[Bluescale]

Quote:

Originally Posted by TheDac I think you're right. Your observations match my tests. Due to my setup, I'm only able to test on one of the plasmas right now, but considering it's a VT25, I'll venture to say these observations are accurate.

Thanks for testing. It's good to know that it's not just me or my set. I have an ST30, so it's not surprising we are seeing similar behavior. The Display3 flys through the calibration with extremely stable readings. Tomorrow I'm going to try my SXRD upstairs, just for the heck of it.


After looking at the bars jump around long enough, I was able to get a decent instinctual feel for compensating for the jumpiness. Attached is the calibration I completed a few minutes ago. Interestingly, the measurements for the primaries and secondaries were very stable, so I was able to dial those in pretty quickly. Note that the color measurements are using 75IRE color patterns.


All patterns were generated from my video processor. Since my processor only generates windowed patterns, I'm going to take another pass shortly using AVSHD to generate APL patterns to see what sort of results I get.

 

Full Calibration - Duo 1st Pass.zip 6.1142578125k . file

 

[Bluescale]

And here is my first attempt using the APL patterns on HCFR

 

Full Calibration - APL 1st Pass.zip 4.8193359375k . file

 

[JohnAd]



Quote:
Originally Posted by TheDac /forum/post/21724743


First off, a big Thank You! to John. I've gotten an i1Display Pro recently, and I was quite frustrated with the apps available (or not available) for it. Reactivating the project is awesome !

Firstly welcome to the forums

That has to set some kind of record for best first post.

Quote:
Originally Posted by TheDac /forum/post/21724743


Now on to my initial experience with the new version of HCFR. I'll be a little verbose, and hopefully that will help John with the development, so bear with me here.

Indeed it does. I'll reply to the relevant parts but the high level of detail is very useful on working out what's going on

　

Quote:
Originally Posted by TheDac /forum/post/21724743


I have 2 colorimeters: a DTP-94 that I've had for a very, very long time, and an i1Display Pro retail (i1D3).

I suppose it's inevitable that multiple meters might have issues, but very useful to test with this config.

Quote:
Originally Posted by TheDac /forum/post/21724743


I started the program, with both devices plugged in, selected the Argyll device, and I suddenly got the "Incorrect Driver - starting communications with the meter failed with severe error" message. Both devices were running with the XRite drivers. I went back to the sensor selection screen, selected the DTP-94 device, and voila, the program started up just fine. Took a full set of readings, and everything looked OK.

Here's what I think happened, the argyll driver spotted the 2 meters and assigned port 1 to teh DTP-94 and port 2 to the i1d3, when you started the meter on port 1 you got the incorrect driver error on teh DTP-94 but at thios point had you selected port 2 the i1d3 should have worked.

Quote:
Originally Posted by TheDac /forum/post/21724743


But, my interest was with the i1D3. So I installed the Argyll drivers, first just for the i1D3, then for the DTP-94 as well. I tried all possible combinations between them, but the result was the same: HCFR would work fine with the DTP-94 on both drivers, but it wouldn't work with the i1D3 on either driver.

This is odd, were you playing with the meter number at this stage? When you say the DTP-94 was working was it working via the argyll driver or the HCFR one?

Quote:
Originally Posted by TheDac /forum/post/21724743


so I decided to tinker with it. I commented a couple of lines of code where the program is bypassing HID devices (lines 80-81 in usbio.c):

There may well be an issue with trying to use the i1d3 with the supplied drivers. We may need to look into this later, for now use the x-rite drivers.

Quote:
Originally Posted by TheDac /forum/post/21724743


The i1D3 has some slight variations - I think due to the speed it reads the patches - I timed it at exactly 10 seconds for the entire grayscale, primaries and secondaries set. That's 17 patches in 10 seconds, just over half a second for each. The variations were just slight in my case, so I don't worry about it. Of note is that the first 3-4 gray patches were considerably slower, just over a second, while the light grays went by very, very fast. The DTP-94 was consistent as well (though wrong, but that's a device problem).

Yes the latest code for the i1d3 changes the way it calculated the integration time at lower light levels which should give better results at the expense of slightly longer read times.

Quote:
Originally Posted by TheDac /forum/post/21724743


1) if you use an i1D3, make sure you disable "Confirmation message" under Advanced -> Preferences -> Advanced. The device will not wait for your confirmation, and it will start reading, giving errant readings. By the time you click OK, the generator has almost already finished going through the grayscale.

Sounds like a bug, thanks

Quote:
Originally Posted by TheDac /forum/post/21724743


2) in a couple of instances, the Argyll meter selection did not recognize the i1D3 because the cap on the lens was on. In my case, I usually keep it on until just before I run the process. The odd thing is that I can't replicate the issue. It happened a couple of times, and then it just went away.

Keep and eye out for this happening again, the cap should not affect recognition

　

Quote:
Originally Posted by TheDac /forum/post/21724743


3) if you have multiple meters on your computer, you may get strange problems. In my case I was running at times with both devices attached. The Argyll selection screen would not populate the device name after it was selected, and you might not know which device is selected, since the selection screen only displays the meter number, not

the name.

Yes, there are problems with multiple meters in 3.0.0.0, this will improve...

Quote:
Originally Posted by TheDac /forum/post/21724743


4) as I just mentioned the Argyll selection screen does not always populate the name of the device once you make your selection and close the window. I was able to consistently replicate that. I would go to the selection screen, select meter 2 (the i1D3), click OK, and the Sensor would show just "Argyll Meter" no matter how long I waited. I would go back in the selection screen, simply click OK, and now the name would appear under sensor, i1Display3. The meter was still selected the first time around, and it was working fine, but the sensor data was just not being updated.

The way detection works will change a lot in teh next version, you should select the meter by name from the list.

Quote:
Originally Posted by TheDac /forum/post/21724743


5) I have an odd issue when both meters are running on the XRite drivers, and both are plugged in. I suppose this is due to the enumeration of the devices in the original list. Here's the issue: the i1D3 is assigned meter 2; meter 1 seems to be assigned, but it generates the "Incorrect Driver - starting communications with the meter failed with severe error" message, if selected. If I unplug the DTP-94, the i1D3 is assigned to meter 1, and it works fine. I am speculating that somehow Argyll still detects the DTP-94 as a valid device when it's connected, but it can't initialize it with the XRite drivers.

Yes it does, if you install the supplied drivers for this device it should work, can you try with just the DTP-94 plugged in and with the supplied drivers for that and selecting argyll from the meter list (not the dtp-94)

　

Quote:
Originally Posted by TheDac /forum/post/21724743


6) you CANNOT have 2 active calibration windows with the same meter selected. The original calibration window will work fine, but the second window will give you the "Incorrect Driver ..." message

Good point, I'll need to fix this for the next version.

　

Quote:
Originally Posted by TheDac /forum/post/21724743


7) the "Argyll Meter Property Page" screen (sensor selection screen) will default to the previously selected meter number. If you plug/unplug meters, or unplug them and then plug them in another USB port, the assigned meter number can, and most likely will, change. If you ever selected a meter higher than 1, the screen will never default back to meter 1, unless you specifically select it.

That should be fixed in the next version the who meter number thing will be hidden from you.

Quote:
Originally Posted by TheDac /forum/post/21724743


10) with the i1D3, if you get the message "Meter is in incorrect position", it means that the cap is on the lens. You need to open it up before you take any readings

Good to know thanks.

 

[JohnAd]



Quote:
Originally Posted by Bluescale /forum/post/21726167


That's the display I have, and any readings at 40 IRE and below are completely unstable. My theory is the speed of the display3 coupled with dithering on the plasma is causing the problem. There needs to be some sort of algorithm to compensate for that.

There are some changes to the way the integration time is calculated in the newer code from ArgyllCms, we'll see if his makes things better, if not then there are a few things we can try tweaking.


It's also worth pointing out that support for using the plasma adjustments is in the pipeline but probably won't be in the next version maybe the one after that.


John

 

[Bluescale]



Quote:
Originally Posted by JohnAd /forum/post/21727624


There are some changes to the way the integration time is calculated in the newer code from ArgyllCms, we'll see if his makes things better, if not then there are a few things we can try tweaking.


It's also worth pointing out that support for using the plasma adjustments is in the pipeline but probably won't be in the next version maybe the one after that.


John

That's good to know, thanks.

 

[vega509]

glad to see this project resurrected.


the i1d2 needs to use the argyl driver, install it from device manager.


the CIE diagram in Windows 7 64bit still has no secondary reference points.


thanks John

 

[JohnAd]



Quote:
Originally Posted by vega509 /forum/post/21728349


the i1d2 needs to use the argyl driver, install it from device manager.

Thanks for that, I've been testing with the i1d2 as I 've got a few duff ones lying about the place, there is a chance that it will work with the filter libusb drivers which will mean that you can keep the x-rite driver but still use the argyll code, it needs abit mores testing though.

Quote:
Originally Posted by vega509 /forum/post/21728349


the CIE diagram in Windows 7 64bit still has no secondary reference points.

Give me a chance

It's on the list....


John

 

[703]

How am I supposed to install the included drivers? Do I need to replace windows drivers installed by x-rite with them? (i.e update drivers under device manager).


If so, does it mean the x-rite applications may no longer work correctly? Such as EyeOne Match?


I assume ArgyllCms doesn't work the same way as the old how HCFR originally access the device, just via a .dll dropped in the main dir?


Thanks

 

[JohnAd]



Quote:
Originally Posted by 703 /forum/post/0


How am I supposed to install the included drivers? Do I need to replace windows drivers installed by x-rite with them? (i.e update drivers under device manager).


If so, does it mean the x-rite applications may no longer work correctly? Such as EyeOne Match?


I assume ArgyllCms doesn't work the same way as the old how HCFR originally access the device, just via a .dll dropped in the main dir?


Thanks

For most meters at the moment yes you'll have to replace the drivers, the i1d3 is the main exception. And as you point out the x-rite software won't work until you put the x-rite drivers back again. There are a couple of ways I'm looking at to avoid this driver dance but they are not ready yet.


As a small bonus you can now try out argyllcms for display profiling as we share drivers with that.


John

 

[wl1]



Quote:
Originally Posted by JohnAd /forum/post/0



For most meters at the moment yes you'll have to replace the drivers, the i1d3 is the main exception. And as you point out the x-rite software won't work until you put the x-rite drivers back again. There are a couple of ways I'm looking at to avoid this driver dance but they are not ready yet.


As a small bonus you can now try out argyllcms for display profiling as we share drivers with that.


John

Thanks for re-starting this. I have got my i1D2 responding in your new version. It has a few things different to how I recall the hcfr got left, but maybe I got left behind at the very end. (eg patterns available via tabs).


I didn't have time yesterday, but if I do a Plasma calibration, is the software going to give me good results? As in, no worse than old hcfr in its processing/calculations?


I don't want to make changes which make it worse, though I can verify the software etc without making changes - it depends on your software status.

 

[zoyd]

Doing some testing today, a couple of quick notes on installation (32 bit x86 winxp).


- i1pro argyll driver installed with no problems


- D2 was missing the Wincoinstall*.dlls so I had to install those from my argyll package


- Both i1pro and D2 argyll devices were recognized by HCFR and readings from both are stable.


- Can't tell if hi-res mode on i1pro is working (spectrum graph is blank)


- Every time I start a measure there is a pop-up "save calibration data". I don't recall this as normal, can I turn it off?


- If I start a continuous measures with one probe I can't stop it with the run/stop button. This button stays as green right arrow. I can press it again and it will restart the measures or I can switch to the other probe and start measures there.


Will do some profiling on my plasma in a bit.


Thanks for the new toy!

 

[JohnAd]



Quote:
Originally Posted by wl1 /forum/post/21731198


I didn't have time yesterday, but if I do a Plasma calibration, is the software going to give me good results? As in, no worse than old hcfr in its processing/calculations?

It should behave the same at the previous versions n, the options to average lots of reading have gone but I'd expect the numbers returned from the meter to be the same.


John

 

[JohnAd]



Quote:
Originally Posted by zoyd /forum/post/21731548


Doing some testing today, a couple of quick notes on installation (32 bit x86 winxp).


- i1pro argyll driver installed with no problems

Great, thanks for the i1pro report

Quote:
Originally Posted by zoyd /forum/post/21731548


- D2 was missing the Wincoinstall*.dlls so I had to install those from my argyll package

Yes missed those in the installer, thanks for spotting that.


- Both i1pro and D2 argyll devices were recognized by HCFR and readings from both are stable.

Quote:
Originally Posted by zoyd /forum/post/21731548


- Can't tell if hi-res mode on i1pro is working (spectrum graph is blank)

Probably doesn't make much difference without the spectrum mode, that will be in probably the next but one release.

Quote:
Originally Posted by zoyd /forum/post/21731548


- Every time I start a measure there is a pop-up "save calibration data". I don't recall this as normal, can I turn it off?

Yes, probably something odd in the way I set up the new meter, I'll look into it.

Quote:
Originally Posted by zoyd /forum/post/21731548


- If I start a continuous measures with one probe I can't stop it with the run/stop button. This button stays as green right arrow. I can press it again and it will restart the measures or I can switch to the other probe and start measures there.

Hmmm, not seen this, probably some odd issue with multiple meters, I'll see if I can reproduce.


Thanks


John

 

[zoyd]

Thanks John. So far both meters are working great on argyll drivers with HCFR. I ran the i1pro through grayscale and primaries/secondaries on my plasma and results are the same as the x-rite drivers. Profiling the D2 against the i1pro also works fine. The hi-res switch is working as well, I measure a consistent deltaE shift of 2.3-2.5 for red and and 1-1.2 for green. This is the same thing I measure using the argyll package in hi-res mode.


Two things that are different with these drivers:


1. i1pro 0%-10% stimulus readings are more stable than the x-rite drivers. Are integration times longer?


2. The D2 will not return readings below 0.028 ftL. Below that it returns nothing so I'm guessing there is a math precision problem. The x-rite drivers will return values down to 0.005 ftL.


The previous problem I mentioned about not being able to stop free measures was user error.

I didn't have the correct window focus...


edit: The D2 problem also appears using spotread.exe at the command line. If I try to read a dark pattern it returns "Spot read failed due to a communications error"

 

[smokarz]

hi zoyd,


i just bought the i3 display pro from amazon. should be able to do some testings by tues/weds.


would you suggest i install the x-rite driver or the argyll driver for the display 3?


thanks

 

[JohnAd]



Quote:
Originally Posted by smokarz /forum/post/0



would you suggest i install the x-rite driver or the argyll driver for the display 3?

stick with the x- rite one


John

 

[smokarz]

thanks John.


i looked at the x-rite website but didn't see a download for device driver.


i guess i just install the driver that would come with the packaged cd?

 

[Bluescale]



Quote:
Originally Posted by smokarz /forum/post/21732436


thanks John.


i looked at the x-rite website but didn't see a download for device driver.


i guess i just install the driver that would come with the packaged cd?

That's what I did. Then the software prompted me to install an update. You only need to do the update if you want to use the profiler on yor computer monitor.

 

[Bluescale]



Quote:
Originally Posted by wl1 /forum/post/21731198


Thanks for re-starting this. I have got my i1D2 responding in your new version. It has a few things different to how I recall the hcfr got left, but maybe I got left behind at the very end. (eg patterns available via tabs).


I didn't have time yesterday, but if I do a Plasma calibration, is the software going to give me good results? As in, no worse than old hcfr in its processing/calculations?


I don't want to make changes which make it worse, though I can verify the software etc without making changes - it depends on your software status.

With the i1D2, I get very stable results on my plasma.

 

[zoyd]

@John: I found the problem with low level light readings and the D2. Apparently there is a timeout problem at low levels with some versions of libusb. I reinstalled the one packaged with argyll all is well. ( see this )