Greenwich Sunset...

[Remorhaz]

Headed down to nearby Manns Point at Greenwich on Sydney Harbour with my daughter to check out sunset. There were far too many clouds for a nice sunset so I headed around the point to see if we could frame the city in the distance being hit by the setting suns rays. Quickly setup with this fallen tree, the rocks (and oysters ) and this pillar out in the water (made from railway line) for foreground interest.

You'll notice immediately that one of these (the second) has been flipped horizontally - I did this with the idea that it might be nice if the tree leads in from the bottom left corner pointing out towards the city in the top right third.

I'd be keen to know if you think this works or not? and any other feedback on the composition generally and/or which of the two shots you might prefer (excluding the horizontal flip) - they are a similar "exposure" but vastly different exposure times due to before and after sunset times?

About 15 minutes before sunset - 16-35/4 at 23mm, 1.6 sec @ f/11 and ISO 100 with the Heliopan CPL and Lee 0.9 grad ND



and this is about 15 minutes after sunset - 16-35/4 at 26mm, 124 sec @ f/11 and ISO 100 with the Heliopan CPL and Lee 0.9 grad ND

[Matt. K]

Rodney
It warms my heart to know that there are photographers out there mulling about in an attempt to catch a great sunset... Reminds me I need to get out more often.
I like the second shot more than the first but I'm not sure why? It has something to do with that leaning peg, which seems to have a lot of power for such a small element in the image. It is darker than most of the other elements so that gives it extra weight/push/pull. But apart from that I'm not sure why #2 seems to be a more complete image. Your image is a good example ofhow important a foreground element is in land/sea/scapes.
OK...looked again. It may have something to do with the way we scan text from left to right. In #1 the peg points to the right and when we scan into it there is little resistance to the eye as they are both travelling in the same direction. In the flipped image we scan into the peg which is now leaning into our direction of travel and creats more resistance. The eye crashes into it. This seems to work in the images favour.
Anyone else want to deconstruct this?

[chrisk]

I think the tree trunk is too bright personally.
I haven't heard the idea about flipping an image like you did. So bloody clever !

[Murray Foote]

In the second one, given that the skyline will be recognisable to Sydney people, why not just flip the sea and tree part of the image? I'd be inclined to try to correct the perspective of the buildings too, though that will make it narrower at the top which may or may not be a problem.

[Remorhaz]

Thanks for the suggestions guys.

Matt - thanks for your insightful breakdown and analysis
Chris - I'll see how it looks with the tree reduced in exposure a little
Murray - could be onto something - although that would move the city to the left side of the image and might make that side a little heavy. I did try correcting the building lean quickly in LR before but I couldn't seem to make it much better (and the image lost quite a bit of the sides - too much in the end). Maybe PS's puppet warp or something might do a better job (if I could figure out how to use it properly ).

[gstark]

Matt,

I like the second shot more than the first but I'm not sure why? It has something to do with that leaning peg, which seems to have a lot of power

It leans to the left and has a lot of power? Princess Jullia is looking for something exactly like this.

OK...looked again. It may have something to do with the way we scan text from left to right.

That would be my guess as well. The direction of the lean allows the eye to follow the trunk down, in the same direction.

[Murray Foote]

Maybe PS's puppet warp or something might do a better job (if I could figure out how to use it properly ).

My first thought would be content aware fill to reinvent what you've lost. Also, since the city is going to be on a different layer than the harbour and the tree, you could convert the city layer to a smart object and scale it up.

[zafra52]

I prefer the first to the second; maybe the colours are
warmer. I also agree that the foreground tree is on the
bright side and detracts rather than adding to overall picture.
At the risk of being too picky, I think you should correct
the distortion of the buildings in the horizon; they lean
to either side giving an almost surreal look to both pictures,
which I believe it is not your intention.

[Remorhaz]

My first thought would be content aware fill to reinvent what you've lost. Also, since the city is going to be on a different layer than the harbour and the tree, you could convert the city layer to a smart object and scale it up.

I prefer the first to the second; maybe the colours are warmer. I also agree that the foreground tree is on the bright side and detracts rather than adding to overall picture. At the risk of being too picky, I think you should correct the distortion of the buildings in the horizon; they lean to either side giving an almost surreal look to both pictures, which I believe it is not your intention.

Thanks Zafra and Murray - I've revisted this (finally) and just used some puppet warping to fix the city lean without loosing too much of the rest of the image, plus some darkening down of the tree, etc

[Murray Foote]

I think that's much improved.

I don't think I've ever used puppet warp except just to test it out. Free transform would probably have done the same thing here, too.

[zafra52]

I also think this version is a big improvement and in large
print would look superb.