Possibly. If you look at the RBG plot I posted you can see that the colors that Derek measured are much closer between the i1pro and D2 than in Tom's test.

Tom: Did you try more than one D2? I don't recall.

RIch
And I almost forgot...in the same thread, Tom's D2 tracked the i1 Pro very well for LCD and DLP FP/RP displays, but the 2 plasmas were way off. So my "bum" D2 theory is shaky.

-JD

The thing that is throwing me for a loop in this discussion is the D2/i1Pro graphs that TomHuffman posted. They seem to be much more out of line than Bill's results (even considering that we don't have Tom's xyY values).


Could it be that Tom has a bum D2 (or conversely, that Bill has a good one)?

-JD

Don't forget that the numbers you quote from Tom were at 100 IRE, if you look at the mean over 30-90 IRE the results are not as severe:

R: 6% D2 > i1pro
G: 3% D2 < i1pro
B: 4% D2 < i1pro


Tom said he would post the xyY values of that plus some additional measurements but hasn't done so yet.

And I almost forgot...in the same thread, Tom's D2 tracked the i1 Pro very well for LCD and DLP FP/RP displays, but the 2 plasmas were way off. So my "bum" D2 theory is shaky.

-JD
I don't think it's a "bum" D2 either, I think what we are seeing is that the D2 is not specifically calibrated for use with a plasma which has a very different SDF then any other display out there, especially a very strong green peak at 540 nm and strong red lines near the edge of the photopic response function. The interesting questions are under what conditions this inaccuracy is noticeable, does it exceed the manufacturer's accuracy claims, and is there any thing a user can do about it. As you probably can tell I am highly skeptical on that last point.

Don't forget that the numbers you quote from Tom were at 100 IRE, if you look at the mean over 30-90 IRE the results are not as severe:

R: 6% D2 > i1pro
G: 3% D2 < i1pro
B: 4% D2 < i1pro


Tom said he would post the xyY values of that plus some additional measurements but hasn't done so yet.

Do you think it is significant that these are actually:

R: +6% D2 > i1pro
G: -3% D2 < i1pro
B: -4% D2 < i1pro

EDIT: Never mind - I just noticed your "<" and ">" symbols :rolleyes:

Question about HCFR's ANSI measures. It asks for 4 measures, 100IRE, 0IRE....and then an ANSI 1 and ANSI 2.

I'm assuming the ANSI measures are coming from the traditional ANSI 'checkboard' patterns (like on getgray). Is ANSI 1 suppose to the be the 'white' or 'black' area of the checkerboard? With ANSI 2 being the opposite?

I had asked about that feature in a PM and was told to ignore it for now.

Question about HCFR's ANSI measures. It asks for 4 measures, 100IRE, 0IRE....and then an ANSI 1 and ANSI 2.

I'm assuming the ANSI measures are coming from the traditional ANSI 'checkboard' patterns (like on getgray). Is ANSI 1 suppose to the be the 'white' or 'black' area of the checkerboard? With ANSI 2 being the opposite?

Colin,
I think you are talking about the dynamic range (contrast ratio) measurements right? I remember reading somewhere that the ANSI numbers refer to the black and white squares in the checkerboard pattern like on getgray as you said, and that the ANSI contrast ratio is a smaller number that the regular (max) contrast ratio that vendors quote. Not sure about the digital values for the ANSI ratio, maybe they are 235 and 16? Max contrast ratio on the 9UK plasma is an interesting thing. To get maximum value you need to use a small window so power control doesn't limit the brightness. I just use the getgray 100% window, but even this is probably too large. The hard part is what to use for the black measurement. I think it should be 0%, i.e., btb. The darkest screen you can get without turning off the display basically. SOmetimes it seems like mine is a little bit brighter than total black for a while (5 secs) then drops down to black after this. Have you seen anything like this?

RIch

Hi Rich --

For full-blown 100IRE reading (non-ansi), ya, I know you need to use the window to avoid the power limiter.

But for ANSI, I'd think measuring it using the full checker patter (whether it hits power limiter or not) is the point isn't it? -- for a 'realistic' ansi measure of what white level would be like in a real viewing environment?

Hi Rich --

For full-blown 100IRE reading (non-ansi), ya, I know you need to use the window to avoid the power limiter.

But for ANSI, I'd think measuring it using the full checker patter (whether it hits power limiter or not) is the point isn't it? -- for a 'realistic' ansi measure of what white level would be like in a real viewing environment?

That sounds like what I remember reading pretty much. Maybe someone with more experience could chime in. I only take the max contrast ratio measurement myself.

Hi Rich --

For full-blown 100IRE reading (non-ansi), ya, I know you need to use the window to avoid the power limiter.

But for ANSI, I'd think measuring it using the full checker patter (whether it hits power limiter or not) is the point isn't it? -- for a 'realistic' ansi measure of what white level would be like in a real viewing environment?

yes for ANSI use the checkerboard. I don't remember which is first, black or white, but if you do it wrong you'll get weird numbers out and then just do it the opposite.

Hi Colin
Question about HCFR's ANSI measures. It asks for 4 measures, 100IRE, 0IRE....and then an ANSI 1 and ANSI 2.

I'm assuming the ANSI measures are coming from the traditional ANSI 'checkboard' patterns (like on getgray). Is ANSI 1 suppose to the be the 'white' or 'black' area of the checkerboard? With ANSI 2 being the opposite?
Yes, that's it. Ansi 1 and Ansi 2 are checkboard patterns, with inverted black and white squares. You can measure any square first, a black or a white. The software consider the lowest level as black, and the highest as white. So, you can measure all positions with the same patterns.

We think it is not really a very interesting measure, so we don't "push" it. Colorimetric probes are rarely sensitive enough to measure correctly modern projectors contrast. Their dynamic is not sufficient for that. Anyway, programming contrast measures is the easiest thing to do, that's why you can find it in any software. But the calculated value cannot be accurate because of probes limitations.

If you really want an accurate value for contrast, you should use a luxmeter...

Regards
Georges

can you recommend a good luxmeter?

Hi Gino
can you recommend a good luxmeter?
Well, in this domain, you often have what you spend for ;) I have an old LX-1108, which is quite expensive (around $120-$130) regarding its precision. But it is the cheapest luxmeter with built-in RS-232 interface, and it is easy to find worldwide... I thought it was a good idea to choose this one because maybe one day, using its RS232 interface, I could connect it to ColorHCFR... But if you want a basic luxmeter, you can buy any 30$ one, which will give same results, without RS232 interface ;)

Note: if you want to measure precisely contrast ratio for a latest-generation projector like HD1, you need really precise and very sensitive instruments. 30$ luxmeter are not able to do that. They are better than any colorimeter, but professional instruments are better suited for this exercise... Anyway, do we really need to know if a contrast ratio is exactly 12575 or 13010 ? :D

Regards
Georges

Is there an English instruction set for building the HCFR Probe?

Hi Gino

Well, in this domain, you often have what you spend for ;) I have an old LX-1108, which is quite expensive (around $120-$130) regarding its precision. But it is the cheapest luxmeter with built-in RS-232 interface, and it is easy to find worldwide... I thought it was a good idea to choose this one because maybe one day, using its RS232 interface, I could connect it to ColorHCFR... But if you want a basic luxmeter, you can buy any 30$ one, which will give same results, without RS232 interface ;)

Note: if you want to measure precisely contrast ratio for a latest-generation projector like HD1, you need really precise and very sensitive instruments. 30$ luxmeter are not able to do that. They are better than any colorimeter, but professional instruments are better suited for this exercise... Anyway, do we really need to know if a contrast ratio is exactly 12575 or 13010 ? :D

Regards
Georges

Georges,
Are you saying the luminance readings obtained with HCFR and a basic colorimeter (like D2) are not as accurate as a $30.00 luxmeter? I don't understand.

Thanks,
Rich

I would also like to know that, Rich.

Maybe a more important question: Is using a Lux going to give more accurate information as it relates to Gamma fitting? If so, sounds like 30 bucks would be worth it. But if something better than a colorimeter is only going to give a slightly more accurate on/off or ANSI contrast ratio reading...i guess i wouldn't care.

Hi HCFR Team

I use a program called "Colorcop" to examine the RGB values of the Grayscale in the HCFR Calibration DVD Disk. It turns out that the RGB values of 0IRE are (0,0,0) and RGB values of 100IRE are (253,253,253).

Shouldn't they be (16,16,16) and (235,235,235)?


For the Red Bar in the Disk, the RGB values are (252,0,0). Shouldn't they be (235,16,16)?

For the Green Bar in the Disk, the RGB values are (0,253,0). For the Blue Bar, the RGB values are (0,0,251). Are they correct?


JM

Hi HCFR Team

I use a program called "Colorcop" to examine the RGB values of the Grayscale in the HCFR Calibration DVD Disk. It turns out that the RGB values of 0IRE are (0,0,0) and RGB values of 100IRE are (253,253,253).

Shouldn't they be (16,16,16) and (235,235,235)?


For the Red Bar in the Disk, the RGB values are (252,0,0). Shouldn't they be (235,16,16)?

For the Green Bar in the Disk, the RGB values are (0,253,0). For the Blue Bar, the RGB values are (0,0,251). Are they correct?


JM

I think you are correct that 0 IRE and 100 IRE correspond to digital inputs 16 and 235 respectively, however, the purpose of calibration is to make 16 indistinguishable from 0 (by adjusting brightness) and 235 indistinguishable from 255 (by adjusting picture if possible). So I am not surprised that you would measure these values when looking at a well calibrated set.

Rich

Colorcop is a PC applet that measures RGB values of anything you put a "dropper" over. Basically the same idea as DigitalColor Meter on a Mac.

Georges,
Are you saying the luminance readings obtained with HCFR and a basic colorimeter (like D2) are not as accurate as a $30.00 luxmeter? I don't understand.

Thanks,
Rich

I think Georges is referring to measuring very high contrast ratios (> 5000) on front projectors where you have to use a diffuser to measure incident light. In that case lux meters are better at the low end since they are broadband (no color discrimination).

Hi Rich
Georges,
Are you saying the luminance readings obtained with HCFR and a basic colorimeter (like D2) are not as accurate as a $30.00 luxmeter? I don't understand.

Well, I can say "yes", and I can say "no". It's only a question of accuracy, relative to light level. A colorimeter is not enough sensitive to give you a measure of black light level. A $30 luxmeter can better measure black, but it is also obviously inaccurate for that. Let's take an example. You want to see the difference between black (7.5 IRE) and very very dark gray (10 IRE). Theoretically, 7.5 IRE is absolute zero, and 10 IRE is 2.7% of full light output. Practically, 7.5 IRE have some lights and 10 IRE have a little more... how much ? very few... and you are exactly in the same kind of difference than the difference between two graduation on your instrument. Sometime the instrument sees the difference, sometime not. And repeated measures are never equal... You see the trouble ? ;)

Then, practically, you can consider the luxmeter is better. But try to measure very low light output with a luxmeter, you will see the measure is not perfectly stable. If you use a colorimeter, it's the same: stability is not perfect. Globally, a colorimeter is accurate enough to measure gamma globally, with a very very little error margin. But measuring exact black level, and use it to compute a contrast ration need much better accuracy, except if you consider acceptable a +/- 30% value ;)

Actually, I don't know if we could have any real benefit using a luxmeter. It's a subject of research, which needs many experiences to know if yes or no there is a benefit.

Regards
Georges

Hi Joseph
I use a program called "Colorcop" to examine the RGB values of the Grayscale in the HCFR Calibration DVD Disk. It turns out that the RGB values of 0IRE are (0,0,0) and RGB values of 100IRE are (253,253,253).

Shouldn't they be (16,16,16) and (235,235,235)?

No, don't worry about it. A DVD is encoded in YUV format, and what you see in your PC is the result of a translation YUV to RGB. What you measure is the result of the translation, and depending on your software or hardware, the scale can be either 13-235 or 0-255, and usually on a PC screen it's 0-255.

Regards
Georges

Actually, I don't know if we could have any real benefit using a luxmeter. It's a subject of research, which needs many experiences to know if yes or no there is a benefit.

Regards
Georges

I don't think a $30 luxmeter would be any benefit. The minimum rated illuminance for these is typically 1 lux. That's equivalent to a (Lambertian) screen luminance of ~.3 cd/m^2 The D2 has a sensitivity a factor 10 better than that (0.02 cd/m^2).

Anyone have recommendation on getting this to work with a macbeth eye one?

I can't seem to get it up and running. I keep getting a dll error and it crashes.

Anyone have recommendation on getting this to work with a macbeth eye one?

I can't seem to get it up and running. I keep getting a dll error and it crashes.

have you copied the specific .DLL file thats in the Macbeth folder and paste in the ColorHCFR folder where the program .exe file is at? You need to do this for it to work..

have you copied the specific .DLL file thats in the Macbeth folder and paste in the ColorHCFR folder where the program .exe file is at? You need to do this for it to work..

Nope, thanks!! I'll give that a try,

Nope, thanks!! I'll give that a try,


Off topic - ..EARTHQUAKE CINENOVA owner here!!!

I’ve received a few questions about calibration files for the HCFR DIY Colorimeter which has prompted me create a document that describes some of the methods and procedures that hopefully will answer some questions about it.

Interested parties can access the pdf document HERE (http://members.aceweb.com/dlarsen/gerbers/HCFR_Calibration_Files.pdf) .

Dave

I’ve received a few questions about calibration files for the HCFR DIY Colorimeter which has prompted me create a document that describes some of the methods and procedures that hopefully will answer some questions about it.

Interested parties can access the pdf document HERE (http://members.aceweb.com/dlarsen/gerbers/HCFR_Calibration_Files.pdf) .

Dave

Nice Job Dave..added this to my arsenal (ahemm..black binder)

Does anyone know if the HCFR probe is more accurate than the Spyder2 probe when using it with an RPTV? Specifically with the Sony KDS-R60XBR2 SXRD.

Does anyone know if the HCFR probe is more accurate than the Spyder2 probe when using it with an RPTV? Specifically with the Sony KDS-R60XBR2 SXRD.


You need to have an existing calibration file that exist with ColorHCFR for your set. If it does, then YES, it will be as accurate as the probe that was used to create the calibration file . You can also use the Spyder to create a calibration file and use the ColorHCFR probe to read down lower, but it wont be more accurate than your Spyder but it will read lower than the Spyder (so you will basically TRAIN the ColorHcfr probe from the SPyder)

In terms of absolute accuracy, it escapes me as to how any instrumentation can be any more accurate than the instrumentation use to calibrate or verify it. (except by dumb luck)

If I used a S2 as reference or to train the HCFR colorimeter from, I can’t reasonably expect it to be anymore accurate than that. As Rich noted, once calibrated, it can perhaps offer better range, sensitivity, or repeatability etc however. If the HCFR sensor is trained and calibrated off something much more ‘accurate’ than the S2 would it then be more accurate than the S2? I dunno. An S2 is my only frame of reference.

I believe the HCFR style sensor is more sensitive down low then MY S2 is. I don’t really trust the data from the S2 in my environment at much below 15 IRE. It just seems to diverge too much and isn’t very repeatable. I trust the data from the HCFR sensor more down low but it is somewhat a case of trust or faith. I currently have no way of knowing which is really more ‘accurate’.

I can say that in the instances that I’ve made calibration files and runs via simultaneous measurements with the S2 and HCFR sensor in parallel at the same time, I’ve seen the delta deltaE <2 between the two meters over the 15-100 IRE range.

Dave

Nice job Dave, and yes, of course, the probe cannot be more accurante than the one use to produce profile files.
I think if offers better range, sensitivity and repeatability than many, but cannot be more accurate...
BTW, We generaly use a Eyeone Beamer to build the given profiles files.

--Patrice

Hi FGM,

I'm sorry but I'm not familiar with the HCFR test patterns, I use the getgray ones but I would imaging HCFR has similar patterns. The way I set color saturation (which I've assumed is your "lightness" setting) is to take a 75% white pattern and record the blue luminance reading using the free measures window. (If you click on the display measures icon when you have free measures running, it will bring up a window which includes the RGB readings). I then put up a 75% blue pattern and adjust saturation (color) until it matchs the blue reading from the white pattern.

hope that helps,

-scott

Hi Scott,

Did you mean adjust colour control as in the attached file?


JM

yes, I use the 75IRE (%) patterns though, sometimes my probe (D2) acts up at 100 IRE.

BTW, We generaly use a Eyeone Beamer to build the given profiles files.
BTW, any chance you will be generating one for the hugely popular Mits HD1000 ?? :D

:cool:

Patrice,

What's the latest on the next version?

Hello.
Anyone know of any calibration software for Mac? (I don't own a PC but would love to calibrate my Sony KDL52-XBR3 with my Eye One Display2.)

Try running Virtual PC.

Patrice,

What's the latest on the next version?

Yes, i was wondering what the major changes are going to be in the next release also.

Thanks guys!
Carmine.

Off topic - ..EARTHQUAKE CINENOVA owner here!!!

I'm not sure what you are referring to?? You own one?

Hi,

It is a bit early to say it all, but I can share we are debating if we will name it v1.3 (as planned) or v2.0 (as features and all changes may indicate it is) ;)

So far I can tell you that beside tacking in account most (if not all) feedback requests, we also have fully revamp GUI and document management, and added a bunch of new features both in the pure measurement area and around it (basic settings, patterns...)

Also looking at a guided mode (beginner/no time to spend on learning this f.....g software).

We are now in advance beta stage, so hopefully some beta version may circulate to testers "soon"...

--Patrice

I'm not sure what you are referring to?? You own one?
look at your signature..its says YOU are a owner also of this amp

look at your signature..its says YOU are a owner also of this amp
Yep, I do and I love it...

I thought it was some smart alec remark like owning Bose, hopefully I was wrong. ;)

Yep, I do and I love it...

I thought it was some smart alec remark like owning Bose, hopefully I was wrong. ;)

Dude - you couldnt pry the Earthquake away from me..so you were WAY OFF..that baby is a beast..

We are now in advance beta stage, so hopefully some beta version may circulate to testers "soon"...
Awesome! Looking forward to it.

Awesome! Looking forward to it.

Hey Gino

Have you done any calibrations with ColorHCFR for your G90's

Hi,

It is a bit early to say it all, but I can share we are debating if we will name it v1.3 (as planned) or v2.0 (as features and all changes may indicate it is) ;)

So far I can tell you that beside tacking in account most (if not all) feedback requests, we also have fully revamp GUI and document management, and added a bunch of new features both in the pure measurement area and around it (basic settings, patterns...)

Also looking at a guided mode (beginner/no time to spend on learning this f.....g software).

We are now in advance beta stage, so hopefully some beta version may circulate to testers "soon"...

--Patrice
Great! I have a new D2 to play around with!

Dude - you couldnt pry the Earthquake away from me..so you were WAY OFF..that baby is a beast..

heheh, this is true. That baby signs..... Although there are some haters out there of this amp. It's my best amp to date....dynamic and gobs of headroom.

Now if I can get my picture to look how my system sounds.... :(

Hey Gino

Have you done any calibrations with ColorHCFR for your G90's

G90's??? I'm blending with 9500LC Ultra's (albeit heavily modified so probably better than a G90)

I have calibrated them with ColorHCFR... worked a treat! Took a long time though

I'm thinking of picking up 2 Lumagen Radiance's for tweaking the gamma and grey scale

I'm wondering if anyone else has noticed this or is it just my equipment.... when running the 10 grey windowed IRE test screens, 0% does something strange with Getgrey and Avia on my Panasonic 50PX60U plasma. The 0% box changes intensity after about 5 or 6 seconds. It starts out barely bright enough to read the text and then starts to fade out and then goes darker so that I cannot read the text. I noticed the same thing with Avia. Is the 0% window supposed to do this?

I'm wondering if anyone else has noticed this or is it just my equipment.... when running the 10 grey windowed IRE test screens, 0% does something strange with Getgrey and Avia on my Panasonic 50PX60U plasma. The 0% box changes intensity after about 5 or 6 seconds. It starts out barely bright enough to read the text and then starts to fade out and then goes darker so that I cannot read the text. I noticed the same thing with Avia. Is the 0% window supposed to do this?

Are you sure you mean with the 0%? There shouldn't be any text or any dithering at all there....everything should be off.

If you're referring to the 5% one....I see the same thing on my 9UK. After about 5 seconds it will drop some brightness dithering.... Doesn't always happen, depends on my low-end brightness.

Turns out I had my brightness set too high.

Has anyone here done a Pio 1140 with HCFR? I might be doing one here soon :)


thx

bob

I've been following the threads on this software as well as the DIY sensor it's designed for. I've been thinking about buidling the sensor, based on parts list dlarsen put together, assuming I can find a way to program the pic.

I understand that to use the sensor and software though, you need to calibrate the sensor, either using another sensor, or using a 'reference file for your display.

I couldnt find a list of display that had a refernce file, so I downloaded the software. I found alits of files for different displays, but I see my display isn't listed. Does this mean I can't use this sensor to claibarte my display, Toshiba 65HX93?

Also, I thought I'd try to learn a little about softwer, but when I went to try the similuted sensor setting, it still seems to want to open a file, unfortunately, I can't seem to find a file to use.

Can anyone post some instructions instructions on how to use the simlated sensor in this software? And is there a place I can search for a reference file for theToshiba 65hx93?

I like the concept these folks have come up with of a DIY sensor and freeware for DIY'ers.

Thanks,
Rick

Has anyone had troubles calibrating an LCD TV (Panasonic 32" or Sony Bravia 32") using HCFR and Spyder2?

I have done my Panasonic plasma, and it looks great, but for some reason on LCD screen's it doesn't work. The grayscale ends up way too red. I am using the S2 in LCD mode with the filter on. Do I have afaulty sensor? How do I know if my sensor is faulty? Given that it works fineon my plasma, why would it be any different for LCD?

Hey Steffe, i too had issues with "reddish" grayscale while using an S2 meter. Not necessarily blaming the meter, but i went out & purchased a Display2 meter & my results were nothing more than astonishing! I personally believe the D2 is a far more accurate meter when paired with either CalMan or HCFR software.

Hope this helps.
Carmine.

Hey Steffe, i too had issues with "reddish" grayscale while using an S2 meter. Not necessarily blaming the meter, but i went out & purchased a Display2 meter & my results were nothing more than astonishing! I personally believe the D2 is a far more accurate meter when paired with either CalMan or HCFR software.

Hope this helps.
Carmine.

But why then are so many people using and talking about Spyder2?'

What type of display were you trying to calibrate?

I have a Sony LCD RPTV. Could not achieve accurate grayscale with my S2 so i purchased the D2.

Carmine.

I'm wondering if anyone else has noticed this or is it just my equipment.... when running the 10 grey windowed IRE test screens, 0% does something strange with Getgrey and Avia on my Panasonic 50PX60U plasma. The 0% box changes intensity after about 5 or 6 seconds. It starts out barely bright enough to read the text and then starts to fade out and then goes darker so that I cannot read the text. I noticed the same thing with Avia. Is the 0% window supposed to do this?

I noticed this as well with 0% window on my 9UK. I think I was using a sony Ns75H DVD player at the time. It's like the black "tries" to be really black by completely turning off the screen after a few seconds. Non of the other windows 10-100% do this.

What did you find regarding having your brightness too high? How was this related and did it go away now?

Thanks
Rich

I noticed this as well with 0% window on my 9UK. I think I was using a sony Ns75H DVD player at the time. It's like the black "tries" to be really black by completely turning off the screen after a few seconds. Non of the other windows 10-100% do this.

What did you find regarding having your brightness too high? How was this related and did it go away now?

Thanks
Rich

I'm glad I'm not the only one who's seeing this. It's making it VERY difficult to get a reading at 0% because the totally dark screen goes even darker, messing up the 0% reading. It causes the software to report incorrect gamma readings. It seems the Spyder2 cannot give accurate readings at 10% either so I am ignoring any reading I get below 20% and using the Contrast/Brightness screen on Getgrey that has the moving vertical black bars to properly set brightness. Oh and another thing.... I get more accurate readings in real time than if I do the 10 step grey-scale reading in HCFR. No idea why. Maybe I have a bad Sypder. I've got a headache right now from all this nonsense. I can't believe how hard it is to calibrate this 50PX60U plasma. Other people seem to have no problem doing it. I'm at the point where I'm ready to shell out the big bucks for an in home ISF. :eek:

I'm glad I'm not the only one who's seeing this. It's making it VERY difficult to get a reading at 0% because the totally dark screen goes even darker, messing up the 0% reading. It causes the software to report incorrect gamma readings. It seems the Spyder2 cannot give accurate readings at 10% either so I am ignoring any reading I get below 20% and using the Contrast/Brightness screen on Getgrey that has the moving vertical black bars to properly set brightness. Oh and another thing.... I get more accurate readings in real time than if I do the 10 step grey-scale reading in HCFR. No idea why. Maybe I have a bad Sypder. I've got a headache right now from all this nonsense. I can't believe how hard it is to calibrate this 50PX60U plasma. Other people seem to have no problem doing it. I'm at the point where I'm ready to shell out the big bucks for an in home ISF. :eek:

I just ignore 0 and 10% in my calibration. Then at the end, look at the darkest screen and see if there is a tint of a color (i.e., dim little red dancing dots) and back that color's offset down a hair.

PS. If you pay for ISF you miss out on the fun of doing this yourself any time you want. I kind of like being able to tweek over time.

I just ignore 0 and 10% in my calibration. Then at the end, look at the darkest screen and see if there is a tint of a color (i.e., dim little red dancing dots) and back that color's offset down a hair.

PS. If you pay for ISF you miss out on the fun of doing this yourself any time you want. I kind of like being able to tweek over time.

That's what I have been doing (ignoring 0 and 10%) but I've been unsuccessful at getting my gamma anything over 2.1 What is the secret? If I reduce brightness so I get a higher gamma reading, I crush my black levels. I've got about 30 man hours into measuring and tweaking, seeing what effect certain service menu settings do to the measurements. It's such a slow process but I think I'm getting there.

Regarding being able to adjust or check your own set anytime you want, it's great to be able to do this. However, there is a limit. According to this post from Datacolor http://www.avsforum.com/avs-vb/showthread.php?p=10330670&&#post10330670
The Spyders have a lifespan of only 2 years after which their readings cannot be trusted. This is very unfortunate.

BTW on the Panasonic line of plasmas, does anyone know what the service menu setting called WB BRT controls? No guesses please... I need someone who knows exactly what this control does to please explain it to me.

That's what I have been doing (ignoring 0 and 10%) but I've been unsuccessful at getting my gamma anything over 2.1 What is the secret? If I reduce brightness so I get a higher gamma reading, I crush my black levels. I've got about 30 man hours into measuring and tweaking, seeing what effect certain service menu settings do to the measurements. It's such a slow process but I think I'm getting there.
BTW on the Panasonic line of plamas, does anyone know what the service menu setting called WB BRT controls? No guesses please... I need someone who knows exactly what this control does to please explain it to me.

Which panny do you have? On the consumer models you can not adjust overall gamma, this is a look-up table in the video processor and is fixed at ~2.2. You can tweak it at the extremes with brightness and contrast. The WB BRT is a sub-brightness control, it gives you finer control over black levels than the user brightness, for example I use it to gradually increase gamma below 20% stimulus from 2.2 to about 2.4.

Wow Zoyd, how you get a gamma like 2.4 is beyond me. I have never been able to get above 2.1 without destroying black detail. I guess I have to live with the gamma I have. I have the TH-50PX60U. Zoyd, on your Panasonic when viewing the Get Grey DVD pattern with the vertical black bars that move horizontally. The ones that show % stimulus of -2 -1 0 +1 +2, are you supposed to see any of those bars or should they all be off? Another thing, should the screen be a dark grey at 0% black? My screen never seems to go totally dark.

There seems to be a lot of talk about calibrating Projectors and Plasma using HCFR, and a variety of sensors....BUT is HCFR designed/suited for calibrating LCD screens?

I was using HCFR and the Spyder2 to try and calibrate my Sony Bravia 32", but kept getting a very red tinge in my grayscale patterns, even though the HCFR software was showing RGB levels were spot on.

So, I went out and invested more money and bought a Display2 sensor, and although the results are slightly better than with the Spyder, they're still off. I still see far too much red in the grays.

What is it? What am I missing? I wonder if its the HCFR software that isn't setup properly to calibrate an LCD display?

Please give me some advice.

Someone said that CNET did a review on the bravia and found that the Green was way out and that they corected it a little by changing the colorspace to medium...but I wouldn't have thought that this would affect grayscale adjustments.

Anyone???

Wow Zoyd, how you get a gamma like 2.4 is beyond me. I have never been able to get above 2.1 without destroying black detail. I guess I have to live with the gamma I have. I have the TH-50PX60U. Zoyd, on your Panasonic when viewing the Get Grey DVD pattern with the vertical black bars that move horizontally. The ones that show % stimulus of -2 -1 0 +1 +2, are you supposed to see any of those bars or should they all be off? Another thing, should the screen be a dark grey at 0% black? My screen never seems to go totally dark.

My panel clips BTB so I don't see anything <0. In a completely dark room I can just see the 2% bar "turn on" so there are a few active pixels but not completely filled in. My luminance reading at 0% is typically ~0.1 cd/m^2. This may appear "gray" when it stands alone but you have to remember that our visual systems have an automatic brightness adjustment so that in very dim environments what appears "gray" will appear quite black when compared to normally lit scenes. For comparison, I have read that typical LCD black levels are ~0.3 cd/m^2 or a factor of 3 worse than good plasmas.

There seems to be a lot of talk about calibrating Projectors and Plasma using HCFR, and a variety of sensors....BUT is HCFR designed/suited for calibrating LCD screens?

I was using HCFR and the Spyder2 to try and calibrate my Sony Bravia 32", but kept getting a very red tinge in my grayscale patterns, even though the HCFR software was showing RGB levels were spot on.

So, I went out and invested more money and bought a Display2 sensor, and although the results are slightly better than with the Spyder, they're still off. I still see far too much red in the grays.

What is it? What am I missing? I wonder if its the HCFR software that isn't setup properly to calibrate an LCD display?

Please give me some advice.

Someone said that CNET did a review on the bravia and found that the Green was way out and that they corected it a little by changing the colorspace to medium...but I wouldn't have thought that this would affect grayscale adjustments.

Anyone???


can you post some of your cal files? The software shouldn't make any difference but the sensor can. The D2 with an LCD should be quite accurate. The software packages just read the sensor output and do relatively simple matrix math to generate the various outputs.

Sure.

Its just that I got a response from someone at Colorvision to my email complaining about inaccurate readings being obtained from the Spyder2 sensor. He responded by saying that the Spyder2express software isn't designed for use with TV calibration. I had a chuckle to myself, and responded by saying that I wasn't using their software, but was using another software package (HCFR) which I did not mention to him. He then got back to me and said that the algorithyms are different for computer monitors that they are for TV screens, and that COlorvision doesn't support the use of non-supported software with their sensors.

How do you have your probe set-up, your primaries and secondaries are really off. Also, what is the calibration chain, getgray-upscaler-panel? Is this the S2 or the D2?

How do you have your probe set-up, your primaries and secondaries are really off. Also, what is the calibration chain, getgray-upscaler-panel?

I am using a Pioneer 676A player connect via component video to my Sony Bravia KDL-32V2000 LCD TV. No upscaler. Yep I was using getgray 75% windows...

What do you mean by how do I have my probe setup?

I am using a Pioneer 676A player connect via component video to my Sony Bravia KDL-32V2000 LCD TV. No upscaler. Yep I was using getgray 75% windows...

What do you mean by how do I have my probe setup?

Is it up close to the screen or at a distance and is this the S2 or the D2?

Its the D2...

right up against the screen....

Here are the results from using the S2....even worse I reckon...

SO I would not see asny difference if I was to buy the Calman software? Just curious, as that was going to be my next step!

I too thought that the software was only representing in graphical terms what the sensor was reading. I didn't think there was any manipulation of the results obtained from the sensor.???

Do I need to use a filter in front of the D2? The Spyder2 has the removable filter, but the display2 is different. Is it built-in and can't be removed? How does it do CRT in that case?

Well, I'm not familiar with LCD's but I've never seen primaries that out of whack and while the S2 and D2 aren't the same they do show excursions in the same ballpark. As you can see you can still get a decent gray scale but I'm wondering if there isn't something related to the highly oversaturated green/cyan/blue that is prompting the over compensation in red that you notice. Do you know if there is any color management system on your panel that you can play with to try and tame that green? You could start by lowering the user "color" setting but that might not be sufficient.

I don't think there is too much I can do with teh color decoder. There doesn't seem to be any service menu options for it...unless I missing it?

CNET mentioned putting the colorspace setting wo wide to reduce the green error by 20%...but I tried that and saw no difference in the CIE chart.

But why would the color decoder affect grayscale adjustments anyway? Im confused...

SO I would not see asny difference if I was to buy the Calman software? Just curious, as that was going to be my next step!

I too thought that the software was only representing in graphical terms what the sensor was reading. I didn't think there was any manipulation of the results obtained from the sensor.???

Do I need to use a filter in front of the D2? The Spyder2 has the removable filter, but the display2 is different. Is it built-in and can't be removed? How does it do CRT in that case?

I haven't used Calman but it wouldn't be any different as far as end results. There might be differences in ease of use, navigation, etc. but you'd have to ask someone who has used it. They also provide more specific probe support with respect to their software on their forum.

You shouldn't need any type of extra filter with an LCD, plasma's are still under investigation. The D2 CRT mode just loads a different calibration table.

But why would the color decoder affect grayscale adjustments anyway? Im confused...

You're right, it shouldn't. But your colors are so out of whack that I suspect it may be influencing everything. I think you need to find someone who knows a bit more about your panel.

Wild guess time but the other thing that might help is to get a cheap upscaler and use the hdmi input if you have one instead of component.

Whats your thinking there?

Well, these panels all do their video processing digitally so eliminating the component A/D step is one less place to screw things up. It's a long shot though, I start with trying to find an expert on your panel.

thanks Zoyd...trick is knowing what the US equivalent is to my model...

I have a question for the HCFR team.

I have built 2 of the HCFR probes now, one I have modded with an optics light booster.

See here the lums concentrator with the probe on the floor below a 60w bulb.
http://i111.photobucket.com/albums/n128/SmokeyJoe_03/008.jpg
http://i111.photobucket.com/albums/n128/SmokeyJoe_03/009.jpg

I have been thinking about further mods to help the bottom end response and accuracy. My thought is, can I add more TAOS sensors in series to capture more signal as voltage? In theory more sensors should add voltage, but can the curcuit handle the extra load? And how would the curcuit handle long length wires attached to the normal sensor contact points with say 3 sensors on each sensor 0 and sensor 1?
The wire I was thinking of using is Cat5 or Cat6 cable as it is twisted pairs so shouldn't suffer interference. I would probably have a plug aswell for convenience.

I was thinking of making a boom about 1 metre wide with 6 sensors spead out evenly on thin extenders. The reason for the boom is to get uniform capture and cut down on probe shadowing.

Rregards
Smokey

Wow. A ‘lums concentrator’. That looks waay cool. What improvements over the baseline have you seen with this?

My thought is, can I add more TAOS sensors in series to capture more signal as voltage? In theory more sensors should add voltage, but can the curcuit handle the extra load?

No, I don’t believe you can. The TAOS TCS230 is effectively a light to frequency converter not a light to voltage. (They do have other configurations though) The output is a square wave whose frequency is directly proportional to the light intensity. Each sensor array has 64 photodiodes. 16 of the photodiodes have an integrated red filter, 16 have a blue filter and 16 have a green filter. The remaining 16 are non-filtered clear.

Each of the four banks of 16 photodiodes is wired in parallel so every capture is already integrated over the 16 discrete photodiodes. Only one color (or clear) can be selected and sampled at any given moment. To wire even more in parallel, I think you’d need to be cracking open the die and hacking on the nano scale. At least with the TCS230. Also, due to a photo diodes intrinsic dark current and thermal noise, I doubt there is much to gain in terms of near-dark signal sensitivity by paralleling (series?) more. Maybe Cryogenic cooling? :D Adding more arrays could integrate a measurement over a larger collection area if there’s an advantage to that but I believe you’d need to expand the PIC and address and sample each array individually and sequentially.

As far a getting more sensitivity, your experiment looks more interesting to me. It looks like a lens (type?) (FL?) attached to a (glass) tapered light-pipe? Advantages? Downside? Materials?

Thanks again. Cool experimentation.

Dave

perhaps mounting the TAOS chip on a peltier cooler would lower the dark current?

:confused:

perhaps mounting the TAOS chip on a peltier cooler would lower the dark current?

:confused:

It would. Dave, that brings up another question. The i1 sensors do an offset correction for dark current, does the HCFR probe have anything similar or do you try and measure it manually and correct after the fact?

The i1 sensors do an offset correction for dark current, does the HCFR probe have anything similar or do you try and measure it manually and correct after the fact?

Not that I’m aware of. I don’t. As the TOAS TCS230 is an integrated digital device (to its outside world) The is really no ‘dark current’ external to it. It will have been factored into the light to frequency conversion. The TAOS datasheet does specify dark frequency at between 2-12 Hz. Also, with regards to something like the lums concentrator above, the is a saturation illuminance as well. You can get too much light and saturate it.

perhaps mounting the TAOS chip on a peltier cooler would lower the dark current?

It sounds like Zoyd thinks it would. Measurable, reasonable or practical though? I’m skeptical. My comment about cryogenically cooling was kinda a joke. My experience with TEC’s is that it really isn’t practical to get more than about 50 deg C worth of cooling out of one. One side need to get hot for the other side to get cool and if you don’t have a REALLY efficient way of moving that heat then the hot side will get so hot the eutectic bonds of the TEC fail. I’m not sure 50-60 deg C would have a dramatic impact on the dark current. Frosting and condensation is also an issue when it gets down near freezing. Perhaps an experiment at below freezing temps or by saturating with cold spray could reveal if that has any measurable impact.

Dave

The TAOS datasheet does specify dark frequency at between 2-12 Hz.
Dave

How does that level compare with a lower end reading, say around 1 cd/m^2? btw, you can get roughly an e-folding reduction in thermally generated dark current for every 7 degrees in silicon detectors.

Are they independent of each other? If I adjust my primary/secondary controls does it throw off the gray scale?


thx

bob

Are they independent of each other? If I adjust my primary/secondary controls does it throw off the gray scale?

thx

bob

Moving primaries changes gray scale, secondaries do not.

My panel clips BTB so I don't see anything <0. In a completely dark room I can just see the 2% bar "turn on" so there are a few active pixels but not completely filled in. My luminance reading at 0% is typically ~0.1 cd/m^2. This may appear "gray" when it stands alone but you have to remember that our visual systems have an automatic brightness adjustment so that in very dim environments what appears "gray" will appear quite black when compared to normally lit scenes. For comparison, I have read that typical LCD black levels are ~0.3 cd/m^2 or a factor of 3 worse than good plasmas.

Thanks. This is how I have mine set too. 2% bar is barely on. But my gamma is still reading <2.1
If I decrease the PICTURE setting this will bring the sontrast ratio down but will this bring gamma up near 2.2? Or should I not get so hung up on my gamma level?

Moving primaries changes gray scale, secondaries do not.


Thanks,

So zero everything and adjust Primaries (and Secondaries) first, and then move on to gray tracking?


bob

Thanks. This is how I have mine set too. 2% bar is barely on. But my gamma is still reading <2.1
If I decrease the PICTURE setting this will bring the sontrast ratio down but will this bring gamma up near 2.2? Or should I not get so hung up on my gamma level?

I wouldn't obsess about it, like I said, for a panel with a fixed gamma look up table there isn't a whole lot you can do other than set your black (and white) levels where you like them.

Thanks,

So zero everything and adjust Primaries (and Secondaries) first, and then move on to gray tracking?


bob

sounds good. Don't forget brightness/contrast and gamma if you got it after gray scale. Lather, rinse, repeat until exhausted.

That's what I have been doing (ignoring 0 and 10%) but I've been unsuccessful at getting my gamma anything over 2.1 What is the secret? If I reduce brightness so I get a higher gamma reading, I crush my black levels. I've got about 30 man hours into measuring and tweaking, seeing what effect certain service menu settings do to the measurements. It's such a slow process but I think I'm getting there.

Regarding being able to adjust or check your own set anytime you want, it's great to be able to do this. However, there is a limit. According to this post from Datacolor http://www.avsforum.com/avs-vb/showthread.php?p=10330670&&#post10330670
The Spyders have a lifespan of only 2 years after which their readings cannot be trusted. This is very unfortunate.

BTW on the Panasonic line of plasmas, does anyone know what the service menu setting called WB BRT controls? No guesses please... I need someone who knows exactly what this control does to please explain it to me.


You *measure* gamma. You can not *change* gamma.

It is a property of your display. Specifically, it is the exponent in the (simplified) equation L=V^gamma that relates screen illumination to input voltage. When you run the grayscale calibration you are just observing the value of this exponent and reporting it in most cases.

To actually change gamma you need to change the contents of the look up table or formula used in your display's video processor. Some sets (like the panasonic 9UK for example) offer different gamma settings that the user can choose. The 9UK offers 2.0, 2.2 and 2.5. I think the consumer pannys might not have this option but I'm not sure.

Rich

You *measure* gamma. You can not *change* gamma.

It is a property of your display. Specifically, it is the exponent in the (simplified) equation L=V^gamma that relates screen illumination to input voltage. When you run the grayscale calibration you are just observing the value of this exponent and reporting it in most cases.

To actually change gamma you need to change the contents of the look up table or formula used in your display's video processor. Some sets (like the panasonic 9UK for example) offer different gamma settings that the user can choose. The 9UK offers 2.0, 2.2 and 2.5. I think the consumer pannys might not have this option but I'm not sure.

Rich


So, adjusting brightness/contrast won't change the gamma (luminance) chart?

thx

bob

I have a question for the HCFR team.

I have built 2 of the HCFR probes now, one I have modded with an optics light booster.

See here the lums concentrator with the probe on the floor below a 60w bulb.
http://i111.photobucket.com/albums/n128/SmokeyJoe_03/008.jpg
http://i111.photobucket.com/albums/n128/SmokeyJoe_03/009.jpg

I have been thinking about further mods to help the bottom end response and accuracy. My thought is, can I add more TAOS sensors in series to capture more signal as voltage? In theory more sensors should add voltage, but can the curcuit handle the extra load? And how would the curcuit handle long length wires attached to the normal sensor contact points with say 3 sensors on each sensor 0 and sensor 1?
The wire I was thinking of using is Cat5 or Cat6 cable as it is twisted pairs so shouldn't suffer interference. I would probably have a plug aswell for convenience.

I was thinking of making a boom about 1 metre wide with 6 sensors spead out evenly on thin extenders. The reason for the boom is to get uniform capture and cut down on probe shadowing.

Rregards
Smokey

Very, very cool construction. But as far as improving low-end accuracy, that depends if the present inaccuracy is caused primarily by the low level of light to the sensor, or by the large variance in light output from the display under test.

If the source of error is mostly display variance than concentrating the available light would be no better than averaging for a longer interval. However, if the primary source of error is that the sensors just can't read low light levels accurately (due to noise floor for example), then you've got a sure fire winner here.

Good luck!

Rich

So, adjusting brightness/contrast won't change the gamma (luminance) chart?

thx

bob

Sure, adjusting brightness/contrast may change the measured gamma of the curve, but at the expense of uncalibrating your display. You know how to adjust brightness. There is really only one way, BTB looks same as video Black. Adjusting contrast (aka gain or picture) I believe will add a linear shift to the luminance curve with minimal effect on its exponential relationship, but I have to look in to this more. I believe Carl will be able to address this. Make contrast too high and you will clip as you know.

The difficult part of this discussion is that different display technologies (and even different brands) have their controls doing different things.

The 9UK's Picture control simply drives the output of the display....it doesn't mess with your Video signal driving....that simply gets mapped (and maximized by setting your weakest video Drive=FC) to your drive circuitry, and your gamma preset will determine this driving behavior.

Find a suitable gamma and stick with it. Anything but S-curve. But the best will be dependant on your source, etc. Find the one that looks the flatest possible on the logarithmic mode. Your lighting conditions will play a roll to, as 2.0 might look quite bad in absolute dark, and 2.5 might look bad in sunshine, etc.

I explained that the best I could (and to the best of my knowledge)....I'm sure it could be worded better.

I wouldn't obsess about it, like I said, for a panel with a fixed gamma look up table there isn't a whole lot you can do other than set your black (and white) levels where you like them.

Ok then. Thanks for all your help. I will stop obsessing about gamma. :)

I will stop obsessing about gamma.While you may be limited in what de-gamma options your display has available, you may have options with your source. If one is using an HTPC, you probably have another exposed and highly adjustable LUT available to you. You can manipulate each of the 256 control points with 8-16 bit granularity. There is also an external low-end degamma booster for analog RGB (VGA) than many of the CRT guy are using and most external video processors and scalers have selectable degamma manipulations.


How does that level compare with a lower end reading, say around 1 cd/m^2 I don’t know. I haven’t put a scope to it nor have the source code for the PIC to do register peeks or ICD. The TAOS datasheet may have close to your answer published as it does list output frequency with respect to the spectrum wavelength and amplitude. The optical power is quantified in terms of uW/cm^2 and the response is quantified in terms of lux/Hz at different wavelengths. I’m unfamiliar with the term e-folding reduction. I’ll need to do some learnin to understand that better and if 5-7 of them would be significant or not. I’m not even sure they are Si, Probably but I suppose they could be GAAS or ??

Dave

I’m unfamiliar with the term e-folding reduction. I’ll need to do some learnin to understand that better and if 5-7 of them would be significant or not.
Dave

sorry for the terminology. e-folding is shorthand for a 1/e reduction in physical systems that exhibit exponential behavior. In other words a temperature drop of 7 degrees would reduce the dark current by ~1/2.7

In other words a temperature drop of 7 degrees would reduce the dark current by ~1/2.7

It seems there must be an assumed unit to the 1 / 2.718 # or other weighted variables at play. % reduction? Most of the photodiodes I’ve played with only have about 300-400 uA of full scale current and this is when operated with a relatively large reverse bias voltage. Much less near zero bias or small forward bias. One 1/e would be about 1K times greater that the full scale amount. Perhaps, I could spice model it and run parametric simulations at various temps if TAOS published the characteristic curves for it.

Dave

Dave, your geek-o-meter reading just went off the scale with that post. :)

I only have a nerdometer. Is there a conversion factor for converting from geek units to nerd units? :D

Dave

Yes, but it involves math with lots of squiggly notation.

Back sort of on topic, when I was toying with building a probe I looked at the TAOS datasheets and got to wondering if there was another way to do this using camera chips like what you would find in a cell phone that might be more spectrally accurate.

Unfortunately from what I could find they are pretty much all lens/filter based and as such subject to ageing effects, etc.

yes, it involves a coeffecient equal to the mm rim-width of your glasses.

It seems there must be an assumed unit to the 1 / 2.718 # or other weighted variables at play. % reduction? Most of the photodiodes I’ve played with only have about 300-400 uA of full scale current and this is when operated with a relatively large reverse bias voltage. Much less near zero bias or small forward bias. One 1/e would be about 1K times greater that the full scale amount. Perhaps, I could spice model it and run parametric simulations at various temps if TAOS published the characteristic curves for it.

Dave

Ahh, Spice models, now you're talking MY language! ;) Does this mean I'm a nerd or a geek?:)

-Steve

It seems there must be an assumed unit to the 1 / 2.718 # or other weighted variables at play. % reduction? Most of the photodiodes I’ve played with only have about 300-400 uA of full scale current and this is when operated with a relatively large reverse bias voltage. Much less near zero bias or small forward bias. One 1/e would be about 1K times greater that the full scale amount. Perhaps, I could spice model it and run parametric simulations at various temps if TAOS published the characteristic curves for it.

Dave
:eek: Much too complicated Dave. Dark current is just the result of thermally generated electrons in the bulk material. This phenomenon can be shown to have an exponential dependence on temperature. Reduce temperature by 7 degrees, reduce electrons by a factor of 1/2.7, simple. The 7 degrees # is typical for all silicon-type detectors, photodiodes, CCD and CMOS.

Ahh, Spice models, now you're talking MY language! ;) Does this mean I'm a nerd or a geek?:)

-Steve

oh god, i remember those.

and i think you're granted both titles if you consider it a first language.

Ahh, Spice models, now you're talking MY language! ;) Does this mean I'm a nerd or a geek?:)

-Steve

Spice models, hmm... you mean like these. (http://upload.wikimedia.org/wikipedia/en/a/ae/PowerOf5.jpg)

lol

Reduce temperature by 7 degrees, reduce electrons by a factor of 1/2.7

Ok. Then it’s like a percentage, No? So 50 deg would be about 2.5%? As I mentioned, using a TEC to get it 50 deg cooler could be very problematic and across the border from practical and reasonable IMO, Not the mention the ice and frost that you’ll be creating unless you can seal it tight to keep moisture out. For 2.5%? Seems a lot for a little but I’ll try the freeze spay when I get a chance. It may be hard to distinguish any near dark improvements from the much larger operational variance that a >50 deg change will cause however. Also, in order to effectively utilize a TEC, some creative (expensive?) mounting would need to be employed to manage and minimize the thermal interface resistance. As it’s packaged in a plastic SOP, the only thermal path would be through the lead frame and bonding wires. Pretty high thermal resistance.

yes, it involves a coeffecient equal to the mm rim-width of your glasses. Thank god for optical plastic! (Or Lasic, but most self-respecting geeks/nerds I know seem to resist Lasic and wear their glasses proudly like a pocket protector)

Spice models, hmm... you mean like these.
Ha! Good one. I was expecting something from the spice mines and giant worms ala Dune. Ahhh… de Spice… Yeaaahhh.

Dave

Spice models, hmm... you mean like these. (http://upload.wikimedia.org/wikipedia/en/a/ae/PowerOf5.jpg)

Ha-Ha!:D Afraid not. (What were their names again? Sassy, Sleezy, Winky, ...?)

For those who may be confused, SPICE is a computer program for modeling and simulating electrical circuits/components. At least 70% of my job is modeling components and circuits.

-Steve

I have nothing to add to this discussion, I am either proud not to know what they heck is going on or I just feel dumb as heck..something like half empty half full feeling..

Ok. Then it’s like a percentage, No? So 50 deg would be about 2.5%?

Dave

No, 50 deg. would be 7 e-foldings or (1/2.7)^7=0.096% (99.904% reduction) A small amount of cooling might be beneficial but it really depends on what the dark current level is compared to the minimum luminance that you want to measure. Do you get any sort of reading with the sensor blocked off?

Zoyd-

As you seem very knowledgeable about optics, what are your thoughts on Smokey Joes optical collection approach? I’ve worked with a few optics lens-heads over the years and most every time I’d approach them with some hair-brain idea I’d have, it usually would get foiled with them muttering on about entendue or such. Seems that law of thermodynamics is a tough law to violate no matter the discipline.

Do you get any sort of reading with the sensor blocked off?
Yea, just about like 0 IRE data from my CRT. (as reported from HCFR probe native data) It’s spitting out non-zero values but its whack and not signal IMO. It not like it hangs or stop outputting if you put you thumb over it.

No, 50 deg. would be 7 e-foldings or (1/2.7)^7=0.096% (99.904% reduction) Wow. 99.9% reduction for 50 deg and the reduction continues exponentially. Ok then. Thanks. I didn’t think the reduction would be expondential as well. I was thinking (1/e)*7

Dave

Zoyd-

As you seem very knowledgeable about optics, what are your thoughts on Smokey Joes optical collection approach? I’ve worked with a few optics lens-heads over the years and most every time I’d approach them with some hair-brain idea I’d have, it usually would get foiled with them muttering on about entendue or such. Seems that law of thermodynamics is a tough law to violate no matter the discipline.

Dave


Looks like he's got something like this. (ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/19880003488_1988003488.pdf) Works like a fiber optic guide does using total internal reflections. And yes, etendue must be preserved (at least in this universe) but there's no hanky panky going on with these concentrators. Looks like you can easily get a factor of 10 concentration.

Yeah it basically works like a funnel of light, since the angle of incedenance on the side walls is low the light will change direction rather than penetrating the wall.
Sure there is some losses, but they are lower than the gathering potential since the surface area goes from 45mm dia down too 8mm, which equals a surface area of 1590square mm down to 50square mm. A ratio of 31.64.

It is made of 45mm Diameter acrylic plastic. I lathed it down at an angle and polished the surface. I roughed up the narrow end so as to defuse the final light output.

You could do this with any probe.

It works like any optics device that tries to gather light, ie when you look to the stars and want to see a star with more intensity you need a bigger collector.

Now you could do this in several ways, one being a lens. I pulled one out of an old overhead projector. It worked, but had an issue of focusing. But then I thought like the lens of the projector which changes the direction of light I just need to change the light direction I don't need to focus, so hence the funnel.
The majority of light emission will bounce off glass at 30degrees and less. You can prove this too yourself quite easily by grabbing a torch at night and shine it at a window. Slowly change the angle you point the torch at the window surface, at about 30 degees you will see a large portion of the beam shine onto the adjacent wall.

Since the idea of banking up the sensors is a no go, I might refocus(pun) on my funnel. Stage 2 will be a 16:9 cone shape so the small end will cover both sensors on the board.

The whole idea came from my search for a better black point measurement and adjustment. Currently we use the calibration DVD's to set brightness, but I kinda feel that like calibration itself we can only resolve a closeness by feel and vision, setting brightness could have the same error as balancing RGB by eye.

The other reason I built this probe is for my work on DIY screens. If you check my sig you will see my work on DIY screens to this end.

SO I would not see any difference if I was to buy the Calman software?
From my personal experience with Calman and the S2, the only difference you'll notice is the gamma curve # since you're using "Display w/ Black Compensation". Calman goes with a calculation (more traditional I think) similar to "Camera Standard" in HCFR, which is what I always use if taking measures with HCFR.

Just my 2 cents ;) .

Hey Smokey Joe-

Thanks for the additional information. I can’t quite tell from your photos but is the center cone of the acrylic hogged out in the middle in a V shape like the drawing that Zoyd posted, or is it a solid cone? It seems like the drawings that Zoyd posted utilizes the walls to achieve TIR. Also, is that a lens on the top entrance pupil?

On your idea of banking up the sensors, I hope you didn’t let my comments dissuade you. Even though it didn’t seem feasible to me, I’d hope the HCFR guys comment as I know your question was posed to them.

I’ve followed your DIY screen work. I’ve been impressed with your methodology and your experiments and it seems you certainly don’t let what has been previously been ‘known’ to stop you from further investigations, experiments, and progress. If and when that law of thermodynamics is violated with free energy cold fusion, anti-gravity, zero friction, or free beer for everyone, it will probably be by someone thinking way outside the box and challenging what is a ‘known fact’ to the rest of the community. Hell, some people think Tesla had everything but the free beer part figured out decades ago. :)

Dave

PS- Nice lookin cat and photography.

Edit- I see the turned cone in a solid acrylic cylinder now. The acrylic is so clear where in contacts the face plate that I missed it. It looks like about a 3/8” or so thru-hole was drilled first?

Yeah I used a dyecast alloy case instead for my HCFR probe with the same pattern layout as the plastic varient. By eye the whole thing is deceiving, as so our perception on reality...a whole other issue.

What you see is a outer tube of plastic which only holds the cone shape too 90 degees off the sensor. The cone itself is solid, the collector end is flat and polished, the other small end is roughed up with sandpaer to defuse. The small end protrudes through the case and sits away from the sensor by roughly 5mm.

http://i111.photobucket.com/albums/n128/SmokeyJoe_03/008.jpg

There is a possiblity that there could be RGB imbalances cause by the plastic, but when you look at the average IRE0 and IRE10 figures and you tighten the result down by very large amounts any imbalance is totally blown away by the tightened measurement.

The photo doesn't lie does it. The camera is the independent witness to the enhancing ability of a simple optics collector that requires no focusing.

btw I read somewhere(can't remember where) that clear acrylic plastic distorts RGB balance less than glass.

There is a possiblity that there could be RGB imbalances cause by the plastic, but when you look at the average IRE0 and IRE10 figures and you tighten the result down by very large amounts any imbalance is totally blown away by the tightened measurement.If there is any imbalance, it seems it would be closely related to the material and could be quantified, characterized and largely calibrated out.

OK. NOW I finally get what your photo shows. I just noticed a shot glass sitting on our counter that has a pretty deep V shape to it, I wonder if I can get it to TIR? What’s the refractive index for gin? :)

Dave

Noob question: I have the Pantone Eye-One Display 2, I assume that this software supports it?

And being a noob, I would need to hook up my LCD TV to my computer? :D

Noob question: I have the Pantone Eye-One Display 2, I assume that this software supports it?

And being a noob, I would need to hook up my LCD TV to my computer? :D

It is supported. You just need to copy your EyeOne.dll file to the HCFR directory.

You don't need to hook your LCD TV to your computer if you have a calibration disk you can use with the software (Avia, DVE, GetGray, etc.)

Do the "Eye-One Monitor and Eye-One Pro" work with this? Asking on behalf of a friend in Spain

Do the "Eye-One Monitor and Eye-One Pro" work with this? Asking on behalf of a friend in Spain

i1pro is supported, what's an i1 monitor?

Moving primaries changes gray scale, secondaries do not.


Does moving primaries change secondaries or are they independent?


bob

i1pro is supported, what's an i1 monitor?
The i1 Monitor is a discontinued product. It included an "Eye One (not "Pro") Spectrophotometer", that looks like an Eye-One Pro, but can only read in an emissive mode.

It is supported. You just need to copy your EyeOne.dll file to the HCFR directory.

You don't need to hook your LCD TV to your computer if you have a calibration disk you can use with the software (Avia, DVE, GetGray, etc.)

Thanks, I'll let you guys know how my calibration goes. I'll probably hook up to my computer and use the software rather then the calibration disk to take advantage of my Display 2 sensor.

Anyone from HCFR want to comment about my questions in post 2348?

Hi,

I will try to answer you later on, but I don't think it'll be that easy...

--Patrice

Anyone from HCFR want to comment about my questions in post 2348?


Hello.

As dlarsen said, TAOS output is frequency but not voltage so it is not possible to stack several sensors.

One possibility to improve dark accuracy would be to increase measurement time so that measure is integrated on more output signal periods.
Current maximum measurement time is 10s for each component, which leads to 30 s for complete measurement.

If you're interested in testing others configurations, let me know the duration you're interested in ;)

Good thread for HCFR feature requests?

Figure it can't hurt to throw out a couple.

1. Editable tables everywhere, not just the grayscale cells.

2. Deviations from temperature goal computations....like the "+/- 156K from 20IRE-100IRE"

3. Maybe use 75% intensity window patterns for the "view images"....or maybe simply allow resizing of the 'test color' box.

4. In 'combined histogram for free measures', allow the ability to fine tune the zoom....so add a 'values' choice like the normal rgb histogram.

I had some others....I'll think of them later.

All those are available in upcomming v2.0 ;)

--Patrice

Ok thanks Ki, Laric.

I am not sure adding more time measure will improve the 0IRE or 10IRE response if that time doesn't record any data or data that has alot of variation error. Hence my reason for looking at alternatives like stacking or optics to improve the bottom end collection to raise the values above any noise.

Actually what I personally would like with version 2 is the deviation and/or max min values for the measures.

My optics device really does help at the other end, I plan to try some other improvements to this idea. ALthough I need to get an IR filter as I struck problems whilst trying to calibrate a projector with a heater operating in the room.

One question you didn't answer is my different query on having long lengths of wire like Cat5 between the sensor and the board. The idea being I can get the sensor closer to the screen surface with the lowest effect of shadows.

Smokey

All those are available in upcomming v2.0 ;)

--Patrice

beautiful. i can't wait.

thanks patrice.

My Samsung should arrive today and I have question on use ColorHCFR in conjunction with my Eye-One Display 2 probe.

My first question since its going to be hooked up to my PC, do I need to preset my video card resolution to match the LCD TV prior to hookup?

Do I basically follow the onscreen instructions using ColorHCFR once its starts? My only other experience is the software that came with the Display 2, which was step-by-step and screen to screen followthrough.

Any other tips or specials things to do?

Thanks for the help guys. :)

Ok thanks Ki, Laric.

I am not sure adding more time measure will improve the 0IRE or 10IRE response if that time doesn't record any data or data that has alot of variation error. Hence my reason for looking at alternatives like stacking or optics to improve the bottom end collection to raise the values above any noise.

Actually what I personally would like with version 2 is the deviation and/or max min values for the measures.

My optics device really does help at the other end, I plan to try some other improvements to this idea. ALthough I need to get an IR filter as I struck problems whilst trying to calibrate a projector with a heater operating in the room.

One question you didn't answer is my different query on having long lengths of wire like Cat5 between the sensor and the board. The idea being I can get the sensor closer to the screen surface with the lowest effect of shadows.

Smokey

As TAOS output is frequency square signal 0 - 5V, using wire between sensor and board should not be a problem.

As TAOS output is frequency square signal 0 - 5V, using wire between sensor and board should not be a problem.

Ki_, have you measured output frequency vs. temperature in the dark? I'm curious what the dark current level is on these devices or even if it is measurable.

In addition to the already existing white points, can we have a manual override on white points if we want that flexibility in tweaking with Ver2.0??

Oh i remember what i was going to add.....how about LCH derivations?

I'm working my way through this thread, but haven't read all 2400 posts. Forgive me if these are FAQs...

I have a CRT projector -- 3 tubes, 3 sources of light. The measured color temp can vary wildly depending on where the sensor (Spyder2 Express) is pointing; skew it towards the blue CRT and the temp goes up, towards red and the temp goes down. I also have an old Minolta TV2150 colorimeter, and with that one I always aimed the sensor for max luminance (ft-Lamberts), assuming that was centered on the projector. Questions:

* Does HCFR have a "light meter" or "colorimeter" mode that just dynamically measures and displays xyY? In addition to being interesting, this would help me aim my Spyder2.

* Is there a better way to get accurate color readings with a CRT projector?

* Can HCFR display English units? I'm used to ft-Lamberts, not cd/m^2...

* ...umm... there was at least one other question... :)

Thanks!
Gary

I'm also trying to workout the proper way to measure a CRT projector. I believe I have it it pointed almost perfectly straight-forward and the readings appear to be close to correct. If/when I get the projector dialed in close to HCFR references and it still looks off then I might have to try and reading off the screen instead for my measurements. I would guess this might be an issue because of the Spyder's issue with low light output sources.

I'm working my way through this thread, but haven't read all 2400 posts. Forgive me if these are FAQs...

I have a CRT projector -- 3 tubes, 3 sources of light. The measured color temp can vary wildly depending on where the sensor (Spyder2 Express) is pointing; skew it towards the blue CRT and the temp goes up, towards red and the temp goes down. I also have an old Minolta TV2150 colorimeter, and with that one I always aimed the sensor for max luminance (ft-Lamberts), assuming that was centered on the projector. Questions:

* Does HCFR have a "light meter" or "colorimeter" mode that just dynamically measures and displays xyY? In addition to being interesting, this would help me aim my Spyder2.

* Is there a better way to get accurate color readings with a CRT projector?

* Can HCFR display English units? I'm used to ft-Lamberts, not cd/m^2...

* ...umm... there was at least one other question... :)

Thanks!
Gary


hey gary,

to do real-time measurements click the green play button in the measures window and then click the upper left icon "display measures information button"

you'll get the most accurate colortemp the further you are from the PJ and best would be to measure off the screen as close to screen as possible for max signal. Or take a CT reading off the screen and then turn it around and adjust pointing to get the same CT.

SI units in this version, maybe next will include english.

Visually, what's the difference between calibrating to REC 709 vs 601? I understand white will still be D65 but in real world viewing what differences would you see?

I've only used 601 because I watch SD and HD and was told to.

thx

bob

Visually, what's the difference between calibrating to REC 709 vs 601? I understand white will still be D65 but in real world viewing what differences would you see?

I've only used 601 because I watch SD and HD and was told to.

thx

bob

hey bob,

As you point out it won't affect your gray scale calibration. Also, if you can't adjust your primaries it won't make any difference which you use, you'll just get the wrong CIE diagram displayed when you look at your primary locations. IF you can adjust your primaries and you adjust them to the wrong colorspace, your colors will be wrong. The primary noticeable difference between r601 and r709 is green. Again, if you aren't adjusting your primary locations, which is true for most displays, it won't matter.

My TV has a menu called "color management" that allows phase and gain adjustments of the primaries and secondaries. I presume that I can use the phase to correct the hue, but since gain isn't saturation I don't think that's quite so useful (although I can use the primary gains to match RGB with the intensity of the corresponding components of white, it's not useful for color gamut correction).

My question is this - should I leave the phase alone, or try to line up the primaries with rec 601/709 even though the saturation is wrong? Would I try to move them onto a line extending straight out from D65 and passing through the actual rec color point? HCFR kind of draws those lines in, although for some reason the one that goes toward red doesn't pass through the rec color point (bug?).

My TV has a menu called "color management" that allows phase and gain adjustments of the primaries and secondaries. I presume that I can use the phase to correct the hue, but since gain isn't saturation I don't think that's quite so useful (although I can use the primary gains to match RGB with the intensity of the corresponding components of white, it's not useful for color gamut correction).

My question is this - should I leave the phase alone, or try to line up the primaries with rec 601/709 even though the saturation is wrong? Would I try to move them onto a line extending straight out from D65 and passing through the actual rec color point? HCFR kind of draws those lines in, although for some reason the one that goes toward red doesn't pass through the rec color point (bug?).

I think the best approach in doing color management is to minimize overall deltaE on the u'v' diagram. Other people have pointed out that in this process minimizing the secondaries is more important than the primaries since most natural image content lies in the secondary gamut. btw concerning your statement that "I can use the primary gains to match RGB with the intensity of the corresponding components of white, it's not useful for color gamut correction" This technique gets the correct saturation for your current primary locations in the expanded gamut, if you can use the separate RGB CM gain controls to match the r709 gamut you should do so.

This technique gets the correct saturation for your current primary locations in the expanded gamut, if you can use the separate RGB CM gain controls to match the r709 gamut you should do so.

If I change gain on the separate CM menu, then my saturations no longer match gray scale, so I judge them useless for adjusting gamut....

Wouldn't I want the secondaries to be on the lines of the triangle formed by my actual primaries, not on the lines formed by the rec 709 primaries?

Wouldn't I want the secondaries to be on the lines of the triangle formed by my actual primaries, not on the lines formed by the rec 709 primaries?

This should happen naturally. I think I clearly stated the goal in doing CMS adjustments in my previous post, minimize color differences between your measurements and reference, be it r709 or r601. One recommended trade-off in making these decisions is to weight the deltaE in secondaries more heavily than the deltaE in primaries.

If you are able to change the locations of your primaries then you should align them with the diagram, after that you will have to go back and redo the gray scale.

Ki_, have you measured output frequency vs. temperature in the dark? I'm curious what the dark current level is on these devices or even if it is measurable.

No I haven't done such measurement.

What is the difference? (Besides $70) :)

thx

bob

The S/W that comes with is different. They are the same H/W.

Puh! Finally reached the end of this monster thread! I think I`ll try to build the HCFR-probe. I could only find the building instructions in French?

I will use this to calibrate my Barco 1209S CRT-projector. I already have an old Minolta CL-100 colorimeter that give consistent readings down to about 30IRE. It has not been calibrated for many years, but I have compared it to a newer Minolta at high light conditions. So at least it seems that I can use it at 100IRE to calibrate the HCFR-probe.

Gunnar

I've got:

1. A DLP projector
2. Eye-one LT
3. GetGray DVD
4. A DVD player
5. Laptop
6. A good attitude
7. Lots of patience

From what I understand, I should not use the ambient light filter with the i1 colorimeter and HCFR. Also, I should point the i1 at the screen, at an angle (up towards the center of the screen - making sure I am not pointed towards the shadow from the i1), and very close.

I have reviewed post #708 (of this thread) and the geekwithfamily website for a step by step.

Is all of the above correct?

-thanks

This should happen naturally. I think I clearly stated the goal in doing CMS adjustments in my previous post, minimize color differences between your measurements and reference, be it r709 or r601. One recommended trade-off in making these decisions is to weight the deltaE in secondaries more heavily than the deltaE in primaries.

If you are able to change the locations of your primaries then you should align them with the diagram, after that you will have to go back and redo the gray scale.

I always hate bringing this up, because we just don't seem to be speaking the same language.

My TV has controls that they call "color management". For each of RGB and CMY I have hue and gain controls. With the hue control for green, for example, I can make green more blue or more yellow. The gain control makes green dimmer and brighter.

You state that one criteria is that the intensity of each of my primaries on a color window should equal the intensity of that same primary in a gray window. Ok. I can do that with my color and tint controls. Once that's set, I cannot use the gain settings in my color management controls because that's what they change - they don't change saturation, they change intensity.

So my measured color gamut is outside both the rec. 601 and 709 triangles and there's nothing I can do about that.

So, given that I'm stuck outside the correct color gamut you say you "clearly stated the goal in doing CMS adjustments in my previous post, minimize color differences between your measurements and reference, be it r709 or r601." Well, maybe you did, but I don't have the faintest clue what your definition of "minimize" is.

Is a minimized difference the point of closest approach even though it's off to the side and therefore more yellow than green ought to be, or off to the other side and more blue than green ought to be? Or is it on a line out from D65 where the tint of green has the same proportions of blue and green as the correct color gamut green point? Personally, I'd think it's the latter definition. If that is the correct definition of "minimize", then where is that point on the CIE diagram? I hypothesized that it's on a straight line passing through D65 and the correct gamut green point, but I don't know that that's true. Just as white is a curve, greens with the same ratio of blue/yellow but different saturations probably lie along a curve. But if they do, then how do I visualize that curve?

You refer to deltaE, but where's deltaE for colors in the CIE diagram? All I've seen in HCFR is the deltaE ellipse it draws around my gray points. What am I missing?

And why would changing my primaries change my gray scale? It's color management and at least on my TV is doesn't appear to change the gray scale. After all, it's not actually changing the color of the three color engines, it's just changing the way "green" is decoded in the color portion of the signal to add some yellow or some blue into the decoded value. At least, that's what it seems to be doing.

By minimize I mean minimize deltaE on the u'v' diagram. If you mouse over the primaries/secondaries measurement it will give you the deltaE value. This is a measure of the perceptual error that a particular color has with respect to the reference. However your particular set implements color controls the goal is the same. My comment about gray scale changes pertains to sets that actually change the color of the primaries by mixing in other colors, not many plasmas do this, the only ones I know of are the pioneer elites.

OK, obviously I do not know what I'm doing yet. (No surprise. :))

I tried reading off the screen, as suggested. The values I got made no sense. I was looking at the graphic display of RGB levels while I took real-time measurements. With the grayscale settings I'd set with my Minolta TV2150, HCFR/Spyder2 claimed my color temp was about 2500K !? I tried adjusting RGB to make it happy. I thought the RGB levels would be very useful for this -- just adjust them to the target. But while I could get the blue to the target line, I had green maxed and red at 0 and neither one reached the target.

So then I tried doing readings with both the Spyder2/HCFR and my Minolta. First, just the primaries: (I did this by turning on only the appropriate gun on my CRT and pointing the sensor at the projector.)

R: .665/.331 (Spyder), .668/.330 (Minolta)
G: .331/.580 (S), .310/.584 (M)
B: .150/.058 (S), .138/.056 (M)

Not perfect, especially the x in the Blue and Green, but overall not terrible. So they do seem to be reading roughly the same color coordinates.

Then I tried reading off the screen and pointing at the projector, while displaying a 100 IRE window:

Off screen: .420/.307 (S), .318/.298 (M)
At projector: .315/.350 (S), .294/.394 (M)

(The at-projector Minolta readings weren't D6500 because of the previous tweaking. But this experiment seems to indicate my screen is affecting my color a lot -- even with the Minolta there is a significant change between at-proj and at-screen.)

The Spyder and Minolta were not even close. I could pass off the "at the projector" differences by saying they're not aimed at the same point -- even though I aimed both of them for max luminance. But reading off the screen is even worse, and that's supposed to be more reliable.

Why am I getting totally different readings, even off the screen, when the primaries measured fairly close with the two colorimeters!?

I think the Minolta is reasonably accurate, especially for these high lum levels. What am I doing wrong? Do I need to calibrate the Spyder? (But the primaries were OK!?) How can I calibrate it without a calibrated source of white/black/R/G/B ? Maybe I can use the Minolta as a reference, but how would someone do it if he didn't have another colorimeter??

Finally: the real-time "Measurements" window takes continuous measurements, but very slowly. It's hard to use it to tweak the grayscale when it only takes one sample every 3-4 seconds, even at IRE100. I'm sure it would be much slower at IRE 30. I tried turning off the "long exposure for low IRE" and "average multiple readings" options but that didn't speed it up at all, even for IRE100. Is it possible to get a faster response? My Minolta takes a reading about every .7 seconds, which makes it a lot easier to adjust the settings.

Thanks,
Gary

By minimize I mean minimize deltaE on the u'v' diagram. If you mouse over the primaries/secondaries measurement it will give you the deltaE value.

Cool! Mouse-over.... Whoda thunk it. Do you have a quick layman's explanation of what the u'v' CIE diagram is trying to tell me versus the one I normally see?
Roy

Cool! Mouse-over.... Whoda thunk it. Do you have a quick layman's explanation of what the u'v' CIE diagram is trying to tell me versus the one I normally see?
Roy

the u'v' diagram is designed to better represent our perception of color differences, a change in the u' coordinate in one part of the graph should be equally noticeable as a change in u' in another part of the graph. You'll notice that reds take up a lot more area than greens in this diagram compared to the normal xy diagram, that's because we are much more sensitive to changes in red than green. This can also be seen if you have a green primary outside the gamut in the normal diagram it will move closer to the reference in the u'v' diagram. If you were able to perfectly align each color to reference it wouldn't matter which one you use but when there are errors, the u'v' diagram gives a better representation of the real-world effect of the error, how we perceive it.

hey gary,

A couple of quick questions first, is the spyder in lcd or crt mode, is the filter/baffle on or off?

When pointing at the screen do you have it as close to the screen as possible and pointed to avoid shadow?

If you are using crt mode, baffle off and pointing at the screen you will probably have to construct a light baffle. A simple way to do this is just spray paint the cardboard tube from a roll of toilet paper with flat black paint and tape it onto the sensor.


I tried turning off the "long exposure for low IRE" and "average multiple readings" options but that didn't speed it up at all, even for IRE100. Is it possible to get a faster response? My Minolta takes a reading about every .7 seconds, which makes it a lot easier to adjust the settings.



This is a limitation of the sensor, because the S2 uses a diffuser (the white disk you see under the filter) it sees less light than the typical reflectance meter and needs longer integration times to get a decent reading.

CRT mode, filter off.
I have the sensor about 6-7" from the screen, tilted up about 20-30deg to avoid shadows. (But I have no idea how wide the Spyder's view-cone is. The readings don't seem to change much if I tilt it up or down a little, so I think it's not seeing the shadow.)

I don't understand the purpose of the light baffle. It's supposed to limit the sensor's view cone? Why -- to avoid seeing the shadow?

Re the diffuser &etc: yes, I can understand the sensor being slow. But regardless of the sensor, I thought telling HCFR to "take long shots, integrate many readings" would take longer than "don't take long shots, don't integrate many readings." Why doesn't it? (Actually I think the "take long shots" option only applies to low-IRE readings, and I was reading 100 IRE, so maybe it didn't affect these readings. But the "integrate many readings" part should still apply?)

Gary


EDIT: Hm. I must have done something wrong last night. I just did a quick re-measure with the Minolta and with HCFR. The readings are still different between the two -- .321/.352 for Minolta, .407/.353 for Spyder -- but at least now I'm getting nearly identical readings off the screen and off the projector.

So I guess I still need to understand how to calibrate HCFR?

In the Eye one parameters menu it has an option under 'Black Level' to 'calibrate internal sensor offsets' Do I need to do that??

Also, there are only Eye one-LCD and CRT. I have a DLP. I assume I should choose LCD. Is that right?

...............and...........what does it mean to 'put the eye one on a planar surface' ...just a flat surface????

Please help

...............and...........what does it mean to 'put the eye one on a planar surface' ...just a flat surface????

Please help


yes, as to not allow any light into it.

I think I figured it out. I just did what it said to do. I put it on a flat surface, and from what I understand, the felt around the outside of the sensor closes out all the light.

Thanks primetimeguy

Finally: the real-time "Measurements" window takes continuous measurements, but very slowly. It's hard to use it to tweak the grayscale when it only takes one sample every 3-4 seconds, even at IRE100. I'm sure it would be much slower at IRE 30. I tried turning off the "long exposure for low IRE" and "average multiple readings" options but that didn't speed it up at all, even for IRE100. Is it possible to get a faster response? My Minolta takes a reading about every .7 seconds, which makes it a lot easier to adjust the settings.

In addition to the S2 being inherently slow, HCFR uses a lot of CPU resources when running continous measures - there's a lot going on in this mode. I don't know what CPU you are using, but my PIII 1Ghz is almost unbearingly slower than my P4 2.4 Ghz notebook when running continous measures.

As for the measurement differences, when was the last time the Minolta was calibrated?

-JD

CRT mode, filter off.
I have the sensor about 6-7" from the screen, tilted up about 20-30deg to avoid shadows. (But I have no idea how wide the Spyder's view-cone is. The readings don't seem to change much if I tilt it up or down a little, so I think it's not seeing the shadow.)

I don't understand the purpose of the light baffle. It's supposed to limit the sensor's view cone? Why -- to avoid seeing the shadow?


The S2 view-cone is the entire hemisphere, 2pi steradians. It is designed to be a contact probe for emissive displays and placed directly on the screen so it will view a small area. By adding a light baffle in your set-up you can limit it's field-of-view to a smaller area of the screen. Also try LCD mode with the filter/baffle installed to see if it agrees better with the minolta.


Re the diffuser &etc: yes, I can understand the sensor being slow. But regardless of the sensor, I thought telling HCFR to "take long shots, integrate many readings" would take longer than "don't take long shots, don't integrate many readings." Why doesn't it? (Actually I think the "take long shots" option only applies to low-IRE readings, and I was reading 100 IRE, so maybe it didn't affect these readings. But the "integrate many readings" part should still apply?)


yes, the check boxes apply to low IRE readings only. If you want to increase the integration time at high IRE, change the read time from 300 ms to something like 600 ms.

So I guess I still need to understand how to calibrate HCFR?


You do not calibrate the software, it's just processing what the probe tells it.

the u'v' diagram is designed to better represent our perception of color differences, a change in the u' coordinate in one part of the graph should be equally noticeable as a change in u' in another part of the graph. You'll notice that reds take up a lot more area than greens in this diagram compared to the normal xy diagram, that's because we are much more sensitive to changes in red than green..Thanks for that concise explanation, I never did understand the u'v' diagram, but now I get it!

zoyd and others

Does anyone besides jdbimmer have any experience with this probe? Also known as DPT94?

http://www.xrite.com/documents/literature/en/L11-143_%20DTP94_en.pdf



thx

bob

zoyd and others

Does anyone besides jdbimmer have any experience with this probe? Also known as DPT94?

http://www.xrite.com/documents/literature/en/L11-143_%20DTP94_en.pdf



thx

bob

YES..I have that also..

YES..I have that also..


So, how do you like it with HCFR? Do you others to compare to?


thx

bob

So, how do you like it with HCFR? Do you others to compare to?


thx

bob
I think Rich has a Mits RP?? TV, not a plasma.

bob - technically your D2 is the replacement to the DTP-94

Thanks for that concise explanation, I never did understand the u'v' diagram, but now I get it!

i love the u'v'. it's interesting that the the standard mode is how the 'eye ball sees it'....but the u'v' is how the 'brain sees it'. which is why the u'v' is good to use for adjustments.

kraz actually recommends using LCH to make these decisions....it's a derivation beyond u'v'. There's an awesome excel sheet at the end of the main calibration sticky.

i'm hoping the hcfr crew would auto-crunch some of these numbers for us...i'm lazy :)

bob - technically your D2 is the replacement to the DTP-94
Unfortunately, I think the DTP94 got axed after all the Xrite-MonacoSys-GMB acquisitions...competing products offered by the same company doesn't fit the marketing/cost savings targets of a merger.

-JD

I think Rich has a Mits RP?? TV, not a plasma.


correct CRT-rearPJ and DLP (fp) -both mits

zoyd and others

Does anyone besides jdbimmer have any experience with this probe? Also known as DPT94?

http://www.xrite.com/documents/literature/en/L11-143_%20DTP94_en.pdf



thx

bob

I used it on my RP CRT and it works great.

Cool! Mouse-over.... Whoda thunk it.
Roy

Darn! Mouse-over only shows deltaE for the primary and secondary measurements. When you do continuous measurement the current point shows up on the CIE diagram but mouse-over only shows its coordinates.... I guess that makes sense 'cause they'd have to guess which point to compute deltaE against, but it sure is inconvenient....
Roy

Darn! Mouse-over only shows deltaE for the primary and secondary measurements. When you do continuous measurement the current point shows up on the CIE diagram but mouse-over only shows its coordinates.... I guess that makes sense 'cause they'd have to guess which point to compute deltaE against, but it sure is inconvenient....
Roy

secondaries report deltaE over here.

bob - technically your D2 is the replacement to the DTP-94

colin,

Made by the same company? Which would you get for $130?

thx

bob

colin,

Made by the same company? Which would you get for $130?

thx

bob

Not really...i believe some aquisitions put the DTP-94 in the same product line at GMB. They've discontinued it because they've deemed the D2 the one they're going to keep. I own neither. If someone owns both perhaps they can report any likes/dislikes of either. The HCFR crew seems to think both are quite good according to the first page of this thread. Again, none of the 'cheap' sensors are perfect for plasma....it seems they haven't designed the cheapos to handle the conditions typical of a plasma. I'd suggest using zoyd's offset matrix to compensate...it should get you as close as you can until they make a cheap sensor better suited.

secondaries report deltaE over here.

Yeah, primaries and secondaries. But start a continuous measurement and bring up the CIE diagram. You'll see a yellow dot that is the current "continuous measurement" point. It moves in real time as you adjust the color settings of the TV. If it showed deltaE you could easily find the minimum.

Instead you have to take a primary/secondary measurement, then look at the CIE diagram, then lather rinse and repeat....

Yeah, primaries and secondaries. But start a continuous measurement and bring up the CIE diagram. You'll see a yellow dot that is the current "continuous measurement" point. It moves in real time as you adjust the color settings of the TV. If it showed deltaE you could easily find the minimum.

Instead you have to take a primary/secondary measurement, then look at the CIE diagram, then lather rinse and repeat....

ah ok, didn't realize you were talking about the real-time measure. You'd have to add a way to specify what your targer color is for that to work, or maybe have it start reporting dE when it was in the vicinity of one of the colors.

In addition to the S2 being inherently slow, HCFR uses a lot of CPU resources when running continous measures - there's a lot going on in this mode. I don't know what CPU you are using, but my PIII 1Ghz is almost unbearingly slower than my P4 2.4 Ghz notebook when running continous measures.My laptop says it's a "1.8 GHz Mobile, 1.2 GHz." So apparently it's not a lot faster than your PIII.

As for the measurement differences, when was the last time the Minolta was calibrated?Unknown. :o Not since I've owned it. But given what the Minolta tells me and what the Spyder/HCFR tells me, I'm much more inclined to believe the Minolta. The S2 numbers were just crazy.

The S2 view-cone is the entire hemisphere, 2pi steradians. Aha. So the baffle limits its effective view to a known area, avoiding the shadow. Of course, it will further reduce the light coming in, making it even slower, *sigh*...

Also try LCD mode with the filter/baffle installed to see if it agrees better with the minolta. Hm. OK...

You do not calibrate the software, it's just processing what the probe tells it.Then what are the "Conversion Matrix" and "Calibration" tabs for? (Sensor -> Configure) Aren't those calibrating the software to work properly with what the probe tells it?

Then what are the "Conversion Matrix" and "Calibration" tabs for? (Sensor -> Configure) Aren't those calibrating the software to work properly with what the probe tells it?

Those are for training one probe against another, so for instance you could measure with a reference probe like the i1pro and then with an unknown or inaccurate probe and generate a new sensor matrix for the unknown probe. Again, the software just reads the X,Y,Z values that the sensor sends, it doesn't care if those values are accurate or not.

OK. So if I believe the Minolta is accurate, I could use it as a reference and adjust the conversion matrix to get the Spyder to produce the same readings. Looking at the help it appears that you can do this with two supported sensors, or with one supported sensor (Spyder2) and a reference (Minolta). I'll try that.

The calibration process wants R/B/G primaries and 0/100 IRE. Would it be OK to enable only the R/G/B tube desired (e.g. R tube for R primary) ? Or is it important to provide a Red signal to all 3 tubes? In theory only the Red should light up, but...

Gary

The calibration process wants R/B/G primaries and 0/100 IRE. Would it be OK to enable only the R/G/B tube desired (e.g. R tube for R primary) ? Or is it important to provide a Red signal to all 3 tubes? In theory only the Red should light up, but...

Gary

yes you can create the RGB readings for cross-calibration that way. HCFR will automatically calculate the new sensor matrix for two supported probes (using advanced->Adjust XYZ coordinates). I'm not sure how you are going to do it with the minolta though because you can't manually enter the minolta's reading into the program, you will have to wait for version 2 for that. (coming soon)

My laptop says it's a "1.8 GHz Mobile, 1.2 GHz." So apparently it's not a lot faster than your PIII. That's the Intel Pentium M 1.8 GHz CPU, which is a low voltage chip designed for notebooks that performs like a P4 2.x + GHz. It has more than enough processing power for the task, so any speed issues you see are occuring at the colorimeter.

nate-

Continuous measures is a nonstop reading of whatever IRE is displayed on screen. The way to use it is for example: put up 20IRE or 20% if using GetGrey on your screen. On combined histogram window, look at RGB Levels, if out of calibration you should see red, green and blue lines. Adjust your R-Bright, G-Bright and B-Bright until they converge onto a single line. Keep an eye on the color temp, you want it at 6500K, which you can see at the bottom of the same window. Then put up 80 or 90IRE/% and repeat, this time adjusting RGB-Contrast. Once done hit Stop.

You cannot adjust Primary/Secondary colors with this method. Are all those menus per input?

Wont the sensor be hot, so that the readings will be off using this continious reading?

Wont the sensor be hot, so that the readings will be off using this continious reading?

The sensor is always running continuously, as soon as you plug it into the usb port. Given that you should always allow your probe to stabilize before taking critical measurements. A good rule of thumb is to let it warm up for 1/2 hour, especially the S2 since it's response varies with temperature more than the other probes.

Have the actual S2 on the screen with both the S2 and TV on for that 1/2 hour. just clarifying

In all honesty, I've done some measuring without 'warming up'...I haven't seen anything too wild. But I do it because I'm told :)

Have the actual S2 on the screen with both the S2 and TV on for that 1/2 hour. just clarifying


yes, but it's just a general guide line. Some time try a gray scale with no warm-up time and compare that to one after 20-30 minutes and see if there is any difference.

So for the upcoming v2.0 of HFCR, witch sensor should i go for? I've read to page 30 and then jumped to p.75, and suddenly the so called d2 came up :).

For calibrating both plasma panels, pj (dlp/lcd), and lcd panels, is the dtp-94/monaco xr the way to go, instead of the Display 2 LT? I really dun like the "lockup" thing mentioned with the D2. This scares me away from it. Maybe this isn't a big issue, really?

How much better is the i1 pro over these other sensors? Is it worth the extra money?

About the faulty measurements with a warm sensor. Over at calman they mention a recalibrate after 10min of use with the dtp-94. The D2 doesn't need recalibration.

As you might remember, I'm in a "round robin" sharing of a D2 probe with other Panny 9UK owners. I had the probe for 2 weeks and ran many cals on my Panny 9UK. I was one of the few that had the high red clipping issue and could not get above 94% at 80 IRE (stopped going up at 14 or so).

I still haven't been happy with the color and have been waiting to see how Pdawg, who did not have the high red problem, but discovered that after many UM and SM adjustments that the Getgray 5% and gradient ramps had a definite reddish and slightly greenish tinge throughout the range. I too have noticed this on my HDMI and component inputs and today decided to eyeball the gray ramps to eliminate the red tinge. To do so, my high red went from +24 to -1 and the low red went from +10 to -3.

When switching to real world viewing (DVD via component and HDDTV via HDMI), and keeping the calibrated settings in standard mode with the eyeball settings in cinema mode, I could toggle between the 2 and lo and behold grays look grey not reddish gray; browns look brown not reddish brown, beige looks..... you get the idea.

I have been watching different DVD'd and DTV programming all afternoon and finally I am seeing what a (closer) good grey scale should look like!

Pdawg has the probe now so I can't give you any measurements but my eyes say that this is better than it has been.

So I don't know if the probe has gone out of cal. or it's user error, but for now I'm finally seeing decent color.

Might just have to go for ISF :)


bob





now has the probe) would do in the SM

From what I understand, I should not use the ambient light filter with the i1 colorimeter and HCFR. Also, I should point the i1 at the screen, at an angle (up towards the center of the screen - making sure I am not pointed towards the shadow from the i1), and very close.


Apologies if I missed this earlier. When using an Eye-one Display2 or LT, should one use the diffuser pointed towards the pj or no diffuser and pointed towards the screen? Or can one do either depending on whether they want to take the screen into account?

TIA

No diffuser, reading off the screen is how it is normally used.

yes, but it's just a general guide line. Some time try a gray scale with no warm-up time and compare that to one after 20-30 minutes and see if there is any difference.I did some experiments last night and noticed considerable changes from one read to another -- not so much with the Spyder2, which was mounted on a tripod, but with my Minolta probe, which I was hand-holding. I saw changes as much as +/-0.010 in both x and y. That can easily make a 500K difference in the measured color temp. I don't THINK this was due to warmup changes, since the Spyder2 didn't seem to change. I think it was due to positioning the probe differently each time. So even though I was reading off the screen, I was seeing a lot of sensitivity to probe positioning. Which makes me wonder how you can measure the projector and reliably get repeatable results.

I'll have to experiment with it some more to make sure. It may have actually been warmup changes. I'd get measurement X with the Minolta, and it would be reasonably stable even if I moved the probe, and then 5 minutes later I'd get measurement Y. But it was still doing this after the projector & colorimeters had been on for an hour...

And I may have to give up on the Spyder2. This thing is SLOOOWWW. So slow that I don't see how you could use it to make adjustments to zero in on 6500K: tweak colors, wait 2 minutes for a read, tweak, wait 2 minutes, tweak... I checked the CPU usage while HCFR was doing continuous measurements, and ColorHCFR.exe was running between 0% and 1% CPU usage. So it's not my laptop that's making the thing take 20-120 seconds per reading. The HCFR probe is a LOT faster, right?

bob - changing the grayscale by eye isn't going to do much...by turning down red so much, you're probably pushing toward blue or blue/green quite hard. Blue is less offensive-looking on the grayscale, which is why you don't 'appear' to have any color shifts in your grayscale. The 9UK's have bouncy grayscales when @ D65, that's just how they are. It doesn't take from the fact that you should attempt to get D65. When you get your D2 back, use try using zoyd's offset matrix to correct for the D2's inaccurate grayscale (it is quite off compared to an i1Pro).

Regarding the round-robin sharing of the D2, have you guys seen TomHuffman's posts regarding plasma IR and the D2? Could that be contributing to your red issue spongeBob?

I have the dtp-94 and have done 2 plasmas (fujitsu couple years old and a brand new samsung 5054). Both dialed in to D6500 according to the instrument. I saw no issues by eye.

If you can see a red tinge, and you think you're at D65k, then somethings wrong. Could you have an issue with your plasma? You do mention red runout. Might there be a hardware (in the panel) issue?

Best,
jeff

Regarding the round-robin sharing of the D2, have you guys seen TomHuffman's posts regarding plasma IR and the D2? Could that be contributing to your red issue spongeBob?

I have the dtp-94 and have done 2 plasmas (fujitsu couple years old and a brand new samsung 5054). Both dialed in to D6500 according to the instrument. I saw no issues by eye.

If you can see a red tinge, and you think you're at D65k, then somethings wrong. Could you have an issue with your plasma? You do mention red runout. Might there be a hardware (in the panel) issue?

Best,
jeff

Jeff -- check out this thread, link. (http://www.avsforum.com/avs-vb/showthread.php?t=836076)

It contains the best analysis on the S2/D2/i1pro performances on plasma yet. zoyd and others put in a lot of work for the community to see/understand the behaviour. Even a solution....which I've been enjoying for awhile now.

Green/Blue is overpowering his red. This is easily fixed by using the SM's grayscale adjustments...(setting red as weakest rather than green).

Regarding the round-robin sharing of the D2, have you guys seen TomHuffman's posts regarding plasma IR and the D2? Could that be contributing to your red issue spongeBob?

I have the dtp-94 and have done 2 plasmas (fujitsu couple years old and a brand new samsung 5054). Both dialed in to D6500 according to the instrument. I saw no issues by eye.

If you can see a red tinge, and you think you're at D65k, then somethings wrong. Could you have an issue with your plasma? You do mention red runout. Might there be a hardware (in the panel) issue?

Best,
jeff

Hi Jeff,

The D2/plasma problem is not due to any excess of near-Infrared radiation but due to an inaccurate calibration. Both Tom and I have shown excess red (and reduced blue) readings from the D2 and Tom has also shown that adding a NIR filter does nothing. Whether calibrating with it in this state produces noticeable errors by eye is not certain, the enhanced red reading will force you to add blue/green to offset and the temperature difference is ~100-200 K (to the cool side). I've posted a correction matrix in the thread colin pointed out or one could just shoot for a color temperature of ~6350K to compensate.

bob - changing the grayscale by eye isn't going to do much...by turning down red so much, you're probably pushing toward blue or blue/green quite hard. Blue is less offensive-looking on the grayscale, which is why you don't 'appear' to have any color shifts in your grayscale. The 9UK's have bouncy grayscales when @ D65, that's just how they are. It doesn't take from the fact that you should attempt to get D65. When you get your D2 back, use try using zoyd's offset matrix to correct for the D2's inaccurate grayscale (it is quite off compared to an i1Pro).

colin

I will try the matrix change but pdawg has done that and SM adj, *and* still has the obvious red tinge in the gray patterns. I know I prolly ended up bluish but watching DVD and DTV the grays and browns don't have the red tinge.

Question for you and others:

After calibrating to D65, do you see any visible red tinge using getgray patterns (5% and gradual)?

thx

bob

colin

I will try the matrix change but pdawg has done that and SM adj, *and* still has the obvious red tinge in the gray patterns. I know I prolly ended up bluish but watching DVD and DTV the grays and browns don't have the red tinge.

Question for you and others:

After calibrating to D65, do you see any visible red tinge using getgray patterns (5% and gradual)?

thx

bob

Bob-

I know Colin can/will answer this question but just to give you one now, I had asked him this and he said due to the spikes in red around D6500, some of his bars look a bit discolored as well...I believe it is just the way it is with the 9UK...

colin
After calibrating to D65, do you see any visible red tinge using getgray patterns (5% and gradual)?

thx

bob

The 5% pattern always looks slightly reddish to me (10% and up, no). In the ramp on my plasma I can notice a bit of color pop up here and there but that is a limitation of 8 bit color.

The 5% pattern always looks slightly reddish to me (10% and up, no). In the ramp on my plasma I can notice a bit of color pop up here and there but that is a limitation of 8 bit color.

So you don't see any color at 10IRE and above? Maybe I'm misunderstanding your statement...

So you don't see any color at 10IRE and above? Maybe I'm misunderstanding your statement...

correct, and my deltaE for 10%-100% < 3

zoyd has a 600u. From all reports I've read so far, it appears the 9UK has a slightly more bouncy grayscale. This doesn't mean you can't achieve an accurate grayscale.

I've zoomed into 90%-110% view here, which is pretty extreme....but it shows you how bouncy it is. Where you see red spike at 60ire and 75ire, you can actually observe on the 5% ramp pattern. But in all honesty....if you watch from where I watch (about 12ft), it's extremely hard to see on the ramp pattern....and when you actually add color to the mix, this difference is gone....all dE is below 4 in this case....I forget the magic number....but I believe if your grayscale is below 6dE is all spots, it's 100% impossible to observe in real-world material....material with color applied that is.

edit - this is my current-best vga run...i'm still working on small compromises to get red/blue down @10ire.

Jeff -- check out this thread, link. (http://www.avsforum.com/avs-vb/showthread.php?t=836076)

It contains the best analysis on the S2/D2/i1pro performances on plasma yet. zoyd and others put in a lot of work for the community to see/understand the behaviour. Even a solution....which I've been enjoying for awhile now.

Green/Blue is overpowering his red. This is easily fixed by using the SM's grayscale adjustments...(setting red as weakest rather than green).

Actually, red is set to FC. The red tinge is obvious to my wife when watching real world material too.


thx

bob

Actually, red is set to FC. The red tinge is obvious to my wife when watching real world material too.



thx

bob

right, sorry i should clarify. setting red to FC so that you can maintain a red at high IREs doesn't mean that there won't be bounciness in the grayscale.


As for being able to see it in real-world material....I'd wait until you've done a grayscale with a correction matrix.

Although I can tell you that your D2 is overreading red....so you're going to end up increasing red when running a correction based balancing.

.

but I believe if your grayscale is below 6dE is all spots, it's 100% impossible to observe in real-world material....material with color applied that is.

.


Would that mean viewing gray material( color applied?) like walls/floors, etc? They would still show a slight tinge of red?


bob

Would that mean viewing gray material( color applied?) like walls/floors, etc? They would still show a slight tinge of red?


bob

no.

Remember that typical tv shows are far from what you'd call 'reference material'. most are crap. some channels at least put 'some' effort into broadcasting a good channel, *some* material on discovery HD, HDNET, and a couple of movie channels come to mind. But in general, even my SDTV looks quite natural after calibration.

Even when you enter DVDs, etc., you have to remember what you're watching....if you saw it in the movie-theater, try to remember what it looked like there!

For example.....the movie Traffic.....great movie...but if you just pop it in, you're going think your TV is TRASH.....that's because the director wanted to use filters to differenciate the locations. Washington scenes use a blue filter. Mexico scenes have a yellow/orange filter, etc. That's the way it was intended!

in short- don't use the movie 'Traffic' for color balance evaluations :)

thanks zoyd and colin... I missed the elite + HCFR thread... I had seen Tom's info, and say zoyd's latest, but missed most in between. Sorry to distract from the current thread...

I do agree that the dtp-94 is a really good probe. I contemplated the D2, but the issues with it make me glad I got what I got... If anyone has a good source for dessicant packets and new dtp-94's please let me know ;-)

Best,
jeff

no.

Remember that typical tv shows are far from what you'd call 'reference material'. most are crap. some channels at least put 'some' effort into broadcasting a good channel, *some* material on discovery HD, HDNET, and a couple of movie channels come to mind. But in general, even my SDTV looks quite natural after calibration.

:)


I totally agree, But... we scanned through many channels and DVD's and saw the red "push" on all when changing modes. I'm looking forward to recalibrating :)


bob

I totally agree, But... we scanned through many channels and DVD's and saw the red "push" on all when changing modes. I'm looking forward to recalibrating :)


bob

Bob,

The easiest way for me to separate gray scale problems from other things is to watch some B&W material from DVD or AMC. If you see the red problem there then it is in the gray scale, if you don't, then it is elsewhere, like an oversatured red primary.

Bob,

The easiest way for me to separate gray scale problems from other things is to watch some B&W material from DVD or AMC. If you see the red problem there then it is in the gray scale, if you don't, then it is elsewhere, like an oversatured red primary.


I clearly did see it on TCM B/W. This is switching from standard (cal settings, high red 24) to cinema (high red -1).

I realize that higher blues can fool the brain but it wouldn't "create" red tinge by switching back and forth could it?

bob

I clearly did see it on TCM B/W. This is switching from standard (cal settings, high red 24) to cinema (high red -1).

I realize that higher blues can fool the brain but it wouldn't "create" red tinge by switching back and forth could it?

bobAre you by any chance calibrating the component inputs using GetGray/DVE/Avia and then applying those settings to HDMI for your HD cable box? From my experience, the settings for each input are very different.

With regard to adjusting by eye, I can usually spot too much red just by staring and looking for it in the background of the GetGray main menu, as long as green and blue are fairly lined up, and without switching to another setting on the W/B menu. However, when you switch back and forth between differently calibrated settings you can easily fool yourself (your brains color adjustments) into seeing one color or another. I always think things look green when I do that :). I was reading a recent article on new technology that allows counting the number of different cones and rods in the eye (rather than in cadavers were the pigment is lost). The study found huge high/low variances in the number of these cones in individuals. However, all these individuals had similar levels of color perception. So it would seem that most color perception occurs in the brain.
EDIT: Link to article:http://www.sciencedaily.com/releases/2005/10/051026082313.htm

If you want to really make yourself crazy, check out some of the color perception illusions here:http://www.echalk.co.uk/amusements/OpticalIllusions/illusions.htm

This software is so great! I thought that Spectrum feature didn't work yet, but I just realized it does with the i1 Pro. I am not sure if it does anything for calibration, but it is so cool!

Here's D65 (or close to it) on my 8Uk:
http://i155.photobucket.com/albums/s312/jdbimmer/Spectrum8UK.jpg
**Geek Alert**

That is cool.

It would be interesting to compare the spectra of different plasma models.

That is cool.

It would be interesting to compare the spectra of different plasma models. I agree. Too bad I don't have a few here :) . But I do have a RPTV CRT, Direct view TV and PC monitor CRT's, and a PC LCD monitor to compare, well, for the heck of it.

Wow, it displays the primaries and secondaries too.

Here's Green (hey what's with that red in my Green! :eek: )
http://i155.photobucket.com/albums/s312/jdbimmer/GreenPrim.jpg

CYAN:
http://i155.photobucket.com/albums/s312/jdbimmer/SecondaryCyan.jpg

Okay, I must be annoying some of you by now, so I'll stop... :o

Wow, it displays the primaries and secondaries too.

Here's Green (hey what's with that red in my Green! :eek: )


That red is panasonic's attempt to balance the overwhelming green phospor emission. I just wish they'd add controls to adjust that peak, it would allow us to get the color spot on.

Here is red, if we could add a little blue to it we could pull it closer to reference.

http://i64.photobucket.com/albums/h173/prairie_2006/spec1.jpg

Zoyd,

I understand that the i1 Pro reads those 36 points shown on the graph to determine the color coordinates, but if HCFR can show them, does that mean it passes XYZ (or xyY) and these values as well to the software, and the values are stored in the CHC files?

-JD
EDIT: I see that the CHC file is about 5K bytes bigger for the i1, so I guess it does.

Thanks..

Here is red, if we could add a little blue to it we could pull it closer to reference.Even though our PDP glass is the same, I am still amazed that the Red primaries on your 600u are pratically identical to mine. Very cool.

Zoyd,

I understand that the i1 Pro reads those 36 points shown on the graph to determine the color coordinates, but if HCFR can show them, does that mean it passes XYZ (or xyY) and these values as well to the software, and the values are stored in the CHC files?

-JD
EDIT: I see that the CHC file is about 5K bytes bigger for the i1, so I guess it does.

yes, it passes both. I've actually looked at the spectra for the red coordinate in both cinema mode and vivid mode and found that in vivid mode there are very weakly elevated levels of blue and green and this moves the coordinate closer to rec709. The shift is very noticeable, 0.011 in x and decrease deltaE from 17 to 5. Now if only I find out where in the firmware these adjustments are stored. :rolleyes:

Are you by any chance calibrating the component inputs using GetGray/DVE/Avia and then applying those settings to HDMI for your HD cable box? From my experience, the settings for each input are very different.

No, my DVD player has component and HDMI out. BUT, I am plugging the DTV HR-20-700 into the HDMI slot and using those settings. How much the DVR is altering the color, I'll never know. And most of my problems are with the Sat signal, not the DVD.

With regard to adjusting by eye, I can usually spot too much red just by staring and looking for it in the background of the GetGray main menu, as long as green and blue are fairly lined up, and without switching to another setting on the W/B menu. However, when you switch back and forth between differently calibrated settings you can easily fool yourself (your brains color adjustments) into seeing one color or another.

So, I could be making up the red when switching from a cooler (bluer) preset?


If you want to really make yourself crazy, check out some of the color perception illusions here:http://www.echalk.co.uk/amusements/OpticalIllusions/illusions.htm


OK, everyone must check out that link. Do the color perception illusions but don't bet your life like I did. To prove it's real, we made a "funnel" out of black paper and isolated the colors in question and they don't change !!! Freeking awesome :)

bob

yes, it passes both. I've actually looked at the spectra for the red coordinate in both cinema mode and vivid mode and found that in vivid mode there are very weakly elevated levels of blue and green and this moves the coordinate closer to rec709. The shift is very noticeable, 0.011 in x and decrease deltaE from 17 to 5. Now if only I find out where in the firmware these adjustments are stored. :rolleyes: Does that mean that Vivid mode might be a better choice for viewing/calibrating gray scale? In the 8UK, I can use Vivid mode and select 2.0, 2.2, 2.5 rather than Vivid's default "S-Curve" gamma.

Does that mean that Vivid mode might be a better choice for viewing/calibrating gray scale? In the 8UK, I can use Vivid mode and select 2.0, 2.2, 2.5 rather than Vivid's default "S-Curve" gamma.

no, you can get the same gray scale out of any mode, vivid won't be any better or worse than the other modes. The fact that the red coordinate is closer to reference means that red will be closer to correct in an image that contains color. Also, this was measured on my 600U, don't know if the 8UK does the same thing.

No, my DVD player has component and HDMI out. BUT, I am plugging the DTV HR-20-700 into the HDMI slot and using those settings. How much the DVR is altering the color, I'll never know. Same here, I have no idea what my SA8300HD cable DVR is doing, I can only calibrate the HDMI from DVD and hope for the best.
So, I could be making up the red when switching from a cooler (bluer) preset? Possibly..

OK, everyone must check out that link. Freaky, isn't it? :)

no, you can get the same gray scale out of any mode, vivid won't be any better or worse than the other modes. The fact that the red coordinate is closer to reference means that red will be closer to correct in an image that contains color. Also, this was measured on my 600U, don't know if the 8UK does the same thing. I think I understand - the gray scale would end up the same, but red on a stop sign or a painted barn for example, would be a "truer" representation of that red.

I think I understand - the gray scale would end up the same, but red on a stop sign or a painted barn for example, would be a "truer" representation of that red.

yes, wherever the primary chromaticities end up you can always find a combination to form the D65 white point. But if they aren't at the r709 points the colors will be wrong when forming a color image.

btw, that link is super cool, the spinning gears and crawling skin are creepy.

yes, wherever the primary chromaticities end up you can always find a combination to form the D65 white point. But if they aren't at the r709 points the colors will be wrong when forming a color image.

.


So, you can be at D65 and still have the colors not right? I thought that was true for 6500K vs. D65, but that's what I get for thinking :)


bob

thanks zoyd and colin... I missed the elite + HCFR thread... I had seen Tom's info, and say zoyd's latest, but missed most in between. Sorry to distract from the current thread...

I do agree that the dtp-94 is a really good probe. I contemplated the D2, but the issues with it make me glad I got what I got... If anyone has a good source for dessicant packets and new dtp-94's please let me know ;-)

Best,
jeff


dessicant packs?


bob

Freaky, isn't it? :)
Very cool. I copied the image into Photoshop and checked with the dropper tool just to be sure. It's indeed the same color in all three illusions. This raises the important and often overlooked issue of bias light behind the set. If you don't have a D65 light behind your set, you will throw off the perception of all the colors displayed! Any lamp next to your set will have an impact.

dessicant packs?


bob


Packets of silica you put in a sealed container with the probe to reduce humidity. The filters used in these probes gradually change their spectral response when exposed to water vapor over long periods of time.

Very cool. I copied the image into Photoshop and checked with the dropper tool just to be sure. It's indeed the same color in all three illusions. This raises the important and often overlooked issue of bias light behind the set. If you don't have a D65 light behind your set, you will throw off the perception of all the colors displayed! Any lamp next to your set will have an impact.


Good idea, Carl. We made a small isolation tube out of black paper and the color didn't change!


bob

So, you can be at D65 and still have the colors not right? I thought that was true for 6500K vs. D65, but that's what I get for thinking :)As zoyd said, no matter where the primaries are, you can find a combination that will produce D65, but other colors will be off if the primaries aren't right.

As a thought experiment, imagine an extreme primary error: when you tell your display to show red, it shows green; green shows blue; and blue shows red. Obviously most colors are going to be totally whack! Stop signs are green, grass is blue, the sky is red. But you can still combine those three primaries to get a D65 white.

So the ability to get a D65 white doesn't mean your colors are right. Most displays will be a whole lot closer than the extreme example above, but they can still affect the colors that are displayed. Unless your primaries are dead on the SMTPE standards, the colors you see won't be quite the same as the colors the camera saw.

For a real-world example I'm familiar with: most CRT projectors need color filtering on the green and (to a lesser extent) red. The "native" color produced by green phosphor is too yellow. See the attached pic for the measured xyY for my Marquee 8500 CRT projector, without color filtering and with several different color filters. Without filtering you can nail D6500 no problem, but the colors look muted and dead. (To me, anyway, but many people run 8500's without filtering.) Adding filtering can move the green primary much closer to the correct value, and colors & skintones look much more realistic.

As zoyd said, no matter where the primaries are, you can find a combination that will produce D65, but other colors will be off if the primaries aren't right.

As a thought experiment, imagine an extreme primary error: when you tell your display to show red, it shows green; green shows blue; and blue shows red. Obviously most colors are going to be totally whack! Stop signs are green, grass is blue, the sky is red. But you can still combine those three primaries to get a D65 white.

So the ability to get a D65 white doesn't mean your colors are right. Most displays will be a whole lot closer than the extreme example above, but they can still affect the colors that are displayed. Unless your primaries are dead on the SMTPE standards, the colors you see won't be quite the same as the colors the camera saw.

For a real-world example I'm familiar with: most CRT projectors need color filtering on the green and (to a lesser extent) red. The "native" color produced by green phosphor is too yellow. See the attached pic for the measured xyY for my Marquee 8500 CRT projector, without color filtering and with several different color filters. Without filtering you can nail D6500 no problem, but the colors look muted and dead. (To me, anyway, but many people run 8500's without filtering.) Adding filtering can move the green primary much closer to the correct value, and colors & skintones look much more realistic.

So if my primaries and secondaries are off on my 9UK, I'll never get accurate color even with a decent gray scale? This is my experience so far. Maybe the new Panny (750) will be better?

bob

So if my primaries and secondaries are off on my 9UK, I'll never get accurate color even with a decent gray scale? This is my experience so far. Maybe the new Panny (750) will be better?

bob

hey bob, I thought you had a pio elite for awhile, on those you can plunk the primaries wherever you want. The new panny's will have the same problem as the old, until they add CMS controls we are stuck with where they think we want 'em.

So if my primaries and secondaries are off on my 9UK, I'll never get accurate color even with a decent gray scale? This is my experience so far. Maybe the new Panny (750) will be better?

bob

An accurate D65 grayscale corrects your color of black-gray-white. But it does have an effect on color.

Think of it like preparing your canvas for the painting.

There are many factors that go into accurate color....grayscale is key...it's the 'canvas'. Primary color accuracy is also imporant....as that determines what the color of red/blue/green actually start with when painting your picture.

Then there's color decoding, think of this as the 'inbetween'....it's the choices (decoding) made to mix primaries to create other colors (how much red, blue, and green go in this color?, etc.) That's why color decoding determines your secondary colors, etc.

hey bob, I thought you had a pio elite for awhile, on those you can plunk the primaries wherever you want. The new panny's will have the same problem as the old, until they add CMS controls we are stuck with where they think we want 'em.


I did but returned it. Didn't have the "pop" the 9UK does.

The new 750 is supossed to add CMS and full user WB.


bob

An accurate D65 grayscale corrects your color of black-gray-white. But it does have an effect on color.

Think of it like preparing your canvas for the painting.

There are many factors that go into accurate color....grayscale is key...it's the 'canvas'. Primary color accuracy is also imporant....as that determines what the color of red/blue/green actually start with when painting your picture.

Then there's color decoding, think of this as the 'inbetween'....it's the choices (decoding) made to mix primaries to create other colors (how much red, blue, and green go in this color?, etc.) That's why color decoding determines your secondary colors, etc.

The way I like to think of it is that the gray scale is the "lamp" that illuminates the canvas, too warm and it will give the picture a sepia, 40 watt bulb cast, too cold and you get that winter, brilliant sun illuminated look.

So if my primaries and secondaries are off on my 9UK, I'll never get accurate color even with a decent gray scale? Yes. Suppose your DVD wants to display a green that lands exactly on the "green" corner of the SMPTE-C color chart. To properly display that color, your display should show 100% green, 0% red and 0% blue. If your green primary is correct -- that is, if it is ALSO exactly on the green corner of the SMPTE-C triangle -- then you'll see exactly the same color that the camera saw. But if your display's "green" primary is actually blue, or pink, or yellow-green, then you're going to SEE blue, pink, or yellow-green. You can't fix that with the grayscale because that's what your display thinks "green" is.

The way I like to think of it is that the gray scale is the "lamp" that illuminates the canvas, too warm and it will give the picture a sepia, 40 watt bulb cast, too cold and you get that winter, brilliant sun illuminated look.Great analogy. So if you illuminate the picture with a proper "white" light, your colors should look right. But if you go to paint some green grass, and your green paint is actually blue, or pink, or yellow-green... then your painting will look sick. That's what happens when your primaries are off.

The way I like to think of it is that the gray scale is the "lamp" that illuminates the canvas, too warm and it will give the picture a sepia, 40 watt bulb cast, too cold and you get that winter, brilliant sun illuminated look.

i like that...even better.