View Single Post
  #12  
Old April 9th, 2018, 09:24 AM
Doug Kerr Doug Kerr is offline
Senior Member
 
Join Date: May 2006
Location: Alamogordo, New Mexico, USA
Posts: 8,565
Default

Hi, Asher,

Quote:
Originally Posted by Asher Kelman View Post
Tell us about Chroma 4:4:4

I am not knowledgable as to these specifications and the benefits!
I do not know the significance of this as a capability of a printer.

That notation (and it works in a very odd way) describes the way in which the chrominance information (which determines the "color", in the sense that civilians would use the term - that is, not embracing luminance) of a pixel is recorded in (among others) JPEG format. The designation "4:4:4" means that the chrominance information is specified for each pixel of the image. In contrast, the designation "4:2:0" means that the chrominance information is specified only once for 4 pixels.

(For our purposes here, we can think of chroma as a synonym for chrominance.)

In that case, the "resolution" of the chrominance information is (in both horizontal and vertical directions) half that that of the luminance information. And the decoder in effect "interpolates" to estimate the chrominance for the pixels for which it is not specified in the file.

This approach is called "Chrominance Subsampling". The object is to reduce the data content of the encoded file (and thus the file size). The scheme exploits the fact that in human vision the resolution is less for chrominance than for luminance.

An intermediate mode is designated "4:2:2". In that, the chrominance information is only specified for every two pixels. In that case, the resolution of the chrominance information in the horizontal direction is half that of the luminance information, but in the vertical direction is the same as the luminance.

Conceptually, the use of the 4:4:4 arrangement, compared to the more common 4:2:2, in a JPEG file could give better color rendition in a delivered display or print.

Note that as I have described it, the notion of chrominance subsampling pertains only to the image as encoded in a JPEG file (and some other formats, I think, as well). It does not pertain to the delivery of the "decoded" image to a display unit ("monitor") in the traditional forms (VGA, for example).

But perhaps it has some implication on the more modern types of interface to the display unit (such as HDMI). And then there might be the issue of whether the display unit can in fact exploit the "more refined" chrominance information of the 4:4:4 form.

Best regards,

Doug
Reply With Quote