Sony 930, 950 Knee aperture, Knee correction, Knee point and knee slope, Level depend, Limiter

9	compared in numbers, it does not appear in the
	specifications of the brochure. However, it is very
	important for evaluating picture quality intuitively.

Usually, it is only called “DTL (detail)”.

IT/FIT

In an IT type CCD, beside the pixels there are columns of cells dedicated for vertical transfer. The electronic signals are transferred from the pixels to the columns of cells during vertical blanking. Finally, the signals reach a row of cells driven by a horizontal block for horizontal transfer, from where the signals are output as video signals. Although there is a cover to protect the cells from light leakage, sometimes the leakage still happens under a strong light, which causes smear.

To avoid this phenomenon, in an FIT type CCD, there is additional storage added to the CCD which enables the vertical transfer to perform much more quickly so to minimize the chance of light leakage even under a strong light. However, this additional storage increases the cost.

Recently, there has been great advances in IT CCD technology and the gap between IT and FIT has been dramatically reduced.

Knee aperture

Contrasts always become worse when compressing in the direction of the bright level during knee correction. Knee aperture is used to improve this point to emphasize the contour to which the knee correction applies.

Knee correction

The dynamic range of a natural scene is very large, but that of a video standard is limited. When we use the iris to adjust for the subject photographically, the bright portion of the picture will become almost completely white. However, when we focus on the bright portion, the subject for photography will become very dark. Knee correction is a function to solve this problem by compressing the level of the bright area to meet that in the video standard so as to make the dynamic range seem larger.

Knee point and knee slope

Knee point is the starting point of the knee correction process, and the knee slope is its compression rate. Usually, knee slope is the peak value under an environment of 200% incident light.

Knee saturation

Because knee correction performs independently in R/ G/B channels in a conventional knee circuit, the R/G/B percentage of the portion to which the knee correction applys will change, and thus the color phase and saturation change. This is called “white clip” or “wash out.” Knee saturation is used to balance the coloring to avoid this phenomenon.

Level depend

Just like Crisp reduces noise, level depend is a circuit that decreases the DTL value in dark areas.

Limiter

A limiter is a circuit used to constrain the DTL peak value that may ruin the picture quality when the luminance difference is too big that it will generate too large of a DTL.

Linear matrix

The linear matrix is actually a circuit used to make the color reproduction look better, although some textbooks mention that it is used to correct the difference between camera chromaticity and that of

HDC-900/950/930 Series Product Information Manual

the signal standard. It seems that the linear matrix should be more important to emphasize the visual appeal of the picture apparently rather than just be theoretically determined by some coefficients.

Mix DTL/NAM DTL

Recently, most of the cameras create V DTL from all three channels of R/G/B. Depending on how to apply these three values of V DTL to the add operation, there are two types of mix: the Mix DTL – a simple sum of the values and the NAM DTL – a non-additive mix of the values. Both of them have pros and cons. The Mix DTL is not always better than “out of green”, while the NAM DTL helps make a picture sharper but sometimes it makes the S/N become worse or adds the DTL opposite to the luminance. The NAM DTL does not apply to H DTL, because it will increase aliasing.

Multi matrix

Multi matrix is used to easily implement color reproduction that complies with color matching and visual appeal between different cameras by adjusting the matrix coefficient to match the color phase. For instance, it is possible to change the color phase and saturation of red color only. In most of Sony’s cameras, it works with 16 axes for color phase on a color plane like a vector, each of which has its own matrix coefficient to perform the linearity corrections between each axis. Normally, each axis needs six matrix coefficients (that means 96 matrix coefficients are needed). However, the number of coefficients can be reduced to two for each axis in most of Sony’s cameras using some dedicated software - color phase and saturation.

On-chip-lens

To obtain a better quality picture, -the more pixels the better for a CCD. However, increasing the number of pixels means decreasing the pixel size. When the pixel size becomes smaller, less incident light can be gathered, which degrades the sensitivity. The on-chip- lens is a micro lens put on each pixel to help increase light gathering so to improve the sensitivity.

Optical filter

There are two types of optical filters, Neutral Density (ND) filter and Color Temperature Conversion (CC) filter. The former is like a pair of sunglasses for the camera to avoid the iris to be stopped down too much under a very strong light, while the latter works like a pair of colored glasses to compensate for the color temperature beyond the white balance compensation range. Both of them are integrated on a built-in filter disk, can be manually or electronically controlled and can be remotely controlled using the remote controller for the electronic filter.

Optical low pass filter

Shooting with a CCD camera means sampling a picture via the use of the pixels on a CCD. An optical low pass filter is made of optical crystal to cut off the unnecessary high frequency that might cause beat interference caused by the relation between the number of pixels and the pattern of the shooting subject during the sampling.

Out of green

The older version cameras create V DTL from the G channel and add it to each channel to carry out image enhancement. This method helps use a less number of delay lines, but does not perform well in some cases such as a deep red flower picture. Some cameras create V DTL from R/G channels.

Chapter 9 Glossary – Terms and Definitions 174