View Full Version : x.v. Color. Great color space, or Greatest color space?
Jacob Mason March 29th, 2007, 08:14 PM I've heard some interesting things about the new x.v. color space introduced by Sony and Mitsubishi.
So far, Sony has been producing camcorders that can capture in this color space and naturally they are also making their HDTVs to display the x.v. color space for proper viewing of the x.v. material.
The problem I see with this is retaining that extra color through any given compression scheme, be it Sony's HDV or AVCHD. Wouldn't the footage have to be fully uncompressed video in order for this new color standard to have any significant impact on the image quality and then it would still require the x.v. capable monitor/TV for the images to be displayed accordingly, correct?
I'd like to know what some of you think about this new color standard.
Will it be adapted by others, and does anybody know for certain if the x.v. color space is only beneficial when shooting uncompressed?
Art Willig March 31st, 2007, 06:15 PM I'm shooting with it on in my HC5 and will adjust in post if need be. This way, I'm gathering the most data regardless.
John Miller March 31st, 2007, 09:00 PM According to Wikipedia,
"xvYCC or Extended-gamut YCC is a color space used in video electronics, supporting 1.8 times as many colors as the sRGB color space[1]. An xvYCC display, along with Deep Color to reduce posterization, can display the full range of colors viewable by the human eye — these two technologies form the color specification of HDMI 1.3. xvYCC was specified by the IEC in October 2005 and published in January 2006 as IEC 61966-2-4. The wider spectrum of LED backlighting replacing cold cathodes has enabled this extension of the LCD display color gamut."
Hmmm....the full range of colors viewable by the human eye. If that's true, that will be quite something. However, I'll believe it when I see it - particularly the very deep indigoes that I have seen SCUBA diving in the caves of Gozo. Or indeed the very brilliant colors of various underwater life forms. Ultraviolet radiation and consequential fluorescence add a lot to the experience.
Jacob Mason March 31st, 2007, 11:24 PM From what I've been reading, the new x.v. color space should not be confused with "deep color".
Deep color increases its color accuracy from 8 to 16 bit but still stays within the RGB or YCbCr boundaries.
The new x.v. or xvYCC color space goes beyond those boundaries to meet or even exceed what the human eye can see.
Either way, they're both very interesting.
And I believe that since it's based primarily off of the HDMI 1.3 standard, then it would only have an impact on the image quality when shooting uncompressed, and viewing on a monitor that was compatible.
Mikko Lopponen April 5th, 2007, 04:06 AM First they need to build lcd's that actually have good color reproduction with rgb. No use in xvcc with regular flourescent backlight lcd's.
Dave Blackhurst April 5th, 2007, 10:12 AM Doing some test shooting with an HC7, with x.v.color on, it looks pretty good to me when downloaded into Vegas7. Don't see it on the camera LCD/VF IMO, but the colors looked better overall when compared to my HC1 and 3. It is subtle, but I do think you're getting a bit more to "work with" - it will be interesting to see if it develops or goes the way of BETA...
Paulo Teixeira April 5th, 2007, 01:21 PM One thing I will say about this feature is that it’s a gamble. Meaning if you keep it on you will be adding up to 1.8 times the color but then again, the picture quality may actually be worse if you hook it up to a standard HD TV. So basically they advise you to keep it off if you don’t have an HDMI 1.3 complaint TV.
From what I researched so far, you will see HDMI 1.3 complaint TVs from Sony, Samsung, JVC, Sharp, Panasonic and Toshiba this year from around June to September for most models depending on the manufacturer. I don’t know about Pioneer but it doesn’t matter because as is, their top plasma screen will have a 1 million to 1 contrast ratio.
Besides Sony, JVC was also going to use the new color standard on the HD7 but decided to use it on future HD camcorders. I’m thinking that maybe the reason they didn’t use it is because it would have cost a lot more money to implement. This is why the HC7 cost a little bit more than the HV20. To most people the HD7 is too expansive as is.
Like everybody else, I wish there were engineers on this board that can answer any question concerning xvYCC that people may have.
Glenn Chan April 6th, 2007, 01:03 AM From my understanding of this...
1- Displayable colors will ALWAYS be limited by the color gamut of the display device. Most display devices cannot produce extremely saturated colors. No commercial device can produce individual wavelengths of light, which is the most saturated/pure you can get.
2- xvYCC can define imaginary colors... with negative values, it can specify colors even more saturated than a single wavelength of light. And since displays can't produce imaginary colors, you won't see them.
I suppose it "[exceeds] what the human eye can see"... but that doesn't really mean anything in practice.
3- For some info on xvYCC, check out the Rec. 1361? transfer function described in Charles Poynton's book. And you can take advantage of the ITU's 3 free downloads and get Rec. 1361.
4- When you go wide gamut (with the xvYCC scheme or any other scheme), you are now describing a much larger # of colors. Or, if you have the same bit depth, you have less precision in the colors you specify.
To maintain the same precision and to describe a much larger # of colors, you need to increase the bit depth. So you need to jump from 8-bit to 10-bit. (AFAIK, deep color refers to 10-bit support; and 12-bit.)
5- Wouldn't the footage have to be fully uncompressed video in order for this new color standard to have any significant impact on the image quality and then it would still require the x.v. capable monitor/TV for the images to be displayed accordingly, correct?
Your footage doesn't have to be uncompressed. You can compress it 10-bit + whatever codec.
6- For the display to actually display wide gamut colors, it needs to be a wide gamut display and it needs to support HDMI 1.3. Normal gamut displays (with HDMI 1.3) will simply clip colors that exceed its range.
7- I'm not sure if you can record xvYCC colors onto tape. The recorded signal will be different than Rec. 709 and Rec. 601 video; it will decode but the levels will need correction (unless the NLE supports xvYCC and gets the levels right). It might be that xvYCC is only available on HDMI out, where the HDMI 1.3 compliant display knows how to decode the signal correctly.
Charles Hurley April 10th, 2007, 11:24 PM The xv colorspace allows for more colors not more saturated colors. Most dispays are already capable of saturating a signal far past the point of reality. I'm sure this is a simple progression away from rgb/yuv and analog in general especially as companies begin to phase out 3 ccd systems (read RGB/YUV) in favor of CMOS.
Glenn Chan April 12th, 2007, 10:15 AM It does allow for more saturated colors to be specified (i.e. very pure colors like single wavelengths, the light from a laser, etc.). It is up to the monitor whether it can display those colors or not.
Charles Hurley April 14th, 2007, 01:54 AM Fair enough. Then what is the difference between 24 bit color and XV?
Glenn Chan May 21st, 2007, 06:49 PM Charles: Sorry, somehow I missed your point (I tend to skim the forums and not read everything).
"24 bit" presumably refers to bit depth... i.e. 8 bits for each channel.
"Deep Color" in HDMI 1.3 refers to bit depths higher than 8-bits... i.e. 10 and 12and 16 bits/channel.
xvYCC requires higher bit depths. Since it describes a bigger range of colors, the spacing between each color is bigger / you have less precision. To make up for that, you need to move to a higher bit depth.
|
|