or Connect
AVS › AVS Forum › Display Devices › Display Calibration › HCFR - Open source projector and display calibration software
New Posts  All Forums:Forum Nav:

HCFR - Open source projector and display calibration software - Page 86

post #2551 of 3436

Hi,

 

the "unsupported function" message is gone which is already great.

But clicking the calibrate button nothing happens.

I add a new calibration sheet, open the property page and click calibrate -> 

the following is written to the log:

Argyll Debug - dtp92_fcommand: command '0216CF^M' returned '<00>', value 0x0

 

i click it a second time and this is written to the log:

Argyll Debug - dtp92_fcommand: command '10103RM^M' returned '<02>', value 0x2

Argyll Debug - dtp92_fcommand: command '0216CF^M' returned '<00>', value 0x0

 

Dont know if that helps you.

 

As i am not an expert, is it right to have the DTP94 face a total dark spot, like my desk when calibrating? That's what the old driver asked me to do.

Also is the temperature option not supported in the argyll driver?

 

Thank you for your efforts again!

post #2552 of 3436
Thread Starter 
Yes, that will help me debug. Does the same thing happen for both crt and lcd display modes? And to clarify, you see no message boxes at all after clicking calibrate? You should at least see "No calibrations needed"

btw, the dark current temperature drift correction is in the meter firmware so it's always "on"
Edited by zoyd - 11/9/13 at 5:30am
post #2553 of 3436

i just checked it. yes, the same for crt, lcd and generic modes. No message boxes. When clicking the calibrate button the options window doesn't even go to "not responding" state. Sometimes i can see the mouse arrow flickering once for a fraction of a second as if it did something.

 

Again to clarify that im doing it right. When i click the "calibrate meter" button, what is the driver expecting me to do with the DTP94? facing some absolute dark area or some test pattern?

post #2554 of 3436
Thread Starter 
Quote:
Originally Posted by unknown577 View Post

i just checked it. yes, the same for crt, lcd and generic modes. No message boxes. When clicking the calibrate button the options window doesn't even go to "not responding" state. Sometimes i can see the mouse arrow flickering once for a fraction of a second as if it did something.

Again to clarify that im doing it right. When i click the "calibrate meter" button, what is the driver expecting me to do with the DTP94? facing some absolute dark area or some test pattern?

The program will give you instructions if needed but I need to figure out why you are not getting any pop-up boxes at all. Like the D2 I don't think the DTP94 requires any calibration to perform properly.
post #2555 of 3436
Thread Starter 
Quote:
Originally Posted by unknown577 View Post

i just checked it. yes, the same for crt, lcd and generic modes. No message boxes. When clicking the calibrate button the options window doesn't even go to "not responding" state.

Can you save your colorhcfr.exe and replace with the one attached and try again. Let me know what messages are display when you press calibrate using Display type CRT Display and Display type LCD Display thx. ColorHCFR.zip 1990k .zip file
post #2556 of 3436

i click "calibrate meter" -> message box "calibration instructions" "place cap on instrument put on dark surface or white ref" -> i click ok -> takes a while -> "refresh measurement perform..." "no discernable refresh rate found" -> i click ok -> "calibration complete" "device is now calibrated. if the device requires it return to the correct measurement position".

 

the same for crt, lcd and generic

 

FYI: need to leave now. will be back tonight and check if you want me to test anything else.

post #2557 of 3436
I just took the new build for a test drive AND see zoyd it was never about the spectrals during the beta testing it was the low intergration time that made things funky.

I noticed with the munki .05 -1. sec work best.But i'll leave it parked in 1 second for better low reading stability.

Thanks for your time and effort.
post #2558 of 3436
Thread Starter 
Quote:
Originally Posted by unknown577 View Post


FYI: need to leave now. will be back tonight and check if you want me to test anything else.

Thanks for the report, try the attached. The expected behavior for LCD mode is ask for a dark calibration and finish and for CRT mode first a dark calibration and then a refresh mode with the message "Provide and 80% gray patch". Let me know how it goes.

I did a little more research on the dtp94 and yes they do need period offset calibration to get the best dark measurements, so when we get it working you should do the dark calibration every 15-20 minutes.

ColorHCFR.zip 1989k .zip file
post #2559 of 3436
Thread Starter 
I was able to borrow a colormunki spectrometer for testing and it did very well compared to my i1pro 2, also very nice for printer profiling.


post #2560 of 3436
Quote:
Originally Posted by zoyd View Post

I was able to borrow a colormunki spectrometer for testing and it did very well compared to my i1pro 2, also very nice for printer profiling.

Cool, hopefully they all roll out of the production line that accurate. I have no way to verify if mine is accurate or not, hopefully it is and there isn't much variance unit to unit. Zoyd, have you found that you need to calibrate it every 15-20m by providing a black patch or is it not needed for this meter? (I usually just leave it on the screen and turn the dial to calibration and also close the sensor slide door) One other question, how reliable and trustworthy are the low light reading? I read it's not too reliable sub 5 cd/m2 which would mean anything under 30% gray.
post #2561 of 3436
Quote:
Originally Posted by Barsk View Post

Sorry, I have found these options now. I used the drop down list which covers most options, but not all...

If I can suggest something from these findings then it would be that it would make better sense to have the primaries and secondaries sat sweep in the droplist instead of these separate six we have now. Simplifies the droplist and pulls in what you most usually need. My 2 cents.

This post was about changing the GUI in respect to selecting saturation sweeps. I was dead wrong about this. The droplist is for choosing which datatable to view and the menus are ther for choosing which to run or in cases a combination of them. So leave it as it is, as it is working just fine now and making good sense. Sorry about the confusion.
post #2562 of 3436
Thread Starter 
yeah, I thought I'd let you figure it on your own. smile.gif quick tip: if you want quick access to all the saturation sweeps check Extended measures and Saturation measures toolbars from top menu View
post #2563 of 3436
Thread Starter 
Quote:
Originally Posted by fairchild99 View Post

Cool, hopefully they all roll out of the production line that accurate. I have no way to verify if mine is accurate or not, hopefully it is and there isn't much variance unit to unit. Zoyd, have you found that you need to calibrate it every 15-20m by providing a black patch or is it not needed for this meter? (I usually just leave it on the screen and turn the dial to calibration and also close the sensor slide door) One other question, how reliable and trustworthy are the low light reading? I read it's not too reliable sub 5 cd/m2 which would mean anything under 30% gray.

That's the advantage of a spectrometer, it doesn't matter if the hardware has different sensitivity unit to unit, all that matters is that the in-house calibration process is consistent and that is a lot easier than producing matched optical performance for colorimeters.

People tend to exaggerate the problems with low light measurements - I've looked at the low-light accuracy of the colormunki spectro and it's perfectly fine down to 1 cd/m^2 for grayscale measurements. I do this by profiling a D3 against the spectro and then comparing the measurements of the two using a 10% pattern which is just under 1 cd/m^2 on my display. The measurements agree almost exactly:

xyY
D3 0.315, 0.324, 0.977
CM 0.316, 0.325, 0.975

The CM red channel has a small amount of noise at that level, +/- 1%

I had previously come to the same conclusion regarding the i1pro here.
Edited by zoyd - 11/9/13 at 11:41am
post #2564 of 3436
Cool thanks as always for the insight. I guess making changes to the 10 IRE w/b and gamma will be pointless on my VT60 as that might be reading too low. My 10% read at 0.69 cd/m2 on my P50S60 using 80 contrast in Cinema.
post #2565 of 3436

i checked your new executable.

 

now it looks like this:

 

i click "calibrate meter" -> message box "calibration instructions" "place cap on instrument put on dark surface or white ref" -> i click ok -> takes a while -> "calibration complete" "device is now calibrated. if the device requires it return to the correct measurement position".

 

The same for all three modes.

 

What do you mean by ".. and then a refresh mode with the message ... "? What is a refresh mode?

 

 

Seems to me like everyone uses a i1 display pro by now otherwise someone would have come across this bug already. maybe i should also get one hehe.

Is the d3 better than the DTP94? Does it need recalibration every x minutes?

And another general question, what's the advantage of creating a correction matrix for one of the mentioned colorimeters and perform the calibration only with it? What about just using the i1pro2?

post #2566 of 3436
Quote:
Originally Posted by zoyd View Post


If you use a display pro and want to use refresh mode in conjunction with with an edr file choose refresh mode and sync your probe using the calibrate button first, then choose your edr display type.

That shouldn't be needed. If you reload the .edr's using ArgyllCMS V1.6+ oeminst.exe, then the resulting .ccss's will set the refresh mode appropriately for the type of display. So plasma should set refresh mode automatically.
post #2567 of 3436
I was calibrating all day today and for some reason HCFR stopped reading color space.... It stopped showing color checker targets, saturation targets, color checker measures, saturations, etc... It just shows me a white triangle with D65 in the middle. Just a minute ago everything was fine and then suddenly it stopped showing! I restarted my PC, reinstalled HCFR, unplugged my colorimeter from USB, closed it, opened again, etc. HCFR shows gray scale just fine, but the CIE diagram is now empty, so I can't calibrate saturations!

What is going on?
post #2568 of 3436
Thread Starter 
Quote:
Originally Posted by gwgill View Post

That shouldn't be needed. If you reload the .edr's using ArgyllCMS V1.6+ oeminst.exe, then the resulting .ccss's will set the refresh mode appropriately for the type of display. So plasma should set refresh mode automatically.

What I needed to do to get the refresh calc part of the calibration flow (as opposed to on first measurement) was change the refresh calibration flag to a non-deferable one. I then call for instructions using inst_calt_available. In that scenario only setting the display type to refresh generates instructions to do inst_calc_emis_80pc. All the other display types including the crt and plasma edr based return inst_ok from m_meter->calibrate.
post #2569 of 3436
Thread Starter 
Quote:
Originally Posted by unknown577 View Post


i click "calibrate meter" -> message box "calibration instructions" "place cap on instrument put on dark surface or white ref" -> i click ok -> takes a while -> "calibration complete" "device is now calibrated. if the device requires it return to the correct measurement position".

The same for all three modes.

What do you mean by ".. and then a refresh mode with the message ... "? What is a refresh mode?

That just means the probe is not asking for a refresh mode calculation, it would only do this when display type is set to crt but don't worry about it. It looks like your dark offset calibration is working now, I will update the package.
Quote:
Seems to me like everyone uses a i1 display pro by now otherwise someone would have come across this bug already. maybe i should also get one hehe.
Is the d3 better than the DTP94? Does it need recalibration every x minutes?
And another general question, what's the advantage of creating a correction matrix for one of the mentioned colorimeters and perform the calibration only with it? What about just using the i1pro2?

The D3 is a good investment if you don't have an i1pro or other spectrometer, it's reasonably accurate on common display types using the spectral correction sets (.ccss) The advantage of profiling a colorimeter, especially the D3, is speed and low light accuracy.
post #2570 of 3436

Well if i set display type to crt it doesn't ask for what you mentioned. It is just the same like for lcd and generic display.

 

I have a brand new i1pro2. Would you recommend to sell the dtp94 and get a d3? Does that one need frequent recalibration?

post #2571 of 3436
Thread Starter 
Quote:
Originally Posted by unknown577 View Post

Well if i set display type to crt it doesn't ask for what you mentioned. It is just the same like for lcd and generic display.

I have a brand new i1pro2. Would you recommend to sell the dtp94 and get a d3? Does that one need frequent recalibration?

That depends on how often you calibrate and would use the D3. If you are only touching things up a few times a year what you have is fine. If you're calibrating a lot of different displays and using it constantly the D3 would shorten your sessions and it does not need to be re-profiled often. It also doesn't require frequent dark recalibrations.
post #2572 of 3436
Thread Starter 
Quote:
Originally Posted by MonarchX View Post

What is going on?

right click on the diagram and enable the display options you require. And don't ask me how you got it in that state, I have no idea.
post #2573 of 3436

I see, well i don't care if measuring takes 10ms or 2 seconds. What is important for me is accuracy. But not having to constantly recalibrate would be a nice to have thing.

 

So, you think it is correct that i dont see that "Provide and 80% gray patch" message in crt mode?

post #2574 of 3436
Thread Starter 
Quote:
Originally Posted by unknown577 View Post

I see, well i don't care if measuring takes 10ms or 2 seconds. What is important for me is accuracy. But not having to constantly recalibrate would be a nice to have thing.

So, you think it is correct that i dont see that "Provide and 80% gray patch" message in crt mode?

I am checking with Graeme on this point but in either case it won't affect your calibrations.
post #2575 of 3436

Ok, well my calibrations won't be affected anyway, cause i don't own crt TVs, plasmas or crt projectors.

I just want to make sure we dont overlook something, i just like to help you build a stable software.

By the way, great work!

 

Do you actually have some kinda bug tracking or ticketing system? And i was wondering if this is a closed software project or if it s possible for the community to help contribute to the code?

post #2576 of 3436
Quote:
Originally Posted by MonarchX View Post

How do I know if I did or did not use correction matrices?

I originally used HCFR 3.0.4.0 WITHOUT a correction file to calibrate my non-LED LCD HDTV. This time, I used version 3.1.0.0 and selected D3_GENERIC_CCFL_LCD correction file. My results with correction file were different from the results without correction file. My reds without correction file were at 100% at each grey scale %, but with correction file they were at 110% at each grey scale %. So, I re-calibrated my HDTV so that my reds are @ 100% for each grey scale %, utilizing correction file.

The result is that I no longer have that pinkish/reddish effect on my screen! However, I had to use my own eyes when calibration 0-10% because this version of HCFR does not provide a reading for it and even if it did, my colorimeter would not provide an accurate result.

I think the GUI needs to adress this problem. The fact that the user is presented with a choice of correction matrices at start and then in Argyll another set of choices confused the hell out of me for sure. Once I skipped the matrix in the first dialog and went for the ccss correction in dialog two instead things started to get right for me. The thing is, as already mentioned that you need to select ONE of three possible correction methods. The legacy HCFR one in dialog one, ccss correction in dialog two or correction matrix in a tab in dialog two. I think in order to make it intuitive and avoid pitfalls the GUI should only allow one to make ONE choice of these.

I do not see any reason why one would like to use a combination. Or?

I also suggest that the combining of the refresh/no-refresh inside the ccss select list violates standard GUI "guidelines" as how GUI's should behave. Or at lest how people expects them to. A select list is for selecting ONE thing and one thing only. If you need to select two different things - make two select lists.

I also suggest that calibrating of the sensor is made as a third dialog after one has selected all options in the Argyll meter dialog. If the sensor needs calibrating - then popup the "provide a 80% white patch" or whatever is needed according to selected meter AFTER the user has pressed Ok on the Argyll dialog. That will make the whole startup process a step-by-step or wizard-style thing that prevents chances of doing the wrong thing or in the wrong order regardless if you are a pro or a newbie with HCFR.

Other than that, I have been using the 3.1.0.0 version extensively last night trying to nail down my CMS issues with my VT30 . Did six full measurements with grayscale, primaries/secondaries and full saturation sweeps for 4% APL windows, 10% APL windows, and 10% windows, black background. This to learn how the TV responses to different pattern types and see how to best calibrate this non-linear beast! Then calibrated at 75% Rec 709 with a 75% A/75% S pattern and finally got very close to a perfect picture. I can report NO bugs found, and perfectly stable readings at low IRE throughot my whole session. Good job Zoyd and John!
post #2577 of 3436

Speaking of low IRE (sorry to bother you again with my DTP94b ^^), when using an old HCFR version with the none argyll driver i had a checkbox in the meter property dialog asking if i wanted to average reads an low ire conditions. I guess now that's covered by the "integration time" dropdown, right? But that option is not available to me (greyed out). Is this correct?

Maybe someone could clarify that for me.

Thanks in advance!

post #2578 of 3436
Thread Starter 
Quote:
Originally Posted by unknown577 View Post

Ok, well my calibrations won't be affected anyway, cause i don't own crt TVs, plasmas or crt projectors.
I just want to make sure we dont overlook something, i just like to help you build a stable software.
By the way, great work!

Do you actually have some kinda bug tracking or ticketing system? And i was wondering if this is a closed software project or if it s possible for the community to help contribute to the code?

No formal bug tracking, code is open source. If you are not interested in development but would like to contribute you could help update the wiki pages.
post #2579 of 3436
Thread Starter 
Quote:
Originally Posted by unknown577 View Post

Speaking of low IRE (sorry to bother you again with my DTP94b ^^), when using an old HCFR version with the none argyll driver i had a checkbox in the meter property dialog asking if i wanted to average reads an low ire conditions. I guess now that's covered by the "integration time" dropdown, right? But that option is not available to me (greyed out). Is this correct?
Maybe someone could clarify that for me.
Thanks in advance!

It's preferable to use the term "video level" instead of the now defunct "IRE".

This option went away with the update to Argyll-only meter code. It's on my list to add it back in for probes without adaptive integration capabilities.


3.1.0.1 is available

Changelog
post #2580 of 3436
Quote:
Originally Posted by zoyd View Post


No formal bug tracking, code is open source. If you are not interested in development but would like to contribute you could help update the wiki pages.

 

Nice, I just checked it out. I would like to help developing a bit in the future. Can you recommend any material to better get into the whole topic?

New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Display Calibration
AVS › AVS Forum › Display Devices › Display Calibration › HCFR - Open source projector and display calibration software