introduced with new features that have to be figured out?? It comes off as a little bit elitist to dismiss a lens or feature like SF just because one can afford $1,000 L glass and an $800 piece of software that can sort-of reproduce the in-camera effect. Or to act as though the SF effect was sssooooooo 1980's and haven't we all moved on to the next hot trendy thing already. I like the unique challenge of something like SF, and find this a refreshing thread to read in the place of the zillion "My bokeh is better than your bokeh" threads. Now THAT gets a little bit old after a while!

Very true!

I'm going to try and get hold of one of these now

Rudeofus wrote in post #7860717
There are three things which confuse me (being an old grumpy analog shooter):

1. Why are people proud to have "mastered" the 85L but unwilling to learn how to work the 135SF? Is it just because the 135SF looks outdated and costs a fraction of the 85L?
   
2. Why are people so upset if a lens needs figuring out anyway? At least for hobbyists like me it's part of the game. And if a pro is unwilling to learn he'll be obsolete pretty quickly.
   
3. I always hear that DSLRs are so great because you can try around. But as soon as there is a lens that takes practise or multiple shots people complain.

Very confused ...

I hope you aren't talking about me. My analog old-fart credentials are pretty good!

I wouldn't have minded learning the effects of this lens had I liked the rendering of the effects. That rendering is important to me but I certainly don't expect others to share my own preferences on the matter. By the same token, I had visualized an effect, and the lens didn't match my visualization. That is an easier position to defend artistically, it seems to me, than spending a considerable amount of time learning a lens just because one should be willing to learn. I have lots of lenses that teach me lots of things, but in the end they must serve my visualization, not dictate it.

And it's not a matter of being professional--pros are less likely to waste time on something that doesn't match their visualization than are amateurs. For them, "time is money" is not just a cliche.

Rick "seemingly a lot less upset than those complaining of my being upset" Denney