What The H*** Is Gamma? And How Do I Use It?

[rfsIII]

Could anyone explain what "gamma" means in terms that a regular photographer can understand?

I ask because in Nikon Capture NX-D there is a control for gamma. Moving the points definitely changes the look of the photo but I couldn't describe to you how it changes.

Is there a way to explain gamma in terms of what adjusting the curve does to a photo and where and when it might be preferable versus just a plain old curves adjustment.

The CambridgeInColour definition says gamma defines the relationship between a pixel's numerical value and its actual luminance and talks about the difference between the human eye and the sensor, but that doesn't provide any insight into how to make use of the controls.

And if possible does anyone know why the scale goes from 6.0 to 0.05.

Thanks!


The Gamma panel from Nikon Capture NXD software.

 

[Lan]

The CambridgeInColour definition says gamma defines the relationship between a pixel's numerical value and its actual luminance and talks about the difference between the human eye and the sensor, but that doesn't provide any insight into how to make use of the controls.

^ Techically, that

It's difficult to put it into simple terms because it's not a simple concept.

It's easiest to think of it as a curve that's applied to a digital image in order to display it. In most cases the effect is not dissimilar to moving the midpoint on a standard Curves adjustment.

 

[GCam]

In film, every film type has a characteristic curve that determines how each grain will respond to exposure, depending on how much it receives, and I guess you could say sections of the film. When you change the gamma, you change the characteristic curve, or the slope of the film, o rof the particular file you are editing. Changing the gamma will change the contrast of the file to more contrast or less, basically. I'm not sure it's even a necessary step in editing since you can do the same thing, but this does add more contrast or less to every section of the film or file. gc

 

[Ernie Misner]

Not sure if this is technically correct, but when I use it in CNX-2, it increases or decreases brightness of the image without affecting color contrast. Much like using curves in PS and changing the blend mode to luminosity.

Try darkening or increasing contrast a lot using curves and watch the colors get dark and crunchy, HDR like. (some might like that). Then make the same change using the gamma curve and the colors should remain more neutral.

Correction, they are calling that the LCH (master lightness) curve in Capture NX-2 but it looks the same as what you have in NX-D.

I think that can work the other way around also. In PS for example if changing the saturation a lot, the luminosity (brightness) will be affected. But if you change the blending mode to luminosity while doing that, the luminosity remains more neutral.

--
Ernie Misner
http://www.flickr.com/photos/erniemisner/
"I bet we'll find hints of extraterrestrial life before 2030" - Jamie Shreeve, Popular Science

 

[walking dude]

Could anyone explain what "gamma" means in terms that a regular photographer can understand?

Midpoint.

I ask because in Nikon Capture NX-D there is a control for gamma. Moving the points definitely changes the look of the photo but I couldn't describe to you how it changes.

Moving the slider to the left brightens the overall image. Moving the slider to the right darkens the overall image.

Is there a way to explain gamma in terms of what adjusting the curve does to a photo and where and when it might be preferable versus just a plain old curves adjustment.

Using your curves image below, when you place a point in the very center (midpoint) and drag down as in your example, you are darkening the midtones, same result as moving the midpoint (gamma) slider to the right. Push that point upwards and now the midtones are getting brighter, same as moving the midpoint (gamma) slider to the left.

In short, place a point in the center of the curve or use the slider, whichever works for you. That is adjusting the gamma. Put additional points on the curve and adjust them up or down and now your just making a plain old curves adjustment. ;-)

The CambridgeInColour definition says gamma defines the relationship between a pixel's numerical value and its actual luminance and talks about the difference between the human eye and the sensor, but that doesn't provide any insight into how to make use of the controls.

And if possible does anyone know why the scale goes from 6.0 to 0.05.

It's on my 'things I don't need to know' list. :-D

Thanks!


The Gamma panel from Nikon Capture NXD software.

 

[JohnWheeler]

I am not sure this link will help yet it tells you all about gamma (and the various ways that is used): https://www.cambridgeincolour.com/tutorials/gamma-correction.htm

You will know all about gamma yet I am not sure that answers your question. With the above information in the link, it will at least enable you to ask your questions more specifically.

Hope this helps (yet not sure it does)

 

[Pixel Poetry]

All I know is this is what happens when you increase Gamma.







--
PX

 

[hurtsme]

I love it!!!!

 

[Sailor Blue]

The human eye and film both respond in a similar way to varying brightness in a scene. The shadow details and highlight details aren't distinguished very well from each other but the middle tones are easily distinguished from each other.

When you graph the response of the human eye or film you get an "S" shaped curve.

The shadows aren't distinguished well and tend to blend together so that part of the graph has a very low slope. That area of the curve is called the toe of the curve.

The middle brightness tones are easily distinguished from each other and that region of the graph takes on an almost linear shape. The gamma is the slope of this area in the type of graph used.

The highlights aren't and tend to blend together so that part of the graph has a very low slope. That area of the curve is called the shoulder of the curve.

Here is a graph of the density for B&W film vs the base 2 root of the amount of light. The brighter the light the more exposure the film got so the darker the film, i.e. the higher the density, after developing.



Digital camera sensors have a completely flat response to the amount of light hitting the sensor. There is no toe or shoulder and the slope is 1, i.e. double the amount of light and the response doubles.

To give you an image that makes sense to our human vision your camera or RAW converter takes this flat response and converts it into a curve much like the above one.

There are limits at which we stop being able to distinguish tones from each other so extending the graph beyond those points even if the data is recorded isn't necessary.

Using the curve shown in your post allows you to change the gamma for specific areas of the image, thereby changing the contrast in those areas, which means you can change the amount of tone separation seen in the image for those areas.

For instance, if you increase the slope in the toe of the curve you can see more separation between the shadow details. This is what we call "opening up the shadows". Of course you could decrease the slope to purposely block up the shadows.

Alternatively if you reduce the slope of the highlights then you can see more separation between the highlight details, i.e. you are "pulling back the highlights". Increasing the slope would move the White Point in toward the mid-tones.

Increasing or decreasing the slope in the middle changes the contrast of the mid-tones.

--
Living and loving it in Pattaya, Thailand. Canon 5DS R & 7D - See the gear list for the rest.

 

[rfsIII]

I can buy that....

The CambridgeInColour definition says gamma defines the relationship between a pixel's numerical value and its actual luminance and talks about the difference between the human eye and the sensor, but that doesn't provide any insight into how to make use of the controls.

^ Techically, that

It's difficult to put it into simple terms because it's not a simple concept.

It's easiest to think of it as a curve that's applied to a digital image in order to display it. In most cases the effect is not dissimilar to moving the midpoint on a standard Curves adjustment.

 

[rfsIII]

Thanks

I am not sure this link will help yet it tells you all about gamma (and the various ways that is used): https://www.cambridgeincolour.com/tutorials/gamma-correction.htm

You will know all about gamma yet I am not sure that answers your question. With the above information in the link, it will at least enable you to ask your questions more specifically.

Hope this helps (yet not sure it does)

--
John Wheeler

 

[rfsIII]

Can’t argue with that.

All I know is this is what happens when you increase Gamma.



--
PX

 

[rfsIII]

Allows you to fine tune the separation of values in various parts of the image. Great explanation.

The human eye and film both respond in a similar way to varying brightness in a scene. The shadow details and highlight details aren't distinguished very well from each other but the middle tones are easily distinguished from each other.

When you graph the response of the human eye or film you get an "S" shaped curve.

The shadows aren't distinguished well and tend to blend together so that part of the graph has a very low slope. That area of the curve is called the toe of the curve.

The middle brightness tones are easily distinguished from each other and that region of the graph takes on an almost linear shape. The gamma is the slope of this area in the type of graph used.

The highlights aren't and tend to blend together so that part of the graph has a very low slope. That area of the curve is called the shoulder of the curve.

Here is a graph of the density for B&W film vs the base 2 root of the amount of light. The brighter the light the more exposure the film got so the darker the film, i.e. the higher the density, after developing.



Digital camera sensors have a completely flat response to the amount of light hitting the sensor. There is no toe or shoulder and the slope is 1, i.e. double the amount of light and the response doubles.

To give you an image that makes sense to our human vision your camera or RAW converter takes this flat response and converts it into a curve much like the above one.

There are limits at which we stop being able to distinguish tones from each other so extending the graph beyond those points even if the data is recorded isn't necessary.

Using the curve shown in your post allows you to change the gamma for specific areas of the image, thereby changing the contrast in those areas, which means you can change the amount of tone separation seen in the image for those areas.

For instance, if you increase the slope in the toe of the curve you can see more separation between the shadow details. This is what we call "opening up the shadows". Of course you could decrease the slope to purposely block up the shadows.

Alternatively if you reduce the slope of the highlights then you can see more separation between the highlight details, i.e. you are "pulling back the highlights". Increasing the slope would move the White Point in toward the mid-tones.

Increasing or decreasing the slope in the middle changes the contrast of the mid-tones.

--
Living and loving it in Pattaya, Thailand. Canon 5DS R & 7D - See the gear list for the rest.

 

[RHWeber]

In Capture One the curve tool under base characteristics adjusts the gamma for RAW files. The program offers five choices including auto. If you look at the histogram as you select different curves you can see how it is changing the Gamma curve. C1 bases the curves in the profile for the camera that took the picture. However you can manually alter the profile if you like.

 

[Sailor Blue]

In Capture One the curve tool under base characteristics adjusts the gamma for RAW files. The program offers five choices including auto. If you look at the histogram as you select different curves you can see how it is changing the Gamma curve. C1 bases the curves in the profile for the camera that took the picture. However you can manually alter the profile if you like.

All image files, including raw files, are just a file of numbers that are derived by converting the analog signals collected by the camera's pixels into digital numbers. Think of this file as a big spreadsheet table with RGBG numbers for each pixel in raw files, or simply RGB numbers for other formats.

The image data in the file must be interpreted to produce a viewable/printable image. This is true of any format, raw, JPG, TIFF, PSD, etc.

Raw data numbers must be normalized to adjust for the different responses of the pixels to red, blue, and green light and for the different responses at different color temperatures. Finally the RGBG data must be demosaiced to produce an image with only RGB pixels.

When any raw converter produces a viewable/editable image from the raw data it produces an image with a curved foot for the shadows, a linear mid-tones region, and a curved shoulder for the highlights. The model used for this "S" shape conversion differs from company to company, producing different looking results.

This differing "S" conversion, plus different color choices for the conversion, explain why some photographers prefer the initial image from one raw converter/editor over another. In reality it should be how easy it is to reach a final version that meets your inner artist's desires that is important, not the initial viewable version.

For JPGs the raw converter is built into the camera or it can be part of a raw image converter/editor.

Capture One may say that they adjust the gamma for RAW files but in fact all that they are doing is changing the conversion model for the viewed image, which is what happens when you change the Contrast slider in Lightroom or any other raw file converter/editor. You can also change the gamma of the viewed image by using Curves, Exposure, Highlights, etc. in any raw file converter/editor.

 

[CraigVMN]

All I know is this is what happens when you increase Gamma.

Damn, beat me to it

 

[Mark Hollister]

Adobe should add a AHB (Anti-Hulk Button), so that when enabled, Photoshop refuses to do anything that will turn you into a Hulk.

 

[Chaplain Mark]

All I know is this is what happens when you increase Gamma.



--
PX

That looks more like Gampa than Gamma, sorry...

--
Thank You,
Chaplain Mark
-----
'Tis better to have a camera and not need one than to need a camera and not have one.
--------------
In pursuit of photographic excellence.

 

[Gesture]

Not sure about modern ones, but older Pentax DSLRs had a menu choice to "impose that S curve" in camera, something many of us almost intuitively do as one of our first editing steps.