Hyper Real attempt from a newbie.. Comments?

[KongKurs]

Hi there!

Mostly, I've been aiming to create realistic HDR's, but with the cloudy boring weather in Denmark these days, I just couldn't resist trying for a bit hyper-real..

And the hyper-real editing is just so damn fun!

Please, feel free to comment on this.

Metered exposure:

Edited in Photomatix, and post-processed with Viveza 2 and curves tool in CS5:

What do you think? Overdone?
I'd just grab the chance of asking a question to the experts out there: I'm always shooting at 200 ISO for HDR, but the Photomatix-processing seems to add quite a bit of noise to the tonemapped image - anyone experienced this? I shoot with 1 EV intervals always..

 

[Provo]

Sometimes cloudy days give you the best contrast skies.
I think you did ok the image edit does look better then the flat original one that's for sure
the water does have a blue color cast on there that can use a reduction since the ice is
white not blue. But not bad a good practice is to try to create HDR is any weather environment & available lighting
this way you can familiarize yourself and get some good reference points as to what will work and what won't.

 

[SlickSalmon]

I shoot with 1 EV intervals always..

We've had this "intervals" discussion a couple times, and there is a range of opinions. My belief is that the 1EV practice started because some higher end dSLRs could only AEB in 1EV increments. Later cameras can do up to 2EV intervals, and this spacing is what is recommended by a number of published authors. The issue with the smaller 1EV interval is that you need more images to cover the same dynamic range. More images = more noise.

 

[KongKurs]

I shoot with 1 EV intervals always..

We've had this "intervals" discussion a couple times, and there is a range of opinions. My belief is that the 1EV practice started because some higher end dSLRs could only AEB in 1EV increments. Later cameras can do up to 2EV intervals, and this spacing is what is recommended by a number of published authors. The issue with the smaller 1EV interval is that you need more images to cover the same dynamic range. More images = more noise.

Hmm.. I've read in David Nightingales "Practical Guide to HDR" that 1 EV spacing is optimal, and that 2 EV results in more noise.. It's even illustrated with an exaggerated example...
Now I'm confused...?

 

[Provo]

I shoot with 1 EV intervals always..

We've had this "intervals" discussion a couple times, and there is a range of opinions. My belief is that the 1EV practice started because some higher end dSLRs could only AEB in 1EV increments. Later cameras can do up to 2EV intervals, and this spacing is what is recommended by a number of published authors. The issue with the smaller 1EV interval is that you need more images to cover the same dynamic range. More images = more noise.

Hmm.. I've read in David Nightingales "Practical Guide to HDR" that 1 EV spacing is optimal, and that 2 EV results in more noise.. It's even illustrated with an exaggerated example...
Now I'm confused...?

LOL I get a kick of reading these threads how the heck is 1ev optimal when 1ev isn't capturing all of the dynamic range.
The problem is that high dynamic range imagery has taken off and when something is popular often or not people see it as a cash cow
and they have to take a stake at it, twisting information around confusing people stick with what work's for you I always say.
If 1ev works better for you then continue to do that but 3 exposures at 1ev step will get it's tail kicked with 3 shots @ ev2 simply because more dynamic range is captured from the scene.
don't believe look at the histogram.
ever take a look at Trey Ratcliff's work
take a look around the site
Stuck In Customs HDR Photography

also take a look at his other site where top hdr processors post there images a site I belong to & have had a few of my images featured as well

have a look at the featured images you will see higher EV and multiple shot's
HDR Spotting - Gallery of HDR Photos and Community of HDR Photographers


here's some images I have done with higher ev as example

 

[SlickSalmon]

I really hate to re-start this discussion because it always gets agitated and confused. The goal of HDR is to capture a wide dynamic range. The central question is how finely should you parse that range. Some guys routinely use 5 shots separated by 1EV; some guys routinely use 3 shots separated by 2EV. It's the same dynamic range, but it's parsed differently. The argument offered for the former is that you get greater control during editing. The argument for the latter is that you get less noise. I don't have David Nightengale's book, so I don't know why he would claim that 1EV spacing leads to less noise. I'm going to have to go buy it, because everyone else I have read says the opposite.

 

[KongKurs]

Some guys routinely use 5 shots separated by 1EV; some guys routinely use 3 shots separated by 2EV. It's the same dynamic range, but it's parsed differently. The argument offered for the former is that you get greater control during editing. The argument for the latter is that you get less noise.

I had to test how less exposures with more EV spacing compares to more exposures and less EV spacing..

(images cropped drastically to show noise)
3 exposures, 3EV spacing:

9 exposures, 1EV spacing:

 

[SlickSalmon]

I just looked back at Ferrell McCollough's book, and he cautions against too many EV steps between frames. I think you've demonstrated that 3EV spacing is too much. I have done direct comparisons of 1EV vs 2EV steps and never seen a difference. But my camera has an algorithm that handles noise from over-exposures, and that must have disguised the true difference.

Interestingly, Michael Freeman writes that "the ideal spacing of exposure settings between frames is 2 stops. HDR software can handle this difference, so there is no advantage in shooting a smaller difference, such as 1 stop".

At any rate, thanks again for providing some actual data!

 

[SlickSalmon]

I've got some data of my own to add. It occurred to me that we have a direct comparison in Shootout #2. These two images are both processed identically with Photomatix Pro, details enhancer, strength = 50, noise reduction off. I clipped out an area of deep shadow where noise should be the worst. I looked at them real closely, and I just don't see a meaningful difference.

4 images, 2EV spacing

7 images, 1 EV spacing

Certainly, as Anders has shown, 3EV spacing is too wide. But, my images seem to show that 2EV spacing is good enough, and there isn't much, if anything, to be gained from tighter 1EV spacing.

 

[Bitter Jeweler]

Just a question from a bystander...don't you get more noise from underexposure than overexposure?

 

[KongKurs]

Certainly, as Anders has shown, 3EV spacing is too wide. But, my images seem to show that 2EV spacing is good enough, and there isn't much, if anything, to be gained from tighter 1EV spacing.

Yup, that seems to be the conclusion..
I'll try that next..

Thanks for the input.

 

[KongKurs]

Just a question from a bystander...don't you get more noise from underexposure than overexposure?

I believe so. The Photomatix FAQ states, that the best way to avoid noise in the darker areas of the photo, is to ensure that the lightest image of your exposure sequence has its shadows in the mid-tones on the histogram.

I shoot with 1 EV intervals always..

The issue with the smaller 1EV interval is that you need more images to cover the same dynamic range. More images = more noise.

Sorry Slick, I just have to reopen this debate.. I couldn't stop wondering about this, so I read a little on the Photomatix FAQ. This is what I stumbled upon:

Regarding the exposure spacing, an increment of 1.5 or 2 E.V. is generally recommended. The advantage compared to an exposure spacing of one-EV is that it will limit the number of shots required to span the dynamic range, and therefore reduce the risks of mis-registrations and ghosting. It will also reduce storage needs and make for faster processing in Photomatix. On the other hand, a one-EV spacing -or lower- does have one advantage, which is to better smooth out noise when merging to an HDR image.

They generally recommend 2EV steps other places in the FAQ though. How do you interpret the above? I read it as the more photos, the less noise. But it could also mean, that if you can capture the dynamic range using 3 1EV photos, that's better than taking 3 2EV photos...

 

[SlickSalmon]

Just a question from a bystander...don't you get more noise from underexposure than overexposure?

I believe so. The Photomatix FAQ states, that the best way to avoid noise in the darker areas of the photo, is to ensure that the lightest image of your exposure sequence has its shadows in the mid-tones on the histogram.

I shoot with 1 EV intervals always..

The issue with the smaller 1EV interval is that you need more images to cover the same dynamic range. More images = more noise.

Sorry Slick, I just have to reopen this debate.. I couldn't stop wondering about this, so I read a little on the Photomatix FAQ. This is what I stumbled upon:

Regarding the exposure spacing, an increment of 1.5 or 2 E.V. is generally recommended. The advantage compared to an exposure spacing of one-EV is that it will limit the number of shots required to span the dynamic range, and therefore reduce the risks of mis-registrations and ghosting. It will also reduce storage needs and make for faster processing in Photomatix. On the other hand, a one-EV spacing -or lower- does have one advantage, which is to better smooth out noise when merging to an HDR image.

They generally recommend 2EV steps other places in the FAQ though. How do you interpret the above? I read it as the more photos, the less noise. But it could also mean, that if you can capture the dynamic range using 3 1EV photos, that's better than taking 3 2EV photos...

I read it that I was completely mis-informed. I always thought that fewer shots (eg. 2EV spacing) meant less noise, but I was wrong. Apparently, more 1EV shots covering the same range leads to less noise (but increases the risk of ghosting). Personally, as I've mentioned before, with my camera and the things I tend to shoot, I don't see any difference between 5 shots with 1EV spacing and 3 shots with 2EV spacing. This is important for me, because I do a lot of hand-held HDR. My camera won't shoot a 5 shot burst; it will only do 3 shots. Hence, 3 shots at 2EV have become my standard. If I've got a tripod and more time (and patience) I might go for more, but it's usually only when I need more dynamic range. There are several forum participants who routinely use tight spacing (Provo, myshkin, Bynx and others), and I certainly can't argue with the quality of their work.

As for comparing a lower dynamic range covered by 3 x 1EV shots rather than 3 x 2EV shots, I can't say. You certainly want to take the opportunity to get the shadows 'out of the noise'. I'm no longer sure how much of an over-exposure you need to do that. I've always been confident that simply making sure the tail of the most over-exposed histogram doesn't touch the left edge was good enough, but maybe I need to be more aggressive than that.

 

[mortovismo]

it will only do 3 shots. Hence, 3 shots at 2EV have become my standard. If I've got a tripod and more time (and patience) I might go for more

I'm in exactly the same boat as you...I have zero patience

 

[myshkin]

in my opinion you guys over think this stuff.