Saturday, March 29, 2014

Things are getting complex!

As I have written before, as an engineer it is easy for me to get distracted by the technical side of photography; and this week has been no different to any other. The problem started with the purchase of my new ‘toy’, the Canon TS-E 24mm F/3.5L lens: one of the sharpest primes out there, plus it comes with bells and whistles, ie a tilt and shift capability.
I was na├»ve to think that ‘just’ buying the lens was it, and, soon, related purchases started cascading into our Amazon account, which generated an insightful question from my wife: “when are you going to start taking pictures, rather than buying stuff”.

And this got me thinking about the complexity of modern digital photography.

At one level digital photography has simplified are photography our lives: point the camera at what you wish to take a picture of and push the shutter button. Assuming, as we are at the simplest level, you have the camera on Programme mode (P-mode), your half-shutter press does auto focusing and you are shooting JPEGs, with in-camera style settings, the camera will do everything for you. Simple; and, abusing the use of some technical analysis, the photograph was captured with just one degree of freedom, ie the choice of the scene, because the camera removed all decisions about focus and exposure from you. As I say, photography at its simplest.

But, as we know, as we develop our craft, complexity creeps into our lives. Instead of using P-mode, we gravitate towards manual (M-mode), we reassign focus to the back-button and become sensitized to not trusting the camera’s exposure meter, ie we need to understand stops, 18% grays and metering zones etc. We also shoot in RAW and spend hours ‘making’ an image in Lightroom and/or Photoshop.

All this means we now need to ‘worry’ about many more ‘degrees of freedom’, some of which are independent variables, with others being inter-related. Complexity has entered our lives.

To illustrate the above, take what just happened to me this last week or so. I convinced myself that my photography would benefit from a tilt + shift lens. But such a purchase ended up just being a trigger for some other purchases: 
  • As a manual focus lens I needed to use the LCD, but, in strong daylight this is impossible to see, hence the purchase of a Hoodman loupe;
  • Although the Hoodman loupe is great, the LCD screen is still ‘low-res’, hence I needed a higher definition way of achieving Live View focus, eg using an external monitor. Luckily I already have CamRanger and this saved the day (and my pocket);
  • The tilt + shift lens is a precision instrument that needs to be used as such. I quickly found out that a tilt + shift lens and a ballhead are not good bedfellows. Although my ballhead is a good one, it was too difficult to achieve the required lens alignment, even using the camera’s built in ‘spirit level’ or an external one. This led me to buy a Manfrotto geared head!

  • The head arrived (BTW one of the best purchases I have made) but then another ‘problem’ arose: because of the diameter of my feisol carbon fiber tripod, I needed to buy an extension block to lift the head above the top of the tripod (here’s the final arrangement, including an optional nodal rail).
Thus, from a single purchase, the tilt + shift lens, I ended up ‘needing’ to purchase several more pieces of ‘stuff’. 

And what about the ‘degrees of freedom’ I now have to contend with, well it goes something like this: 
  • 1 from the scene/composition; 
  • 3 from the head, ie yaw, pitch and roll; 
  • 1 from the nodal rail (if I use it);
  • 1 from the tilt; 
  • 1 from the shift; 
  • 1 from the manual focus, to rotate the plane of sharp focus of the tilt+shift lens; 
  • ‘3’ from the manual exposure (assuming I’m not using Magic Lantern Auto-ETTR).
That is, at least, 11 degrees of freedom! With many of the degrees of freedom being inter-related.

Anyway, enough of the technical ‘stuff’, what about the ‘artistic’ side. Well this (Saturday) morning I decided to go out early (nobody around to disturb me) and try out some architectural photography. I took both of my DSLRs: a ‘normal’ 5DMkIII with the TS-E lens attached and my IR-converted 50D, with a 10-20mm lens.

The attached are two of my first cut processed images from the shoot. The triangular one is the IR shot. As I say, these are quick and dirty processed ones: I really need to spend more time on getting the right vision, eg luminosity and presence. 

In conclusion, the deeper you get into digital photography, the more complex life becomes. Or so it appears. In other words, the common factor in the above it me! That is, I choose to create complexity! It’s that technical streak in me, which I have difficulty suppressing.
“Technical skill is mastery of complexity, while creativity is mastery of simplicity”
Erik Christopher Zeeman (British mathematician)

Tuesday, March 25, 2014

Magic Workflow Refinements

Over the last few days I have been refining my Magic Lantern based workflow, especially with respect to my manual lenses, ie the 14mm Rokinon and my 24mm Canon TS-E. Both these lenses require a little more respect than a ‘normal’ auto-focus lens.

In the case of the Rokinon things are simple, as I mainly use this for maximum depth of field; in which case I use an aperture of, say, F11, to minimize the total depth of focus errors from the convolved deblur and diffraction errors, and, at which setting the hyperfocal distance is in the 5-6ft zone. This hyperfocal distance I have precalibrated on the lens focus ring.

In the case of the TS-E, I consider the scene and where I wish to place the plane of sharp focus, and the resultant focus wedge; visualize where I wish to place my hinge point in feet (J) and tilt the lens according to 9/2J degrees recipe; and finally adjust focus to rotate the plane of sharp focus to where I need it.

So far no magic!

The magic comes these days via the Auto-ETTR and, if required, the Dual-ISO features in Magic Lantern. The A-ETTR alone allows me to forget about exposure, ie I am virtually guaranteed a perfect exposure each time, ie the best tonal data to ‘abuse’ in Lightroom and/or Photoshop.

There is also magic when I need to focus, ie I use the Magic Zoom feature in ML to allows me to focus anywhere in the scene. This is especially critical when you need to rotate the plane of sharp focus in the TS-E lens.

All well and good, especially if you are shooting inside and can see the LCD. But what if you are outside and cannot see the LCD because of the bright sun. This was a real issue for me….until this week’s purchase came from Amazon!

The Hoodman Compact Loupe is an optical viewfinder that fits over my 5DMkIII 3.2" LCD Ddsplay. When you wish to inspect your LCD, eg when trying to obtain critical focus, simply hold the loupe over the LCD and you have a crystal clear and sun-shaded view of you LCD.

Although this blog is more about the technical side of photography, rather than the artistic side, if I can I do like putting the odd image on a post. Although as many know the images tend to be of my home, as that is where I do most of my experimenting. Well today is no exception.

The image below was taken with my TS-E lens as I wanted to see what tilting and shifting would accomplish. The scenario was a simple realtors’ shoot, ie to show off the front of a home that was to be marketed. At 24mm I knew I was not going to be wide enough, so I knew I would be shifting. But what about tilting? Well I had set myself an (experimental) objective to bring a water feature into sharp focus with the house. If I shot at F/11, at 24mm, I knew I needed to set up a hyperfocal focus at about 15 ft (I use a circle of ‘confusion’ of 0.02 in my calculations). This would then ensure everything from about 7ft to infinity was in focus.  The trouble was the (dried up!) water feature was about 5-6 ft from me. To be fair, as I use 0.02, I knew that at 0.03 all should have been OK: certainly for a realtor’s website image. 

But this was an experiment, so I decide to see what my TS-E could achieve.

I positioned the hinge (in space) at about 3 ft below the lens axis, by dialing in about 1.5 degrees of downward tilt. I then rotated the plane of sharp focus towards the horizontal by focusing at the furthest point on the house, ie the roof. This rotated the focus wedge and allowed me to bring everything into focus.

Finally, I composed the image to ensure that when I shifted 12mm, left and right, I would capture the extremities of the house. Thus my TS-E setting were 1.5 degrees of downward tilt and a horizontal shift arrangement of +/- 12mm.

The Magic workflow went as follows:
  • Set the TS-E as above; 
  • Compose the scene, ie check the shift extremes;
  • Level the camera;
  • Set an aperture of F/11 to minimize deblur+diffraction and obtain the sharpest maximum depth of field;
  • Invoke ML A-ETTR and ML sets the optimum shutter speed;
  • Take the unshifted image, then the +12mm and -12mm images; 
  • Ingest into LR; 
  • Set the basic setting in one of the images and sync to the others; 
  • Export from LR to Photoshop pano; 
  • Create the merged image and use the PS Adaptive Wide Angle filter to ‘correct’ as you see fit; 
  • Reimport into LR to ‘finish off’.
OK the image is not visually exciting and needs some drama, ie I need to shoot it at the Blue Hour and with house lights on. But today’s experiment was about magic workflow and the hoodman.

Bottom line: every time I use Magic Lantern I realize this is THE tool every Canon DSLR photographer should be using; and now I have the collapsible hoodman loupe I can always guarantee seeing the LCD and thus be able to use ML's full potential. The magic just got better!