So what's best?

View previous topic View next topic Go down

So what's best?

Post by Plant on 2009-01-22, 14:07

Three or more different exposed shots in the camera.

OR

One shot in the camera (raw) and then make different exposures within Bridge (or similar)?

Obviously for moving subjects, it has to be post processing but what about static landscapes/architecture etc.?

I do both but can't convince myself which is best.

Cheers,

Carl.

Plant

Male Number of posts : 4
Location : Tunbridge Wells, UK
Registration date : 2009-01-22

View user profile

Back to top Go down

Re: So what's best?

Post by SpiffyPix on 2009-01-22, 14:15

Great question, Carl. I've done those faux HDRs (three files created from one RAW image), but I've never had quite as much success as I've had with three separate exposures out of the camera.

With the faux HDRs, I seem to always get a lot more noise (esp in the sky) than I get w/ a 3exp HDR. I did have some decent luck with the photo below (1 shot HDR), but I think it was a fluke. Smile

How about everyone else?

Melanie



EDIT: Hmmmm...for some reason, the right side of my pic got cut off here. Bummer.
avatar
SpiffyPix
Admin

Female Number of posts : 170
Age : 44
Location : New York
Job/hobbies : choo-choo driver
Camera Equipment : Canon 5D Mark II, Canon Glass, Gitzo 'pod
Registration date : 2009-01-21

View user profile http://www.spiffypix.com

Back to top Go down

Re: So what's best?

Post by dddphoto on 2009-01-23, 10:13

Without a doubt multiple exposures will always be better than a single exposure. But you don't need to take anyone's word for it. Just take one of your own multiple exposure images and process it then take just the 0 EV exposure and process it. Now look at the results. The caveat with this is that your original image must have a dynamic range outside what the 0 EV image can capture. After all THAT is what HDR is all about.
avatar
dddphoto

Male Number of posts : 28
Age : 60
Location : Boise, Idaho, USA
Job/hobbies : Cytogenetic Technologist
Camera Equipment : Pentax K10D, DA*16-50
Registration date : 2009-01-23

View user profile http://www.dddphotography.com

Back to top Go down

Re: So what's best?

Post by HDRJunkie on 2009-01-23, 19:02

Technically, single-shot RAW processing is not HDR, but MDR (Medium Dynamic Range). Sometimes it is necessary to do single-shot, like when you are shooting a moving subject. You can get some pretty good results from MDR, but if the subject is too high a contrast, it won't be as good as getting bracketed shots.

As Melanie says, you will get more noise with MDR for several reasons. First, you are pushing dynamic range out of the darker areas, and you know that will get you more noise (just look at the dark areas of nighttime shots). Second, noise is a random thing, and when you combine multiple shots a lot of the noise cancels itself out.

If you are doing single-shot, do some post-processing with a good noise removal program.
avatar
HDRJunkie

Male Number of posts : 48
Camera Equipment : D300, SB-900, 17-55 f/2.8, 18-200 VR
Registration date : 2009-01-23

View user profile

Back to top Go down

Re: So what's best?

Post by Monastic Images on 2009-01-23, 19:18

I agree completely with dddphoto concerning the superiority of multiple images over a MDR image. In my tests, usually side by side comparisons, five images shot at high speed, always produced better tones images than trying to "pump up" a single image.

However, I have gone back and retrieved images shot back in 2002 and 2003, long before I ever heard of HDR, and re-processed those images in Photomatix with really outstanding results. I re-set the exposure value at +1, +2, -1,-2 and combine the five shots using virtually no changes in saturation, gamma, or microcontrast. I prefer to handle those changes in post-processing through Lightroom or NX2. The end results, without exception, retrieved both highlight and shadow detail that was missing in the original photos. Although clearly not HDR quality, those shots are now substantially better.

Monastic Images

Number of posts : 3
Registration date : 2009-01-22

View user profile

Back to top Go down

Re: So what's best?

Post by 10megapixel on 2009-01-25, 18:33

It's definitely going to give you best results when you have multiple exposures, but I have gotten excellent results using a single RAW image and creating multiple exposure images from it. Here's an example of a single image HDR I made...

avatar
10megapixel

Male Number of posts : 29
Age : 45
Location : Indiana
Camera Equipment : Canon G10
Registration date : 2009-01-25

View user profile http://www.flickr.com/photos/kewl_pics/

Back to top Go down

Re: So what's best?

Post by hugh on 2009-01-27, 10:11

Hello all,

If you are only going to process a single image, don't bother with making multiple exposures, just Tone Map the original.

hugh
avatar
hugh

Male Number of posts : 60
Age : 75
Location : Orlando, FL, USA
Job/hobbies : Retired
Camera Equipment : Nikon/Hasselblad
Registration date : 2009-01-27

View user profile http://www.lightechos.com

Back to top Go down

Re: So what's best?

Post by HDRJunkie on 2009-01-27, 13:35

hugh wrote:Hello all,

If you are only going to process a single image, don't bother with making multiple exposures, just Tone Map the original.

hugh

Exactly right! Especially if you are shooting RAW (Highly recommended). All the same dynamic range is in the RAW data, whether you manually create the +/- 1 eV versions or not. I'm not sure about the other programs, but Photomatix will easily create a good MDR picture from a single RAW file.
avatar
HDRJunkie

Male Number of posts : 48
Camera Equipment : D300, SB-900, 17-55 f/2.8, 18-200 VR
Registration date : 2009-01-23

View user profile

Back to top Go down

Re: So what's best?

Post by wattsbw2004 on 2009-01-27, 20:41

You always want to use multiple exposures that cover the entire dynamic range because even though you can change the exposure in a single raw, there is no possible way for it to be able to cover the DR of a scene like multiple raw files can.
avatar
wattsbw2004

Male Number of posts : 29
Age : 29
Location : Texas
Camera Equipment : Sony A700, 18-70, 75-300, Sigma 10-20, Wacom Intuos3 tablet, Photoshop CS4
Registration date : 2009-01-23

View user profile

Back to top Go down

Consider this when looking at this question

Post by imagineit on 2009-01-27, 21:12

When making a single image RAW file, the photographer is limited to the upper and lower limits of the sensor's (D/R), at the specific ISO, further impacted buy their interpretation of what is an 'acceptable' level of noise. If you have a 14 bit camera, the total dynamic range (which may start between 6.5 EVs and 8.5 EVs in most cases) is already clipped by bit 1 (bit one- the point of buried noise) and further clipped by the increase of ISO (on many cameras, although it is much better today than in the past). So, even as a RAW, you will not likely have the full use of the sensor's potential D/R. In very hot climates, over 100+ degrees F, thermally induced noise shows up too, further reducing D/R. In very cold climates, a gain over the sensor's 'normal' mid temperature D/R characteristics occurs, too.

All that being said, consider that with RAW, when you are boosting or reducing exposure, the data being used to calculate the tones in the image file is still subject to the limits of the sensor as outlined above. In may cases that is very few electrons being used to calculate tones and tonal changes, particularly in the highlights. Even so, in an ideal situation, most better camera files processed from RAW can still hold detail with about +2EV 'push' and you can 'pull' the shadows about -1EV.

So, by using single shot approach, you MAY expand detail in highlights and shadows somewhat, or not, depending on a lot of variables. But you may just be inducing noise as you do so, too.

The alternative: making multiple images, records different SECTIONS of the SCENE'S total dynamic range and can be used without inducing the noise and other artifacting of the single shot method since there is NO push, or pull, but rather only BLENDING of well exposed, good dynamic range files to reflect the true detail the lighter and darker areas. Example: in bright sun in the desert, you may have a 14EV scene brightness ratio. The camera, at best, may only cover 8.6 EV. By Making a +2. and a -2., you have expanded the actual recorded data of the scene's D/R to cover 12.5 EV and then by CAREFUL blending of the files, create a file that represents the 12.5 EVs without endangering the scene's tonal ramp and keeping a natural look. With the greater electron count, the tones will be more accurate and you still have not tapped 'pushing' the highest highlight or 'pulling the lowest shadow.

Of course, when printed, the greater number of tones are there, but the distance between them (the ramp of change of tonal values) shortens since the paper (or other display media) has fixed limits of yield (reflectance or transmission) to squeeze in all of the tones, the ramp of change is steeper.

Sorry to be so long on my first post.
Nice site, nice people, behaving well.

Good for all.
john
avatar
imagineit

Male Number of posts : 2
Location : Traveling the USA
Camera Equipment : Nikon, Canon, Megavision, and more
Registration date : 2009-01-27

View user profile

Back to top Go down

Re: So what's best?

Post by saftbill on 2009-01-29, 17:26

Echoing John, a single exposure just won't have all the 1's and 0's that you get with multiple bracketed exposures. Therefore, a single exposure will never have the total DR of multiple bracketed exposures, and will not look as good.

However, due to conditions or subject matter, sometimes a single exposure is all that's possible, and fudging the over/under exposures is all you can do. Sometimes, that's as good as it gets, and it's often an improvement over a 'regular' image. If overcooking the tonemapping melts your butter, then that's good too, but neither technique is HDRI, regardless of what it's called (if a cat has kittens in the oven, you can call them biscuits, but are they actually biscuits?).

One day, camera sensor technology will advance to where all the DR information is available on a single exposure, obsoleting current HDRI techniques and opening HDRI up to subjects and situations we can only dream about right now.
avatar
saftbill

Number of posts : 9
Registration date : 2009-01-29

View user profile

Back to top Go down

Re: So what's best?

Post by saftbill on 2009-01-29, 17:28

Carl, I guess that was my way of saying IMO, get the multiple exposures whenever you can, and singles when it's all you can realistically do.

Cheers,
Bill
avatar
saftbill

Number of posts : 9
Registration date : 2009-01-29

View user profile

Back to top Go down

Re: So what's best?

Post by cswift1 on 2009-03-09, 19:14

I've been playing around with HDR for a while, but usually end up with unsatisfactory results. The other day, I came up with one that was miles ahead of the rest, and I've come up with a hypothesis (probably bogus, but let's go for it anyway).

I'm living in a land of ice and snow now and my other efforts have been with scenes that show a huge range (+/-10 EVs?) - what with the blocked out snow and the darks of evergreens. I would take 7-12 images to try and get it all in - the results basically sucked.

The images I took the other day were of a scene that presented a much small range, and shot +/- 1 EV. The results were much, much better.

Here's the hypothesis: If your scene presents an enormous range of exposures, your final product will not be nearly as pleasing as one where the range has tightened in a whole lot.

I'm sure there are those of you who's experience contradicts this, but is it an axiom that obscene ranges (20EVs?) simply don't play well with Photomatix?
avatar
cswift1

Male Number of posts : 6
Location : Oak Creek, Colorado, USA
Camera Equipment : Nikon D70s
Registration date : 2009-03-01

View user profile http://www.hdrcountry.com/

Back to top Go down

Re: So what's best?

Post by DAVE RHUBERG on 2009-03-14, 08:45

A RESPONSE TO CSWIFT1:

You may be right, but in my experience you may need to make multiple HDR from the same photos.
Save your tonemapping before converting. Go back and tweak for the snow, then go and do a tweak for the trees. You could combine these in PS.

Many HDR problems arise from the opposite, not enough EV. If you've got the exposures in with full capture of good pixels for the whole image, I think you could make an adaptation.

Can you make a good conversion, but not a good detail enhance?
avatar
DAVE RHUBERG

Male Number of posts : 23
Age : 46
Location : Lenoir NC, USA
Job/hobbies : Reading, photography, astronomy, travel
Camera Equipment : Nikon D70, D70s, 80-200, 12-24, 50mm, Modbook computer
Registration date : 2009-03-11

View user profile

Back to top Go down

Re: So what's best?

Post by Sponsored content


Sponsored content


Back to top Go down

View previous topic View next topic Back to top


 
Permissions in this forum:
You cannot reply to topics in this forum