![]() |
|
#21
|
|||
|
|||
|
Well I have a Nvidia graphics card so ATI't thing definitely won't work for me.
|
|
#22
|
|||
|
|||
|
Could this be it? (kind of an old patent by the looks of it:
http://www.patentstorm.us/patents/57...scription.html
|
|
#23
|
|||
|
|||
|
Quote:
![]()
|
|
#24
|
|||
|
|||
|
If it's done on GFX instead of colour management tools of OS then used application shouldn't have any effect to availability of colour space conversion.
|
|
#25
|
|||
|
|||
|
I'm not sure the term "colour management" is properly understood by some in this thread.
A colour managed application only acts differently than a non-colour managed application if there is an ICC profile embedded in the content material. If there is no tag and/or profile embedded in the content, there's nothing to manage. For instance, untagged JPEGs are generally processed as sRGB, because that is USUALLY what they are. When I generate JPEGs from my RAW format photos I specifically tag all of them sRGB. If I export something in AdobeRGB, it must be tagged with the aRGB profile. If I export a JPEG specifically for printing at a print house, I will convert it to the ICC profile of the specific printer and include that ICC profile in the JPEG (options when saving from Photoshop). A non-colour-aware picture viewer would treat both photos as sRGB and the aRGB photo would look way off (pale and dim, generally). A colour-aware picture viewer would read the aRGB tags and be able to translate that into the currently configured display's profile and display the picture with proper colour. There are four ways to translate out-of-gamut colour information and those method must be chosen. So, is any video content actually tagged with colour space information? HDTV, BR and content like that should all be in a standard colour space. The only alternative colour space I'm aware of that's anywhere close to the consumer market is x.v In that case, x.v content would need to be tagged as x.v, and a colour-aware player would have to translate from x.v to the colour space of your TV (if it isn't x.v). I don't think that x.v is a factor in most people using Media Player Classic. Now, calibrating a monitor so that it has proper grey scale, gamma and sRGB response (because sRGB is the computer "standard") will make your monitor as correct as possible. Applications such as games which expect sRGB colour space don't need to be colour managed to benefit from this. If the game expects sRGB and that isn't what your uncalibrated/profiled monitor displays - calibrating it will help regardless if the app knows anything about colour. The applications only needs to be colour managed if different colour spaces come into play (such as sRGB, aRGB, ProPhoto RGB with photos).
|
|
#26
|
|||
|
|||
|
Quote:
Calibration: If you have a non-sRGB monitor you calibrate it's gamma curve like just like you do for any monitor, this does NOTHING to change the inherent color differences. This just ensures that when you give red 185 in you get the appropriate level of red output to match the gamma curve you selected. Nothing at this step adjusts or reacts to color space. Profiling: The next step is profiling your monitor. This reads the actual color space of your monitor. This profile is used by color managed applications to do input to output color space translations. VERY few application do this. Color Management: Some applications do input color management, in that they will read the color space of the JPEG and translate, but they always assume an sRGB output. So they are not doing output color management as needed for different Gamut monitors. irfanView is one such application. It can (if enabled) translate your aRGB tagged files to sRGB for viewing, but it doesn't know about converting it to the color space of the monitor and always assumes sRGB (like most applications). AFAIK, outside of FF3 and a few graphics applications (and Vista desktop, newest Office?), nothing else uses the profile and color space will not be translated. Operation System/Video card fix?: What people are hoping for is something that will handle all the old non managed applications ( 99%+ of software). I am doubtful. Video card: I don't see a video card solution unless you treat everything as sRGB and use the profile to translate, effectively remapping ALL output to sRGB. There are issues with this. Actual profile aware color managed applications will be wrong because they will get double profiled. The second issue is that you have completely remapped your output device into an sRGB device (which is still likely inferior to a real sRGB). So what is the point at all of having wider gamut in this case? The remapping likely impacts display step size and image quality (more likely to have banding etc) and now you can't get the wider gamut at all. Operating System: The OS could probably pull this off because you could tag which applications get treated as sRGB and which are profile aware and passed through, but this is not there yet and who knows if it ever will be. There may be too big a performance hit to do color space translation for gaming. I will probably be out of the monitor market for the next 4 years. I wonder if there will be a good solution for legacy applications even then? Last edited by Snowdog; 08-29-2008 at 12:22 PM..
|
|
#27
|
|||
|
|||
|
Well it turns out that I was only partially right, or mostly wrong
![]() In Firefox 3 on Beasta with color management disabled, the saturation differences showed up between monitors. I then realized I never enabled color management on XP Firefox 3, so I went ahead and did that, and things turned out better. The only problem is that my standard gamut FP241VW shows less saturation because Firefox can only use one ICM profile within the color management app, which in this case is the LCD2690WUXI profile. Using Firefox 3 on NEC and Internet Exploder on the BenQ, things look identical on test images with strong reds and greens, so I guess I will load Firefox on the NEC and Internet Exploiter on the BenQ for consistent browising colors ![]() Additionally on videos the same thing with Media player classic. Skin tones aren't badly affected but strong greens and reds show up as over saturated. Games the same thing. The reds on the Gears of War menu screens are very strong on the NEC, and not so much on the BenQ. It turns out that Windows photo gallery built in to Vista IS color managed. The desktop backgrounds showed up with saturation differences on Beasta as well, just like they did on XP. Either way, it's not a huge deal for me, as I still absolutely love the NEC (as I also have a very good, uniform panel) and with console games the added saturation is not as bad, because I am used to it (as I turn up saturation on the BenQ through the OSD). I plan on loading up Photoshop and seeing how accurate the color is, as I also have (what I believe is) a wide gamut printer in the Epson R1800, so I'll print out some test images and see what I find. But it does seem that while this "gamut translation" is built in to the ATI drivers, it is not specifically enabled. Tamlin mentioned that it is only supposed to work in Beasta so I'll see what I can find concerning this, as I do prefer my ATI 4850/4870 to my 8800GTS (or 9800GTX+ dedicated 24/7 to folding right now), and I don't plan to remove the Radeons for a while. Also the 10-bit color depth is an added bonus, though I'm seriously doubting that it is doing much. Regards.
|
|
#28
|
|||
|
|||
|
Quote:
![]() 10 bit is almost certainly doing nothing unless you can convince me you have a 10 bit data path to your monitor and your monitor is 10 bit aware and accepts 10 bit signals and does the proper thing with them. IOW, no it won't do anything with current LCDs.
|
|
#29
|
|||
|
|||
|
The explanation that I heard is that it uses a form of dithering to "approximate" it. Yay more dithering!
Apparently the Dell 2709W does this as well to give supposedly expanded color space of 1.04B colors. Though I don't know how this is done if it's getting 8-bit input (????) Beasta = Vista = big RAM eating OS with lotsa nice bloatware to enhance efficiency, peformance and productivity. ![]() Quote:
|
|
#30
|
|||
|
|||
|
I really don't wan't my graphics card adding dithering, thanks, but no thanks.
It couldn't be FRC because they don't have the bandwidth to do that. So it would have to be static, which seems like a complete and silly waste. Besides what applications will support this "10bit" output? Basically this is a bunch of marketing smoke and mirrors. It would likely do more harm than good if enabled.
|
|
#31
|
|||
|
|||
|
Yes, out of control marketing with little to no useful real world results. We should be welcome to this new monitor landscape where TV features and marketing continue to pervade.
I will try out software calibration of sRGB mode on the 26" display and see how that works out. Just for fun after I take some super macro photos of text with different sharpness levels. Long weekend crash testing at its best.
|
|
#32
|
|||
|
|||
|
The downside
Quote:
![]()
|
|
#33
|
||||
|
||||
|
Quote:
Quote:
That would be exactly same thing as how TN monitors claim 8 bit colours even though TN LCD panels are only 6 bit accurate... and as I recall you're the one keeping it as perfectly working system. And while actual 10 bit accuracy from source to output device and device capable to it would be better this could still help avoid some of the colour banding problems when adjusting colours/gamma from graphic card's settings. Logically thinking as potential downside this already present FRC might cause somekind interference with monitor's own processing... especially with that for FRC in TNs.
|
|
#34
|
|||
|
|||
|
Quote:
All you can offer is static/spatial dither which is completely pointless. If there was a better shade to use on the graphics in the first place, you would already be using it.
|
|
#35
|
|||
|
|||
|
Quote:
http://www.displayblog.com/2008/06/1...4-lcd-monitor/ My TV has x.v.color so can display up to 16bits per colour. Some Blu-Ray players handle deep color or x.v.color and there is media that uses them.
|
|
#36
|
|||
|
|||
|
Quote:
Last edited by Snowdog; 08-29-2008 at 07:25 PM..
|
|
#37
|
|||
|
|||
|
Quote:
HDR will (and needs to) be able to use the larger colour space. I'm not sure if current HDR implementations can use it though. Art packages and video editing software benefift too. I dont know of any software that can utilise more than a 24bit display available now. Looks good for the future though ![]()
|
|
#38
|
||||
|
||||
|
Quote:
Quote:
Now if there is a trick I am missing, hit me with it. So is there a setting in the card that enables this feature or what? This is the sort of thing I doubt would be in a video card review but if it works that would be exciting stuff.
|
|
#39
|
|||
|
|||
|
Yep, I posted more findings later and rectified my original thoughts.
I then proceeded to do the following: 1) Warm up the NEC for an hour 2) Put it in sRGB mode and wait for a while (half hour or so) 3) Calibrate with BCC 4.1.8 in XP (crashes in Vista 64) 4) Transfer the color profile to Vista from XP and install Long story short, the colors are, as expected, far closer to sRGB than before. Greens are a bit too saturated, and reds are a bit maroon looking. It is similar to using ATI CCC to lower saturation by 20%, but not as fast (as the monitor needs time to "adjust to sRGB). Additionally if the ATI CCC avivo approach works for games, I think I'd prefer that right now. And you are right, none of the included applets other than Windows Gallery/Photo Viewer are color managed in Vista. Quote:
|
|
#40
|
|||
|
|||
|
Quote:
Load a color-profile to the graphic-cards LUT, yes - as you mentioned that would defeat the whole purpose of a wide-gamut display but it's getting increasingly harder to get a non-wide-gamut display anyway. Also you could easily swap out that profile when doing other work. This seems to be possible on ATi cards: http://www.driverheaven.net/vista-ra...r-profile.html Wouldn't that would work to *convert* a wide-gamut to an sRGB display (no, won't be as good as a native sRGB display) and when you don't want it (like when working in PS) you just turn it off or switch profile. Sounds to good to be true so I assume that I've missed something, I hope someone points it out ![]()
|
![]() |
| Thread Tools | Search this Thread |
|
|