What is the purpose of tone compensation?

[Mike Buckley]

Bruce,

As I understand it, loading custom curves into your camera has two primary purposes.

One reason is that doing so minimizes the amount of time required to post-process images. As an example, you may be capturing images at an equestrian event for sale to the riders, their family and friends. You will make more sales if you can get the images online for their review sooner than later. Minimizing the time required to post-process the images, even when doing so using batch processing, is very good in that situation.

The other main reason is that whatever curve is used at the time of capture affects the exposure. If you can nail the curve, you can also nail the exposure. Anything that you can do to achieve the ideal exposure in-camera rather than making adjustments during post-processing will result in both saving time and producing a higher quality image.

[Cantab]

As I understand it, loading custom curves into your camera has two primary purposes.
...
The other main reason is that whatever curve is used at the time of capture affects the exposure. If you can nail the curve, you can also nail the exposure. Anything that you can do to achieve the ideal exposure in-camera rather than making adjustments during post-processing will result in both saving time and producing a higher quality image.

Mike, many thanks for your comments. I'm going to try experimenting with this on flowers, where there are often very subtle gradations of light colours.

[xpatUSA]

By the way, that article mentions that Photoshop Elements has no Curves tool. The Curves tool was present in version 1 of the software and I'm sure ever since then, so it makes me wonder what else might be a bit inaccurate in that article.

Mike, To get actual curves in PSE6 I had to add a filter, IIRC.

It does have that Levels tool, though. Which affects all three channels' luminance at once, AFIK.

Wow, two acronyms in one post :-)

WTH, make that three ;-)

[John Morton]

Great shot of that egret, Christina; I really like that! I think you are right about the extra detail in the feathers, thanks to the low contrast setting in-camera. I am definitely going to try some of this when I have a little time free.

John, I did a brief internet search and my understanding is that in-camera tone adjustments in the Nikon system are like those in the Canon system: they affect jpeg's but have no permanent effect on RAW. If one shoots in RAW, am I correct in assuming that the benefit of having an in-camera reduction of contrast available is so that the image one sees in the LCD [LED monitor?] will give you an on the spot idea what the final result might be if one was to do no postprocessing?

Hi, Bruce;

Doing a little Internet reading to confirm what I remember about camera histograms, it turns out that these are in fact generated from the JPEGs that appear on the LCD screens of dSLRs. So in a roundabout way, how a tone curve affects the JPEG rendering of an image in-camera also determines how the histogram for the image is displayed; and of course, many people use the histogram to set their exposure properly... so as Mike notes:

The other main reason is that whatever curve is used at the time of capture affects the exposure. If you can nail the curve, you can also nail the exposure. Anything that you can do to achieve the ideal exposure in-camera rather than making adjustments during post-processing will result in both saving time and producing a higher quality image.

In a roundabout fashion, then, it seems that having the proper tone curve applied in camera gives the best JPEG results; and having good JPEG results then produces the most accurate histogram for the image, allowing one to more accurately assess the exposure needed to capture a great RAW file.

This brings us to a point that has been nagging at the back of my mind: does varying the tone curve in-camera actually affect the RAW file, or not? The answer to this appears to be, yes it does IF you use the histogram for an image to set the 'proper' exposure.

Any tone curve applied in-camera can be reset for a RAW file; but the exposure for an image is what it is once it is set in the camera.

I might be wrong in this, and I would certainly like to hear if I am; but as far as I can see now, the idea behind setting a custom tone curve for an image seems to be that of getting the best information one can about the best exposure to use for that image.

[Mike Buckley]

Mike, To get actual curves in PSE6 I had to add a filter, IIRC.

That's interesting. I happen to have PSE6 (you and I are probably the only two people who still have it). I have never used the Curves tool on it because I do those adjustments in my primary editor. It was so foreign to me that I was actually really surprised to see that you change the curve by moving sliders. I suppose it's possible that I added a filter to be able to use it but I doubt it. If I had added a filter to do that, you would think that I would have at least recognized how one uses the tool to change the curve.

By the way, I'm rather proud of myself for having gotten through that entire paragraph without using any acronyms. EDIT: Oops! I did use one. Darn!

[Mike Buckley]

Any tone curve applied in-camera can be reset for a RAW file; but the exposure for an image is what it is once it is set in the camera.

The exposure of a RAW file can also be easily changed and there are important reasons to do so. I conducted a test about that when I first began using my Nikon D7000 and its higher ISO capabilities. If I underexpose an image and then use the Curve tool or any other method to brighten parts of the image, I will generate a certain amount of noise when shooting at high ISOs. The higher the ISO and/or the more underexposed the image is, the more noise is generated. However, if I adjust the exposure before making any other edits, less noise is generated. This applies when using Nikon Capture NX2 on Nikon RAW files; I don't know about other combinations of software and RAW files.

[Brownbear]

Thank you everyone for all your replies... As usual, for me this forum is like photography school.

John, in my Nikon I can set the tone compensation and the vivid setting (assuming this means saturation)... It is in the Shooting Menu under the label custom.

Also I shoot jpeg and raw, and when I view the photos the raw photos definately look higher contrast the the jpegs but I can't upload the raw photo as a comparison as photo bucket does not allow me to do so.

[drjuice]

For folks who don't have this capability in camera, you can work with it in at least two pieces of software after you've loaded your RAW image into the software:

1. free for 30 days after which a donation is requested: FastStone Image Viewer (load the RAW image into the preview screen; doubleclick with the zoom icon active to put it into the edit screen; right click on the image to expose the choices menu; select colors; choose which method you want to use; do your thing).

2. not free but relatively inexpensive in my view: Corel Draw! comes with a piece of software called PhotoPaint (load the RAW image into the app; CTRL-T will open the curves window; do your thing; warning: loading RAW may take a bit of time just because of its sheer size, in my case a RAW image is usually around 38MB, not at all like a JPG or PNG which is usually something less than 1MB).

Enjoy!

virginia

[xpatUSA]

That's interesting. I happen to have PSE6 (you and I are probably the only two people who still have it). I have never used the Curves tool on it because I do those adjustments in my primary editor. It was so foreign to me that I was actually really surprised to see that you change the curve by moving sliders. I suppose it's possible that I added a filter to be able to use it but I doubt it. If I had added a filter to do that, you would think that I would have at least recognized how one uses the tool to change the curve.

Thanks Mike,

I went back and looked in PSE6 and did find "Adjust Color Curves" function buried under the "enhance" menu. It's only global, i.e. by curves it means all of them at the same. The little graph is shown but not adjustable by clicking on the line. There's kiddie's presets (thanks, Adobe) and then horizontal sliders to mess with the graph. All great fun but I've never used it myself either.

The one I added is called "Smart Curve" c. 2009 Alois Zingl. Don't recall where I found it. But you can drag the curve, select all or just one channel, select color space (no idea why), save and load settings and it has a histogram display too. Haven't used it much but it looks pretty good.

I reverted to PSE6 after getting irritated with the overly-intrusive Organizer in PSE7 - just couldn't get rid of it!

[xpatUSA]

This brings us to a point that has been nagging at the back of my mind: does varying the tone curve in-camera actually affect the RAW file, or not? The answer to this appears to be, yes it does IF you use the histogram for an image to set the 'proper' exposure.

Any tone curve applied in-camera can be reset for a RAW file; but the exposure for an image is what it is once it is set in the camera.

I might be wrong in this, and I would certainly like to hear if I am; but as far as I can see now, the idea behind setting a custom tone curve for an image seems to be that of getting the best information one can about the best exposure to use for that image.

John,

Not saying you're wrong but I'll try to add 2 cents-worth of clarification (or confusion).

As we all know, the image data in a RAW file is the sensor data. The only processing of the sensor data (Nikon) is noise reduction, ISO gain and their NEF compression, AFAIK. As I understand it, the RAW image data is not changed by any in-camera options. We should know too that a true RAW image is very dark and rarely seen as it is. Has to be decoded, and gamma applied, to make sense on any screen, camera or monitor. The only way to view a true RAW is to use a linear converter, for example DCraw with that option set.

Which means that if someone says they "viewed a RAW image", they really didn't.

However, extra data is included in the file - this is called "metadata" because it's not the image data itself. Metadata includes lots of things like model, date/time, image ID, lens used, exposure details, white balance, and much much much more.

Nikon RAW files can set to also include a JPEG which, on my D50, is BASIC quality only and that could well be the same as what is shown on the LCD. A few experiments with extreme option settings should show whether the JPEG on the LCD includes all those wonderful options so thoughtfully provided by Nikon and which I, for one, keep turned off at all times ;-)

[Mike Buckley]

Nikon RAW files can set to include a JPEG which, on my D50, is BASIC quality only

I believe (not sure) that model is the last one that does not embed a full-size high-resolution JPEG. If you open your RAW file in Capture NX2, make a change and save the file, the software will automatically embed a full-size high-resolution JPEG. This may also be possible using View NX2, which is free. It has been a long time since I have used that software, so I really don't know.

[John Morton]

John,

Nikon RAW files can set to also include a JPEG which, on my D50, is BASIC quality only and that could well be the same as what is shown on the LCD. A few experiments with extreme option settings should show whether the JPEG on the LCD includes all those wonderful options so thoughtfully provided by Nikon and which I, for one, keep turned off at all times ;-)

Hi, Ted;

Yes the use of a tone curve in-camera would not change the RAW file at all; but if a tone curve is used, and this changes the display JPEG and thus the histogram the camera displays for an image; and if a photographer decides to alter their exposure based upon a test photograph, then in a roundabout way we can say that the use of the tone curve in the first test capture causes the photographer to actually change the nature of the RAW information captured in the second image by the very fact that they have changed the exposure parameters based upon the tone curved histogram presented by the first test image.

Apart from that kind of scenario, no I am not saying that the use of a tone curve permanently alters the RAW file it is applied to in-camera because I don't think that it does.

[Mike Buckley]

I am not saying that the use of a tone curve permanently alters the RAW file it is applied to in-camera because I don't think that it does.

The proof of that is that every image has a tone curve applied by the camera. It may be the default tone curve decided upon by the manufacturer, an optional, preset tone curve provided by the manufacturer for the user to select, or a custom tone curve provided by the user. In every case, the tone curve can later be altered while post-processing the image.

[John Morton]

In researching this issue on the Internet, I've come across people who state that the reason RAW files have extra latitude beyond the histogram displayed for any file is precisely because the histogram is generated from a JPEG and therefore can not encompass and display all the information available in the RAW file. Thus when we say "a RAW file has an extra two stops of exposure that can be used" what we really mean is "the JPEG histogram we are using leaves out two stops of information contained in the RAW file." Or so I've been 'told' (?).

My own personal interest in this stems from the camera I am using - a Nikon D700. This camera has fantastic low light capabilities (less amazing now, as the technology continues to advance) that are a exactly what I need when I do work such as concert photography. However, as an old hand with black and white photography, I also prefer having blacks in my images that are actually nice, deep, rich blacks. That's one thing I really love about my film scanner, a Minolta DiMAGE Scan Elite 5400: the 16X multi-scan function produces beautifully deep and noiseless blacks.

The D700 is so good at capturing low light information that he blacks are not anywhere as deep as I prefer them to be! And at the other end of the scale, I often have white areas in my photographs that are clipped (because I tend toward photographing scenes with tonal extremes). So if I can custom design a tone curve which will let me nail exposures with higher contrast in the blacks and lower contrast in the whites, I'll be really happy.

In going over my images for inclusion in the mini competitions, I came to realize that I have been really depending upon HDR to meet my basic imaging requirements; but that what I really need to be doing right now is concentrating upon single frame captures. I think that starting to work with custom designed tone curves might help me to move toward the kind of basic image structure I have been looking for, through using HDR, in my single image captures so I definitely need to put some time into investigating this whole matter of customized tone curves.

[Brownbear]

I wonder using a low contrast tone curve will eliminate chromatic aberration when shooting birds in flight in a high contrast situations?

I use a Nikon D80 and when trying to shoot birds in flight, I almost always have this problem.

[xpatUSA]

Mike,

Just out of interest, I looked at the EXIF, and all the other, metadata in a Nikon D50 NEF file.

There were no tone curves in the metadata but it did have a pointer to a "JpgFromRawStart" address and also a tag "JpgFromRawLength" which had a value of 131072. So any tone curve is in the embedded JPEG "subfile", I reckon.

John,

Have you looked at the DNG profile editor where you can create camera profiles to use with ACR?

I keep looking at the D700 . . can't really justify it for watch pics though!

[John Morton]

Hi, Ted;

Well I must confess to have closely considered a Sigma dSLR because I do like the idea of the Foveon sensor!

However, I do not use Adobe Camera Raw as my converter, favoring Nikon Capture NX2 instead. Nor do I use Adobe's DNG format, either; I just leave my Nikon RAW files in the .NEF format.

I did briefly use Photoshop and Adobe's Camera RAW converter, but I noticed once while shooting a Santa Claus parade that the Adobe converter did a very poor job in the gradated shades of shadowed red on Santa's suit; whereas NX2 rendered the reds smoothly, without banding and without blotching. Not knowing what eles Adobe might no be getting right with .NEF files, I haven't used ACR since.

I do however have Nikon's ViewNX, Capture NX2, and Camera Control Pro 2 - each of which are capable of producing and editing tone curves that can be loaded in those programs or uploaded directly into my D700; so I am definitely going to start playing around with that.

In another thread that concerned a wide histogram plug-in for Photoshop, there is such a plug-in listed and I already have this available to me. This particular plug-in, when applied to a photograph, will actually export a list of every pixel value in the photograph; so I am going to get a bunch of photos together that have the tonal range I am looking for, print out the histogram results for each, and see if I can define what would be a particularly good tone curve for my individual photographic preferences.

[Mike Buckley]

There were no tone curves in the metadata but it did have a pointer to a "JpgFromRawStart" address and also a tag "JpgFromRawLength" which had a value of 131072. So any tone curve is in the embedded JPEG "subfile", I reckon.

That's interesting, Ted. I don't know what those metadata fields mean and I actually had not considered that a description of a tone curve might be included in the metadata. As an example, when I use anything other than an Auto White Balance in my two Nikon cameras, the white balance is reported in the metadata only as "manual" even when I use a Nikon preset white balance such as "Cloudy."

[John Morton]

I wonder using a low contrast tone curve will eliminate chromatic aberration when shooting birds in flight in a high contrast situations?

I use a Nikon D80 and when trying to shoot birds in flight, I almost always have this problem.

Hi, Christina;

No, I don' think that this would help that issue because chromatic aberration has to do with the way that different wavelengths of light focus at different places; so all of that is going to be taking place with the lens and, to a certain extent, with the sensor. Tone curves are only going to affect information already captured by the sensor, so at that point the chromatic aberration has already occurred.

[Brownbear]

Thank you John.