Tilt shift lenses; your thoughts?

Messages
1,756
Edit My Images
No
I was just wondering whether many (any?) people around here used a tilt-shift lens and on a moderately regular basis, I would be interested to read your thoughts/experiences on using one as I'm considering one for my (predominantly) landscape work, though appreciate one lens can be used for so may applications...
[apologies if this has been discussed before, I can't seem to get the search function on here to work - keep getting errors!]
 
Now we have entered the digital age, I can see no pressing need for one.
I certainly can not see the advantages for land scape shots.
 
They look like good fun. There have been a couple of shots on here (tried just searching for tilt shift, and none of the ones I were thinking of turned up).
It is possible (but possibly not as much fun) to reproduce them in photoshop, which is what I do, as I cannot justify the cost of a lens at the moment
 
I've just ordered a 17mm TS-E (Canon) primarily for landscapes. It will provide the equivalent of around 10mm on a FF body at full shift, so will be useful for panos, as there will be no parallax issues. I've done some research and been in contact with David Summerhayes, who wrote this article http://www.luminous-landscape.com/tutorials/focusing-ts.shtml
and he is of the opinion that its possible, with the right scene to get everything in focus from 6" to infinity.
Its not for everyone, but I am looking forward to getting mine as I believe it will provide a lot of creative opportunities.
 
Err, surely that will just be 17mm lens on a FF body, how did you work out that it would be 10mm?, as I though all lenses were given focal length still based on the 35mm/FF system.

I'd love to have a go with a tilt shift, but at the moment there are other lenses that I want first. Although, to be honest until I read this thread I didn't realise the potential of a tilt shift, as all the stuff I'd seen was the 'faux tilt shift' churned out from photoshop.
 
Err, surely that will just be 17mm lens on a FF body, how did you work out that it would be 10mm?, as I though all lenses were given focal length still based on the 35mm/FF system.

It'll be the equivalent AOV achieved from using the maximum shift, one side to the other.

Bob
 
How are they used for landscapes?

That depends on a landscape itself. In general they are not that useful, however in some occasions they may be. The example that I have seen in a few books that comes to mind is when you include a tree in a landscape with roots going all the way to the bottom of the frame and your hyperfocal distance is not close enough to include those. This example is for example in an Ansel Adams Camera book and in a few other sites. Similar sample photo of what I am referring to is here.
 
I was just wondering whether many (any?) people around here used a tilt-shift lens and on a moderately regular basis, I would be interested to read your thoughts/experiences on using one as I'm considering one for my (predominantly) landscape work, though appreciate one lens can be used for so may applications...
[apologies if this has been discussed before, I can't seem to get the search function on here to work - keep getting errors!]

I always wanted to try them (mainly interested in tilt capabilities as shifts are useful mainly for architecture and I don't shoot many of that) but found the price of specialised ones prohibitively expensive. So I got myself an old Pentacon 6 lens (medium format) and tilt adapter from Arax (they also do shift adapters as well) for Nikon F-Mount. So now I have a 90mm F/2.8 tilt lens with 360 degrees movements and up to 8 degrees tilt. Works a treat, costs about 100 quid in total but is bloody difficult to focus ;). For me, to play with, this seems like the best solution without the need to shell a lot of cash for a proper tilt/shift lens.

After using this for all kind of flower shots, I can say that it is easy to use only if you want the "miniature" look (where you tilt to have sharpness in one limited area of the frame only). If you, however, want to extend sharpness by tilting, it is a very hard work - apparently there are books written on how to focus with tilt lenses. The difficulty is that when tilting the DOF is no longer shaped like a sandwiched area (where you have everything in focus between "from" and "to") and is more like a cone or V shaped area which depends on a tilting angle. This makes it hard to predict what will be sharp and what is not. So I always end up with a few test shots and peeping at 100% view on camera LCD to see how that DOF is distributed and then correcting and going through that again and so on until I get it right.
 
I think T&S lenses have less uses with small format digital, and they are fiddly to use. I can see tilt being useful for depth of field in specialised studio photography - still life, product shots etc. And when there are lot of images to produce it's often quicker to do shift corrections in-camera rather than post processing. But there are less uses for landscape - mega depth of field is the only one I can think of. Shift perspective effects can largely be replicated in post processing. Basically diminished applications these days, fiddly, and very expensive.

T&S is invaluable in large fomat photography as depth of field is so narrow you need all the help you can get. Not so much with small format - just stop down a bit more ;) Perspective control is also done very easily in camera, but if you scan the film that is now an easy Photoshop job, too.

Unless you really need the unique way in which T&S lenses swivel the depth of field plane for pictorial effect, they seem to me to be something of a hammer to crack a nut these days.
 
Now we have entered the digital age, I can see no pressing need for one.
....

I suppose that depends whether one's interest is photography or image creation. TV has eliminated the need to be present at sports venues, virtual reality will eliminate the need to visit places of interest and IVF has eliminated the need to copulate but I suspect there will be diehards who'll always continue to do one or more of these things the traditional way.

Bob
 
Not fullt a tilt-shift but I have a lensbaby which gets regular use.
Does this allow shift, or just tilt ?

For panoramic pictures I still see the uses of shift photography being important, just way beyond the amount I could justify spending.
 
thanks for your thoughts all. Interesting discussion I think. If you consider the plate camera+bellows set ups they all tend to come with the ability to tilt and shift the lens on the front, and since a number of landscape photographers, such as Joe Cornish to name one off the top of my head, use a system with this 'in build' flexibility in the camera design I was curious as to it's application into the realm of 35mm (i.e. full frame digital, or cropped even etc). The thinking being perhaps it would be similar?

On the subject of adaptors - rather than a specific t-s lens - my guess would be a bellows set up with a lens mounted on the front and a system specific attachment to mount the camera on the rear. effectively the same as a t-s lens only rather retro approach. it that something widely available? sounds like a cheaper option at least to get started and see about it all...
 
On the subject of adaptors - rather than a specific t-s lens - my guess would be a bellows set up with a lens mounted on the front and a system specific attachment to mount the camera on the rear. effectively the same as a t-s lens only rather retro approach. it that something widely available? sounds like a cheaper option at least to get started and see about it all...

It all sounds nice but in practice there are two obstacles:

1) The bellows that allow tilt and shift movement for 35mm are really rare and usually quite expensive. And I don't know of any of them (even the most flexible ones) that allows full set of movements into all directions without any restrictions

2) Using 35mm lenses or enlarger's lenses your use of the bellows on 35mm cameras will be limited to macro only. To get sufficient coverage to focus at infinity and cover the large area to accommodate for the movements, you'd need to use MF or LF lenses and they simply won't be wide enough on 35mm for landscapes.
 
It all sounds nice but in practice there are two obstacles:

1) The bellows that allow tilt and shift movement for 35mm are really rare and usually quite expensive. And I don't know of any of them (even the most flexible ones) that allows full set of movements into all directions without any restrictions

2) Using 35mm lenses or enlarger's lenses your use of the bellows on 35mm cameras will be limited to macro only. To get sufficient coverage to focus at infinity and cover the large area to accommodate for the movements, you'd need to use MF or LF lenses and they simply won't be wide enough on 35mm for landscapes.

^^^ Spot on (y)

I think Joe Cornish uses a large format field camera for different reasons - quality, working method and a bit of ethos mystique that helps to sell fine art prints ;)

Tilt and shift movements on large format is altogether a different proposition, as the cameras are much larger and you have room to manoevre. Plus the fact that as depth of field is so shallow, you need them!

It would be interesting to compare a Canon 1Ds3 or 5D2 and one of the new Canon T&S lenses with large format. Results aside, the Canon would lose hands down on magic factor, as Canon Bob explained so beautifully in this quote above:
I suppose that depends whether one's interest is photography or image creation. TV has eliminated the need to be present at sports venues, virtual reality will eliminate the need to visit places of interest and IVF has eliminated the need to copulate but I suspect there will be diehards who'll always continue to do one or more of these things the traditional way.

Bob
 
Ah yes I suppose it makes sense on the large format frames to have this, thanks to the shallow dof. Didn't consider that one, good thinking. So, on 35mm it's more of a perpective correction lens ie for architecture and so on rather than for landscapes and the like.
 
Ah yes I suppose it makes sense on the large format frames to have this, thanks to the shallow dof. Didn't consider that one, good thinking. So, on 35mm it's more of a perpective correction lens ie for architecture and so on rather than for landscapes and the like.

I would say it's more for swivelling the depth of field plane, which can not only give a unique look, but also allows you to get the same sort of depth of field you can usually only get with wide angles.

The perspective corrections you can usually do in post - simply straightening coverging lines in the main. But you can do more with montage/panorma techniques. An interesting interior shot like that was posted here recently as a bit of a puzzle picture as to how it was done - I think it might have been by Terrywoodenpic (not certain of that - might try to find it and post a link).

Found the link. Not sure how relevant it is, but if you like toilets :)
http://www.talkphotography.co.uk/forums/showthread.php?t=147492&highlight=terrywoodenpic
 
showphoto.php
[/url][/IMG]

I borrowed Bob's 24mm tilt shift and used it to do some architectural photography for which I have to say, I found it very useful. I know that I can sort converging verticals is PP but then I lose a rather large chunk of the original image. So I found it really useful for sorting those verticals out in camera and retaining as much of the original scene as possible. I'd love to see what the 17mm will do! :)

It's a funny example though because for all I managed to sort out all the verticals, the builders of Manchester Cathedral could not sort out their horizontals. The steps and the windows are hardly parallel.:shrug: Turns out that back in those days they could use a plumb bob for verticals but did not have an equivalent tool for levelling horizontals. Short of rolling coins on surfaces and not many builders had an excess of coins.
 
In your example above Ali, you have not fully sorted the verticals. Was the lens on max shift? This might be a drawback with T&S lenses vs post processing, as there is no limit in post. And you are not usually restricted to using a prime focal length.

Sure you can run out of image when you try to squeeze/pull a picture straight in post, but then you just have to shoot with a wider angle to give you more room.

BTW, that image looks like a perfect candidate for HDR technique so maybe you would have been doing a fair bit of post processing anyway ;)
 
Software corrected never seems to look right, there's always some softening or some artifacts or something that degrades the parts that have been altered.
I doesn't matter much on basic minor corrections, but if you have to redraw half the image it always seems to have an impact.
There's also a yawn factor to consider.
 
How about shifting the plane of focus so it is not parallel to the film plane. Can't do that in post processing.

For a large part of my working life I used Monorail cameras. These have sufficient movements To do any thing with in the capability of a lens.
A tilt shift lens is like have front movements but no back movements.... This is rather a poor compromise.

With a decent Pano program and a Focus fusion program, most things are possible digitally.

The reason you shift a plane of focus is to increase depth of field. However though it works in one plane it makes all other planes worse.
 
....snip....
The reason you shift a plane of focus is to increase depth of field. However though it works in one plane it makes all other planes worse.

I keep reading this Terry and I haven't managed to figure out what you mean....can you elaborate please.

Bob
 
I keep reading this Terry and I haven't managed to figure out what you mean....can you elaborate please.

Bob

In Terry's absence, tilt doesn't increase depth of field per se, it just swivels the plane of sharp focus.

If you were doing a food shot with a selection of dishes spread out over the table in front of you, by tilting the lens down you can move the plane of focus closer to the plane of the table, where you want all the dishes to be sharp from front to back. Then with a moderate f/number you can pull them all sharp. Does that make sense? But in doing this, DoF is reduced in all other planes.

DoF is like a plane of glass in front of the lens, parallel to it. As you increase f/number and with it DoF, the glass becomes thicker (much thicker, like yards and yards thick!). As you tilt the lens, you also tilt the glass.

I don't accept what Terry has said about T&S lenses being compromised by not having movements for the sensor/film, as you do with a monorail large format camera - it's just less convneinet. What you have to do is move the entire camera and lens on the tripod head so that the sensor is angled in the way you want it, then adjust the lens to suit. It's the same difference.

In practice however, you can't do very much with that on a small format camera. Not only is there very little room in there to manoevre the lens, depth of focus is very narrow with short focal lengths so you soon run out of scope. Standard lens on a 5x4in camera is 150mm with a lot more depth of focus (but less depth of field); on 10x8in it's 300mm and you can swing the back by several cms quite easily. Scheimpflug Rule and all that.

Edit: did somebody mention the article in this month's Digital SLR User Magazine (Aug edition)? It's quite good, and you don't often read about T&S lenses in action. I have to say is pretty unconvincing in terms of what you get for your money and the best reason given is extending effective depth of field. It gives you enough DoF in landscapes with a low f/number so you can use a faster shutter speed to stop flowers blowing about. Well, that's going to be handy three times a year. Worth reading for anyone thinking about spending a four figure sum.
 
In Terry's absence, tilt doesn't increase depth of field per se, it just swivels the plane of sharp focus.

If you were doing a food shot with a selection of dishes spread out over the table in front of you, by tilting the lens down you can move the plane of focus closer to the plane of the table, where you want all the dishes to be sharp from front to back. Then with a moderate f/number you can pull them all sharp. Does that make sense? But in doing this, DoF is reduced in all other planes.

DoF is like a plane of glass in front of the lens, parallel to it. As you increase f/number and with it DoF, the glass becomes thicker (much thicker, like yards and yards thick!). As you tilt the lens, you also tilt the glass.

I don't accept what Terry has said about T&S lenses being compromised by not having movements for the sensor/film, as you do with a monorail large format camera - it's just less convneinet. What you have to do is move the entire camera and lens on the tripod head so that the sensor is angled in the way you want it, then adjust the lens to suit. It's the same difference.

In practice however, you can't do very much with that on a small format camera. Not only is there very little room in there to manoevre the lens, depth of focus is very narrow with short focal lengths so you soon run out of scope. Standard lens on a 5x4in camera is 150mm with a lot more depth of focus (but less depth of field); on 10x8in it's 300mm and you can swing the back by several cms quite easily. Scheimpflug Rule and all that.

Er, thanks Richard....I must have my "thick" hat on tonight. I thought I understood the Scheimpflug principle but it's the "all the other planes" bit that's not getting through here. I'm going to have to saddle up one of my T&S lenses and work it through.

Thanks for trying though.

Bob
 
Er, thanks Richard....I must have my "thick" hat on tonight. I thought I understood the Scheimpflug principle but it's the "all the other planes" bit that's not getting through here. I'm going to have to saddle up one of my T&S lenses and work it through.

Thanks for trying though.

Bob

Bob, point your T&S at a wall. If you are square on, everything will be sharp from top to bottom and left to right even at lowest f/number.

If you swivel the lens about its axis top to bottom, the wall will be sharp only across a line in the middle, left to right. The plane of focus is now running through the wall, fore and aft of it, at an angle parallel to the lens.

If you rotate the lens left to right, the wall will be sharp in the middle top to bottom, in the same way.

If you now move close to the wall and aim the camera along it looking away from you, the only part of the wall that will be sharp is the bit at the focusing distance. If you now swivel the lens left to right, you can move the plane of focus closer to the plane of the wall and with a moderate f/number pull it all sharp. Scheimpglug will tell you the optimum angles for sensor and lens to give greatest sharpness, maximising both depth of field and depth of focus.

But as you swing the lens to align more closely with the wall, depth of field on the opposite side will be reduced.
 
I suppose that depends whether one's interest is photography or image creation. TV has eliminated the need to be present at sports venues, virtual reality will eliminate the need to visit places of interest and IVF has eliminated the need to copulate but I suspect there will be diehards who'll always continue to do one or more of these things the traditional way.

Bob

Well said Bob. Count me in as one of the diehards. I still prefer to copulate and photograph the traditional way ( and I believe those who use the alternate way are somehow deficient in their capabilities):LOL::LOL:
 
DoF is like a plane of glass in front of the lens, parallel to it. As you increase f/number and with it DoF, the glass becomes thicker (much thicker, like yards and yards thick!). As you tilt the lens, you also tilt the glass.

A little correction if I may to otherwise excellent explanation. The plane of class analogy is only applicable to the case where lens is parallel to the projected image plane (i.e. sensor or film). When the lens is tilted, this "plane of glass" changes shape and becomes non-uniform. It's shaped like a roof of the house ("V" shaped) with "thickness" decreasing towards the direction of the tilt and bottom of the camera and increasing away from it. It is not parallel to the lens in this case. The image below illustrates it:

vcdof3.gif


The image comes from this page from site I referenced before where you can get free books about view camera focusing.
 
cheers for that link. seems to fully understand it all a little research is required!
 
Back
Top