List of all types of Image Noise and Artifacts

This is a list of imaging terms I have been collecting off the web, if anyone can add to it or better help describe it it

would be great. I am using my own definition of Noise and Artifacts here-it is most likely many would not like this

distinction, but I find it usefull in classifying these affects.

 

 

By NOISE I mean any non-image affects that occure between the sensor chip and wrighting of the RAW file.

This includes the cameras circuitry and any algorythens used to make the resulting file(s)-anything that can

negatively affect the fidelity of the recording process from the moment the light is recorded to the moment the digital

negative is produced.

 

By ARTIFACTS I mean any class of non-image changes introduced by the lenses or post-processing of the RAW

file. As the term implies, the effects of these various manipulations of the RAW file can leave visible affect (artifacts)

within the image information throughout the workflow.

 

"Purple Fringing" usually refers to a typical digital camera phenomenon that is caused by the microlenses.

 

"Blooming" is a type of noise where an excess of charge flow over to surrounding pixels, brightening or overexposing

them in the process.

 

"Pixilation Noise" (jaggies/aliasing) is due to insufficient image information/resolution for the required output. This

type of information loss is part of the restrictions of the CMOS chip used to record the image.

 

"Pixilation Artifacts" (jaggies/aliasing) is information loss idue to post-production down sampling.

 

"Chromatic aberration" or "color fringing" is an artifact caused by the camera lens not focusing different wavelengths

of light onto the exact same focal plane.

 

"Maze Artifacts" are moiré patterns caused by the camera's internal image processing to generate "maze" like

patterns. If this is post-processing of the RAW file in-camera, I think it should be consdered an Artifact, if it is done

as part of making the RAW file it should better be described as noise.

 

"Moire Patterns" If a scene contains areas with repetitive detail which exceeds the resolution of the camera (1), a

wavy moiré pattern can appear. This may be a type of noise, but a unique class for sure.

 

"Sharpening Halos" the affects of ﾓDigital resolution" where sharpness is faked by making the edges more contrasty.

 

"Luminance Noise" fluctuations in luminance across different channels-monochromatic grain, similar to film.

 

"Color Noise" or "chromaﾔ noise is not luminance noise, but I do not have a good definition here?

 

"Posterization" is an artifact that occurs when an image's apparent bit depth has been decreased so much that it

has a visual impact. a loss of continuous tone information.

 

"Color fringing" is this really different than "Chromatic aberration"?

 

"Channel noise" is random spikes in long exposure modes.

 

"Barrell distortion" is the distortion of actual pixel information due to the effects of lenses without proper coating.

 

Random noise is characterized by intensity and color fluctuations above and below the actual image intensity. There

will always be some random noise at any exposure length and it is most influenced by ISO speed. The pattern of

random noise changes even if the exposure settings are identical

 

Fixed pattern noise includes what are called "hot pixels," which are defined as such when a pixel's intensity far

surpasses that of the ambient random noise fluctuations. Fixed pattern noise generally appears in very long

exposures and is exacerbated by higher temperatures.

 

Banding noise is highly camera-dependent, and is noise which is introduced by the camera when it reads data from

the digital sensor. Banding noise is most visible at high ISO speeds and in the shadows, or when an image has

been excessively brightened. Banding noise can also increase for certain white balances, depending on camera

model.

 

"JPEG artifacts" (JPEG compression artifacts) Artifacts in digital images are unwelcome and unnatural elements or

distortions.

 

"Interpolation Artifacts" any type of non-image information used to upsample image date.

 

This is all I could find, do you know of more or have better definition to use?

 

And sorry for not spell checking

There's also two other failures types for pixels.

 

stuck pixel: a pixel that always reads out one color regardless of exposure length or ISO and

 

dead pixel: a pixel that never reads out any information (always is black)

 

 

[["Barrell distortion" is the distortion of actual pixel information due to the effects of lenses without proper coating. ]]

 

Barrel distortion doesn't have anything to do with lens coatings. It has to do with lens design.

Thanks Rob... Did you find the Noise/Artifact distinction useful?

I think it's a good idea to have a list like this for reference. With a little more work (perhaps some examples and clarifications, where needed) this would make for a good addition to the Learn section of photo.net

Could Color noise be an example of a stuck pixel as you describe it?

Note that Apochromatic lenses (APO) are designed to minimize Chromatic Aberration. Even in the black and white days, an APO lens could make a picture look sharper because failure to focus different frequencies of light to the same point caused things to look a touch out of focus.

I think it would be good to separate the list into optical and media/film/camera artifacts before it gets any longer.

<p>

To that end, the six basic optical aberrations (some of which you mention - and good catch on distortion, Rob) are:

<p>

1. chromatic aberration - color fringing caused by focal length varying according to color (wavelength) of light;

combated by using optical materials with various refractive indices

<p>

2. spherical aberration - blur of point images caused by focal length varying at different distances off-axis;

that is, the lens doesn't have the same FL at the center and edge of the optics (not the image); perfectly

spherical optical surfaces cause this defect but are easiest to manufacture

<p>

3. coma - radial blurring increasing off-axis; related to spherical aberration

<p>

4. astigmatism - caused by cylindrical asymmetry ("lopsidedness") in the optics (the optics are therefore not

radially symmetric; contrast with spherical aberration)

<p>

5. distortion - focal length (magnification) varies radially across the image (contrast with spherical

aberration)<br>

5.1. barrel distortion - FL increases toward center of image<br>

5.2. pincushion distortion - FL increases toward edge

<p>

6. field curvature - the focal surface is not a plane, leading to best focus in the image center or in a ring

around the center; some specialty cameras (e.g. Schmidt cameras) incorporate curved film or plate holders

What I have found in many of the posts on the web and in many "top" books, is a lack of consistent language over some important imaging issues-this is just a great conversation to have!

 

Now, if your thinking of dividing the class of Artifacts into a sub groups I think it makes good since because the Artifacts list did not seem to fit nicely into one single exhaustive class of its own, but do you think the noise class of affects are clear enough given that it is describing the narrow group of affects relating to imaging ability of the sensor chip to the output of a digital negative. The whole class would describe, well, the range of the sensor's fidelity-if it changes of not, and the quality of the circuitry/software producing the RAW file.

Nice list!

 

Chroma noise is when you have bogus RGB pixels in dark areas, rather then uneven white/gray/black.

Film shows chroma noise, especially expired 800 speed film.

 

Color fringing is a superset of purple fringing. Telephoto lenses often show green or blue on one side

of objects and purple or red on the other. I'm not sure why digital cameras suffer primarily

from purple fringing in the corners, and not other colors.

Hi

 

to perhaps corelate results of a search, some examples on this discussion too

 

http://www.photo.net/bboard/q-and-a-fetch-msg?msg_id=00PC75

 

and some noise examples on my own site

 

http://home.people.net.au/~cjeastwd/digital/inCameraArtifacts.html

Chris Eastwood: I do not think your opening definition of "Noise" is exclusive enough to work with a good classification of imaging issues: quote: "When is noise, not noise? Answer: When its a camera generated processing artifact :-)". Here you could argue for most every event, both software related and hardware that is used to translate light information into digital information (RAW). Here the distinction breaks down easily, with no room to classify by, distinctions, anything except for maybe the lens affects.

 

However, I think you went on to describe the "Digital Negative" VS "JEG in-camera post-processing" correctly. Also your title hints that artifact is a better label here. My definition limits noise to the events between the sensor and RAW- the conversion of light information into the first instance of digital information. In most every field noise is anything that degrades a message between the source of the message to a receiver. In the case of a digital camera, I think the source of the message is the sensor's capture of light information and the receiver is the resulting RAW file. I will have to read your article much closer because I think your working definition between the two is the best.

Also, you said "pixelization is where you can distinguish individual pixels, typically this is from zooming in too much." I think better use should be one that describes a low resolution image due to small sensor size or by down-sampling. Although this too does not set well with me either, because of two reasons: first: by limiting it to the sensor's range it seems to be a type of noise; Secondly, if this is also an effect of down sampling, we have to consider another form of pixilation that is also a form of Artifact.

Here is another Main Grouping scheme:

 

Noise: Any non-image information resulting from the interpertation of light information by the sensor and into digital information contained in the RAW file.

 

Artifacts: Any post-processing affects left over in the image file.

 

Aberrations: all affects on the color and image due to optical issue within the lense or enviormental conditions.

More imaging isues found on B&H Photo

 

Absolute resolution

Image resolution as expressed in horizontal versus vertical pixels. (E.g. 1600x1200 Pixels is the absolute resolution,

and is also expressed as 2.1 Megapixel)

 

Blowout

Blowout is caused by overexposure, which results in a complete loss of highlight details. With the exception of RAW

files captured within 2-stops of the correct exposure, 'blown' highlights are difficult, if not impossible, to correct after

the fact.

 

Dark Current

Pixels collect signal-charges in the absence of light over time, which can vary from pixel to pixel. The result is known

as dark current, or more commonly as noise.

 

Fringing

Fringing, commonly associated with less-expensive lenses, describes the 'bleeding' of color along the edges of

contrasty portions of a digital image. Fringing often shows up as cyan blurring on one side of a contrasty object

complimented by red or magenta blurring on the opposite side of the object.

 

Optical Resolution

The physical resolution at which a device can capture an image. The term is used most frequently in reference to

optical scanners and digital cameras.

 

 

Pincushion Distortion

An optical distortion, common in less expensive lenses, where parallel lines on the horizontal and/or vertical plane

bow inward. Pincushion distortion is the opposite of barrel distortion.

 

 

Red-Eye

Red-eye is the term used to describe the reddened pupils of the eyes that sometimes occurs when photographing

people or pets with an electronic flash. The red color appears when the pupil of the eye is dilated, usually in a low

light environment when the light of the flash strikes the rear portion of the eye and illuminates the blood vessels

located in the rear portion of the eye. Red-eye can often be avoided by placing the flash further than 6" from the

camera lens.

 

Vignetting

Fall-off, or darkening of the edges of a photographic image due to the inability of a lens to distribute light evenly to the

corners of the frame. While correctable with filtration using on-camera center-weighted neutral density filters or

electronically in Photoshop, it is often used as a creative device to direct the eye back to the center of the frame.

Sensor Dust is another

Sensor dust is an unavoidable, but easily controlable, annoyance. This article offers suggestions on minimzing dust build-up in camera, as well as tips to testing for, and remove sensor dust in your camera. Also explored are different methods to digitally eliminate dust from finished images using various photo editing software.