pritchiedotcom
lvl.4
Flight distance : 125738 ft
United Kingdom
Offline
|
"any pro who frowns on it is not a true pro but just someone who doesn't properly understand how it's done."
Very true.
There are lots of folks who get paid to take photos (more due to who they know and business skills) and therefore call themselves professionals and not due to their craft or artist skills.
Calling themselves professional does not mean they are good at what they do or their work is good either.
HDR, High Dynamic Range is good. The technique is not new (a lot of people relate it to being something belonging to new software age) and was invented by a very clever french chap (call Le Gray I believe) in the late 1800s (he even looked cool with his goatee).
Unfortunately HDR these days is also linked to grungy (high clarity) images that have been created poorly by newbies who think its different and new in the search for the holy grail of individuality within photographic world full of smart phones.
More senior photographers therefore generally frown upon it and it's all tarnish with same "juvenile", "in poor taste" and “artificial” label every time you mention the three letters HDR. It's now also unfortunately been jumped onto by marketing departments of smart devices.. (Sorry, but photography is a personal art form, if your device does it, it's just mass produced without taste or personal style)
Smart photographers will know it is a way of over coming the limitation of all camera hardware.
The human eye can see about 7 stops of light (static, I've heard quoted from 6.5 to 14), however it can see over 40 stops of light dynamically! A camera can capture around 10 stop of light.
A “stop” is a unit of light intensity, as in half or double the light intensity,
1/250 lets in half the light a shutter speed of 1/125 does.
As a camera can capture around 10 stops of light, detail can be lost in dark or light areas (under and over exposed). Unintentional detail loss is frowned upon by anyone who is any good. (I'll assume you know this and why as you're using HDR) .
Taking a HDR. There are many way of completing this successfully.
You need a starting exposure point for your brackets exposures. This can be the middle (whole frame roughly equal 18% grey), or at a point at either side of the range if some detail loss is intended.
If in the middle, you take bracketed exposures +/- 1 stop. The more exposures the better your final image will be. Using 7 exposures with 1 stop steps will create a smoother (less halos) image than 3 exposures with 3 stop steps. The higher the dynamic range of the scene the more stops will be required to fully capture it (7 x 1/2 gives you 13 stops, 7 x 1 gives you 16 stops. You can measure the range using a spot light meter (like the metering used in a DSLR) Measure the darkest area in the scene and then the lightest area in the scene (not directly at the sun as this may damage your eyes and your camera, but a bright area within the scene). (CAUTION always avoid looking directly at the sun with your camera, think of it as a magnify glass)
Once you know the light and darkest areas you can workout the range of stops required, your middle point etc). To start with you'll probably be too slow and as the sun rises the dynamic range/exposure required will change too. With experience, you'll learn to guess and pre-empt a little adjustment/extra steps before its required.
You can either start with your darkest exposure and bracket up (useful if you're not sure longer shutter speeds are not going to avoid unacceptable movement blurring)
ISO, use the lowest you can as noise is one of HDR weaknesses.
aperture, avoid going over f16 if you can, as your images will suffer from something called diffraction. The ideal max aperture depends on the camera and its sensor (16Mpx FF is roughly f16, 36Mpx FF is roughly f11). Yes, going over a certain aperture depending on sensor size and sensor resolution will make an image softer not sharper. (See Huygens–Fresnel principle)
Take RAW images and import these into your software directly.
Some selective editing can be done before merging, if advantageous, (eg. noise reduction)
however mostly it will cause more problems. Avoid adjusting the exposure as software normally reads the meta data to calculate the stop gaps between exposures (so you're creating expected results). Avoid curve adjustments as you'll just get a compounded result in the final image.
Personally I'd just avoid pre-editing unless you know there is a clear reason to do it and it justifies the extra work involved. If you do edit before merging, save edited files using a high bit range (Tiff 16bit or DNG).
The trick to a good HDR is to make it look as natural a possible, unless your gritty effect adds to your intended image's message, don't just do it cos you feel it looks different and “cool” (cos most of us will have been there, done it, and now look back and thinking “what was I doing? its so puerile”)
A good technique for making a HDR image more natural while maintaining a high dynamic range is to place the middle exposure as a layer over the final HDR image and blend it back in until you have the final scene you wish.
Avoid movement in scenes, (not just the camera), but the branches of trees, cars, boats. Movement creates shadows, some are easier removed than overs (tree movement is very hard to remove). Pick
windless day, early morning.
I also find using luminosity masks with 5 brackets is good, similar effect, more work, better results in my opinion for landscapes. HDR I find better for large luminosity ranges though (like inside cathedrals)
Final note (as I could go on and on), I sometimes use a term of a “composite high luminosity image” as it's not as cliché as HDR.
I hope some of my rambling makes sense and helps. Though I find my students find practice lessons more helpful.
Sorry for any gamma spelling mistakes (no time to re-read it)
Pritchie
PS long exposure. white means over exposed. exposure is a four element triangle (yes four elements)
- Shutter
- Aperture
- ISO
- Intended exposure value (this can be ignored for you)
So you (or your camera) is roughly aiming for a natural balanced exposure over the frame
Say this is 1/60 at f4 at ISO 400 for your scene.
however you wish to slow the shutter manually for long exposure. This means the sensor will be exposed to the light for longer than required. So you need to balance the triangle so all the imaginary values still equal a balance exposure. (the exposure you'd get using auto on your camera)
So 1/60 to 2 seconds means there is time for 7 stops of extra light to pass through causing your sensor to saturate to white.
To compensate you can adjust;
- the aperture to create a smaller hole for the light to pass through so less light passes through during the 2 seconds. Lets say adjust this from f4 to f16, that means 4 stop less of light can pass through to the sensor.
- now lets make the sensor less sensitive by changing the ISO from 400 to 100 (best quality setting). that 2 more stops of light (2 + 4 = 6 stops in total).
We still have one more stop of light hitting the sensor than required for a balanced exposure, and this is where the ND filter comes in.. add a 1 stop ND filter and you have now reduced the amount of light entering by 7 stops so creating a balance exposure for 2 seconds.
So for long exposures, find the correct exposure, adjust the shutter to the required long exposure time, then adjust the aperture and ISO to create the required exposure value. Add ND filters if you reach the limits of your camera settings.
I hope this helps too.. :-)
All the best
Again, sorry for any gamma/spelling mistakes. |
|