Hey all, so I have been reading about the new blur tools in CS6 and how they can mimic the shallow depth of field of a fast aperture lens much more realistically than anything before. Now of course a fast aperture lens gives you the ability to achieve much higher shutter speeds when shooting at say F1.4 vs the standard F2.8 of a quality zoom, but my question still stands:

Are there folks out there deciding not to shell out the big bucks for F1.4 glass, or deciding to sell their expensive 1.4 glass, because they can get fast enough shutter speeds most of the time with F2.8 lenses and can then mimic the shallow DOF with photoshop? I have some awesome F2.8 zooms, and am teetering about whether to plunk down the cash for some F1.4 primes in addition. I could see the photoshop route being advantageous in that you won't miss focus as often with the original shot, or get part of the subject in focus but some out of focus, due to the shallow DOF of say F1.4. You can then be sure your subject is sharp and blur the background as needed with these new CS6 tools. Looking forward to hearing thoughts and opinions about whether F1.4 glass is less necessary now with this new and pretty dang advances post processing software in CS6.

Best -L

I'm sure you could achieve a very similar blur effect with postprocessing, but I see 2 reasons for sticking with fast primes

1. Reduced processing time. Sure, you could do this for selected shots, but do you really want to spend the time/effort required to do this for shot after shot - are you a shallow DOF junkie? If you're only doing it for a couple of shots, the post-processing route may make sense. How much is your time worth?
Some would argue that the fast glass has a character that is difficult to replicate.

2. Shutter speed in low light. As you say, sometimes you need the large aperture to keep your shutter speed up. Though with the improved high iso performance of newer bodies, f/2.8 may be enough. You can also add flash.

The increased DOF from shooting at f/2.8 does give you more leeway in case of missed focus though.

While the diff between 2.8 & 1.4 is quite significant in both light gathering and DOF ... the diff between 1.7 or 1.8 and 1.4 isn't nearly so much. Unless you have a specific need for the speed of a 1.4, the 1.8 version can be a marked improvement over a 2.8, yet at a cost that isn't nearly as prohibitive. Adding a little PP blur from a 1.8 is likely to have a more natural look than the need for more blur from a 2.8 as a strategy.

There are a few other things to note regarding ultra fast glass is that it will have to be heavier & larger, and for some that means a slower performing AF (more glass to move).

But, that being said ... I don't think I'd opt for a steady diet of 2.8 + PP if shallow DOF was gonna be "my thing".



Shooting with larger apertures requires better focusing skills. Replicating natural looking DOF requires better PP skills.

Choose your poison.

I was messing with this the other day in OnOne's Focal Point trial version. Supposedly it is one of the better tools for this but in a multidimensional type scene with two subjects at different distances from camera, then a very branchy tree behind them and then a building beyond that I was having trouble making it look convincing to my eye.

Funny thing is that sometimes people who aren't photographers (and even some who are) sometimes think that you photoshopped in actual bokeh that came from using a wide aperture. I posted one pic on Facebook a while back and a friend said, "nice pic, did you use a blur?" whatever that means!

I have also posted some actual tilt-shifted photos that have prompted photographers to comment on the "photoshopped tilt shift"

brad

thanks for weighing in guys, much appreciated on the insights. I will have to see how important the fast shutter speeds of 1.4 are for me as opposed to my 2.8s. Thanks! -L

I tried a free download once and I did not like producing blur in post. It is OK on occassion but not a regular way to go. I get good bokeh from F1.8 lenses like the new Nikon 85mm F1.8G.

Advantage of the software approach: you could have more DOF in the part that you want to be sharp than by using a very wide aperture at capture (unless you use a Carl Zeiss lens).

Disadvantage of the software approach: it doesn't know the actual distances of each photo element relative to the subject and cannot be relied upon to get the blur right - either in terms of which elements to blur or in terms of different amounts of blur for different distances.

- Alan

williamkazak wrote:
I tried a free download once and I did not like producing blur in post. It is OK on occassion but not a regular way to go. I get good bokeh from F1.8 lenses like the new Nikon 85mm F1.8G.

+1

I am lucky to have both a Canon 50/1.0L and a Canon 200/1.8L
I can affirm that needed post work is minimal.

I have never, for one second, regretted purchasing fast glass.

The bokeh on my 85mm 1.4 is something I have yet to see a piece of software produce.

This shot is from my most recent wedding, I was actually just editing it before taking a break and browsing the forums...

Fast glass is so nice when you nail the focus and don't get veiling flare. Nice. I had the Nikon 50mm F1.2 AIs and that was very difficult to do with an F3 body.

DOF is a matter of focal length, aperture and subject distance. This was taken at f8. You don't always need real fast glass.

runamuck wrote:
DOF is a matter of focal length, aperture and subject distance. This was taken at f8. You don't always need real fast glass.

absolutely, I'll admit only 1.8 glass I have is 50/1.8 (Nikon) but I have great shots with 70-200/2.8 when at 2.8 and minimum focal lengths. Even more so with 105/2.8 since it's a macro lens.

I'm just not good enough in photoshop nor do i enjoy post-processing to speak to that method. LR4 does all I need to do.

Of course, like all things photography, one day I'll have that fast glass. . . .

You can do it in Photoshop but it looks fake to trained eyes no matter how good you are at Photoshop. However, for your average visually illiterate Joe6Pack clients, they can't tell the difference.

Purists might like the effect of wide apertures but I suspect a lot of people are equally happy with software versions of bokeh and with some experimentation particularly using layers results can be achieved quite quickly.

It is also more predictable.......there is always some element of luck shooting wide open but stop down a bit and the image is more likely to be spot on so far as focusing goes.......it allows one to fine tune the image later...at leisure.

I have a theory that software is the future, not lens design. There are several reasons for thinking this.....but my experience with LR is what convinced me.

My first DSLR was a 10D and processed in LR1. Then I got LR3 then LR4. I go back and re-do some of the early files.....and each LR progression gives better images..the original ones look quite poor by comparison.

What will software be able to do along with even better sensors in say 5 yrs time?

In not too distant future, software will do any visual effects you fancy automatically on either still or motion. Today's professional retouchers will have to find a new profession by then.

http://vimeo.com/22223687

http://www.youtube.com/watch?v=zkH8snwmfMA

mshi wrote:
In not too distant future, software will do any visual effects you fancy automatically on either still or motion. Today's professional retouchers will have to find a new profession by then.

http://vimeo.com/22223687

http://www.youtube.com/watch?v=zkH8snwmfMA

I think we are close in opinions......I don't doubt the f1.2 lenses have a special feature.....but I can get close to duplicating something similar with not much effort on my computer.....and I wonder how many pros have clients that will recognize the subtleties of difference and want to pay the extra for the financial investment.

These lenses are leftovers from the film / pre software era and while not knocking their aesthetic merits I am suggesting these merits are now available to competent Photoshop users ..at a fraction of the cost.

And as mentioned above.....in 5 yrs?