Old barn Opinions on overall brightness?

[ccphoto]

Just wondered if three levels of "gray" with a tinge of blue would help to separate the sky from the roof and the foreground snow. Also gave a bit more color pop to the barn wood and foreground grasses/twigs, etc.

[tbob]

Chris; I will add this to the "grist for the mill" in considering options for processing. Often I have to revisit the image several times as things slowly percolate. I try not to wander too far astray, always like to keep a couple of versions stored to refer to, I find it helps to prevent me doing some very strange things.

[george013]

Trevor,

Your blue and green channels are seriously clipping.

George

[xpatUSA]

Depending on how much can be recovered at each end, make say 5 to 7 images from the raw and fuse ** them?

** Like Enfuse or TuFuse ...

Found this for the Mac. There may still be others ...

http://software.bergmark.com/enfuseGUI/Main.html
.

[tbob]

Trevor,

Your blue and green channels are seriously clipping.

George

Which version? I checked the last one I did, albeit I did some further processing so it may be changed from the posted versions, and there is minimal to no clipping. The peak touches the upper margin of the histogram, still a sharp point. RBG clipping is something I will have to educate myself on, more concerned with shadow and highlight clipping.

[Manfred M]

Thanks for high key definition(s). I struggled with what exactly constitutes a high key image as they seemed so variable. Maybe because there are two forms?

The decision has been made. My wife prefers the less "high key" version. Since I defer to her in all things artistic, there is no point in proceeding further (Or at least showing her any other versions)

The "classic" definition of High Key comes from early television where the broadcast equipment could not handle could not handle significant contrast. These scenes were lit so that the key (main) light and fill light were quite balanced approaching a 1:1 lighting ratio. This means that low key lighting is almost shadow free.

A more modern (photographic?) view of high key lighting tends to carry on the look with very little shadow, but also tends to have subjects that are predominately light tones (often white) but will have some darker toned elements. Proper exposure with minimal clipping is still the aim in these images.

[Manfred M]

Which version? I checked the last one I did, albeit I did some further processing so it may be changed from the posted versions, and there is minimal to no clipping. The peak touches the upper margin of the histogram, still a sharp point. RBG clipping is something I will have to educate myself on, more concerned with shadow and highlight clipping.

Trevor - what you uploaded is not necessarily the same file that sources like TinyPic, Facebook, etc. links display. These sites can and do rework your uploads to save space on their servers. I've uploaded images that show no clipping when I prepare them for upload, but when I download them the storage site, they sometimes do show problems with highlight and / or shadow details.

RGB clipping is nothing more than one of the channels clipping before the others, When all of the channels clip we get pure white highlights or pure black shadow detail, but when just one or more channels clip we get an inaccurate colour representation.

[xpatUSA]

Which version? I checked the last one I did, albeit I did some further processing so it may be changed from the posted versions, and there is minimal to no clipping.

Last one you did looks OK - very little clipping at all.

The peak touches the upper margin of the histogram, still a sharp point.

Does your histogram auto-scale? If so, one peak will always reach the top and go no further; meaning that somewhat less can be deduced from it.

RawTherapee also allows an absolute view where counts can disappear off the top, FWIW.

RBG clipping is something I will have to educate myself on, more concerned with shadow and highlight clipping.

Well worth doing, IMO. There are indeed some wrinkles in RGB views ...

[george013]

Which version? I checked the last one I did, albeit I did some further processing so it may be changed from the posted versions, and there is minimal to no clipping. The peak touches the upper margin of the histogram, still a sharp point. RBG clipping is something I will have to educate myself on, more concerned with shadow and highlight clipping.

I toke the first one. If you've the shot still on your card put it in the camera and check the histogram. There're 4 different histograms: brightness, red, green and blue. The brightness is ok. But the blue and green aren't. I would check if it's on the original to, otherwise it's due to pp. When I'm well informed the camera has a threshold of 3 or 4 to play save.

George

[xpatUSA]

Originally Posted by tbob Which version?

I toke the first one. If you've the shot still on your card put it in the camera and check the histogram. There're 4 different histograms: brightness, red, green and blue. The brightness is ok. But the blue and green aren't. I would check if it's on the original to, otherwise it's due to pp.

I have no idea why George is still talking about post #1 when the rest of us are talking about Trevor's post #15 !!!

Just so's we're all clear, here are two of the histogram options in RawTherapee, 1) auto-scaled and 2) not auto-scaled:

When I'm well informed the camera has a threshold of 3 or 4 to play save.

[Manfred M]

I toke the first one. If you've the shot still on your card put it in the camera and check the histogram. There're 4 different histograms: brightness, red, green and blue. The brightness is ok. But the blue and green aren't. I would check if it's on the original to, otherwise it's due to pp. When I'm well informed the camera has a threshold of 3 or 4 to play save.

George

I'm not sure why looking at the histograms in the camera, rather than the ones supplied in PP software would something one would do, especially if we are looking for clipping issues. The camera histogram uses jpeg data and is quite conservative. For instance, when my camera shows clipping in a channel, when I get the image over to my PP software, I will find that the raw data will have between 1 and 1-1/2 eV of latitude in both the shadows and highlights areas.

The PP software will show clipping as it interprets the raw data more correctly.

[george013]

I'm not sure why looking at the histograms in the camera, rather than the ones supplied in PP software would something one would do, especially if we are looking for clipping issues. The camera histogram uses jpeg data and is quite conservative. For instance, when my camera shows clipping in a channel, when I get the image over to my PP software, I will find that the raw data will have between 1 and 1-1/2 eV of latitude in both the shadows and highlights areas.

The PP software will show clipping as it interprets the raw data more correctly.

I don't know what software Trevor has or what it's able to. I even don't know if he shot jpg or raw. Did he shoot raw then there might be the possibility to change color balance. Otherwise he could use a gray point to correct the colors.
Clipping info is always based on the jpg, more precisely the rgb raster image. And Nikon is using a threshold like I wrote in the former post.
And I still don't know if he checked for blinkies only for brightness or for the individual channels. Mostly brightness will be enough but with all that snow, close to white, it might be important to check the individual channels.

George

[tbob]

I don't know what software Trevor has or what it's able to. I even don't know if he shot jpg or raw. Did he shoot raw then there might be the possibility to change color balance. Otherwise he could use a gray point to correct the colors.
Clipping info is always based on the jpg, more precisely the rgb raster image. And Nikon is using a threshold like I wrote in the former post.
And I still don't know if he checked for blinkies only for brightness or for the individual channels. Mostly brightness will be enough but with all that snow, close to white, it might be important to check the individual channels.

George

I use Capture One for processing.

As for exposure, I shoot almost exclusively landscapes and my technique is to set the aperture I want (I use Aperture priority as my default), use one focal point to get the focus I want, lock the focus and move the same spot to something approximating grey to let the camera set the exposure . If I get more "blinkies" than I feel I can deal with in post processing, in areas important for the image, I adjust the EV and reshoot. Same process if the shadows are excessively clipped. Has never occurred to me to check RGB clipping as an option. I am not even sure if I even need to worry about it in the field. My brain is already pushed to the limit with focus, depth of field, exposure (or my crude approximation) and composition.

[george013]

I use Capture One for processing.

As for exposure, I shoot almost exclusively landscapes and my technique is to set the aperture I want (I use Aperture priority as my default), use one focal point to get the focus I want, lock the focus and move the same spot to something approximating grey to let the camera set the exposure . If I get more "blinkies" than I feel I can deal with in post processing, in areas important for the image, I adjust the EV and reshoot. Same process if the shadows are excessively clipped. Has never occurred to me to check RGB clipping as an option. I am not even sure if I even need to worry about it in the field. My brain is already pushed to the limit with focus, depth of field, exposure (or my crude approximation) and composition.

So you didn't check the individual channels. I think it's quite important with all that snow. Pay attention to it.

Another thing I was wondering about is that spot metering. You spotmeter on a subject that should me something about grey and then you correct based on blinkies in the white. Why not just matrix? Much more accurate.

A wise man said "The only way to never make mistakes is to do nothing."

George

[JBW]

I use Capture One for processing.

As for exposure, I shoot almost exclusively landscapes and my technique is to set the aperture I want (I use Aperture priority as my default), use one focal point to get the focus I want, lock the focus and move the same spot to something approximating grey to let the camera set the exposure . If I get more "blinkies" than I feel I can deal with in post processing, in areas important for the image, I adjust the EV and reshoot. Same process if the shadows are excessively clipped. Has never occurred to me to check RGB clipping as an option. I am not even sure if I even need to worry about it in the field. My brain is already pushed to the limit with focus, depth of field, exposure (or my crude approximation) and composition.

and I thought manual macro was complicated

[DanK]

Hmm. I think this really boils down to one issue, in my simplistic thinking: you have too much almost entirely detail-free white. Seems to me that the only two things you can do are to darken it and reduce the amount of it.

I certainly wouldn't add a color cast.

Re this:

Clipping info is always based on the jpg, more precisely the rgb raster image.

That's just wrong. I think a correct version of this would be: "any histogram reflects a rendering of the image." The in-camera processing needed to produce a jpg will often shift the histogram to the right. I don't shoot Nikon, but with Canon, you can minimize this problem by picking the "faithful" picture style and setting all processing parameters to zero. That only approximates the raw converter's histogram, which is the best indication of whether you have more clipping than you can recover, but I find that on my Canon bodies, this approach is usually close enough to let me avoid clipping without going much darker than is really necessary.

Another thing I was wondering about is that spot metering. You spotmeter on a subject that should me something about grey and then you correct based on blinkies in the white. Why not just matrix? Much more accurate.

I think the right answer about metering is "whatever works for you." However, in a scene like that, I personally wouldn't use matrix metering because it takes away control. I would use spot metering, but on the snow. I would start by spot metering on a bright spot in the snow and then opening up roughly two stops. That tells you what exposure will get you whites that aren't blown out. You can play around with how much to open it up to turn the gray into a suitable white. I would then spot meter on the barn to see how far off it is. If the metering based on the snow would put the barn into such deep shadows that pulling it up would be noisy--or worse, cause clipping at the bottom--I'd bracket exposures.

[tbob]

So you didn't check the individual channels. I think it's quite important with all that snow. Pay attention to it.

Another thing I was wondering about is that spot metering. You spotmeter on a subject that should me something about grey and then you correct based on blinkies in the white. Why not just matrix? Much more accurate.

A wise man said "The only way to never make mistakes is to do nothing."

George

I will try the matrix setting, see if I am comfortable with the additional step of going from setting a focal point and then shifting the exposure point to resetting to matrix metering. Not that complex and may be a useful change to my current routine.

[george013]

I will try the matrix setting, see if I am comfortable with the additional step of going from setting a focal point and then shifting the exposure point to resetting to matrix metering. Not that complex and may be a useful change to my current routine.

I'm not sure if I understand you well. It's not an additional step, just set the camera on matrix and correct based on your diagram. Point and shoot, examine and eventual do it over.
When Dan says he's using spot and correct it, he's doing the same: get a start setting and correct it. But he knows the snow must be on the right site of the histogram so he can adjust it in advance with 2 stops. But he still has to check it.

George

[DanK]

But he knows the snow must be on the right site of the histogram so he can adjust it in advance with 2 stops. But he still has to check it.

to be very careful, yes, but just doing two stops will generally work.

The difference is that in the case of matrix metering, the result will be affected by factors that are irrelevant to the problem, which is the variance in tonality. For example, the size of the barn (more generally, the proportion of the frame that is dark) will affect the camera's guess. If you spot meter and open up two stops, you can be certain that you won't blow the whites, regardless of the size of the barn.

My film camera for years was a Canon FTb, which only had spot metering. Nikons at the time had only center-weighted averaging, if I recall correctly. So some of what underlies my preference is that I grew up using spot metering and am comfortable finding the tone in the scene off which I want to meter. I am also used to some work-arounds--e.g., meter off your palm and open up one stop. However, the other reason I resort to spot metering--and sometimes multiple meterings--is because of the level of control it provides.

[xpatUSA]

to be very careful, yes, but just doing two stops will generally work.

The difference is that in the case of matrix metering, the result will be affected by factors that are irrelevant to the problem, which is the variance in tonality. For example, the size of the barn (more generally, the proportion of the frame that is dark) will affect the camera's guess. If you spot meter and open up two stops, you can be certain that you won't blow the whites, regardless of the size of the barn.

My film camera for years was a Canon FTb, which only had spot metering. Nikons at the time had only center-weighted averaging, if I recall correctly. So some of what underlies my preference is that I grew up using spot metering and am comfortable finding the tone in the scene off which I want to meter. I am also used to some work-arounds--e.g., meter off your palm and open up one stop. However, the other reason I resort to spot metering--and sometimes multiple meterings--is because of the level of control it provides.

+1!
I am a spot metering guy too, although my own foray into "real photography" was with a Praktica MTL3 armed with a battery-powered Cadmium Sulfide cell which was probably just a scene-averaging device - my, my, how much have we have progressed since then!

In spite of all the wonderful "matrix" metering with an ever-increasing number of points, what you're still getting is the opinion of some gentleman from a far-off land as to what you're actually looking at, even though he is not there looking over your shoulder.

Sux, IMHO.

Trevor, don't even bother with it.