[JMitchell]

So.....not sure if any of your remember a shot I posted from Kew Gardens a few months ago, a HDR made by Artizen or some other program. Well, I went back to the two shots and did it manually, and I was so pleased with the result that I think I'm never going to use a program again. The detail in the tree is very cool, and is the only way the camera can get all those colour tones in one image (accurately)
I know the sky is gone, but that is at least an accurate representation of what it looked like
And...the building is wonky, not my camera, and that's the truth.

Check it:




TFL,

Jack

[Rob Barron]

Can't say this is a good reason to not use a program to do this HDR Jack, you have got the tonal range in the true for sure but remember, a good HDR shot is usually a combination of at least 3 shots and where possible 4, 5 or more. Try doing that lot manually and you'll struggle!

Sorry, despite your protestations, that building LOOKS wonky and this is a photograph, not a record of the scene per se. What I mean by that is the old adage 'If it looks wrong, it is wrong!' and there is no queston that building looks wrong (I assume the people next to it were standing wonkily as well? )

Trees are extremely difficult to get a really accurate HDR shot from simply because it has to be absolutely flat calm or else the different photos will always have some movement of the leaves, especially to the edge of the canopy, so they will look less than sharp. If they are further away that doesn't matter but if they are close up then it becomes noticeable.

I applaud you for giving it a go here but I would definitely look to using something like Photomatix Pro or Photoshopor whatever and using a minimum of three shots, roughly 2 stops a part of 5 or 6 shots at 1 to 1.5 stops apart. The pictures obtainable with practice can be phenomenal and very difficult to get close to with manual processing.

This is not a criticism of your work at all jack, you have done a decent job here using two photos manually. I just mean that method has severe limitations and those were reached here.

Cheers,
Rob