I have yet to try any HDR stuff. Is there any benefit to using a "merge" feature rather than using masking to combine two different layers?
-
Welcome to the new forums! Please read this first. For known issues we are working to resolve, click here.
Beginner's question - how to do HDR?
- Thread starter nikond80hdr
- Start date
No, you just merge the RAW files out of camera to get a 16 bit image which maintains the same WB as the original, then you use WB, NR, and Curves and the resultant image. Think of a merged RAW.
Repeat, you can use ISO AS WELL* . Working on shutter speed alone can force you to slower shutter speeds.
GaryDeM
Veteran Member
the problem would come not from the raw setting. but using different raw conversions for each shot prior to the hdr software. if youn would like to take the 5 shots? in awb(on a sunny day), then convert then to cloudy for all 5, then go to it. they would be the same as far as the hdr software is concerned.
I don't use the dri pro software, it looks pretty limited to me. The example I gave of dropping 3 stops for a cloudy sky is because that is typical of the difference between the sky and the land underneath it
All HDR is extending the range of the sensor, if we had 32EV sensors we would not only need no HDR but we would not even need to set an exposure. That day will come!
But until it does we must use more than exposure if the scene contains more DR than the sensor will measure.
Let's bury the issue of blending Vs HDR photo merge, while I can see that there have been plugins and techniques that do a simple blend of 2 shots which are simpler than more sophisticated algorithms which do tone mapping over more images with greater range, the fact of the matter is that both PSP X2 and PS CS3 now have built in HDR photo merge algorithms, so who would consider buying/using simpler blending techniques?
Let's get down to the real issue, how many shots. I can load as many as I like, but I know it does not necessarily improve the image. When I do setup shots on tripods then I will always shoot loads of shots at different exposures, as everyone does. Generally speaking including narrowly spaced intermediate shots will have negligible effect on the end result.
When doing hand held shots, however, I find it better to use few. Hand held shots require alignment and allthougth the software can handle this, the operation is never (cannot be) perfect, so including more images makes the output softer.
You can cover a large range with a few shots. For example, you mentioned the sun that has just set behind a hill, here is such a shot (the forground was in twighlight so I had to up the ISO on my crappy camera, that is why it is noisy).
To shoot this handheld I first decided the FOV and noted a couple of points on the image that intersected reference points in my viewfinder, so that I could keep more or less the same view. I spot metered a sky shot and a another on the hillside.
I then threw in a few extra shots in between. When I came to merge the shots, although the program did a pretty good job of aligning the images (here BTW, I would point out that PSP X2 seems much better than PS CS3 in this respect), excluding the intermediate shots yielded pretty much the same tone but the image was sharper, particularly the detail of the trees. Using the 2 extremes, which were more than 5EV apart, was too much, but by including just one of the intermediate exposures was sufficient to make the difference. Adding others just worsened, basicly.
So by all means take as many shots as pratical, but including all of them in a merge may be counterproductive, use as few as possible is my motto!
Auto bracketing would not have worked well in such an extreme shot, but when taking ordinary shots in sunlight, I sometimes just stick -2EV exposure comp. and switch the drive to bracketed (which I normally have preset to 5 shots at 1 EV difference). That is more than sufficient to include the shadows under the trees, for example.
alanr0
Veteran Member
On many cameras, increasing ISO will reduce dynamic range, so you may need an additional (slower shutter speed) exposure in any case, to capture the dynamic range you need. You are more likely to see this if you shoot raw, and/or use a camera such as the KM A200 or Pentax K10D with fixed gain between sensor and ADC.
Reduced dynamic range is less likely to be a concern with some of the Canon DSLRs, where the read-out noise (expressed in electrons or photons) increases at low ISO, so the dynamic range is similar at ISO 50 and ISO 400, and only 1/3 stop better at ISO 100: http://www.clarkvision.com/imagedetail/evaluation-1d2/index.html
Compare this with the Nikon D200 which is noisier at high ISO, but maintains fairly consistent read-out noise down to ISO 200, with best dynamic range at ISO 100: http://www.clarkvision.com/imagedetail/evaluation-nikon-d200/index.html
Cheers.
--
Alan Robinson
On current models this does not happen much until you are over ISO 400. Some camera actually get slightly worse at lowwer ISO.
But lets get to the real world and the OP's D80, moving from 100 to 400 he would lose perhaps 0.2EV shadow range in DR and gain 2EV at exposure for the same shutter speed (assuming he had reached the lowwest he is prepared to use.
Noise difference would be negligible, especially as HDR tends to attenuate noise effects.
Gaining +2EV exposure for a -0.2EV loss in DR seems like a good deal to me.
alanr0
Veteran Member
My comments were based on Roger Clark's measurements of D200 raw performance, which show 1.6 stops less dynamic range at ISO 400 compared with ISO 100. I made no assumptions about limits on shutter speed. I believe the sensor technology of the D80 is similar to that used in the D200. 0.4 EV improvement in shadow noise for 2 EV degradation in highlight clipping is barely worth the trouble. http://www.clarkvision.com/imagedetail/evaluation-nikon-d200/index.html
You may be correct for in-camera JPEG processing. In fact the Imaging Resource review shows a reduction in noise at ISO 800, presumably as a result of more agressive noise reduction. http://www.imaging-resource.com/PRODS/D80/D80IMATEST.HTM
Similarly, Phil's tests show little variation in JPEG dynamic range between ISO 100 and 1600 http://www.dpreview.com/reviews/NikonD80/page19.asp
Cheers
--
Alan Robinson
Keyboard shortcuts
- f
- Forum