Re: Real world isis ver i1 iO test
Re: Real world isis ver i1 iO test
- Subject: Re: Real world isis ver i1 iO test
- From: Martin Weberg <email@hidden>
- Date: Fri, 5 Mar 2010 00:02:40 +0100
Hi Derek
These are calculations using my development version of ValiProof.
Colour difference formula is CIE dE 2000.
iO_p1m1 vs.p1m2
Min 0.12
Max 1.89
Mean 0.61
StDev 0.20
Median 0.60
MAD 0.13
Skewness 0.56
Kurtosis 1.82
iO_p2m1 vs. p2m2
Min 0.04
Max 2.14
Mean 0.80
StDev 0.33
Median 0.77
MAD 0.21
Skewness 0.57
Kurtosis 0.71
iO_UV_p1m1 vs. UV_p1m2
Min 0.01
Max 7.03
Mean 0.27
StDev 0.54
Median 0.12
MAD 0.05
Skewness 6.67
Kurtosis 59.4
iO_UV_p2m1 vs. p2m2
Min 0.01
Max 2.58
Mean 0.14
StDev 0.21
Median 0.09
MAD 0.04
Skewness 6.16
Kurtosis 49.8
iSis_p1m1 vs. p1m2
Min 0.0
Max 0.73
Mean 0.06
StDev 0.06
Median 0.04
MAD 0.02
Skewness 4.34
Kurtosis 29.9
iSis_p2m1 vs. p2m2
Min 0.02
Max 0.72
Mean 0.16
StDev 0.09
Median 0.15
MAD 0.05
Skewness 1.85
Kurtosis 5.52
iSisUV_p1m1 vs. p1m2
Min 0.0
Max 0.32
Mean 0.03
StDev 0.02
Median 0.03
MAD 0.01
Skewness 3.84
Kurtosis 28.7
iSisUV_p2m1 vs. p2m2
Min 0.02
Max 0.71
Mean 0.16
StDev 0.09
Median 0.14
MAD 0.05
Skewness 1.76
Kurtosis 5.31
Now, some analysis:
iO UV included
Distribution of colour differences are relatively normally
distributed. No difference between paper1 and paper2. Mean/StDev:
0.61/0.20 and 0.80/0.33
iO UV excluded
Distribution of colour differences are skewed and peaked (using robust
stats). No difference between paper1 and paper2. Median/MAD: 0.12/0.05
and 0.09/0.04
iSis UV included
Distribution of colour differences are skewed and peaked (using robust
stats). Some difference between reading of paper1 and paper2. Unclear
if due to paper or instrument. Median/MAD: 0.04/0.02 and 0.12/0.05
iSis UV excluded
Distribution of colour differences are skewed and peaked (using robust
stats). Some difference between reading of paper1 and paper2. Unclear
if due to paper or instrument. Median/MAD: 0.03/0.01 and 0.14/0.05
iSis UV included vs. iSis UV excluded
No difference between readings.
iO UV included vs. iSis UV included
iSis is more consistent between readings.
iO UV excluded vs. iSis UV excluded
iSis is more consistent between readings for paper 1. No difference for paper 2.
Now over to the difficulties in interpreting these results.
It is unclear what introduces the iO's higher differences. Is it due
to less measurement accuracy or due to possible different paths while
reading catching test chart non-uniformity? Let's assume the latter,
that the iO takes two slightly different paths and the iSis takes the
exact same paths. Then, accuracy could be the same for both
instruments and still give results like above. The two different iO
readings would be valid representations of the test chart.
What do you others think?
Martin Weberg
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Colorsync-users mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden