Home Theater Forum and Systems banner

41 - 60 of 77 Posts

·
Registered
Joined
·
178 Posts
It is a fact that the software is now the determining factor of the cube calibration, wereas previously it was the skill of the calibrator.

In my opinion, CalMAN produces better results and has more customization options for those targets. I'm willing to post hard data to back that opinion up. CalMAN and Chromapures results should be able to be cross checked with each other, no need for HCFR.
OK, so again I ask in what way is Calman producing better results than Chromapure? The rest is just rhetoric (my HFCR comment was just a little joke BTW ;) ). However, I think in fairness to this being the Chromapure thread, it would be better to create a fresh thread to compare the two such this be done.




The point, Tom, is precisely that gamma may vary over the dynamic range of individual colors. The question is how much better is it to use the 125 point system than to use a 6 point system. As Joel pointed out, this would vary with displays. I would like to see the difference between a 6 point calibration and the 125 point calibration, with the data on the points not calibrated from the 6 point method compared to the adjustments that the 125 point system makes. A measure of variance across the points on each color might be meaningful. It would demonstrate the value, if significant, of the 125 point system, and/or be a measure of color accuracy (or lack thereof) in a display.

I'm not sure if you are aware, but Chromapure can perform the 125 point calibration as well, not just a 6 point one. My apologies if I've misunderstood your post, but I read it that you don't realise this.
 

·
Registered
Joined
·
71 Posts
OK, so again I ask in what way is Calman producing better results than Chromapure? The rest is just rhetoric (my HFCR comment was just a little joke BTW ;) ). However, I think in fairness to this being the Chromapure thread, it would be better to create a fresh thread to compare the two such this be done..
I would assert it's better from the exceptionally low average dE's and the from the relative speed that it attained the dE's.

dE being the best metric we have for color accuracy.

At some point, the results get close enough to perfect that absolute accuracy would be considered equivalent. There are a few other qualities, ability to preserve as much light output at possible, speed and configurability come to mind.

But sure if you want to start another thread we can do that. I have quite a few displays and a couple projectors that I can test on.
 

·
Registered
Joined
·
178 Posts
Yes, I think another thread would be fairer given this is the Chromapure one. I'd be interested to see the same display and same meter corrected with Chromapure if that's possible (I'm assuming you have a sample of your rival's software). I got some pretty low dEs when I did my TV using the 125 point Chromapure, but my TV is a little unstable at low levels, so I'll post back once I do my JVC X35 in the next few weeks. However my meter may not be as accurate, which was part of my point above, though if the dEs are below 2 (or is it 3?) then isn't that below what we can see anyway?
 

·
Registered
Joined
·
71 Posts
Yes, I think another thread would be fairer given this is the Chromapure one. I'd be interested to see the same display and same meter corrected with Chromapure if that's possible (I'm assuming you have a sample of your rival's software). I got some pretty low dEs when I did my TV using the 125 point Chromapure, but my TV is a little unstable at low levels, so I'll post back once I do my JVC X35 in the next few weeks. However my meter may not be as accurate, which was part of my point above, though if the dEs are below 2 (or is it 3?) then isn't that below what we can see anyway?
3 is generally what we say for moving content. If you butt fit two swatches against each other you can see down to about 0.5 or so.

Also if you had one shade of gray at a dE of 2.5 toward red, and the next shade was 2.5 towards cyan you would most certainly notice the 5 dE swing between the two shade although they are both technically within 3.
 

·
Registered
Joined
·
11 Posts
The point, Tom, is precisely that gamma may vary over the dynamic range of individual colors. The question is how much better is it to use the 125 point system than to use a 6 point system. As Joel pointed out, this would vary with displays. I would like to see the difference between a 6 point calibration and the 125 point calibration, with the data on the points not calibrated from the 6 point method compared to the adjustments that the 125 point system makes. A measure of variance across the points on each color might be meaningful. It would demonstrate the value, if significant, of the 125 point system, and/or be a measure of color accuracy (or lack thereof) in a display.
I don't know what "gamma may vary over the dynamic range of individual colors" means. Any time you measure color at anything less than 100% stimulus, then gamma comes into play. But there is nothing mysterious about this. You can calibrate using a fixed power law gamma or a variable gamma method, such as BT.1886, but this is all a relatively simple calculation.

Whether a 124-point calibration offers a meaningful improvement over a 6-point calibration depends entirely on the display. Some displays are quite linear, so that having calibrated at the gamut boundary the colors inside the gamut line up quite nicely. For other displays, the very act of using its CMS to calibrate 6 colors results in colors inside the gamut that are quite inaccurate. Others are just natively non-linear (Sharp Elites). It all depends on the display involved and how well its CMS is designed.

By way of example, I had previously heaped praise upon the Sharp XVZ-20000's color performance because of what I thought was its great CMS. However, I later discovered after measuring inside the gamut that it substantially undersaturated color even while colors at the gamut boundary measured perfectly. This was easily visible with regular program material.
 

·
Registered
Joined
·
33 Posts
Yes, I think another thread would be fairer given this is the Chromapure one. I'd be interested to see the same display and same meter corrected with Chromapure if that's possible (I'm assuming you have a sample of your rival's software). I got some pretty low dEs when I did my TV using the 125 point Chromapure, but my TV is a little unstable at low levels, so I'll post back once I do my JVC X35 in the next few weeks. However my meter may not be as accurate, which was part of my point above, though if the dEs are below 2 (or is it 3?) then isn't that below what we can see anyway?
Hi Kelvin,

Your point about the limit of 'visible threshold' is a good one because as you know I have a Duo so cannot use 125 point calibration.

Do you have any 'Before' and 'After' results yet.

I ask because I find that my use of 75% sat patterns has shown my cheapo CCFL LCD to be very linear when calibrated this way compared with rubbish results at 100%.
 

·
Registered
Joined
·
178 Posts
Hi Kelvin,

Your point about the limit of 'visible threshold' is a good one because as you know I have a Duo so cannot use 125 point calibration.

Do you have any 'Before' and 'After' results yet.

I ask because I find that my use of 75% sat patterns has shown my cheapo CCFL LCD to be very linear when calibrated this way compared with rubbish results at 100%.
I have some 125 results but they are from my Sony LCD TV which is a bit unstable at low %. I only use it as a test bed to save projector lamp hours, but as I'm replacing the HD350 soon I haven't bothered doing the 125 point on that as I'll wait until I get my X35.

I did find that when using the old type CMS on my TV that it worked better using 75% like you. However, my yellow then did something odd at 100% which the 125 point of course corrected, but as there is little 100% yellow in 'real' content it seemed a good compromise.

If I compared some paused scenes of Transformers (with plenty of yellow in) then I could see quite a difference between the 125 point and the 8 point setting (I saved two memories in my Lumagen with the same greyscale and only the 125/8 point CMS settings different). If I'm honest though, it was only that I knew the 125 point measured better as I think in a blind test I wouldn't know which was which. ;) At least with the 125 point calibration I know that I'm seeing it as close to how it's meant to look, which is probably the whole point.
 

·
Registered
Joined
·
11 Posts
OK, so again I ask in what way is Calman producing better results than Chromapure?
In fact, I am not even sure what "better" would mean here. Most of the heavy lifting in the process is done by the hardware, the Lumagen and the meter. The Lumagen generates all of the test patterns, makes all of the adjustments, and outputs the corrected result. The meter reads the before and after data.

What does the software do?

1. It instructs the Lumagen to generate the correct test pattern.
2. It records a reading from the meter.
3. It converts the xyY data into RGB and then compares the uncorrected RGB value with the target RGB value and assigns a dE.
4. It determines whether the measured dE meets a user-specified target. If so, it moves on to the next color. If not, it proceeds to the next step.
5. It uses a proprietary algorithm to determine what correction is needed to get the RGB output closer to the target values and then instructs the Lumagen to apply that correction.
6. It takes another reading and then just repeats 3-5 until the target dE value is met and continues to move through the series of colors until all have been corrected.

It was A LOT of work to implement this, but most of these steps are quite straightforward, if also rather labor intensive. Only step 5 is really tricky and just involves much trial-and-error until you get it right. In fact, there is no difference in principle between the 124-point auto-calibration and the standard 6-point auto-calibration we have had for over a year. It is just a lot more colors and a lot of new Lumagen commands to master. The basic process is the same.

What would constitute doing this "better"? Speed? That is largely a function of the meter, how far out the display is--the further out it is the more corrections are required--and options the user selects--the lower the dE target, the more corrections are required. The efficiency with with the software implements all of this is relevant as well, but no one who is being honest and knows what they are talking about would ever accuse CP of being slower in this regard than the competition. Rather, it is quite the reverse. Accuracy? That is just a function of the dE target selected and the inherent accuracy of the meter. BTW, selecting a CIE94 or CIEDE2000 dE target below 1.5 is sort of pointless. Any lower dE value is almost certainly achieved by only extending the completion time of the calibration without resulting in visibly improved color.

All that is left is the UI and and ease-of-use.
 

·
Registered
Joined
·
178 Posts
Thanks Tom, that makes a lot of sense and re-enforces what I had in the back of my mind but wasn't able to get down so succinctly.

Somewhere I have some old HFCR results that show a 'perfect' result when I used my Lumagen to calibrate my HD350 manually at 11 points greyscale and WRGBCYM CMS but using my old i1LT (D2) sensor. The charts looked pretty, but the picture didn't, so having dEs nearly at 0 didn't count for much in my case, hence my comments about being more limited by the meter (D3 thesedays which is an improvement).

I know that Calman has been improved since I first looked at it, but I've always found Chromapure seemed easier to use and understand, which for an amateur is a big consideration. Before autocal the real time gamma section with the real time RGB was pretty much all I needed to do my greyscale. The actual greyscale 'post calibration' measurements were just to confirm what I already had adjusted.
 

·
Registered
Joined
·
33 Posts
Historically I bought both Calman and CP at reasonable costs so have merely kept them both up to date.

I like both softwares for different reasons.

I agree with Kelvin that CP's attraction is the simplicity of its operation (which incidentally does not mean that the Professional Calibrator should not use it).
However again I agree with Kelvin that Calman has improved significantly in many ways in appealing to the novice / amateur user rather than needing a degree in technical understanding before its capability can be explored.

Certainly both are now far more logical with their workflows compared with my first attempts using HFCR.

Both have contributed substantially to my understanding of 125 point calibration which because I do not have a Lumagen (Duo owner) I have envied from a distance.

For some time I have struggled with a Toshiba CCFL LCD that simply would not fully saturate the Blue colour, in fact shifting its hue seems to be also problematic at its near 100% saturation value.

Tom helped me analyse my ACM results (via Chromapure Pro) that probably my Panel would be better if 75% saturation patterns were used for Color Gamut calibration.
I have done this for some time now using the AVS Disc to generate the patterns (lately with CP via its own 75% Duo generated patterns, thanks Tom).

I was confident my calibration was good but being an inquisitive person I kept wondering what the 125 point calibration would do for my Panel.
This would be difficult unless I bought a Lumagen just to get this capability.

Then Joel helped solve this dilemma for me by introducing the Colorchecker capability into Calman 5. This was exactly the tool I needed to check my calibration throughout the full gamut because its 24 patterns were spaced to check exactly that.

So I had results (in terms of dE values) for each of the 24 patterns and an average figure across them and a maximum value for the one needing attention. Like Kelvin I was concerned about the 'Visible Threshold' so was confident my results would hold up against a 125 calibration.
Unfortunately nobody seemed to be using this tool to publish their results so comparison was still difficult.

Joel has helped with this by providing moving and static side by side 'noticeable' dE limits.

I now have my answers and would advise any Duo owners to explore all calibration options of their Displays before they decide to purchase a Lumagen for its 125 point capabilty only, they may be disappointed.

Whatever your choice of software I think you will enjoy it and if you are like me there is no need to choose.
 

·
Registered
Joined
·
178 Posts
Nice to see a reasoned and honest reply PE06MCG. :clap: It's debatable as to whether you'd get more of an overall improvement if you replaced the TV rather than the VP, though I did notice that there is an ex display Mini3D on the UK AVforum (I can PM you a link if need be) :devil: for £1099. You'd maybe get a fair bit of that amount back for your Duo too. Not sure if you're a non 3D viewer like me, but the Mini3D can process greyscale and CMS for 3D signals rather than just pass them through as the Duo does, so there are a few reasons to spend a bit more, but I do see your point.
 

·
Registered
Joined
·
33 Posts
Nice to see a reasoned and honest reply PE06MCG. :clap: It's debatable as to whether you'd get more of an overall improvement if you replaced the TV rather than the VP, though I did notice that there is an ex display Mini3D on the UK AVforum (I can PM you a link if need be) :devil: for £1099. You'd maybe get a fair bit of that amount back for your Duo too. Not sure if you're a non 3D viewer like me, but the Mini3D can process greyscale and CMS for 3D signals rather than just pass them through as the Duo does, so there are a few reasons to spend a bit more, but I do see your point.
Thanks Kelvin but I think me and my Duo are well suited.

My TV is problematic but trying to solve its problems has taught me a lot. Don't think changing my Display hasn't crossed my mind though (She Who Must Be Obeyed may have something negative to say about that though).

Perhaps I should have said that I have lived through virtually all the Duo's early firmwares, some of which were decidedly scary but the hardware as well as the current firmware has not given me a moments trouble.

Not a 3D person but can see advantages in both VP's (a bit like Chromapure v Calman do you think?).

I suppose that I was trying to infer that perhaps Displays are not as non linear as we have been led to believe? My Cheapo Toshiba just wanted calibrating at 75/75 to find its sweet spot maybe other more expensive ones would benefit from this approach.
 

·
Registered
Joined
·
178 Posts
I think this is a big part of calibration and of course where a good pro earns his money: Knowing the little quirks of certain displays and what point works best to calibrate at and probably many other things. The thing with being an enthusiastic amateur is that with limited displays you can concentrate on getting the max out of them. I haven't owned a plasma TV or a DLP projector so I haven't 'wasted' any time learning about them, but if it's a JVC DILA or a Sony LCD, then I've read as much as I can about them, tested and practised to get the best out of them. Even my 5 year old 40" stays in the living room as it's still giving a great picture (with the Lumagen of course;) ) and I like to think that my HD350 was giving near enough as good a picture as many less carefully set up more expensive projectors. :) It sounds like you've maxed out your Toshiba too, so you know it's giving it's absolute best now.
 

·
Registered
Joined
·
309 Posts
ChromaPure 2.5.4 is now available for download.

ChromaPure 2.5.4 Release Notes

New Features

Added support for the Lumagen 17x17x17 (4913 colors) LUT calibration

Added support for the Colorimetry Research CR-100 professional colorimeter.

Added support for the Minolta CL-200 colorimeter

Added a Test Gamut report

This report is available to all ChromaPure Standard users. It measures a broad sample of the gamut (215 colors) and generates a report that describes performance throughout the gamut. Use this to determine whether the display requires an LUT calibration. Running this report only requires an active meter and signal generator. The built-in signal generator may be used.

Improvements

Inspired by the need to support the new 17^3 Lumagen LUT, we implemented a much faster method for advanced auto-calibration. The new method reduces the time required to run an advanced auto-calibrate session by more than half. Using an i1Display Pro the time required to cycle through one pass of the color matrix is now approximately

125 Colors: 6 minutes
729 Colors: 28 minutes
4913 Colors: 194 minutes

When using a Klein K-10a or CR-100 the time required is now
125 Colors: 3 minutes
729 Colors: 13 minutes
4913 Colors: 90 minutes

We added support for additional gamma targets in the Gamma module.

Bug Fixes

When selecting the 75% of Rec. 709 Reference Gamut, the Advanced Color Management module remains fixed at Rec. 709.

Known Issues

If, for whatever reason, you want to run Advanced Auto-cal more than once, we recommend that you close ChromaPure and then re-start between attempts. There is a as-yet-unidentified problem when performing multiple runs of advanced auto-cal during the same calibration session. We are investigating the problem and hope to have a fix in a future release.
 

·
Registered
Joined
·
309 Posts
ChromaPure 2.5.5 is now available for download.

ChromaPure 2.5.5 Release Notes

2.5.5 primarily fixes bugs and refines previously released features.

New Features

We have added an OLED (White) mode to the list of PRO corrections that are used with the Display 3 PRO and other colorimeters.

Improvements

We increased the precision of the DVDO TPG video generator by upgrading from percentage values to 16-bit video code values.

Advanced Auto-calibrate now detects the specific Lumagen model and activates only those color matrix sizes supported by the model.

The speed of grayscale/gamma calibration in Lumagen auto-cal has been increased.

We provided a new reset option to Lumagen auto-cal that leaves the grayscale/gamma setting unchanged for those who prefer to do this manually. If you select this option, then grayscale/gamma auto-cal is not available.

We added an option to the Gamut Test Report that allows use without a signal generator. The user will provide test patterns from a Blu-ray or DVD disc and the application will automatically take measurements at user-defined intervals.

We added a progress bar to the Gamut Test Report.

The window size used in the Built-in test patterns was reduced to about 6% to avoid gamma errors near 100%.

Bug Fixes

We have fixed the initialization problems with the i1Pro 2.

We have fixed the Front Projection mode initialization problem with the i1D3.

We have fixed a problem with the AccuPel in which it would freeze after displaying test patterns selected from signal generator toolbar.

We fixed gamut errors in Advanced Color Management when EBU was selected as the reference gamut.

We fixed formatting errors in the Advanced auto-cal dE report chart that would show numbers exceeding 100%.

We fixed the white reference value the in ColorChecker module when using the Built-in test pattern generator.

We fixed dE errors that appeared when running Advanced Auto-cal with BT.1886 selected as the target gamma.
 

·
Registered
Joined
·
309 Posts
ChromaPure 2.5.5.1 is now available for download.

ChromaPure 2.5.5.1 Release Notes

This contains a few critical stability and performance updates.

Because this is the last 2.x release, we recommend that all ChromaPure users upgrade to this version.
 

·
Registered
Joined
·
309 Posts
ChromaPure 2.5.5.19070 is now available for download.

Release Notes

''The only functional difference is that we set it so that if you select the "leave grayscale alone" option in adv. auto-cal, it does not set the Lumagen to 12-point, thus erasing any 21-point calibration.

My recommendation is to let auto-cal do the grayscale/gamma. When it is finished, you can go into the Lumagen, change to 21-point (this does not erase a 12-point--it just interpolates the in between points), and then make minor adjustments if/where necessary.''
 
41 - 60 of 77 Posts
Top