Upgrade path for establishing architecture photographer

[sharperstill]

I left my job as a press photographer some 5 years ago after close to 20 in that game. My intention was to move cities and establish myself as an architecture photographer. I bought the latest fandangled Canon at that time, which turned out to be the 5D Mk III, having been a Canon user since '98.
It's been a tough slow road to make clients and get myself established and known in my new town, not to mention actually learning architecture photography (I left the media industry with a bunch of portraits, photos from world-level sporting events, studio work, & food, and news of course, but absolutely not one image that could be classified as architecture in my portfolio).
Zoom forward 5 years and I've shot for architects that have recently won awards for their work (that featured my work). I've also had issues where I've discovered the limitations of the 5D's sensor and have sometimes had real issues getting the results that I want.
So I'm now looking at the upgrade path forward and think I have two viable options:
1) wholesale transition from Canon to Nikon (D850)
2) use my existing lenses on the Sony A7 R III
3) use my TS lenses on the Sony A7RIII via an adapter and replace my anon primes/zooms with Sony equivalents.
Both have pros and cons. Some concerns I have, and that I invite responses to, are;
- Canon TS lenses perform better than their Nikon equivalents ergo the hybrid solution might be better for overall IQ
- Lens adapters create a situation less favourable than a single manufacturer 'ecosystem'
- selling a whole kit 2nd hand and replacing it with a kit of new is obviously more expensive
- battery life issues on mirrorless and absence of optical viewfinder
- use of mirrorless with adapted lenses for non-architecture work (ergonomics of camera), yes I still do some corporate and portraiture work from time to time

Something about the 'hybrid kit' makes me feel nervous, although maybe I'm just being old & stubborn (I'm not that old).

Nikon are very keen to 'ease my transition' and have offered some load gear which I should get in a week or two. It's seemingly more difficult to get the Sony for trial purposes, at least at the moment. I have 'placed some calls'.

Jon

[Wolfman]

You can rent the Sony a7r3 and an adapter for your Canon t/s lenses at https://www.lensrentals.com to give it a try. Being able to use the Canon t/s lenses is a big plus over the Nikon.....though the D850 is a very nice camera also. I think the Sony a7r3 is versatile enough to cover the other subjects you mentioned you want to shoot. The battery life on the a7r3 has improved over the a7r2.

[NancyP]

Keith Cooper of Northlight Images, UK, has loads of useful information - he's a professional architectural photographer and Canon user. Reviews, blog, etc. Canon has refreshed its line of TS lenses recently for longer focal lengths, and the TSE 17 and TSE 24 v.2 are highly popular workhorses.

http://www.northlight-images.co.uk/

As for use of Canon lenses with Sony FF A7R III body and adapters, I bet someone here has done this.

[Shiftworker]

For architecture I wouldn't bother with buying an A7rIII - get a 2nd hand A7rII for less than 1/2 the price as the image quality will be just the same. The Metabones adapter will work with your existing Canon lenses very well. The 24mm TS-EII is superb and better than the Nikon equivalent esp when shifted and you can add a 1.4 MK3 EF extender to get a 35mm shift lens with virtually no loss of quality. The 24mm will likley be your 'workhorse' lens. The 17mm is good but not quite as good as the 24mm and I only use mine when I'm in a tight spot. The new 50mm TS-E is allegedly superb but it's very expensive and a good alternative is the old Hassleblad 50mm FL_E (the one with the distance optimisation control) with a shift adapter. The Sony may feel a bit small and feeble in your hands esp with bigh shift lenses so one option is to get a good cage for it esp one with a support foot for the Metabones adapter.

[Jack Hogan]

Hi sharperstill,

By sensor limitation I assume you mean DR and number of pixels, right?  In architecture aliasing and moirè can be an issue, so having smaller pixels with an anti-aliasing filter can be useful.  How about the 5Ds then?  This way you also get to save on new lenses.  The 5Ds' main weakness is DR but that can be easily compensated for by bracketing.

Just a thought,
Jack

[BernardLanguillier]

One data point is that the Nikon 19mm T/S is the best wide T/S currently available, not sure if it is wide enough for you.

Overall, I guess that it is really about the EVF vs OVF aspect.

If you can afford to wait a few months, you may also consider waiting for Canon and Nikon to reveal their mirrorless strategy. This may help figure out the long term potential of both brands relative to Sony?

Cheers,
Bernard

[NancyP]

One thing that generally doesn't get tested for tilt-shift lenses is performance off axis related to the degree off axis. I don't doubt that Nikon has hit it out of the park with this new 19mm PC (tilt-shift) lens, but I was curious to see if there were reviews with systematic evaluation of off-axis performance - didn't find such a review immediately. I daresay that it is difficult to do the usual resolution / distortion / vignetting tests using standardized charts (at least for tilt - shift ought to be easier to test). WWRD? What Would Roger (Cicala) do?? Cicala has the best lens testing platform and explanations arounds (LensRental blog) - geekalicious!

If OP goes for keeping his Canon 24mm TSE II and 17mm TSE, he could use the lens frames with Arca-compatible foot, in order to keep lens in one position and shift or tilt camera around the lens. Why bother? This might be handy if using one of the Sony cameras with adapter.
http://www.northlight-images.co.uk/ppl-tse-adapter-lens-shift-review/
http://www.northlight-images.co.uk/rogeti-tse-frame-review/

[sharperstill]

Hi sharperstill,

By sensor limitation I assume you mean DR and number of pixels, right?  In architecture aliasing and moirè can be an issue, so having smaller pixels with an anti-aliasing filter can be useful.  How about the 5Ds then?  This way you also get to save on new lenses.  The 5Ds' main weakness is DR but that can be easily compensated for by bracketing.

Just a thought,
Jack

Yep, I've just spent 2 days blending a bunch of exposures together on a dark-skinned house. It's a very inefficient way to go about post-processing. Dealing with the artefacts in foliage (even when it hasn't moved much) is a tremendous pain in the ----.
I've only had a few bad instances of moire but again it makes the shoot less profitable by the time taken to deal with the issue and the (sometimes) inferior result.
Not sure I'd consider the 5Ds just yet.
Strongly leaning toward the Sony at the moment, although still wanting to test it out first. Small Rig have an interesting looking L-plate that has an optional Metabones support.

Shiftworker, yes I agree that Canon seems to have the edge on Nikon for the TS lenses and the decision to make a 24mm and a 19mm seems odd. I have a new 1.x teleconverter and use it with my TS 24 for 'across the road' architecture shots. My 17 TS is the 'bathroom lens'.

Bernard, I have a friend fairly high up in Canon Australia and he is completely tightlipped about any future mirrorless from them. I could end up waiting a long time and who's to say Canon will get their sensors performing as well as Nikon and Sony?

[BernardLanguillier]

Bernard, I have a friend fairly high up in Canon Australia and he is completely tightlipped about any future mirrorless from them. I could end up waiting a long time and who's to say Canon will get their sensors performing as well as Nikon and Sony?

Then you just have to wait until the Nikon announcement.

Or just get the Sony and move forward next week.

Cheers,
Bernard

[Joe Towner]

Not sure I'd consider the 5Ds just yet.

Why not?  It's a great camera, has a very high resolving power and it will match your current workflow.  Plus the sensor is in line with your current Mk III and we should see a refresh soon as it's 3 years old now.

[BAB]

Any capable camera and lens will do get business first then you can buy several systems and not feel the purchase!

[sharperstill]

Why not?  It's a great camera, has a very high resolving power and it will match your current workflow.  Plus the sensor is in line with your current Mk III and we should see a refresh soon as it's 3 years old now.

If the sensor is "in line" with my current sensor then what's the point?

[Shiftworker]

Why not?  It's a great camera, has a very high resolving power and it will match your current workflow.  Plus the sensor is in line with your current Mk III and we should see a refresh soon as it's 3 years old now.

It's DR significantly lags behind Nikon and Sony high MP sensors which seems to be the OP's problem with current Canon's sensors ( it was also my reason to change )

[TonyVentourisPhotography]

I was never fond of the 24ts-e... though v2 was way better.  The 17 I liked a lot.  I switched from a canon system using those lenses to a cambio tech cam and an old phase one p45.  To me the easier use of filters and better shift range and performance was worth it.  Plus pricing was decent on the older gear.  The Medium format files still look better to me than 36mm sensors regardless the resolution.  I like the clarity as well as the colors and look of the ccd backs.  But that’s just me.  I am also biased towards the larger ratio files 4:3 instead of 3:2 when shooting architecture.

I’ve wanted to upgrade to a higher res mf back but I’m just not sold on the latest.  And I would also lose out of my current shift range.  I actually have many great shots using 20mm of shift with the sk 35xl lens.  That was the other selling point for me to leave from the tse lenses.  I was never happy with them past 8mm or so. 

To be honest...I’ve been using my e-m1 mk2 with the Leica 8-18 lens for a lot of my architecture clients lately.  The lens is absolutely great and can shoot directly into bright windows with no issues at all.  I wish there were good shift solutions at the wide end for that system but so far nothing surprisingly.

[mcbroomf]

On adapting the Canon TSE lenses to Sony; beware of adapters that are just a little too thin, the Metabones in particular as that is a likely choice.  This does not matter with lenses that have unity focus (all elements moving together) but it does with internal focusing lenses, and wide angle lenses appear to be affected worse.

I found this out when looking for a fast UWA lens for wide field astro.  I bought a Rokinon 14mm SP 2.4 lens after trying several others.  It's not available in Sony FE mount and bought a Canon so that I could get EXIF data.  I tested it with both MC11 and Metabones adapters.  The image was almost perfect with the MC11 (and I now use it wide open for astro with an MC11), but with the Metabones it was absolutely terrible in the corners. 

I ended up shimming both of my Metabones adapters and now get equivalent performance from them.

[alatreille]

On adapting the Canon TSE lenses to Sony; beware of adapters that are just a little too thin, the Metabones in particular as that is a likely choice.  This does not matter with lenses that have unity focus (all elements moving together) but it does with internal focusing lenses, and wide angle lenses appear to be affected worse.

I found this out when looking for a fast UWA lens for wide field astro.  I bought a Rokinon 14mm SP 2.4 lens after trying several others.  It's not available in Sony FE mount and bought a Canon so that I could get EXIF data.  I tested it with both MC11 and Metabones adapters.  The image was almost perfect with the MC11 (and I now use it wide open for astro with an MC11), but with the Metabones it was absolutely terrible in the corners. 

I ended up shimming both of my Metabones adapters and now get equivalent performance from them.

This is the first I've ever heard of this!

Sent from my LG-H873 using Tapatalk

[Shiftworker]

On adapting the Canon TSE lenses to Sony; beware of adapters that are just a little too thin, the Metabones in particular as that is a likely choice.  This does not matter with lenses that have unity focus (all elements moving together) but it does with internal focusing lenses, and wide angle lenses appear to be affected worse.

I found this out when looking for a fast UWA lens for wide field astro.  I bought a Rokinon 14mm SP 2.4 lens after trying several others.  It's not available in Sony FE mount and bought a Canon so that I could get EXIF data.  I tested it with both MC11 and Metabones adapters.  The image was almost perfect with the MC11 (and I now use it wide open for astro with an MC11), but with the Metabones it was absolutely terrible in the corners. 

I ended up shimming both of my Metabones adapters and now get equivalent performance from them.

I've had to shim all 4 of my Metabones adapters for thickness and for face parallelism unfortunately. The manufacturing quality of the rear flange seems to be the culprit as the face that mates with the adapter body is quite crudely finished. Also the Canon shift mechanism isn't that accurate in terms of parallelism and even set to zero their is often still some tilt going on which can throw one part of the image off in sharpness. An engineers surface plate, dial gauge and micrometer are very useful tools for getting the best out of these lenses and adaptors.....

[douglevy]

Have you considered a tech camera with an older digital back setup? If I shot architecture, this would be my plan.