www.dpreview.com
www.dpreview.com

Photographic Mission Impossible

whiteheat | Photo Techniques | Published Jul 18, 2012

Mission Impossible or Not?

I have set myself a photographic challenge.  Nothing extraordinary about that as many people take up photographic projects, assignments, subjective themes, etc.  What’s different about this one is the technical complexity involved, which may mean that it is not particularly achievable or that the results may be poor at best.

Challenge Outline:

To photograph my brother’s house from an observation tower (actually it’s the Telestra Tower in Canberra – located 11.5 miles (18.5Km) away!  My brother’s house is half-way up a hillside facing the tower, which itself sits atop the crest of a hill (Black Mountain).  After all, it seems kind of cool to have a photograph of your house taken from a well known spot an enormous distance away.

So what are the problems involved with this?

I did attempt this exercise last Christmas and found I had certain problems which I will now outline and along with my current thinking on how to overcome or ameliorate these problems for when I try again next time around.  Please feel free to comment, especially with any ideas for improving any part of what I’m trying to do.

The first problem is trying to find and identify my brother’s house.  No easy task from that distance I assure you.  However, it is theoretically possible.  I say theoretically, because in my first attempt at this exercise I failed to find it due to insufficient experience.  Naturally, there is a line of sight between the observation tower and my brother’s house.  What now seems to be the best course of action, is to use a lensatic compass to take a bearing from the house to the tower.  Note the bearing down and then work out the reciprocal bearing.  The reciprocal bearing should in theory give me the line to look along to find the house once I am at the tower.  Again, theory is great, but the reality is that I still have a whole hill side of houses to choose from.

The final location identification device must therefore be a beacon of some description.  For this, I am going to use a 500w directional halogen lamp, of the type you see highway maintenance crews using.  I would have to set it on the house balcony facing the tower and switch it on just before making the trip to the tower.  Take a pair of binoculars along to first locate the house/beacon, 8 or 10 power should be sufficient.  Any more powerful and they become difficult to use hand held.  Now, the combination of compass bearing and high power light and binoculars, should allow me to pinpoint the house, which leads to what apparently became the second problem.

In my first attempt, I originally used a Sigma 100-300mm F4 with a 2x teleconverter, but making out any individual house at that distance was almost impossible, as each house was about the size of a pinhead.  From a very base practical level, I am going to have to guess how strong a telephoto lens I can use or at least be guided by cost.  Now, I intend to use a Vivitar Series 1, 800mm catadioptric lens (that’s a mirror or reflex lens for those unfamiliar with the term) coupled with a Vivitar 3x teleconverter, to give a 2400mm focal length.  I am still extremely doubtful whether this will be enough to capture and fill around 90% of the frame with my brother’s house (large detached, 5 bedroom).  On an APSC camera (currently my Sony A300), this could rise to a 3600mm of focal length taking the 1.5x multiplication in to account.  Still not enough me thinks.

The third problem is related to critical focusing.  These ‘high power’ catadioptric lenses have an extremely small dof.  Adding a teleconverter makes the dof almost non existent, so much so that manual focusing (the only type of focussing possible with this lens type) is incredibly difficult.  For example, just resting a finger tip on the lens is enough to throw the focus out.  Added to this, it is impossible to know at what point you hit peak focus because peak sharpness is reached and passed within a fraction of a millimetre of rotational movement.  Additionally, IQ will be substantially below what you might expect with a normal refractive lens, so much so that the subject never attains an ‘acceptable’ level of sharpness as most people would come to expect.  This is to be expected to some degree because of the lens type and because the teleconverter will degrade IQ.

However, there are two other compounding issues which dramatically affect focus and IQ.  First, in my initial attempt at this exercise, I was using a tripod.  However, it being a light travel tripod, it could not cope with the windshear and gusting that was prevalent at the time.  The result was both macro and micro vibrations that are magnified by any high power telephoto lens.  The result is an unsharp image.

The second issue to affect IQ was mirror slap, which again is not usually noticeable in lower focal length lenses but at 900mm focal length equivalent, it becomes an issue, which again leads to unsharp images.  Needless to say, use a remote shutter release.  I am hoping that the acquisition of the coming Sony FF SLT (Full Frame Single Lens Translucent) along with a much more substantial tripod will alleviate these last issues.  I am hoping that the focus peaking feature will help obtain peak focus point and the tripod and SLT design will eliminate camera movement from wind and mirror slap.  However, this remains to be seen. 

The fourth problem is that of weather.  Weather conditions are crucial to this working out or not.  The problem is that only a relatively narrow set of weather conditions could be classed as satisfactory for shooting, with only one very slim band of weather conditions being classed as near optimum for shooting.  The majority of the times, the weather will be less than satisfactory.  Wind above a certain speed, will be deleterious to IQ even with a more solid tripod.  Rain or high atmospheric moisture content will cut visibility down, usually to the point that the target subject will be greatly obscured if not totally invisible. 

Conversely, as soon as the sun comes out, heat haze will definitely be a problem when trying to capture a subject at such a distance, which would cause a varying degree of IQ loss depending on the severity and intensity of the haze.  A light breeze would be ideal to blow heat haze away but nothing stronger as that would start to move the camera around.  It all comes down to how lucky I am going to be on the day.  Ideally, I would like a bright, overcast day at around 20 to 24 degrees C, very low humidity with just a light breeze at most.  I know this is a bit of an ask, but these are the conditions that I suspect are optimum (not proven though) and perhaps the only conditions under which I stand a chance of getting a “reasonable” shot.  My expectation is that “reasonable” would be the best I could ever hope to achieve and even that would be under the best conditions possible all the way round.

Other observations include the following:

Clean the sensor before shooting.  I did not do this on the first attempt at this exercise, with the resultant dust motes and dark spots appearing on the images.  Much cursing and beating of fists on the ground occurred as a result.  In fact, it would not do any harm to clean all optics involved as well.

You have to use the viewfinder as Live View won’t cut it in this instance, at least not with the screen resolution available to me at the time.  The next FF SLT may be a quantum step better, but any, and I mean any, camera movement will mess up Live View bad due to such high telephoto power massively magnifying any movement so that it is impossible to keep a static picture on the rear view screen.  As an aside, I have since found that using an eyecup on the viewfinder, significantly improves my ability to survey any scene presented.  I would be over the moon if the next FF SLT had an extended viewfinder (by an extra 5cm), allowing enough space between my face and the back of the camera for my nose.  Also, it will be useful to buy an eye patch for whichever eye will not be put to viewfinder use.  This really does save some eye/eye lid muscles from having to work overtime.

Make sure you shoot in RAW as there will be a lot of post processing involved.  Just prior to my first attempt, I lent the camera to the wife and set it up to shoot jpegs only – just for her convenience.  I forgot I had set the output to jpeg only until after I made the first attempt.  At which point, I cursed mightily and vowed to get the wife her own camera, which has now been achieved courtesy of a Sony A35 (great camera by the way – very impressive).

I am still trying to find nice high spots to try out the mirror lens and also with/without the teleconverter attached to see what results I can get.  However, with my current camera and tripod setup, I still have camera movement and mirror slap issues.  This said, I still think I won’t have near enough telephoto power to achieve what I’d like.  It would be nice to find a local high spot from which I can find and focus on a target 18Km or more away.  Around where I am now, it’s as flat as a pancake, so unlikely to happen anytime soon.

Further thoughts on object achievability:

After doing some further research, it has become apparent that in order to get the subject (the house) to fill 90% of the frame on a FF sensor camera, at that distance, I would need a lens that had a focal length of around 40 meters, that is, 40,000 mm!  Clearly, this is not a feasible option.  Therefore, as of this time, I can not see any alternative to using a teleconverter on a catadioptric lens, which currently would give a focal length of 800mm x3 = 2400mm.  This would fill approximately 5% of the frame, which would not adequately fulfil the mission objective.

Continuing on from the above, use a geared head.  It was extremely difficult to position the camera to get a subject dead centre of the frame.  The centre of the frame will be the best spot to place the subject for this exercise.  The dead centre of any lens should be the sharpest portion of the lens, which should translate to a sharper image within this part of the frame.  This should prove hugely significant when enlarging the image later on.  The only way to accurately and precisely place the subject at the required point in the frame, is to use a geared head.  Next time, I will try the junior geared head from Manfrotto.

I can only hope that with a high resolution sensor and extermely accurate focusing with as sharp an optical combination (lens plus teleconverter) that I can obtain, that I will be able to significantly enlarge the resultant image and still obtain a reasonable image quality.  However, my hopes are not high, even though I should have eliminated camera movement and got the sharpest image through the available optics that I could get.

Any useful tips and comments on this mission, especially on how to get a much higher focal length and an improved image quality, that could help me obtain the end result would be greatly appreciated.

Update 25 January 2013:

Over Christmas, I set up my equipment to test the viability of the equipment configuration.  It seemed appropriate that I capture the Telstra Tower from my brother's house first, before then attempting to capture my brother's house from Telstra Tower, a much more daunting task.

Anyway, the Manfrotto 575B heavy duty tripod was set up, Manfrotto 405 Geared Head attached atop the tripod.  The two Vivitar 3x teleconverters were screwed together and then screwed to the T-mount adapter at one end and then the Vivitar Series 1 800mm catadioptric lens at the other.  The whole assembly was attached to the Sony A99, which was then mounted to the geared head.

After much traversing and alteration of elevation, switching between Full Frame and APSC mode and costant twisting of the focus ring, I finally got the Tower in focus as best I could - although this was a mission in itself, taking a full 8 minutes to find the target and then get a focus I deemed was the best I could get.  The results I thought were impressive given the distance involved, the amount of glass to be passed through and the amount of light lost through the 6 EV light loss because of the 2 teleconverters.  The equivalent aperture worked out to be F64 with a 35mm equivalent focal length of 8,400mm. The results are visible here:

http://www.flickr.com/photos/89625133@N02/sets/72157632601610173/detail/

When I went to the Tower the next day and set up, the lighting conditions were good, that is, very bright overcast but with wind gust.  The wind gust was good from the point of view of blowing the heat haze away but not so good for trying to keep the whole camera and tripod ensemble from moving around.  No matter, there were subsidences in the wind which allowed good shots to be taken.

Unfortunately, I was unable to surmount the problem of locating and identifying my brother's house.  Even though I had taken a compass reading on the tower, so that I could just follow the line of the reciprical heading when at the tower and even though I had switched on a 500w halogen lamp on my brother's balcony and pointed it at the Tower, I still could not find my brother's house.  The binoculars I was using were faulty - I was getting a double image displayed and there were totally ineffective at pin pointing the house.  From this point of view the mission was failure.  However, getting as far as I did with this experiment, I am quite confident that it is feasible to capture an image of the house and that the quality will be acceptable for the conditions under which it would be composed.  I will make another attempt again next year, if circumstances permit.