If this is your first visit, be sure to check out the FAQ by clicking the link above. You may have to register before you can post: click the register link above to proceed. To start viewing messages, select the forum that you want to visit from the selection below. |
|
|
Thread Tools | Display Modes |
#1
|
|||
|
|||
Histogram and highlight clipping in DSLR's
Hi,
I did an experiment yesterday. I took a photo, and my set my camera to show clipped highlights. About half of the image was blinking as being clipped. I then opened the resulting file in software, and asked it. Almost none of it was clipped. Now the software will only mark it clipped if it's 254 or above (my setting). I then changed the setting to about 245, and then it really matched what my camera LCD was showing me. This was a rough estimate - I didn't try other numbers to really pin it down. This bothers me. My purpose of looking at clipped highlights is just to know if detail has been lost - and my camera is a bit conservative about this. I'm losing some flexibility in increasing the exposure because I don't _really_ know if too much is being exposed. It was somewhat of a shock to realize that only a very tiny portion of the image was "really" clipped, while the camera was showing it as half of the image (which would normally force me to reduce the exposure somewhat unnecessarily at times). I don't really have that good a feel for the numbers. Is 254 brightness that much more than 245? I'm otherwise happy with my camera (Olympus E-500). I'm just curious if people have tried figuring this out in other DSLR's (at least the ones that show clipped highlights). Is this a common issue? -- Don't drink and park, accidents cause people. /\ /\ /\ / / \/ \ u e e n / \/ a w a z anl |
#2
|
|||
|
|||
Histogram and highlight clipping in DSLR's
Mueen Nawaz wrote:
Hi, I did an experiment yesterday. I took a photo, and my set my camera to show clipped highlights. About half of the image was blinking as being clipped. I then opened the resulting file in software, and asked it. Almost none of it was clipped. Now the software will only mark it clipped if it's 254 or above (my setting). I then changed the setting to about 245, and then it really matched what my camera LCD was showing me. This was a rough estimate - I didn't try other numbers to really pin it down. This bothers me. My purpose of looking at clipped highlights is just to know if detail has been lost - and my camera is a bit conservative about this. The difference between 254 and 245 is not all that conservative. I'm losing some flexibility in increasing the exposure because I don't _really_ know if too much is being exposed. So meter it. :-) It was somewhat of a shock to realize that only a very tiny portion of the image was "really" clipped, while the camera was showing it as half of the image (which would normally force me to reduce the exposure somewhat unnecessarily at times). I don't really have that good a feel for the numbers. Is 254 brightness that much more than 245? That depends on why you need it. You might try opening up a worrisome image in something like Photoshop and asking it to show you only the parts with a value over 245 to see how much detail is available. In PS, you'd do Image - Adjustments - Levels, make sure the Channel is set to RGB, and then drag the black Output slider to 245. I'm otherwise happy with my camera (Olympus E-500). I'm just curious if people have tried figuring this out in other DSLR's (at least the ones that show clipped highlights). Is this a common issue? |
#3
|
|||
|
|||
Histogram and highlight clipping in DSLR's
"Mueen Nawaz" wrote in message ... Hi, I did an experiment yesterday. I took a photo, and my set my camera to show clipped highlights. About half of the image was blinking as being clipped. I then opened the resulting file in software, and asked it. Almost none of it was clipped. Now the software will only mark it clipped if it's 254 or above (my setting). I then changed the setting to about 245, and then it really matched what my camera LCD was showing me. This was a rough estimate - I didn't try other numbers to really pin it down. This bothers me. My purpose of looking at clipped highlights is just to know if detail has been lost - and my camera is a bit conservative about this. I'm losing some flexibility in increasing the exposure because I don't _really_ know if too much is being exposed. It was somewhat of a shock to realize that only a very tiny portion of the image was "really" clipped, while the camera was showing it as half of the image (which would normally force me to reduce the exposure somewhat unnecessarily at times). I don't really have that good a feel for the numbers. Is 254 brightness that much more than 245? I'm otherwise happy with my camera (Olympus E-500). I'm just curious if people have tried figuring this out in other DSLR's (at least the ones that show clipped highlights). Is this a common issue? -- Don't drink and park, accidents cause people. /\ /\ /\ / / \/ \ u e e n / \/ a w a z anl What an interesting subject you have raised! I'm gonna replicate what you did later today. Your camera is playing safe and there is nothing wrong with that. DSLRs do, on the whole, I find, err on the side of caution when assessing exposure compared to a point and shoot camera. This is because a) DSLR users (traditionally) have been more inclined to post process their pictures and b) because the larger sensors in DSLRs have more dynamic range so shadow detail can be recovered very successfully. Ian Digital Photography Now The online magazine and community for anyone interested in taking digital pictures! http://dpnow.com DPNow forum and users photo gallery: http://forum.dpnow.com Win an Olympys E-330 Live View DSLR during December: http://dpnow.com/3241.html |
#4
|
|||
|
|||
Histogram and highlight clipping in DSLR's
On Sun, 17 Dec 2006 14:57:51 -0600, Mueen Nawaz wrote:
I did an experiment yesterday. I took a photo, and my set my camera to show clipped highlights. About half of the image was blinking as being clipped. I then opened the resulting file in software, and asked it. Almost none of it was clipped. Now the software will only mark it clipped if it's 254 or above (my setting). I then changed the setting to about 245, and then it really matched what my camera LCD was showing me. This was a rough estimate - I didn't try other numbers to really pin it down. This bothers me. My purpose of looking at clipped highlights is just to know if detail has been lost - and my camera is a bit conservative about this. I'm losing some flexibility in increasing the exposure because I don't _really_ know if too much is being exposed. It was somewhat of a shock to realize that only a very tiny portion of the image was "really" clipped, while the camera was showing it as half of the image (which would normally force me to reduce the exposure somewhat unnecessarily at times). I don't really have that good a feel for the numbers. Is 254 brightness that much more than 245? I'm otherwise happy with my camera (Olympus E-500). I'm just curious if people have tried figuring this out in other DSLR's (at least the ones that show clipped highlights). Is this a common issue? I can't answer your question, partly because my camera doesn't show clipped highlights, but I have a question for you. The display that shows clipped highlights has *much* lower resolution than the sensor, so how many display points are used to indicate clipping? Does each one have to light up even if only a single pixel has reached 254 (or 245)? Or only if, say, 5% of the pixels (under its jurisdiction, as it were) have clipped? If Olympus lights up any of these points used to show where clipping has occurred, even if only a single pixel has saturated, then it's possible for the clipping histogram to indicate half of the image has clipped, when it might actually be only 1% of the total number of pixels. But if that 1% is randomly distributed across the sensor, due to the way that the saturated pixels are monitored, the camera could indicate that half of the regions it's monitoring have one or more saturated pixels. That might not be the way the average photographer would want the camera to warn about clipping, but it might be what some finicky pros would want. |
#5
|
|||
|
|||
Histogram and highlight clipping in DSLR's
ASAAR wrote:
I can't answer your question, partly because my camera doesn't show clipped highlights, but I have a question for you. The display that shows clipped highlights has *much* lower resolution than the sensor, so how many display points are used to indicate clipping? Does each one have to light up even if only a single pixel has reached 254 (or 245)? Or only if, say, 5% of the pixels (under its jurisdiction, as it were) have clipped? If Olympus lights up any of these points used to show where clipping has occurred, even if only a single pixel has saturated, then it's possible for the clipping histogram to indicate half of the image has clipped, when it might actually be only 1% of the total number of pixels. But if that 1% is randomly distributed across the sensor, due to the way that the saturated pixels are monitored, the camera could indicate that half of the regions it's monitoring have one or more saturated pixels. That might not be the way the average photographer would want the camera to warn about clipping, but it might be what some finicky pros would want. Interesting way to look at it. In this particular case, the truly clipped area was just a tiny area - not spread around, and well within the resolution of the camera LCD. But I perhaps should check with a number like 250 and see if the "clipped" areas are scattered across the image - in which case your explanation may make sense, and the 245 number may be off. (Anyway, even if what you say is true, I think the lesson is the same: Not to overrely on the histogram/clipped highlight indicators). -- "Class, please! If you don't learn Roman numerals, you'll never know the dates certain motion pictures were copyrighted." -- Mrs. Krabappel in The Simpsons. /\ /\ /\ / / \/ \ u e e n / \/ a w a z anl |
#6
|
|||
|
|||
Histogram and highlight clipping in DSLR's
I tend to be a result person. What the camera or software believed was clipped is only a tool to get to and end product. It is all about what the end product looks like. I did not see any comments from you about what the end product looked like. If you don't see the difference in the end product, then I suggest that you are getting far too hung up on the technical aspects of photography and missing the real real core of the subject the art of photography. You are letting the tools dictate the results not your heart. On Dec 17, 3:57 pm, Mueen Nawaz wrote: Hi, I did an experiment yesterday. I took a photo, and my set my camera to show clipped highlights. About half of the image was blinking as being clipped. I then opened the resulting file in software, and asked it. Almost none of it was clipped. Now the software will only mark it clipped if it's 254 or above (my setting). I then changed the setting to about 245, and then it really matched what my camera LCD was showing me. This was a rough estimate - I didn't try other numbers to really pin it down. This bothers me. My purpose of looking at clipped highlights is just to know if detail has been lost - and my camera is a bit conservative about this. I'm losing some flexibility in increasing the exposure because I don't _really_ know if too much is being exposed. It was somewhat of a shock to realize that only a very tiny portion of the image was "really" clipped, while the camera was showing it as half of the image (which would normally force me to reduce the exposure somewhat unnecessarily at times). I don't really have that good a feel for the numbers. Is 254 brightness that much more than 245? I'm otherwise happy with my camera (Olympus E-500). I'm just curious if people have tried figuring this out in other DSLR's (at least the ones that show clipped highlights). Is this a common issue? -- Don't drink and park, accidents cause people. /\ /\ /\ / / \/ \ u e e n / \/ a w a z anl |
#7
|
|||
|
|||
Histogram and highlight clipping in DSLR's
|
#8
|
|||
|
|||
Histogram and highlight clipping in DSLR's
Ed Ruf Ed Ruf (REPLY to E-MAIL IN SIG!) wrote:
Does your camera have individual channel histogram/cluipping warning, or is it just combined luminance. In software I'll bet you looking at individual channel levels, not the combined luminance value. The camera only warns on the overall luminescence - although it does show individual channel's histograms. The software was showing the warning for overall luminescence - not for an individual channel. As I said in another post, there was quite a bit of detail that I managed to extract from the area the camera thought was clipped. -- "I'm a lawyer." "Honest?" "No, the usual kind." /\ /\ /\ / / \/ \ u e e n / \/ a w a z anl |
#9
|
|||
|
|||
Figured it out
I should have guessed it earlier. The camera shows clipped highlights based on the JPEG it creates (I shoot in RAW + JPG). I was looking at the RAW file on the computer. When I loaded up the JPEG and compared the highlights, it matched up really well. Now if I'm ever in the mood, I'll set the contrast setting on the camera to a minimum and then compare what it says and what the software says. When I think about it, I almost always shoot raw, and I'll check with the RAW software folks to see if they make use of my contrast, saturation, etc settings in the camera. If not, I might as well set it low on the camera anyway. -- Youthful Figu What you get when you ask a woman her age. /\ /\ /\ / / \/ \ u e e n / \/ a w a z anl |
Thread Tools | |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
RE- DSLR's that take AA batteries | Clive | Digital Photography | 2 | August 1st 06 12:49 AM |
SLR lens on DSLR's | Zeno | Digital Photography | 46 | October 21st 05 10:15 AM |
DSLR'S, very disappointed | [email protected] | Digital Photography | 135 | April 13th 05 07:58 PM |
DSLR'S, very disappointed | [email protected] | Digital Photography | 0 | April 6th 05 10:02 AM |
highlight with C1 Converter | leo | Digital Photography | 7 | October 9th 04 03:53 AM |