This topic is in the 2013 archive. You can read it and like its posts here. It is read only.
Archive 2013So, Uh...Canon shadow noise?
p.8 #1So, Uh...Canon shadow noise?
form wrote:
You want me to have a racing accident in my souped up ford mustang?
Do you have a Ford Mustang?
You want me to have a racing accident in my souped up ford mustang?
Do you have a Ford Mustang?
Jun 03, 2013 at 12:41 PM
p.8 #3So, Uh...Canon shadow noise?
We could have made a small fortune on those shots... 
Jun 03, 2013 at 12:52 PM
p.8 #4So, Uh...Canon shadow noise?
Probably doesn't have a Mustang because the black tires had too much chroma noise. 
Jun 03, 2013 at 01:24 PM
p.8 #5So, Uh...Canon shadow noise?
Most tires are black. Do you just feel an uncontrollable urge to find niggling ways to insult me?
Jun 03, 2013 at 01:30 PM
p.8 #6So, Uh...Canon shadow noise?
..................where ETTR would not be beneficial.
------------------------------------------
ETTR is always possible. EPTR is what's bad. Nobody has said to Expose Past The Right.
------------------------------------------
ETTR is always possible. EPTR is what's bad. Nobody has said to Expose Past The Right.
Jun 03, 2013 at 01:30 PM
p.8 #7So, Uh...Canon shadow noise?
What is "past" the right? Does right have a right-er?
I'll add a literal case: When part of the face is in sun and part in shade, how do you decide you are "exposing to the right?"
I'll add a literal case: When part of the face is in sun and part in shade, how do you decide you are "exposing to the right?"
Jun 03, 2013 at 01:34 PM
p.8 #9So, Uh...Canon shadow noise?
Please post RAW files of the two original images.
Jun 03, 2013 at 01:47 PM
p.8 #10So, Uh...Canon shadow noise?
form wrote:
Most tires are black. Do you just feel an uncontrollable urge to find niggling ways to insult me?
Most tires are black. Do you just feel an uncontrollable urge to find niggling ways to insult me?
Jun 03, 2013 at 01:48 PM
p.8 #12So, Uh...Canon shadow noise?
snapsy wrote:
Nice demonstration. I am curious how well the D800 handles apparently/seemingly-blown highlights. If I'm reading the EXIF data correctly it looks like the exposure for the outdoors was at 1/25th second, f/5.6, ISO 100 and the exposure for the interior was at 1/6th second with the same f/5.6 and ISO 100? If this is accurate, a 2-stop difference in exposure duration, then I am curious how well the D800 can bring down the 2-stop overblown highlights in the single shot 1/6th second exposure version without blending in the 1/25th second outdoor exposure.
My point is this...there are two ways to approach ambient-light backlit scenes, expose for the highlights and try to, um, rassle-up the shadows, or expose for the shadows and try to, uh, rassle-down the highlights.
Any chance I could trouble you to post-process the original 1/6th exposure so we can see how well the D800 image data preserves those seemingly blown highlights in their RAW files? Or, simpler, still, an exported version using the RAW's native Linear Gamma of the 1/6th second single, unblended, exposure?
Methinks there is more here than meets the eye!
Thanks in advance!
Jimmy GHighlight recovery is principally a function of metering decisions rather than sensor differences since it relates to how close the brightest highlight with detail the photographer wants retained is metered relative to the saturation point on the sensor.
Hi snapsy, and thanks for taking the time to reply,
Your technical point about metering decisions is all fine and well but what I was hoping you could demonstrate with a Linear Gamma version of your 1/6th second image was that there is, indeed, an awful lot of highlight detail that can be recovered in our RAW files if we take the time to look. Going back to the OP's sample image with the backlit couple I was hoping to offer him another suggestion for that sort of lighting condition beyond my original High-Key suggestion...shooting for highlight recovery.
Anyhoo, thanks for your feedback, maybe I should have been a bit more upfront on what I was looking to accomplish with my request.
Thanks again!
Jimmy G
jimmy462 wrote
Hi snapsy,Nice demonstration. I am curious how well the D800 handles apparently/seemingly-blown highlights. If I'm reading the EXIF data correctly it looks like the exposure for the outdoors was at 1/25th second, f/5.6, ISO 100 and the exposure for the interior was at 1/6th second with the same f/5.6 and ISO 100? If this is accurate, a 2-stop difference in exposure duration, then I am curious how well the D800 can bring down the 2-stop overblown highlights in the single shot 1/6th second exposure version without blending in the 1/25th second outdoor exposure.
My point is this...there are two ways to approach ambient-light backlit scenes, expose for the highlights and try to, um, rassle-up the shadows, or expose for the shadows and try to, uh, rassle-down the highlights.
Any chance I could trouble you to post-process the original 1/6th exposure so we can see how well the D800 image data preserves those seemingly blown highlights in their RAW files? Or, simpler, still, an exported version using the RAW's native Linear Gamma of the 1/6th second single, unblended, exposure?
Methinks there is more here than meets the eye!
Thanks in advance!
Jimmy G
Hi snapsy, and thanks for taking the time to reply,
Your technical point about metering decisions is all fine and well but what I was hoping you could demonstrate with a Linear Gamma version of your 1/6th second image was that there is, indeed, an awful lot of highlight detail that can be recovered in our RAW files if we take the time to look. Going back to the OP's sample image with the backlit couple I was hoping to offer him another suggestion for that sort of lighting condition beyond my original High-Key suggestion...shooting for highlight recovery.
Anyhoo, thanks for your feedback, maybe I should have been a bit more upfront on what I was looking to accomplish with my request.
Thanks again!
Jimmy G
Jun 03, 2013 at 01:54 PM
p.8 #13So, Uh...Canon shadow noise?
?
The whole point of exposing to the right is to deal with high contrast scenes.
Why is that hard to understand?
Expose until your lightest tones are almost clipped, then raise the shadows slightly if needed. For me, it means about 1/2 stop over, and shadows raised about 1/2 stop for high contrast California landscape lighting. I don't see form's problem, other than trying to make cameras do something they aren't designed to do.
Get used to the idea that some shadow areas don't have any significant content or detail.
One can't base their work on gear that doesn't exist.
The whole point of exposing to the right is to deal with high contrast scenes.
Why is that hard to understand?
Expose until your lightest tones are almost clipped, then raise the shadows slightly if needed. For me, it means about 1/2 stop over, and shadows raised about 1/2 stop for high contrast California landscape lighting. I don't see form's problem, other than trying to make cameras do something they aren't designed to do.
Get used to the idea that some shadow areas don't have any significant content or detail.
One can't base their work on gear that doesn't exist.
Jun 03, 2013 at 01:58 PM
p.8 #14So, Uh...Canon shadow noise?
jimmy462 wrote:
Your technical point about metering decisions is all fine and well but what I was hoping you could demonstrate with a Linear Gamma version of your 1/6th second image was that there is, indeed, an awful lot of highlight detail that can be recovered in our RAW files if we take the time to look. Going back to the OP's sample image with the backlit couple I was hoping to offer him another suggestion for that sort of lighting condition beyond my original High-Key suggestion...shooting for highlight recovery.
Anyhoo, thanks for your feedback, maybe I should have been a bit more...Show more →
Jimmy, here is a raw histogram of one of the windows for the interior exposure of the two exposure photos, plus what it looks like with 100% highlight recovery in LR4. Note that the red and blue channels are clipped.
http://horshack.smugmug.com/photos/i-W2bbttq/0/O/i-W2bbttq-O.jpg
http://horshack.smugmug.com/photos/i-h83tS7G/0/XL/i-h83tS7G-XL.jpg
Your technical point about metering decisions is all fine and well but what I was hoping you could demonstrate with a Linear Gamma version of your 1/6th second image was that there is, indeed, an awful lot of highlight detail that can be recovered in our RAW files if we take the time to look. Going back to the OP's sample image with the backlit couple I was hoping to offer him another suggestion for that sort of lighting condition beyond my original High-Key suggestion...shooting for highlight recovery.
Anyhoo, thanks for your feedback, maybe I should have been a bit more...Show more →
Jimmy, here is a raw histogram of one of the windows for the interior exposure of the two exposure photos, plus what it looks like with 100% highlight recovery in LR4. Note that the red and blue channels are clipped.
http://horshack.smugmug.com/photos/i-W2bbttq/0/O/i-W2bbttq-O.jpg
http://horshack.smugmug.com/photos/i-h83tS7G/0/XL/i-h83tS7G-XL.jpg
Jun 03, 2013 at 02:30 PM
p.8 #15So, Uh...Canon shadow noise?
form wrote:
My experience with highlight recovery is that the overexposure highlight recovery gradation is much more abrupt (in LR4) with the D800 than the Canon 5d2. I most often noticed wedding dresses or white suits that were maybe 1 stop or less overexposed, but which would have an intense, difficult-to-remove light blue tint when the highlights were brought down. By contrast, the 5d2 had no such tint when blown highlights were recovered - either the detail was there or it wasn't, but there was no tinting like that. I also felt - without proving it personally - that the highlight ceiling...Show more →
Hi form,
Thanks for offering some of your experiences here on this, it helps to flesh out for those of us trying to help here where it is you are experiencing some limitations with the technology.
I'd like to suggest a different approach here for you in dealing with the issues you're encountering. This whole thread is all over the place with a variety of issues being brought up, a lot of cross-talking, and, well, a lot of frustrations going around. So, I'm not sure how it is that you are able to sift through all the chaos for some of the answers you seek.
Case in point, your above reply to me. The subject of "the downsides to highlight recovery" is worthy of a thread unto itself. Oh, say, something along the lines of..."Hey folks, I've been using highlight recovery in backlit situations on my 5D3 and have been encountering odd blue-tints in my results. I've tried a variety of things (giving examples) and nothing seems to work. What have you folks been doing?" And then that thread will populate with all sorts of suggestions, tips, techniques, etc. Instead, this great topic for discussion (which would benefit not only you but the whole community) is buried amidst the ruble of this aimless thread.
So, here's my suggestion when for posting for help..."focus on your subject matter".
Yeah, it was that bad.
Anyhoo, my suggestion for your above situation...color casts in mixed ambient lighting I address by blending two color temperature corrected versions of the same image to suit my tastes. But, according to my tastes, shadows are "cooler" in the real world, just as late day sun is "warmer", and I take a light hand to those adjustments. My adjustments and technique may not suit your needs. Um, what do you do?
And with that, some more pearls of my experience go to a watery grave beneath the sediments of some more, uh, (forgive me) Canon Forum Noise.
Best to you,
Jimmy G
jimmy462 wrote
Nice demonstration. I am curious how well the D800 handles apparently/seemingly-blown highlights. If I'm reading the EXIF data correctly it looks like the exposure for the outdoors was at 1/25th second, f/5.6, ISO 100 and the exposure for the interior was at 1/6th second with the same f/5.6 and ISO 100? If this is accurate, a 2-stop difference in exposure duration, then I am curious how well the D800 can bring down the 2-stop overblown highlights in the single shot 1/6th second exposure version without blending in the 1/25th second outdoor exposure.Hi form,
Thanks for offering some of your experiences here on this, it helps to flesh out for those of us trying to help here where it is you are experiencing some limitations with the technology.
I'd like to suggest a different approach here for you in dealing with the issues you're encountering. This whole thread is all over the place with a variety of issues being brought up, a lot of cross-talking, and, well, a lot of frustrations going around. So, I'm not sure how it is that you are able to sift through all the chaos for some of the answers you seek.
Case in point, your above reply to me. The subject of "the downsides to highlight recovery" is worthy of a thread unto itself. Oh, say, something along the lines of..."Hey folks, I've been using highlight recovery in backlit situations on my 5D3 and have been encountering odd blue-tints in my results. I've tried a variety of things (giving examples) and nothing seems to work. What have you folks been doing?" And then that thread will populate with all sorts of suggestions, tips, techniques, etc. Instead, this great topic for discussion (which would benefit not only you but the whole community) is buried amidst the ruble of this aimless thread.
So, here's my suggestion when for posting for help..."focus on your subject matter".
Yeah, it was that bad.Anyhoo, my suggestion for your above situation...color casts in mixed ambient lighting I address by blending two color temperature corrected versions of the same image to suit my tastes. But, according to my tastes, shadows are "cooler" in the real world, just as late day sun is "warmer", and I take a light hand to those adjustments. My adjustments and technique may not suit your needs. Um, what do you do?
And with that, some more pearls of my experience go to a watery grave beneath the sediments of some more, uh, (forgive me) Canon Forum Noise.
Best to you,
Jimmy G
Jun 03, 2013 at 02:40 PM
p.8 #16So, Uh...Canon shadow noise?
What a waste of time. I just spent a portion of my life reading this thread, hoping to offer solutions or hear other people's ideas... and if the op is not willing to pony up the RAW files so we can see what/if anything might resolve the issue then I'm done with this thread.
I don't like "attack the poster' messages, and find myself feeling uncomfortable contributing to it, but clearly the op could use some help since he fails to understand the basic concept of exposing to the right.
So does he really want any help, or is he just bitching because this forum provides him the opportunity to do so?
Whatever. Moving on to the next topic....
I don't like "attack the poster' messages, and find myself feeling uncomfortable contributing to it, but clearly the op could use some help since he fails to understand the basic concept of exposing to the right.
So does he really want any help, or is he just bitching because this forum provides him the opportunity to do so?
Whatever. Moving on to the next topic....
Jun 03, 2013 at 02:56 PM
p.8 #17So, Uh...Canon shadow noise?
snapsy wrote:
Anyhoo, thanks for your feedback, maybe I should have been a bit more upfront on what I was looking to accomplish with my request.
Thanks again!
Jimmy GJimmy, here is a raw histogram of one of the windows for the interior exposure of the two exposure photos, plus what it looks like with 100% highlight recovery in LR4. Note that the red and blue channels are clipped.
http://horshack.smugmug.com/photos/i-W2bbttq/0/O/i-W2bbttq-O.jpg
http://horshack.smugmug.com/photos/i-h83tS7G/0/XL/i-h83tS7G-XL.jpg
Hi snapsy,
Great, thanks for posting this up! A few caveats first...I'm not a LR user but my understanding is that you are working with a 2.2 Gamma Curve adjusted interpretation of the original RAW file's 1.0 Gamma, and I'm not sure how one can access the Linear 1.0 Gamma view of that RAW image in LR. (Perhaps someone can chime in here... ) I seem to get better recovery going the Linear Gamma route, YMMV. I access the Linear Gamma for my Canon RAWs in DPP, again, not sure how others do it.
Regardless, this example suffices enough to make my point. Comparing your out-the-window "recovered highlight results" with the out-the-window results from the 1/25th second exposure...
Orig lower exposure image
...we can see that the highlight clipping is quite close between the two (though slightly worse on the recovered 1/6th image). I'm thinking, quite usable in a pinch.
So, in the run-and-gun scenarios of shooting a wedding where one just might not be able to grab a second exposure for the backgrounds, I would consider shooting for proper exposure on my shadowed subjects and trying to rassle-down my highlights rather than shooting for my highlights and trying to rassle-up my shadowed subjects (read: MOB & FOB). Or, shoot for a compromise between the two extremes as Gunzorro and others have suggested. Or, go "high key" and walk away.
And, for those times when no matter what we shoot for there was no nice, workable solution to get what we wanted...well, belly up to the bar, I'll buy you one while you tell me your tales of woe and I'll regale you with my tales of lost fortunes spent on the myriad slides and negatives that went straight to the trash can...and you better have a few spare fins for some buy backs.
Best,
Jimmy G
jimmy462 wrote
Your technical point about metering decisions is all fine and well but what I was hoping you could demonstrate with a Linear Gamma version of your 1/6th second image was that there is, indeed, an awful lot of highlight detail that can be recovered in our RAW files if we take the time to look. Going back to the OP's sample image with the backlit couple I was hoping to offer him another suggestion for that sort of lighting condition beyond my original High-Key suggestion...shooting for highlight recovery.Anyhoo, thanks for your feedback, maybe I should have been a bit more upfront on what I was looking to accomplish with my request.
Thanks again!
Jimmy G
http://horshack.smugmug.com/photos/i-W2bbttq/0/O/i-W2bbttq-O.jpg
http://horshack.smugmug.com/photos/i-h83tS7G/0/XL/i-h83tS7G-XL.jpg
Hi snapsy,
Great, thanks for posting this up! A few caveats first...I'm not a LR user but my understanding is that you are working with a 2.2 Gamma Curve adjusted interpretation of the original RAW file's 1.0 Gamma, and I'm not sure how one can access the Linear 1.0 Gamma view of that RAW image in LR. (Perhaps someone can chime in here...
) I seem to get better recovery going the Linear Gamma route, YMMV. I access the Linear Gamma for my Canon RAWs in DPP, again, not sure how others do it.Regardless, this example suffices enough to make my point. Comparing your out-the-window "recovered highlight results" with the out-the-window results from the 1/25th second exposure...
Orig lower exposure image
...we can see that the highlight clipping is quite close between the two (though slightly worse on the recovered 1/6th image). I'm thinking, quite usable in a pinch.
So, in the run-and-gun scenarios of shooting a wedding where one just might not be able to grab a second exposure for the backgrounds, I would consider shooting for proper exposure on my shadowed subjects and trying to rassle-down my highlights rather than shooting for my highlights and trying to rassle-up my shadowed subjects (read: MOB & FOB).
Or, shoot for a compromise between the two extremes as Gunzorro and others have suggested. Or, go "high key" and walk away.And, for those times when no matter what we shoot for there was no nice, workable solution to get what we wanted...well, belly up to the bar, I'll buy you one while you tell me your tales of woe and I'll regale you with my tales of lost fortunes spent on the myriad slides and negatives that went straight to the trash can...and you better have a few spare fins for some buy backs.
Best,
Jimmy G
Jun 03, 2013 at 03:58 PM
p.8 #18So, Uh...Canon shadow noise?
http://www.joeyallenphoto.com/IMG_1842.CR2 - Image 1 taken with 5d2, NOT 5d3 - however IMO results are basically identical at these ISOs
http://www.joeyallenphoto.com/IMG_1449.CR2 - Image 2, 5d3 photo
http://www.joeyallenphoto.com/IMG_1530.CR2 - Extra 5d3 photo at ISO400 with high dynamic range, sunlight on the bride but nobody else. This was also the only time she went over like this with the groom's parents.
http://www.joeyallenphoto.com/IMG_5091.CR2 - Another 5d3 photo, ISO100, pretty high contrast environment, the b&g just went under shade and I was in manual without having made adjustments to compensate between this shot and the last one about 1 second prior. Note pushing 2 stops and the effect on the groom's suit (not a black suit, but a grey one).
Edited on Jun 03, 2013 at 04:13 PM · View previous versions
http://www.joeyallenphoto.com/IMG_1449.CR2 - Image 2, 5d3 photo
http://www.joeyallenphoto.com/IMG_1530.CR2 - Extra 5d3 photo at ISO400 with high dynamic range, sunlight on the bride but nobody else. This was also the only time she went over like this with the groom's parents.
http://www.joeyallenphoto.com/IMG_5091.CR2 - Another 5d3 photo, ISO100, pretty high contrast environment, the b&g just went under shade and I was in manual without having made adjustments to compensate between this shot and the last one about 1 second prior. Note pushing 2 stops and the effect on the groom's suit (not a black suit, but a grey one).
Edited on Jun 03, 2013 at 04:13 PM · View previous versions
Jun 03, 2013 at 04:03 PM
p.8 #19So, Uh...Canon shadow noise?
It's not an expose to the right situation. It's a high contrast situation. Showing us a raw might help but ultimately he is hitting a photographic law here: you can get bright highlights and dark shadows on the same shot without having problems.
The only way he can get past this photographic law is by lifting the shadows in shot - fill flash.
He is getting shadow noise because he is pushing shadows. He is pushing shadows because he isn't filling.
Joey, respectfully, learn to fill. Get to the ceremony a minimum 30 minutes beforehand and nail down your settings. Your life will change.
It really is that simple. It's just one of those photographic rules that you need to employ. Luckily on M, A, or, T, the 5D3 defaults to fill mode so you can dial in your manual exposure and the camera does a half decent job. You can chimp and ride the exposure as necessary, bearing in mind that a little over is usually recoverable. If you want it more reliable then embrace it manually.
Edited on Jun 03, 2013 at 04:41 PM · View previous versions
The only way he can get past this photographic law is by lifting the shadows in shot - fill flash.
He is getting shadow noise because he is pushing shadows. He is pushing shadows because he isn't filling.
Joey, respectfully, learn to fill. Get to the ceremony a minimum 30 minutes beforehand and nail down your settings. Your life will change.
It really is that simple. It's just one of those photographic rules that you need to employ. Luckily on M, A, or, T, the 5D3 defaults to fill mode so you can dial in your manual exposure and the camera does a half decent job. You can chimp and ride the exposure as necessary, bearing in mind that a little over is usually recoverable. If you want it more reliable then embrace it manually.
Edited on Jun 03, 2013 at 04:41 PM · View previous versions
Jun 03, 2013 at 04:07 PM
p.8 #20So, Uh...Canon shadow noise?
Well we all know HTP is the death of shadow chroma from the start so lets begin by throwing out the HTP CR2...
Jun 03, 2013 at 04:26 PM