Profiling eye 1 disp pro with colormunki photo spectro - AVS Forum
Forum Jump: 
 
Thread Tools
Old 03-30-2014, 05:28 PM - Thread Starter
Member
 
Andrei_VVB's Avatar
 
Join Date: Mar 2010
Posts: 191
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 44 Post(s)
Liked: 16
Hi guys! I have an Eye1 disp pro. I might get a colormunki spectro and I would like to know if profiling with this will get me some benefits. My logic says that it will. I ask this because I will rent the munki so I want to be sure that this solution will be a good one for a DIY calibration(I can't hire a pro in my country).
Second, I would like to be helped with the exact procedure of how I do this profiling with the latest HCFR, position of the meters, what patterns etc with my Panasonic VT60 plasma.
And last, after profiling I use only the colorimeter for everything(including CMS) or only for wb, gama and spectro for CMS?
Thank you very much for your help!
Andrei_VVB is online now  
Sponsored Links
Advertisement
 
Old 04-01-2014, 06:17 PM
Newbie
 
tvnovus's Avatar
 
Join Date: Apr 2011
Posts: 4
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Andrei,

I use my "munki" to profile my display pro. Probably not as good as some of the other spectros, but should help to improve color. I don't use HCFR, but with calman and chromapure the software will make the corrections for you.

Once profiled, use the display pro for all your measurements. Lot of choices on the "best" patterns. I like both standard windows and APL in the 5-10% range. If you don't have a generator try one of the discs found here on the site.

ControlCal has also been a great help to me.

Good Luck!
Andrei_VVB likes this.
tvnovus is offline  
Old 04-01-2014, 10:18 PM - Thread Starter
Member
 
Andrei_VVB's Avatar
 
Join Date: Mar 2010
Posts: 191
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 44 Post(s)
Liked: 16
When I asked about patterns I meant what color and white to use for profiling. In Michael Chen's videos I saw 75% White and primaries for Calman and Chromapure. For HCFR it seems that I have to take some initial grayscale and colours measures and use that as a reference like someone advised me in the hcfr thread.
Thanks for your response!
Andrei_VVB is online now  
Old 04-02-2014, 04:07 AM
Newbie
 
tvnovus's Avatar
 
Join Date: Apr 2011
Posts: 4
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
I am not Mr. Chen, but it seems I saw somewhere here on the forum that the "colormunki" was less accurate at lower levels, especially blue. I have been using 100% white, red, green, blue. This helps assure a better sample for the accuracy range of the meter.

The Mascior disc has a great chapter for profiling.
Andrei_VVB likes this.
tvnovus is offline  
Old 04-02-2014, 11:50 AM
AVS Special Member
 
Michael TLV's Avatar
 
Join Date: Feb 2000
Location: THX/ISF Calibrationist/Instructor, Calgary, AB, Canada
Posts: 6,774
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 28 Post(s)
Liked: 138
Greetings

Even my mother does not call me Mr. Chen.biggrin.gif The police however ... is another story.

The limitations for the entry level spectrals comes in at stuff less than 30% ... But it depends on the device. 10% and 20% on some flat panels are plenty bright for the device.

doing a profile at 75% levels is just fine or else we have even bigger problems if 75% is a problem.

regards
Andrei_VVB likes this.

Michael Chen @ The Laser Video Experience
ISF/THX/TLV Video Instructor
The Video Calibration Education Hub - www.TLVEXP.com

Michael TLV is offline  
Old 04-03-2014, 02:01 AM
AVS Special Member
 
ConnecTEDDD's Avatar
 
Join Date: Sep 2010
Location: Athens, Greece
Posts: 2,433
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 131 Post(s)
Liked: 511
Quote:
Originally Posted by Andrei_VVB View Post

When I asked about patterns I meant what color and white to use for profiling. In Michael Chen's videos I saw 75% White and primaries for Calman and Chromapure. For HCFR it seems that I have to take some initial grayscale and colours measures and use that as a reference like someone advised me in the hcfr thread.
Thanks for your response!

Hello,

It doesn't matter what Stimulus Levels WRGB Patterns you will use during your meter profiling if you have a JETI / Photo Reaseach / Minolta.... but If you have i1PRO 1/2 or ColorMunki Spectro: these meters are returning chromaticity but they are not so accurate reading low luminance levels, for example Blue Pattern during the profiling @ Luminance Levels Below ~8 cd/m2.

i1PRO1/2 ColorMunki Spectro, is not so accurate on everything below ~10cdm/2..It will return to you chromaticity readings but it will be fake.

You can run a test to see where your i1pro1/2 doesn't return trusted readings.

Profile your colorimeter with using your spectro and run a 21-Step Grayscale with your spectro and with your profiled colorimeter.....you will see that at 50%-100% grayscale both meter will return to your similar readings but lower you will see that i1pro2 will not track well your colorimeter readings.....At the point you will see that problem is starting, check the Luminance level of that point and you will see at what cdm2 is the limit you can trust at the future

Examples / Why we have to always profile with 100% Amplitude patterns with i1pro1/2:

If you have a Display with White @ 120cd/m2 with Gamma 2.2:

If you perform profiling with 100% Amplitude Patterns then you will have:

100% White @ 120 cd/m2
100% Blue @ 8,6 cd/m2

If you perform profiling with 75% Amplitude Patterns then you will have:

75% White @ 63,5 cd/m2
75% Blue @ 4,5 cd/m2

If you have a Projector with White @ 14fl = 48cd/m2 with Gamma 2.2:

If you perform profiling with 100% Amplitude Patterns then you will have:

100% White @ 48 cd/m2
100% Blue @ 3,4 cd/m2
100% Red @ 10,2 cd/m2

If you perform profiling with 75% Amplitude Patterns then you will have:

75% White @ 25,4 cd/m2
75% Blue @ 1,8 cd/m2
75% Red @ 5,4 cd/m2

After some of my post and discusion at past in CalMAN 5 Topic, CalMAN changed the default software Stimulus Level from 75% to 100% for that reason.



BTW This Meter Profiling Check layout page is useful for users to check if their meter profiling correction table is accurate.

I'm sure you will be surprised when you will find out that your familiar meter profiling procedure you followed since now maybe has create some not valid tables for some reason but you haven't realized that this is be possible because you have always followed the correct way but you haven't checked it.

Use this Reference Meter Page to take WRGB measurements using your Spectro and after that run your meter profiling procedure. (re-measure again the WRGB patches using your spectro and colorimeter)

Later, use the Profiled Meter Page to measure the WRGB Patches with your Profiled Meter to see if the correction table is accurate vs. your reference meter readings.

You will found that sometimes it will need to re-perform your meter profiling 2-3 times to get you accepted results. Count also your meter repeatability/display instability here that maybe produce inaccurate meter correction tables.

This workflow page is included to CalMAN 5 Workflows for Ted's LightSpace CMS Calibration Disk Users.

Anyone who has time he can design something similar for his own custom workflow, since I'm giving this out to my disk users only. wink.gif

Ted's LightSpace CMS Calibration Disk Free Version for Free Calibration Software: LightSpace DPS + CalMAN ColorChecker
S/W: LightSpace CMS, SpaceMan ICC, SpaceMatch DCM, CalMAN 5, CalMAN RGB, ChromaPure, CalPC, ControlCAL
Meters: JETI Specbos 1211, Klein K-10A, i1PRO2, i1PRO, SpectraCAL C6, i1D3, C5
ConnecTEDDD is online now  
Old 04-03-2014, 03:25 AM - Thread Starter
Member
 
Andrei_VVB's Avatar
 
Join Date: Mar 2010
Posts: 191
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 44 Post(s)
Liked: 16
All right. So is safer to use 100% patterns. If that will eliminate a possible bad profiling I see no problem with that.
Now, the question for me is: how do I do that with HCFR? I read these WRGB paterns with spectro. Then what? I need some help here smile.gif
I repeat, I saw in those great videos of MR Chen:) how's done with calman and chromapure but I don't have those.
Andrei_VVB is online now  
Old 04-03-2014, 09:56 AM
AVS Special Member
 
Michael TLV's Avatar
 
Join Date: Feb 2000
Location: THX/ISF Calibrationist/Instructor, Calgary, AB, Canada
Posts: 6,774
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 28 Post(s)
Liked: 138
Greetings

Safer to use 100% patterns? Not necessarily ...

If your contrast is set wrong, the 100% patterns could be color shifted or clipped ... which is not representative of the rest of the image.

Where is most of the content that you watch? is it at 100% levels? Or is it closer to the mid range ...?

75% is used because it is closer to the midrange and more representative of the real picture content range.

Regards

Michael Chen @ The Laser Video Experience
ISF/THX/TLV Video Instructor
The Video Calibration Education Hub - www.TLVEXP.com

Michael TLV is offline  
Old 04-03-2014, 10:50 AM
AVS Special Member
 
ConnecTEDDD's Avatar
 
Join Date: Sep 2010
Location: Athens, Greece
Posts: 2,433
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 131 Post(s)
Liked: 511
Quote:
Originally Posted by Michael TLV View Post

Greetings

Safer to use 100% patterns? Not necessarily ...

If your contrast is set wrong, the 100% patterns could be color shifted or clipped ... which is not representative of the rest of the image.

Where is most of the content that you watch? is it at 100% levels? Or is it closer to the mid range ...?

75% is used because it is closer to the midrange and more representative of the real picture content range.

Regards





Absolutely correct about Clipping, the user has to check that using a similar Contrast Pattern and a Color Channel Clipping Pattern like the above ones, before starting any profiling session.

Ted's LightSpace CMS Calibration Disk Free Version for Free Calibration Software: LightSpace DPS + CalMAN ColorChecker
S/W: LightSpace CMS, SpaceMan ICC, SpaceMatch DCM, CalMAN 5, CalMAN RGB, ChromaPure, CalPC, ControlCAL
Meters: JETI Specbos 1211, Klein K-10A, i1PRO2, i1PRO, SpectraCAL C6, i1D3, C5
ConnecTEDDD is online now  
Old 04-03-2014, 12:14 PM
AVS Special Member
 
HDTVChallenged's Avatar
 
Join Date: Jan 2002
Posts: 8,578
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Liked: 146
Quote:
Originally Posted by ConnecTEDDD View Post



If you have a Display with White @ 120cd/m2 with Gamma 2.2:

If you perform profiling with 100% Amplitude Patterns then you will have:

100% White @ 120 cd/m2
100% Blue @ 8,6 cd/m2

If you perform profiling with 75% Amplitude Patterns then you will have:

75% White @ 63,5 cd/m2
75% Blue @ 4,5 cd/m2

I'm sure this is a stupido question but why does this matter? Are you saying that the i1Pro/Munkis can manage to read the "blue" component accurately enough on "White," but then somehow falls down when reading the same "blue" component alone? By this logic, *every* white/grey reading below 100% would be "suspect." confused.gifbiggrin.gif
HDTVChallenged is offline  
Old 04-03-2014, 01:09 PM
AVS Special Member
 
ConnecTEDDD's Avatar
 
Join Date: Sep 2010
Location: Athens, Greece
Posts: 2,433
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 131 Post(s)
Liked: 511
Quote:
Originally Posted by HDTVChallenged View Post

Quote:
Originally Posted by ConnecTEDDD View Post



If you have a Display with White @ 120cd/m2 with Gamma 2.2:

If you perform profiling with 100% Amplitude Patterns then you will have:

100% White @ 120 cd/m2
100% Blue @ 8,6 cd/m2

If you perform profiling with 75% Amplitude Patterns then you will have:

75% White @ 63,5 cd/m2
75% Blue @ 4,5 cd/m2

I'm sure this is a stupido question but why does this matter? Are you saying that the i1Pro/Munkis can manage to read the "blue" component accurately enough on "White," but then somehow falls down when reading the same "blue" component alone? By this logic, *every* white/grey reading below 100% would be "suspect." confused.gifbiggrin.gif

100% Blue Patch means Blue wth 100% Saturation @ 100% Stimulus Level (Amplitude)

75% Blue Patch means Blue wth 100% Saturation @ 75% Stimulus Level (Amplitude)

I'm not spliting the White reading to RGB Components to my above posted data's.

I hope this makes sense now smile.gif

Ted's LightSpace CMS Calibration Disk Free Version for Free Calibration Software: LightSpace DPS + CalMAN ColorChecker
S/W: LightSpace CMS, SpaceMan ICC, SpaceMatch DCM, CalMAN 5, CalMAN RGB, ChromaPure, CalPC, ControlCAL
Meters: JETI Specbos 1211, Klein K-10A, i1PRO2, i1PRO, SpectraCAL C6, i1D3, C5
ConnecTEDDD is online now  
Old 04-04-2014, 01:54 AM
AVS Special Member
 
HDTVChallenged's Avatar
 
Join Date: Jan 2002
Posts: 8,578
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Liked: 146
Quote:
Originally Posted by ConnecTEDDD View Post

100% Blue Patch means Blue wth 100% Saturation @ 100% Stimulus Level (Amplitude)

75% Blue Patch means Blue wth 100% Saturation @ 75% Stimulus Level (Amplitude)

I'm not spliting the White reading to RGB Components to my above posted data's.

I hope this makes sense now smile.gif

Not really. I'm trying to understand why you would think the radiant spectral power of the "blue" patch (primary) would be any different than the radiant spectral power of the "blue" component in the white patch? (Allowing for variation due to the "current" CCT of white. Edit: scratch that ... CCT doesn't matter here. )

A couple of follow up questions: How much of the "blue" spectrum falls under the y-bar observer function? And secondly, how does this effect the calculated "Y" value of the blue patch? In other words, I'm asking can the Y value of white really be directly compared to the Y value of "blue?"

On my LCD, the SP of "blue" peaks at about 1/2 the peak power of green ... but the blue primary spreads out over almost 2.5 times the spectrum width of the green primary, so once all the math settles it *may* be that blue the primary actually contains a wee bit more radiant power than the green primary.

Perhaps I'm over thinking this ... but I don't think the REC709/601 luma (Y) relationships can be used as a reliable predictor of i1Pro accuracy in the way you propose. OTOH, There could be something else going that would explain the i1Pro vs Jeti discrepancies you are seeing.
HDTVChallenged is offline  
Old 04-04-2014, 03:56 AM
AVS Special Member
 
ConnecTEDDD's Avatar
 
Join Date: Sep 2010
Location: Athens, Greece
Posts: 2,433
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 131 Post(s)
Liked: 511
Quote:
Originally Posted by HDTVChallenged View Post

Quote:
Originally Posted by ConnecTEDDD View Post

100% Blue Patch means Blue wth 100% Saturation @ 100% Stimulus Level (Amplitude)

75% Blue Patch means Blue wth 100% Saturation @ 75% Stimulus Level (Amplitude)

I'm not spliting the White reading to RGB Components to my above posted data's.

I hope this makes sense now smile.gif

Not really. I'm trying to understand why you would think the radiant spectral power of the "blue" patch (primary) would be any different than the radiant spectral power of the "blue" component in the white patch? (Allowing for variation due to the "current" CCT of white. Edit: scratch that ... CCT doesn't matter here. )

A couple of follow up questions: How much of the "blue" spectrum falls under the y-bar observer function? And secondly, how does this effect the calculated "Y" value of the blue patch? In other words, I'm asking can the Y value of white really be directly compared to the Y value of "blue?"

On my LCD, the SP of "blue" peaks at about 1/2 the peak power of green ... but the blue primary spreads out over almost 2.5 times the spectrum width of the green primary, so once all the math settles it *may* be that blue the primary actually contains a wee bit more radiant power than the green primary.

Perhaps I'm over thinking this ... but I don't think the REC709/601 luma (Y) relationships can be used as a reliable predictor of i1Pro accuracy in the way you propose. OTOH, There could be something else going that would explain the i1Pro vs Jeti discrepancies you are seeing.

i1PRO's are loosing accurancy at their readings when the measured patch has low luminance, it's not about blue or red or green, it's about measured Luminance of any patch.

For higher performance below 10cdm/2 you need a better spectro.

One day I will measure 21-Point Grayscale + 20-Step of RGB to show you where the i1PRO1/2 is loosing it's tracking vs. my JETI 1211

Ted's LightSpace CMS Calibration Disk Free Version for Free Calibration Software: LightSpace DPS + CalMAN ColorChecker
S/W: LightSpace CMS, SpaceMan ICC, SpaceMatch DCM, CalMAN 5, CalMAN RGB, ChromaPure, CalPC, ControlCAL
Meters: JETI Specbos 1211, Klein K-10A, i1PRO2, i1PRO, SpectraCAL C6, i1D3, C5
ConnecTEDDD is online now  
Old 04-04-2014, 11:08 AM
AVS Special Member
 
HDTVChallenged's Avatar
 
Join Date: Jan 2002
Posts: 8,578
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Liked: 146
Quote:
Originally Posted by ConnecTEDDD View Post

i1PRO's are loosing accurancy at their readings when the measured patch has low luminance, it's not about blue or red or green, it's about measured Luminance of any patch.

For higher performance below 10cdm/2 you need a better spectro.

One day I will measure 21-Point Grayscale + 20-Step of RGB to show you where the i1PRO1/2 is loosing it's tracking vs. my JETI 1211

You need to review how spectro's actually work. "Luminance" (Y) has no meaning in this context. "Luminance" (Y) only matters to the human eye/brain. All spectros "care about" is the actual physical intensity (radiant power) falling upon their sensors. "Luminance" (Y) is not an actual physical property, it's a virtual construct of the CIE 1931 Standard Observer.

But yes we know already know that the i1Pro is *not* a Jeti 1211. There may be discrepancies between the two meters, but that has nothing to do with "Luminance" (Y) of individual patches. smile.gif
HDTVChallenged is offline  
Old 04-04-2014, 11:09 AM
AVS Special Member
 
PlasmaPZ80U's Avatar
 
Join Date: Feb 2009
Posts: 7,291
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 112 Post(s)
Liked: 204
Quote:
Originally Posted by HDTVChallenged View Post

Not really. I'm trying to understand why you would think the radiant spectral power of the "blue" patch (primary) would be any different than the radiant spectral power of the "blue" component in the white patch? (Allowing for variation due to the "current" CCT of white. Edit: scratch that ... CCT doesn't matter here. )

A couple of follow up questions: How much of the "blue" spectrum falls under the y-bar observer function? And secondly, how does this effect the calculated "Y" value of the blue patch? In other words, I'm asking can the Y value of white really be directly compared to the Y value of "blue?"

On my LCD, the SP of "blue" peaks at about 1/2 the peak power of green ... but the blue primary spreads out over almost 2.5 times the spectrum width of the green primary, so once all the math settles it *may* be that blue the primary actually contains a wee bit more radiant power than the green primary.

Perhaps I'm over thinking this ... but I don't think the REC709/601 luma (Y) relationships can be used as a reliable predictor of i1Pro accuracy in the way you propose. OTOH, There could be something else going that would explain the i1Pro vs Jeti discrepancies you are seeing.

like Ted mentions it's simply because blue is much darker than white (red, green, and blue at the same time)... simply put, the i1pro is more accurate measuring brighter patches than dimmer ones (and blue is the darkest color measured during profiling)

EDIT: also, this is why SpectraCal updated CalMAN 5 not too long ago to make the default % stim for profiling 100% from the previous 75%
ConnecTEDDD likes this.
PlasmaPZ80U is offline  
Old 04-04-2014, 11:10 AM
AVS Special Member
 
HDTVChallenged's Avatar
 
Join Date: Jan 2002
Posts: 8,578
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Liked: 146
Quote:
Originally Posted by PlasmaPZ80U View Post

like Ted mentions it's simply because blue is much darker than white (red, green, and blue at the same time)... simply put, the i1pro is more accurate measuring brighter patches than dimmer ones (and blue is the darkest color measured during profiling)

See above (#14,) please. smile.gif

It's certainly true that the i1Pro does better with more "light." The problem here is assuming that the i1Pro (or any spectro) sees light the same way we humans do.
HDTVChallenged is offline  
Old 04-04-2014, 11:13 AM
AVS Special Member
 
PlasmaPZ80U's Avatar
 
Join Date: Feb 2009
Posts: 7,291
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 112 Post(s)
Liked: 204
Quote:
Originally Posted by HDTVChallenged View Post

See above (#14,) please. smile.gif

actually it's well documented the i1pro1/2 accuracy falls off when measuring dark patterns... which is why most profile with colorimeters

you're overthinking this
ConnecTEDDD likes this.
PlasmaPZ80U is offline  
Old 04-04-2014, 11:16 AM
AVS Special Member
 
HDTVChallenged's Avatar
 
Join Date: Jan 2002
Posts: 8,578
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Liked: 146
Quote:
Originally Posted by PlasmaPZ80U View Post

actually it's well documented the i1pro1/2 accuracy falls off when measuring dark patterns... which is why most profile with colorimeters

you're overthinking this

The *POINT* point here, is that the blue patch is not really as "dark" (to a spectro) as it's Y value would suggest. smile.gif The "Z" value would be a more appropriate measure of the physical "energy" contained in (or radiated from) the blue patch.

Again ... Standard Observer applies to humans ... not mechanical/electrical devices.
HDTVChallenged is offline  
Old 04-04-2014, 11:24 AM
AVS Special Member
 
ConnecTEDDD's Avatar
 
Join Date: Sep 2010
Location: Athens, Greece
Posts: 2,433
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 131 Post(s)
Liked: 511
Quote:
Originally Posted by HDTVChallenged View Post

Quote:
Originally Posted by ConnecTEDDD View Post

i1PRO's are loosing accurancy at their readings when the measured patch has low luminance, it's not about blue or red or green, it's about measured Luminance of any patch.

For higher performance below 10cdm/2 you need a better spectro.

One day I will measure 21-Point Grayscale + 20-Step of RGB to show you where the i1PRO1/2 is loosing it's tracking vs. my JETI 1211

You need to review how spectro's actually work. "Luminance" (Y) has no meaning in this context. "Luminance" (Y) only matters to the human eye/brain. All spectros "care about" is the actual physical intensity (radiant power) falling upon their sensors. "Luminance" (Y) is not an actual physical property, it's a virtual construct of the CIE 1931 Standard Observer.

But yes we know already know that the i1Pro is *not* a Jeti 1211. There may be discrepancies between the two meters, but that has nothing to do with "Luminance" (Y) of individual patches. smile.gif



Hey, this is display calibration forum, we are measuring displays, so the Luminance is the correct terminology.

BTW if you want to see where your i1PRO is loosing it's accurancy, use your colorimeter as a reference meter and create a meter correction table for i1PRO, after that make that correction table active and read 21-step grayscale with both meters and compare later the charts. You will see at what measured luminance levels the i1PRO is loosing the tracking.
Iron Mike likes this.

Ted's LightSpace CMS Calibration Disk Free Version for Free Calibration Software: LightSpace DPS + CalMAN ColorChecker
S/W: LightSpace CMS, SpaceMan ICC, SpaceMatch DCM, CalMAN 5, CalMAN RGB, ChromaPure, CalPC, ControlCAL
Meters: JETI Specbos 1211, Klein K-10A, i1PRO2, i1PRO, SpectraCAL C6, i1D3, C5
ConnecTEDDD is online now  
Old 04-04-2014, 11:27 AM
AVS Special Member
 
HDTVChallenged's Avatar
 
Join Date: Jan 2002
Posts: 8,578
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Liked: 146
Quote:
Originally Posted by ConnecTEDDD View Post


Hey, this is display calibration forum, we are measuring displays, so the Luminance is the correct terminology.

Completely missing the point ... again. biggrin.gif
HDTVChallenged is offline  
Old 04-04-2014, 11:27 AM
AVS Special Member
 
PlasmaPZ80U's Avatar
 
Join Date: Feb 2009
Posts: 7,291
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 112 Post(s)
Liked: 204
Quote:
Originally Posted by ConnecTEDDD View Post

if you want to see where your i1PRO is loosing it's accurancy, use your colorimeter as a reference meter and create a meter correction table for i1PRO, after that make that correction table active and read 21-step grayscale with both meters and compare later the charts. You will see at what measured luminance levels the i1PRO is loosing the tracking.

I think that''s the easiest way to see this. Sometimes a little less theory and a little more practice is a better teacher. (In other words, you can learn sometimes better by just trying things out instead of theorizing).
ConnecTEDDD likes this.
PlasmaPZ80U is offline  
Old 04-04-2014, 11:33 AM
AVS Special Member
 
HDTVChallenged's Avatar
 
Join Date: Jan 2002
Posts: 8,578
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Liked: 146
Quote:
Originally Posted by PlasmaPZ80U View Post

I think that''s the easiest way to see this. Sometimes a little less theory and a little more practice is a better teacher. (In other words, you can learn sometimes better by just trying things out instead of theorizing).

You can also be "right" for the "wrong" reason(s.) The premise (hypothesis) presented by Ted is completely invalid.

PS: I really was trying to be "gentle" on this one. wink.gif
HDTVChallenged is offline  
Old 04-04-2014, 12:16 PM
Advanced Member
 
spacediver's Avatar
 
Join Date: May 2013
Location: Toronto
Posts: 943
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 120 Post(s)
Liked: 75
This thread raises an interesting issue.

We often hear that the i1 pro suffers loss of accuracy at low luminances. But, as HDTVChallenged rightly points out, luminance is the radiant power of the source (per nm) multiplied by the spectral luminous efficiency function of the human visual system (see the black curve in image below).



What this basically means is that the human visual system is not equally sensitive to radiant power across the spectrum. In photopic conditions (so ignore contribution from rods), we are most sensitive to around 555 nm, and below 400 nm and above 700 nm, our sensitivity tapers off dramatically. The quantity luminance takes this weighting function into account.

The thing is, a spectroradiometer is pretty much equally sensitive to all wavelengths that it detects. It measures spectral radiant power, and then software is used to calculate luminance, by multiplying the raw spectral radiant output by the luminous efficiency function.

It therefore doesn't really make sense to say that a spectrophotometer suffers a loss of accuracy at low luminance. Rather, it suffers a loss of accuracy at low radiance.

At the green end, much less radiant power is required to make something visible, compared to at the blue end. This means that a spectrophotometer can read green blue accurately at much lower luminances than it can at the blue green end.

This means that we should really discuss the spectro's loss of accuracy as a function of radiance, and then we can figure out what that means for luminance depending on the spectral source we're talking about.
spacediver is online now  
Old 04-04-2014, 12:52 PM
Advanced Member
 
spacediver's Avatar
 
Join Date: May 2013
Location: Toronto
Posts: 943
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 120 Post(s)
Liked: 75
Quote:
Originally Posted by ConnecTEDDD View Post


BTW if you want to see where your i1PRO is loosing it's accurancy, use your colorimeter as a reference meter and create a meter correction table for i1PRO, after that make that correction table active and read 21-step grayscale with both meters and compare later the charts. You will see at what measured luminance levels the i1PRO is loosing the tracking.

yep, this would be a good way to do it, but you'd get different results depending on whether you're measuring white, blue, red, or green. And it would also vary from display type to display type.

If you're only concerned with white point balance, however, then using a grayscale pattern will be fine to find this "critical luminance".
spacediver is online now  
Old 04-04-2014, 04:12 PM
AVS Special Member
 
HDTVChallenged's Avatar
 
Join Date: Jan 2002
Posts: 8,578
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Liked: 146
Quote:
Originally Posted by spacediver View Post

This thread raises an interesting issue.

At the green end, much less radiant power is required to make something visible (to humans), compared to at the blue end. This means that a spectrophotometer can read green accurately at much lower luminances than it can at the blue end.

This means that we should really discuss the spectro's loss of accuracy as a function of radiance, and then we can figure out what that means for luminance depending on the spectral source we're talking about.

I was with you up to this last part. I think perhaps you have the bolded part above backwards ... If you are using "luminance" in the CIE1931 SO sense. Again, Spectro's don't care about "luminance." Blue is always going to have a *much, much* lower luminance (Y) value compared to green ... That's just a pure, mathematical consequence of the CIE 1931 SO model.

To get back to what originally got my "geek" on: If we accept that an i1Pro can't accurately measure a 75% Stimulated blue patch, then it follows that it can't accurately measure the corresponding 75% Stimulated White patch either (even though the SO luminance of the white patch is *much* higher than that of blue.) Blue radiance is a part of White radiance. If blue is wrong, then white is also wrong ... Thus by the Ted's original hypothesis, the i1Pro can't be trusted for any patch *other* than 100% Stimulated White. And we know from observations and comparisons that this is simply not the case. QED. Are we done here? smile.gif

PS: I do see the possibility that the i1Pro's sensors may not be "equally" sensitive at all wavelengths or bands. This is why I said there could be something else going on that explains the "loss of accuracy" observations.
HDTVChallenged is offline  
Old 04-04-2014, 05:02 PM
Advanced Member
 
spacediver's Avatar
 
Join Date: May 2013
Location: Toronto
Posts: 943
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 120 Post(s)
Liked: 75
Quote:
Originally Posted by HDTVChallenged View Post

I was with you up to this last part. I think perhaps you have the bolded part above backwards ... If you are using "luminance" in the CIE1931 SO sense. Again, Spectro's don't care about "luminance." Blue is always going to have a *much, much* lower luminance (Y) value compared to green ... That's just a pure, mathematical consequence of the CIE 1931 SO model.

You're right, I got it backwards. Spectro will be able to read lower luminances more reliably at the blue end than the green end. Thanks for the correction.

Quote:
Originally Posted by HDTVChallenged View Post

To get back to what originally got my "geek" on: If we accept that an i1Pro can't accurately measure a 75% Stimulated blue patch, then it follows that it can't accurately measure the corresponding 75% Stimulated White patch either (even though the SO luminance of the white patch is *much* higher than that of blue.) Blue radiance is a part of White radiance. If blue is wrong, then white is also wrong ... Thus by the Ted's original hypothesis, the i1Pro can't be trusted for any patch *other* than 100% Stimulated White. And we know from observations and comparisons that this is simply not the case. QED. Are we done here? smile.gif

This is where my knowledge breaks down, but perhaps we'd need to factor in the relative perceptual contribution of the primaries to achieving D65. If it turns out that blue is relatively less important, nit for nit, than green is (in other words, if we can tolerate larger errors in blue's luminance without moving our delta E too much), then perhaps the situation changes.

Quote:
Originally Posted by HDTVChallenged View Post

PS: I do see the possibility that the i1Pro's sensors may not be "equally" sensitive at all wavelengths or bands. This is why I said there could be something else going on that explains the "loss of accuracy" observations.

I think factors such as the gap between the individual detectors on the array, and the bandpass filters used for each detector determine the overall spectral responsiveness. But from what I understand, the worst that can happen is that the function will be slightly undulating across the spectrum (totally flat would be ideal).

be back in 24 hours, gotta run!
spacediver is online now  
Old 04-04-2014, 11:38 PM
AVS Special Member
 
HDTVChallenged's Avatar
 
Join Date: Jan 2002
Posts: 8,578
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Liked: 146
Quote:
Originally Posted by spacediver View Post

You're right, I got it backwards. Spectro will be able to read lower luminances more reliably at the blue end than the green end. Thanks for the correction.

Whew! I thought I missed something in the math there. smile.gif
Quote:
This is where my knowledge breaks down, but perhaps we'd need to factor in the relative perceptual contribution of the primaries to achieving D65. If it turns out that blue is relatively less important, nit for nit, than green is (in other words, if we can tolerate larger errors in blue's luminance without moving our delta E too much), then perhaps the situation changes.

Usually, the 4-color target patches are encoded the same way the SMPTE color bar charts using the specific REC709 or 601 defined relationships between R,B,G and W. So, if you look at the spectrum of W it should equal the spectrum(s) of R+B+G (assuming a flat greyscale) ... If you take W and subtract away R and G you should be left with the spectrum of the B primary. Cyan's spectrum, for instance, would be the sum of B and G primary spectrums, etc, etc. (assuming all the color decoding a/o CMS controls are adjusted correctly.) Point being that the blue primary represents 1/3 (ish) of the total W spectrum (radiantly speaking.)

In any case, I do agree with Ted that one should give the i1Pro/Munki's as much light to work with as possible. That's just best practice, regardless of reasoning. wink.gif

PS: For the record on my last i1Pro reference reading the Y value for blue was 8.353, the Z value was 102.23 vs white's Y of 113.95Nits and Z of 116.548. Obviously, Z measurements are not equivalent to Y, but it should be clear that there's plenty of available radiant energy to be measured in the blue patch.
HDTVChallenged is offline  
Old 04-05-2014, 12:46 AM
AVS Special Member
 
HDTVChallenged's Avatar
 
Join Date: Jan 2002
Posts: 8,578
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Liked: 146
OK ... Here's my alternate hypothesis as to why the i1Pro looses accuracy at lower White/Grey luma levels. As we drop the stimulus level from 100% toward zero, the *entire* spectral radiance level drops ... spectral peaks and valleys flatten to where the total spectral radiance "curve" tends toward a flat-line. Thus it becomes more and more difficult to "separate" X, Y and Z mathematically ... intensity differences between each 10nm band tend to get smoothed out, so as the total radiance level drops the ratios between the integrated values for X, Y and Z would each tend toward a "fixed" value. Specifically, the ratios between the areas under each of the three SO weighting functions. So to my feeble brain it's probably a combination of sensor sensitivity, "bandwidth" (10nm), and last but not least, limitations built into the SO mathematical model itself.

(Smacks head) So I suppose luma *could* be used indirectly as an accuracy threshold. It's just that you can't use the same luma value threshold for white and blue (or red, or green.) That being said, I'm looking at my LCD's spectrum of white at 100% and 70% stimulus and I don't see anything "alarming" between the two graphs ... peaks and valleys all there ... 20% on the other hand ....

This is a case where a series of pictures would go a long way ... unfortunately, I don't have them to post, so I must leave the visualization up to the imagination.

My brain hurts now. smile.gif

Sorry for edits ... brain race condition ...

Final Edit: Anyway, my take-away from all this "thinkin'" is that if your i1Pro (or Munki) can read a 30% stimulated white patch accurately, it shouldn't have *any* significant problem reading a 30% stimulated blue patch. So for profiling, it's really comes down to dealers choice ... use your 100% stimulated patches or use your 75% patches ... "it's all good, man" - potential clipping issues aside. wink.gif
HDTVChallenged is offline  
Old 04-05-2014, 02:48 PM
Advanced Member
 
spacediver's Avatar
 
Join Date: May 2013
Location: Toronto
Posts: 943
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 120 Post(s)
Liked: 75
Quote:
Originally Posted by HDTVChallenged View Post

Usually, the 4-color target patches are encoded the same way the SMPTE color bar charts using the specific REC709 or 601 defined relationships between R,B,G and W. So, if you look at the spectrum of W it should equal the spectrum(s) of R+B+G (assuming a flat greyscale) ... If you take W and subtract away R and G you should be left with the spectrum of the B primary. Cyan's spectrum, for instance, would be the sum of B and G primary spectrums, etc, etc. (assuming all the color decoding a/o CMS controls are adjusted correctly.) Point being that the blue primary represents 1/3 (ish) of the total W spectrum (radiantly speaking.)

yep, I remember taking spectral readings of my FW900 phosphors and seeing that the primaries added up nicely.

Here's more of what I'm getting at, and it might be interesting to do this as an experiment.

Find the threshold luminance at which the spectro can read off a blue primary's chromaticity accurately. Lower the luminance to half of this level, and note how noisy the delta E readings are. Let's call this value Noise A.

Create a d65 pattern whose blue component matches that threshold luminance.

Scale down the luminance of this d65 pattern to half, and note how noisy the delta E readings are. Let's call this value Noise B.

It may be the case that Noise A and Noise B are not equivalent, even though they represent the same amount of instrument noise. The reason for this is that perceptually speaking, d65's chromaticity may be more resilient to noise in the blue component, than blue's chromaticty is. Remember, delta E is in the perceptual domain, and the perceptual scale varies depending on where you are in the CIE xy space. So equivalent instrument noise may result in varying perceptual noise depending on what chromaticity coordinate your working with.

I should probably do this experiment myself tongue.gif
spacediver is online now  
Old 04-05-2014, 02:51 PM
Advanced Member
 
spacediver's Avatar
 
Join Date: May 2013
Location: Toronto
Posts: 943
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 120 Post(s)
Liked: 75
Quote:
Originally Posted by HDTVChallenged View Post

OK ... Here's my alternate hypothesis as to why the i1Pro looses accuracy at lower White/Grey luma levels.

just curious: why are you using the term luma here? Luma is an engineering approximation to CIE L*. It's more akin to lightness than luminance.
spacediver is online now  
 
Thread Tools


Forum Jump: 

Posting Rules  
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off