Reading reviews, there's a lot about CR, CR, and CR, with numbers to back up observation. When it comes to colors, the talk is mostly about accuracy. I don't often hear a reviewer says "X has nicer colors than Y". I guess that's too subjective and may be meaningless. With plasmas, you hear claims of the display being having 1 billion colors or whatever, but with PJs, it's mostly about CR and lumens.

I bring this up b/c I have a 1080UB home now and I can't shake the impression the Panny AE2000 had (going from memory) "better" colors -- more saturated, deeper maybe? I just remember the colors on the Panny really wowing me. I don't know if it's my imagination or I was prejudiced by a comment someone else made or what.

My question is is this even possible? Or will 2 PJs with D7 panels have pretty much the same colors once properly calibrated. Or did Panny do something better? Certainly the Epson has better CR so why shouldn't another PJ have better colors (however that's defined)?

Some projectors like LCDs/LCoS often have a colour gamut where the primaries/secondaries (RGBYCM) exceed the colour gamut of the system you're using, so if you're using NTSC and REC601, greens may look too green because a strong green is being shown as too saturated - it's gone beyond the point that it should be at for a full green (i.e 0 to 10 is the gamut, but yours goes from 0 to 15, with 11 to 15 being greener than it should be, so a 10 will be shown as a 15 instead).

This is often shown in the good reviews with a CIE chart showing the standard gamut and the gamut the projector can produce. When you can see that you will know if the display will be showing over saturated colours or not.

Some projectors have a means to move the primaries/secondaries but quite often you have no control over it unless the colour saturation control works correctly. Then you can reduce the saturation of all the colors and reign them in closer to their correct positions. However, not all colours are equally over saturated so reigning in one color to correct it may leave the others under saturated.

An external video processor with full CMS can correct it but that's more expense you may not wish to incur.

Having billions of colours is only useful if you have source material that can take advantage of that. Currently most things are 8 bit colour so that's all your display will have to render. I don't know if any can stretch the source colour to the extended colours or what that would look like.

Gary

This is often shown in the good reviews with a CIE chart showing the standard gamut and the gamut the projector can produce. When you can see that you will know if the display will be showing over saturated colours or not.

What affect does luminance dimension have on the CIE ? Using your example what happens if the CIE for a color should be"10" and the PJ CIE gamut measures "10" for but the luminance measures "15" instead of "10"?

What impact would this have on image even if the CIE and meaured gamut appear very close?

D

Reading reviews, there's a lot about CR, CR, and CR, with numbers to back up observation. When it comes to colors, the talk is mostly about accuracy. I don't often hear a reviewer says "X has nicer colors than Y". I guess that's too subjective and may be meaningless. With plasmas, you hear claims of the display being having 1 billion colors or whatever, but with PJs, it's mostly about CR and lumens.

I bring this up b/c I have a 1080UB home now and I can't shake the impression the Panny AE2000 had (going from memory) "better" colors -- more saturated, deeper maybe? I just remember the colors on the Panny really wowing me. I don't know if it's my imagination or I was prejudiced by a comment someone else made or what.

My question is is this even possible? Or will 2 PJs with D7 panels have pretty much the same colors once properly calibrated. Or did Panny do something better? Certainly the Epson has better CR so why shouldn't another PJ have better colors (however that's defined)?
CR stands for...???

CR stands for...???
OK! I've just got it!!!
CR = Contrast Ratio!!! :)

Hi Dazzer,

Good question :)

At a guess I think that if the luminance is high then you may have a gamma problem, i.e an 's' shape curve that will be flat at the top rather than a curve going from low to high smoothly from black to white. I seem to remember seeing a 3D representation of all those relationships so I'll see if I can find it and post it here.

I find threads like this a little depressing, because I think color performance is very important and these make it clear how little attention is paid to it.

First, the Epson 1080UB has one of the best CMSs in the business built right in to the unit. It is theoretically capable of nearly perfect color performance.

Second, color performance, unlike almost any other aspect of display performance, is very precisely quantifiable. It is perhaps the LEAST subjective.

Each color is specifically defined in terms of a xyY value by the gamut in which the material was mastered (SD or HD). Those values can be measured by instrumentation, and the amount of error can be expressed in dE units, a numerical figure that describes amount of color error.

Agreed and despite that ,Tom, several highlly regarded forum members feel that it is way down the list and should be subjectified.

One vo my favorites, it dismisses color accuracy quest based on what joe six pack thinks is right;

"Color accuracy is important...to Bob Sorel. Most marketing including "brilliant color" wants colors beyon the normal color specs. That's what 99.9% of the population wants."

Art

Originally posted by Tom Huffman:

I find threads like this a little depressing, because I think color performance is very important and these make it clear how little attention is paid to it.


I agree with you that color is quantifiable in accordance with SMPTE standards. Standardized color itself, therefore, is not subjective. People's preference for a certain color image, however, is. Although many of us on this forum prefer colors in accordance with the SMPTE standard and an image calibrated accurately to a D65 grayscale, many others do not.

Perhaps because of the way many manafacturers and dealers opted to display their images, oversaturated or with contracst turned way way up, consumers became used to these images and expected displays to look like that. As Greg Rogers often likes to state, many people prefer an "oversaturated" color gamut. It is also not uncommon to hear people say that a properly calibrated display looks "washed out" or other some other negative comment.

The question, therefore, becomes how do we judge color preference? A simple answer would be, we can't. A reviewer, therefore, can only report on quantifiable standards and whether a particular projector is capable of displaying that standard, which is extremely important to many us. A useful report, like those written by Greg Rogers, can tell us the projector's ability to change color gamut but stating that the projector is inferior because it cannot display a particular SMPTE standard, for some may be misleading, because, like in the case of the RS1 many consumers may prefer the incorrect gamut to the correct one.

Judging color, therefore, in my opinion comes down to viewing impressions by each particular consumer. That is why it is so important before making any significant purchase to view the product in a proper environment. What is unfortunate in our industry, however, is that there are so few places where products are displayed properly. As a result, I do not believe the general public have been provided an opportunity to discover how far FP technology has come.

I'm not even sure what the 'correct' colors are--SMPTE-C, Rec 709, Rec 601, ... ? The first 2 are mentioned most frequently, but I think Rec 709 has a larger gamut than SMPTE-C, at least that's my inference from various discussions. (It's clear I don't know what I'm talking about, but would like to!)

Just to return to the original post about the AE2000 compared to the 1080UB: If you change modes on the AE2000, the 'Cinema 1' mode has quite a different 'colour' to, say, the 'Colour 1' mode, yet if you look in the menus the colour/tint/brightness/contrast are (out of the box anyway) at the same values. In other words, the projector's colour settings are 'hidden' within these modes. Not that they can't be adjusted using the user's menu controls, but the base line is set. Therefore the 1080UB may well have the same D7 panels, but the circuitry and 'hidden' settings within may well give quite a different result.

The problem I have about 'accurate' colour is one of reference. My eyes seem to compensate themselves quite well enough. I have sometimes started to watch a film not realising the mode was 'Cinema 1' instead of 'Colour 1' (which is reported as being closer to correct OOTB). Only when I flick back to the 'correct' setting can I see the change and 'Colour 1' does look more 'natural' than 'Cinema 1'. In isolation it is hard to tell. Where as with brightness and contrast settings, I can more easliy tell if they aren't quite right, particularly if the black level is too high.

All IMHO of course and some might berate me for not being able to tell when I have perfect colour....

Just put your Panny on 'Color1' and it will look the same - because it nails the REC709/D65 standard. You must be having it on one of the blownout color presets - yes it is more colorful - but that is inaccurate - those colors are not how the media is to be reproduced.

Your brain has been programmed that overblown color and contrast is how TV should look - just leave it on the 'color1' preset for two weeks before switching back - you will then reprogram your brain how films on video media should look - and not be able to watch anything else again. It takes a while to see what you have been missing - luckily you don't need to pay ISF to see it in the first place - the right preset is close enough - but it might make you a videophile that wants it ISF to be just perfect.

The colors have nothing to do with the technology of the panels - it has to do with the choices made of optical color filters and any digital corrections. Without them the panels themselves would make a B&W picture.

People have preferences for all kinds of things. I would never (well, almost never) engage in debate about personal preferences. It's mostly a waste of time. If you prefer a bluish white, bright reds, and an oversaturated color palette, then go with god. You can reserve the right to like what you like and I'll reserve the right to have my own opinion about your tastes.

My point wasn't about preferences, but about the fact that so many people are not even aware that color can be discussed in any terms other than personal preference, despite the fact that other aspects of performance (contrast, sharpness, light output) are routinely assumed to be based on objective standards.

To address Millerwill's question: SD (SMPTE-C) and HD (Rec. 709) each includes specific, objective, and clearly defined definitions for white and the primary and secondary colors. In the same way that most displays fall short of those standards with respect to white and require grayscale calibration to D65, so too they often fall short with respect to the primary and secondary colors. The Epson 1080UB is a perfect example of a display that provides all of the tools in the user menu to fully address these problems.

It was the fact that the owner was completely unaware of the tremendous added value this feature gives the 1080UB that I found depressing, but that's just me.

Each color is specifically defined in terms of a xyY value by the gamut in which the material was mastered (SD or HD). Those values can be measured by instrumentation, and the amount of error can be expressed in dE units, a numerical figure that describes amount of color error.One of the problems I see is that HD material is largely mastered on monitors that have SMPTE-C primaries. And I believe the main reason HD material is mastered that way is to save money. That is, the places doing the mastering already have those monitors, so mastering HD with HD primaries on CRTs would require buying new monitors with HD primaries, while continuing to use the older monitors they already own doesn't. In that case there are colors that are "correct" based on the mastering that was done, but still may be quite different than what was originally filmed just because of the limitations in the primaries of those monitors (and not because somebody made a choice for that scene to pull a color in). For instance, a car that is a strong red may get filmed and then mastered on those monitors with a push toward orange because of the SMPTE-C primaries having a red that is toward orange from REC.709 or even compared to what the color of the car actually was while filmed.

The current situation definitely isn't ideal. For most of the HD material we get SMPTE-C primaries are what is correct based on the mastering done on those CRT monitors, but that means missing many colors that exist in real life, just because of those monitors. I personally prefer using REC.709 primaries, but anybody who sets up that way is doing something different than the vast majority of CRT monitors used to master the material.

When something is artificially limited (like colors have been due to the limited primaries of SMPTE-C) I think it is natural that some would prefer less limitation, even when that isn't correct for the material.

Tom Huffman,

Do you normally watch HD with SMPTE-C primaries, or REC.709 primaries (when you have control of the primaries)?

--Darin

Darin,
I agree with your point regarding reds in SMPTE-c but the fact that there are standards including 709, still, there are standards. This allows one to get closer to the way the material was intended to be viewed.

Point being even if we feel that the standards need updating color accuracy is important.

Art

I am aware that the 1080UB (I have the Home version) has, besides the color modes presets and "skin tone" -- RGB gain and offsets to adjust gray scale, and RGBCYM hue/saturation/brtness adjustments. Is that enough to correct any color problem? (For ex, I think mine has a bit of red push.)

I think these videos help explain my prefs (mite sound or ignore the commentary):

http://www.youtube.com/watch?v=Fx_ayoLAZKw
http://www.youtube.com/watch?v=TZYT6DXL08s&NR=1
http://www.youtube.com/watch?v=kW1gazuwgzM&NR=1

I prefer the colors of the display on the right, which happens to be the Panny. (of course we don't know how the displays are set up and how the camera angle influenced the video and what the bulb hours were, etc. etc.)

Anyway, I think I'm one of the rubes who likes to have a bit oversaturated colors. However, I do NOT find increasing color saturation on the 1080UB gives me more pleasing colors, but more unnatural and garish.

I guess what I'm asking is whether

A) there's something in the AE2000 with regards to colors that the 1080UB cannot match, like Panny uses a better color processor or whatever. (I read that Panny consulted with Hollywood colorists, fwiw.), or

B) the Epson can do anything the Panny can with respect to colors -- even if what I want isn't technically correct. It's just a matter of calibration.

The current situation definitely isn't ideal. For most of the HD material we get SMPTE-C primaries are what is correct based on the mastering done on those CRT monitors, but that means missing many colors that exist in real life, just because of those monitors. I personally prefer using REC.709 primaries, but anybody who sets up that way is doing something different than the vast majority of CRT monitors used to master the material.

Do you normally watch HD with SMPTE-C primaries, or REC.709 primaries (when you have control of the primaries)?Darin:
No the situation isn't ideal, but you do the best you can do. It won't be ideal until (if ever) the gamut of film and the gamut of commercial video are the same. Until then, calibrating to the standards in which the material was mastered yields by far the best results.

As for myself, I make judgments on a case-by-case basis. I find that a lot of HD material on broadcast looks better in SMPTE-C and that most Blu-ray content looks better in Rec. 709. Note: the differences are about 9 dE, so were are not talking about night and day here, but it is a perceptible difference.

Interestingly, I find Rec. 709 to be much more forgiving in that even with material that looks better when viewed with a SMPTE-C calibrated display, the difference is quite small when viewed with a Rec. 709 calibrated display. On the other hand, when material looks better when viewed with a Rec. 709 calibrated display, shifting to SMPTE-C looks VERY wrong to me. It gives the image a greenish tint that I find especially disturbing. YMMV.

Darin,
I agree with your point regarding reds in SMPTE-c but the fact that there are standards including 709, still, there are standards. This allows one to get closer to the way the material was intended to be viewed.

Point being even if we feel that the standards need updating color accuracy is important.Depends on what "intended to be viewed" is applied to. If the intention of the original material was a REC.709 red, but it was mastered on a CRT monitor with SMPTE-C primaries, which way was it intended to be viewed and why? Basically, a person could argue that it was intended to be viewed with SMPTE-C primaries, but if that was only because the mastering house didn't want to spend money to master on a monitor with REC.709 primaries, then a person could also make a good argument that the object which was originally as red as red is in REC.709 was intended to be viewed with REC.709 red, and was only deviated from that to save money.

Do you view with SMPTE-C primaries or REC.709?

--Darin

Until then, calibrating to the standards in which the material was mastered yields by far the best results.Okay, you just said the above, but then said that you view most Blu-ray content with Rec. 709 primaries (separate from the decoding matrix). I believe that the vast majority of both HD DVD and Blu-ray content is mastered on monitors with SMPTE-C primaries, so going with the primaries used for mastering would mean using SMPTE-C and not Rec. 709 for most of that content.

Just to be sure, I'm assuming in the above that you were talking about the primaries, and not switching to Rec. 601 decoding.

--Darin

Okay, you just said the above, but then said that you view most Blu-ray content with Rec. 709 primaries (separate from the decoding matrix). I believe that the vast majority of both HD DVD and Blu-ray content is mastered on monitors with SMPTE-C primaries, so going with the primaries used for mastering would mean using SMPTE-C and not Rec. 709 for most of that content.

Just to be sure, I'm assuming in the above that you were talking about the primaries, and not switching to Rec. 601 decoding.The matrix math is exactly the same for SMPTE-C and Rec. 709. The targets are defined by the primaries and the white point. You get different results only because the primaries are different.

I have no way of knowing how any individual film was mastered, so all I can base my viewing preferences on is which of the 2 standards subjectively look better to my eyes. As I said, Rec. 709 seems more forgiving so that may play a role.

The important point is that the differences between SMPTE-C and Rec. 709 are really, really small compared to the types of deviations from EITHER standard that we see with many modern digital displays. For this reason, I think that emphasizing these differences distracts from the much more important issue of reasonably accurate colorimetry in general.

Let's get that right first and then worry about which of these two standards we calibrate to.

Depends on what "intended to be viewed" is applied to. If the intention of the original material was a REC.709 red, but it was mastered on a CRT monitor with SMPTE-C primaries, which way was it intended to be viewed and why? Basically, a person could argue that it was intended to be viewed with SMPTE-C primaries, but if that was only because the mastering house didn't want to spend money to master on a monitor with REC.709 primaries, then a person could also make a good argument that the object which was originally as red as red is in REC.709 was intended to be viewed with REC.709 red, and was only deviated from that to save money.

Do you view with SMPTE-C primaries or REC.709?

--Darin

No, I get it but since they are mastering with SMPTE-C now, this is the most logical standard irrespective of what the reason the mastering facilities had for going that way.

Right now I'm using SMPTE C but am considering 709 for the better looking reds (IMO).

Don't get me wrong ,I'm not married to any of these standards but after sitting with an RS2 the other night, and seeing some very intersting colors that literally I'd never seen any place including crimson Ferraris. I think that having some sort of reasonably accurate color set to some baseline is a good goal.

Art

The important point is that the differences between SMPTE-C and Rec. 709 are really, really small compared to the types of deviations from EITHER standard that we see with many modern digital displays. For this reason, I think that emphasizing these differences distracts from the much more important issue of reasonably accurate colorimetry in general.



Yes !

Art

I have no way of knowing how any individual film was mastered ...If the content came from any of the 6 major studios I think you can be pretty sure that it was mastered with a CRT monitor using SMPTE-C primaries (although many of these things are also checked on other displays, like for near black performance). At least from last time I asked somebody who works in this area.

If you knew that the content was mastered with SMPTE-C, but it looked better displayed with Rec. 709 to your eyes, which would you use?

I do think many displays go way too far, but I would lean toward Rec. 709 for playback even if mastered on a CRT monitor with SMPTE-C when that was done to save money instead of moving forward to the HD standard that is out there.
Right now I'm using SMPTE C but am considering 709 for the better looking reds (IMO).That's what I would probably do, although if you can set things up so you can easily choose between them, that would probably be best. One nice thing about DLPs is that it is pretty much just math once just a few measurements are taken and the Planar 8150 has the ability to easily choose between multiple sets of primaries. Besides auto there is SMPTE-C, Rec. 709, and native. On the one I saw the only visible difference between Rec. 709 and native was that native had the red out a little bit further.

I was surprised when I found out that the RS2 had pushed red further than even the RS1 did. I think they went too far and consider it one of the advantages of the RS1 (although the RS2 has some other advantages).
I think that having some sort of reasonably accurate color set to some baseline is a good goal.I agree. I'm just not willing to tie myself to the more orangish reds of SMPTE-C for HD.

--Darin

Reading reviews, there's a lot about CR, CR, and CR, with numbers to back up observation. When it comes to colors, the talk is mostly about accuracy. I don't often hear a reviewer says "X has nicer colors than Y". I guess that's too subjective and may be meaningless. With plasmas, you hear claims of the display being having 1 billion colors or whatever, but with PJs, it's mostly about CR and lumens.

Like others have stated, colors are definately quantifiable. The issue here is that the definition of color quality isn't as easy to fully understand, as CR or resolution. People tend to put most weight in the specs that they understand, rather than the specs that are actually important. In that regard, "1 billion colors" says absolutely nothing at all about the color quality of a plasma. When we talk about color accuracy, it's exactly the same as discussing 5000:1 vs 15.000:1. Like contrast, the perfect scenario can't be debated, however, if two pj's have different strong points, it certainly can be debated which is better. For instance, which PJ has better contrast: The one with 15.000:1 on/off and 300:1 ansi, or the one with 5000:1 on/off and 900:1 ansi? Similarly, which display has the "best" colors: The one with oversaturated primaries but correct color decoding, or the one with accurate primaries but incorrect color decoding?

In other words, if you think one projector has "nicer colors", it is quite simply mostly down to some difference in measurable color performance, that may or may not be correctable (though some areas of what you might call color performance is related to things like shading, banding etc).


My question is is this even possible? Or will 2 PJs with D7 panels have pretty much the same colors once properly calibrated.

Did you see both properly calibrated, in the same room with the same material?

An analogy for you: Deliberately oversaturating all movies is like only listening to techno remixes of all music genres. While I'm sure there are some people that only listens to Bach if it has an "oomph oomph oomph" in the background, to most people it's at best a gimmick that's very funny for about 10 seconds. For some reason, it's okay for a lot of people to watch "techno remixes" of movies.

There's actually been a very good example of this recently, some dude made a remix of Pink Floyd, Another brick in the wall pt.2. I heard it on the radio a dozen times, and hated it every time. Then I heard the original (which I've listened to hundreds of times) on the same radio station, and the first thing that came to mind was "hmm, that's slow, did they turn down the pitch control?" What I'm saying is: If your display is oversaturating the colors, all you will get out of it is change your own perception of the point where colors look saturated. As soon as you get used to more saturated colors, the impressiveness of it all fades away, and you'll be stuck with the errors that the oversaturated colors inevitably introduce.

When we talk about color accuracy, it's exactly the same as discussing 5000:1 vs 15.000:1.I see one difference. Real life can have huge CRs (I can get over a million to 1 just by going outside and then going inside and closing the door in a dark room or walk-in closet). But colors are being artificially limited from what real life contains in order to save money, limitations of the technology, etc. A red that is oversaturated compared to SMPTE-C may be exactly the color the object had while filmed or the color that would have been seen while mastering if it hadn't been artificially limited because of what are basically standard definition primaries on monitors used for mastering.

As has been stated, things can definitely be taken too far. But I don't see artificially pulling the colors in from real life as the same as trying to get the technology to be capable of the kinds of huge CRs that are present in real life. I see artificially limiting the colors because of the limitations of the monitors used (especially those with SD primaries) as more like artificially limiting the CRs, like purposely lowering the ANSI CR so that it isn't any higher than what mastering monitors have. Or purposely limiting the on/off CR similarly. There is the one difference between the situations in that the standards for CR don't artificially limit them, where the standards for color do, though. Maybe because they needed to for one, but didn't see that need for the other at the time the standards were written.

--Darin

When switching between Rec. 709 and HD1 primaries the difference depends on the source and goes from very obvious to rather small. When I start to dislike skin colors it's time to switch to Rec. 709. But some films are graded such that also with Rec. 709 the skin looks yellowish or greenish or whatever. And with HD1 colors it looks then 'sick'.
But when I watch "Cars" I prefer the cartoon colors of the HD1. And for some desaturated films as well.
The whole color gamut issue is for me not really relevant outside of skin color and some nature shots or other familiar subjects unless I know exactly what the master looks like (I usually don't). If a sweater has this red or that red etc. does not make or break a film for me. That stuff is manipulated so much with (digital) grading anyway that one quickly learns to accept that there is no fundamental right or wrong here, only choices made by whoever grades this and gets it approved. I don't like general tints in a film that should not be there, but that is not the same as more saturated primaries. Basically ignorance is bliss.
The sooner we have cinema color gamut and sources the better, but till then what we see is often not what the (film) original looks like, whether we use SMPTE-C, Rec 709 or HD1 primaries.

I'm not even sure what the 'correct' colors are--SMPTE-C, Rec 709, Rec 601, ... ? The first 2 are mentioned most frequently, but I think Rec 709 has a larger gamut than SMPTE-C, at least that's my inference from various discussions. (It's clear I don't know what I'm talking about, but would like to!)


The gamut difference between SMPTE C and Rec 709 primaries is far smaller than the differences between the primaries of various displays and the calibration of displays with CMS. In practice, it is not nearly as big a problem as improper encoding and decoding matrix selection and display calibration. Rec 601 does not even specify primaries. There is only one gamut specification that should be used for video currently and those are the primaries in Rec 709. If all displays were calibrated to these and the correct encoding and decoding matrix is selected, there would be little debate about the matter. In practice we have to make trade-offs, but getting as close to Rec 709 primaries as possible is the best practice and will produce the best results in most cases.

I am sceptical of the arguments made about rigidly adhering to color standards.

The most common argument made in support of existing standards is that, when adhering to then, we view material "as it was intended to be seen." The implication here is that a film's director, cinematographer and producer had something to do with that DVD or bluray disk that we are playing at home. This is not true. The quality of film transfers varies greatly in many ways including color. Plus, the standars are flawed because all the colors in film can never be captured by them! While the director, cinematographer or producer clearly had some intent about how a film would look on the screen, people in the industry that post here claim that that intent rarely makes it into the theater, much less into our homes. Thus, adhering to SMTPE C or REC 709 will NOT enable us to view the film as it was intended.

So, I ask the following question: what benefit are we getting by adhering to these standards? People here admit to playing SMTPE C material in a REC 709 gamut to get better reds. How is a projector manufacturer widening the gamut even further any different? I can't see how it is.

Manufacturers should offer presets that have many different color options. We are getting imperfectly mastered material. If someone thinks it looks better with a wider gamut, who is anyone else to argue with that preference.

It is small-minded in the worst sort of way to declare that color performance is objectively measured when the material that you are viewing is imperfectly mastered. I, for one, would rather watch colors that look good than colors that are accurate but bad looking.

Lawguy: I heartily agree about how nice it would be if pj's had several preset color gamuts to choose from. When I saw JVC's vp with the RS2 at the CES this past Jan, it was really nice just to be able to switch between them (Rec 709 and the RS2's original 'expanded' gamut). Seems extravagant to have to get an entire vp for this though. It's my hope that the next RSn will have this, but I haven't heard any promising rumors to this effect.

It's my hope that the next RSn will have this, but I haven't heard any promising rumors to this effect.

Sadly, nor have I.

So, I ask the following question: what benefit are we getting by adhering to these standards? People here admit to playing SMTPE C material in a REC 709 gamut to get better reds. How is a projector manufacturer widening the gamut even further any different? I can't see how it is.
Think of it this way - when a transfer is made from film to high definition media, compromises in the color palette are made (fewer colors are available). When a display gamut is correct, you will at least be seeing the "least damaged" version of color reproduction. When a larger than standard gamut is used, it DOES NOT reverse the gamut compression used in mastering, but rather expands the colors in a way that has no relationship to the original palette. That is, color "compression" is a controlled methodology, but the expansion used with larger than standard gamuts is somewhat random (in relation to the original compression, that is), and in fact it actually imparts color differences that did not previously exist. Using an expanded gamut insures that EVERY color reproduced is now incorrect (except gray), with the colors closest to gray being the least affected and the colors closest to the extreme ends of the gamut (red, green, and blue) being the most affected. But make no mistake about it, when you watch an uncorrected RS-1, EVERY color is wrong, not just the extreme greens, reds, and blues. It is just that the extreme RGBs are the easiest for people to see.

I am sceptical of the arguments made about rigidly adhering to color standards.

The most common argument made in support of existing standards is that, when adhering to then, we view material "as it was intended to be seen." The implication here is that a film's director, cinematographer and producer had something to do with that DVD or bluray disk that we are playing at home. This is not true. The quality of film transfers varies greatly in many ways including color. Plus, the standars are flawed because all the colors in film can never be captured by them! While the director, cinematographer or producer clearly had some intent about how a film would look on the screen, people in the industry that post here claim that that intent rarely makes it into the theater, much less into our homes. Thus, adhering to SMTPE C or REC 709 will NOT enable us to view the film as it was intended.

So, I ask the following question: what benefit are we getting by adhering to these standards? People here admit to playing SMTPE C material in a REC 709 gamut to get better reds. How is a projector manufacturer widening the gamut even further any different? I can't see how it is.

Manufacturers should offer presets that have many different color options. We are getting imperfectly mastered material. If someone thinks it looks better with a wider gamut, who is anyone else to argue with that preference.

It is small-minded in the worst sort of way to declare that color performance is objectively measured when the material that you are viewing is imperfectly mastered. I, for one, would rather watch colors that look good than colors that are accurate but bad looking.

The reason to adhere to standards is to have a reference and minimize variance in the reproduction chain. Once you have a display that is on standard, and you understand the likely variances in the production and distribution, you can better deal with them. Having the display closer to correct makes the worst variations less objectionable.

The issue of standards for primaries is simply overstated. There is many times the difference in an uncalibrated or poorly designed display than the difference between Rec 709 and SMPTE RP 145 primaries. The big variances come in saturation in sources, and matrix selection.

Why adhere to the standards? Because they look better most of the time. Even rather naive consumers can see this. A common question that I ask clients who have no understanding of color science and calibration is to look at the next sports even that they watch. Do the Gator orange and blue and the grass on the field look like they do when you go to the Swamp for the games? Does the orange look orange or more pink than it should. Can you tell the difference between LSU purple and Gator blue? Between Miami aqua and Jacksonville teal? The answer is usually a light going off and they get it. After calibration they are almost always happy to see things more as they would in real life.

Does that mean that the display is then perfectly natural and that every broadcast and disc is correct? Absolutely not! We still see cameras that are clearly not balanced properly, SD that does not get upconverted and corrected to the proper HD coefficients, and multiple sources that do not match. This is where the user's judgement and tolerance come in. An educated consumer with a well calibrated and well designed set will understand what he is seeing and whether to make further correction or not. He can make informed decisions. Anyone who ignores the variance between sources and the ability of the user to adjust or not as he prefers is simply being idealistic at the expence of enjoyment of the system.

Arguing that we should not adhere to standards is, however, far worse IMO than the previously mentioned idealism. Going there is simply chaos. I see a clear pattern of frustration with that chaos that is relieved to a large degree with the knowledge, understanding, and expanded toolkit that we give our calibration clients to provide for improved viewing experiences.

I am sceptical of the arguments made about rigidly adhering to color standards.

The most common argument made in support of existing standards is that, when adhering to then, we view material "as it was intended to be seen." The implication here is that a film's director, cinematographer and producer had something to do with that DVD or bluray disk that we are playing at home. This is not true. The quality of film transfers varies greatly in many ways including color. Plus, the standars are flawed because all the colors in film can never be captured by them! While the director, cinematographer or producer clearly had some intent about how a film would look on the screen, people in the industry that post here claim that that intent rarely makes it into the theater, much less into our homes. Thus, adhering to SMTPE C or REC 709 will NOT enable us to view the film as it was intended.

So, I ask the following question: what benefit are we getting by adhering to these standards? People here admit to playing SMTPE C material in a REC 709 gamut to get better reds. How is a projector manufacturer widening the gamut even further any different? I can't see how it is.

Manufacturers should offer presets that have many different color options. We are getting imperfectly mastered material. If someone thinks it looks better with a wider gamut, who is anyone else to argue with that preference.

It is small-minded in the worst sort of way to declare that color performance is objectively measured when the material that you are viewing is imperfectly mastered. I, for one, would rather watch colors that look good than colors that are accurate but bad looking.

Oh my ! Just because there are imperfections in the chain doesn't mean an added layer will improve things. The other night I saw an RS2 where a Ferrari looked almost purple ! I saw a guys shirt that was so oversaturated that it looked like it was back lit.

I'm not saying that SMPTE is perfect, in fact, I believe that it is not, but it is an excellent place to start especially when those are the colors used on the monitors that the film industry employs.

Of course ,saying that is right or rigidity to those standards is correct is not what I'm saying but throwing out the baby with the bath water is far worse IMO.

Art

Oh my ! Just because there are imperfections in the chain doesn't mean an added layer will improve things.
Art
Excellent! If we believe in deviation based on personal taste/judjement where are we going to stop. It would be an endless battle. :)

I see one difference. Real life can have huge CRs (I can get over a million to 1 just by going outside and then going inside and closing the door in a dark room or walk-in closet). But colors are being artificially limited from what real life contains in order to save money, limitations of the technology, etc.

I agree, and this may have something to say in the fact that most consumers don't understand the concept of accurate color. It's easy to understand the concept of "the higher the contrast number is, the better". Not having a number that should just be as high as possible, seems to bother some people. For some reason "more is better" is easier to comprehend, than "more accurate is better". I guess the argument here is that "accurate" contrast ratio equals "infinity", so the closer you get to "infinity", the more accurate it gets.

What bothers me is that there seems to be a lot of people who think it's extremely important to display the resolution "accurately" (as in 1080P without scaling), but still don't agree that it's important to display color accurately. Resolution is perhaps more equal to color than contrast - it is limited by the system, and it's (in theory at least) more important to match the system specs than to have resolution "as high as possible".

When switching between Rec. 709 and HD1 primaries the difference depends on the source and goes from very obvious to rather small. When I start to dislike skin colors it's time to switch to Rec. 709. But some films are graded such that also with Rec. 709 the skin looks yellowish or greenish or whatever. And with HD1 colors it looks then 'sick'.
But when I watch "Cars" I prefer the cartoon colors of the HD1. And for some desaturated films as well.
The whole color gamut issue is for me not really relevant outside of skin color and some nature shots or other familiar subjects unless I know exactly what the master looks like (I usually don't). If a sweater has this red or that red etc. does not make or break a film for me. That stuff is manipulated so much with (digital) grading anyway that one quickly learns to accept that there is no fundamental right or wrong here, only choices made by whoever grades this and gets it approved. I don't like general tints in a film that should not be there, but that is not the same as more saturated primaries. Basically ignorance is bliss.


While I respect the right to have whatever opinion on this you want, I think the "ignorance is bliss" statement is a knife in the heart of the film makers. Yes, the color of a shirt will be different in the film than it was in real life. But that's part of film making! Just because that shirt doesn't exist in real life, doesn't change the fact that the film maker had a reason for making the shirt a certain color. And I certainly don't like your take on color grading, grading the film is part of the film making, and a very important one in my opinion. Good or bad grading can seriously impact the experience of watching a movie.

Basically, I believe your opinion on colors suggest that you watch the image quality for the sake of the image quality itself, and not for the sake of the movies. Stop watching image quality, and start watching movie making. Stop watching differences in the projector's color gamut, and start watching differences in color grading of the movies. To do this, you will need a display that is accurate, to be able to properly see the differences. If you do this, I believe you will get better experiences out of the movies, which is what it's all about anyway.

The sooner we have cinema color gamut and sources the better, but till then what we see is often not what the (film) original looks like, whether we use SMPTE-C, Rec 709 or HD1 primaries.

That's not the point. In theory at least, the film maker should have monitored the change from cinema to HD, and made sure that within the scope of HD, the movie still looks "right" (from a movie making perspective). Recreating the cinema look isn't the goal of watching HD, when we're watching HD, we have created a new reference, and the movie maker will - some more than others - have done whatever he can to make the movie look the most like cinema, within the limits of HD. The _only_ way the movie maker could make use of a wider gamut of your display, is to deliberately grade the HD version of the film such that it will look right on a display with a wide gamut. This obviously is a path we don't want to go down, as an HD1 and an HD100 doesn't have the same kind of wide gamut, for instance.

The most common argument made in support of existing standards is that, when adhering to then, we view material "as it was intended to be seen." The implication here is that a film's director, cinematographer and producer had something to do with that DVD or bluray disk that we are playing at home. This is not true. The quality of film transfers varies greatly in many ways including color. .

I agree, but why make it worse? Adhering to standards doesn't mean that every movie looks perfect. It means that every movie looks _as good as it can_. There's a big difference in those two statements.



So, I ask the following question: what benefit are we getting by adhering to these standards? People here admit to playing SMTPE C material in a REC 709 gamut to get better reds. How is a projector manufacturer widening the gamut even further any different? I can't see how it is.

See my previous statement. The monitor with SMPTE C phosphors is a working tool. The "accurate" representation isn't nescessarily identical to what the grader sees on his monitor, if the grader deliberately undersaturates the color a little bit, because he knows that if he does this, it will look better on actual displays.

I think we all agree that the quality of movies vary quite a bit, and that there isn't a single setup that will show every single movie with perfect colors - whichever way you like to see your movies, the differences between each film will still be there. Widening the gamut doesn't change the fact that some look better than others.

One more point, that may be quite important here: People deliberately using HD gamut even though they know the monitor in the studios used SMPTE-C, are _NOT_ deliberately deviating from the standards. If you watch HD material, HD gamut IS the standard. The _studio_ is deviating from the standards. Did the studio take this difference into account when grading the movie? We don't know. So, we cannot adjust for the limitations the studio may have been introducing. We can only try to not introduce any further limitations - and that is actually done by using HD gamut. Personally, I don't care what monitor the studio uses. They put out a product that is spec'ed for HD gamut, so that's what I will use. If it looks wrong, then their product is wrong, not my monitor.

A question for those who change gamut based on how you believe the film was graded: How often do you watch a film with your PJ set for SMPTE-C gamut, and actually think the film looks wrong, vs. how often you watch a film with the pj at HD gamut, and actually think the film looks wrong? Honestly, did you change your set to SMPTE-C because it looked wrong with HD gamut, or because you read on AVS that the monitors in the studios use SMPTE-C?

Manufacturers should offer presets that have many different color options. We are getting imperfectly mastered material. .[/QUOTE]

NO! Manufacturers should offer a preset that is _accurate_. If they did this, maybe the studio's who don't master their products perfectly, will step up. As the market is now, why bother with grading the HD version properly, if noone watches it the way we graded it to look anyway? We can _only_ get better masters by adhering to the standards. Deliberately deviating from the standards will only make the problem worse.

If someone thinks it looks better with a wider gamut, who is anyone else to argue with that preference.

This goes both ways! If I think it looks better with an accurate representation, why should you decide that this shouldn't be possible on my display?

Anyone can do whatever he wants. However, anyone should have the _option_ of watching the films accurately, so that he can choose for himself if he wants to change that look. How do you know if you prefer a different look than the original, if you have no way of watching the original look? I have _no_ problem with a wide color gamut option of a PJ, as long as I have a button to push, to make the colors look at least somewhat accurate. And I don't believe any of the pro-accurate people will disagree with this. However, since the accurate representation is what the movies are designed for, this _should_ be a standard choice for anyone, which one could then deviate from as he chooses, based on personal preferences. It would be a benefit to everyone, even the one preferring deliberately wrong colors, to have the option of seeing what the source actually looks like. That would only make it _easier_ for you to change the color of your set to make it look the way YOU want.



It is small-minded in the worst sort of way to declare that color performance is objectively measured when the material that you are viewing is imperfectly mastered.

But how do you know if the material is imperfecly mastered, if you're not watching it accurately??

Does anyone disagree with this statement: Whichever way you like to display your movies, you will never be able to get a better quality than the source has to offer.

Using an expanded gamut insures that EVERY color reproduced is now incorrect (except gray), with the colors closest to gray being the least affected and the colors closest to the extreme ends of the gamut (red, green, and blue) being the most affected. But make no mistake about it, when you watch an uncorrected RS-1, EVERY color is wrong, not just the extreme greens, reds, and blues. It is just that the extreme RGBs are the easiest for people to see.
That is technically right, but refers to the compromised Rec 709 version. I'm much more interested in comparing the original version with cinema colors to the Rec. 709 version and the Rec. 709 version displayed with other primaries (such as on the HD1). I'm aware that using more saturated primaries does not bring back the cinema colors per se but at the same time I have trouble believing everything now just got worse than compared to the Rec. 709 primaries. Has anybody actually analysed representative film material and common ways to create the Rec. 709 version and numerically compared the perceptual color differences between the original and Rec. 709 with HD primaries and for example HD1 primaries? What are the results? Has anybody looked into alternative ways to display the Rec 709 master with oversaturated primaries instead of the normal linear addition of primaries? For example one could non linearly interpolate between Rec 709 primaries and HD1 primaries depending of how far a color is perceptually away from the white point. So skin color would be far less influenced by the oversaturated primaries than highly saturated colors. How would that affect average color differences compared to the cinema master? Any studies on this?

I'm aware that using more saturated primaries does not bring back the cinema colors per se but at the same time I have trouble believing everything now just got worse than compared to the Rec. 709 primaries.
I don't have time to answer this right now (I am on my way to work), but it can be proven beyond doubt mathematically. Every color in the entire REC 709 has a point on the x,y axes, and if you compare the correct values to those from the RS-1's gamut you will find that EVERY single point is different except those that fall on gray (from 0 to 100 IRE, assuming that the user has taken the time to precisely calibrate his grayscale).

While I respect the right to have whatever opinion on this you want, I think the "ignorance is bliss" statement is a knife in the heart of the film makers.

That was not the intended message...

Yes, the color of a shirt will be different in the film than it was in real life. But that's part of film making! Just because that shirt doesn't exist in real life, doesn't change the fact that the film maker had a reason for making the shirt a certain color. And I certainly don't like your take on color grading, grading the film is part of the film making, and a very important one in my opinion. Good or bad grading can seriously impact the experience of watching a movie.

That was not my point. My point was that what is in front of the lens and what is coming out at the end of grading is only losely connected since the world is basically recreated in the grading phase. So I can not look at a picture and say: Hey, in real life this car model has this shade of red but here on my monitor it does not, so my display/disk etc. is screwed. My enjoyment of the film is ruined. I have no absolute reference for this. All I have is my own perception and the knowledge that IF the disk was mastered to follow the standard AND my display is following the standard and IF the film makers supervised the HD master I see an approved approximation of the original. I still don't see the original.
Or to put it differently, I think adhering to standards is necessary and has its place but it's not the end of all if the standard in itself is incapable of giving me the original colors. It's ok to try out different things and chose parameters that work better for oneself depending on what one is watching. After all, no harm is done to the source and nobody else is forced to watch it this way. That's a lot more than I can say of sharpening and noise reduction ON THE SOURCE forced down everybody's collective throat.

Basically, I believe your opinion on colors suggest that you watch the image quality for the sake of the image quality itself, and not for the sake of the movies.

No. I'm saying I'm not interested in being holier than the pope when I'm starting with a flawed standard. I think it's ok to try out different things at home in the color department to find out what works best for oneself. That's all.

I don't have time to answer this right now (I am on my way to work), but it can be proven beyond doubt mathematically. Every color in the entire REC 709 has a point on the x,y axes, and if you compare the correct values to those from the RS-1's gamut you will find that EVERY single point is different except those that fall on gray (from 0 to 100 IRE, assuming that the user has taken the time to precisely calibrate his grayscale).
It is different from the corresponding Rec. 709 color. It is not necessarily different from the original cinema color. For example if cinema colors were transformed to Rec. 709 colors in such a way that the primaries were made to correspond and all other colors follow accordingly the transformation would be an invertible function (assuming no precision issues with bit depth). And if you had cinema color primaries at home you could in theory recreate them from the Rec 709 HD master. I know colors are not transformed like this. But still, they way colors are transformed, how far away are the Rec 709 colors from the original on average and how far away are for example HD1 colors using the HD master away on average, using linear interpolation, optimised non linear interpolation? Any studies done?

But make no mistake about it, when you watch an uncorrected RS-1, EVERY color is wrong, not just the extreme greens, reds, and blues.

That is probably correct Bob, but I would argue that defects in the mastering process leads to the same result when you compare the dvd/bluray version with the film version. Both are incorrect as compared to the original!

Why adhere to the standards? Because they look better most of the time.

Are you not now using a subjective criteria to justify an objective standard?

I agree that many people feel this way and their opinions are valid. However, those who prefer oversaturated colors (and I would argue that those of you watching SMPTE C material in a REC 709 gamut are among them) are, like you, just trying to get achieve an image that looks good to them.

Oh my ! Just because there are imperfections in the chain doesn't mean an added layer will improve things. The other night I saw an RS2 where a Ferrari looked almost purple ! I saw a guys shirt that was so oversaturated that it looked like it was back lit.

I'm not saying that SMPTE is perfect, in fact, I believe that it is not, but it is an excellent place to start especially when those are the colors used on the monitors that the film industry employs.

Of course ,saying that is right or rigidity to those standards is correct is not what I'm saying but throwing out the baby with the bath water is far worse IMO.

Art

I don't disagree with you, Art. Colors that stand out to the eye as wrong harm, don't help image quality.

I agree, but why make it worse? Adhering to standards doesn't mean that every movie looks perfect. It means that every movie looks _as good as it can_. There's a big difference in those two statements.

I don't agree. I believe that the standards are a form of shield against imperfect mastering. Adhering to them means that you will never see a color that stands out because, by definition, NO COLOR CAN STAND OUT. You are simply toning down your color palette. Does this lead to better image quality? Sometimes, but certainly not always.

You can accept this as an imperfect compromise and you might even have the better of the argument. But, let's not try to pretend that this is objectively the right way to do things.
My solution to this dilema is, as I previously wrote, having manufacturers include different gamut choices. Let the viewer decide which looks best on a given source. Hopefully one day we will have material mastered in larger gamuts, the mastering process improves, and we can really display things as they were intended to be in our homes.

I don't agree. I believe that the standards are a form of shield against imperfect mastering. Adhering to them means that you will never see a color that stands out because, by definition, NO COLOR CAN STAND OUT. You are simply toning down your color palette. Does this lead to better image quality? Sometimes, but certainly not always.

You can accept this as an imperfect compromise and you might even have the better of the argument. But, let's not try to pretend that this is objectively the right way to do things.
My solution to this dilema is, as I previously wrote, having manufacturers include different gamut choices. Let the viewer decide which looks best on a given source. Hopefully one day we will have material mastered in larger gamuts, the mastering process improves, and we can really display things as they were intended to be in our homes.

I think the main point everyone is making is the adhering to the standards is good because it provides the same starting point to everyone. Some may find it looks best this way, others may find it doesn't. Beauty is in the eye of the beholder. This much is certain - it will appear as the director intended (assuming he had control of the entire mastering process). If a movie looks 'ugly' even with accurate colors, maybe it was meant to be ugly - not everyone is a supermodel.

Example: Two movies from the same director (Steven Spielberg): Saving Private Ryan (SPR) and Jurassic Park (JP). Two very different movies with very different looks. SPR is a gritty war drama and has a very muted/desaturated palette - everything looked pretty drab. JP is a fantasy/adventure in a lush jungle - very vivid/saturated palette. With a color accurate display, SPR will look drab and JP will look very colorful. With an oversatured display, SPR will look 'colorful' and JP will look very, very colorful. Maybe a lot of people will prefer a bright and colorful SPR, but they should be able to at least see the accurate version first to provide a starting point.

Expanded gamut is a good thing provided a CMS (or user selectable presets) is available. The viewer can than be free to be their own 'director' and change the color as they see fit while still having the ability to view the default correct version (the creator of the original material gets to define what is correct).

Essentially, I think what you're saying is choice is always better than no choice, and I agree. But I also believe that there is a correct choice, and that there's nothing wrong with prefering something incorrect. It's like OAR vs pan-and-scan.

This much is certain - it will appear as the director intended (assuming he had control of the entire mastering process).

False! If anything is apparent, it is that directors can't even control how their films appear in theaters much less home theaters.

If this were true, it would change alot, but sadly, it appears not to be.

Also, saving private Ryan does not appear colorful on an uncorrected RS1.

You are right that I think that choices should be made available.

False! If anything is apparent, it is that directors can't even control how their films appear in theaters much less home theaters.


What's false? I'm simply saying that if the display is accurate, then you'll be able to see what the director intended. Of course the director has no control over how his material is displayed, but that doesn't change the goal of accurately displaying what he wanted. All he can do is provide the source. The burden of finding an accurate display falls on us the viewer. That could mean finding a theater that's calibrated per the standards, or buying (and calibrating to standards) an accurate display for our personal home theater.

What's false? I'm simply saying that if the display is accurate, then you'll be able to see what the director intended. Of course the director has no control over how his material is displayed, but that doesn't change the goal of accurately displaying what he wanted. All he can do is provide the source. The burden of finding an accurate display falls on us the viewer. That could mean finding a theater that's calibrated per the standards, or buying (and calibrating to standards) an accurate display for our personal home theater.

I think that perhaps you don't understand what I am writing.

The point is that the intention of the director is lost between the time he/she releases the studio master and the time the film makes its way onto dvd/bluray. We are all dealing with flawed copies. You can choose to display that flawed copy as it was intended, but that says nothing about the director's intention.

Any studies on this?
Genoa (http://www.genoacolor.com/technology.asp?pageId=81) has done some studies along those lines.

I'm just curious, how many people reading and/or posting in this thread who are defending the use of a wider than standard gamut own ANY other projector than the RS-1/RS-2?

I'm just curious, how many people reading and/or posting in this thread who are defending the use of a wider than standard gamut own ANY other projector than the RS-1/RS-2?

Bob, I'm surprised. Let's attack the message, not the messenger.

...The implication here is that a film's director, cinematographer and producer had something to do with that DVD or bluray disk that we are playing at home. This is not true...I am told that in a lot of cases the Director/DP is directly involved/supervises the telecine process. Peter Jackson (LOTR) had contractual requirements that he produced the DVD versions. Francis Ford Copolla's contract stipulated that his company produced the Dracula Blu Ray Master to ensure that his vision was correctly recreated in that Media.

In those cases the only way to see what the Director created is to have an accurate display.

I am told that in a lot of cases the Director/DP is directly involved/supervises the telecine process. Peter Jackson (LOTR) had contractual requirements that he produced the DVD versions. Francis Ford Copolla's contract stipulated that his company produced the Dracula Blu Ray Master to ensure that his vision was correctly recreated in that Media.

In those cases the only way to see what the Director created is to have an accurate display.


If you have a source for those assertions I would love to see it.

The forum has recently been frequented by an industry source that has told a very different tale of how these things are done.

I agree that if what you are saying is true than those films would look most accurate when viewed in whatever gamut they were mastered in.

I think that perhaps you don't understand what I am writing.

Agreed :)


The point is that the intention of the director is lost between the time he/she releases the studio master and the time the film makes its way onto dvd/bluray. We are all dealing with flawed copies. You can choose to display that flawed copy as it was intended, but that says nothing about the director's intention.

Case 1: Ideally, the director's intention is preserved all the way to the final media. If so, then the only way I can see for certain what the director intended is if I have an accurate display.

Case 2: Everything goes out the window by the time it reaches the final media. The only way to see what the director intended is if my display exactly offsets the flaws introduced during the mastering process.

I really don't know what happens between the studio master and release on dvd/bluray. Therefore, I will optimistically (and perhaps naively) assume that case 1 is the more likely scenario and represents the best chance for me to see what the director intended*.

Perhaps some insiders can comment in general how much the final dvd/bluray releases deviate from the director's intent.

I have the feeling that I'm still not understanding your point. Maybe I'm actually agreeing with you :). Bottom line - I think standards are useful and important to provide the option to see the director's intent. There maybe a weak link somewhere else in the chain that mangles the directors intent. The solution is to fix the broken link, not change a different link to compensate.

*I expect the chances of this are improved when the movie is an important (blockbuster) release. The studio/director will pay more attention to quality control with big $$$ on the line.

Bob, I'm surprised. Let's attack the message, not the messenger.

Bob's question might give a little insight into possible bias.

Art

Bob, I'm surprised. Let's attack the message, not the messenger.
I am not attacking anyone, but rather pointing out that the issue of color accuracy was never challenged in this forum AFAIK prior to the introduction of the RS-1. The RS-1 is the poster child for inaccurate color, and being as popular as it is, has generated debate on both sides of this issue. Prior to the RS-1, no one seemed to care all that much (though there was some criticism of the Sony Ruby's colors by people like me :) ), and due to the extreme nature of the RS-1's oversaturated primaries, people have found all kinds of reasons to support inaccurate color. I just wonder how many people who don't own the RS-1 or RS-2 feel as strongly as RS-1/RS-2 owners do. Personally speaking, the color performance of the RS-1 (when I owned it) certainly brought color accuracy to the forefront of my attention and has made me painfully aware that not all display manufacturers feel the same as I, but I was truly surprised at how many people are not bothered by, or actually enjoy, the JVC colors.

Most projectors don't have accurate color. No Marantz models do, for instance. Sony's didn't until recently. Others are perhaps closer, but still inaccurate. Of the major manufacturers, only really Infocus tried to adhere to standards (look where that got them).

The point is that most projector owners have been blissfully living with juiced colors for a long time. Only recently has it become an issue. The most likely reason is that the RS1 became tremendously popular and that has caused a backlash.

So, here we are discussing the RS1 again, when I thought that the point of this thread was to talk about color accuracy and whether it is important or not important. I didn't start this thread.

All of you who admit here that you watch SMPTE C material with a REC 709 gamut are trying to have it both ways. How is it not wrong to watch material on the WRONG gamut? Does it make you feel better that you are adhering to a standard, albeit the wrong one for the material that you are watching?

I don't have time to answer this right now (I am on my way to work), but it can be proven beyond doubt mathematically. Every color in the entire REC 709 has a point on the x,y axes, and if you compare the correct values to those from the RS-1's gamut you will find that EVERY single point is different except those that fall on gray (from 0 to 100 IRE, assuming that the user has taken the time to precisely calibrate his grayscale).This is also true for stuff mastered on a monitor with SMPTE-C primaries, but displayed with Rec. 709. If you are going to watch material that you know was mastered on a monitor with SMPTE-C primaries and that this artificially limited the color more compared to the film version than would have been the case if the mastering house had bought an HD monitor with Rec. 709 primaries, which set of primaries would you use to watch the material? Of course assuming you could choose the primaries?

What if material that was mastered with a monitor with SMPTE-C primaries looked better to your eyes using Rec. 709 primaries to play it instead of SMPTE-C primaries? Would you watch it with SMPTE-C primaries?

At least one person here who is on the side of sticking strictly to standards has said they use Rec. 709 even for material that was mastered with SMPTE-C and I'm wondering if you (Bob) limit yourself to the reds of SMPTE-C because the mastering houses mostly have decided to save money by not moving forward to mastering monitors with Rec. 709 primaries.

As far as what people own (I have an RS1), of course that decision is based partially on our opinions about color gamuts for many of us. If I demanded orangish reds without the ability to have even Rec. 709 reds I would value a display with more orangish reds (like SMPTE-C) for that. I personally like the choices the Planar provides, but would probably use Rec. 709 for playback of HD a whole lot more than SMPTE-C. And for Cars I would probably go for Native, at least on the 8150 I saw, where red in Native was even a little more red.

IMO, Rec. 709 is a reasonable compromise. But I also don't fool myself into thinking it isn't somewhat arbitrary to use Rec. 709 for primaries when the source wasn't done with Rec. 709 primaries.

--Darin

Bob's question might give a little insight into possible bias.

Art


I admit that I am biased.

I am biased in favor of allowing people to decide for themselves how things should look in their home theater.

I am a fan of the RS1, but not of its color. I had always dialed down the color control so that I would not see those offensive greens and yellows. I recently bought a Lumagen HDQ and have been playing with the gamut control to try to improve the RS1's color. None of this is any secret. I believe that I have been very open about it.

Since the beginning I have stated that manufacturers should have lots of presets for different gamuts. This way, the viewer has a choice about what looks best.

This approach is for some reason, not acceptable to some among us. They believe that there is only one way for color to look (unless they happen to make the decision to use a different color gamut than the one the material is mastered in). I am not sure why some people find the concept of freedom of choice so disturbing.

No Marantz models do, for instance.
Agreed. What's your point? Even though the Marantz models are nowhere near as oversaturated as the JVC designs, I would not buy one unless Marantz decided to provide accurate color, a CMS, or an accurate preset.
Others are perhaps closer, but still inaccurate.
Many companies have found it a simple task to supply an accurate preset so that the user can have the option of watching either accurate or juiced colors as he wishes.
Of the major manufacturers, only really Infocus tried to adhere to standards (look where that got them).
Are you suggesting that accurate color is what caused IF's financial woes? :rolleyes: C'mon, it is well known that IF's woes started in the business projector department where the lion's share of their sales were generated.
The point is that most projector owners have been blissfully living with juiced colors for a long time.
And Sony first and then JVC second saw fit to juice them up far more than ever. This is something I would term "color de-evolution".
So, here we are discussing the RS1 again, when I thought that the point of this thread was to talk about color accuracy and whether it is important or not important. I didn't start this thread.
It was not my intention to bash the RS-1 - I simply wanted (and still want) to know how many people support expanded gamuts other than RS-1/2 owners, that's all. We can stop the RS-1/2 talk right here and now...no problem.
Does it make you feel better that you are adhering to a standard, albeit the wrong one for the material that you are watching?
You are comparing differences of completely different magnitudes of error. The difference between REC 601 and REC 709 is tiny compared to the differences introduced by overly large gamuts created by some display manufacturers - not even in the same league, but if that's the best you got, then so be it.
If you are going to watch material that you know was mastered on a monitor with SMPTE-C primaries and that this artificially limited the color more compared to the film version than would have been the case if the mastering house had bought an HD monitor with Rec. 709 primaries, which set of primaries would you use to watch the material? Of course assuming you could choose the primaries?
Yes, I understand the dilemma, and personally I choose REC 709 under those circumstances, but my point was simply that the magnitude of error between REC 709 and REC 601 is far smaller than that of either 601 or 709 and the expanded gamuts being used in some display devices. So to that end I watch inaccurate colors (well maybe, as I don't know which ones were mastered using REC 601 primaries and which ones were mastered using REC 709), but the inaccuracies generated by some displays are just too much for me to tolerate. I guess we all have our breaking points, and though I was able to somewhat live with the Ruby's expanded gamut (in the normal gamut setting mode), the RS-1 was the straw that broke the camel's back for me.
This approach is for some reason, not acceptable to some among us. They believe that there is only one way for color to look (unless they happen to make the decision to use a different color gamut than the one the material is mastered in). I am not sure why some people find the concept of freedom of choice so disturbing.
I'm all in favor of freedom of choice. I just have to wonder why JVC took it away from us by NOT providing an accurate preset. This would have been a simple matter (if not in the RS-1, then certainly in the RS-2) and would have satisfied all of us. After Greg's review of the RS-2, I would have gone back and bought one if I didn't need to buy a $4k VP to repair the color.

And BTW, Lawguy, I really wish you had the opportunity to do an instant A/B of an uncorrected RS-1 to a corrected one for a week or two...You just might change your mind...;)

I'm all in favor of freedom of choice. I just have to wonder why JVC took it away from us by NOT providing an accurate preset. This would have been a simple matter (if not in the RS-1, then certainly in the RS-2) and would have satisfied all of us.As I understand it, it wouldn't have been a simple matter because it would have required changing hardware. Like adding a chip to do the processing, since there wasn't enough processing power in the chips in there. I would also like to see them give us more choices without having to add an external processor, but I'm not sure if we'll see it even in whatever their next model is, just because of the issue of needing more processing power internally to do it.

--Darin

And BTW, Lawguy, I really wish you had the opportunity to do an instant A/B of an uncorrected RS-1 to a corrected one for a week or two...You just might change your mind...;)


I lived with and loved my Sharp 12k, which had color that is said to be close to some standard. I have nothing against accurate color. I just don't like it when some among us (not you) repeatedly scold others for prefering something else.

I think most of us would agree that it is OK to deviate from these standards so long as it makes a subjectively positive improvement. For you (and for me) the RS1 goes too far. I fixed that issue for me by turning down the color control. That didn't work for you. Both of us agree that we all should have choices.

Like Darin, I really like that new Planar and the choices that it gives. This should be the model for all manufacturers to follow.

. . . I'm not sure if we'll see it even in whatever their next model is, just because of the issue of needing more processing power internally to do it.

--Darin

I bet that you are right even though I can't imagine why it should be so. Many less expensive projectors have full CMSs at this point. Others have lots of different generally-accurate presets.

My money is on them pushing and/or bundling that RSVP2 (which apparently only has SMPTE-C color space).

OK, I would put to some of you who like the RS2 colors or any projector with significantly inaccurate colors,do you not feel that some sort of reasonable standardization is positive ?

Art

I'm all in favor of freedom of choice. I just have to wonder why JVC took it away from us by NOT providing an accurate preset. This would have been a simple matter (if not in the RS-1, then certainly in the RS-2) and would have satisfied all of us. After Greg's review of the RS-2, I would have gone back and bought one if I didn't need to buy a $4k VP to repair the color.

JVC's current line up includes four models two with accurate color gamuts, they are:

Model MSRP
DLA-RS1X $5,995
DLA-RS1X/RSVP2 $9,690
DLA-RS2U $7,995
DLA-RS2U/RSVP2 $11,690

These prices compare favorably to competing Marantz, Sim, and Projection Design models. Unfortunately, JVC made two marketing mistakes, one offering consumers a choice by separating the VP and the projector, and not coming out with the projector and VP bundle first, forcing early adopters to cough up the extra cash and maximizing their profits.

I admit that I am biased.

I am biased in favor of allowing people to decide for themselves how things should look in their home theater.

I am a fan of the RS1, but not of its color. I had always dialed down the color control so that I would not see those offensive greens and yellows. I recently bought a Lumagen HDQ and have been playing with the gamut control to try to improve the RS1's color. None of this is any secret. I believe that I have been very open about it.

Since the beginning I have stated that manufacturers should have lots of presets for different gamuts. This way, the viewer has a choice about what looks best.

This approach is for some reason, not acceptable to some among us. They believe that there is only one way for color to look (unless they happen to make the decision to use a different color gamut than the one the material is mastered in). I am not sure why some people find the concept of freedom of choice so disturbing.

I'm not against choice or choosing. I'd be more against not being able to choose and that choice being so visibly wacked.

Art

I'm not against choice or choosing. I'd be more against not being able to choose and that choice being so visibly wacked.

Art

I can't speak about the RS2, but on the RS1, there was a pretty simple solution: just turn down the color control.

Although many people did this and were satisfied with the results (me included), we were continuously scolded and told that we had made the colors worse, even though they appeared better to all of our eyes.

This was not an ideal solution, but it was a compromise that many made and happily live with.

OK, I would put to some of you who like the RS2 colors or any projector with significantly inaccurate colors,do you not feel that some sort of reasonable standardization is positive ?

Art

Ideally, films would be mastered for the home in a wide enough gamut and the intent of the filmaker would be captured in the mastering process.

Why would any projector manufacturer sell a projector that did not conform to that spec? Color could be as muted or saturated or oversaturated as Best Buy wanted for display purposes and still please those who wanted true color accuracy.

As I understand it, it wouldn't have been a simple matter because it would have required changing hardware. Like adding a chip to do the processing, since there wasn't enough processing power in the chips in there.
Darin, I can understand (sort of) why JVC might not have been able to add the preset with the RS-1, but certainly 1 year later it should not have been an issue with the RS-2, don't you think? Instead of adding a color correct preset, JVC chose to alter their current preset to be even more oversaturated. That decision told me all that I needed to know.

I can't speak about the RS2, but on the RS1, there was a pretty simple solution: just turn down the color control.

Although many people did this and were satisfied with the results (me included), we were continuously scolded and told that we had made the colors worse, even though they appeared better to all of our eyes.


The colors looked better because, overall, using perceptual measures they were better. It just took a few a while to accept that fact. Cine4home also began shipping their calibrated units with the color control set to -10.

I would like to comment on the belief Infocus has always had near perfect color. I always chuckle when proponents of accurate color bring up Infocus as an example for others to follow. Perhaps this is true with their latest offering but the 4805, IN72, IN76 which were always touted as having accurate color were far from it. Using perceptual measures the 4805 (72/76 similar) measures worse than the JVC. The Infocus green is more undersaturated from rec.601 than the RS1's green is oversaturated from rec.709. BUT very rarely were the greens complained about and often used as a poster child for accurate color.

Measured and corresponding dE do not take into account how these color differences manifest themselves with regard to how offensive these may be perceived, they only offer degrees to which color changes are noticeable. So why do I chuckle, proponents of color who use Infocus as an example of great color are using subjective preferences for color evaluation at a level equal to those they believe they share little with it's just that the same level of error appears more correct to them.

Moral of story, every-one's full of crap :D, just kidding. As long as there's a strong correlation between oversaturated color and sales revenue (or some other profit generating feature combined with oversaturated color) and the available "affordable" fixes are exercises in compromise greater study with color errors would be helpful. Not just when these changes are noticeable but when/where these color errors approach offensive and not in a vacuum but as it relates to video and the reproduction of skin tones, sky, water, blood, trees, grass, bananas etc. I think Infocus green vs JVC green is an example which highlights the possible benefits of such.

Oh, and something general I've been meaning to ask for sometime: Why is Sony marketing nasal spray on this site??

The colors looked better because, overall, using perceptual measures they were better. It just took a few a while to accept that fact. Cine4home also began shipping their calibrated units with the color control set to -10.

Of course they looked better. That is why when people posted that this solution made the situation even worse, it became very frustrating. Perhaps the colors measured worse, but they certainly looked very nice overall, still do.

I did not know those things that you wrote about Infocus. Thanks for sharing.

I also agree that we are all generally full of crap only I'm not kidding.:)

Perhaps the colors measured worse, but they certainly looked very nice overall, still do..

They actually do measure better.

I also agree that we are all generally full of crap only I'm not kidding.:)

You know how I hate to offend anyone.:)

... I really wish you had the opportunity to do an instant A/B of an uncorrected RS-1 to a corrected one for a week or two...You just might change your mind...;)

Bob, At the CES this past Jan I actually did see JVC's demo of their new vp. They had a calibrated studio monitor set up right next to the TV so that one could see that the SMPTE colors coming from the RS2 through the vp were identical to the monitor. And one could simply switch between the SMPTE gamut and the original 'expanded' color gamut.

And ... yes, I liked the SMPTE colors better! They were more subdued and realistic. Though I have been quite pleased with my RS1, including its colors (and I also, like most owners, have the Color control set to -10--and I note that the CES demo did not have it set this way with the 'expanded' gamut), I would certainly like the choice of having various preset gamuts to choose from.

I've been hoping that the next RSx machine would have such presets, so I regret the gist of the thinking I'm hearing here about this.

Darin, I can understand (sort of) why JVC might not have been able to add the preset with the RS-1, but certainly 1 year later it should not have been an issue with the RS-2, don't you think? Instead of adding a color correct preset, JVC chose to alter their current preset to be even more oversaturated. That decision told me all that I needed to know.

I don't think 1 year is a very long time for product development, especially integrating a key component like a processor chip.

I don't know for certain, but I can speculate that the RS2 doesn't have major design differences from the RS1. The RS2 could've achieved improved CR without changing it's electronics, architecture, and optical system by:

- selecting LCOS panels with the highest native CR; the rest go to the RS1x
- having a more agressive fixed iris
- selecting brighter bulbs to compensate for the smaller iris

This sort of mixing/matching of components to create a higher performance product with essentially the same design is a common practice in manufacturing. It takes some time to optimize volume and scrap balance between the higher and lower tier products, which may explain why it took a year for the RS2 to be introduced.

You may still disagree and believe that JVC had all the opportunity to make things the way you'd like but chose not to, but I hope that I've shown you some arguments that there may be alternative explanations.

My solution to this dilema is, as I previously wrote, having manufacturers include different gamut choices. Let the viewer decide which looks best on a given source. Hopefully one day we will have material mastered in larger gamuts, the mastering process improves, and we can really display things as they were intended to be in our homes.

I don't believe the viewer should have to decide anything, I believe the viewer should be able to setup his system once, and then watch whatever he wants without having to change anything at all. And I do believe this is exactly the purpose of the standards.

False! If anything is apparent, it is that directors can't even control how their films appear in theaters much less home theaters.

If this were true, it would change alot, but sadly, it appears not to be.

That's the whole point of the discussion! If you adhere to the standards in your home theater, the directors _can_ control how their films appear in your home theater. And this is the _only_ way directors can ever get this control.

Let's say, for instance, that the HD1 has DCI gamut on purpose. Now the director wants to make use of this expanded gamut, so he grades the HD version of his film so that if you watch it on an HD1 (or any other display that has DCI gamut), it will look right (meaning as close to the original as possible). This does NOT mean that it is better to deviate from the standard - this means that we have now _changed_ the standard, so now the standard involves displaying HD sources with DCI gamut. In that case, my point is still valid: We should adhere to the standards, which now means having a gamut like the HD1 does.

However, I don't believe any director had sone this, and I certainly don't believe that directors should deliberately compensate for the fact that consumers don't have accurate sets. As soon as we go down that path, all hope of watching the movies as close to the original as we can, is lost forever.

The point is that the intention of the director is lost between the time he/she releases the studio master and the time the film makes its way onto dvd/bluray. We are all dealing with flawed copies. You can choose to display that flawed copy as it was intended, but that says nothing about the director's intention.

How do you know which way to set up your display to compensate for these flaws, prior to watching the movie?

This approach is for some reason, not acceptable to some among us. They believe that there is only one way for color to look (unless they happen to make the decision to use a different color gamut than the one the material is mastered in). I am not sure why some people find the concept of freedom of choice so disturbing.

I can only speak for myself, but I am NOT against the freedom of choice, and neither are most of the people speaking out in this forum for accurate color. On the contrary, we _want_ the freedom of choice, to make it possible to choose accurate colors. As many have stated already, the issue we have with the HD1 is not that it has inaccurate color out of the box. It is the fact that we do NOT have the freedom to choose a more accurate gamut. If we want to have the benefits of the HD1 in areas like contrast, we have to live with one set choice of color. How is this any different than if it had one, ACCURATE set choice of color?

Like Bob says: If the HD1 had fairly accurate color out-of-box (or even desaturated color), we wouldn't even have this discussion. I truly believe that anyone trying to justify inaccurate colors, would have been perfectly fine with accurate Rec.709 colors, if that was what the HD1 had delivered. YES, you have the freedom of choice, but if you were given the freedom of actually choosing accurate Rec.709 color, and gave it a shot for, say, a week, I truly believe that's what you would end up with. Every single person I have talked to or discussed with on forums, that wants freedom of choice more than just one, accurate choice, are people who _don't_ have the choice of accurate color. Freedom of choice is fine, but if one of the choices are accurate color, the other choices quite quickly becomes moot.

All of you who admit here that you watch SMPTE C material with a REC 709 gamut are trying to have it both ways. How is it not wrong to watch material on the WRONG gamut? Does it make you feel better that you are adhering to a standard, albeit the wrong one for the material that you are watching?

Rec 709 is NOT the wrong gamut for HD material, even if it was mastered using a SMPTE-C monitor. If there is an issue with this difference, the issue isn't in my display, and yes that makes me feel better. I don't believe I can fix problems that's inherent in the source. If an HD source doesn't look right on a Rec.709 monitor, that's a problem with the source.

Like I said earlier, I believe people put way too much emphasis on the fact that studio monitors use SMPTE-C. I believe the "bottom line" isn't the grading monitor, that's just a working tool. The "bottom line" is the screening cinema of the studio, so what's interesting is which gamut they use in the screening room, not on the color grading monitor. If the movie doesn't look right in the screening room, they will (should) go back to the grading monitor and change it.

In any case, I will NOT be forced to change gamut according to which movie I'm watching, and I don't see that as an advantage in a system. I want to be able to slip a disc in the player, press play, and enjoy the movie, and know that if I'm not watching the movie the way it was intended to look, or at least close to it, I'ts a shortcoming of the source, not the display.

I have nothing against accurate color. I just don't like it when some among us (not you) repeatedly scold others for prefering something else.

I scold others for preferring something else, without knowing what "accurate" looks like. A _lot_ of people do this, and heartily defend their right to display the colors inaccurately, without having ever seen accurate.

That's the whole point of the discussion! If you adhere to the standards in your home theater, the directors _can_ control how their films appear in your home theater. And this is the _only_ way directors can ever get this control.

I feel like I keep repeating myself about this one point: in reality (as opposed to theory) adhering to standards does not mean that you are watching a film as the director intended. I am sure that there are exceptions, and I am sure that some DVDs or Blurray disks do accurately capture the filmmaker's intent, but as a general rule, the moment the filmmaker releases the studio master, the filmmaker loses control of it.

This fact is so generally accepted that magazines like WSR, when they review DVDs or Bluray, typically comment on the quality of the transfer. There are numerous threads on these boards about which releases qualify as reference quality and which do not.

I hate to be the one to break the news but the world is not perfect and neither are the sources that we watch.

Freedom of choice is fine, but if one of the choices are accurate color, the other choices quite quickly becomes moot.


No argument from me on this. Manufacturers should give people the ability to choose.

Like I said earlier, I believe people put way too much emphasis on the fact that studio monitors use SMPTE-C. I believe the "bottom line" isn't the grading monitor, that's just a working tool. The "bottom line" is the screening cinema of the studio, so what's interesting is which gamut they use in the screening room, not on the color grading monitor. If the movie doesn't look right in the screening room, they will (should) go back to the grading monitor and change it.

You can't really mean this.

If what you are saying is true, you might as well assume that studios understand that most TVs have oversaturated colors so they test the material on oversaturated TV sets and, if it doesn't look right on them, they go back to the SMPTE C monitor and fix it. That would mean that the real intent could only be captured if a dvd was watched on a set with oversaturated colors. I don't think that you mean to suggest that, do you?

I don't believe the viewer should have to decide anything, I believe the viewer should be able to setup his system once, and then watch whatever he wants without having to change anything at all. And I do believe this is exactly the purpose of the standards.

I agree and with as many stylized films that are various colors (extreme examples Underworld blue, Matrix green ,King Kong yellow orange) one would go crazy without standards. I like the idea of being close to the colors that were intended by just turning my projector on putting in a disc and sitting down.

Art

I can't speak about the RS2, but on the RS1, there was a pretty simple solution: just turn down the color control.

Although many people did this and were satisfied with the results (me included), we were continuously scolded and told that we had made the colors worse, even though they appeared better to all of our eyes.

This was not an ideal solution, but it was a compromise that many made and happily live with.

If you look at RS2 color at least it isn't simply saturation issues.

Art

Ideally, films would be mastered for the home in a wide enough gamut and the intent of the filmaker would be captured in the mastering process.

Why would any projector manufacturer sell a projector that did not conform to that spec?

Good question, can you answer that one for me since that is exactly what is happening.

art

If you look at RS2 color at least it isn't simply saturation issues.

Art

What is it? I know that red moved further away from where it should be.

What is it? I know that red moved further away from where it should be.

It's not just saturation but the x/y position of the color is significantly pulled toward purple. Like I said , a Ferrari isn't crimson.

Art

It's not just saturation but the x/y position of the color is significantly pulled toward purple. Like I said , a Ferrari isn't crimson.

Art

When we talk about x/y position we are talking about saturation.

When we talk about x/y position we are talking about saturation.
Not just saturation, but also hue. Two dimensions.

When we talk about x/y position we are talking about saturation.
It depends on where that x,y position is. If it is a lateral move from a line drawn from white to the proper x,y coordinate, then it is an error in hue. If it is on the same line but just further away from white, then it is oversaturated. If it is both further away and skewed laterally, then both the hue and saturation are affected.

Not just saturation, but also hue. Two dimensions.


Right but from what Art had posted, it seemed like there was no connection between x/y and saturation. I had heard that the hues on the RS2 were almost perfect.

I feel like I keep repeating myself about this one point: in reality (as opposed to theory) adhering to standards does not mean that you are watching a film as the director intended. I am sure that there are exceptions, and I am sure that some DVDs or Blurray disks do accurately capture the filmmaker's intent, but as a general rule, the moment the filmmaker releases the studio master, the filmmaker loses control of it.

This fact is so generally accepted that magazines like WSR, when they review DVDs or Bluray, typically comment on the quality of the transfer. There are numerous threads on these boards about which releases qualify as reference quality and which do not.

I hate to be the one to break the news but the world is not perfect and neither are the sources that we watch.

I never said they are. Actually, a page ago I said specifically that displaying colors accurately, doesn't mean that all movies will look perfect, it means they will look as good as they can under the circumstances. There is no way changing the gamut of the set changes the fact that movies will look different, and some will look better than others. I don't see what the difference in transfers has to do with how we set the colors, since these differences will be there no matter how you set the colors.

You may argue that it should be possible to change the gamut according to each movie, but how would you know how to set it? Do you really want to change settings during the course of a movie? I certainly don't.

If what you are saying is true, you might as well assume that studios understand that most TVs have oversaturated colors so they test the material on oversaturated TV sets and, if it doesn't look right on them, they go back to the SMPTE C monitor and fix it. That would mean that the real intent could only be captured if a dvd was watched on a set with oversaturated colors. I don't think that you mean to suggest that, do you?

I need to clear something up here: Are we talking about watching DVD material with rec.709 colors, or watching HD that was mastered with a SMPTE C monitor, using rec.709 colors?

I definately believe NTSC DVD should be watched using SMPTE C gamut, and HD should be watched using rec.709 gamut. In practice, you often have to choose one for both format, but this is not ideal. If an HD source is poorly mastered, i.e. by using a monitor with SMPTE C and not taking this difference into account (which, as Bob points out, really isn't much of an issue anyway since it doesn't make that much difference), then the issue is with the source.

If the producers make HD material using SMPTE C gamut, they're the ones doing something wrong. If we change our sets to compensate, we remove the possibility of correcting the problem where it stems from: The source. If we change the display to adjust for errors in the source, we don't correct the problem - we postpone the problem to the next movie.

The director may have used a SMPTE C monitor for the actual monitoring, but the final product is (or should be) graded such that it will look "right" on a rec.709 monitor, not on the SMPTE C monitor that was used for grading. Even if the working tool is SMPTE C, the final product is (or should be) designed for rec.709. In reality, I believe it will be graded such that it works on either, since the workflow in practice would be: Grade it to look properly on SMPTE C monitor, check the result in the screening room with rec.709 (and probably varous other sources), and if it doesn't look right there, go back to the SMPTE C monitor and change whatever looked wrong.

The reality is that the movies are graded to look "right" on displays varying to some degree. The display doesnt have to be 100.000000000% correct, for the movie to look "right". If the display looks "roughly right", the movie will look right. SMPTE C vs rec.709 is "roughly right", rec.709 vs. HD1 is not.

I must leave a tiny discalimer at this point: My understanding of the workflow of production studios are from small, independent studios. I wouldn't know how the major studios work.

The reality is that the movies are graded to look "right" on displays varying to some degree. The display doesnt have to be 100.000000000% correct, for the movie to look "right". If the display looks "roughly right", the movie will look right. SMPTE C vs rec.709 is "roughly right", rec.709 vs. HD1 is not.

If you believe this to be true, even to the degree that you appear to, how can you rule out the possibility that people making DVDs make them knowing that most TV sets are shipped with oversaturated colors and therefore set colors assuming that they will look good on those sets? This suggests that watching them to spec would be wrong because colors would be undersaturated if viewed that way. This perhaps explains why many people view accurate colors are cold and undersaturated. Perhaps they are?

If you believe this to be true, even to the degree that you appear to, how can you rule out the possibility that people making DVDs make them knowing that most TV sets are shipped with oversaturated colors and therefore set colors assuming that they will look good on those sets?

I can't. That certainly IS a possibility. However, I have no way of knowing if that is the case when playing the movie, so all I can do is play it properly, and if it doesn't look right, then the transfer is bad, and we should bash the publisher for spitting out crappy releases. If we allow the publishers to stray from the standards, and just adjust accordingly on the display, we get chaos, because we won't know how to adjust our set. If that scenario happens, then we have lost. This is a scenario that we must not allow to happen, because there is no way we could correct for that. It would be the same as giving up on the whole idea of watching the movies with even remote accuracy. Fortunately, the people I have talked to so far are pretty hard-fastened in the belief that their job is to make the movie look right when displayed right, and if people don't display it right, that's their own problem.

I'm still not sure if we're talking DVD or HD material here, but: DVD's are mastered for SMPTE C, and should be watched as such. HD is (or should be) mastered for rec.709. If the producer has a rec.709 monitor, all he has to do is make sure it looks right on that monitor. If he only has a SMPTE C monitor available, he will use this monitor (since he has no choice), but make sure that it looks right using a rec.709 display. I.e. he deliberately makes it look "wrong" on his set (in reality, they just inherently know that they shouldn't fully saturate certain colors, etc.) Basically, even though the monitor used is SMPTE C, being an HD release the proper gamut to use is still rec.709, and that's what the release was designed for. So, watching it with rec.709 gamut isn't oversaturating. HD should always be watched with Rec.709 colors, regardless of whatever the producer used for monitoring.

EDIT: Just read my previous post again, and I realize that what I wrote made it seem like the producers grades the movies to make them look pretty good on a _wide_ range of displays. That wasn't really what I meant to say, at least color-wise, they (as far as I'm being told) use pretty strict setups - however, it is still my understanding that the final evaluation of the resulting colors are done in the screening room, not in the color grading room, and because of this, it is not really that important to know which monitor they are using to work with, as this is not the final result.

Once again, this is based on what I'm being told by people in the industry. If anyone knows otherwise, feel free to jump in.

When we talk about x/y position we are talking about saturation.

The intensity of a color can be altered without moving it toward or away from a primary.

The crimsom Ferraris is one thing the clothing that looked correct but glowing is another.

Of course ,just turning the color down will coorect some of this but if only one primary is exaggerated then it introduces other issues.

If I had my way I'd certainly choose over saturated before under since that later just sucks the life out of the picture.

Art

Not just saturation, but also hue. Two dimensions.

Yes !

If I had my way I'd certainly choose over saturated before under since that later just sucks the life out of the picture.

Art

I think most people agree to this, and this is why most people, if they had to make the choice due to limitations in the set, would prefer to display all material using Rec. 709, rather than watching all material using SMPTE C. This is more a practicality than a choice of oversaturating the colors. As much as I want to watch movies accurately, there are limits as to how far I want to go in practical terms to get there. Having to change settings during the film is a big deal for me. Perhaps it's _because_ I do that for a living, that it's something I really don't want to have to do when I watch movies in my spare time... And my biggest gripe with having to change settings is still the fact that I won't know how to set the colors until I'm done watching the movie, and by that time it won't matter anymore... What we want is to find one setting that works best with everything. In my opinion, rec.709 colors does this, in theory and in practice, so why deliberately go for anything else?

I still feel that people like Lawguy (nothing personal!) are looking for arguments where they really arent nescessary to make. If Rec.709 works, why are we looking for reasons to use something else?

The intensity of a color can be altered without moving it toward or away from a primary.

The crimsom Ferraris is one thing the clothing that looked correct but glowing is another.

Of course ,just turning the color down will coorect some of this but if only one primary is exaggerated then it introduces other issues.

If I had my way I'd certainly choose over saturated before under since that later just sucks the life out of the picture.

Art

If the RS2 is like the RS1, its hues are very good. Greg Rogers calls both close to perfect.

The intensity that you were seeing is probably oversaturation, because out of the box, I think the Y (lightness) of the RS2 is also pretty good.

When you turn down the color control on the RS2, it lowers the Y value. The colors will still be oversaturated, but they will probably not stand out like you mentioned.

I still feel that people like Lawguy (nothing personal!) are looking for arguments where they really arent nescessary to make. If Rec.709 works, why are we looking for reasons to use something else?

No offense taken.

Sometimes more vibrant colors are subjectively better looking. This suggests that Rec. 709 may not be working as well as everyone likes. That is perhaps why manufacturers make sets that have colors that are oversatured when viewed from any existing specification. That is why consumers prefer oversaturated colors. That is why newer specs like x.v. color and Deep Color are emerging. That is why HDMI 1.3 contemplates these newer color gamuts.

Everyone recognizes the problem. None of the pieces are in place to correct it yet.

If the RS2 is like the RS1, its hues are very good. Greg Rogers calls both close to perfect.

The intensity that you were seeing is probably oversaturation, because out of the box, I think the Y (lightness) of the RS2 is also pretty good.

When you turn down the color control on the RS2, it lowers the Y value. The colors will still be oversaturated, but they will probably not stand out like you mentioned.


Nay, not it. Like I said ,the intensity is off but so is the hue . Now again my experience is limited but just saying what I saw.

Nay, not it. Like I said ,the intensity is off but so is the hue . Now again my experience is limited but just saying what I saw.


You have one up on me becasue I have never seen one.

Nay, not it. Like I said ,the intensity is off but so is the hue . Now again my experience is limited but just saying what I saw.

You mentioned earlier Ferrari reds looked crimson (ie: blue tint), assuming the error isn't isn't in the source I'd first assume the projector is in desperate need of calibration.

No offense taken.

Sometimes more vibrant colors are subjectively better looking. This suggests that Rec. 709 may not be working as well as everyone likes. That is perhaps why manufacturers make sets that have colors that are oversatured when viewed from any existing specification. That is why consumers prefer oversaturated colors.

It is a well known fact, that if you show to pictures to the average consumer, he will most likely point to the most saturated one as "best". However, this doesn't mean that the consumer actually prefers oversaturated colors in everyday use. I calibrate several tv's every week, and get the response from customers. It is _extremely_ rare that people don't find the post-calibration picture better, and most often the same oversaturated colors that made them purchase the set in the store, is the very reason they want the set calibrated in the first place. Manufacturers know this. They don't produce sets with oversaturated colors because it's what the consumers think is best. They do it because it's what the consumers choose when they're watching the in-store display wall. But they do this without realizing what the impact will be on everyday use.

Why JVC chose to oversaturate the colors on a projector, that doesn't need to make itself seen on a video wall, is beyond me. People buy it for the contrast, not for the colors. If anything, people buy it _despite_ the colors, and I seriously believe JVC would have sold more if it had rec.709 colors.

It's like shagging Amy Winehouse. It's probably fun once or twice, but do you seriously want to marry her?

That is why newer specs like x.v. color and Deep Color are emerging. That is why HDMI 1.3 contemplates these newer color gamuts.

x.v.Color is completely different! This is a new color system, and should be treated as such. If the source is x.v.Color, the display should certainly still adhere to the standard, just the NEW standard. Obviously, if the source is x.v.Color, it should be displayed as such. This is completely different than displaying the image using different primaries than the system was intended to use.

One very important thing in this regard: x.v.Color uses rec.709 primary reference points (AFAIK), but expands the color gamut beyond these reference points. If you add x.v.Color to a JVC, it will still have oversaturated color. If you catch my drift, you can actually have a display that has oversaturated primary reference points, but is not capable of displaying colors that are fully saturated out to the limit of x.v.Color. This display would have undersaturated color extremes, but oversaturated colors up to the limit of the display.

Basically, if you had an x.v.Color source, and wanted to make use of the JVC's expanded gamut to display a wider gamut, resembling the original gamut of the film, you would still need to calibrate the JVC to match the rec.709 primary reference points. If you don't, you won't benefit from the extended gamut in the souce at all. So basically, even if the source includes a wider gamut, you still need to match the primary references.

(Deep Color is greater bit depth, and has nothing to do with this discussion)



Everyone recognizes the problem. None of the pieces are in place to correct it yet.

Replacing the HD standard with a standard that allows the gamut of the source to match the original film source, still requires accurate primaries. That doesn't change anything, and like you say none of the pieces are in place. What we need to do to be able to make use of i.e. x.v.Color in the long run, is to have displays today that have a wide color gamut (to be able to support x.v.Color), BUT has a rec.709 mode to be used for HD sources. In this way we can "sneek" displays capable of x.v.Color in through the back door, and when a sufficient number of displays are available, actually sending x.v.Color sources to market will finally provide us with the ability to display the full color palette of the movies. But as long as the movies we watch are HD, the palette will be limited in the production, and there is no way of getting them back by expanding the colors inside the display. This limitation can only be removed by the producers, and they can only remove it when the displays provide the ability do do so. Which is first: The chicken or the egg?

Basically, the perfect display has four gamut modes: SMPTE C for NTSC sources, EBU for PAL sources (not important stateside), Rec.709 for HD sources, and "Rec.709 expanded" for x.v.Color sources (in practice you choose Rec.709 gamut, then toggle x.v.Color on/off)

To stress my point: I most certainly do want to be able to display the most saturated color of film, that are normally limited by the HD system. I just don't believe it is possible as long as the source is HD, we need to change the system to see these colors, if we want to do it without sacrificing color fidelity. Sacrificing the most saturated colors of the film, is not a choice of the display, it's a choice of the system.

Why JVC chose to oversaturate the colors on a projector, that doesn't need to make itself seen on a video wall, is beyond me. People buy it for the contrast, not for the colors. If anything, people buy it _despite_ the colors, and I seriously believe JVC would have sold more if it had rec.709 colors.

Strange that JVC's new vp (the RSVP2) that has 'corrected' color gamuts for the RS1 and RS2 says that the 'corrected' colors are SMPTE-C, not Rec 709. From your posts here and above, this seems weird.

Basically, the perfect display has four gamut modes: SMPTE C for NTSC sources, EBU for PAL sources (not important stateside), Rec.709 for HD sources, and "Rec.709 expanded" for x.v.Color sources (in practice you choose Rec.709 gamut, then toggle x.v.Color on/off)


I would also throw in a few custom gamuts that could help with different kinds of faulty source material.

Strange that JVC's new vp (the RSVP2) that has 'corrected' color gamuts for the RS1 and RS2 says that the 'corrected' colors are SMPTE-C, not Rec 709. From your posts here and above, this seems weird.

Maybe it just IS weird?

I'm not saying I have the final truth to everything in the world, I'm just saying how I see things. I have no experience with or knowledge about the JVC VP, but obviously there are different opinions as to what "accurate" actually is, so why shouldn't JVC's technicians have a different opinion on this. To my knowledge, the Samsung's that have been designed by Joe Kane, use Rec.709 for HD inputs, so just because a manufacturer says this or that is correct, doesn't make it so. If I had to choose between trusting JK and trusting JVC, I would have to choose the former. They can't both be right.

EDIT: As I said, I don't have experience with the processor. From this, it sounds like the gamut is not adjustable, it just provides a "corrected" setting? Perhaps they just chose to cheap out and provide one "corrected" gamut for everything. Even if HD should be watched with SMPTE-C gamut, what about PAL?

I would also throw in a few custom gamuts that could help with different kinds of faulty source material.

You can have 200 custom gamuts, as long as those four are present ;-)

SMPTE C vs rec.709 is "roughly right", rec.709 vs. HD1 is not.Looking at Greg's review for the RS1, by his measurements it looks like at least for red the RS1 primary is closer to Rec.709 than the Rec.709 red is to SMPTE-C red.
To my knowledge, the Samsung's that have been designed by Joe Kane, use Rec.709 for HD inputs, so just because a manufacturer says this or that is correct, doesn't make it so. If I had to choose between trusting JK and trusting JVC, I would have to choose the former. They can't both be right.The Samsungs let the user decide which to use and I believe that Joe uses SMPTE-C (and recommends SMPTE-C) even for HD for now because of most of the mastering monitors having SMPTE-C primaries. Ken Whitcomb could probably verify that.

--Darin

You mentioned earlier Ferrari reds looked crimson (ie: blue tint), assuming the error isn't isn't in the source I'd first assume the projector is in desperate need of calibration.

I'd say that too but it had just been done, within the confines of the projectors adjustments ,not with an external VP.

Art

I'd say that too but it had just been done, within the confines of the projectors adjustments ,not with an external VP.Given the controls that the projector has, I wonder if it would have had the push toward blue for the Ferrari with default settings. Trying to calibrate one thing can screw up another when there aren't proper controls.

Do you remember what show the material came from? I would like to try it on an RS2 and other projectors at some point.

--Darin

The Samsungs let the user decide which to use and I believe that Joe uses SMPTE-C (and recommends SMPTE-C) even for HD for now because of most of the mastering monitors having SMPTE-C primaries. Ken Whitcomb could probably verify that.



In that case I haven't been paying attention during classes... I don't remember him recommending SMPTE-C for HD. I think I have come to the point where I should verify my sources before continuing this discussion, and maybe go back and watch some HD with different gamuts, it's been a while since I tried that last... I may be biased in that most of what I watch is either HD or PAL, not much NTSC.

Anyway, the main point of the discussion - to me at least - was never SMPTE C vs Rec.709, but accurate vs. non-accurate. Even if I'll have to eat my words regarding SMPTE C vs Rec.709, I still stand by every point I made regarding the importance of having one set standard to go by, even if sources in practice may be compromised for different reasons.

Don't get me wrong ,I'm not married to any of these standards but after sitting with an RS2 the other night, and seeing some very intersting colors that literally I'd never seen any place including crimson Ferraris. I think that having some sort of reasonably accurate color set to some baseline is a good goal.

Art
Art, I don't doubt for a second that you saw exactly what you say. I could only guess at what could have been wrong with that projector (a bad digital color decoder, a problem with a panel or panel driver, ???). But there is no way the Ferraris are anything close to crimson on my RS2. That is born out by the color measurements that show that the hue of red is nearly perfect, it is just oversaturated from the Rec. 709 red. So I totally believe you, but I suspect something was wrong with the projector you saw. (Calibration-wise the hue control would have to be quite a ways off, so I doubt that was it. I'm guessing it was somehow defective.)

Art, I don't doubt for a second that you saw exactly what you say. I could only guess at what could have been wrong with that projector (a bad digital color decoder, a problem with a panel or panel driver, ???). But there is no way the Ferraris are anything close to crimson on my RS2. That is born out by the color measurements that show that the hue of red is nearly perfect, it is just oversaturated from the Rec. 709 red. So I totally believe you, but I suspect something was wrong with the projector you saw. (Calibration-wise the hue control would have to be quite a ways off, so I doubt that was it. I'm guessing it was somehow defective.)

Great to hear since other than the color the projector looked fantastic.

Art

Not tracking the D65 or uncalibrated grayscale can shift the hue.

Not tracking the D65 or uncalibrated grayscale can shift the hue.Yes, but I doubt to the extent implied by Art's crimson observation for a color so close to the red primary.