Focusing with tilt

[Grayhand]

I must admit that I become a bit confused by some of the treads here on this site about focusing regarding tilt and the use of a ground glass.
(It might be due to the fact that I am often, and easily confused, according to my son).
I have been using manual focusing on different system since about 2005. First on a 1Dsl and 1Dsll with Contax lenses.
But here the view finder was not optimal for that..

Then I got a RZ Proll + a D-model and focusing become some thing completely different.
Especially after I adjusted the position of the ground glass!
I modified a focus loupe for the camera and it become even more easy to do.

The I began to use a 4x5 camera and, using ground glass was even more easy. + with a loupe!
In one of my confused moments I replaced my RZ with a AFDll camera.
Not much more in focus in the studio after the shift compared to before
And I am still carrying about 15 kg around on my back out in the nature.

In this period I also got a Sinar P 8x10 camera.
And a Hartblei 45mm tilt/shift lens from this forum.
And 90% of my photos is manually focus.

But what I never really has understood is the priniple of focusing tilt after a number out of a calculation or table.
I am to easily confused of my subject in front of the camera to be able to do that.
And I almost never has my camera aligned horizontal. I mostly point it downwards when I am using tilt.

I will use the picture below o describe how I focus with tilt.
And I will use my Hartblei lens as an example.

First I mace a ruff composition in my view finder. And, the camera is pointing downwards a bit.

I focus ruffly in the middle of the picture.
Then I apply about 1-2 degrees of tilt depending on how much the camera leans forward.
Check focus in center.

Then starts the process of adjusting tilt.

I look at the foreground to my view finder and make (if necessary) small adjustment to focus.
Then with the foreground in focus I adjust my grip on the focus ring so I now have my index finger at 12 clock.
And since I use f5.6 for focusing it means that one of the small "tabs" on the aperture ring is just under my finger, also at (almost)12 clock.

Then I shift my view point to a position in the distant focus plane that is easy to see.

Now I rock the focus ring back and forward to get this distant point in optimal focus.
Then I feel where my finger are in regard to the "tab" on the aperture ring:
If it is past 12, then I must tilt a little bit more. "Under Tilt"
But, if it is before 12, then I have to much tilt "Over Tilt"

I adjust tilt accordingly to my result, many by half a degree or so, and then do the focus check again.

After 2 or 3 adjustments both close and distant focus point now lays on the same distant mark on the focus ring.
Any adjustment of the focus will now shift the plane of focus (ruffly) perpendicular to my chosen plane of focus.

The I ponder what f-stop I should use, it will decide how "thick" my focus plane will be.

Then I take the photo, move the focus out and in again, take one more photo, and if the composition feels good,
I repeat the refocus/take photo process one or two more times.

With this technique I will walk away from a scene with 40 - 60% good photos, depending on my f-stop.
And it takes much longer to describe than to actually do it.

On my Sinar P4x5 or 8x10 I have the benefit of using a loupe which increase the accuracy a lot of the process.
Yes, I use the same process described above on my large format cameras.

It is easy. But it requires one act of fait:
You just have to believe that you actually can use the ground glass for focusing :chug:

 

[darr]

I shoot MF because I love the lenses and the slower pace to capture. Tilt has never been that important to me. I find what I want when I am 'one with my gear' and not wrapped up in it. You make some good points. I do love my ground glass and loupe.

 

[paulraphael]

Here's a good tutorial on various various view camera focusing techniques. Some of it may be helpful.

 

[Grayhand]

Well, I feel a greater need to tilt when I use digital compared to film.
With film diffraction is much less of a problem for me.

Regarding the link, it is more or less what I do to focus with tilt. Nothing really new there.
I don't think "Scheimpflug teori" any more. I just visualize my plane of focus and place it where I want it.
But this method, as in the link, requires one thing, that your ground glass is in its proper place.
I am surprised how often I find ground glass that is mounted wrong
So I always check carefully, and if needed shims the ground glass.

What I really is saying is that I don't understand when somebody says that it is not possible to focus their technical camera by the ground glass.

Give me an PhaseOne 260, a Arca Rm3di with an sliding adapter with a properly adjusted ground glass,
and a loupe that has its focus adjusted to the proper plane, and I will ensure you that I can use the ground glass to focus it.

Of course I understand the problem with ground glass if you are to some degree visually impaired.
I am getting older my self, which not directly is improving my eye sight :shocked:

And on the 260 you could actually use the screen on the back to check your results quick and easy.
Trying that with my P45+ is just an exercise in frustration

Ray

 

[torger]

I also find a quite large need for tilting, due to the high resolution. Better optimisations of the available depth of field can often be had with tilt.

However, some have a different shooting style, always focus flat and let some parts of the image be visibly out of focus, they may find it more valuable that the plane of focus is ultra-sharp and super-high resolution than that the whole image is perceived as equally sharp. I have got the impression that since the introduction of pancake cameras (which often lack tilt or have it is an expensive add-on) this way to shoot has become more popular.

Focusing with tilt with a longer focal length (~70+mm) and closer subjects is very easy to do on the ground glass, you don't even need that much magnification. With very wide angles say a Schneider 35mm the edges is very dark (like a 90mm on 4x5" I guess) so it becomes hard to look at the edge, in that case measuring distance to the ground (taking into account camera=sensor tilt) and setting tilt from a table can work better is my experience.

Simply looking at the camera from the side and visualizing a right-angled triangle between sensor, lens and ground, adjusting lens tilt so the tip hits the ground work well when no table is at hand.

Flat-focusing at a flat subject like a wall, or focus at infinity can be quite difficult to do with high precision with a ground glass, much more difficult than the nice tilted scene you show in the attached photo. In those cases I have found that a very high magnification loupe is helpful (I use 20x). Very wide-angles are a bit more difficult still due to visible focus breathing and vignetting, those are most easily focused close to center, I sometimes do a "focus and recompose" in those cases, i e point the camera towards the object to focus at so I can focus close to center of the lens, where it's bright and no disturbing focus-breathing, and then compose afterwards. The deep DoF masks any shifting due to recomposing.

I use these techniques when I focus a landscape scene:

How to focus a landscape scene

 

[jotloob]

We have been at this point of discussion before .

http://www.getdpi.com/forum/medium-format-systems-digital-backs/43598-iq-backs-live-view.html

Have a look to the last page of that thread .

There was very little response . Now , 4 weeks later we get some very intensive discussions about the topics FOCUS and TILT and SWING with technical cameras .
I am surprised .
I appreciate the essays from Jae Moon , Torger and Grayhand as well as the contributions from others .
It seems , that we have some real "TILT specialists" here and I carefully read all contributions and try to understand .
But , I must admit , I do NOT understand all formulas and posts and have a more confused understanding of TILT/SWING than before . I find the described techniques all very very complex .
As mentioned in an other thread , I have attended a TILT/SWING workshop at LINHOF in Munich some years ago . That was when I was shooting LF most of the time .
I got along very well with the Rodenstock TILT/SWING slide rule from Walter E. Schoen and GG focusing .
But since I shoot MFD I feel absolutely lost with TILT/SWING .

I will keep reading all contributions , but up to now , I still feel , its all too complicated .

 

[paulraphael]

One (of many things) to consider when working with a good digital back that has high S/N performance, diffraction isn't the monster we were all taught in Optics 101.

Up until f16 or so, it isn't actually obliterating detail at the frequencies we care about; it's just reducing MTF. Which means it's reversible. You can stop down to f11 with confidence that 100% of the detail that's apparently lost can be recovered, especially if you use a sharpening tool that makes use of deconvolution (like LR, or Photoshop's smart sharpen filter, or any of the advanced sharpening plugins).

Even at f22, diffraction is probably a much smaller deal than defocus blur.

 

[torger]

Yes I'd say many are a bit too afraid of diffraction. However while 100% reversible in theory it cannot be done perfectly in practice. The low frequency component in the airy pattern can cause a bleed from light into dark areas, a non-linear contrast loss that is hard to recover. Deconvolution sharpening algoritms only looks at a small blur spot, as it would introduce too many errors trying to reverse it all.

So at some point there's just too much diffraction to be able to sharpen it well. With current tools I'd say that f/11-f/16 is fine for 6um pixels (60 megapixel backs) probably hard to distinguish on a print after proper sharpening, f/22 there will be some degradation, and f/32 is quite difficult to work with.

In many scenes tilt allows for a solution with a larger aperture and thus less diffraction. In cases with very close foreground (often due to a low tripod) I sometimes combine tilt and small apertures, say 3-4 degree tilt and f/22 with that you can do magic DoF .

 

[Grayhand]

Nice link Torger, I have not seen that article of you before.

Jürgen, I think that, in a practical demonstration, I could quickly show you how simple it is to focus by the ground glass on medium format,
maybe 30 minutes and then you have it clear. And it will be done without any help of external tools or tables. Ones you get it, it become irritatingly simple

I ones shot a brick wall.
First att F11 and then at F22 with my Rz and P45.
And I was really surprised how large effect diffraction had on the quality :shocked:

My base line quality requirement for my photos today is that they should be possible to print up to 1x2 meters,
and that I should be able to stand half a meter from the print without crying.
If I don't want everything front to back in sharp focus, then I must be able to control focus.
And if I also want to control the plane of focus, then I must have even more control.

And regarding the way I work with MY equipment, manual focusing on the ground glass is the ONLY way that I will meet MY requirements.

And, I have NO opinion on how other people work with THEIR own equipment, I always suppose they do what suits THEM best.
I hope that was clear enough

I only become confused when some one says it is not possible to work the way I do

 

[paulraphael]

So at some point there's just too much diffraction to be able to sharpen it well. With current tools I'd say that f/11-f/16 is fine for 6um pixels (60 megapixel backs) probably hard to distinguish on a print after proper sharpening, f/22 there will be some degradation, and f/32 is quite difficult to work with.

Your estimates agree closely with my own experience, and also with the basic theory.

Here are some diffraction MTF charts that also consider sensor MTF (theoretical) and defocus blur.

Anywhere you've got over 10% MTF and not a lot of noise, you should be able to recover a lot of detail.

 

[danlindberg]

Grayhand, your approach is very similar to mine when I used my Linhof SuperTechnica 4x5 or the Fuji GX680III in the past.

I just want to add that concerning diffraction I have found that I can push a symmetrical lens atleast one stop further compared to a retrofocused lens.

 

[torger]

I just want to add that concerning diffraction I have found that I can push a symmetrical lens atleast one stop further compared to a retrofocused lens.

Very interesting... I've also got a sense that it differs between lenses how hard diffraction seems to hit, but I have never really made a proper test of it to see if it's a true difference or if it is only in my head . In (simplified) theory it should be no difference, but perhaps there is some explanation why there seem to be?

Maybe it's about how lenses are optimised rather than if they are retrofocus or not? Many of the Schneiders are optimized for f/11 while I think the Rodenstocks are optimized for f/8 (?).

 

[shortpballer]

Very interesting... I've also got a sense that it differs between lenses how hard diffraction seems to hit, but I have never really made a proper test of it to see if it's a true difference or if it is only in my head . In (simplified) theory it should be no difference, but perhaps there is some explanation why there seem to be?

Maybe it's about how lenses are optimised rather than if they are retrofocus or not? Many of the Schneiders are optimized for f/11 while I think the Rodenstocks are optimized for f/8 (?).

I had thought the rodis were optimized for wide open? On their site it says they recommend you stop down as little as possible. Instead they recommend the use of tilt to keep everything in focus...

I could be wrong, but I am almost always shooting wide open unless I need to stop down to get more in focus.

 

[torger]

I had thought the rodis were optimized for wide open?

Could be, I don't know, only know that the more modern Rodenstock is optimized for something larger than f/11, and that the less modern Schneider "working aperture" is f/11. (My own shooting style is better suited for f/11 working aperture, so I prefer the Schneider.)

 

[paulraphael]

According to Rodenstock they perform best stopped down. Apo Sironar Digital at f11, the newer designs at f8. Wide open performance is impressive though.

 

[jotloob]

Jürgen, I think that, in a practical demonstration, I could quickly show you how simple it is to focus by the ground glass on medium format,
maybe 30 minutes and then you have it clear. And it will be done without any help of external tools or tables. Ones you get it, it become irritatingly simple

Thank you Ray for your kind demonstration offer . I think I will come to Sweden and get you show me your workflow . :thumbup:

 

[Grayhand]

It is interesting to take one step backwards when discussing lenses.

My most boring lens for my Mamiya AFDlll cameras is the AFD 80mm.
I bought two of these cameras, so I got 2 of does lenses.
And both of thous give non-exiting results.
Yes, if I zoooooom deep in they are sharp and good, just like my old 120 macro.

But, I am much more interested to how thing looks when I have printed my standard 1x2 meter
(or what I get depending on how I crop) and nailed it to a wall. (Then diffraction also become less of an issue).
My two favorite lenses for my AFD are at the moment the 120 macro and the 150 AF.

But I am still lamenting the utter lack of my RZ system and its lenses since I sold that system two years ago
Those lenses was special ..

Ray

 

[Grayhand]

Thank you Ray for your kind demonstration offer . I think I will come to Sweden and get you show me your workflow . :thumbup:

No problem, I am quit often in south of Sweden.

And I also gladly take any excuse to go to the south of Denmark

And in combination with some nice nature wiev,
then its a pice of cake, or was it pancake, as in camera :chug:

Ray

 

[jotloob]

Should we have a TECH CAMERA TILT CLUB ? ? ?
I have just joined the tilt world . Will it be an inferno ? ? ?

 

[Grayhand]

Should we have a TECH CAMERA TILT CLUB ? ? ?
I have just joined the tilt world . Will it be an inferno ? ? ?

Splendid idea :thumbup:

Why not name the club "INFERNO"

Then all here at GetDPI Medium Format will feel right at home

Ray