JVC RS1/HD1 CC filter tweak - AVS Forum
Forum Jump: 
Reply
 
Thread Tools
post #1 of 132 Old 04-09-2007, 10:31 AM - Thread Starter
AVS Club Gold
 
Hughman's Avatar
 
Join Date: Jun 2004
Location: Canada
Posts: 2,247
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 12
After reading a little about using filters to cut down RGB levles when calibrating which maintains contrast ratio as apposed to cutting the levels digitally I thought I'd give it a try. I went on the cheap and purchased square glass filters from Cokin which were half price at the local shop, the filters slide into a housing therefore they can be stacked if neccessary. Taking a look at the relative RGB levels when using the projectors High temp and very blue mode it was clear substantial cuts in blue and green were required but not equally. I needed to bring blue down more than green in about a 1:3 ratio to get the two to match.

Orange filters absorb blue and green with the emphasis on blue so felt orange may be a perfect match but this still left how much total absorbtion required to bring the levels down to red. The outlet I was at only had one orange filter so the decision was easy and as it turns out was a little too aggressive therefore it was required to bring red down to match to new levels of blue/green so some lumens were needlessly lost but that's ok in my system as the screen way too bright anyway. As you can see from the new RGB histogram with the filter installed using the pj's high temp with all RGB levels at default maxed out positions I think I got pretty lucky. Don't think I could ask for a better result from a $15 filter with respect to bringing down blue and green.

During my readings I came across posts which indicated that while the filters may increase or actually maintain contrast there could be a hit in the representation of the colors.

So the big question, how much potential does this filter tweak have for increasing contrast ratio over calibrating only digitally. I don't have a dedicated light meter but am using an Eyeone Display sensor through HCFRcolorimater software. The sensor is placed on a tripod about 18 inches from the lens of the PJ to bring up black levels the meter can comfortably measure. I measured the CR for the PJ with 100% white calibrated to .2 dE from D65 using the RGB controls only plus another CR measurement using the native High Temp with the orange filter in place, 100% white then further digitally adjusted to within .3 dE from D65. My result are:

No filter D65 CR: 14038
D65/filter Hi CR: 17016
Hi temp default: 17125
LL
LL
Hughman is offline  
Sponsored Links
Advertisement
 
post #2 of 132 Old 04-09-2007, 11:04 AM
AVS Addicted Member
 
krasmuzik's Avatar
 
Join Date: Oct 1999
Location: NewPort, VA
Posts: 11,270
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
You need to measure Luv (xyY) of color - not just uv (xy). Is that pre/post filter gamut or a reference gamut vs. post filter? Does your software have dE targets for color? The CIE gamut chart is poor for visualizing color error difference as it covers the entire color space which is several hundred dE wide. You can see as small as a 3dE difference - which is not going to show on a gamut chart. This is evident in your greyscale appearing to be on target on the gamut chart - but clearly is 10-20dE off on the dE chart.

Here are charts of an FL-D filter on a SP7210 - in these older charts I was not measuring L for color as I do now - but you can see the impact on delC (chroma) and delH (hue). Basically what these charts say is the WMYR gamut got more red. These older charts do not record contrast/lumens - but if I recall it was something like 30% more contrast for 15% less lumens - and if I used white peaking and abused the gamma curve it was a 100% boost to contrast.

http://krasmuzik.biz/reviews/SP7210.pdf

Your max potential would be whatever contrast ratio it has at it's native maxed out color temp - presuming you find the filter that optically calibrates it.
krasmuzik is offline  
post #3 of 132 Old 04-09-2007, 11:35 AM - Thread Starter
AVS Club Gold
 
Hughman's Avatar
 
Join Date: Jun 2004
Location: Canada
Posts: 2,247
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 12
Hello Krasmuzik,

The gamut is post filter, they grayscale off due to there being some ambient light in the room therefore I spent no time on the entire grayscale, I made note of it somewhere in my OP. I'll be spending more time on it later tonight.

My software does provide color dE but alone probably won't mean a whole lot. I didn't run a reference measure without filter, the reference I have was performed last week but plan on running the entire series of measurements again tonight for so the two can be compared. I have the xyY values but will post them with dE from the measrements taken tonight.

I'm not quite at the max potential due the requirement to pull red down digitally quite a bit but am reasonably close. I measured CR at high temp last week in the mid 18's, but again that was another measurement session with a different set of variables ie: pitch black room.

You have to admit though the filter I picked was a stroke of good luck.

Appreciate further input when I get a full set of measrements.
Hughman is offline  
post #4 of 132 Old 04-09-2007, 11:40 AM
AVS Addicted Member
 
krasmuzik's Avatar
 
Join Date: Oct 1999
Location: NewPort, VA
Posts: 11,270
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Yes dE for color is not the whole story - but an overall error - be it hue, chroma or lightness. I use LCH for charting - but use dE (RMS) for overall grading. LCH is perceptually weighted so that the error units are balanced in each of the LCH dimensions.
krasmuzik is offline  
post #5 of 132 Old 04-09-2007, 11:40 AM - Thread Starter
AVS Club Gold
 
Hughman's Avatar
 
Join Date: Jun 2004
Location: Canada
Posts: 2,247
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 12
Krasmuzik,

What are the dE ring values in your posted charts. Now I see what you mean regarding screwing the colors.
Hughman is offline  
post #6 of 132 Old 04-09-2007, 11:44 AM
AVS Addicted Member
 
krasmuzik's Avatar
 
Join Date: Oct 1999
Location: NewPort, VA
Posts: 11,270
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
10dE, 30dE, 50dE...I had to standardize on 60dE as my chart limit - indeed I have measured some that go beyond that limit, but any further you lose the detail around the bulls eye!

dE is actually a sphere not a ring - but too confusing to represent errors in 3D - so the gamma deviation for greyscale and color lightness errors is charted seperately.

On my website I describe the charts - I convert dE (RMS) to percent by subtracting from 100 to assign grades. So roughly the rings are excellent, adequate or very poor.

http://www.krasmuzik.biz/reviews/reviewformat.htm

FYI this chart format is Copyright KRAS Muzik, LLC 2005-2007.

The colorspaces used however are CIE 1976 standard - so anyone is free to generate the numbers - I just do not want to see my chart format showing up in calibration software.


In this SP7210 example the tint control cannot fix the problem - the FL-D does not affect the Cyan hue while it pushes the Yellow and Magneta tints towards red (+delH is counterclockwise hue error, -delH is clockwise hue error). Only a proper CMS can fix that - which Infocus does not have. Be interesting to plot these charts showing color/tint control extremes - maybe next time I calibrate.
krasmuzik is offline  
post #7 of 132 Old 04-09-2007, 01:23 PM
AVS Addicted Member
 
noah katz's Avatar
 
Join Date: Apr 1999
Location: Mountain View, CA USA
Posts: 20,629
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 301 Post(s)
Liked: 167
Hugh,

That's a nice 20% increase in CR. How visible is it with program material, and especially, blackouts?

Noah
noah katz is offline  
post #8 of 132 Old 04-09-2007, 01:42 PM - Thread Starter
AVS Club Gold
 
Hughman's Avatar
 
Join Date: Jun 2004
Location: Canada
Posts: 2,247
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 12
Noah,

Too early to tell, I'll be viewing tonight with a black room. In theory this tweak should cure the bright corners somewhat by reducing ultimate black levels thereby lowering our sensitivity and also I believe that the bright corners is native blue heavy light leaking through therefore the filter should also reduce this blue excess of the native lamp color shining through again making the corners less visible if not invisible.
Hughman is offline  
post #9 of 132 Old 04-09-2007, 04:31 PM
Advanced Member
 
dreamer's Avatar
 
Join Date: May 2000
Location: Walnut, CA, USA
Posts: 974
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 81 Post(s)
Liked: 34
Did this Orange filter tame the neon oversaturated greens ?

*********************
Kirk Ellis
JVC DLA-RS1 VuTec 122" Screen
Harmon Kardon AVR 247 Parasound L&R Amp
Psycoustic Mark III L&R Towers, Boston Center
Energy Take 5 Surrounds, HSU Research Sub
dreamer is online now  
post #10 of 132 Old 04-09-2007, 05:36 PM - Thread Starter
AVS Club Gold
 
Hughman's Avatar
 
Join Date: Jun 2004
Location: Canada
Posts: 2,247
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 12
I've updated my first post with a new set of contrast ratios which are very close to what I measured earlier but now included is the CR at High Temp full output. Regarding full output at high temp, I found that with both sets of RGB adjustments at default 0 the panels were not fully driven, the offset controls allow levels to be increased to which I set both green and blue and +6 before the onset of clipping.

Attempting to calibrate with the filter in place has proved a little frustrating. For some reason there's a bit of a seesaw effect to reds response the fulcrum about in the middle of the grayscale. The best I could do provides a dE error at 20% of about 7.5 and 3.9 at 100%, the middle grays all approach dE of .75. Calibrating without the filter provides much more linear dE of under 1 until about 30% at which point blue takes off providing dE of 12 and more at 20% and lower. Not a big deal but the dark grays start out warm and get cooler the closer to 100% you get.

The filter has increased gamma slighly to an average of 2.24, the highest I could muster without crushing blacks sans filter was 1.15. there is a slight anomaly here though, with filter in place green and blue track together at above 2.24 red has taken it's own course however at a lower gamma of about 2.18.

Regarding color, here's the before and after xyY numbers for primaries and secondaries. I'll also include the dE as laid out in the CIEu'v' chart. Forget it I can't seem to do a list that is legible, if someone could provide instruction for lists I'd appreciate it. For now here's the dE numbers for rec709 standard. Keep in mind the dE are high do to the JVC's large color Gamut.

....N/F..Filter
R 19.5 21.5
G 38.0 34.6
B 45.9 41.5
Y 20.7 21.1
C 19.4 17.5
M 15.9 18.4

No filter average dE: 25.567, with filter average dE: 25.767.

So to sum up thus far effects of using this filter:

1- increased contrast ratio of 21% from 14038 to 17016

2- Decrease in light output of about 25%, this filter is more aggressive than required but works wonders in my too bright setup.

3- Significant decrease in bright corner issue, one of my corners actually took up about 1/5 of the screen. Not only has the perceived area been reduced to about half that the color of remaining area is less offensive.

So far so good.
Hughman is offline  
post #11 of 132 Old 04-09-2007, 05:42 PM - Thread Starter
AVS Club Gold
 
Hughman's Avatar
 
Join Date: Jun 2004
Location: Canada
Posts: 2,247
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 12
Quote:
Originally Posted by dreamer View Post

Did this Orange filter tame the neon oversaturated greens ?

Green moved slightly, but am not sure how noticeable it is. I'm not too sure that the neon and oversaturion are the same issue. Oversaturation would make the colors deeper in color, the neon or glowing issue seems more of an color intensity of brightness issue. Maybe they go hand in hand but I'm not entirely convinced as yet.
Hughman is offline  
post #12 of 132 Old 04-09-2007, 06:30 PM
AVS Addicted Member
 
krasmuzik's Avatar
 
Join Date: Oct 1999
Location: NewPort, VA
Posts: 11,270
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Your GCB gamut has less error - your CYR gamut has more error- this is why your average did not change much - so it has a tradeoff really since it is just a shift in balance for the color errors.

The color intensities must be way off from color decoder requirements - as your Blue did not look that far off on the gamut chart like Green is. Possible your sensor is just not good at reading dark blues...did you get as close to lens as possible to increase light output?

Be interesting to break down the xyY raw data to LCH color space and see what the errors are - but I don't know how to post a list either! Maybe attach an Excel file? Since LCH can all be over/under target- you get a better picture looking at it. Does dL, dC, dH each shift from +20 to -20 for no net dE change - but a large perceptual change?
krasmuzik is offline  
post #13 of 132 Old 04-09-2007, 07:41 PM - Thread Starter
AVS Club Gold
 
Hughman's Avatar
 
Join Date: Jun 2004
Location: Canada
Posts: 2,247
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 12
I'm looking for a colorspace converter, found one but it's not performing the said tasks, I'll keep working on it. I can't even post an excel, recently upgraded my computers and they don't even have the basic work spreadsheet on them anymore, what a rip!

I only got close to the lens when measuring contrast ratio, all other measurements the meter faced the HP screen. I've performed dozens of measures and the results of blue are always consistant which provides me some faith it's capable of, at the very, reading the light level. I've noticed anything about 1 cd/m2 the cut-off for faithfull repeatability, the 75% blues always fall above that. Full resolution CIE charts show the errors between red green and blue about equal visually, I suppose if the filter is moving the primaries non-relative to each other this may be causing the error in secondaries as it's not as JVC intended. Since the JVC does not have a blue screen mode I'm using filters to adjust the color decoder which as you is not the most accurate method but all three colors (utilizing the squint method) look bang on, better than I've ever witnessed.

I'll do my best to covert xyY to LCH but not right now, all this typing is cutting into my drinking time and that's something I'll never get back.
Hughman is offline  
post #14 of 132 Old 04-09-2007, 09:17 PM - Thread Starter
AVS Club Gold
 
Hughman's Avatar
 
Join Date: Jun 2004
Location: Canada
Posts: 2,247
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 12
Just realized I haven't provided any info on the filter. It is a Cokin P029 85A Color Conversion Resin Filter. Here's a photo of the filter in the separate housing.
LL
Hughman is offline  
post #15 of 132 Old 04-09-2007, 09:30 PM
AVS Special Member
 
lovingdvd's Avatar
 
Join Date: Jul 2002
Posts: 8,943
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 97 Post(s)
Liked: 46
Quote:
Originally Posted by Hugh2 View Post


So to sum up thus far effects of using this filter:

1- increased contrast ratio of 21% from 14038 to 17016

2- Decrease in light output of about 25%, this filter is more aggressive than required but works wonders in my too bright setup.

3- Significant decrease in bright corner issue, one of my corners actually took up about 1/5 of the screen. Not only has the perceived area been reduced to about half that the color of remaining area is less offensive.

So far so good.

Thanks this is very informative. Can you take a before/after shot of your CIE chart? Have you measured ANSI CR before/after? Word is that CCF filters may have an adverse affect on ANSI CR so it would be useful to know if that applied in this case. Lastly, can you discern a difference that the 21% increase in on/off CR makes?
lovingdvd is online now  
post #16 of 132 Old 04-10-2007, 07:00 AM - Thread Starter
AVS Club Gold
 
Hughman's Avatar
 
Join Date: Jun 2004
Location: Canada
Posts: 2,247
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 12
Kras,

I found a conversion calculator which converted my xyY values to LCHuv: But at this point can not find the standards for LCH?? Here's a "list" of numbers
  • ________ L__________C_________H

NO FILTER
  • R ____ 45.033943 158.863377 13.149837
  • G____ 55.213175 106.808380 128.961929
  • B____ 21.341089 93.334756 264.908006
  • Y____ 71.084812 92.959002 78.415043
  • C____ 61.953332 65.682267 208.892535
  • M____ 49.453629 113.891737 311.832498

FILTER
  • R ____ 34.703447 122.402773 13.011432
  • G____ 55.056884 104.487092 128.029202
  • B____ 16.815835 73.543663 264.908006
  • Y____ 62.418050 81.711387 85.416330
  • C____ 58.01901 56.013365 188.930635
  • M____ 38.669600 93.224710 306.413141


The changes in Lightness, Chroma, and Hue are rather interestng. Can you steer me towards the standards and perhaps a Delta calculator. I can do the calcs using LAB but not for LCH. An obvious question, where chroma is >100 and the model for chroma is 0-100 is the LCH chroma space smaller than CIE space.

Here's a link to the Bruce Lindbloom converter I used:
http://brucelindbloom.com/index.html?ColorCalcHelp.html
Hughman is offline  
post #17 of 132 Old 04-10-2007, 07:10 AM - Thread Starter
AVS Club Gold
 
Hughman's Avatar
 
Join Date: Jun 2004
Location: Canada
Posts: 2,247
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 12
Here are the xyY coordinates

Filtered

R...660....336...8.351
G...293....704..22.137
B...139....049...2.264
Y...448....550...30.894
C...206....336...25.981
M...317....146...10.468

100% white Y 46.826

No Filter

R...661....337...14.566
G...286...712....23.137
B...139....049....3.357
Y...463....535...42.311
C...193....291...30.348
M...334....158...17.965

100% white Y 61.782

Rec.709 standard (Tom, I stole this from your post)

Rec. 709-------x-------y-------Y
R--------------0.640---0.330---0.213
G--------------0.300---0.600---0.715
B--------------0.150---0.060---0.072
C--------------0.225---0.329---0.787
Y--------------0.419---0.505---0.928
M--------------0.321---0.154---0.285
White----------0.313---0.329---1

If I'm reading this correctly then based on the filtered results I should be cranking up the color brightness control?? Using DVE's RGB filters the color brightness appear almost perfect.
Hughman is offline  
post #18 of 132 Old 04-10-2007, 08:39 AM
Member
 
neilher's Avatar
 
Join Date: Dec 2005
Posts: 62
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Hugh,
Is that a Cokin plastic filter?????????, are you getting any distortion from the heat on the plastic ??????????? How far are you in front of the lense????????? I thought about using a ND filter, buts its glass.
Thanks
neilher is offline  
post #19 of 132 Old 04-10-2007, 08:48 AM - Thread Starter
AVS Club Gold
 
Hughman's Avatar
 
Join Date: Jun 2004
Location: Canada
Posts: 2,247
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 12
Quote:
Originally Posted by neilher View Post

Hugh,
Is that a Cokin plastic filter?????????, are you getting any distortion from the heat on the plastic ??????????? How far are you in front of the lense????????? I thought about using a ND filter, buts its glass.
Thanks

It's labelled as resin, feels just like glass. The filter is about 3/4 inch in front of the lens with no heat issues.
Hughman is offline  
post #20 of 132 Old 04-10-2007, 08:58 AM - Thread Starter
AVS Club Gold
 
Hughman's Avatar
 
Join Date: Jun 2004
Location: Canada
Posts: 2,247
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 12
Ok, some more numbers to look at. I located a Delta E calculator for LAB colorspace which appears to resemble closely LCH so I converted all my xyY measurements to LAB. I then calculated a different reference set of values for the non-fltered and filtered setup using the luminance value of each.

Here's the calculator:
http://www.colorspan.com/support/tools/deltae.asp

Here's the Delta E chart:

............Filtered.........Non-filtered
R.............7.34..............19.58
G...........31.2...............77.43
B.............4.72...............6.24
Y...........31.18.............41.68
C...........11.5...............17.99
M............4.93................3.67

Ave.......15.145...........27.765
Hughman is offline  
post #21 of 132 Old 04-10-2007, 09:07 AM
AVS Special Member
 
Gary Lightfoot's Avatar
 
Join Date: Jun 2000
Location: Surrey, UK
Posts: 4,487
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 31 Post(s)
Liked: 44
You may notice some more lens flare and reduced ANSI with the resin filters. I would try a HMC Hoya equivalent as they have less effect on the image compared to the Cokin ones. I noticed that when I tried a Cokin FL-D filter so for long term use I now use Hoya.

Gary

Quote:
Originally Posted by elmalloc
Who says Cameron is "right" and why do we care about him so much - lol!

I trust Gary Lightfoot more than James Cameron.
Gary Lightfoot is offline  
post #22 of 132 Old 04-10-2007, 09:21 AM - Thread Starter
AVS Club Gold
 
Hughman's Avatar
 
Join Date: Jun 2004
Location: Canada
Posts: 2,247
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 12
Hello,

I just picked up something very inexpensive for an experiment. I''ve been enjoying the improvement, mainly in reduced foot Lamberts and better blacks, but I've have yet to notice any negatives with the filter. I plan on purchasing the two lighter versions of this Cokin, if I can get the brigtness just right I'll not even need to adjust the red more than a few clicks plus I can throw those on to compensate for lamp brigtness loss if desired.
Hughman is offline  
post #23 of 132 Old 04-10-2007, 09:25 AM
AVS Special Member
 
MikeSRC's Avatar
 
Join Date: Dec 1999
Location: San Clemente, CA
Posts: 5,921
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 4 Post(s)
Liked: 11
Beyond measurement, have you noticed any changes in hue? Often your reds will look orangey with an 85 series filter.

BTW, here's a chart of how the 85 series filters affect the spectrum.
LL

Enjoy!

Mike
Surf Remote Control
Contributing reviewer and THX calibrator, ProjectorReviews.com
MikeSRC is offline  
post #24 of 132 Old 04-10-2007, 09:38 AM - Thread Starter
AVS Club Gold
 
Hughman's Avatar
 
Join Date: Jun 2004
Location: Canada
Posts: 2,247
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 12
I haven't noticed any shift of red hue to orange, if anything it appears deeper red but this I'm sure just a product of reduced brightness. I really can't see any negatives being introduced by using this filter yet. As time progresses that's likely to change, overall the picture is stunning but it was kinda like that before just a little too bright.
Hughman is offline  
post #25 of 132 Old 04-10-2007, 10:06 AM - Thread Starter
AVS Club Gold
 
Hughman's Avatar
 
Join Date: Jun 2004
Location: Canada
Posts: 2,247
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 12
Quote:
Originally Posted by lovingdvd View Post

Thanks this is very informative. Can you take a before/after shot of your CIE chart? Have you measured ANSI CR before/after? Word is that CCF filters may have an adverse affect on ANSI CR so it would be useful to know if that applied in this case. Lastly, can you discern a difference that the 21% increase in on/off CR makes?

The CIE chart is esentially in number form a few posts up, it's easier to discern differences this way. I have not measured ANSI CR but I definitely don't perceive any drop and contrarily the perception is an improvement.

It's difficult to say if I can perceive the additional CR, the increased CR is accompanied by lower foot lamberts on the screen and both give the same perceived end result. As average scene brightness increases my rooms ability to support either contrast ratio decreases or disappears and eye bias takes over. There is a definitely a new sense of palpability and depth to the picture which could be attributed to the increase in CR but there's so many variable it's difficult to attribute this solely to CR though I'd like to.
Hughman is offline  
post #26 of 132 Old 04-10-2007, 10:43 AM
AVS Addicted Member
 
krasmuzik's Avatar
 
Join Date: Oct 1999
Location: NewPort, VA
Posts: 11,270
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Hugh2

The color equations are documented on that site.

Did you normalize Y (L) to 100 for white with each measurement set - that is the only way you can compare - white is always LCH (100,0,0). I suspect since your dE is less with the filter - you are using the white without filter as your reference. You need the filtered dE based on white with the filter, likewise nonfiltered dE based on white without the filter.

And I would use lens readings - while you may have repeatability this could be the software doing some averaging - you want accuracy for these primary measures which means more light on the sensor. Or just size down to a 2' image on screen.

Your sensors absolute accuracy is not all that important - what matters is the relative accuracy so that the delta measures are meaningful. Unless you want to compare to the references rather than just w/w.o. filter.

LAB and LCH have the same dE - it is take the root of the sum of the squares of the three target differences for each color. Make sure you use LCH though and not LCh for that to work.

dC, dL, dH are just the difference from each other - or the reference target however you want to express it.
krasmuzik is offline  
post #27 of 132 Old 04-10-2007, 10:55 AM
AVS Addicted Member
 
krasmuzik's Avatar
 
Join Date: Oct 1999
Location: NewPort, VA
Posts: 11,270
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Looking at your data the Chroma of Red increases and Blue decreases while Green was same and the Hues of the secondaries are perceptually changed as well as their Chroma reduced. Chroma includes Lightness - so you have to look at that to figure is the color stronger because it is brighter or more saturated.
.
.
.

Never mind I will recomment once you have indicated the data is properly normalized - WTF this is more interesting than taxes - I'll spreadsheet your data for you.
krasmuzik is offline  
post #28 of 132 Old 04-10-2007, 11:25 AM
AVS Special Member
 
shodoug's Avatar
 
Join Date: Dec 2000
Posts: 1,262
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by Hugh2 View Post

I can't even post an excel, recently upgraded my computers and they don't even have the basic work spreadsheet on them anymore, what a rip!

Hugh,

Not sure from your post, but in case it might be helpful there is an open source office suite called Open Office.

You can download it here.

I just played around a little, but it is free, seemed intuitive, and the non-computer geek I installed it has continued to use it, and exchange documents with users of MS Office, on a regular basis.

The tough part was figuring out which program was the spreadsheet, which was the word processor, etc.

Just a suggestion, i realize that it might not be feasible on a work computer administered by someone else, etc.

Doug
shodoug is offline  
post #29 of 132 Old 04-10-2007, 11:27 AM - Thread Starter
AVS Club Gold
 
Hughman's Avatar
 
Join Date: Jun 2004
Location: Canada
Posts: 2,247
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 12
Quote:
Originally Posted by krasmuzik View Post

Hugh2

Did you normalize Y (L) to 100 for white with each measurement set - that is the only way you can compare - white is always LCH (100,0,0). I suspect since your dE is less with the filter - you are using the white without filter as your reference. You need the filtered dE based on white with the filter, likewise nonfiltered dE based on white without the filter. .

Yes, I normalized my reference to the Y of each measurement set. Didn't catch that little oversight till about half-way through my calculations though, but I did catch it.

Quote:
Originally Posted by krasmuzik View Post

And I would use lens readings - while you may have repeatability this could be the software doing some averaging - you want accuracy for these primary measures which means more light on the sensor. Or just size down to a 2' image on screen..

Your sensors absolute accuracy is not all that important - what matters is the relative accuracy so that the delta measures are meaningful. Unless you want to compare to the references rather than just w/w.o. filter.

With the meter facing the screen I get excellent repeatability even at 20% and moreso when measuring primaries so I sort of trust the results. I've read a few posts about meters accuracy vs amount of light though that I will heed your advice and get more into the meter to see if anything changes.

I'm not going to claim my measuring equipment is reference quality but overtime of using I do get a sense that it's at the very least consistant. Barring significant LAB dE differences when pointed towards the lens I'm pleased with the measured performance of the lens, and more importantly the image looks really big and juicy.
Hughman is offline  
post #30 of 132 Old 04-10-2007, 11:29 AM - Thread Starter
AVS Club Gold
 
Hughman's Avatar
 
Join Date: Jun 2004
Location: Canada
Posts: 2,247
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 12
Quote:
Originally Posted by krasmuzik View Post

Looking at your data the Chroma of Red increases and Blue decreases while Green was same and the Hues of the secondaries are perceptually changed as well as their Chroma reduced. Chroma includes Lightness - so you have to look at that to figure is the color stronger because it is brighter or more saturated.
.
.
.

Never mind I will recomment once you have indicated the data is properly normalized - WTF this is more interesting than taxes - I'll spreadsheet your data for you.

LOL, I've been procrastinating that all well. If it wasn't for procrastination nothing would get done. Thanks for the help.
Hughman is offline  
Reply Digital Hi-End Projectors - $3,000+ USD MSRP

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