Video Color Correction

 

 

Just what is "color balance" you ask ?

In short:  the blacks are pure black - The whites; pure white......  & skin tones appear normal without a distracting abnormal hue (like purple skin for example)..  Put another way: the color the camera "saw" & what the camcorder recorded, should accurately represent the color our eyes & brains "saw" at the time of the taping. However, what a color camera "see's" and what our visual cortex "see's", is often (or more accurately:)  not the same.
 

Our brains perform color balance corrections automatically, dynamically, & over a wide dynamic range -  and with far greater precision than anything man has yet created. Thus for our human engineering creations, we must rely on something far more basic, such as Chip Charts and Skin Tone Reference Charts.  Compared to what our brains come pre-wired to do so easily & naturally,  camera engineers must rely  the most simplistic of calibration reference schemes.

 

Porta-Pattern Log Chart

The ability to gauge proper skin tone is subjective.  Obviously, someone who is "color blind" could never properly align a color camera or make chroma adjustments so that what goes out over the air looks  even remotely somewhat normal.

To better quantitatively measure the color balance, a monochrome "Chip Chart" was used. Since it was a monochrome chart ranging from black to white, any deviation from a balanced Red, Green or Blue channel output level at any point on the individual response curves, for each chroma channel, could easy be measured and adjusted on just an oscilloscope.

When properly white balanced, the logarithmic chip chart should
appear as black & white with no color tones in the shadows, midranges
or highlights. This test chart is used to verify proper setup of the
 white balance calibration as well as equal response from the RGB
channels across the full dynamic range of the camera.
  Porta-Pattern Corp

 

 

 

 

BBC Test Card F

Though never popular in the US, Color Test Card F is still used by the BBC and
over 30 nations around the globe. It was developed by George Hersee of the BBC.
Since virtually every broadcast engineering dept used this test card as a standard color reference in PAL countries over many decades, his Daughter Carole Hersee is without question, the most televised gal in the UK ! .........  

....and all the time you thought it would be someone far more popular & recognizable ! - - - Nope: Not so !)

This would make for a GREAT trivia question !

Anyways, Test Card F was unique, in that  RGBYMCBW colors could be set on a vector scope to "land" the colors in the vector "boxes" and yet provide a skin tone reference as well as basic resolution and linearity.

For most engineers, the skin tone of how Carole appeared along with the overall background, was the final deciding factor in achieving the best color purity of what went out over the airways .

BBC

 

 

BBC 61P Flesh Tone Reference

BBC Chart 61P - Flesh Tone Reference

For the United States and other NTSC countries, Porta-Pattern published  BBC Test Card 61P - the Flesh Tone Reference which  became the defacto standard by nearly all NTSC broadcasters.

This gal image is no doubt the one that has been in front of TV cameras in NTSC countries than anyone else's.

No network or individual station in the US I'm aware of, didn't use Porta-Pattern's BBC Flesh Tone Reference 61P in conjunction with a chip chart..

Thus for US NTSC broadcasters, skin tone itself was the final "acid test" for determining proper color balance.  Yet whether Test Card F or 61P, accurate reproduction of skin tone was always the final acid test... 

The reason ? 

Flesh tones are the most difficult  for any camera to accurately reproduce....  and any equipment in "the chain" so to speak, can introduce it's own chroma aberrations. More importantly; It is also often the sole reference by which all end viewers judge the color balance.

Thus all 3 charts became the world standards for the calibration of broadcast color.

 

 

The Video Color Corrector

 

So why you ask,  all the need for chip charts and skin tone references  ?Color Vectorscope Displaying NTSC SMPTE Color Bars

Well, the human eye and more importantly; the way our brains process visual information is nothing short of miraculous.  No film, mainframe computer or video camera - no matter what the cost or level of sophistication, can yet to this day (or in the foreseeable future for that matter) come even remotely close to the combination of sensitivity, resolution, wide dynamic range and the ability of our brains to automatically "white balance" any scene with incredible accuracy and speed ...           From the ambient light and resulting color temperatures produced from a brilliant red sunset, to the soft low light illumination of a full moon, our brains dynamically process the information in real time and always get the colors right......  a virtually PERFECT white balance EVERY time !  

We walk from a brilliantly lit room to the outside moonlight, and without even being aware, our brains have made dramatic color balance corrections without us even being aware of such a radical color temperature swing !  We don't even have to think of it !

Alas, such is not the case with our camera creations...  Not even Close !   - Even with even today's most sophisticated & expensive video cameras.....    I dare anyone to name a single camera that can accurately capture the color of a full moon reflecting off the water or thru a forest canopy each and every time (or almost ever for that matter)......     Yet our brains do it on a repeatable accurate basis - each and every time.... The "software" routine that performs this amazing feat, for each of us, has been hard wire encoded into each of our DNA.....

Cameras be they film or video, cannot hold a candle to how our brains process visual information....  Our best engineers are yet no match for this true engineering feat.... Our best and brightest engineers are simply far outclassed...  Thus we must resort to "Romper Room" engineering tactics to crudely get the colors even basically correct in the case of video cameras.

To do so; early color cameras had a manual white balance procedure to be followed to insure correct color balance of the scene. Often times, the procedure was skipped, or changing lighting conditions occurred before a new white balance was made. Later cameras had auto white balance circuitry which worked ok....  most, but not all of the time. Whether  manual or auto, attempting to white balance on a scene with no true white in frame, would result in a flawed color balance.  Other possible causes are faulty camera setup, excessively hot or cold operating temperatures resulting in unpredictable camera operation, insufficient lighting, low battery or improper filter selected.  Three tube cameras needed to be constantly calibrated using a grey scale logarithmic chip chart to ensure equal dynamic response from each red, green or blue tube channel, lest a color imbalance ensue.

CCD cameras, though immune to the varying responses of analog tube cameras such as the Saticon and Plumbicon's; CCD's are not immune to the camera operator failing to properly white balance when the illuminating color temperature of the scene has changed, and are thus subject to introducing severe color imbalances when the "auto" circuitry cannot properly identify what should be identified as "white".....  Also, inexpensive consumer cameras often have an "auto" white balance that is effective in most circumstances, but sadly cannot adapt to all situations. Unfortunately, many consumer cameras have no manual white balance over-ride, and are thus subject to introducing severe color imbalances when the "auto" circuitry cannot properly identify what should be a true white object in the scene.

Fluorescent lighting often results in a green tint, as fluorescent tubes put out a lot of "green". Often an orange tint will result from aged film if the camera is used in a telecine (film to video) application. Both are common problems .

Other color imbalances are often introduced by the chroma processing circuits of any equipment in the signal path.  Feeding a calibrated set of color bars and then using a vectorscope, allows accurate measurement and setup. All professional edit master tapes will (or at least should have) a calibrated set of color bars and audio tone at the start of the tape.

Whatever the cause, the video in need of color correction could exhibit an overall hue of almost any color.  It also may be exhibited only in the whites or sometimes only in the shadows or mid-tones, depending upon the nature or cause of the imbalance. Objects that should be pure white may have a false color tint in the highlights, or the items that are shades of grey or black may have a sickening color hue in the shadows. Color correction will offset the errors, thus re-balancing the colors.  Color correction has saved many a production from having to be re-shot over the years.

At best, re-shoots are an expensive if not embarrassing endeavor; and at worst, the footage was a one time opportunity that can't be repeated. Thus color correctors perhaps have saved countless lives of producer's seriously contemplating suicide.

The color corrector is a valuable tool in the restoration of vintage video, where color imbalances are common. 

Though dramatic improvements in color balance are often possible, there are limits.....     Color balancing as it's name implies, involves "trading off" two colors for the benefit of the one that is weak .... or boosting the weak R, G or B chroma channel itself so as to achieve a "balance".  In cases for example where the master tape is so far out of color balance there is NO red channel signal to speak of, neither is there is enough RED information remaining to boost....  nor can GREEN or BLUE be reduced enough to re-balance it.....     In all but the extreme cases just mentioned, color rebalancing can usually make dramatic improvements !

In the case of transferring film to video tape or DVD, a color corrector can easily re-balance most color imbalances due to either film degradation; or as often the case, a sloppy transfer having been made where no color correction or proper white balance was employed. Most of the inexpensive film to video transfer services do not use broadcast quality cameras and thus color imbalances are commonplace. (If it's a "Super Deal", you can almost automatically assume a broadcast quality camera was not used for the transfer).... Though we do not currently offer film to video transfers, we can color correct the video of what you did have transferred, often making an dramatic improvement.

 

Last Modified: June 16, 2012

Since 1978 - Copyright 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011, 2012, 2013 Video Interchange

Page copy protected against web site content infringement by Copyscape

Terms and Conditions on using this site


Video Interchange
92 George Luce Rd
Waldoboro, ME - Maine   04572

207-832-5064   -   mail@videointerchange.com

Northeast Region - New England

Home WebCam Weather Site Map Site Search Price List Film Transfers Video History Vintage Video Video Tape Video Recovery Damaged Tape About Dvd's Color Correction Pal Conversions Obsolete Formats Vintage Audio Audio Formats Audio History Wire Recorder Audio Restoration Vintage 78s Audio Clips Audio Q&A Printing/Scanning Tech Tips Reference Tape Conversion FAQ Links Wanted For Sale Cats PDP-9 Classmates Synchronicity Coal Stoves Energy Losses Waldoboro About Us Site Security Payments Privacy Shipping Contact Us Contact Form