ColorHCFR v2.0 Available... - Page 3

Just "a bug"?... Vista is more like a swarm of bugs....

I'd like to try HCFR, but I heard earlier versions were not meant for plasma calibration. Is that still the case?

The earlier version was fine for plasmas. The issue was that depending on what particular probe you used the measured results were not necessarily as accurate as they could be.

There is now a specific mode for using the Display2 probe for plasmas.

I have used it on mine and the results are quite good.

best place to get a deal on a display 2?

thx

bob

j&r

Actually it's much less at dgr. $129 after shopping cart discount and free shipping. It's the Display LT which is the exact same sensor as the Display 2.

http://www.digitalgraphicsresources....ROD&ProdID=865

The only match I get at jr is this one, for $169 plus shipping;

http://www.jr.com/JRProductPage.process?Product=4027183

As far as I can tell the sensors in the Eye-One Display LT and the Eye-One Display 2 are identical, with the only differences being in the software provided.

Hello Losha

Yes, this error was present since the very beginning of the project (even in the very first beta versions)... we didn't find it because it does not exist when compiling in debug mode :roll:

Anyway, I found it, and it will be ok in next release. But I don't think it's a showstopper, nobody complained about it before

Anyway, I think there are many new other things which have not been noticed. We changed our matrix computation formulas, to improve calibration files and XYZ adjustment calculus accuracy. For people using two probes, a brand new controlled mode has been added, which recompute in real time XYZ adjustment matrix to make measures as accurate as possible, using two probes.

And believe me or not, some internal refactoring will allow us to develop new things soon

Georges

From a simple "user" viewpoint I think that the usability is much better.

The biggest factor here is the improved measurement speed and no more hanging.

The new window system also has all of the information you need in one place, which is a good usability feature.

As far as improved accuracy goes, I don't have a reference so I can't really comment on this one. I do know that after remeasuring and fine tuning with 2.0 the picture still looks fantastic.

I have noticed these and the training of probes is amazingly accurate now - great job. And there appears to be a saving/loading of the XYZ matrix which is very convenient. There are also a number of new options which I haven't tried yet, but with more time I will.

I tried them and I think the Test Patterns->Display Patterns are excellent! . Request for 2.x Display Patterns: A small window option (% of screen input like in view images) for the full screen colors too?

-JD

Possible bug or I am doing something wrong?: Open 2 CHC files and set one as Reference and then open the RGB levels histogram for the non-reference. The dashed "Reference Measure DeltaE" line values are incorrect and go off the scale (disappear) towards the lower % amplitude. This worked correctly in 1.22.

-JD

Hi
Can you post both chc files that show this problem ?

Ki_,
Attached are the 2 chc files and one jpg of what I see on the non-reference RGB Levels graph. You can see the Reference Measure dashed line only on the right (where it is within the scale). I set the i1 as the reference for the DTP94 measures. Please excuse the sloppy low amplitude measures (I had averaging off).

Thanks,
JD

EDIT: If I check the "Do not use Luminance in DeltaE" preference, the Reference Measure dashed line looks correct as it did in 1.22. ?

 

RGB_reference.zip 85.7431640625k . file

I have a question on the HCFR Calibration Disk Ver. 2. On the HCFR downpage, it says the "iso image" is 60MB. I have downloaded the file successfully, but the size is only about 15MB? Is my download corrupted or it is actually 15MB?

JM

Your download must have timed out - it is 63MB.

jdbimmer, thanks for the measurement.
It seems to be a bug. We'll look at it quickly.

What is the "set reference" option?

Some suggestions which makes free measures easier:
1) in the main measurement window, include the DeltaErrror underneath the three colour bars.
2) in the combined graph for free measurements, there are scale figures to the left. Put these scale figures to the right as well since when you have done some measurements, the new values appear from the right side, scrolling left.
3) in the combined graph for free measurements, implement some scale adjustment for Luminance as well.

This might just concern me but...
on the DVD with test images, make the window patterns slightly larger and move them slightly to the right.
Now my (lefthandside oriented) OSD menu for Gain/Bias almost covers the small rectangle and positioning the probe is kind of tricky.
Maybe supply two different sets of patterns: one for OSDs on the left and one for OSDs on the right...
Edit: or maybe display one small window rectangle to the left and another one to the right or does that defeat the purpose of using windows??

Thanks!

I have also experienced this on my RS1/Radiance combo. The magenta dE number does not seem to correlate to the xyY value.

White Y = 74.52
Magenta:
x = .323
y = .154
Y = 20.523
dE = 22.5
delta xy = .002
delta luma = -3.3%

Mike

What is the difference between the LT and the "regular" model?

What is a DTP-94 or a Monaco X-Rite Optix XR ?

thx

bob

The dE calculation for the secondaries in v.2 looks broken to me. It returns numbers hat are MUCH too high. The primaries seem OK.

It is really very minor. Thanks,
-JD

Hi Tom

I confirm your diagnostic. Delta E calculus using luma is wrong for secondaries. It is right for grayscale and primaries. I already fixed it for next release, with the other issues.

Sorry for inconvenience.
Georges

Using YdE, my yellow and cyan dE are not far off of the expected, if at all. Like Mikela, magenta is the one that is way out of bounds. That's one reason why I missed it before, I didn't check all of them. If yellow and cyan dE calculations are off, it doesn't seem to be by much.

i.. ehh... have a little problem here

testpatters
use - & + (numpad)

well .. i don't have a numpad

Hello jvincent

The entries in .ini file are the same, you can "hack" dark reading average mechanism to average many readings on any level for Eye One.

[EyeOne]
MinRequestedY=10
MaxReadLoops=8

Above are default values when those lines do not exist. The program loops until the sum of all measured Y is at least 10 (or any other value), or until the number of measures reaches 8 (or any other value). It you put a very high value in MinRequestedY, like 1000000, you will always have the number of measures defined in MaxReadLoops, even for high level readings.

The exact same entries exist under sections [DTPSensor] and [SpyderII], with the same meaning, but for the other probes

Georges

Thanks for confirming that the feature still exists. I like to use it (I use 3 reads) for all light level readings just to average out some of the measurement to measurement differences.