Architecture. Shift lens compared to post processing test.

[Reilly Diefenbach]

You have over corrected and the shot has lost how it looked in reality

Mmm no. The tower is absolutely no way racked to the right as you show it, otherwise it would have collapsed long ago.

I'm just glad you're not building my house :^)

Ever seen the leaning tower of Pisa?

I have been up to the very top of the Tower of Pisa.

Where was your church shot taken? What town?

Fivizzano in the Tosco Emilian Apennines

Thank you. The handy thing about Google is that you can look up just about anywhere.

It looks as if in your "corrected with software" version below you grabbed the middle of the frame and moved the building over but forgot to bring the tower along for the ride. Leaning Tower of Pisa? Mm no.




Corrected




San Paolo from Google




San Paolo 2

Absolutely no sag or bend in the tower. It's as straight and plumb as it was in 1148 A.D. even allowing for the typical lazy photographer's pyramidal rendering.

 

[Nigvo]

You have over corrected and the shot has lost how it looked in reality

Mmm no. The tower is absolutely no way racked to the right as you show it, otherwise it would have collapsed long ago.

I'm just glad you're not building my house :^)

Ever seen the leaning tower of Pisa?

I have been up to the very top of the Tower of Pisa.

Where was your church shot taken? What town?

Fivizzano in the Tosco Emilian Apennines

Thank you. The handy thing about Google is that you can look up just about anywhere.

It looks as if in your "corrected with software" version below you grabbed the middle of the frame and moved the building over but forgot to bring the tower along for the ride. Leaning Tower of Pisa? Mm no.


Corrected


San Paolo from Google


San Paolo 2

Absolutely no sag or bend in the tower. It's as straight and plumb as it was in 1148 A.D. even allowing for the typical lazy photographer's pyramidal rendering.

You lift two awful keystone ridden shots from internet, taken with cheap barrel distorted lenses, an pass this off as proof.

Well, I have seen this church for real, and took shots from several angles with a well corrected professional lens on a level tripod. I know what this church looks like in real life.

Here is another test shot, with just vertical shift, but with very weak two point perspective, that does not please, as the horizontals do not look good. But shows the building as it really is. I have screen grabbed the picture with a grid, so as you can see the doors and windows are vertical, as is the drainpipe in the far right.

The tower leans. Period.



Here are the perspective lines and a key to explain two point perspective. 2 point perspective is a basic drawing skill I learnt in my early days in the design office.

The lens has described the building perfectly with little distortion



--
“Everyone sees what you appear to be, few experience what you really are.”
- Niccolo` Machiavelli.
https://nigelvoak.blogspot.com/

 

[Nigvo]

The different shooting points changes nothing.

What can I say ?

We can not conclude anything from any of your tests. None of them are valid.

The tests in the OP were from the same point and the same lens.

 

[Reilly Diefenbach]

Fivizzano in the Tosco Emilian Apennines

Thank you. The handy thing about Google is that you can look up just about anywhere.

It looks as if in your "corrected with software" version below you grabbed the middle of the frame and moved the building over but forgot to bring the tower along for the ride. Leaning Tower of Pisa? Mm no.


Corrected


San Paolo from Google


San Paolo 2

Absolutely no sag or bend in the tower. It's as straight and plumb as it was in 1148 A.D. even allowing for the typical lazy photographer's pyramidal rendering.

You lift two awful keystone ridden shots from internet, taken with cheap barrel distorted lenses,

You have no idea what lenses were used.

an pass this off as proof.

Yes, indeed, the converging verticals tally perfectly when corrected with one Transform Skew:


San Paolo Corrected

Yours is the only one of dozens that shows the tower leaning umpteen degrees to the right.

You would also have us believe that the left wall is way out of plumb as compared to the right:


?

Boy, those old builders sure were incompetent. They probably need to rope off the area before someone gets crushed :^)

Well, I have seen this church for real, and took shots from several angles with a well corrected professional lens on a level tripod. I know what this church looks like in real life.

You are just having a lot of difficulty getting your picture to reflect the reality of the scene.

Here is another test shot, with just vertical shift, but with very weak two point perspective, that does not please, as the horizontals do not look good. But shows the building as it really is. I have screen grabbed the picture with a grid, so as you can see the doors and windows are vertical, as is the drainpipe in the far right.

The tower leans. Period.

No, it doesn't, as any one of a dozen or so images easily available on Google will prove beyond any doubt:

https://www.google.com/search?q=san...ECAEQIQ&biw=1266&bih=626#imgrc=h2g4EAib6iFcwM

You would have us believe that rigid bricks and mortar at the top of the tower could somehow bend that far without collapsing. Good luck with that.

Here are the perspective lines and a key to explain two point perspective. 2 point perspective is a basic drawing skill I learnt in my early days in the design office.

The lens has described the building perfectly with little distortion

The tower is way off, and both walls are racked to the left.

It's the internet. Anyone can get up on their little soapbox and proclaim victory. Feel free to present your subjects any way you wish, as long as I don't have to look at them.

 

[Nigvo]

So, I took your picture taken by some random photographer and adjusted by you to how you think the shot should look.

Unfortunately straitening the tower has skewed other parts of the picture. I have marked up the door and a building to the far right and brought the door frame into the vertical.

Oh dear! The tower does in fact lean, just as my photograph taken with a perfectly level camera and a high quality corrected lens shows. The wall on the lest leans outwards badly and not inwards, as we see in my photograph and how I witnessed it in person.

This corrected picture is pure fantasy.

I have included a grid to make things clearer.

It would be better to study some optical theory and how lenses project an image, before coming back again.

BTW I am a structural engineer and can assure you that with a lean of 2°, that tower is quite safe as the centre of gravity walls remains within the thickness of the wall.


Your shot















--
“Everyone sees what you appear to be, few experience what you really are.”
- Niccolo` Machiavelli.

NIGEL'S RAMBLES

nigelvoak.blogspot.com

 

[Reilly Diefenbach]

Your perception of building geometry is as highly problematic as your silly "tests." Harmless enough, I suppose. Make as big a mess as you please, no one will care. Just don't come anywhere near my building site. Over and out.

For anyone who cares about correct architectural geometry and wants to learn how to do it really well without spending 3-4 grand on a tilt shift lens which wouldn't get it all in anyway, I have a tutorial on my Flickr page.


Tartu U

 

[John Sheehy]

Your perception of building geometry is as highly problematic as your silly "tests." Harmless enough, I suppose. Make as big a mess as you please, no one will care. Just don't come anywhere near my building site. Over and out.

For anyone who cares about correct architectural geometry and wants to learn how to do it really well without spending 3-4 grand on a tilt shift lens which wouldn't get it all in anyway, I have a tutorial on my Flickr page.

Your sample is 32MP, and the amount of correction is unstated. So, while I don't see much resampling/aliasing artifacts in your image, they could easily be visible with a low-res sensor that outputs very aliased capture, along with a huge perspective correction. There was a time when people were using sensors with only a few MP, and with a very sharp lens, one would be a lot better off with optical shift, as sharp, low-res images do not resample gracefully.

Although noise is mostly subliminal in a shot like yours, someone might do perspective-challenged photography with some action in low light, and in those cases, optical shift will keep the noise grain from getting wider and deeper at the top of a building compared to the bottom.

 

[Nigvo]

You keep very quiet about my correction to your "corrected" image, so I take it your "correction" was badly done.

You use the usual tactic of changing the subject with a personal attack thrown in for good measure.

Just admit I was right all along, and your "corrected" image was wrong.

Now tell me, why do serious architectural photographers use shift lenses? It is essential gear for these photographers.

 

[Reilly Diefenbach]

You keep very quiet about my correction to your "corrected" image, so I take it your "correction" was badly done.

My correction, although imperfect as it must be, having been taken as a single frame is far better than what you've made with your tilt shift.

You use the usual tactic of changing the subject with a personal attack thrown in for good measure.

You started this thread in order to pontificate on the virtues of tilt shift lenses and heap scorn on "cheap lenses." I replied. I can get a tack sharp shot of a building with a $120 nifty fifty any day of the week. Way sharper in the corners than a 11-12mm shifted lens.

Just admit I was right all along, and your "corrected" image was wrong.

Hint to Nigvo: every shot from close distance will be more or less wrong, especially if I didn't take it. I prefer less wrong and will take the time to make the picture as accurate as possible. Does anyone care about good geometry any more? Maybe not, but I always will.

Now tell me, why do serious architectural photographers use shift lenses? It is essential gear for these photographers.

I have zero interest in what anyone else is using, only what will get the job done to my standards. A tilt shift can be useful in very limited circumstances, but shooting up from 50 feet away simply swaps one bad distortion for another, as we see over and over again:




Distortion



Almost every shot of a large building will entail close quarters and shooting up. Not what you show in your "tests" out in the country with plenty of room to level the camera.

Both of the above pics are very wrong. Those photogs who have spent large dollars on these T/S lenses are seldom inclined to finish the job with appropriate corrections in software, you being a prominent example. They take the shot and put it up half finished. Fine with me, my stuff will sell better. Their loss and my gain.