Pages: [1]   Go Down

Author Topic: Dynamic Range  (Read 3258 times)

  • Guest
Dynamic Range
« on: November 30, 2004, 09:17:43 am »

A single blown channel is not necessarily a problem, particularly in an area with no detail.

BUT, if you aren't shooting Raw you are depriving yourself and your clients of much of the image quality that your system is capable of.

Michael
Logged

MatthewCromer

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 505
Dynamic Range
« Reply #1 on: November 30, 2004, 11:52:14 am »

Why in the world aren't you shooting RAW?  You'd have to threaten me with execution to get me to shoot in jpeg mode!
Logged

BJL

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 6600
Dynamic Range
« Reply #2 on: November 30, 2004, 02:03:21 pm »

Quote
Why in the world aren't you shooting RAW?
If you are asking me (which I know you were probably not!) I would answer that I am an amateur, as in doing it for enjoyment, not out of necessity of paying the bills or meeting anyone else's quality standards, so I can afford to choose my trade-offs of convenience against image quality, and I do not enjoy spending a lot of extra time processing each RAW image.

I imagine that even a professional who takes a lot of photos might prefer avoiding a significant amount of unnecesary extra "non- picture taking, non- eating and seeing the family" time, if and when in-camera JPEG conversion can give good enough results. If absolute best available image quality were always needed, few professionals would ever have stooped to using 35mm film.

So to me, the question is if and when the best JPEG is close enough to RAW in quality, and this seems to vary greatly between cameras. The E-1 that I use has an unusually low compression ratio of about 2.8:1 in its best ("SHQ") JPEG option, and perhaps because of this, the few RAW vs SHQ JPEG comparisons I have made with it show no differences. But that might hold only for "normal scenes" where sharpness and color accuracy are the main issues, not high contrast scenes (sorry, "scenes of high subject brightness range"). Hence my curiosity.
Logged

BJL

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 6600
Dynamic Range
« Reply #3 on: November 30, 2004, 10:38:18 am »

I have started doing some experiments with scenes of high subject brightness range, but maybe someone has already done it better. Can one improve dynamic range in JPEGs by shooting with in-camera contrast set to a low (minus) value, in conjuction with careful metering of highlights? My first rough observation is that you still have to use about the same reduction of exposure to avoid blown highlights even with low contrast settings, so the contrast setting mostly helps to bring the mid-tones and shadows up from their initially "underexposed" state.

Hopefully, the effect is compressing the range from mid-tones to highlights, similar to what the shoulder of a film emulsion does. Since final images almost all end up in 8-bit formats (JPEG or 8-bit TIFF), the dynamic range has to be fit into that 8-bit per channel format at some stage, so maybe good in-camera JPEG conversion is enough in many cases.


I realize that using RAW and 16 bit mode editing before final 8-bit conversion gives more options, such as local contrast adjustments like dodging and burning.
Logged

Tony Collins

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 68
Dynamic Range
« Reply #4 on: November 30, 2004, 08:34:42 am »

I have been sorting out some pictures for a stock agency. Their requirement is for the max DR without blown highlights or blocked shadows. Some of my pics have a good luminance spread on the histogram, but may have one channel blocked, typically blue. I am using a 10D with parameters all at zero. My back catalogue is all .jpg.

Does a modest amount of single channel blocking make an image un-usable? (Or blowing one in highlights?). I would give the example of a landscape with a very blue sky. If blue is blown there is no detail to see anyway.

Would turning down contrast and saturation give me more headroom?

Not suggesting I do it but would a low contrast lens reduce DR?

I know I am going to have to bite the bullet and start exploring RAW to squeeze out more DR in future but is the no blocking no blowouts rule a counsel of perfection for most scenes photographed digitally?
Logged

didger

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2030
Dynamic Range
« Reply #5 on: November 30, 2004, 09:37:08 am »

Yeah, raw for sure no matter what.  As for the dynamic range issue, that's a serious issue indeed and shooting raw and blending dual conversions only gets you rather limited improvement.  However, is there anything that prevents you from shooting bracketed images (under and overexposed) to blend later?  This is pretty easy and bomb proof whenever you can shoot static subjects with a tripod.  There is now even a super easy and effective and FREE automatic Photoshop blending action that my brother created for me because I have a huge number of image pairs for blending.  It's the layer mask technique in Michael's blending article, but automated with an action.  
Here it is
Logged

BryanHansel

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 369
    • www.bryanhansel.com
Dynamic Range
« Reply #6 on: November 30, 2004, 11:01:17 am »

You may want to check out this thread over at digital preview.  It is mainly about the D70 and D100, but seems to address exactly what you are attempting to do, if I understand what you have written correctly.

http://forums.dpreview.com/forums....1239112
Logged
Bryan Hansel
[url=http://www.paddling

DiaAzul

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 777
    • http://photo.tanzo.org/
Dynamic Range
« Reply #7 on: November 30, 2004, 01:18:34 pm »

Quote
Not suggesting I do it but would a low contrast lens reduce DR?
When a lens is described as low contrast it is typically refering to local contrast and not global contrast - as seems quite typical in photography the same words crop up to describe multiple phenomena.

Global contrast is what you are looking to control (ie contrast across the whole image) and the only way to capture the entire dynamic range of a scene is to (a) By a MF camera with a high dynamic range sensor or ( bracket multiple exposures.

A low contrast lens on the other hand refers to the attenuating impact of a lens's MTF (modulation transfer function) on the high spatial frequencies in the image. In plainspeak, a lens will blur very detailed parts of the image, the amount to which it causes this fine level blurring depends on the quality of the lens. The more blurring the lower the perceived contrast of the image produced. Sometimes this is quoted as a sharper (high contrast) versus a softer lens (low contrast).

So to answer your question, the answer is no. All you would end up doing is reducing fine detail in your image without reducing the overall dynamic range of the scene.
Logged
David Plummer    http://photo.tanzo.org/

Tony Collins

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 68
Dynamic Range
« Reply #8 on: December 01, 2004, 06:03:44 pm »

Like BJL I have stuck to jpeg to try and keep screen/life time balanced. I am already well behind with keeping my jpegs sorted and organised. It would be nice to have the opportunity to select RAW after the fact whilst reviewing the image as can be done on my G3 for those obvious keepers and "difficult" subjects. As for blending, I find that most of my pictures haven't been plotted but are spontaneous, even accidental. The other weekend I had taken a pretty boring beach/coast/sky scene on an island off Britanny when a "Deuche" rounded the bend ahead and came bouncing towards me. I swung around and snapped, nice pic with the old car, coast road and a typical French pointy roofed cottage. When I got home and back to my screen I discovered that the previous coastal shot overlapped and three minutes later with the help of Panorama Factory's default settings I had a lovely seamless pano encapsulating France and the Breton islands. It wouldn't have happened with a tripod.  

Thanks to Didger tho' for sharing his brother's efforts with us . I do try and bracket and with distant subjects it is possible to hand-hold well enough to superimpose exposures.
Logged

Jonathan Wienke

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 5829
    • http://visual-vacations.com/
Dynamic Range
« Reply #9 on: December 01, 2004, 06:31:08 pm »

Quote
Can one improve dynamic range in JPEGs by shooting with in-camera contrast set to a low (minus) value, in conjuction with careful metering of highlights?
Yes. The extra latitude may show up in the shadows instead of the highlights, but the contrast adjustment controls how much of the sensor's dynamic range ends up in the final converted file. I made a custom tone curve for the 1Ds that gives me about an extra stop in JPEGs, with most of it in the shadows. Typically the default tone curve throws away between 1 and 2 stops of the sensor's dynamic range during in-camera conversion. Adjusting camera contrast to minimuum will push this closer to 1 stop than 2.
Logged
Pages: [1]   Go Up