CalMan 5 Release Notes and Discussion - Page 9

what display types/light sources do you have?

2 LG 55LHXs, and 1 LG 55LV9500. (3 Years trying to get the 55LHXs calibrated - Calman 3, to 4, some ChromaPure, and finally 5 - pulled out most of my hair during this time - especially with comments that `They're suppose to have that Yellow tinge!' Went outside, looked at the nice Blue Sky with White Clouds - never saw any Yellow Tinge there!) Now with 5, NO Yellow Tinge - Bold Colours, wonderful Detailing, and B&W is B&W, without the aged look! Sign me as a Happy Camper!

Could it be you have just gotten better at doing the calibration and it has nothing to do with the software?

It's the software. As mentioned earlier in this Forum, someone, I believe from Spectracal, said the i1Pro gives more precise calibrations with Calman 5. As also mentioned previously, I could not calibrate the RGB Cuts in the Service Menu White Balance adjustments with Calman 4, only the Gains. Calman 5 I can. (Possibly because the i1Pro under Calman 5 is more accurate?) The profiling of the i1Pro to the Spectracal C6 results in almost the same readouts as the i1Pro. In Calman 4, it was like I was dealing with 3 different meters with 3 different results. The LG 55LHX is a very hard set to calibrate - just ask ChadB. In talks with him, he was going through some very tough procedures trying to get the proper results. Calman 3 - constantly bouncing values, Calman 4 - problems with the Red, plus Yellow Tinges. ChromaPure: Greyscale values on the mark, but rising Gamma from 10IRE to 100IRE. Finally had to calibrate using the Gamma screen. Calman 5 - no bounce, no Red problems, no Yellow Tinge when the calibation is done, CMS doesn't end with yucky looking PQ, and I finally have a great looking set. Are there problems, yes, but minor in comparison to past softwares used.
Please note: I calibrate the hard way, not with AutoCal.

good info.
I noticed different numeric values of gray scale in V4 and V5 using same meter and pattens.
Where you had poor results in V4- I have them in V5.
hmmm.
Done the hard way too.

That sounds like a quote from THE GODFATHER movie...

alright Derek, so I just joined "the family" as a C5 E member... whenever you get a chance drop me "an offer I can't refuse" (free Colorbox etc.)...

could you elaborate? I don't understand why one version of CalMAN would provide significantly different readings from another version. Aren't the meters just reporting the raw XYZ or spectral data they measure off the display? Does this only occur under a very specific set of conditions?

Guy I quoted said it.
There is a difference in meter numbers from V4 to V5.
He believes V5 is more accurate.
My 1 and Only cal with V5 on a cal'ed panel, gave me red and yellow
tinges with a wacky gamma.

Which selection in the Meter Menu did you take?
Was recommended to use the Wide Gamut (White LED), but found it drove my i1Pro wacky. Dropped back one on the Menu to just LCD Direct View (White LED)
Edited by p5browne - 10/3/12 at 3:42pm

I'm just beginning to use v5. Downloaded 5.0.3.9something and was lookkng to see if I will be able to create a data table with the correct data rows in the order I need them to be in so the data can be copied and pastee into a spreadsheet.

Opening a data table is fine. Deleting rows works fine. But when I try to Add Row... nothing happens. I highlight a row, right click, the pop-up menu appears, I select "Add Row" and nothing happens. In v4, when you do that, you get a second pop-up with all the row/data options, you select the one you want and it is inserted into the data table. Is not being able to add a row a bug or am I doing something wrong?

You can export data directly to excel in CalMAN 5, right click the tab and you should have the choice to setup exactly what you need to export out.
These choices save an load with the application, so you're preferences will stick for working with any workflow.

I would use the properties panel for the data grid, that lets you easily select, remove and re-order rows (the same control is part of the export data window).

Can you export/import data from one workflow to another in v5 Enthusiast? Such as from the Enthusiast Workflow to the Quick Analysis Workflow?

Import is not yet supported.

not sure if this already on the list of being addressed in one of the future updates, but would love to see a simple feature in the software that when you change the dE formula on a graph that it also updates the number on the section (avg dE - max dE) to that particular dE formula...

current default is dE2000, and if u change it to dE1976 on a graph the bars change but not the avgdE / max dE values, they still show dE2000 values... this behavior seems to be consistent in the Quick Analysis workflow and Color Cube 3D LUT workflow...

- M

That wouldn't be a simple thing to implement as each chart and field is discrete. I do have some ideas how we could handle that, but it would require a significant engineering investment.

If this would be implemented hopefully it would be as an option. One of the strengths of CalMAN is the discrete functionality of the data charts/fields.

Absolutely, if we added it, it would be as another choice you could make while designing the workflow, we'd never take away flexibility.

I was thinking about sections in workflows where you put a specific dE value on top of a specific chart so it is pretty easy what to update, examples:

1) Quick Analysis > Grayscale: CollectionAvg dE2000

I can change the dE display through the properties inspector to dE 1994, it updates the dE 2000 text string (top of the chart) to dE 1994 I can even edit the on screen text "CollectionAvg dE2000" to whatever I like through the string editor in the properties panel... it also updates the bars when I change dE formulas, and since that means you have stored the data you could simply run the selected dE values through the corresponding dE formula and update the CollectionAvg dE2000 string to the selected dE formula and then output the average dE value right behind it... optionally (or additionally) you could add another column at the luminance table at the bottom and make the last column "Avg" and display the values there for all selected dE formulas...

2) Quick Analysis > Color Checker: 4th chart at the bottom (dE 2000)

is another example where this could be done real simply as all data is there, the values are just not updated on screen when you change the dE formula...


I personally prefer numbers as reference values to keep track of improvements rather than looking at a chart and trying to figure out which bar is closer to dE 3...

I like Calman 5 better than 4 so far and I am getting better repeatability results with a C6 for some reason.

One potential suggestion for the GUI to make it less tedious (and actually there are many different ways to accomplish this):
On the grayscale and gamut color readings, it would be less tedious if instead of having to click the mouse or direct shortcut keys to change which reading you are about to do (like next from 20IRE to 30IRE), then it would be nice to have a NEXT hotkey or make a quick-link bar instead of those large buttons. Reason is because many of use our laptops' pointer without mice, and it is annoying to have to move the mouse back and forth and click stuff. Even a drop-down box as an alternate might be good so we can cycle through quickly with the keyboard from gray-scale straight through the gamut in the same order as the AVS test disk.

I might suggest following a standard bottom quick-bar GUI (a customizeable quick-access bar at the bottom), like how in an MMO game you would pick a spell, as well as a key that cycles through those choices. This could almost be like a customizable "sub-flow" for each workflow on each form. Sorry to refer to a game for an example, but when it comes to GUI's sometimes games make a lot more sense than apps when you think about it.

Also, it would be nice if there was a way to run the final reports from only preliminary data on the calibration instead of needing to be completely done to get a good report. I know some of this stuff is not so much useful for pro calibrators, but it can be useful for reviews and snapshoting PDF's when we are in a rush and do not want to do a full calibration (hence we may want to emphasize one point in a review and use that in a PDF). I know I can just do screenshots of any screen, but it's still nice to have it the other way as well.

Just some thoughts (I haven't used it that much, so some of this stuff might already be there and I didn't look hard enough, so my apologies if I am not making strong points...)
Edited by coderguy - 10/5/12 at 6:23pm

CalMAN 5 keyboard shortcuts:

Alt +: // read 0% grayscale

Alt+F1: // read 10% grayscale

Alt+F2: // read 20% grayscale

Alt+F3: // read 30% grayscale

Alt+F4: // read 40% grayscale

Alt+F5: // read 50% grayscale

Alt+F6: // read 60% grayscale

Alt+F7: // read 70% grayscale

Alt+F8: // read 80% grayscale

Alt+F9: // read 90% grayscale

Alt+F10: // read 100% grayscale

Alt+Left: // previous source index

Alt+Right: // next source index

Alt+Up: // single reading

Alt+Down: // continues read

Alt+D: //Design mode


F4: // init meter

F5: // single reading

F6: // continuous read

F7: // low ActionString = "METER.READSINGLEGRAYSCALE(30)";

F8: // high ActionString = "METER.READSINGLEGRAYSCALE(80)";

F9: // stop

F10: //Read All

ESC: Stop

Too klunky with combo keys on laptops in the dark (remember we are in the dark), I like my suggestion better (sorry).

We've had most of these shortcuts in all the way back tor v3. They started as compatability with ColorFacts.

I'm not saying to remove shortcuts, I'm just wondering why there isn't a cycle-next option or a sub-flow quickbar on the bottom you can adjust. There are many reasons beyond what I mentioned, for instance I have a remote clicker I can use, but if there is no cycle option I cannot use it because the shortcut keys are directly tied to one setting instead of cyclable through all points. I know people with all the fancy pattern generators and automated stuff do not need any of this, but I think a lot of us don't have all that stuff.

I'm trying to work through a Cube calibration following the new sequence with 5.0.3.9whatever and a Radiance XE 3D processor. The Cube workflow defaults to dE2000 so I left it there. When I autocal with 21-steps, the grayscale dEs are 1.5 or lower. When I do the Cube CMS cal, the dEs are 2 or lower. Thinking everything was fine, I quit CalMAN 5 and on the way out it sets the Radiance to the 21-step ramp pattern. I was very surprised to see visible color in some of the steps... a green-tinted step at 85$, orange at 50%, magenta at 35%, etc. I'd say about 12 or 13 of the 21 steps look neutral while the others have colot tints that are visible in the 21 step ramp pattern. The visiblility of the color tints in the ramp pattern isnt't awful, but they aren't difficult to see - they are certainly more obvious than 'just barely noticeable' but not terrible. The colors in some of the steps is far more obvious than the dE2000 error numbers (1.5 or less) would lead you to believe.

Is there something about dE2000 that leads to these issues? Should I be using a different dE? I normally use dEuv for grayscale. If I change to dEuv, will that change the way the grayscale auto-calibration is done? In otherwords, will dEuv errors become the "new standard" for the grayscale autocal?

I can make the steps that have visible color tints look neutral if I manually tweak the grayscale but I need an easier way to make 0.1 adjustments to the DDC controls... the + and - signs make changes or .3 or .4. But I need a resolution of "0.1" in order to make the ramp pattern steps all look neutral. I can do that by dragging the DDC slider, but it's VERY jumpy. Is there some easier, more precise way to adjust DDC controls in 0.1 increments?

If I don't notice the tints in the grayscale steps until AFTER the Color CUBE CMS was done, I presume I would have to Reset the Cube from the DDC window before going back to tweak the grayscale adjustments? And once I was happy with the grayscale adjustments (confirm no visible color in any of the 21 steps), I could then do the Color CUBE autocal. Is that right?

Is it possible that once I have the grayscale steps all looking neutral and I run the color CUBE CMS then view the 21-step ramp again that I could see some steps with a color tint? If that happens, is there a way to fix the grayscale to remove the color tints?

I notice when doing the grayscale autocal, some of the messages I was seeing on-screen (wish I'd written them down, but I didn't) refer to the cube or matrix calibration. What's up with that? What does that mean in terms of what I can or can't do manually to fix problems the autocal leaves behind?

While waiting for some new panel display to show up for review, I'm working with a 2008 Samsung plasma. Is it normal to see 109% white fall off the orange guide line when you are using the workflow page (not in Cube, I think it was somewhere else) that has steps of 90%, 100%, 109% and several more between 100 and 109. The CalMAN/Radiance combo is sending fairly small window patterns... I guess I never checked 100%-109% on a plasma before to notice if the luminance limiting plasmas do is measurable with small window patterns. I do not see one particular color fall off as you'd expect if one of the colors was "running out of gas" at the brightest steps... all 3 primaries remain parallel to each other,but the lines they make aboive 100% have slightly less slope that the orange line that indicates ideal luminance (2.2 gamma target for this panel).

I think that's all the questions so far - the time investment in v5 is starting to get significant so I need to get to the point where this new calibration method starts to sink in -- thanks in advance for any insights that will help accelerate the learning curve.

Doug, what display were you calibrating?

No you shouldn't be seeing visible errors, my guess is if you remeasured you'd see larger errors that would indicate there is an issue.
Also you can flip back and forth between dEuv and dE2000 in the chart. You will undoubtly see 5-15% have higher dEuv, but their should be minimal difference in the dEuv and dE2000 for 50% up.

In other words the problem you saw was unrelated to dE formula.

The next update, not todays, will have 0.1 increment steps for the incrementers. Also todays update did make some other adjustments to the usability of the DDC controls. Personally I use the text entry method.
That's about right, you might be able to tweak after the cube is done, but the steps won't always line up so a 75% pattern after it comes out of the cube might only be 72% going into the grayscale controls so you'd be in a situation where both the 70% and 75% control have equal effect over the pattern.

Early lumagen cube firmwares had serious issues with this. Make sure you are running the latest firmware. The interpolation effect on grayscale has been minimal for the last several builds.
I'm not sure exactly what you saw, but the messages you get during the autocal are all just status messages describing the procedure (as long as you get a completed successfully popup at the end).

It's totally normal to get a little droop as it runs out of steam going to 109%, as long as the lines are tracking in parallel you aren't getting a color shift so it's too big of a deal if they are off gamma by a bit. In fact I believe we have a DeltaL chart underneath to show precisely how much luminance error is being measured.

For the time being, I'm using a circa 2008 Samsung A550 plasma. It can be made fairly accurate with the internal controls and nearly perfect with manual Radiance calibration using 21 step grayscale, 2.2 gamma, and 6-point CMS. 100% blue and 100% red are both SLIGHTLY displaced towards green (the measurement dots still touch the white reference square at the blue and red points, but the dot is not centered in the white box) so you get get them "perfect". But you can get all the color luminances accurate with the internal (or Radiance) controls.
Edited by Doug Blackburn - 10/6/12 at 10:32am

Thanks for the info, Joel. I am using 091512 Firmware in the Radiance. I see 091612 is avaialble now, but it indicates only a bug fix for copying a cube CMS to other CMSes so I think I should be up-to-date-enough with the Radiance firmware that it's not involved in the residual colors I saw in the ramp pattern.

Next time I get into the calibration, I'll make some more detailed observations and record the on-screen messages I saw during grayscale cal that referred to cube/matrix calibration. I also need to see if there are residual colors in the ramp pattern after grayscale cal and before the CMS Cube cal. At least now I know what the tripping points are and can be more aware of where a problem might crop up as I move through the steps.

I did save the last calibration session so I can use another history tab to re-measure grayscale to see if that result produces different dEs than the original grayscale calibration.

The "problem" is: I can also see exactly the same issues with my plasma TV, CalMAN 5 and my Radiance XE. The "funny" thing: with FORMER firmware versions of the Radiance (BEFORE CUBE was implemented), the grayscales looked perfect on my plasma TV. I could see that issues after updating the Radiance XE to new CUBE firmware. So I think it's the price we have to pay for such a small cube. Ususally you should use at least a cube size of 16. The Radiance uses 5 ... IMHO too less.

That's true. But also the newest firmware version has issues with that. You can easily check it by yourself: use the grayscale/gamma autocal with a Radiance firmware version WITHOUT cube support and compare it to the newest one WITH cube support. Then you can easily see that the garyscales / gamma look better with the older Radiance firmware.

100% agree !!! You made my day I am "struggling" about this with sotti every new CalMAN version. I know about Dereks' and sottis' explanations, but it's a truth, that I can calibrate nearly perfect grayscales when using the remote controls: Harry told so, you told so, some other guys too. The autocal doesn't produce such perfect results. Maybe the 0.3 steps are the reason for that