How low can the i1 display pro read - Page 2 - AVS Forum
Forum Jump: 
 3Likes
Reply
 
Thread Tools
post #31 of 51 Old 06-20-2014, 09:38 PM - Thread Starter
Advanced Member
 
spacediver's Avatar
 
Join Date: May 2013
Location: Toronto
Posts: 772
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 39 Post(s)
Liked: 65
Quote:
Originally Posted by TomHuffman View Post
My point exactly. Grayscale should be about chromaticity errors only. If you want to consider luminance errors, then look to gamma. Perhaps I misunderstood what the previous poster was referring to by "absolute Y W/O Gamma".
Sorry I misunderstood you.
spacediver is online now  
Sponsored Links
Advertisement
 
post #32 of 51 Old 06-20-2014, 10:26 PM
AVS Special Member
 
HDTVChallenged's Avatar
 
Join Date: Jan 2002
Posts: 8,309
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Liked: 108
To clarify: HCFR users should use dE76 (Luv) with "relative Y" if you want to iron out the greyscale as much as possible. Alternatively, one could just minimize the delta-xy readings, with the "recommended" dE setting.
HDTVChallenged is offline  
post #33 of 51 Old 06-21-2014, 03:23 AM
AVS Special Member
 
zoyd's Avatar
 
Join Date: Sep 2006
Location: Planet Dog
Posts: 4,431
Mentioned: 6 Post(s)
Tagged: 0 Thread(s)
Quoted: 78 Post(s)
Liked: 299
Quote:
Originally Posted by HDTVChallenged View Post
To clarify: HCFR users should use dE76 (Luv) with "relative Y" if you want to iron out the greyscale as much as possible. Alternatively, one could just minimize the delta-xy readings, with the "recommended" dE setting.

yes, the default behavior is the same as what's used in CM, ArgyllCMS, and LS - the grayscale is treated just like any other color, and color differences are calculated in a color space where L* = 100 is at reference white.

If you wish to use a different luminance reference (to amplify grayscale errors for real-time adjustments for example) you should use the relative Y option although the absolute values obtained are of questionable perceptual relevance. In my opinion the relative Y numbers do not correlate well with perceived errors for very low luminance patches (say 10% video level and below). Another alternative is to use the "gray world" assumption option, this will assign L* = 100 to the 15% linear light level which is a typical average luminance level in film and video reproduction.
zoyd is offline  
post #34 of 51 Old 06-21-2014, 05:16 AM
AVS Special Member
 
zoyd's Avatar
 
Join Date: Sep 2006
Location: Planet Dog
Posts: 4,431
Mentioned: 6 Post(s)
Tagged: 0 Thread(s)
Quoted: 78 Post(s)
Liked: 299
Quote:
Originally Posted by TomHuffman View Post
...

The CIEDE2000 for this pair of colors is 7.3.

or,
...

The CIEDE2000 for this pair of colors is 9.4. Ordinarily, using this method would result in a much lower dE score, but in this particular instance both the chromaticity and the luminance have large errors.
yes, choosing the relative Y option will yield dE = 7.3 for this measurement and choosing the absolute with gamma will yield dE = 9.4.

The absolute Y w/o gamma option sets the target luminance equal to the measured luminance to remove the L component of the error vector. This yields better correlation with perceptual errors than the relative Y option for most of the levels. But I always recommend to switch to the full error vector for final proofing after gamma adjustments to get the full picture. This particular measurement is not a good example to use because as you point out the luminance error is very high and driving the error vector more then is typical.
zoyd is offline  
post #35 of 51 Old 06-21-2014, 10:13 AM - Thread Starter
Advanced Member
 
spacediver's Avatar
 
Join Date: May 2013
Location: Toronto
Posts: 772
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 39 Post(s)
Liked: 65
When I have time I'll take some more measurements after Argyll gamma correction and post the delta Es with the more stringent settings.
spacediver is online now  
post #36 of 51 Old 07-16-2014, 03:40 PM - Thread Starter
Advanced Member
 
spacediver's Avatar
 
Join Date: May 2013
Location: Toronto
Posts: 772
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 39 Post(s)
Liked: 65
Quote:
Originally Posted by gwgill View Post
Using the ArgyllCMS driver in it's default mode, it will integrate for 20 seconds, and the lowest non-zero value of Y seems to be about around 0.002 cd/m^2.
That's exactly what my black level reads with the i1 display pro - 0.002 cd/m^2.
spacediver is online now  
post #37 of 51 Old 07-16-2014, 07:10 PM
Senior Member
 
<^..^>Smokey Joe's Avatar
 
Join Date: Aug 2006
Location: Christchurch, New Zealand
Posts: 367
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 5 Post(s)
Liked: 39
0.002 cd/m^2 +/- what?

Masterpiece Calibration Ltd
Christchurch NZ
<^..^>Smokey Joe is offline  
post #38 of 51 Old 07-16-2014, 07:14 PM - Thread Starter
Advanced Member
 
spacediver's Avatar
 
Join Date: May 2013
Location: Toronto
Posts: 772
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 39 Post(s)
Liked: 65
Quote:
Originally Posted by <^..^>Smokey Joe View Post
0.002 cd/m^2 +/- what?
that was the reading I got when I measured my min luminance on my display: 0.002 cd/m^2. It didn't give me a confidence interval or anything. Not sure what you mean!
spacediver is online now  
post #39 of 51 Old 07-16-2014, 07:47 PM
Senior Member
 
CalWldLif's Avatar
 
Join Date: Feb 2011
Posts: 389
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 9 Post(s)
Liked: 23
0.002 cd/m^2

isn't that @ .0007 ftl
CalWldLif is offline  
post #40 of 51 Old 07-16-2014, 08:11 PM
Senior Member
 
<^..^>Smokey Joe's Avatar
 
Join Date: Aug 2006
Location: Christchurch, New Zealand
Posts: 367
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 5 Post(s)
Liked: 39
Quote:
Originally Posted by spacediver View Post
that was the reading I got when I measured my min luminance on my display: 0.002 cd/m^2. It didn't give me a confidence interval or anything. Not sure what you mean!
Detection alone is not accuracy, detection is where a signal starts returning a value, but it is rare that detection limit and repeatable accuracy are the same. Hence +/- (value) around the around the measured value. At the detection limit the +/-(value) typically grows in error.

For reference equipment the error is typically around <2% of the sweet spot, for example the Jeti 1211 Spec is ± 2 % @ 100 cd/m² and illuminant A

So that is 98cd/m² ~ 102cd/m² of the reading value

Error can get worse around detection limits, it could be 10~20% or more. Reference equipment can hold 2% to close to the detection limit, however there is always a roll off.

The reported spec for the i1D3 is 0.1cd/m² but be sure the accuracy at 100cd/m² is by chance as good as a reference device.

So far no one has shown stats with proof that anything below 0.1cd/m² is anything but signal noise.

Masterpiece Calibration Ltd
Christchurch NZ
<^..^>Smokey Joe is offline  
post #41 of 51 Old 07-16-2014, 08:12 PM - Thread Starter
Advanced Member
 
spacediver's Avatar
 
Join Date: May 2013
Location: Toronto
Posts: 772
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 39 Post(s)
Liked: 65
Quote:
Originally Posted by CalWldLif View Post
0.002 cd/m^2

isn't that @ .0007 ftl
More like 0.00058 ftl, but yes

And it might well be lower - I'm hoping to get a minolta LS-100 to test it!
spacediver is online now  
post #42 of 51 Old 07-16-2014, 08:21 PM - Thread Starter
Advanced Member
 
spacediver's Avatar
 
Join Date: May 2013
Location: Toronto
Posts: 772
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 39 Post(s)
Liked: 65
Quote:
Originally Posted by <^..^>Smokey Joe View Post
Detection alone is not accuracy...
True, but if I'm understanding Zoyd's measurements correctly (reported earlier in this thread), the precision of the instrument is around that level.

Quote:
Originally Posted by <^..^>Smokey Joe View Post
So far no one has shown stats with proof that anything below 0.1cd/m² is anything but signal noise.
I'll take some near black measurements tonight from the first few consecutive video levels. The results might speak to how much of the readings are due to noise.
spacediver is online now  
post #43 of 51 Old 07-16-2014, 09:42 PM - Thread Starter
Advanced Member
 
spacediver's Avatar
 
Join Date: May 2013
Location: Toronto
Posts: 772
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 39 Post(s)
Liked: 65
So at night, my black level is apparently too low to register, so I loaded up a LUT that raises the black level - here are some readings (in cd/m^2):

0%: unable to read
1%: 0.002
2%: 0.008
3%: 0.017
4% 0.027

Highly unlikely that anything below 0.01 let alone 0.1 is noise.

Last edited by spacediver; 07-16-2014 at 09:47 PM.
spacediver is online now  
post #44 of 51 Old 07-16-2014, 10:11 PM
Senior Member
 
CalWldLif's Avatar
 
Join Date: Feb 2011
Posts: 389
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 9 Post(s)
Liked: 23
Quote:
Originally Posted by spacediver View Post
So at night, my black level is apparently too low to register, so I loaded up a LUT that raises the black level - here are some readings (in cd/m^2):

0%: unable to read
1%: 0.002
2%: 0.008
3%: 0.017
4% 0.027

Highly unlikely that anything below 0.01 let alone 0.1 is noise.
wow. what tv?
CalWldLif is offline  
post #45 of 51 Old 07-16-2014, 10:30 PM - Thread Starter
Advanced Member
 
spacediver's Avatar
 
Join Date: May 2013
Location: Toronto
Posts: 772
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 39 Post(s)
Liked: 65
Quote:
Originally Posted by CalWldLif View Post
wow. what tv?
It's a computer monitor: Sony GDM-FW900 - 24 inch widescreen high resolution CRT, often used in professional video studios up until a few years ago.
spacediver is online now  
post #46 of 51 Old 07-16-2014, 11:44 PM
Advanced Member
 
-Hitman-'s Avatar
 
Join Date: Nov 2004
Location: Manchester, UK
Posts: 642
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 30 Post(s)
Liked: 55
Had my I1D3 for a couple of months, on my Pioneer KRP-600m, I get between 0.00045fl and 0.0004fl.
-Hitman- is online now  
post #47 of 51 Old 07-17-2014, 02:30 PM
Senior Member
 
<^..^>Smokey Joe's Avatar
 
Join Date: Aug 2006
Location: Christchurch, New Zealand
Posts: 367
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 5 Post(s)
Liked: 39
Quote:
Originally Posted by spacediver View Post
So at night, my black level is apparently too low to register, so I loaded up a LUT that raises the black level - here are some readings (in cd/m^2):

0%: unable to read
1%: 0.002
2%: 0.008
3%: 0.017
4% 0.027

Highly unlikely that anything below 0.01 let alone 0.1 is noise.
Look, I'm not trying to pop your balloon, or anyone elses for that matter. Trouble is this, measured value and intended value can be two intirely different things. The intended value is the reference level, just because you measure something at the detection limit of the detector doesn't actually mean that it returns a value resembling the intended reference value, you might even be able to repeat the measured value, but it could still carry 20% error. The light source might not even output the correct reference value, how do you know unless you have ultimate reference systems.

What I'd call reference here is a CS2000 or PR670, intergration spheres and pure/controlled light sources, even my Jeti 1211 isn't good enough because it isnt sensitive enough for this low level.

The Device is rated at 0.1 cd/m^2 probably because this is where intended values and measured values become an an acceptable difference. Below that detected values measured will fall foul of compounded error, where you could state that something has been measured, but the actual error is unknown. You can't even place a dE value upon it.

Masterpiece Calibration Ltd
Christchurch NZ
<^..^>Smokey Joe is offline  
post #48 of 51 Old 07-17-2014, 02:55 PM
Member
 
1forsnow's Avatar
 
Join Date: Mar 2010
Location: Chicago
Posts: 171
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 2 Post(s)
Liked: 38
Quote:
Originally Posted by -Hitman- View Post
Had my I1D3 for a couple of months, on my Pioneer KRP-600m, I get between 0.00045fl and 0.0004fl.
Dnice modified? I didn't think stock 600m's got that low, only 500m's did.

sorry for being off topic....
1forsnow is offline  
post #49 of 51 Old 07-17-2014, 03:03 PM - Thread Starter
Advanced Member
 
spacediver's Avatar
 
Join Date: May 2013
Location: Toronto
Posts: 772
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 39 Post(s)
Liked: 65
Quote:
Originally Posted by <^..^>Smokey Joe View Post
Look, I'm not trying to pop your balloon, or anyone elses for that matter. Trouble is this, measured value and intended value can be two intirely different things. The intended value is the reference level, just because you measure something at the detection limit of the detector doesn't actually mean that it returns a value resembling the intended reference value, you might even be able to repeat the measured value, but it could still carry 20% error. The light source might not even output the correct reference value, how do you know unless you have ultimate reference systems.
I thought we were talking about precision here, not accuracy.

Quote:
Originally Posted by <^..^>Smokey Joe View Post
What I'd call reference here is a CS2000 or PR670, intergration spheres and pure/controlled light sources, even my Jeti 1211 isn't good enough because it isnt sensitive enough for this low level.
Keep in mind that the Jeti 1211 is a spectroradiometer, which is a class of instrument generally less sensitive than a colorimeter.


Quote:
Originally Posted by <^..^>Smokey Joe View Post
The Device is rated at 0.1 cd/m^2 probably because this is where intended values and measured values become an an acceptable difference. Below that detected values measured will fall foul of compounded error, where you could state that something has been measured, but the actual error is unknown. You can't even place a dE value upon it.
When it comes to precision, however, I think we have some very solid data indicating that the precision of the instrument is around 0.002 cd/m^2.
spacediver is online now  
post #50 of 51 Old 07-17-2014, 03:16 PM
Advanced Member
 
-Hitman-'s Avatar
 
Join Date: Nov 2004
Location: Manchester, UK
Posts: 642
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 30 Post(s)
Liked: 55
Quote:
Originally Posted by 1forsnow View Post
Dnice modified? I didn't think stock 600m's got that low, only 500m's did.

sorry for being off topic....
Tweaked yes but not by D-Nice.


That measurement was taken as an average from calman5 with auto sync, sync on - I either get 0004 or an error code in CM like "E2 XX" instead of measurement data from the D3.

Last edited by -Hitman-; 07-17-2014 at 03:22 PM.
-Hitman- is online now  
post #51 of 51 Old 07-17-2014, 03:41 PM
AVS Special Member
 
LarryInRI's Avatar
 
Join Date: Jun 2007
Posts: 5,258
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 4 Post(s)
Liked: 144
Quote:
Originally Posted by spacediver View Post
So at night, my black level is apparently too low to register, so I loaded up a LUT that raises the black level - here are some readings (in cd/m^2):

0%: unable to read
1%: 0.002
2%: 0.008
3%: 0.017
4% 0.027

Highly unlikely that anything below 0.01 let alone 0.1 is noise.

Different TV but additional data for the Display Pro (Display 3.)
The Official ChromaPure thread

Larry
LarryInRI is offline  
Reply Display Calibration

User Tag List

Thread Tools
Show Printable Version Show Printable Version
Email this Page Email this Page


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