So, how would you focus this?

[Patrick Dixon]

The diagram's fine, and the picture looks sharp .... but I can't believe you missed the sunset/sunrise there. I think you're just making it easy on yourself - focusing in nice, bright, light!

Thanks everyone for the discussion. I sort of know this stuff, but it helps to hammer it home. If the light returns I'll head back and try and do better.

[timparkin]

Tim's diagram seems to show the DOF at a fairly large aperture. When you get to around f/22, you actually start to see a much larger angle.

I'm not sure you'll get that much depth of field unless you are using a very large circle of confusion. I've been wanting to play around with a web widget that helps to visualise this stuff for a while so I spent a few hours today creating something that should help. I've added an image for the background that roughly represents this example shot..

http://www.timparkin.co.uk/static/focus/index.html

Have a play. Any comments or suggestions gratefully received..

Tim

[Joanna Carter]

I'm not sure you'll get that much depth of field unless you are using a very large circle of confusion. I've been wanting to play around with a web widget that helps to visualise this stuff for a while so I spent a few hours today creating something that should help. I've added an image for the background that roughly represents this example shot..

http://www.timparkin.co.uk/static/focus/index.html

Have a play. Any comments or suggestions gratefully received..

Hmmm. Somehow, your calculator doesn't reflect my real world experience. No, I can't give you chapter and verse on my micro-tilt technique but, if you care to meet up sometime, I can certainly demonstrate it to you.

Mind you, try a tilt of 0.5°, focal length of 90mm, f/32 and bellows of 90.5mm. I can't see where the back of the Dof is because the hinge point is so low it isn't visible on your calculator.

[timparkin]

I'm not sure you'll get that much depth of field unless you are using a very large circle of confusion. I've been wanting to play around with a web widget that helps to visualise this stuff for a while so I spent a few hours today creating something that should help. I've added an image for the background that roughly represents this example shot..

http://www.timparkin.co.uk/static/focus/index.html

Have a play. Any comments or suggestions gratefully received..

Hmmm. Somehow, your calculator doesn't reflect my real world experience. No, I can't give you chapter and verse on my micro-tilt technique but, if you care to meet up sometime, I can certainly demonstrate it to you.

Mind you, try a tilt of 0.5°, focal length of 90mm, f/32 and bellows of 90.5mm. I can't see where the back of the Dof is because the hinge point is so low it isn't visible on your calculator.

Can you copy the link from the textarea so I can get a reference.

I have checked a few of the settings to confirm that they match Merklinger and all seems to be OK.. It makes sense too as once you get to microtilt territory, you are tending towards just plain old hyperfocal. There can't be some magic discontinuity between 0.0006 degrees and 0 degrees.

e.g. http://static.timparkin.co.uk/static/fo ... 32&coc=0.1

For this settings were lenstilt is 0.05 degrees, the result is almost exactly the same as plain old hyperfocal.

I should also add that there are a couple of rendering bugs so it may be better to use the link above and use the firefox browser if possible.

Tim

[Patrick Dixon]

I've added an image for the background that roughly represents this example shot..

http://www.timparkin.co.uk/static/focus/index.html

I look much slimmer in your calculator Tim - thank you very much!

I couldn't get the calculator to work though - Firefox on XP. The sliders wouldn't move or update the settings.

[timparkin]

I've added an image for the background that roughly represents this example shot..

http://www.timparkin.co.uk/static/focus/index.html

I look much slimmer in your calculator Tim - thank you very much!

It's a pleasure

Tim

[Emmanuel Bigler]

http://www.timparkin.co.uk/static/focus/index.html

Thanks Tim for the nice Scheimpflug simulator.
And thanks to Joanna to define Le Poil de Chien as the definite unit of measurement for small tilt angles.

I like very much the first diagram posted here, showing that even with 1 degree of tilt angle, the DOF wedge is very diffrent from the good old vertical & parallel planes.
So you can imagine, with a 28mm lens on a small sensor, what happens if you want to apply tilts with a good wooden field camera, where the zero click stop for tilts takes at least 2 degrees to get off the zero position
I think I'll shamelessly borrow the diagram in a similar form to add to my article on the subject
(the English version is not up to date, this will be a good opportunity to make the upgrades)
the French version
http://www.galerie-photo.com/profondeur ... pflug.html
the English version (was posted to Tuan Luong's web site a long time ago) and has to be updated, so do not read it
http://www.galerie-photo.com/profondeur ... glish.html

By a strange coincidence on the French galerie-photo forum we have a discussion about : is the Rodenstock circular Scheimpflug slide rule useful or not in the field ? So I have taken the liberty to cross-post to here.
http://www.galerie-photo.info/forumgp/read.php?2,20990
My answer would be : the Rodenstock circular slide rule is probably useless, mostly for very small tilt angles, but at least you can carry it in your LF photo backpack to the remotest parts of, say, the outer Hebrides, without need of any kind of portable computer
However, and very seriously, a set of printed diagrams like Tim's, re-computed for your usual focal lengths and preferred circles of confusion and f-stops ("F/22 and be there") could be a real help in the field. For use in the British Isles, those tables & diagrams on paper should of course be made waterproof

[timparkin]

http://www.timparkin.co.uk/static/focus/index.html
Thanks Tim for the nice Scheimpflug simulator.

I like very much the first diagram posted here, showing that even with 1 degree of tilt angle, the DOF wedge is very diffrent from the good old vertical & parallel planes.
So you can imagine, with a 28mm lens on a small sensor, what happens if you want to apply tilts with a good wooden field camera, where the zero click stop for tilts takes at least 2 degrees to get off the zero position

I've added angle markings on my ebony and can get 0.5 degrees fairly accurately. My standards appear to be within 0.1mm for 150mm lens and inf..

I think I'll shamelessly borrow the diagram in a similar form to add to my article on the subject
(the English version is not up to date, this will be a good opportunity to make the upgrades)
the French version
http://www.galerie-photo.com/profondeur ... pflug.html
the English version (was posted to Tuan Luong's web site a long time ago) and has to be updated, so do not read it
http://www.galerie-photo.com/profondeur ... glish.html

I'll be customising this in various ways over the coming weeks/months so that you get a long scale view (i.e. out to a few km) and also switchable backgrounds for different focussing probelms. I'll also add slider for camera height above ground and also camera tilt.

However, and very seriously, a set of printed diagrams like Tim's, re-computed for your usual focal lengths and preferred circles of confusion and f-stops ("F/22 and be there") could be a real help in the field. For use in the British Isles, those tables & diagrams on paper should of course be made waterproof

If you have any ideas for the sorts of things you would like to see, let me know. I'll license the code as Creative Commons free for non-commercial.

Tim

[timparkin]

If anybody is interested, I've added contour maps to my diagram to show 'lines per millimeter' and also to take into account diffraction.

http://static.timparkin.co.uk/static/fo ... 32&coc=0.1

According to this, the 0.3 degree version has approx 3 lines pairs per millimeter across the flowers at the front and between 5 and 7 line pairs per mm

http://static.timparkin.co.uk/static/fo ... 32&coc=0.1

This shows that you acheive at least 10 lppm at the tips of the flower and up to 40 lppm over the heart of the flower. The background is between 15 and 40 lppm too.. The main sacrifice is that the top of the trees in the foreground where you have possible between 1 and 3 lppm across the leaves at the very top of the frame. Also, as the flowers go into the distance towards the hedge, they are sitting about 6 lppm.

For a good 8x10 print, you need 15-20 lppm -- for ultra sharp 16x20's you want 30lppm and on a 24x30 inch print you will need 40 lppm in order to get pin sharp detail.

Tim

p.s the tool is still in beta mode and may be a little slow as It's building the 'heat map' over a grid of 10,000 points. Fortunatley it caches the results so when you hit the same settings for a second time, it should be instant. If you change the settings, the heat map dissapears, you can click on the link at the bottom right to get the heat map back from the current settings.

[Patrick Dixon]

I just thought I'd mention it's not a hedge .... it's actually a bank - the far bank of a deepish gully running water across the woodland.

I've been back this evening and had a go with minimal front tilt and a higher pov, so I'll post the result(s) once I've sorted developing and scanning out. I have some e6s too, but I'm not sure how they'll turn out as the last batch came out of the CPE2 with quite a blue cast. Not too sure what I did wrong ...

[Patrick Dixon]

Angulon 90mm/f6.8, Plus-X Pan, 1sec @f22 1/2

[Joanna Carter]

For use in the British Isles, those tables & diagrams on paper should of course be made waterproof

Hang on a minute. I've been in "la belle Bretagne" in winds of 130kph and sudden squalls where all I could do was protect the camera with the dark cloth and stand, without anything waterproof for myself whilst it passed…

[timparkin]

Angulon 90mm/f6.8, Plus-X Pan, 1sec @f22 1/2

same again roughly although I'd include a bit of swing to get the left-middle flowers in focus.. the top right isn't as important from my view . I'd like to get that arch of leaves and branch nice and sharp and then allow the far background to drop out of focus a bit...

Something like this?

http://static.timparkin.co.uk/static/fo ... 32&coc=0.1

[Emmanuel Bigler]

Hang on a minute. I've been in "la belle Bretagne"

Ooooops ! Sorry Joanna ! I should have said : for use in the British Isles, including of course the Channel Islands, plus Normandy & Brittany....

----------------

Comments and suggestions to Tim for his simulation tool.
It is nice to be able to define fuzzy Depth Of Field (DOF) limits, this is actually one of the most difficult points when speaking about DOF.

There is an interesting diagram in a recent white paper by Dr. Nasse from Carl Zeiss
Camera Lens News N° 35, April 2010

http://www.zeiss.com/C12567A8003B8B6F/E ... keh_en.pdf

explaining what happens when some digital pre-processing is applied to an image ; see page 23 of the pdf, about through-focus MTF plots, an unusual way to plot MTF curves, but actually the best tool for discussing about DOF limits.
The smooth transition, the gradual blur, that we are familiar with in analogue photography has no reason to be different in digital imaging, but manufacturers of consumer-grade digital camera apply tons of secret sharpening algorithms to make the image more flattering before delivering the final JPEG to the customer. Hence the strange behaviour of the sharp/blurred transition in some digital cameras, this unusual transition being called sometimes "razor-sharp" with very little gradual transition.
Silicon itself nor the sampling process in the Bayer array are not at all the culprits, but secret pre-processing sharpening algorithms, definitely, explain the trick.

There is something that you can easily achieve with your simulation programme, namely a comparison of DOF for various focal lengths and formats for the same scene. This is one of the ever-returning controversial discussions, a plague on Internet Photo-forums
Basically you'll find many people arguing that you have more DOF with smaller formats. In fact those who say that never make side-by-side comparisons with good prints made from a large format film. Including good analogue prints with 0% of digital processing.

My understanding is that when you scale focal formats, focal lengths, circles of confusion, and working apertures properly, DOF limits are the same whatever the format and focal lengths might be, at least when you try to get the ultimate image quality, (like here of course) ; what changes dramatically are tilt angles which become minuscule with smaller focal lengths.
Without any calculations, the idea is that the wedge-shaped DOF limits are directly dependent on the hyperfocal distance, so when the hyperfocal distance is the same for different formats, classical DOF limis and Scheimpflug DOF are the same. Keeping the hyperfocal distance unchanged from one format to another corresponds to a situation where you scale the circle of confusion in direct proportion of the format, this is justified if the final enlargements are of same size ; and if you scale the working aperture following the best aperture of your lens. It happens that for smaller formats, the best f-stop scales like the format as well, top-notch digital view camera lenses have a best f-stop around 5.6 or 8, not 22. Both scaling combine will keep the hyperfocal distance unchanged. Hence the DOF limits unchanged in object space. But tilt angle will decrease in proportion of the focal length....

People using tilt & shift lens on a 24x36 mm silicon sensor, or on a medium format sensor, could be tempted to ignore tilt angles because they |b]have[/b] to be so small. In fact my conjecture, and you'll easily prove it by a proper simulation, is that if you want to achieve exactly the same image quality and DOF limits with a medium format sensor, taking into account a same large final print of utmost quality, you cannot ignore tilt angles with smaller formats, and even the nice 0.5 degree that you can achieve with your Ebony(TM) will not be sufficient.
In other words, manageable tilts and scheimpflug settings are, to my opinion, a nice privilege of the large format camera.
For smaller formats, if you really want to compete with the large format approach for sharpness in wedge-shaped portions of your beloved landscape scene, either you have to tilt precisely by 0.1 degree or so, or you can alternatively forget about Scheimpflug and use a stack of digital images focused on different planes ; and you merge them with a suitable software. Digital imaging and computer pre-processing by brute force methods, whatever you might think, prove every day their efficiency... for the commercial photographer.

I prefer good old elegant methods like the one-and-only setting of the LF camera that delivers the perfect image in one shot without cheating with computers afterwards. (except for dust removal for scanned transparencies...)
But I'm not a commercial photographer !!

[dave_whatever]

For smaller formats, if you really want to compete with the large format approach for sharpness in wedge-shaped portions of your beloved landscape scene, either you have to tilt precisely by 0.1 degree or so,

I think on most smaller format tilt lenses the tilt movement is geared in some way, so they do a have a lot more chance of getting 0.1 degrees than I do on a wooden LF camera. This probably also explains why some people who just shoot say 6x7/9 landscapes use geared monorail setups.