|
 |
|
|
| View previous topic :: View next topic |
| Author |
Message |
mp20748
Joined: 12 Sep 2006
Posts: 5681
Location: Maryland
TV/Projector: 9500LC Ultra / Super 02 and 03 VIM
|
 Link Posted: Wed Mar 25, 2015 12:06 pm Post subject: |
 |
|
| CIR Engineering wrote: | | Luuk neele wrote: | | one thing has been bugging me for a long time though, why on earth does the JVC have a brown tint all over its image? |
I think I know what it is you are seeing. I personally call it a magenta tint. However, I don't think evaluating the color response of a projector on a monitor is at all helpful. I have used Casino Royale ever since the first BD was released because it was the first BD I saw that really had reference quality video (some problems, but still very good). The start of the second chapter with the cobra fighting the mongoose, when they pan over the spectators, on most digital projectors and TV's this will have a magenta cast to the light that should not be there...
That said, this is a problem with almost every digital projector that I calibrate. Many consumer grade digitals including JVC, Sony, and especially Epson work their CMS in the HLS color space (hue lightness saturation) instead of in the RGB color space. What results is a projector's color gamut looking perfectly like Rec709 on paper, but having large areas inside of the color gamut that are not correctly mapped resulting in nonlinear color response and incorrect colors. In fact, most digital projector's color gamut is very nonlinear even when the six primary colors are set perfectly.
This is what I have really loved about the Lumagen Radiance processors on most digital projectors. With a finite element system like the 729-point CMS, the nonlinearity can be almost completely removed from any given digital projector. If you get a chance to see a JVC with a Radiance processor that has been calibrated properly using the 125-point CMS (or higher) you will not see this magenta (brown) tint anymore and the colors will be much closer to what you see on a CRT.
Just for reference however, the DPI, Christie, and Sim2 projectors have perfectly linear color response. This is because the CMS works in RGB space and has also been properly implemented. With those projectors when you calibrate the six primaries at full saturation all colors between the primary colors and white will be nearly perfect. I have very good calibration equipment and I have proven this empirically and repeatedly with DPI, Christie, and Sim2. When I install a Radiance on said projectors and run an auto calibration on them, Chromapure will literally make no changes to any of the 729 points. In other words, the color error deltaE is less than 2 for all of the 729 points. You will never see this on most other digital projectors.
So while most digitals can be made to look good in their calibration reports with a 6-point CMS, there is much more to a perfect calibration.
craigr |
It's moments like this when Craig will take the time to post on something you're also familiar with, but neither can put in words or have the depth of knowledge to fully understand and elaborate on.
A lot here has been mentioned in several of the tech classes I had to take. And it also also explains why I think is just plain stupid to use screenshots for comparison. Unlike CRT, the color system used in modern day digital projectors are just too complex for it to be so easily captured by a camera, or to assume it's correct based on simple measures after the finals. It is also one of the reason I got away from being a calibration expert.
One DLP projector that I've had to service and maintain on sites, unlike a CRT projector that has very few actual color adjustments (8 or so) it had upwards of 32. There are at least 4 different RGB color adjustments in each projector. There are also color pixel alignment adjustments where to have to do pixel phase alignment for both the overall image as well as the colors. It can be done basically. but in order to get digitals to align right on some meters and produce near perfect white, there was a lot to do. Digital technology just does not behave as easy as CRT technology. That does not mean one is better, only that one could require more attention to get right. And the main reasons why these DLP's had so many adjustments is because of multi-screen matching.
And this is also why digitals do not show well their colors thru screenshots. They are way too complex to be able to use a point and shoot camera and expect good results. LCD is a little different, but the other digital technologies is get worse.
Try this. Bring up a colorful image on your digital. take a screenshots, and then watch what happens with the colors. They will mute every time. The difference in the shot is huge in comparison to what's on the screen. And this is proven with screen-grabs. They will never have the proper color representation of what's on the actual screen. We delt with this on a large video teleconference system that I was called in on. The very expensive HD cameras did well in capturing everything in the room, but when also trying to capture what's being displayed on the screen through the VTC System, the colors were muted and flat looking. So this is when we found out about what's on the screen being able to be captured using cameras. Digitals like the DPI and Christies perform a little better because they more easily produce a more linear color response, and that's mainly because of TI's DLP technology. CRT is best, with Lcos requiring maybe more work. But none of them should be used through the eye of a cheap camera.
But Craig makes a very good teaching on the tint issue by saying there is a lot more going on, and why you can't judge through a monitor. I would go further and say, it's not just the monitor, it's also the camera, the process of converting to file, and from file to further processing for the monitor. Which in that process, the original has lost its most critical timings (digital) for proper color rendering. On the projectors I've mentioned, they had adjustments that allowed you to make things a little better (phase), but because of proper pixel timings (not necessary with CRT) you could not get things pefect. The use of a camera and monitor image processing system will never be correct with digital images. Just simply look at what the camera/monitor does to the colors from what is on the actual screen for yourself
|
|
| Back to top |
|
 |
Andreas21
Joined: 02 Oct 2013
Posts: 582
|
| Link Posted: Wed Mar 25, 2015 8:49 pm Post subject: |
|
|
| mp20748 wrote: |
Try this. Bring up a colorful image on your digital. take a screenshots, and then watch what happens with the colors. They will mute every time. |
This depends on a lot of factors and different cameras will give you different results, it is no problem to take a shot that looks to have way to much color and wise versa.
_________________
http://www.minhembio.com/21Andreas
|
|
| Back to top |
|
|
CIR Engineering
Joined: 25 Aug 2008
Posts: 4264
Location: Chicago USA & Berlin Germany
|
| Link Posted: Wed Mar 25, 2015 10:01 pm Post subject: |
|
|
| Andreas21 wrote: | | mp20748 wrote: |
Try this. Bring up a colorful image on your digital. take a screenshots, and then watch what happens with the colors. They will mute every time. |
This depends on a lot of factors and different cameras will give you different results, it is no problem to take a shot that looks to have way to much color and wise versa. |
I am sitting here thinking about this as someone who really hasn't looked at many of the screen shots posted. However, if you look at a spectrogram of a digital projector you always find that the output peaks very high at specific wavelengths or at a specific wavelength. Where the primary output is positioned on a spectrogram depends on the bulb technology. Generally speaking, LED is the narrowest and Xenon (or metal halide) has the broadest spectrum. Indeed, I seem to recall that metal halide is the closest of all bulbs to that of sunlight, but they are unfortunately not used in any projectors that I am aware of. Bulb spectra is a little outside what I normally look at so forgive me if I am wrong on any of this.
If you look at the sun, the spectral output is what we can consider reference and is very broad.
CRT does have a broad spectrum that is much closer to sunlight as well. I wonder if certain digital projectors might look as though their colors are muted when photographed because their spectral output does not match the digital camera's abilities to pick up the spectrum that is being emitted from the projector. I also wonder if this is why when I take digital photos of the real world the colors often don't look like what they did when I took them.
Not sure what I am getting at here, but I was just thinking and figured it was food for thought.
One thing's for sure, most digital projectors suffer from excessive colors when they are uncalibrated or not in their correct mode for cinema and TV. The color gamuts that digital projectors can create are far richer and cover more of the visible color spectrum than CRT's are capable of (without diminishing light too much). That said, both CRT and digital can easily produce the full Rec709 color gamut (with correct filtration on the CRT). At the moment the larger gamut digital is capable of is useless and deleterious to the image for the time being and for BD and other common sources in a home theater. If, or when, we get a new color gamut standard, the large gamut potential of digital projectors will become necessary, but for now to make colors correct based on Rec709 digitals usually need a heavy handed CMS to dial back the gamut saturation.
craigr
_________________
*NEW JETI 1501-HiRes 2nm Spectroradiometer
JETI 1211 Spectroradiometer
Photo Research PR-650 Spectroradiometer
Klein K10-A Colorimeter
X-Rite i1Pro2 Spectroradiometer & Spyder Colorimeters *For JVC auto-calibration when Klein & Jeti are not applicable
Murideo Fresco SIX-G HDMI 2.x Multimedia Generator
Murideo Fresco SIX-A HDMI 2.x Analyzer
*NEW Light Illusion ColourSpace XPT Version β Color Calibration Software
Light Illusion LightSpace XPT Pro Version 10.x Color Calibration Software
*NEW OMARDRIS JVC Software Patch to use K10-A and Jeti with JVC OEM AutoCal Software!
Sencore CR7000 CRT Tube Analyzer / Rejuvenater
Authorized Dealer for Lumagen & just about everything worth buying 
www.CIR-Engineering.com - craigr@cir-engineering.com
Phone: 865-405-6892
|
|
| Back to top |
|
|
CIR Engineering
Joined: 25 Aug 2008
Posts: 4264
Location: Chicago USA & Berlin Germany
|
| Link Posted: Wed Mar 25, 2015 10:11 pm Post subject: |
|
|
| Andreas21 wrote: | With the latest JVC line the colorspace (also grayscale/gamma is very good) is almost 100% linear in most of the ones I have seen. And you should really check out the eeColor as it in another league than the Lumagen that has alot of problems. After some tests I threw out my Lumagen when I discovered that it has major flaws, and I bought the eecolor and it does not have the proseesing power the Lumagen has but it is a very good external CMS(it is only a CMS) with up to 65x65x65 point 3D lut.
And the screenshots Diddern and I has posted is calibrated with the eecolor and has no colorshift towards magenta like you claim. Mine is corrected at 4913 Points, and has deltaE less than 1 on all points, and grayscale/gamma is "perfect". It is calibrated with high end equpment like you have and the calibrator is very experienced. |
JVC has gone leaps and bounds with their CMS since the RS20. Every year they have made the CMS perform better and better. That said, what JVC is doing is applying more and more bandaids to their algorithms to force the color to become more linear than I ever thought it was capable of natively. JVC has done the best job of the major consumer grade projectors in creating a viable CMS (Sony is absolutely abysmal and Epson is also rather terrible). If I was going to run a consumer grade digital projector without an external CMS I would choose the JVC because it does, in reality, do a very decent job now.
I really don't think many of the JVC's ever had much issue with grayscale or gamma response with respect to luma. They do quite well with luma response. Their only issue has been where digital domain chroma information is translated back to actual color. JVC hasn't ever looked undersaturated to me, just not quite correct.
I really would love to post more right now as this conversation is actually stimulating  Unfortunately my wife has to go to the Dr so I now have to watch my son until he goes to sleep.
I'll try and drop back in when I have free time.
Thanks guys,
craigr
_________________
*NEW JETI 1501-HiRes 2nm Spectroradiometer
JETI 1211 Spectroradiometer
Photo Research PR-650 Spectroradiometer
Klein K10-A Colorimeter
X-Rite i1Pro2 Spectroradiometer & Spyder Colorimeters *For JVC auto-calibration when Klein & Jeti are not applicable
Murideo Fresco SIX-G HDMI 2.x Multimedia Generator
Murideo Fresco SIX-A HDMI 2.x Analyzer
*NEW Light Illusion ColourSpace XPT Version β Color Calibration Software
Light Illusion LightSpace XPT Pro Version 10.x Color Calibration Software
*NEW OMARDRIS JVC Software Patch to use K10-A and Jeti with JVC OEM AutoCal Software!
Sencore CR7000 CRT Tube Analyzer / Rejuvenater
Authorized Dealer for Lumagen & just about everything worth buying
www.CIR-Engineering.com - craigr@cir-engineering.com
Phone: 865-405-6892
|
|
| Back to top |
|
|
Andreas21
Joined: 02 Oct 2013
Posts: 582
|
| Link Posted: Wed Mar 25, 2015 10:36 pm Post subject: |
|
|
| CIR Engineering wrote: | | Andreas21 wrote: | With the latest JVC line the colorspace (also grayscale/gamma is very good) is almost 100% linear in most of the ones I have seen. And you should really check out the eeColor as it in another league than the Lumagen that has alot of problems. After some tests I threw out my Lumagen when I discovered that it has major flaws, and I bought the eecolor and it does not have the proseesing power the Lumagen has but it is a very good external CMS(it is only a CMS) with up to 65x65x65 point 3D lut.
And the screenshots Diddern and I has posted is calibrated with the eecolor and has no colorshift towards magenta like you claim. Mine is corrected at 4913 Points, and has deltaE less than 1 on all points, and grayscale/gamma is "perfect". It is calibrated with high end equpment like you have and the calibrator is very experienced. |
JVC has gone leaps and bounds with their CMS since the RS20. Every year they have made the CMS perform better and better. That said, what JVC is doing is applying more and more bandaids to their algorithms to force the color to become more linear than I ever thought it was capable of natively. JVC has done the best job of the major consumer grade projectors in creating a viable CMS (Sony is absolutely abysmal and Epson is also rather terrible). If I was going to run a consumer grade digital projector without an external CMS I would choose the JVC because it does, in reality, do a very decent job now.
I really don't think many of the JVC's ever had much issue with grayscale or gamma response with respect to luma. They do quite well with luma response. Their only issue has been where digital domain chroma information is translated back to actual color. JVC hasn't ever looked undersaturated to me, just not quite correct.
I really would love to post more right now as this conversation is actually stimulating Unfortunately my wife has to go to the Dr so I now have to watch my son until he goes to sleep.
I'll try and drop back in when I have free time.
Thanks guys,
craigr |
This is quite the opposite of most calibrators I know mean, and the JVC internal CMS does not work very well according to them and according to one of the best calibrators in Norway the CMS in the latest JVC´s is actually a step back. But with the latest model most of them are quite linear so it is possible to make them quite good with the internal CMS, but if you are unlucky and get a "bad" the internal CMS does not make the job easy. As I am not a calibrator I can not tell you why.
I have used Lumagen for many years with different Sony projectors and I am now using a eeColor with my X500 and it is to me a much better CMS than the Lumagen, with better results on screen.
_________________
http://www.minhembio.com/21Andreas
|
|
| Back to top |
|
|
Spanky Ham
Joined: 22 Mar 2006
Posts: 5643
Location: Comedy Central
|
| Link Posted: Thu Mar 26, 2015 3:41 am Post subject: |
|
|
|
One of the problems with a narrow spectrum light source, as Craig alluded to, is everyone sees color differently. Therefore the image could look different depending on the person. With a broad spectrum, the differences are subdued.
|
|
| Back to top |
|
|
carpfisher
Joined: 30 Nov 2006
Posts: 19
|
| Link Posted: Fri Mar 27, 2015 9:16 pm Post subject: |
|
|
|
Sorry if this has already been mentioned, but I've only read the last couple of pages. What a lot of people forget to turn off on their cameras is the Auto White Balance. This is where the camera automatically adjusts the RGB in the saved picture according to how 'off' it thinks white might be. This very easily throws off images which have one dominant colour. Ever wondered why your pictures of sunsets look lame compared to how you remember the real thing? AWB working its evil. Turn it off and it will transform your pictures. A lot of point and click cameras might not even allow you to do this, but my Sony does. If you leave it on, a screenshot will never look right.
|
|
| Back to top |
|
|
kal
Forum Administrator
Joined: 06 Mar 2006
Posts: 17860
Location: Ottawa, Canada
TV/Projector: JVC DLA-NZ7
|
| Link Posted: Fri Mar 27, 2015 9:48 pm Post subject: |
|
|
| carpfisher wrote: | | Sorry if this has already been mentioned, but I've only read the last couple of pages. What a lot of people forget to turn off on their cameras is the Auto White Balance. This is where the camera automatically adjusts the RGB in the saved picture according to how 'off' it thinks white might be. This very easily throws off images which have one dominant colour. Ever wondered why your pictures of sunsets look lame compared to how you remember the real thing? AWB working its evil. Turn it off and it will transform your pictures. A lot of point and click cameras might not even allow you to do this, but my Sony does. If you leave it on, a screenshot will never look right. |
Even better is to shoot RAW and use a grey card for reference. That's what I do. For HT screenshots, you can do this easily by putting up any greyscale IRE window or fullscreen pattern.
Kal
_________________
Support our site by using our affiliate links. We thank you!
My basement/HT/bar/brewery build 2.0
|
|
| Back to top |
|
|
|
|
|
|
|
|
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
You cannot attach files in this forum
You can download files in this forum
|
Forum powered by phpBB © phpBB Group
|
|
Forum FAQ
Search
Memberlist
Photo Albums
Register
Profile
Private Messages
Log in
