Edingurh at dusk

[Ike Harel]

Edinburgh skyline from Calton Hill at dusk.
Most notable monuments are seen here: The Castle, The Balmoral, and Scott-monument, among other interesting building in this beautiful city.

Canon 60D, EF28-135, RAW, F5.6, 1/100, ISO125.
Picture date: Oct. 21st, 2016, at 17:16 local time.
Ike


*** Realized after posting, that the system reduce size of the original photo to 59kb / 700 pixels, and some parts appearing unsharpened. any clue how to avoid this?

[Urbanflyer]

I am struck by the great number of steeples. Thanks for sharing!

[Shadowman]

Nice shot.

[FootLoose]

Nice light in this scene, Ike.
I don't know about your posting problem. I link from Flikr and have no problems.

[DanK]

I link from Smugmug and also have no problems. I tell smugmug to provide a link to an image resized to 1024, give or take.

This is interesting lighting, but I frankly don't find the image very strong, and I couldn't figure out what you were trying to do. You didn't expose (or process) the non-sky part either dark enough to create a silhouette or light enough to show much of the interesting detail. This in-between exposure also produces and image that is low in contrast even though it extends the full range of the histogram.

If you wanted a silhouette, that would be easy to do in post. Bringing out shadow detail would be harder because the shadow areas are so dark, hence very low in terms of data content. One option when shooting at times like this is to bracket two or three exposures.

[Manfred M]

When I see a back-lit shot like this, the first thing that pops into my mind is to open up the shadows a bit to compensate for the backlighting. When shooting into the sun (even when it is hidden behind clouds) the camera's metering system will usually underexpose the image. Opening up the shadows will generally improve this type of shot.

[DanK]

When I see a back-lit shot like this, the first thing that pops into my mind is to open up the shadows a bit to compensate for the backlighting. When shooting into the sun (even when it is hidden behind clouds) the camera's metering system will usually underexpose the image. Opening up the shadows will generally improve this type of shot.

I agree, but it might be helpful to separate two different things. You noted the tendency of camera's metering systems (at least in most modes) to underexpose in backlit situations. However, that's not all that is going on here. This image has about as much range in luminance as the camera can handle. At least in the processed jpeg, the histogram is beginning to clip at both ends. So treating this as simple underexposure and increasing exposure wouldn't work. One option is to brighten the shadows while NOT brightening the rest of the image, which is of course what you did.

The reason I suggested bracketing as an alternative is that most of the area other than sky is scrunched in the bottom of the histogram, with values < 80 or 100.

[Manfred M]

I agree, but it might be helpful to separate two different things. You noted the tendency of camera's metering systems (at least in most modes) to underexpose in backlit situations. However, that's not all that is going on here. This image has about as much range in luminance as the camera can handle. At least in the processed jpeg, the histogram is beginning to clip at both ends. So treating this as simple underexposure and increasing exposure wouldn't work. One option is to brighten the shadows while NOT brightening the rest of the image, which is of course what you did.

The reason I suggested bracketing as an alternative is that most of the area other than sky is scrunched in the bottom of the histogram, with values < 80 or 100.

Agreed Dan - the metering system is the primary cause of the issue, but this is a symptom in some ways. When it comes to image making, it is all about the mid-tone values. Anchor the black point and white point and then work the mid-tones to get the image to look right.

[xpatUSA]

*** Realized after posting, that the system reduce size of the original photo to 59kb / 700 pixels, and some parts appearing unsharpened. Any clue how to avoid this?

As others have mentioned, try not to let the system mess with your image. This can be avoided by placing your image somewhere on the net that lets you access your image's URL and then typing or pasting the URL in between "img brackets" thus:

(img)your-image-URL(/img)
comment: use square brackets in your post, not parentheses

Above may or may not be easy to understand, feel free to ask questions.

Main point being that you will know what size the image will be and you can apply proper sharpening "for the web" beforehand - as opposed to discovering what the system did to your image after the fact ...

[Ike Harel]

Thank you all, guys, learnt something very important.
As I always try not to "play" with natural outcome of any picture - hence, do not play-out any PP to a point which I do not understand the "know-how-to-do", so the outcome of the picture is open to a difference of opinions. That is OK.
I have 7 shots from this very same point at different moments during dusk, could post darker shadows, could post brighter image from earlier hour, and could off-set backlight of golden marvelous glow - but this one shows the town in its most natural light and colors during an autumn afternoon (minus the lessening effect of sharpness when uploaded).

Maybe I have to elaborate my technique, but I am not going to become a photoshop artist. The fun is when taking the pictures - not when sitting by the PC and processing to a satisfactory outcome.
Thanks again, sincerely,
Ike

1.Earlier from same spot at 17:04 local time
2.Later at blue-hour at 18:09 local time.
Ike

[DanK]

Ike,

To preface this: my view is that this is a hobby for most people, so people should do whatever they like. I'm certainly not suggesting that you should necessarily do what I do.

However, I think your post raises a fundamental issue that has been discussed numerous times on this site. I'll recap this, in case it is useful. Just delete if this is all stuff you have thought about already or if you just don't find it helpful.

You wrote:

As I always try not to "play" with natural outcome of any picture

There is nothing "natural" about a SOOC (straight out of the camera) image. EVERY digital image has been processed. You get to choose how it's processed, not whether it is processed.

When you shoot with a Canon camera, you have to pick a "picture style", or create your own for the camera to use. That picture style is a postprocessing menu made up by engineers at Canon, in an effort to come up with styles of images that will work for many people much of the time. These styles differ in terms of contrast, sharpening, color balance, saturation, etc. The camera applies these recipes blindly, without any consideration of the characteristics of the capture. You can take the same image numerous times, using different picture styles, and you will get images that look different.

The alternative is to do this oneself. Some people become extremely sophisticated in their use of editing tools, but you can take control without doing that. For example, even if you limit yourself to the basics--tonal adjustments, vibrance/saturation, contrast, and sharpening, you can get a great deal of control. I'm personally somewhere in the middle. I don't--or at least very rarely--do really sophisticated editing and don't have the skills to do a lot of it, but I process all of my own images.

For example, whatever picture style you used, in combination with the tonality in the scene and the metering, resulted in a very low-contrast image. I'm guessing that when you stood there, you didn't perceive the vista as lacking in contrast. You can change that to better match what you saw.

In addition, sometimes nature just doesn't cooperate. For example, the lighting may suck. Then, using basic editing tools can compensate.

In case this is helpful, I'll post a few pairs of images.

Here's an example where the editing wasn't too dramatic, although it took a considerable amount of work. The biggest problem was the low-contrast lighting (common in Bergen, where I took this), which also made the colors look washed out. The sky and flag were distracting, so I got rid of them. I cropped the bottom in part to remove the disembodied heads.





This next pair is much more extreme in terms of the impact of editing. This is a night shot in an old neighborhood in Boston. The 'natural' image had lots of problems: wrong white balance, overly bright areas distracting the viewer in some areas, lines from car lights that had to be removed, etc. Key areas are underexposed because I didn't want to blow out the highlights. I kept myself to an ISO of 200, with a shutter speed of 20 seconds; with that ISO on my camera, I had enough shadow detail to get away with this. (I have no recollection why I didn't stick with ISO 100. Perhaps I was trying, unsuccessfully, to keep the exposure short enough that I wouldn't have a car driving by.) These are quite common problems in urban night photography--e.g., white balance is a problem because sodium vapor lamps are so yellow compared with some other lighting, and artificial lights will often give one too big a dynamic range.





People here post criticisms in an effort to help people improve their images. I posted the second of each pair on this site, and I received some useful criticism. However, if I had posted the first of each pair, people would have rightly pointed out the problems I tried to remove in processing, such as lack of contrast, lack of sharpness, distracting details, unpleasant white balance, etc.

Dan

Dan

[Manfred M]

Ike - what you have written is something I have heard many times from photographers. What you do with your images is certainly up to you, but your going in assumption of a "natural outcome of any picture" is simply incorrect. There is no such thing.

Your digital camera records data. If you shoot in JPEG mode some camera engineer has come up with an algorithm as to how that data will be interpreted and output. Your camera settings (even the default settings), are what determine that outcome. Regardless of the settings, the camera will do a number of operations to the image; sharpening, noise reduction, addition of contrast, etc. to modify the data that the camera has recorded.

When I use Photoshop, or any other tool, process an image, I do the same thing that the camera engineer has done; interpret the data that the camera has recorded. The main difference is that I am not limited by what the camera's algorithm has output, but rather can remember what I saw and make the image look more like what I remember seeing. I can also disregard what I saw and do something quite over the top. In my edit, I used my knowledge of the human visual system and how it works to create a view that is more likely similar to what you actually saw when you looked out over Edinburgh, as the human visual system tends to be more keyed to the mid-tone values.

Nothing I did to your image took a lot of effort; the output you see took me less than a minute with your original image. I would have likely done something quite similar to any image I captured had I been standing beside you taking this shot at the same time you were as, in my view, it is more likely a better interpretation of what you saw than your camera output.

[Ike Harel]

Ike - what you have written is something I have heard many times from photographers. What you do with your images is certainly up to you, but your going in assumption of a "natural outcome of any picture" is simply incorrect. There is no such thing.

Your digital camera records data. If you shoot in JPEG mode some camera engineer has come up with an algorithm as to how that data will be interpreted and output. Your camera settings (even the default settings), are what determine that outcome. Regardless of the settings, the camera will do a number of operations to the image; sharpening, noise reduction, addition of contrast, etc. to modify the data that the camera has recorded.

When I use Photoshop, or any other tool, process an image, I do the same thing that the camera engineer has done; interpret the data that the camera has recorded. The main difference is that I am not limited by what the camera's algorithm has output, but rather can remember what I saw and make the image look more like what I remember seeing. I can also disregard what I saw and do something quite over the top. In my edit, I used my knowledge of the human visual system and how it works to create a view that is more likely similar to what you actually saw when you looked out over Edinburgh, as the human visual system tends to be more keyed to the mid-tone values.

Nothing I did to your image took a lot of effort; the output you see took me less than a minute with your original image. I would have likely done something quite similar to any image I captured had I been standing beside you taking this shot at the same time you were as, in my view, it is more likely a better interpretation of what you saw than your camera output.

Dear Manfred,
I shoot RAW and edit by Canon-DPP (latest version) but never change or play with colors or texture. The edit you had done, gave a green tone to the roof-tiles that never exist in any roof-top in Edinburgh. Hope you will visit the beautiful city Edinburgh, and find 50 tones of grey, literally, all beautifully mingling.
Good evening,
Ike

[Ike Harel]

" The sky and flag were distracting, so I got rid of them" …
The sky added good sense of depth, and the flag is no bother at all - why removed or should be removed? I can agree with most of your notions mentioned, Dan, but removing existing objects, elements or changing colors (as wrote to Manfred about green roof tiles) - this has nothing to do with "camera engineered settings" as you wrote, this is merely the taste of the beholder.
By the way, my camera mostly set to "Standard" mode.
Ike

[Manfred M]

Dear Manfred,
I shoot RAW and edit by Canon-DPP (latest version) but never change or play with colors or texture. The edit you had done, gave a green tone to the roof-tiles that never exist in any roof-top in Edinburgh. Hope you will visit the beautiful city Edinburgh, and find 50 tones of grey, literally, all beautifully mingling.
Good evening,
Ike

Ike - Actually I have spent a few days in Edinburgh. I assume that the roofing material is slate.

I only changed the contrast / micro-contrast and mid-tone values in the image, so any "green tones" have come from from your edit / white balance settings in the posted image. Working with a posted jpeg rather than the raw data means that the white balance / colour temperature settings have been "baked in" to the image file.

When I explore the roof colours in your original posting, the cyan colour cast is there, but because the colours are so dark and our visual system is keyed to the mid-tones, we don't see them as readily as in a lighter version. If your DPP has a colour sampler tool (I've never used that software), check the roof colours with it.

[Ike Harel]

I thanks everybody for the suggestion, but I have to quite here - Japan is now playing masterfully over Belgium 2:0 .
Bye,
Ike

[DanK]

removing existing objects, elements or changing colors (as wrote to Manfred about green roof tiles) - this has nothing to do with "camera engineered settings" as you wrote, this is merely the taste of the beholder.
By the way, my camera mostly set to "Standard" mode.
Ike

Canon doesn't publish full specs for its picture styles--or at least I haven't found them--but they have published enough to indicate that "standard" is not "unaltered". Standard boosts saturation and sets the color balance to be quite vivid. (You can't read that from the settings in the menu, as the baseline value of 0 shown on the camera's menu is not the same from one style to another.) As one Canon publication noted:

The standard image looks crisp, like a successful snapshot, and the colour tone and saturation are set to obtain vivid colours.

If you want the camera to do nothing with color, you have to use neutral.

If you use most raw processors, this setting won't matter. With DPP, it does. If I am right, DPP reads the picture style setting from EXIF and applies that recipe itself in its initial rendering of the image. Of course, if you know all of the changes it makes, you can reverse them in DPP.

[Ike Harel]

Canon doesn't publish full specs for its picture styles--or at least I haven't found them--but they have published enough to indicate that "standard" is not "unaltered". Standard boosts saturation and sets the color balance to be quite vivid. (You can't read that from the settings in the menu, as the baseline value of 0 shown on the camera's menu is not the same from one style to another.) As one Canon publication noted:



If you want the camera to do nothing with color, you have to use neutral.

If you use most raw processors, this setting won't matter. With DPP, it does. If I am right, DPP reads the picture style setting from EXIF and applies that recipe itself in its initial rendering of the image. Of course, if you know all of the changes it makes, you can reverse them in DPP.

Thank you Dan, I will take a good prolonged reading on your notes.
Ike