SOme Mits (CRT) do the same thing. so the best thing to do, is go in and out of the Service Menu. Do a full run, then highlight one IRE of concentration, lets say 100IRE, then click on Display Meaurement and look at your color bars, and lets say Red is 90, Green 101 and Blue 123..Go to your DRIVE or Contrast (whatever parameter to handle the brightside of the scale), and then decrease blue some and increase Red some and save/exit, then do a CONTINUOUS READING on it, to see what has change..its a pain in the butt but its the only way to go...

Low 0 through 30
Mid will be 40 thru 60-70
High 80 and up

I will give that a try. The only thing that concerns me right now is that as far as I know, the only way to get back into the service menu is to power the display down, enter key sequence, and power it up. So there will be a lot of powering off and on. Thanks for explaining exactly what to do, it's a big help.



I have the 6070 and had a similar issue with the shade of green. Your's at least ends up with a good yellow. ( Your green actually looks close to the original spec for NTCS green. http://www.imagingscience.com/cie1931_phosphor.htm)

I also noticed that the set does not stay in "warm/lo" setting when in the service menu. I did as the other poster suggested and made changes in the service menu and then exited to take new measurements.

I think the comment to "calibrate to the mid-grayscale" is indicating that if you calibrate the gray scale while in the service menu, you are actually working with the mid setting. Maybe the other poster was suggesting to use the mid setting for viewing after inherently calibrating it in the service menu. I would stick with the plan in the previous paragraph if it were me though.
Yesterday 08:16 PM

Ok, that makes sense. Do you know if there is a way to get back into the service menu after you have exited it without having to power the display down and back on again? Also, I've been unable to find any gamma setting in the service menu, have you? Are there any other settings (besides RGB High & Low) in the service menu that you used when calibrating your 6070?

BTW, would you mind posting your HCFR file? If you don't care, I'd like to see how your diagrams look.

Thanks for your help.

.....
Ok, that makes sense. Do you know if there is a way to get back into the service menu after you have exited it without having to power the display down and back on again? Also, I've been unable to find any gamma setting in the service menu, have you? Are there any other settings (besides RGB High & Low) in the service menu that you used when calibrating your 6070?

BTW, would you mind posting your HCFR file? If you don't care, I'd like to see how your diagrams look.

Thanks for your help.

I haven't found a way to do anything with the gamma in the service menu. There is some mention of gamma relative to using the rs232 serial interface in the service manual, but no description of what is does.

I also had to power off to get back to the service menu.

In this post are some of my earlier HCFR files:
http://www.avsforum.com/avs-vb/showthread.php?p=9474114&&#post9474114

also:
http://www.avsforum.com/avs-vb/showthread.php?p=9481388&&#post9481388


re: one of your ealrlier questions, the service manual says the ABL item is "setting of the power comsumption". It seemed to increase brightness slightly, but I thought I also started seeing some image retention with higher settings so I went to back to the original value.

I just spent this past weekend calibrating my TV with a DTP94 for the very first time and I must say, as a frequent user of FLOSS this is probably the best quality I have seen, so kudos to you HCFR guys!

However what good would FLOSS be without some feature requests :) So here is mine. I really liked the continuous mode to calibrate my grayscale on-the-fly but one thing I wish I could do was set the values of the max/min for the axis like you can when you view non-continuous measurements. Currently I can only choose a pre-defined range for RGB and dE (I believe 50-150 and 0-5 respectively.) When you are getting close to fine tuning RGB to 100 I would really like to define a min/max of 95-105 for example.

This item is under construction in the Help menu. Can someone please point me to step by step instructions for using the software? With over 1700 posts in this thread it is a bit daunting :confused: I have the Display LT and will be calibrating a Mitsubishi HD1000U DLP front projector. Any help getting started would be much appreciated.

thanks,
Jason

Hi! Laric. I know that in your soft I can change the coordinates of the primary colors, but no the secondary colors.

I read this:

http://img297.imageshack.us/img297/6239/avsforumgb7.jpg


This one is from a document of Sillicon Optix (about Yamaha dpx 1300). The full text is here:

http://www.siliconoptix.com/contentEngine/contentDocuments/Yamaha_DPX-1300_-_Widescreen.pdf

Well I ask to myself, if it isn't very interesting that I/we can change the secondary colors too (in HCFR soft).


Regards

This item is under construction in the Help menu. Can someone please point me to step by step instructions for using the software? With over 1700 posts in this thread it is a bit daunting :confused: I have the Display LT and will be calibrating a Mitsubishi HD1000U DLP front projector. Any help getting started would be much appreciated.

thanks,
Jason

The closest thing to "step by step" instructions that I have come across is THIS (http://geekwithfamily.com/2006/12/16/home-theater/howto-hcfr-free-video-projector-calibration-software/).

And post 708 might also help you out. (http://www.avsforum.com/avs-vb/showthread.php?p=9165873&&#post9165873)

Reading thru the entire thread does help tho. Reading the posts and downloading the charts to see what the posters are discussing has helped me out.

Good luck

The closest thing to "step by step" instructions that I have come across is THIS (http://geekwithfamily.com/2006/12/16/home-theater/howto-hcfr-free-video-projector-calibration-software/).

And post 708 might also help you out. (http://www.avsforum.com/avs-vb/showthread.php?p=9165873&&#post9165873)

Reading thru the entire thread does help tho. Reading the posts and downloading the charts to see what the posters are discussing has helped me out.

Good luck

Thanks much!

This item is under construction in the Help menu. Can someone please point me to step by step instructions for using the software? With over 1700 posts in this thread it is a bit daunting :confused: I have the Display LT and will be calibrating a Mitsubishi HD1000U DLP front projector. Any help getting started would be much appreciated.

thanks,
Jason


Do you speak Spanish?

Here (http://www.mundodvd.com:81/foro/viewtopic.php?t=49366) there is a "megapost" about this soft.

Regards

Do you speak Spanish?

Here (http://www.mundodvd.com:81/foro/viewtopic.php?t=49366) there is a "megapost" about this soft.

Regards

Wish I did- that looks like some post!

CJ

Thanks for this Whisper2, and hopefully we can add spanish for next release... Except your screen shot will have to be done again ;) (due to language and also lots of changes :D)

--Patrice

I speak Spanish :D !

Do you speak Spanish?

Here (http://www.mundodvd.com:81/foro/viewtopic.php?t=49366) there is a "megapost" about this soft.

Regards
I do not unfortuantely. thanks though.

Yo tambien! Pues, lo leo mejor que lo hablo. Gracias a Whisper2. (y, espana me encanta--vivio en Salamanca por algunos meses hace vientecinco anos).

And glad to hear that the DTP-94 (Optix XR) is your favorite probe. I was wondering about its merits relative to the EyeOne Display 2

I speak Spanish :D !

Great! You can make a monumental contribution to AVS by translating that web page for us.;) No rush, noon tomorrow is good enough for me.:D

I speak Spanish :D !

Donde esta la biblioteca.. :D

Great! You can make a monumental contribution to AVS by translating that web page for us.;) No rush, noon tomorrow is good enough for me.:D
:eek: !!!!

You could search for a browser translator plugin. I use it on Firefox and works great. Of course, I don't use it for spanish ;) ...

:eek: !!!!

You could search for a browser translator plugin. I use it on Firefox and works great. Of course, I don't use it for spanish ;) ...

Y tu te llamas Hispano..que tragedia, que sin verguenza, que lastima,que que..ah - ya se me olivida lo que hiba decir...puro broma amigo...

Noon tomorrow, Mexico time?? works for me too..no apures, sin prisa..

despues -

EL BORICUA DE CHICAGO!!! GO BEARS!!!

Acho brodel, El plugin es para los que no hablan o leen espanol ;) ...

Claro, Bears pa'lante...! Solo porque los Colts le ganaron a mis Patriotas :eek: !!

Definition of Saturation and Saturation/color control + Using HCFR to adjust color/saturation setting
I just want to clarify some terminology and concept. Please correct me if I am wrong.
The true definition of saturation of a primary or a secondary is the measurement of how pure is the color. It is indicated by the position of the color along the line between white and the reference cie color point. The closer to white the less saturated and if it is beyond the reference point than it is oversaturated.
The saturation/ color control dosen't really change the true saturation(unless with extreme settings) but rather change the intensity(luminance) of the color. So I can have red primary that is bang on with the cie reference and still have sun burn skin tone because the color/saturation is not adjusted correctly.
Than there is the issue of color decoder problem such as red push. If one use the blue color filter to adjust the color/saturation setting of a display you assume that the other color intensity(luminance)will be set right by the decoder. But we know there are decoder problem such as red push. Therefore there are red and green filter to be used with the avia disc to detect decoder error.
I have come up with a way to adjust color/saturation and detect red push using HFCR instead of the 3 different color filter. Using a standard SMPTE standard test image I measured the RGB value of the 75%(ire) white, red, blue, and green primaries. Theoretically the RGB value of the white should matches the corresponding value of the primaries when the color/saturation control is set properly and there is no decoder error. So what I do is adjust the color/saturation control until the B value of the 75% white is equal to the B value of the 75% Blue Primary. Than I compare the R value of the white and R value of the Red Primary. If the R value of the Red is higher than the R value of the White than there is red push. Because this is a decoder error and unless one have ways of changing the decoder matrix within the display one can only use the color/ saturation control to tune the red push down knowing that one will screw up the blue and green . Because red decoder error shows up in skin tone and is very obvious while a less intense blue is less obvious. I therefore use the color/saturation control to bring the R value of the red close to the R value of the white. This method have worked for me with my 42" westy which shows significant red push with VGA input. All comments are welcomed.

Acho brodel, El plugin es para los que no hablan o leen espanol ;) ...

Claro, Bears pa'lante...! Solo porque los Colts le ganaron a mis Patriotas :eek: !!


this Colts vs Bears is too local for me..

klam-

How are you measuring the RGB parts of white? Could you explain what/how you're measuring?

It's quite easy. Just use one of the standard SMPTE test image you find in any test disc. Turn on continuous measurement and open the measurement information tab which is the left most tab and you'll see the RGB value of the color patch your are measuring. The SMPTE test image is the image with the long and short color bars one use to adjust "color/saturation" and tint with the blue filter.
http://en.wikipedia.org/wiki/SMPTE_color_bars

Great! You can make a monumental contribution to AVS by translating that web page for us.;) No rush, noon tomorrow is good enough for me.:D

Just use bablefish.


Just pick the option where you can put the URL for that page into the lower box, and then specify what language you want it translated into for that page using bablefish.

http://babelfish.altavista.com/babelfish/tr

Thanks for this Whisper2, and hopefully we can add spanish for next release...
--Patrice



In Spain there a lot of people who are very interesting in this software and because I was ill (I have to stay at home) I've got a lot of free time for the moment, I made a "short" guide of your software for Spanish people. :)

Well Laric, What about "add the changing the secondary colors" ?:D


Except your screen shot will have to be done again (due to language and also lots of changes )


No problem... ningun problema todo es cuestion de tiempo..!! It's only more time



......
And glad to hear that the DTP-94 (Optix XR) is your favorite probe. I was wondering about its merits relative to the EyeOne Display 2


I prefer an Eye One beamer but "no money"!!.... :(

I've got three probes: Spyder2, DTP94 and the HCFR probe. I can't say nothing about "Display 2" because I haven't got it... No obtante, digo que como la DTP94 se descataloga, mejor es comprar una Eye..lee de nuevo esta en letra negrita ..read again!! :D


Donde esta la biblioteca..


La biblioteca esta en el link anterior, pero por lo que he podido leer de tus "post" tu no la necesitas (sabes un monton!!).Si quieres ayudar a la traduccion del soft...al español en el futuro.. eso si, cuida las faltas de ortografia... :D


Great! You can make a monumental contribution to AVS by translating that web page for us. No rush, noon tomorrow is good enough for me.


:eek: :eek: do you known.... how long it does take to me to make that post? and it isn't finished......yet :eek:



Regards everyone

[QUOTE=
I prefer an Eye One beamer but "no money"!!.... :(

I've got three probes: Spyder2, DTP94 and the HCFR probe. I can't say nothing about "Display 2" because I haven't got it... No obtante, digo que como la DTP94 se descataloga, mejor es comprar una Eye..lee de nuevo esta en letra negrita ..read again!! :D

[/QUOTE]


Yes, as a buyer you have to be concerned about the DTP94 being discontinued, and how long support will be provided. I already have one, so it is a moot point for me.

Anyone tried training the DTP94 to the Eye One beamer? Since it is a technique that seems to significantly improve the color accuracy of the ColorHCFR probe and the Spyder2, it would be interesting to see how it improves the Monaco color performance.

Sorry if this has been asked before, but i'm not really sure how to search for it:

I've read many posts saying that the Spyder II isn't the most accurate of meters. Is there some sort of "correction values" type of thing that can be applied to measuments? The Spyder is the only probe I have, so I can't use any learning features.

I realize the word "correction" implies that the Spyder measurements are wrong, so I guess should ask if there is a way to get 'more accurate' readings with this meter. :)

Yes, as a buyer you have to be concerned about the DTP94 being discontinued, and how long support will be provided. I already have one, so it is a moot point for me.

That's the point... I love DTP94, but it's over, and about Display 2, I think it's "the" probe for HCFR soft ( original HCFR probe, too but it needs a calibration file for each proj).


Anyone tried training the DTP94 to the Eye One beamer? Since it is a technique that seems to significantly improve the color accuracy of the ColorHCFR probe and the Spyder2, it would be interesting to see how it improves the Monaco color performance.


I think it will be available in the upcoming release (1.3)


Regards

Well, user interface and data storage will be improved in v1.3, but you can allready "train" in v1.22.

Eyeone Display 2 / LT is the best value for money... (IMHO)
(beside our probe, of course :D :D :D)

--Patrice

Definition of Saturation and Saturation/color control + Using HCFR to adjust color/saturation setting
I just want to clarify some terminology and concept. Please correct me if I am wrong.
The true definition of saturation of a primary or a secondary is the measurement of how pure is the color. It is indicated by the position of the color along the line between white and the reference cie color point. The closer to white the less saturated and if it is beyond the reference point than it is oversaturated.
The saturation/ color control dosen't really change the true saturation(unless with extreme settings) but rather change the intensity(luminance) of the color. So I can have red primary that is bang on with the cie reference and still have sun burn skin tone because the color/saturation is not adjusted correctly.
Than there is the issue of color decoder problem such as red push. If one use the blue color filter to adjust the color/saturation setting of a display you assume that the other color intensity(luminance)will be set right by the decoder. But we know there are decoder problem such as red push. Therefore there are red and green filter to be used with the avia disc to detect decoder error.
I have come up with a way to adjust color/saturation and detect red push using HFCR instead of the 3 different color filter. Using a standard SMPTE standard test image I measured the RGB value of the 75%(ire) white, red, blue, and green primaries. Theoretically the RGB value of the white should matches the corresponding value of the primaries when the color/saturation control is set properly and there is no decoder error. So what I do is adjust the color/saturation control until the B value of the 75% white is equal to the B value of the 75% Blue Primary. Than I compare the R value of the white and R value of the Red Primary. If the R value of the Red is higher than the R value of the White than there is red push. Because this is a decoder error and unless one have ways of changing the decoder matrix within the display one can only use the color/ saturation control to tune the red push down knowing that one will screw up the blue and green . Because red decoder error shows up in skin tone and is very obvious while a less intense blue is less obvious. I therefore use the color/saturation control to bring the R value of the red close to the R value of the white. This method have worked for me with my 42" westy which shows significant red push with VGA input. All comments are welcomed.

I always adjust the color control with blue/red/green filter (from DVE). Never before, I thought this possibility. It's a very good option


Regards

"I know that Color HCFR is not supposed to run on Windows 98, but I have used it with the internal image generator on an old computer (96mbRAM) running Windows 98, and it has been working as long as I don't get too many windows open at one time. I like using the old computer because I don't have a laptop and I don't want to carry around my XP desktop. I just tried using the old computer with GetGray generating the patterns from a dvd player and it keeps crashing when I start to read a greyscale. I was just wondering if you have any idea why it would work with "view images", but crash when using dvd generator?"

UPDATE
Nevermind--the problem is solved. I reinstalled ColorHCFR directly from setup file to old computer. I had copied 1.22 over from my XP desktop because I had 1.21 on old computer. Problem disappeared. I might have had something to do with copying files over instead of doing a fresh install.

I've read many posts saying that the Spyder II isn't the most accurate of meters. Is there some sort of "correction values" type of thing that can be applied to measuments? The Spyder is the only probe I have, so I can't use any learning features.

I realize the word "correction" implies that the Spyder measurements are wrong, so I guess should ask if there is a way to get 'more accurate' readings with this meter.

if there were some universal constants available which would improve the accuracy, why wouldn't the manufacturer use them top start with? :D You see, errors can be both positive and negative, higher and lower.... if you apply a positive correction from a "negative error" sensor to a sensor with a "psitive error", you make the error worse !!! :eek:

;)

Question:

Could someone please explain why it is necessary to calibrate the eye-one sensor, and what exactly calibrating the sensor entails? I find the directions provided in the HCFR software to be a bit vague.

The instructions direct placing the sensor on a “planar” surface. What’s a planar surface? Does that mean a flat surface? Does the flat surface need to be horizontal? Does ambient light effect the calibration? Would it be best to calibrate in a perfectly dark room, or does that not matter?

Any clarification on this would be very much appreciated.

Thanks

If it's like the EyeOne I have, it should have come with a "holder" that has a little white plate in it. I have the Photo version, but they should all have a calibration plate so they can be zeroed out before taking readings...

I believe that you can either use the calibration plate or the "lens" cap. I remember reading that using the cap was the better approach since it will do a better job of blocking ambient light.

I believe that you can either use the calibration plate or the "lens" cap. I remember reading that using the cap was the better approach since it will do a better job of blocking ambient light.

I think its really just a matter of blocking all the light so it is a true dark reading. Its not actually looking at the plate, for instance (as I understand it).

I think its really just a matter of blocking all the light so it is a true dark reading. Its not actually looking at the plate, for instance (as I understand it).

Thank you lovingDVD. This is finally starting to make sense. So correct me if I'm wrong, but it is not necessary, then, to take the sensor off the tripod to do the calibration. It's only necessary to cover the sensor with something to eliminate any light from getting in.

Thank you lovingDVD. This is finally starting to make sense. So correct me if I'm wrong, but it is not necessary, then, to take the sensor off the tripod to do the calibration. It's only necessary to cover the sensor with something to eliminate any light from getting in.

The EyeOne slips into and out of a mount easily, and that mount is what's attached to the tripod. At least, that's how it is on mine. At any rate if you can find a way to block all of the light and not damage the probe then that may work (i.e. don't stick your finger over the opening :) . To be sure you can come up with your technique and take some measurements, then recalibrate the dark reading in the cradle and retake the measurements and make sure they are identical.

Orwellflash , afaik, the software should run on 98SE, what did not runs is our probe (driver) and many other probes that don't have proper USb support under 98...

Vista is out today, time to shift ;) (BTW, same apply here, software works on Vista, but I wish you good luck to find probes drivers ;))

--Patrice

My Display LT came only with a white plastic protective cover which does not look like it would do a very good job of blocking out ambient light. No mounts, holders, cradles, caibration plates or lens caps.

Although I like the idea of a lens cap because it has a lip around the edge of it which should help to block light. I think I'll dig through my camera gear tonight and see if I can find a lens cap that fits snuggly over the diameter of the suction cup area. I think a black, plastic lens cap should do the trick.

As far as how I mount my eye-one to my tripod... I should take a picture. You all might get a kick out of it.

Question:

Could someone please explain why it is necessary to calibrate the eye-one sensor, and what exactly calibrating the sensor entails? I find the directions provided in the HCFR software to be a bit vague.

The instructions direct placing the sensor on a “planar” surface. What’s a planar surface? Does that mean a flat surface? Does the flat surface need to be horizontal? Does ambient light effect the calibration? Would it be best to calibrate in a perfectly dark room, or does that not matter?

Any clarification on this would be very much appreciated.

Thanks

This is a *really good question*. I was wondering the same thing. I am sure HCFR can answer this. I have interpreted "planar" surface to mean any horizontal flat surface. I never thought that it had to be dark as well. Since the calibration step takes about 1/10th of a second, I assumed it couldn't possibly be measuring light levels in this time, especially low levels, which have a longer averaging window. So what is it doing?

There is also a box at new file startup that says

Black Level
> Calibrate internal sensor offset

What is this doing and is it the same as "calibrate sensor"

What exactly happens during calibration?

Thanks,

RIch

As long as the "planar surface" is flat, smooth, and does not not pass light, the foam ring surrounding the display sensor will prevent light leak from the sides. No other devices should be necessary to ensure the sensor sees "0" light level.

Well, user interface and data storage will be improved in v1.3, but you can allready "train" in v1.22.

Eyeone Display 2 / LT is the best value for money... (IMHO)
(beside our probe, of course :D :D :D)

--Patrice

Well..Im not going to hold your word on this..just a few versions ago it was the DTP-94 now its the Eyeone Display 2, next then its going to be the Spyder2000, then back to the DTP-1000, then the EyeThree, etc...and then you'll say "beside our probe, of course"..hahahah

now im tempted to get an Eyone Display 2 / lt

anyone know know where the best price is for this thing??

There are quite a few Ebay stores selling the DisplayLt for $120 US. This is the best I have seen it for.

Orwellflash , afaik, the software should run on 98SE, what did not runs is our probe (driver) and many other probes that don't have proper USb support under 98...

Vista is out today, time to shift ;) (BTW, same apply here, software works on Vista, but I wish you good luck to find probes drivers ;))

--Patrice


Thanks. Driver for DTP94 seems good. My problem was apparently due to a corrupted file or some other problem that happened when I copied the program over from an XP computer. Anyway, it ran fine after a reinstall from setup.exe. The only crash I had was when I turned off the display without unplugging the probe. Now that I know not to do that :), it's smooth sailing. I don't even want to think about Vista yet!

I just downloaded this program..and was wondering what is the best NTSC disc to use for test pattern?...getgray, DVE, AVIA ?

Also why does it always come up to LCD when I choose the spyder2...and should I change it?

Thanks
RayJr

Well..Im not going to hold your word on this..just a few versions ago it was the DTP-94 now its the Eyeone Display 2, next then its going to be the Spyder2000, then back to the DTP-1000, then the EyeThree, etc...
Well, DTP94 is pretty nice (you can have a look on probes presentation here http://www.homecinema-fr.com/forum/viewtopic.php?t=29838162.
BUT, afaik it is pretty hard to find this probe, it seems deprecated... Also the Eyeone Display 2/LT had a recent price drop...
So, at present, I think the Eyeone Display is probably the best price/value/result deal...
This doesn't mean others are bad, especially if you allready own a probe !
(Except Spyder II :D --> Stop... this is a joke :p )

and then you'll say "beside our probe, of course"..hahahahThat was a joke, our probe have it's own pros and cons and will probably disappoint many occasional users. ;)

--Patrice

Hi Rich
This is a *really good question*. I was wondering the same thing. I am sure HCFR can answer this. I have interpreted "planar" surface to mean any horizontal flat surface. I never thought that it had to be dark as well. Since the calibration step takes about 1/10th of a second, I assumed it couldn't possibly be measuring light levels in this time, especially low levels, which have a longer averaging window. So what is it doing?

Well, I just used the same messages than demo application in GretagMacBeth SDK. And it isn't perfectly clear...

Basically, you have three distinct cases:
1 - calibrating the Eye One Pro/Beamer/Design (the famous spectrophotometer). This one needs to calibrate its black level offsets, very regularly, to achieve a correct accuracy in dark measures. This operation consist to measure a true black, and the probe keeps measured values in memory to perform fine adjustment during next measures. To be reasonably accurate, you must calibrate this probe at least every 10-15 minutes, and to be as good as possible, calibrate before every measure (it's really painful ;) ).

2 - calibrating the Eye One Display in LCD mode. As above, it needs to measure a perfect black. Just close light input and calibrate black level. Anyway, this measure is much more stable than with spectrophotometer. You can calibrate the probe once, and go on measuring.

3 - calibrating the Eye One Display in CRT mode. This is a little different operation. Instead of measuring black, you measure a white, produced by the CRT to measure. This operation consists in measuring the interval between light peaks (the monitor frequency). Of course it cannot be done while measuring black ;) So, you must do this calibration once, before any measuring, and of course every time you change your CRT frequency. Now, I'm not used to measure CRTs with this probe... I suppose it would be better to calibrate manually the probe in LCD mode (to adjust black levels) then switching to CRT mode and recalibrate to adjust the frequency measure... It is not described in documentation, but technically I suppose it may be a good idea to make both calibrations for CRT. Just my opinion ;)


There is also a box at new file startup that says

Black Level
> Calibrate internal sensor offset

What is this doing and is it the same as "calibrate sensor"

What exactly happens during calibration?

It's exactly the same operation: you can launch it manually with this button.

Regards
Georges

I just downloaded this program..and was wondering what is the best NTSC disc to use for test pattern?...getgray, DVE, AVIA ?

Also why does it always come up to LCD when I choose the spyder2...and should I change it?

Thanks
RayJr

GetGray is my favorite, all of them will work, but as far as ease of use, then go with GetGray

My Mitsubishi HD1000 has a "Setup" option with Auto, OFf 3.75% and 7.5%
The manual says Select OFF to make black lighter, Select 3.5% or 7.5 % to make black intense.

Should I select 7.5% Set Up?

From Calman's Calibration Guide:
"Strictly speaking, NTSC-based displays (both high definition and standard definition) use a 7.5
IRE pedestal (or “set-up”) for black. In other words, the TV will produce its black level when it
sees a signal of 7.5 IRE (~54mV). This makes the common practice of equating IRE levels to
percent stimulus very confusing (e.g., is 10IRE 10% or 2.7%?).
Both Joe Kane Productions’ Digital Video Essentials (DVE) and Ovation Multimedia’s AVIA
Guide to Home Theater (AVIA) assume the presence of set-up in their North American (NTSC)
editions. When calibrating a display, it is important to know whether your display has, and test
patterns assume, the presence of set-up. Non-North American users will need to change the setup
toggle for HDTV calibrations."

Thanks,
Jason

Well, DTP94 is pretty nice (you can have a look on probes presentation here http://www.homecinema-fr.com/forum/viewtopic.php?t=29838162.

--Patrice


Wow! Michel obviously has the curiosity and discipline of a good scientist. Very impressive. Looks like Eye One Display2 is the best choice for someone buying a probe today. No need to throw away those DTPs, Spyders, or HCFRs either;)

Think this answers my question about training to the Pro--now just the small matter of getting one to train to:eek:

Question

Shouldn't my probe give consistent results from one measurement to the next even though there obviously is error involved? On continuous measurements, my delta E jumps around between, say, 0-6 on an RGB luminance measurement. I know it has that error range, but shouldn't it be consistent in its error?

Wow! Michel obviously has the curiosity and discipline of a good scientist. Very impressive.
Yes, and we are lucky and proud to have in on board ;)
(You can search for his posts there, there some others pretty usefull (imho)).

6 Delta E is pretty high difference, you usually have 1-2 dE variation... a bit more with SpyderII, could be 4 or 5 dE (depends conditions due to wide angle and if you measure LCDs or not)

--Patrice

Is it normal to have it bounce around in continuous reading mode, every time a reading is taken? BTW, this is happening on a direct view CRT, using DTP probe. I am wondering if my probe is malfunctioning. I may be overstating the size--it might have been in a range of 3-4 Dela E, 4% luminance for a given color.

Yes, and we are lucky and proud to have in on board ;)
(You can search for his posts there, there some others pretty usefull (imho)).

6 Delta E is pretty high difference, you usually have 1-2 dE variation... a bit more with SpyderII, could be 4 or 5 dE (depends conditions due to wide angle and if you measure LCDs or not)

--Patrice

Patrice:

I really love this product. However, sometimes the readings take a very long time and I am wondering if this is normal or just that I'm running on an older pentium 3 laptop. It is especially evident in continuous measure mode. I typically see about one new measure every 3 seconds or so, then for no apparent reason it stops and takes about 30 seconds until the next data point is displayed. Occasionally, but less frequently, it crashes altogether and I can not even kill the application from task manager. Have to power cycle. I notice similar behavior in the 10-step grayscale meas, but it has never crashed here.

1 Clue. I have found that the problem is definitely triggered if I increase the display brightness in real time. For example, if I am in continuous measure mode on the 20% pattern and all is well, then I switch to the 80% pattern the software will stop plotting for a good 15-30 seconds or crash entirely. If I hit "stop meas" before changing the pattern brightness and then "start meas" after changing it seems to work fine. Something is wrong when there is an input level shift perhaps.

Thanks for any help

Rich

I have done literaly thousands of continous measure when demo-ing the software, never crashed... But that was using our probe...

I don't have other probes personaly, will ask other people if they have experienced similar things...

I know S2 is sometimes a bit touchy and can have slow response, anyway not sure that will kill app...

Can you tell me what probe you use ?

--Patrice

I have been using the DTP-94 with ColorHCFR to calibrate my Panasonic TH-50PHD8UK Plasma. I know that the consensus on the Spyder is to use LCD mode for everything but CRT, but does this apply to other sensors like the DTP-94?

In LCD mode, the DTP-94 sensor gives higher values for Red (relative to Green and Blue) than it does in CRT mode. Given that plasma uses something loosely resembling CRT RGB phosphors, will CRT mode provide more accurate results? To my eyes, calibrating using LCD mode results in somewhat greenish/yellow grays and muted colors.

Also, can anyone explain what the technical difference (internal filters, calculations, etc) is for the sensor's LCD and CRT modes?

Thanks,
John

I have done literaly thousands of continous measure when demo-ing the software, never crashed... But that was using our probe...

I don't have other probes personaly, will ask other people if they have experienced similar things...

I know S2 is sometimes a bit touchy and can have slow response, anyway not sure that will kill app...

Can you tell me what probe you use ?

--Patrice

Patrice:
I am using the Eye One Display LT. Performance is excellent other than this sporadic slow response and occasional crash issue.

Thanks
Rich

Thanks Rich, I know Georges is investigating...

--Patrice

Lyckman!!!! Your the MAN!!! Problem solved!!!!

For ANYONE having trouble with this meter locking the HCFR software use the .dll I have attached. It is version 1.2.0.0 and completely solved my lockup issues with the DTP94!


Thanks Lyckman and Adam!!! :) I just had the same problem. Have been using Color HCFR on an older computer, just put it on new XP machine, it kept locking up when I tried to use the DTP94. I downloaded your new driver, and it fixed the problem. I was going crazy until I remembered seeing your post a while back. Don't know what the problem was but sure am glad it is fixed. BTW, file link is post 1233 on page 42 of this thread.

Did anyone ever decide if using the HCFR PAL pattern disc with a DVD player that can play PAL on an NTSC TV is going to result in a trustworthy test pattern, or is the PAL to NTSC conversion a source of uncertainty that should be avoided?

just got the new diplay lt today and I tell you a lot faster than my previous spyder. I do have 1 question for display lt user do u use crt mode or lcd mode I am trying to calibrate a plasma tv. also is it necessary to attach the ambilight cover on the sensor?

Thanks Lyckman and Adam!!! :) I just had the same problem. Have been using Color HCFR on an older computer, just put it on new XP machine, it kept locking up when I tried to use the DTP94. I downloaded your new driver, and it fixed the problem. I was going crazy until I remembered seeing your post a while back. Don't know what the problem was but sure am glad it is fixed. BTW, file link is post 1233 on page 42 of this thread.Good thing it worked out for you too! :)

Maybe a note regarding this issue in the first post in the thread would save some trouble..... ;)

// Lyckman

I am currious to try out HCFR colorimeter and this application. I have never done calibration before, but one has to start somewhere right? I can do soldering so this DIY project seems right to me, however I am struggling to find the way to program 18F2550 . I have following programer from the old days:
http://www.adteknik.se/english/circuits.htm
This however has only 16 holes, and mic has 28 legs ...

As Laric mentioned that microchip can be programmed with any JDM PIC programmer, which seems to have less than 28 legs, I got some hope that maybe it is possible to use my programmer in some way. Any advice????

If not, it would be great to be able to by preprogrammed chip including the box for 1$ as Laric quoted the price somewhere.... . If there are more users wanting to try this out, but not able to get chip programmed or even find all parts locally, it would be great to buy collection of all parts from somebody (with preprogrammed chip) and then work on the project. I know this would be difficult administration, but I am ready to do prepayment to take away the risk at least.

Sure, if you can afford it and don't want to bother with DIY, it is the way to go, no doubt.

We acheive figures of less than $30 (all parts including 5m USB cable) for the DIY, this was for group buying to build 100 probes.

A DIYer should probably be around $50 assuming he have the tools to do it's own PCB and PIC programmer... If you add those, you're better of going commercial ;)

Parts are pretty easy to find, let me know what cause you troubles... I assume Farnell refs are internationnal but I haven't checked. The box ($1 !) could be hard to find (outside France) but any other could do it...

Surface mounts sensors are mandatory because they aren't available in other form :D

I have hand build about 20 probes (prototypes and finals) and they are no brainer.

--Patrice

What PCB programmer you are reffering to? I thought we only need to program microchip, which is marked as PIC. What am I missing here?

just got the new diplay lt today and I tell you a lot faster than my previous spyder. I do have 1 question for display lt user do u use crt mode or lcd mode I am trying to calibrate a plasma tv. also is it necessary to attach the ambilight cover on the sensor?

I use LCD mode on my plasma. I think CRT mode is only for old fashioned CRT didplays as it has something to do with locking on to the sync rate. Pretty much never use CRT is what I've been reading.

Do not attach the white plastic cover. (It is useful for storage however)

Rich

I don't have a plasma, so TIFWIW ;), but technically a plasma screen is like a zillion little CRT's, it emits light from electrically stimulated phosphers. CRT's and Plasma's are the only current display technologies that actually emit light.... so my inclination would be to use CRT mode on a plasma set....

:cool:

Hi,

What PCB programmer you are reffering to? I thought we only need to program microchip, which is marked as PIC. What am I missing here?That is because of my bad english...

You should read (note the commas):

DIYer should probably be around $50, assuming he have the tools to do it's own PCB, and PIC programmer...

You need a way to build the PCB.
and you need a PIC (JDM is fine) programmer.

--Patrice

Why build the colorimeter, when you can buy a spyder 2 express for like $60.
Is the HCFR meter that much better?
Also do you leave the bottom cover on the spyder when you use it..looks like a blue filter?

Thanks
RayJr

A blue filter should block infra red, correct? My hunch says you would need to use it with any display that can emit IR light. Any rear projection set that uses a high insensity lamp or a plasma screen. Experts please correct me if I'm wrong.

I have to say that I've been through 1000's of pages on this forum through the years but I'm just feeling especially lazy today (and I don't know that I've ever done this) but would someone be so kind as to point me to the "Instructions for Dummies" for using this software?

I just ordered a Spyder2. I've put it off and put it off but with 3 projectors, 2 plasmas, a professionally calibrated DLP and an LCD, I need to do something. I'm hoping this is my answer.

Thanks in advance guys. This software really looks great (I've been bouncing in and out of this thread reading since its inception).

-bob

Well, there is one in french and one in spanish.... Here is a brief list of steps from Richl:

1) Adjust brightness/contrast
2) Adjust color/tint
3) Perform grayscale, primary/secondary colors reading (HCFR)
4) Adjust primary/secondary colors if display/projector has controls to do so.
Use CIE chart to see progress (of course you need to re-measure P/S colors
first).
5) Adjust gamma if needed to 2.2.
6) Calibrate grayscale, adjusting R/B offset/cut (brightness) and gain/drive
(contrast) as close as possible to coordinates of x=0.3127, y=0.3290 (D65). Use
Green as your reference, adjust only if necessary. Keep an eye on Delta E as you
want it under 4, as close as possible to 0.
7) Recheck step 1, 2 and gamma.

This may be helpful: http://geekwithfamily.com/2006/12/16/home-theater/howto-hcfr-free-video-projector-calibration-software/

I can send you my "best of" selection from this thread in MSWord format, but it is 61 pages long! :D

My hunch says you would need to use it with any display that can emit IR light. Any rear projection set that uses a high insensity lamp or a plasma screen.

not a plasma, it is emmisive like a CRT....

:cool:

Well, there is one in french and one in spanish.... Here is a brief list of steps from Richl:

1) Adjust brightness/contrast
2) Adjust color/tint
3) Perform grayscale, primary/secondary colors reading (HCFR)
4) Adjust primary/secondary colors if display/projector has controls to do so.
Use CIE chart to see progress (of course you need to re-measure P/S colors
first).
5) Adjust gamma if needed to 2.2.
6) Calibrate grayscale, adjusting R/B offset/cut (brightness) and gain/drive
(contrast) as close as possible to coordinates of x=0.3127, y=0.3290 (D65). Use
Green as your reference, adjust only if necessary. Keep an eye on Delta E as you
want it under 4, as close as possible to 0.
7) Recheck step 1, 2 and gamma.

This may be helpful: http://geekwithfamily.com/2006/12/16/home-theater/howto-hcfr-free-video-projector-calibration-software/

I can send you my "best of" selection from this thread in MSWord format, but it is 61 pages long! :D

I think you better do step 5 before step 1 because adjusting your display gamma setting will change brightness/contrast.

Also, when doing the grayscale, I start by using only two test levels (20% and 80%). Turn on continuous measurement mode and go to the RGB tracking plot.

Here's what I do:

In continuous meas mode.

(1) Set input to 20% windowed gray pattern.
(2) adjust red bias until red trace is on top of green on the plot.
(3) adjust blue bias until blue trace is on top of green on the plot.
(4) repeat 2 and 3 as necessary to get all three colors at the same value. You will also see the lowest deltaE now.

(5) Set input to 80% windowed gray pattern.
(6) adjust red gain until red trace is on top of green on the plot.
(7) adjust blue gain until blue trace is on top of green on the plot.
(8) repeat 6 and 7 as necessary to get all three colors at the same value.

(9) set input back to 20% and repeat steps 1-4.
(10)set input back to 80% and repeat steps 5-8.

(11) repeat 9-10 until there is no longer any need for adjustment.

You now have basic grayscale calibrated. You can move to other input levels to see if you can tweek a bit more now.

Now plot the other graphs and take a look

-Rich

I am not sure where you can buy Spyder 2 Express for 60$. Here in Norway, where I live it costs from 150$ and above.... Therefore I would prefer to build one, but programming that chip.... Can'd find the way to do it.

How is it possible to program with programmers having less than 28 holes??? JDM has also less than 28 holes, right? and Laris says it would work...

I have this programmer:
http://www.adteknik.se/english/circuits.htm

Would this work somehow?

Another one of those secret clubs, I see. LOL Like CIH. Pulling gnats' teeth. :)

Thanks for the posts and the basic how-to. It's hard for me to believe that for something as elaborate as this software is, that someone hasn't done and "Idiot's Guide" or something.

Maybe it's because I haven't got my Spyder yet so I don't understand but everything I've read over the years says using this stuff is complicated as hell.

Just doesn't seem like 12-14 bullets in a list are going to help someone like me too much. Maybe I just don't get it and maybe I will once I get my colorimeter. I guess we'll see. :)

Another one of those secret clubs, I see. LOL Like CIH. Pulling gnats' teeth. :)

Thanks for the posts and the basic how-to. It's hard for me to believe that for something as elaborate as this software is, that someone hasn't done and "Idiot's Guide" or something.

Maybe it's because I haven't got my Spyder yet so I don't understand but everything I've read over the years says using this stuff is complicated as hell.

Just doesn't seem like 12-14 bullets in a list are going to help someone like me too much. Maybe I just don't get it and maybe I will once I get my colorimeter. I guess we'll see. :)

Well, we are treading where no user was meant to go after all. Most knowledge here comes from word of mouth. Free outstanding SW like HCFR is unusual and gives the impression that somebody ought to be documenting this stuff only because it's so polished. Just got to ask questions for now and read.

Well, we are treading where no user was meant to go after all. Most knowledge here comes from word of mouth. Free outstanding SW like HCFR is unusual and gives the impression that somebody ought to be documenting this stuff only because it's so polished. Just got to ask questions for now and read.

Well, anyone who has written software knows that the greatest program in the world is undone by the lack of a user's guide. That's why good practice is to write the guide as you develop the software.

What's amazing to me is that even if you read this forum there's no real explanation of what steps to take. In the list that is quoted above, there's no mention whatsoever of what steps are done with what tools. It's as though folks want to keep this arcane.

Can I do contrast/brightness with HCFR, or do I have to eyeball it? What about color/tone? Why? If I use AVIA/DVE, what screens do I use and/or what limitations are involved? If HCFR is just for grayscale, then why doesn't someone come out and say this in big bold letters? If it can do more, then what are they? What buttons do I click in HCFR to do the kinds of things listed in the standard calibration steps?

Here is were I got my spyder 2 express for $62.99
http://www.amazon.com/ColorVision-Spyder2-Express-Win-Mac/dp/B000ES4PYU
Hope this helps.

Later
RayJr

Yep, but unfortunately : "Shipping: Item can be shipped only within the U.S."

Too bad...

--Patrice

I just got my Spyder2. The manual says to use the blue filter for LCD displays and remove it for CRT displays. CRT displays do not emit IR do they?

Well, anyone who has written software knows that the greatest program in the world is undone by the lack of a user's guide. That's why good practice is to write the guide as you develop the software.

What's amazing to me is that even if you read this forum there's no real explanation of what steps to take. In the list that is quoted above, there's no mention whatsoever of what steps are done with what tools. It's as though folks want to keep this arcane.

Can I do contrast/brightness with HCFR, or do I have to eyeball it? What about color/tone? Why? If I use AVIA/DVE, what screens do I use and/or what limitations are involved? If HCFR is just for grayscale, then why doesn't someone come out and say this in big bold letters? If it can do more, then what are they? What buttons do I click in HCFR to do the kinds of things listed in the standard calibration steps?

Along the same lines (though not specific to this thread), is CIE adjustment. What does one shoot for if you have a display with primaries that cannot ever hit the reference points? If one primary is off, do you still target the reference for the other two? Where should the secondaries go in these cases? Very few displays can nail each reference right on.

There are discussions splattered over this and other threads, but after months of reading thousands of thread, I have found no one place to describe how to adjust, what to look for, and the acceptable tradeoffs. And what info there is seems to be more focused on gray scale adjustment then on color management - both are important especially as more displays come with 6 color CM adjustments.

Anyone want to tackle an adjustment primer?

-Dave

I use LCD mode on my plasma. I think CRT mode is only for old fashioned CRT didplays as it has something to do with locking on to the sync rate. Pretty much never use CRT is what I've been reading.

Do not attach the white plastic cover. (It is useful for storage however)

Rich


Rich,
I haven't tried ColorHCFR yet, but intuitively, I would think that you would want to use the CRT mode. Plasma displays are emissive and use phosphors, very similar to CRTs. LCDs use a backlight and polarizers that make the detected light very directional; using the LCD mode may cause significant errors.

-Steve

With all this talk of CRT, I have to ask, are you specifically talking about old CRT's, or does this include RPTV's that use CRT guns as well? Reason being, I could only get readings on low IRE's using a Spyder2 in CRT mode (with the filter off), and not LCD, but I've been told previously to use LCD. As someone recently mentioned, it would, in fact, get a nice explanation as to what the software does differently in these modes, as well as if anyone whom has a good understanding of the Spyder 2 explain what the filter is for, and when to use it.

Anyone want to tackle an adjustment primer?

-Dave


Well, it's not a primer (and I'm not qualified to create one even if I had the time) but here is a discussion on this topic by some of the most knowledgeable people using these forums: http://www.avsforum.com/avs-vb/showthread.php?t=586691 It says it's for dummies, but ranges pretty far afield.

Also, I found this post useful for contrast and brightness settings: http://www.avsforum.com/avs-vb/showthread.php?t=585527

My understanding is that, in general, you use a disk (GetGray, Avia, DVE) to set brightness/contrast and color/tint. You use Color HCFR to set gamma, greyscale and primary/secondary colors if you happen to have the controls to do it.

I think you better do step 5 before step 1 because adjusting your display gamma setting will change brightness/contrast.

Also, when doing the grayscale, I start by using only two test levels (20% and 80%). Turn on continuous measurement mode and go to the RGB tracking plot.

Here's what I do:

In continuous meas mode.

(1) Set input to 20% windowed gray pattern.
(2) adjust red bias until red trace is on top of green on the plot.
(3) adjust blue bias until blue trace is on top of green on the plot.
(4) repeat 2 and 3 as necessary to get all three colors at the same value. You will also see the lowest deltaE now.

(5) Set input to 80% windowed gray pattern.
(6) adjust red gain until red trace is on top of green on the plot.
(7) adjust blue gain until blue trace is on top of green on the plot.
(8) repeat 6 and 7 as necessary to get all three colors at the same value.

(9) set input back to 20% and repeat steps 1-4.
(10)set input back to 80% and repeat steps 5-8.

(11) repeat 9-10 until there is no longer any need for adjustment.

You now have basic grayscale calibrated. You can move to other input levels to see if you can tweek a bit more now.

Now plot the other graphs and take a look

-Rich

Just so you don't feel unappreciated, I found your post to be very useful. :D

Along the same lines (though not specific to this thread), is CIE adjustment. What does one shoot for if you have a display with primaries that cannot ever hit the reference points? If one primary is off, do you still target the reference for the other two? Where should the secondaries go in these cases? Very few displays can nail each reference right on.

There are discussions splattered over this and other threads, but after months of reading thousands of thread, I have found no one place to describe how to adjust, what to look for, and the acceptable tradeoffs. And what info there is seems to be more focused on gray scale adjustment then on color management - both are important especially as more displays come with 6 color CM adjustments.

Well, there is nothing you can do really, a very very few number of projectors have those adjustements.
If you don't have then the CIE simply display your set Gammut and that is you have to deal with it and ajust all others parms to have do to the best you can.

--Patrice

Rich,
I haven't tried ColorHCFR yet, but intuitively, I would think that you would want to use the CRT mode. Plasma displays are emissive and use phosphors, very similar to CRTs. LCDs use a backlight and polarizers that make the detected light very directional; using the LCD mode may cause significant errors.

-Steve

I think it's been stated pretty clearly by the authors that you use LCD mode for plasma (at least on the eye one display lt probe)

Patrice: please correct if I'm wrong.

Thanks
Rich

Well, there is nothing you can do really, a very very few number of projectors have those adjustements.
If you don't have then the CIE simply display your set Gammut and that is you have to deal with it and ajust all others parms to have do to the best you can.

--Patrice

Nah - you missed my point. If I can adjust color management (either through available adjustments or via service menu), the question what should be the goal if there is no way to hit all the points of the CIE triangle? I can adjust everything with my PRO1140, but some points simply cannot be hit due to plasma limitiations (i.e. green).

ColorHCFR software is great and obviously took a great deal of effort. The next step, it seems to me, is to offer detailed tuning tips and goals given the less than perfect displays that most of us own. I've read a dozen times the link that Orwellflash provided, along with many others - and all it does is raise more questions for me.

I promise not to divulge any ISF trade secrets!

-Dave

With all this talk of CRT, I have to ask, are you specifically talking about old CRT's, or does this include RPTV's that use CRT guns as well? Reason being, I could only get readings on low IRE's using a Spyder2 in CRT mode (with the filter off), and not LCD, but I've been told previously to use LCD. As someone recently mentioned, it would, in fact, get a nice explanation as to what the software does differently in these modes, as well as if anyone whom has a good understanding of the Spyder 2 explain what the filter is for, and when to use it.
I'd suggest looking at the Calman forums since they have done countless testing with it, and also worked closely w/ the manufacturer. Start here (http://www.datapopuli.com/forums/viewtopic.php?t=350) and of course look at the other threads as well ;) .

Dbower, I see what you mean now, and yes we know these tips and bits (as well as a startup guide) are highly demanded, just don't have time to improve software, do DVD disk, documentation, etc... We just need time :D
(There is very valuable threads on the French site, again need time to translate...)

I don't think there is any ISF secrets, and can assure you all knowkedge people put in software (and posts !) are shared...

As for SpyderII lcd/crt, I have no idea, I'm not using it. I will ask the guys that use it... and as Jose spot, there is other very valuable information sources.

--Patrice

Nah - you missed my point. If I can adjust color management (either through available adjustments or via service menu), the question what should be the goal if there is no way to hit all the points of the CIE triangle? I can adjust everything with my PRO1140, but some points simply cannot be hit due to plasma limitiations (i.e. green).

-Dave
Your goal should be to hit all points but if you can't then try to bring them as close as possible. Not much else to it. As far as your green goes, that's pretty normal on digital displays, nothing you can do about it.

any suggestions on how I can fix red on my plasma

You are pretty good, what's wrong with your red ?
(It would be easier if you can also do primaries and secondaries measures)

--Patrice

I get spike on 10 ire and 20 to 40 not quite perfect yet.

It's pretty difficult to get an accurate reading at that low IRE more than anything. Also, try bringing brightness down a bit to bring the gamma curve closer to reference(20-50IRE). Then bring red brightness up a bit if still needed.

It's pretty difficult to get an accurate reading at that low IRE more than anything. Also, try bringing brightness down a bit to bring the gamma curve closer to reference(20-50IRE). Then bring red brightness up a bit if still needed.

I will try that later and thanks for the advice. I do not know if you remember but before I was getting spikes from 50 to 90 ire on luminance I have found out the problem for you lg plasma owners I just turned off the xd engine feature and that fix the problem.

jazyfelon -

Both Laric and HDHolic are correct..I too do not see a problem with Red, its actually in whole, as HDHolic mention, try lowering your brightness, just a click or two(at most) and do a read, I think that is the only thing you will need to do.

For your greyscale, 20 IRE has a Delta E of 5, then 30IRE has 3, then 40IRE has 4, you might see that 20IRE and 40IRE seems very much in line with each other and there is a dip at 30IRE to the line of 6500K, you might want to tweak 20IRE and bring it down to delta E 3-4, this will probably bring 30IRE at delta e 4 or 5 but 40IRE might go down to 2-3, so you might be better hitting the two and sacrifice the 30IRE but not much..My Mits 3000 behaves in the manner, where 20 and 40 seems to be inline but 30 moves opposite from them so I sacrifice one for 2..this is being a little anal and you probably wouldnt even notice that much of a difference but nevertheless - save this settings you currently have and see if it behaves like i think..Even so, dropping brightness will probably require you to tweak a little the greyscale..

one more question regarding color and tind I adjust my color and tint by using dve color filter. is this the proper way or should I adjust it by trying to hit the cie triangle and adjusting color and tint that way.

Yes, that's the proper way. There's no way of adjusting color/tint with cie chart, not made for that.

You don't use tint/color to adjust cie position of the primary. you can use the tint to move the secondary around the triangle clockwise or counterclock wise.
Read my post #1781 about how one can calibrate color control usiing the program instead of eyeballing with color filters.

Regarding tint: if ColorHCFR reports the secondaries are right on the reference triangle (and white point is spot on), then should the color filter test reflect a near perfect adjustment of tint? Conversely, if the color filter test shows tint needs to be, say, 6 clicks to red, then could you use ColorHCFR to adjust the secondaries to allow the tint control to be at 0?

-Dave

Hi! I need an expert !! (Laric ..richlo...HDholic... anyone).


I've got two probes: Spyder2 and DTP94. This week, I made tests to measure speed and consistent of the results. I made four different measures with the Spyder2: (check the Spyder2 file)

"nada"= uncheck Spyder2 parameters
"extendend"= only check "extended..." Spyder2 parameters.
.......


I compared these numbers and I found the same results. Well, there are some different, but they are very little.


I repeated the same with DTP94. Now I only took two measures:
"nada"= uncheck "avergare..."
"average"= check this option.


I compared numbers (DTP) and I found right results.

My problem appears when I compare Spyder2 results with DTP results. All is ok, (well, "inside" the error of Spyder+DTP) except gamma: I always get different gamma: Spyder2 result are always about 0,2 lower.


I studied the "numbers".. and I saw there are so much different in "Y" coordinate. Sometimes, the double!!. I was looking for spec of probes and I got this in HCFR forum:

http://www.homecinema-fr.com/Colorimetre/images_forum/Mlill/200611/12-specs-sondes.png

Error of luminace for Spyder = 1,5
Error of luminance for DTP94= 4

I never should have a error more great of 5,5% (=1,5+4), but I sometimes have a 100% of error.


I've repeated the test in different conditions: I always get a the Spyder2 gamma more lower than DTP94 gamma ( about 0,2) :confused:


The tests were with a LCD monitor . I've repeated a lot. I get the same results.


I ask to me if someone in HCFR team has got these two probes and made the same test...?

This happens to me to have two colorimeters!! :eek:


Regards

Hi,

We have many people having the those two probes... I have asked them to look at this...

The tests were with a LCD monitor . I've repeated a lot. I get the same results.
Anyway, that is the worst case for SpyderII, one of our expert has conduct some tests recently, they are on a thread on french forum...

Here is a try to compare some probes measure scope using patterns of different sizes...

http://www.homecinema-fr.com/Colorimetre/images_forum/Mlill/200611/18-champ-sondes.png

As you can see, facing screen, probes having a small angle have more stable measures response.

LCDs usually emit a lot of side light. If proble viewing angle is wide this disturb measures... S2 is clearly in this case... (as well as our own probe).

You black/dark level measure are clearly too high with S2, hence your results (my guess...)

--Patrice

How do you do continuous measurements with the HCFR software?

Also, how does one increase gamma to 2.2 if it is too low?

You black/dark level measure are clearly too high with S2, hence your results (my guess...)
Yeah, I'd agree with that ;) .

Continuous measurements are started with the "Measures" box. Just hit the green arrow.

As to gamma, you need to adjust your brightness/contrast/cuts/gains. The interaction of those, plus any explicit gamma controls on your set will change the gamma that you measure.

Thanks very much for your answers (public and private!!) , laric and HDholic.. I didn't know what I could do with those results

they are on a thread on french forum...
I do apologize, but I don't understand french...


All the others result are ok

Regards

PD: The curious thing are the specs: theoretical Spyder spec about luminance are better than DTP and these measures are very important to get the gamma (number and shape)

PDD: Next, I have to compare DTP vs HCFR probe . I'm really ver bad guy:D

As you can see, facing screen, probes having a small angle have more stable measures response.

I wonder if it would improve things to mount the probe at the end of a tube of appropriate diameter to fit the face of the probe spacing the probe back a bit from the screen.... the longer the tube, the lower the reading, but maybe more importantly, the less off angle light :D

:cool:

Patrice and gang,

Not sure if you're following the following thread, but I'd be curious if you guys have seen anything similar.

http://www.avsforum.com/avs-vb/showthread.php?p=9674823&&#post9674823

I wonder if it would improve things to mount the probe at the end of a tube of appropriate diameter to fit the face of the probe spacing the probe back a bit from the screen.... the longer the tube, the lower the reading, but maybe more importantly, the less off angle light :D

:cool:


I have implemented something similar to that with very encouraging results.
I am measuring with a DTP94 from just above the pj lens at about 10ft from the screen and readings are very consistent. My "tube" is a truncated pyramid about 4" long and the shape of the base keeps the 16:9 ratios of the screen shape: i.e. the sensors see approx. the full surface of the screen only. All other light reflected from around the room does not get measured (in a direct way i.e.)

Yes, have also experience this, only point anoying being the large impact on sensitivity (lower readings)...

--Patrice

I have a question.
Is there still a large need for programming the PIC 18F2550?
If so, I may be able to help..I just found an old chip programmer that I had from another project...have to dig up the instructions...but I think it will do PIC 18F2550.

Hope it helps.
RayJr

I'm sorry. I know this question has been asked before but I couldn't find the answer.

I'm using the PAL version of the HCFR DVD (I don't believe an NTSC version exists yet) on a US version of the Xbox 360. I don't know if this player officially supports PAL, but the DVD seems to work fine.

The problem (at least, I think it's a problem) is that the test image is very small - perhaps 4"x6" - and it's difficult to align my probe to hit that exact portion of the screen. Should the image be so small?

Supposedly it's suppose to be that size...I think it's wrong since they should cover 50% of the display. I highly recommend GetGray. There's also another thread that has HD patterns that you can use with the 360.

They are 10% in size because you chose the "video projector" option on top menu, you can get 100% by chosing the other option.

BTW, 50% may be easier for most people probe placement, but that makes no difference in measures... (well actually having it small helps on CRTs and related ;))

--Patrice

Yes, have also experience this, only point anoying being the large impact on sensitivity (lower readings)...

--Patrice

Hi, Patrice,
How low were your readings? The lowest I have obtained at 0IRE has been X=0.030,
Y=0.034 and Z=0.035 and they have fluctuated up to X=0.119, Y=0.121 and Z=0.141.
All readings obtained 10ft from the screen.
What kind of 0IRE reading levels are to be expected/normally obtained when reading off the screen reflection? What would be the minimum usable level at 0IRE?
Thanks.
Fermin

They are 10% in size because you chose the "video projector" option on top menu, you can get 100% by chosing the other option.

BTW, 50% may be easier for most people probe placement, but that makes no difference in measures... (well actually having it small helps on CRTs and related ;))

--Patrice

Ah-ha! I don't speak French and just guess what the words meant. I assumed it meant "rear projection", which is what my set is. Thank you!

Found two chips in local store matching HCFR description, which one is right?
PIC18F2550-I/SO
PIC18F2550-I/SP
The only difference in description I found was:
16Kx16 Flash 23I/O 48MHz SO28
vs 16Kx16 Flash 23I/O 48MHz SDIP28
Or would both of these chips work???

Found two chips in local store matching HCFR description, which one is right?
PIC18F2550-I/SO
PIC18F2550-I/SP
The only difference in description I found was:
16Kx16 Flash 23I/O 48MHz SO28
vs 16Kx16 Flash 23I/O 48MHz SDIP28
Or would both of these chips work???

look at this page:
http://es.farnell.com/jsp/search2/browse.jsp?N=500017&Ntk=gensearch_001&Ntt=PIC18F2550&Ntx=&_requestid=768265

Your first is a CMS, your second NO!! (the second is ok)

If you want , I can give to you a list with the "code" of all ( farnell code )

I did four HCFR probes

Regards

Thanks, I found the list on French HCFR pages, but few items there does not contain farnell codes... Would be great to get complete list if possible.

Not sure what CMS stands for, but I understand that 16Kx16 Flash 23I/O 48MHz SDIP28 is the one to buy then, right?

Yes,

Can you tell me what item don't have farnell code ?
I was convienced I put them on all...

--Patrice

let me say that I am newbie. Perhaps those parts are obvious for most of you, but as a newbie I would try to find exact match using farnell code. It was missing on following parts:
R1 - 3.3K
R2 - 150R
R3 - 2.2K
R8, R9 - 27R
Q1 - NPN - BC547

Would be great to find farnell code for box as well, as it is difficult to find a match :) Seriuosly, it would be great to find preprogrammed chip and box to buy, all the rest parts seems to be available around the globe.

let me say that I am newbie. Perhaps those parts are obvious for most of you, but as a newbie I would try to find exact match using farnell code. It was missing on following parts:
R1 - 3.3K
R2 - 150R
R3 - 2.2K
R8, R9 - 27R
Q1 - NPN - BC547

Would be great to find farnell code for box as well, as it is difficult to find a match :) Seriuosly, it would be great to find preprogrammed chip and box to buy, all the rest parts seems to be available around the globe.

CMS= Component montage superficial ( the same as TAOS sensor )

R1 - 3.3K = 9476806
R2 - 150R = 9476385
R3 - 2.2K = 9476652
R8, R9 = 9476644
Q1 - NPN - BC547 = 9207546


Farnell number!!

I did 4 probes with those codes !!


Regards

Yes,

Can you tell me what item don't have farnell code ?
I was convienced I put them on all...

--Patrice
Laric When I made my probes, I had to find some codes farnell because they weren't in the list..

.....but I knew a few ( very very little ) of electronics I didn't have any problem to find the missing components


Regards

great, now I can locate all parts, thanks Whisper2! Now it is only to figure out how to program that chip (or find preprogrammed one).

I first thought that my programmer would not do the job since it does not have enough holes:
http://www.adteknik.se/english/circuits.htm

However JDM programmer, that laric mentioned couple of times in this thread also seem to have less than 28 holes. How can JDM programmer be used then?

ColorHCFR used on AudioHolics.com Panasonic 1080P PDP review! Here (http://www.audioholics.com/productreviews/avhardware/panasonic-1080p-plasma2.php) .

Word spreading on the "Little Program That Could"... :D

great, now I can locate all parts, thanks Whisper2! Now it is only to figure out how to program that chip (or find preprogrammed one).

I first thought that my programmer would not do the job since it does not have enough holes:
http://www.adteknik.se/english/circuits.htm

However JDM programmer, that laric mentioned couple of times in this thread also seem to have less than 28 holes. How can JDM programmer be used then?

Look at this link ...http://www.todotarjetas.com/ , I used T-20 II for my probes ( I didn't buy it, a friend of me had one )

I don't know nothing about the JDM programmer


ColorHCFR used on AudioHolics.com Panasonic 1080P PDP review! Here .

Word spreading on the "Little Program That Could"...


WOW.. thanks for the link...


Regards

ajocius, i did not put the refs on resistor/transistor as these are pretty common componants ! I will update the pages if that helps...

JDM programmers need to be adapt so you can fit the 2550, but this works fine.

--Patrice

ok, I got it: JDM needs to be adapted to fit 2550, other programers might work as well. Just need to find someone having that programmer, seems wrong to buy programmer for one DIY project... Or hope to find somebody able to sell preprogrammed chip. Seems like I am not the only one searching for it...

It seems that i got the gray scale pretty flat, but the colors are off, especially green, yellow and magenta. magenta isn't even on the chart pre calibration. Is there any way for me to bring the colors into line? This is with a Mitsubishi HD1000U DLP front pojector. Shifting color doesn't seem to effect the CIE chart. The gamma is also a little off. I have the option of setting up a user gamma with the abitlity to change high medium and low I believe. Would setting this up allow me to reach 2.2 gamma?
Can somebody please explain how to use constant measuremnt? I had to run the entire 0-100 grayscale measurement everytime I made an adjustment. Very time comsuming.

Any help would be much appreciated.
Great software french guys! :)
Thanks,
Jason

It seems that i got the gray scale pretty flat, but the colors are off, especially green, yellow and magenta. magenta isn't even on the chart pre calibration. Is there any way for me to bring the colors into line? This is with a Mitsubishi HD1000U DLP front pojector. Shifting color doesn't seem to effect the CIE chart. The gamma is also a little off. I have the option of setting up a user gamma with the abitlity to change high medium and low I believe. Would setting this up allow me to reach 2.2 gamma?
Can somebody please explain how to use constant measuremnt? I had to run the entire 0-100 grayscale measurement everytime I made an adjustment. Very time comsuming.

Any help would be much appreciated.
Great software french guys! :)
Thanks,
Jason


Jason -

You need to recalibrate, your HD1000 acts alot like the Mits 3000U which I have. Let me give you some tips

1s - DO NOT USE any RED contrast or Brightness for greyscale, use Green and Blue only, right now your clipping RED because its already MAX out of the box. If you reduce Green you will see Red move up (if your looking at the RGB color Bars). So leave it at ZERO default on both contrast and brightness. Right now, set everything back to zero and restart

2nd. For Gamma - you will see that once you get the above right, that your gamma will track better, then you can fine tweak it with using the USER/GAMMA Cinema.

3rd. After you do a full run of, use the continuos mode to fine tweak it. What you want to do is bring up 30IRE and then pause the pattern, then look below and click the GREEN arrow button in the MEASURE box, this will continously read your pattern. Then highlight your 30IRE results, and click on icon DISPLAY MEASURES INFORMATION and make your adjustment on your set and you will see the color bars move up and down as you make changes, you want 100% on all of it (or as close to it) Disregard once again using RED Contrast and Brightness in the Mits.

4. Recheck contrast and brightness.

5. Run your primary and secondary and it might come alot closer this time around..

6...Oops and for now put your regular contrats and brightness at ZERO

Jason, I'm projecting on a FG Laminate Screen, I fool with the calibrations way too much, but this is my last go. I'm using a Panny XP30 if that helps. Love my PJ replaced a X1. I've adjusted tint back and forth makes some difference. Got a lot of help with Rich and others on this site.

Jason, I'm projecting on a FG Laminate Screen, I fool with the calibrations way too much, but this is my last go. I'm using a Panny XP30 if that helps. Love my PJ replaced a X1. I've adjusted tint back and forth makes some difference. Got a lot of help with Rich and others on this site.


Very well done...Im not sure which gamma setting your using, either Camera at 2.4ish or Display Gamma with black compensation (2.12ish)...Im guessing your projector will use the 2nd gamma I just noted. Your gamma actually is very tight - and looks really good with the exception that it might need to bring it up to 2.2ish..A click here or there of both brightness and contrast (less of it) will probably bring it up..Otherwise, it looks really good calibration number wise..Oopps..looks like you used REGRESSION

How does it look to you?

If your content..please be done with it and take any advice I just mentioned for something down the road when you want to recalibrate in a couple of months.

Txs Rich, color and image look good. Once I realized that this PJ had green push, I was able to adjust the Red and Blue on the CIE chart fairly close and then the tint helped with the secondaries. I was trying not to touch the green as it was recommended but it was necessary to get close. There was another HD1000 calibration posted which was dead on, not sure how that was achieved with the Green push. This replaced my X1, purchased for under $900, got a free bulb, sold the X1 for $250 and discovered HCFR. Needless to say I'm very happy with the purchase and this great software.

Where can I buy the HCFR colorimeter that they recomends?; Im not good enough to build one by myself and maybe someone could help here. Thnaks in advance

You can't buy it--only available through self-build. If you want to buy a probe, current recommendations, as I understand them, are to get an Eye One Display 2 or LT (same specifications).

Here is a good review of several colorimeters (in french-- use babelfish if you need to): http://www.homecinema-fr.com/forum/viewtopic.php?t=29838162

Also good test of Eye One Display2 here: http://www.avsforum.com/avs-vb/showthread.php?t=770737&page=1&pp=30

P.S. Display 2/LT seems to be inaccurate adjusting the gray scale of plasmas, however.

Ok. I'm confused. I'm trying to calibrate my Hitachi LCD RPTV with HCFR and an Eye One. I've set the brightness/contrast with Getgray (great disk; just the moving brightness and contrast screens are worth the price of admission) and I'm addressing the CIE chart. My TV has phase and gain controls for the primaries and secondaries of the color decoder, so I've used that and the 100% color windows to fix my color space. I've been able to get all primary and secondary colors pretty much spot on. But when I check my results with the color/tint display and the DVE blue filter, it's nowhere close. I've tried using the 75% color windows as well as the 100%, and I've tried both rec 601 and 709. 601 looks better, but nothing passes the color/tint test. The most obvious thing is that I have to turn blue down really low.

Isn't it supposed to be at least close even before I attempt to clean up my gray scale? What don't I understand here?
Roy

Try posting your question to this thread, which deals with ColorHCFR calibration:
http://www.avsforum.com/avs-vb/showthread.php?t=786966&page=9&pp=30

Color and tint settings are not set by adjusting color space. Use the GetGray patterns and your filter with your display's color and tint controls for that. I have heard that color space adjustment can make gray scale adjustments easier, but many displays don't allow you to adjust color space, and gray scale adjustment needs to be done regardless of color space issues. I'm new to calibration myself, so you can get more knowledgeable advice on the thread above.

There was also some discussion that the various blue filters are not always very precise. They are certainly better than nothing, but cannot be considered as true references in all cases.

Patrice and all - how is version 1.3 coming along? Any feel for when it may be released?

If I recall correctly someone had posted a feature request for helping set gamma. Basically it was a feature that showed what the target brightness levels were so that one didn't have to calculate this by hand, and can use their gamma controls on the display to set the proper brightness at each IRE step with the help of HCFR.

Is that feature going to be in v1.3? That would be very powerful.

Thanks!

Hi,

So far we have defined what will be in v1.3 and we are currently working on the implementation as well as quite some ergonomic/presentation changes...

It is a bit too early to plan a release date, we hope to have something by and of March... (no garantee here :D)

As for the very feature you mention, I think it is in todo list, not sure thought as it is not my part of things ;)

--Patrice

Hi,

So far we have defined what will be in v1.3 and we are currently working on the implementation as well as quite some ergonomic/presentation changes...

It is a bit too early to plan a release date, we hope to have something by and of March... (no garantee here :D)

As for the very feature you mention, I think it is in todo list, not sure thought as it is not my part of things ;)

--Patrice

Wow. Sounds like some pretty major changes are coming - looking forward to that. Sounds like more of a 2.0 release than a minor version update.

Thanks Orwellflash. I'll do my homework then.

Hi,

So far we have defined what will be in v1.3 and we are currently working on the implementation as well as quite some ergonomic/presentation changes...

It is a bit too early to plan a release date, we hope to have something by and of March... (no garantee here :D)

As for the very feature you mention, I think it is in todo list, not sure thought as it is not my part of things ;)

--Patrice
Patrice,

Has anyone reported an issue where the data points on the CIE Diagram are no longer visible? This has occurred on my Dell notebook, where the CIE reference and measured triangles are visible, and the values display when I hover over the points with the cursor, but the measured dots and reference square points do not appear. On my other PC, everything appears correctly.

I have right-clicked on the diagram and checked all the correct options, and uninstalled/reinstalled ColorHCFR, but this has not fixed the problem.

-John

Patrice,

Has anyone reported an issue where the data points on the CIE Diagram are no longer visible? This has occurred on my Dell notebook, where the CIE reference and measured triangles are visible, and the values display when I hover over the points with the cursor, but the measured dots and reference square points do not appear. On my other PC, everything appears correctly.

I have right-clicked on the diagram and checked all the correct options, and uninstalled/reinstalled ColorHCFR, but this has not fixed the problem.

-John

This happens on my HP laptop but not my desktop. It corrects itself from time to time but more often than not the data points (primary, secondary, and grayscale) are not visible in the CIE chart.

sounds like video driver issues....

Same thing happened to me on a borrowed laptop. It should have had the latest drivers, because it was a fresh XP install and update. Although, it was a VERY old laptop. So it could just be the drivers weren't updated. It wasn't a real problem, I just opened the .chc file on my Desktop...

Yes, that was reported, we suspect some specific setup (drivers ?) but haven't been able to both reproduce it in lab and obviously sort it (if related to app).

--Patrice

Yes, that was reported, we suspect some specific setup (drivers ?) but haven't been able to both reproduce it in lab and obviously sort it (if related to app).

--Patrice

I think I found the problem - my laptop video driver (ATI Mobility Radeon 7500) was set for 16 bit color. When I changed it to 32 bit color, all the CIE data points appeared on the CIE diagram!

très simple!

-John

I think I found the problem - my laptop video driver (ATI Mobility Radeon 7500) was set for 16 bit color. When I changed it to 32 bit color, all the CIE point appeared on the CIE diagram!

très simple!

-John

I tried that on the laptop I was using and no go, but it was a VERY old laptop.

Yup that was the problem, changing from 16bit to 32 solved problem. Not sure why my wife would change the settings on me, time for a little chat. :D

I would like to learn about all this; I dont want to have a degree in how to calibrate tvs/proyectors but know and understand the fundamentals behind all this processes, want to know what Im doing in each part of the process and how everything interacts so.... could you give me some names of books that I can read and learn about colors, calibration, etc?. Thanks

If I recall correctly someone had posted a feature request for helping set gamma. Basically it was a feature that showed what the target brightness levels were so that one didn't have to calculate this by hand, and can use their gamma controls on the display to set the proper brightness at each IRE step with the help of HCFR.

Is there a description somewhere of how to do it by hand? That is, given an arbitrary IRE value, how would I figure out what brightness that ought to correspond to?

Is there a description somewhere of how to do it by hand? That is, given an arbitrary IRE value, how would I figure out what brightness that ought to correspond to?

There are some formulas but it has to do with your gamma offset (which I believe you can now specific with this software with one of the gamma options). I'm not solid on the approach so having this in the software will be a big help.

Forgive me, because I'm at square one and swearing a bloody blue streak. I cannot get the driver installed for the Spyder2 meter. I'm following all the instructions provided in HCFR Colorimeter HTML help, but nothing. It says it can't find the necessary software to complete the hardware instalation. Someone please help, I've been f***ing around with this G-d*** thing for four hours and nothing is working. I'm about ready to send it and my laptop through my plasma screen in one high pitched screaming swing. I feel like a dunce.

thank you.

Forgive me, because I'm at square one and swearing a bloody blue streak. I cannot get the driver installed for the Spyder2 meter. I'm following all the instructions provided in HCFR Colorimeter HTML help, but nothing. It says it can't find the necessary software to complete the hardware instalation. Someone please help,


You need to go to the spyder folder where the spyder software is installed, and copy the CVSspyder.dll from there and paste it in the HCFR folder where the executable is at. If you don't have any of the spyders own included software installed, then you will need to do so in order for it to generate the needed CVSspyder.dll file for you that you need to copy from. That should take care of your problem.

I'm not sure what you mean by "where the executable is at." Where is it precisely? Thanks.

That will depend where you installed it...

It is usualy C:\Program Files\Colorimetre_HCFR

--Patrice

Thanks for your help. That took care of it.

some questions on using HCFR .......

can single measurements be taken on the grayscale chart? ie 20% only

can single measurements be taken on the primary / secondary charts? ie red only

how do I determine which gamma calculation to use?

what is the sensor matrix for?

when calibrating an CRT display, using a Spyder2, should the filter be on or off?

I have seen these issues discussed in this post, but no definitive answers. I have found a few user guides for this software, but no complete guides. The help in the software seems to be for an older version as some of the menu settings are different in the software.

thanks

Hello,

You can get a single measure of anything at any time... Just press F7 or the green arow at center bottom...
But, if I try to understand what you may mean, you cannot chose a primary / secondary / grayscale column and redo that very measure... (update values).

To do so, you have to simply do a single measure and then set the tables to edit mode and input the values (from the single/continous measure tab to your target column...)

As for the Gamma stuff, look my initial post here : http://www.avsforum.com/avs-vb/showthread.php?p=9386895&&#post9386895

You can forget the sensor matrix thing...

--Patrice

when calibrating an CRT display, using a Spyder2, should the filter be on or off?



thanks

Filter on with LCD as the display option, forget CRT option and filter off.

Filter on with LCD as the display option, forget CRT option and filter off.

Why? I use CRT mode with filter off on my plasma with excellent results. I know there are issues with the overall calibration accuracy of the s2 but they are built to be used in crt mode and no baffle/filter with crt displays.

Why? I use CRT mode with filter off on my plasma with excellent results. I know there are issues with the overall calibration accuracy of the s2 but they are built to be used in crt mode and no baffle/filter with crt displays.

There has been many review here that will tell you that with the FILTER ON and LCD mode provides better results. CRT mode in the spyder is for DIRECTVIEW CRT only...now if you getting excellent results and you getting a wonderful pix using this method...be free to continue..

For RearProjector CRT users - (using CRT MODE for DTP-94)

I have had the hardest of time calibrating a Mits55807 CRT tv that was professionally calibrated some years ago. I wanted to tweak the greyscale, when I first ran my ColorHCFR for original settings...the greyscale was hovering over/close around 7500 (d75) and I calibrated it at D65 (x=3127 y = 329)...After calibrating...I was never impressed by the looks, the results where by number - fantastic - but the pix always had the rouge look (magenta in a way)..GUESS WHAT..I realize that at least with this Mits (and let me tell you I have a PERFECT COLOR DECODER professionally done by a proMits calibrator)...Calibrating to D75 is what it should be and NOT D65...This was also an issue with a friends of mine who I did their Mits, I just could never get why I was getting poor viewing results, I thought I just didnt know how to use the COLORPERFECT in the Mits..Now I know...its D75 not D65..if any of you are having the same issues with your RearProjector CRT, calibrate to D75 and see your results..I am BEYOND happy now..

There has been many review here that will tell you that with the FILTER ON and LCD mode provides better results. CRT mode in the spyder is for DIRECTVIEW CRT only...now if you getting excellent results and you getting a wonderful pix using this method...be free to continue..

I didn't realize vega509's display was not directview, in that case I agree lcd mode is appropriate. crt mode is for phosphor based spectral distributions only.

I didn't realize vega509's display was not directview, in that case I agree lcd mode is appropriate. crt mode is for phosphor based spectral distributions only.


It may be directview ;) ..Im just ASS-u-ming its not..DAMN ME

Plasmas I believe are phosphor (or act like it), so it may be okay actually in this case to use CRT mode in this regards...

yes, they are very similar in their spectra. Both use the same rare earth elements although the excitation mechanism is completely different. There are two strong peaks in the red that are not present in lcd's and these peaks get cut-off by the external filter, which is why you don't want the filter in place for these displays. I've got some data on this at the top of thread here (http://www.avsforum.com/avs-vb/showthread.php?t=804346)

cheers

Zoyd,
I'm a bit confused by your statement and the data that you linked to. Your data shows that the plasma display has significant red content outside of the human visual response, and that the filter largely removes this. It seems to me that this indicates that the filter SHOULD be used for plasma displays, otherwise the S2 will measure excessive red. :confused:

-Steve

Zoyd,
I'm a bit confused by your statement and the data that you linked to. Your data shows that the plasma display has significant red content outside of the human visual response, and that the filter largely removes this. It seems to me that this indicates that the filter SHOULD be used for plasma displays, otherwise the S2 will measure excessive red. :confused:

-Steve

Depends, what is the bandwidth of the red measuring detector? Will it include this infarred energy?

Depends, what is the bandwidth of the red measuring detector? Will it include this infarred energy?


It's not clear, but this post (http://www.avsforum.com/avs-vb/showthread.php?p=9759024&&#post9759024) by Zoyd in the same thread shows figures from the Spyder patent that indicate that the region in question is not completely filtered by the internal detector filters. I guess we'll never really know until someone with a reference instrument does a comparison with the Spyder2, with and without the external filter.

-Steve

For RearProjector CRT users - (using CRT MODE for DTP-94)

I have had the hardest of time calibrating a Mits55807 CRT tv that was professionally calibrated some years ago. I wanted to tweak the greyscale, when I first ran my ColorHCFR for original settings...the greyscale was hovering over/close around 7500 (d75) and I calibrated it at D65 (x=3127 y = 329)...After calibrating...I was never impressed by the looks, the results where by number - fantastic - but the pix always had the rouge look (magenta in a way)..GUESS WHAT..I realize that at least with this Mits (and let me tell you I have a PERFECT COLOR DECODER professionally done by a proMits calibrator)...Calibrating to D75 is what it should be and NOT D65...This was also an issue with a friends of mine who I did their Mits, I just could never get why I was getting poor viewing results, I thought I just didnt know how to use the COLORPERFECT in the Mits..Now I know...its D75 not D65..if any of you are having the same issues with your RearProjector CRT, calibrate to D75 and see your results..I am BEYOND happy now..

great info richlo, I too was somewhat dissapointed after calibration of my LCD RPTV. I tried your suggestion last night on one of the inputs, and the picture came to life. I also noticed that the set was a lot easier to get to D75, as opposed to D65.

Hi,

What windows OS will this program work on and what is required for hardware. JUst wondering if my old 133Mhz/W95 laptop will work.

It's not clear, but this post (http://www.avsforum.com/avs-vb/showthread.php?p=9759024&&#post9759024) by Zoyd in the same thread shows figures from the Spyder patent that indicate that the region in question is not completely filtered by the internal detector filters. I guess we'll never really know until someone with a reference instrument does a comparison with the Spyder2, with and without the external filter.

-Steve

I think we are getting off-topic here, but I just wanted to know if this is a valid test for the Spyder2 infrared filtering question:

With the Spyder2 set for LCD mode with the filter on, I ran continuous measures using ColorHCFR - see the attached CHC file. The first four measures are with the S2 pointing up, reading ambient room light. I then pointed my TV infrared remote about 2 inches from the sensor for the next four measures - the red reading shot up from 147 to 166%. I moved the remote closer - to about 1/2 inch - for the last four measures and the red shot up to 198%!

Let me know if I am totally off-base here, but I guess that if a plasma tv emits infrared anything near the same band as an infrared remote, the S2 is will report inaccurate reds.

Yes it will as many other probes (including our one)... And some projectors bulbs did emits IRs... Hence IR filter is pretty interesting sometimes. ;)

--Patrice

I think we are getting off-topic here, but I just wanted to know if this is a valid test for the Spyder2 infrared filtering question:

With the Spyder2 set for LCD mode with the filter on, I ran continuous measures using ColorHCFR - see the attached CHC file. The first four measures are with the S2 pointing up, reading ambient room light. I then pointed my TV infrared remote about 2 inches from the sensor for the next four measures - the red reading shot up from 147 to 166%. I moved the remote closer - to about 1/2 inch - for the last four measures and the red shot up to 198%!

Let me know if I am totally off-base here, but I guess that if a plasma tv emits infrared anything near the same band as an infrared remote, the S2 is will report inaccurate reds.


Thanks JD, that's very interesting. It indicates that even with the external filter, IR can influence the readings. It's still not definitive whether using the filter improves or degrades the accuracy of a plasma calibration, but it does lead me to believe that the filter should be in place.

-Steve

great info richlo, I too was somewhat dissapointed after calibration of my LCD RPTV. I tried your suggestion last night on one of the inputs, and the picture came to life. I also noticed that the set was a lot easier to get to D75, as opposed to D65.


Bingo!!!...I have some thoughts that this is one of those calibration SECRETS not let out to the public...hope I didnt make any enemies, hopefully more friends out of this...

This question should probably go in its own thread, since I noticed this using both HCFR and AccuCal.

Does anyone else prefer D75 over D65? I'm finding calibrating to D65 results in an image with a yellowish cast (too warm). I've calibrated to D65 over and over, changing the gamma settings, etc. and I always end up with a warm image that leaves me thinking "that doesn't look quite right". Note that I'm calibrating a Sony CRT with an Eye One sensor. Any thoughts?


Here is another discovery with D75 over D65...

Thanks JD, that's very interesting. It indicates that even with the external filter, IR can influence the readings. It's still not definitive whether using the filter improves or degrades the accuracy of a plasma calibration, but it does lead me to believe that the filter should be in place.

-Steve

As Patrice mentioned, their probe does this as well, and I got similar - even higher infrared readings - with the DTP-94. It would be interesting to see the results of measurements using a camera infrared filter (one that blocks visible and passes infrared) between these colorimeters and various displays. Then in theory, if it is significant for the display, one could somehow subtract that value from the normal readings?

Even better would be to use an infrared block filter that didn't affect the visible spectrum. Maybe something like this:http://www.schneiderkreuznach.com/tipps/uv-ir_cut_filter.htm

Zoyd,
I'm a bit confused by your statement and the data that you linked to. Your data shows that the plasma display has significant red content outside of the human visual response, and that the filter largely removes this. It seems to me that this indicates that the filter SHOULD be used for plasma displays, otherwise the S2 will measure excessive red. :confused:

-Steve

The human eye response tails off at about 680 nm (the x-bar function in the graph) and I was just pointing out that the s2 filter begins attenuating well before 680 nm so it is acting as a red shaping filter as well as an infrared filter. The plasma spectra shown has all of it's energy within the normal eye response and a large peak at 625 nm that is not present in the lcd I looked at. So I do not think that the filter is primarily an infrared filter but a red shaping filter that should not be used with a plasma due to the 625 nm peak.

For RearProjector CRT users - (using CRT MODE for DTP-94)

I have had the hardest of time calibrating a Mits55807 CRT tv that was professionally calibrated some years ago. I wanted to tweak the greyscale, when I first ran my ColorHCFR for original settings...the greyscale was hovering over/close around 7500 (d75) and I calibrated it at D65 (x=3127 y = 329)...After calibrating...I was never impressed by the looks, the results where by number - fantastic - but the pix always had the rouge look (magenta in a way)..GUESS WHAT..I realize that at least with this Mits (and let me tell you I have a PERFECT COLOR DECODER professionally done by a proMits calibrator)...Calibrating to D75 is what it should be and NOT D65...This was also an issue with a friends of mine who I did their Mits, I just could never get why I was getting poor viewing results, I thought I just didnt know how to use the COLORPERFECT in the Mits..Now I know...its D75 not D65..if any of you are having the same issues with your RearProjector CRT, calibrate to D75 and see your results..I am BEYOND happy now..

Rich, if I did this then my blue line would be above my red and green correct? and then try to reach D75 all the way across on the color temp graph?. I have a panny CRT RPT that I calibrated once and was not real happy with it. I have not tried it again but I think I might give it a go and see what happens. Why D75 and not D65 for the CRT RPT?

Rich, if I did this then my blue line would be above my red and green correct? and then try to reach D75 all the way across on the color temp graph?. I have a panny CRT RPT that I calibrated once and was not real happy with it. I have not tried it again but I think I might give it a go and see what happens. Why D75 and not D65 for the CRT RPT?

just select D75 for the white color in the preferences, and adjust same as for D65.

Rich, if I did this then my blue line would be above my red and green correct? and then try to reach D75 all the way across on the color temp graph?. I have a panny CRT RPT that I calibrated once and was not real happy with it. I have not tried it again but I think I might give it a go and see what happens. Why D75 and not D65 for the CRT RPT?

This is a question Id love to ask the guy who calibrated my projector and this person who calibrated was one of the most know Mits calibrators around...I just know that originally when I ran my greyscale, it came to d75 (or hovering close to it) and the pix still looked pretty good years later but when I re-calibrated to D65 and not only was the diff in cuts and drives was pretty large but the pix changed dramatically and not in a good way...THen after reading MrBob (who post here) stating that he re-calibrated a set done by the same guy who did mine, and also was a Mits (same year model), and that all he did was minor adjustments to it, I figure, let me try to hit d75, AND JUST LIKE THAT..the picture SNAPPED in place...it was a QUANTUM LEAP...not only did the picture come out awesome but my Primaries and Secondaries came in line as before it wasnt..I have no expert reason, just its one of those things I know - my Mits looks way better with D75..give it a shot...nothing to loose, much to gain...I really think there are many secrets not told here because some of the calibrators dont share everything - so this could be one of those things.

btw..all you have to do is change your white point to D75 and calibrate from there

btw..all you have to do is change your white point to D75 and calibrate from there

Ahh gotcha, This is only for CRT RPT's right? I do not want to try this on my projector or do I?

just select D75 for the white color in the preferences, and adjust same as for D65.

Gotcha, I was not thinking :)

This is a question Id love to ask the guy who calibrated my projector
Which Mits PJ was it ??

:cool:

Hi,

What windows OS will this program work on and what is required for hardware. JUst wondering if my old 133Mhz/W95 laptop will work.

I'm using ME and it works just fine.

with all due respect rich, but calibrating to D75 makes no sense. I understand what you're saying, that apparently your set was closer to D75 than D65. Also you say it just didn't look right at D65.

Let's pick this apart with the assumption that what "looks right has to be D65":
* could your instrument be adding in a systematic offset so that D75 is really D65?
* could the software tool be adding in a systematic offset so that D75 is really D65?

my assumption might be bad, but the only real way to figure this out, instead of using anecdotes, is to use another software package and meter, run a grayscale scan for the 4 configurations (2 meters and 2 software packages gives 4 combinations) and see what the measurements are.

Big note here is that IMO D65 is the correct standard to calibrate to. My old dtp-92 (not my current meter) had an issue with green, so that when I thought I was hitting D65, I was way off as verified by the dtp-94.

Best,
jeff

with all due respect rich, but calibrating to D75 makes no sense. I understand what you're saying, that apparently your set was closer to D75 than D65. Also you say it just didn't look right at D65.

Let's pick this apart with the assumption that what "looks right has to be D65":
* could your instrument be adding in a systematic offset so that D75 is really D65?
* could the software tool be adding in a systematic offset so that D75 is really D65?

my assumption might be bad, but the only real way to figure this out, instead of using anecdotes, is to use another software package and meter, run a grayscale scan for the 4 configurations (2 meters and 2 software packages gives 4 combinations) and see what the measurements are.

Big note here is that IMO D65 is the correct standard to calibrate to. My old dtp-92 (not my current meter) had an issue with green, so that when I thought I was hitting D65, I was way off as verified by the dtp-94.

Best,
jeff

Thanks Jeff, it could be...but I am not the only person who has posted that have had better viewing results with D75 instead of D65...and as stated orginally, this was specific to my CRT-Mits55807 - on my MitsDLP 3000U, D65 is perfect and looks natural...Here is how I see it, calibrate to standards (d65) if for some reason, after fine tuning color decoder, brightness, contrast, sharpness and any SVM (although this does not impact the picture to me as much), if it doesnt look good, why not just give D75, seems to work for some people..Industry standards are simple guidelines and nothing else, if you deviate a little and I know that pro calibrators do so in certain situations - in this case I deviated from it and the pix is just as phenomenal as it was when I first had it professionally done...So that you know, I have recently purchased the ColorHCFR probe and will do a comparison, yes, using this software - and then I might take the plunge with Calman (or if someone in the Chicago area wants to meet up and do comparisons - LET ME KNOW - not looking to post here with them - I hate confrontations ;0)


btw... if your thoughts that maybe the offsets with a d75 is actually d65, then that would be scary because then if I calibrate to d65, does it calibrate to D55? :eek: ..Actually I wouldnt care if D65 is D75 for me..I have to get back at you with my results using the ColorHCFR probe and see...if both give me D75, then I will need to go the next step...but i can live with D75 if its offsetting it to d65...

One thing though..Some posted above and didnt like D65 and went with my suggestion to d75 and the individual stated that the picture came to life..and that is exactly how i felt - and Im not sure they where using a DTP-94 so maybe its not the probe but ...stay tune.....

Ahh gotcha, This is only for CRT RPT's right? I do not want to try this on my projector or do I?

I hope nobody is taking my advice that you SHOULD be calibrating to D75, I was pretty specific that D75 for a MitsCRT55807 was what brought the picture to life...I just knew the picture wasnt right, and I have enough experience to know if something is not right...Take my advice lightly and try D75 on your CRT since you were not happy with the viewing results, it may be that yours acts just like mine...Should you use it on your projector? assuming that its not a CRT, maybe not, my Mits3000U front projector is a beaut at D65..so that is why I knew something is not right with my CRT that sits behind it...I continuously reverted back to my original settings from the calibrator, and then I calibrated to d75 from it and increased contrast since it was set too low after aging 5 years...and the pix just SNAPPED into place - I just cant see myself getting rid of it now..I love the CRT look..

maybe for kicks I'll try D75 on my DLP - I can always revert back to my original settings..

Anyone know when v1.3 will be available? Are we looking at a few weeks or months?

D75 is bluer than D65, so maybe that's what you enjoy.

I'd argue that D65 is not just a guideline, but rather the *standard* that the material is mastered to. If you calibrate to D75, then I don't know what you're doing to the image.

Could it be something as simple as that you've got a better gamma when you get to D75 than with D65? That can make all the difference between an image that pops or not. That's what I seem to hear you say.

If the gammas are identical and all that changes is the white point, then I'm more suspicious of a systematic error somewhere (meter and/or software). If gamma also changes then all bets are off.

Note that MichaelLTV says eventhough it's supposed to be 2+2=4, some folks just like 2+2=5 or 7 or 8 or whatever. ;-)
Best,
jeff

D75 is bluer than D65, so maybe that's what you enjoy.

I'd argue that D65 is not just a guideline, but rather the *standard* that the material is mastered to. If you calibrate to D75, then I don't know what you're doing to the image.

Could it be something as simple as that you've got a better gamma when you get to D75 than with D65? That can make all the difference between an image that pops or not. That's what I seem to hear you say.

If the gammas are identical and all that changes is the white point, then I'm more suspicious of a systematic error somewhere (meter and/or software). If gamma also changes then all bets are off.

Note that MichaelLTV says eventhough it's supposed to be 2+2=4, some folks just like 2+2=5 or 7 or 8 or whatever. ;-)
Best,
jeff

Here is something to note

When I calibrated to D65 the deviation from original was MUCH greater than the last calibration I had, now that I did D75, its within a few clicks of the original settings(which makes sense since my original reading was close to D75)..In addition (and i know I stated this previously), MrBob calibrated a similiar set done by the same calibrator and he said the same thing about the greyscale, only a few clicks from the last calibration

Here is my files:

file (55807-9) should be white point D65, ColorSpace709

file(55807-10) should be white point D75 - same colorspace

Gamma tracking a little better on the D65, something I canwork on, but very content visually (left gamma at default at the end from original calibration)

I think that you may be right, D75 (bluerrr)might seem more ENJOYABLE to me at least on this set.


and you make a good point about stating MichaelTLV, even he knows that standards are deviated

I can guarantee you if you ever saw D65 and D75 on my set, you will agree that D75 just seems to make sense - its that much different...

Of course comparing 2 softwares with the same meter, then seperate meters will probably let me know if something else is going on and the process of elimination will take place..Im not much concerned, Ive never called myself an expert - and Ive been giving wrong advice here and there...still lots to learn..

Anyone know when v1.3 will be available? Are we looking at a few weeks or months?
Few weeks... so could be a big month :D

A good thing would be to know what you expect with 1.3...

--Patrice ;)

Just by looking at the measurements, I agree that D75 indeed gives a much more lively picture on this set. Main reason; It's outputting nearly twice the amount of light att 100% stimuli. What puzzles me is that your primaries differs between the two measurements?

As I understand it, this is a CRT-Set. Personally I would aim for gamma ~2.5 instead, which is the "natural" gamma for crt-displays. I have a FP EH Marquee 9500 LC/Ultra myself, and at 2.2 it just looks damn flat to me.

// Lyckman

Few weeks... so could be a big month :D

A good thing would be to know what you expect with 1.3...

--Patrice ;)

Good to hear. I'm mainly wanting the improved meter training setup/wizard and any related fixes to the current implementation (I seem to recall you saying there were some issues with the way it calculates offsets in 1.2x?).

Ok, that will be there ;)

--Patrice

Ok, that will be there ;)

--Patrice

Do you have a handle / fix for the offset calculation issue that needed tweaking?

Hi Lovingdvd

We are still developping the next release (slowly, I confess ;) ). We changed the formula concerning calibration files, offsets and training, to take better care of white readings. It's in beta test now. But we can't deliver any beta, the internals are actually a mess ;)

Regards
Georges

Rich,

Looking at your files, the reason why D75 looks better to you is because overall brightness increased, or at least the cooler color reads brighter.

My CRT RP is calibrated to D65 with overall brightness matching your overall. The issue must be that you are calibrating using "Display w/ black compensation", which as I have stated b4, will give you dimmer picture specially on the mid to low end.

I'd say calibrate to D65 using "Camera Gamma" and see what you get.

after reading the data sheet and white paper on the TAOS sensor used in the HCFR probe, I'm left with the impression that it has a deltaE of about 5.5 inherent in it's design. How can this be accurate enough?

:confused:

after reading the data sheet and white paper on the TAOS sensor used in the HCFR probe, I'm left with the impression that it has a deltaE of about 5.5 inherent in it's design. How can this be accurate enough?I don’t know about that being inherent to the TOAS sensor. I routinely see delta e much less than 5.5 with all the HCFR sensors that I have built. See the following plots from my direct view CRT using a HCFR style probe. <2 deltaE from 35IRE on up and (useful?) color data @ 5 IRE. (Again, this is from a direct-view CRT whose reported CR from HCFR is >82K:1. Data at 5 IRE. I don’t get that from my SpyderII ;)

Dave
http://members.aceweb.com/dlarsen/PICS/HCFR_PROBE_RGBLEVELS.jpg
http://members.aceweb.com/dlarsen/PICS/HCFR_PROBE_TEMP.jpg
http://members.aceweb.com/dlarsen/PICS/HCFR_PROBE_CIE.jpg
http://members.aceweb.com/dlarsen/PICS/HCFR_PROBE_GAMMA.jpg

That is the magic bit ;)

--Patrice

Rich,

Looking at your files, the reason why D75 looks better to you is because overall brightness increased, or at least the cooler color reads brighter.

My CRT RP is calibrated to D65 with overall brightness matching your overall. The issue must be that you are calibrating using "Display w/ black compensation", which as I have stated b4, will give you dimmer picture specially on the mid to low end.

I'd say calibrate to D65 using "Camera Gamma" and see what you get.

I give this a shot..

So is it 2.5 camera gamma or is it 2.2 camera gamma


To me its not the brightness that did it for me but the image looked more natural...but heck let me take some of this advice and give it a go

I don’t know about that being inherent to the TOAS sensor. I routinely see delta e much less than 5.5 with all the HCFR sensors that I have built. See the following plots from my direct view CRT using a HCFR style probe. <2 deltaE from 35IRE on up and (useful?) color data @ 5 IRE. (Again, this is from a direct-view CRT whose reported CR from HCFR is >82K:1. Data at 5 IRE. I don’t get that from my SpyderII ;)

Dave


I don't want this to be interpreted incorrectly so I will say up front that I am not knocking the HCFR sensor (I've never used one), I've used the software and love it. But I do want to point out that precision and accuracy are two different things and Dave's post implies accuracy when he is actually demonstrating precision. The white paper does indicate that the TOAS based sensor has an inherent inaccuracy in chromiticity of something like 5.5 deltaE based on it's design. Unless the HCFR design has somehow overcome that deficiency in design, their probe will suffer from it too. Dave's post demonstrates extremely good stability and sensitivity (or SNR, signal-to-noise ratio) but says nothing about accuracy, to do that you would have to compare to a known reference instrument like a NIST colorimeter. As for the s2 (I have used this one), it is also quite sensitive as I can get stable readings down to 5 IRE with enough patience (long integration times). The s2 design actually has the greater potential for accuracy because you can more closely match the CIE x-bar, y-bar, z-bar function then you can with a TOAS chip. However, the manufacturer does not utilize this potential and sells sensors with mediocre calibrations.


-scott

But I do want to point out that precision and accuracy are two different things and Dave's post implies accuracy when he is actually demonstrating precision.

Yes indeed, I agree. I said this in my webpage…
‘The absolute ‘accuracy’ of the probe will depend on the calibration file and the accuracy of the reference instrumentation used to verify that. As I used a SpyderII to set my reference for my HCFR probe calibration, it won’t be anymore ‘accurate’ than the SpyderII it was trained against. Although it can’t be anymore accurate than the reference, it can perhaps track better or have a better range, better resolution or other benefits relative to the reference meter.’

Dave

Laric, here is my only feature request for 1.3

I really liked the continuous mode to calibrate my grayscale on-the-fly but one thing I wish I could do was set the values of the max/min for the y-axis like you can when you view non-continuous measurements. Currently I can only choose a pre-defined range for RGB and dE (I believe 50-150 and 0-5 respectively.) When you are getting close to fine tuning RGB to 100 I would really like to define a min/max of 95-105 for example.

Yes indeed, I agree. I said this in my webpage…
‘The absolute ‘accuracy’ of the probe will depend on the calibration file and the accuracy of the reference instrumentation used to verify that. As I used a SpyderII to set my reference for my HCFR probe calibration, it won’t be anymore ‘accurate’ than the SpyderII it was trained against. Although it can’t be anymore accurate than the reference, it can perhaps track better or have a better range, better resolution or other benefits relative to the reference meter.’

Dave

Agreed, from the posts I've seen the HCFR sensor can do things the spyder can't in terms of sensitivity. But whether you train it against the s2 or some more accurate sensor, (and the TOAS chip manufacturer's white paper is correct), then there is still the potential of a 5.5 deltaE uncertainty in chromiticity values relative to the reference. That seems like a lot. Here is the relevant paragraph from the white paper:

"When color-corrected TCS230 values are transformed to L*a*b* coordinates
through the optimized matrix of Equation 6, then compared to
the published L*a*b* values of the ColorChecker (as measured by
instrumentation), the average magnitude of the error is about
5.5 delta-E units. In high-quality image reproduction, a delta-E value
of unity is taken to lie approximately on the threshold of perceptibility,
and a value of 2 is taken to be acceptable. A color measurement
instrument is expected to have an error of a fraction of a delta-E."

As for the s2 (I have used this one), it is also quite sensitive as I can get stable readings down to 5 IRE with enough patience

On a display with >80K:1?. I haven't been able to duplicate that with my SpyderII. It tails off around 15IRE and seems pretty much color blind <10. Perhaps I'm not patient enough. I suspect that that the low end wiggles was inherited from the SpyderII used for the initial D65 calibration and I'm confident I could get much better. (Maybe not in terms of 'accuracy' but in terms of stability.) It's tough to measure the color of dark and I don't have a cryogenically cooled spectrophotometer. ;)

Dave

then there is still the potential of a 5.5 deltaE uncertainty in chromiticity values relative to the reference.

Who am I to argue with the TAOS datasheet? All I can really say is that I’ve built and tested 8 of these now and they all perform just like I posted. Seeming much less than 5.5 delta e of uncertainty as reported from HCFR. Maybe I got an exceptionally good batch of sensors from TAOS? I kinda doubt that however...

Dave

Who am I to argue with the TAOS datasheet? All I can really say is that I’ve built and tested 8 of these now and they all perform just like I posted. Seeming much less than 5.5 delta e of uncertainty as reported from HCFR. Maybe I got an exceptionally good batch of sensors from TAOS? I kinda doubt that however...

Dave

I can see from your graph that the probe is extremely stable, peak-to-peak fluctuations are excellent (< 1 delta E above 30 IRE). For comparison, here is my s2 calibration results on my plasma. The fluctuations are higher, about 2 dE p-p, and I do get decent results down to 10 IRE. For reference my black level was 0.068 but the probe can get repeatable measurements down to about 0.017 which would be a contrast ratio of 8000:1 on this display. (Actual contrast ratio is 1800:1) My only point is regardless of what these delta E measurements indicate I can not tell you my absolute uncertainty in delta E, only how stable my sensor is.

Laric, here is my only feature request for 1.3

I really liked the continuous mode to calibrate my grayscale on-the-fly but one thing I wish I could do was set the values of the max/min for the y-axis like you can when you view non-continuous measurements. Currently I can only choose a pre-defined range for RGB and dE (I believe 50-150 and 0-5 respectively.) When you are getting close to fine tuning RGB to 100 I would really like to define a min/max of 95-105 for example.
You can do that now. Right-click on RGB window then Scale>RGB>Values...and input whatever you want.

I give this a shot..

So is it 2.5 camera gamma or is it 2.2 camera gamma


To me its not the brightness that did it for me but the image looked more natural...but heck let me take some of this advice and give it a go
I have my set calibrated to 2.2 otherwise I lose shadow detail.

My only point is regardless of what these delta E measurements indicate…

Yet that is my only frame of reference. The delta e measurements data from HCFR… I called it like I saw it straight from HCFR. It is what it is and I really don’t have the method or the means to slice it any finer.

Dave

I have my set calibrated to 2.2 otherwise I lose shadow detail.

What do you do to change gamma? My set is (at best) a gamma around 2.0, and nothing I do seems to affect that in the least.
Roy

You would use Gamma controls in user menu or service menu in addition to tweaking Brightness/Contrast.

Don't focus on "average" value, set the Gamma window to Logarithmic mode(right-click>logarithmic) and adjust to have as many steps(points) hit as close to 2.22 as possible.

BTW- post your file, we'll take a look at it.

Yet that is my only frame of reference. The delta e measurements data from HCFR… I called it like I saw it straight from HCFR. It is what it is and I really don’t have the method or the means to slice it any finer.

Dave

And I am just providing (hopefully) some context to what those measurements mean to the original poster who was looking for information on accuracy. It's not a knock on you, you know what you are doing, but it can be confusing to people who have not dealt with calibration issues as to what exactly these probes are measuring, how well they are making those measurements, and what the trade-offs are.

-scott

"When color-corrected TCS230 values are transformed to L*a*b* coordinates
through the optimized matrix of Equation 6, then compared to
the published L*a*b* values of the ColorChecker (as measured by
instrumentation), the average magnitude of the error is about
5.5 delta-E units.

However, The HCFR application and probe can create and utilize a unique and custom set of matrix coefficients for the probe based on the display spectum. It doesn’t appear that they use a one-size-fits-all approach for these coefficients. Yet. ;)

Dave

However, The HCFR application and probe can create and utilize a unique and custom set of matrix coefficients for the probe based on the display spectum. It doesn’t appear that they use a one-size-fits-all approach for these coefficients. Yet. ;)

Dave

yes, I'd like to hear from the HCFR folks on what they think their probe accuracy is, given different display spectra and a perfect reference to train with. There are two generic problems with all probes, the first is how well the color information is integrated via filters, direct spectra, or other methods to reproduce the CIE response curves using a calibrated reference input, say D65 etc. The second is after deploying the sensor, how well it deals with sources of different spectral distribution. This is where you get into calibration coefficient corrections, the s2 filter vs. no filter crt debate, etc. The primary *potential* problem I see with the HCFR probe is that there is no way around the 1st problem. Inherent in the design, which consists of 4 broadband measurements, is the fact that it cannot reproduce the CIE response perfectly. The best it can do with the 4 pieces of information it has leads to an estimated accuracy of 5.5 dE using a D65 reference, no amount of calibration coefficient tweaking can fix that.

*potential* problem or not, the data is the data and I can really only comment on what I’m seeing. Personally, I’m much less concerned about absolute accuracy than I am with stability, range and repeatability. I’m more concerned with a flat gray-scale. I suppose I rather have a flat D68(whatever) with low variance rather than a D65(whatever) with higher variance.

I suspect I would notice a 10% variance much more that I would notice a flat 10% error. Just a gut feel though.

As far as potential problems with the probe, I’m more concerned about the effects of IR.
Without an external IR filter, It appears the sensor may get fooled under certain circumstances. There’s a second (false?) response bump up in the IR. It does appear the IR filter would be pretty effective in knocking that out.

Having said that however, I am not seeing an IR problem manifest itself with the CRT and LCD I have used it on. Overall, I’m quite pleased with the results I’m seeing from this probe. I feel it easily bests my SpyderII but I suppose YMMV and I'd have no frame of reference without my SpyderII.

One other thing, I suppose I should be considered biased as I’ve made a few extra (why go through all that effort and expense for just one?) and have offered them up for sale.

Dave

You would use Gamma controls in user menu or service menu in addition to tweaking Brightness/Contrast.

Don't focus on "average" value, set the Gamma window to Logarithmic mode(right-click>logarithmic) and adjust to have as many steps(points) hit as close to 2.22 as possible.

BTW- post your file, we'll take a look at it.

My before calibration and an example attempt with contrast lowered are both posted to the other, calibration files, thread.

My set has a set of 32 RGB values that it calls Gamma. Each is a value between 0 and 3FF, where each value must be larger than the step before it and smaller than the step after it. According to the service manual, it splits "luminance" into 32 adjustable steps, they should never be changed, but if the customer complains that he wants more red in step 12 increase the red value in step 12.

All of which, of course, explains absolutely nothing. I presume they split the RGB 0-255 range up into 32 segments, and that each segment has some influence on the luminance of that segment of the input RGB values. Unfortunately, I have no idea how to successfully change the values. I have a range of Blue that is elevated that I'd like to decrease. So I figured out which segments covered that range of luminance and decreased the B value for those gamma segments. It made absolutely no difference in the 20 step gray scale I measured, but when I looked at a continuous gray ramp I could clearly see a periodic increase in red. So whereas I had decreased blue in one contiguous range of the gamma curve, I had apparently induced a repeated error in the gray scale that occurred in between the 5/10/15, etc. steps I was measuring with HCFR.

So I put the original gamma values back and concluded I didn't have enough information about how gamma works to successfully change the curve. My guess is that the 32 values I can enter are the coefficients of some gamma equation, but without knowledge of the equation I can't tell how to change it without introducing instability outside the range I've altered. I can't find any more documentation about how these things are designed, so I decided I had to live with the curve as provided by the factory.

So basically I've got RGB cuts and gains, and brightness and contrast, and color engines whose gamma has not been calibrated to produce a uniformly increasing luminance - if you look at my calibration files you'll see that Blue increases much faster throughout the middle of the luminance spectrum.
Roy

You can do that now. Right-click on RGB window then Scale>RGB>Values...and input whatever you want.

In the *continuous* window you can do that? I don't have ColorHCFR in front of me now but I swear when looking at the window that is updating in real-time you don't have an option. You do however have the option in the static RGB window once you do your grayscale measurements.

One other thing, I suppose I should be considered biased as I’ve made a few extra (why go through all that effort and expense for just one?) and have offered them up for sale.

Dave

AHA! now we get the juicy part. :) Seriously though, I agree that the single most important requirement for any sensor is stability. In that respect the s2 does appear inferior (I have also seen mention of temperature drifts but I don't know if the HFCR suffers from that). I think on the level of a few dE the accuracy question is purely academic, I just like being precise about these things. ;)

AHA! now we get the juicy part.
Ha! I'm sure glad that I'm the one that disclosed that fact first rather than someone else! (I have also seen mention of temperature drifts but I don't know if the HFCR suffers from that). Interesting comment. I see the DTP-94 has a built in temp transducer. One may notice some extra un-populated circuitry on the HCFR layout I did. Hmmm…. Integrating that feature would require a rework of the PIC’s uCode though.

Dave

The S2 does suffer from ambient temperature drift with as little as a 5 deg change you can get a 5% reading drift. The DTP-94 does have a built in temperature compensator which is good for about 15 deg window outside of that and you will also get drift. The D2 does not have a temperature compensator but I think with it's design it's not susceptible to drift within it's normal operating temperature range.

The S2 does suffer from ambient temperature drift with as little as a 5 deg change you can get a 5% reading drift.

1%/deg is rather high, typical silicon detectors are 0.1%/deg or lower so I suspect it's the S2 cut-on filters shifting, can you tell me what drifts, x, y, or Y?

When I select Eye One - CRT or Eye One - LCD on HCFR, what am I actually selecting?

When I select CRT or LCD on HCFR, what am I actually selecting?
Assuming you are using a SpyderII, I believe you are informing the app if you have the extra filter in place or not.

Dave

to the ColorHCFR team -

How are the calibration files for your probe created? Is there a Known Reference/Probe that is use and then its trained for a specific display? I see currently that there are 3 Mits 3000 files, one which is for the filter(not sure why one would use is filter on this - seems like it wouldnt need one), then there are 2 others? are the other ones the same?

Dave,

Why are two sensors used in the HCFR probe? Does this extend the dynamic range?

-scott

Why are two sensors used in the HCFR probe? Does this extend the dynamic range?My guess would be for better resolution and/or more data samples to average and integrate with. You can set up the SW to use sensor 0, sensor 1, or both as well as different measurement interlacing. I can get similar results using a single sensor (either one) or both. When both are selected they appear additive to me. At least in the Y channel.

I rotated one sensor 90 degrees relative to the other on the layout I did. No real reason other than it just seemed redundant to have them exactly the same. I also plan to make the vignetting aperture (hole in box) smaller in an attempt to limit the viewing angle somewhat. That’s been another reported issue- that it has too large of a viewing angle. I don’t see that problem however. I also plan a smaller hole because it’s a lot easier to make a hole bigger than it is to make it smaller. :D

Dave

You may also find more info on the HCFR site...

I've been getting a French lesson (not THAT kind!) on the side. Funny, Google translate equates the term IRE to ANGER. It took me a bit to catch that. I was seeing ANGER all over that site!

Dave

I muddled my way through Color HCFR and managed to take the readings with my Spider colorimeter. Anyone want to take a look and offer any suggestions on settings to change. It is a Samsun 6187SW DLP set and I do not know my way around the service menu so be kind please.
Thanks
Kenal0

The biggest thing I see in a quick glance is the hump on the EO curve about 90IRE. (That and you’re color temp is a bit high but you may prefer it that way and your display may be much more efficient and brighter there) You’re color temp also falls way off @ <40IRE but that be a function of your meter not getting enough signal (light). What do your eyes tell you? Are your lower/mid blacks looking orangeish/brownish to you?

You might also see if reducing your saturation will tend to pull your R/G primaries (and tag-along secondaries) in but it looks to me you’ll need to get your colortemp lower though to get the secondaries much closer.

I’m not familiar with your display or what adjustments you have available. If you have the right sized hammer, I’d try pounding on the 90IRE bump first.

Dave

On second glance, :D are you sure you're not starting to clip your whites? (Set basic B/C first) Maybe the bump @ 90 is more of a dropping off @ 95 or so. Double check that you aren't clipping whites. Also, a DLP? Do you have a setting for 'white peaking' or 'spoke light recapture'? You may try turning those off.

Dave

I muddled my way through Color HCFR and managed to take the readings with my Spider colorimeter. Anyone want to take a look and offer any suggestions on settings to change. It is a Samsun 6187SW DLP set and I do not know my way around the service menu so be kind please.
Thanks
Kenal0


Actually, what could have happen is that yo read 90IRE and 100IRE with the 90IRE pattern..Try doing another run and make sure you reading each one..See if that hump dissapears..Ive done this before and it always looks like that..

You may also find more info on the HCFR site...

I've been getting a French lesson (not THAT kind!) on the side. Funny, Google translate equates the term IRE to ANGER. It took me a bit to catch that. I was seeing ANGER all over that site!

Dave

I read the translated version, my favorite part was at the end:

"we wish you a good assembly and especially much of pleasure to use this probe with the software of “Colorimeter HCFR”."

Anyway, I didn't find out any more info on why two sensors, maybe the intent was to use one with an external filter and one without or as you said to improve the statistics. It looks like a near-IR filter is required, has anyone done any tests with and without a filter on various displays?

Actually, what could have happen is that yo read 90IRE and 100IRE with the 90IRE pattern..

Good one. I guess that would look a lot like clipped whites, eh? One has to be careful navigating with the DVD menus. I never really put much value on the IR emitter for DVD navigation thing but can see a good case for it there. I’ve only used a DVD for patterns once and only to verify that my video decoder / renderer path levels matched the HCFR automated patterns. I always get waaay frustrated teaching a learning remote.

Dave

It looks like a near-IR filter is required, has anyone done any tests with and without a filter on various displays?

I think a few of the HCFR guys have. I believe that’s why there are variants of the probe calibration files. I had to generate a new cal file when I tried a piece of frosted scotch tape. It likes blue :D

I’ve got a few filters and dichroics but mostly of the hot/cold mirror variety (wants to be at a 45) so not really practical. Honestly, I’m not seeing any of this an issue at all (Again, relative to my SpyderII and my displays) so I not real motived. It doesn’t appear ‘required’ at all to me and the data that I’ve posed is from naked sensors. I don’t know under what conditions and what specific displays this may be a problem. Do you?

TAOS has an eval board on their site that has this tres cool lens/filter/dispersion thingy over it though.

Dave

1%/deg is rather high, typical silicon detectors are 0.1%/deg or lower so I suspect it's the S2 cut-on filters shifting, can you tell me what drifts, x, y, or Y?

It was mostly Y so it's only relative to the duration while you are calibrating. But makes it difficult to compare calibration runs over a period of time like days or weeks. The xy only changed by a small amount around 0.1% like the typical silicon. The biggest issue is when you place a S2 on a direct view LCD or Plasma it can take up to 30 minutes before it will be quiescent with the display. Because of the heat they can produce.

I don’t know under what conditions and what specific displays this may be a problem. Do you?

TAOS has an eval board on their site that has this tres cool lens/filter/dispersion thingy over it though.

Dave

Looking at the response curves of the bare sensor it would be required for any display with significant energy between 680-1100 nm. Not sure which displays this applies to, I've measured my plasma and there is nothing up there, maybe LCD's or other displays with white light sources would benefit.

the eval board does look interesting.

Thanks, I will rerun and make sure I have the correct ire from the dvd.
I will probably wait until I get the Eye One Display Lt. I am sending the Spider back.
The Eye One will work with this software?
Does the Eye one have a tripod mount? If not the how do you attach to the screen?
Thanks

Kenal0

Thanks, I will rerun and make sure I have the correct ire from the dvd.
I will probably wait until I get the Eye One Display Lt. I am sending the Spider back.
The Eye One will work with this software?
Does the Eye one have a tripod mount? If not the how do you attach to the screen?
Thanks

Kenal0

yes, the EyeOne is fast becoming the more popular and reliable probe. Im not a big fan of the Spyder...The EyeOne though is not good for PLASMA displays..So I hear..

For tripod mounting..here is what you need to do

http://www.datapopuli.com/forums/viewtopic.php?t=342

Thanks, I will give it a try. Of the data I uploaded, what would be used to analyze and interpret contrast and brightness. I basically did nothing more than take the readings and upload this file. I liked the idea of Spidertv because it was a no brainer for adjusting the basics but it just would not work correctly on B/C. What is the best way to make the adjustments for B/C?
Thanks and sorry for the newb questions but I have to start somewhere.
Kenal0

Hello SiegeX
I really liked the continuous mode to calibrate my grayscale on-the-fly but one thing I wish I could do was set the values of the max/min for the y-axis like you can when you view non-continuous measurements. Currently I can only choose a pre-defined range for RGB and dE (I believe 50-150 and 0-5 respectively.) When you are getting close to fine tuning RGB to 100 I would really like to define a min/max of 95-105 for example.
If you want to calibrate using continuous measures, seeing precise RGB data, just open the information window (using View menu or toolbar button). This window includes an RGB histogram matching in real time the last measure.

Regards
Georges

Hi everybody

Why two TAOS sensors on HCFR probe ? Now, it's useless. But we internally use it for miscellaneous experiments, and we hope to find a way, using one or two different filters, to get rid of calibration files... Theoretically, it would be feasible, but in actual state of our researches on this subject, accuracy wouldn't be good enough to ensure perfect results on all projectors.

Regards
Georges

dlarsen.... no worries, I appreciate your point of view and still want my sensor.... I, like Derek, have become somewhat of a collector :D

I should explain that I'm an EE (35 years!) and my point here was to try and elucidate some comment from the HCFR guys. I have no doubt that you are getting very stable and repeatable readings... even unit-to-unit.... and that is what intrigues me so much about the unit! repeatable errors are the best kind to have.... they "usually" can be corrected out by training. And if unit-to-unit varience is so small, then a one time training should be reasonable for many instruments :) But my consternation is in the lack of experience in color science and whether these inherent inaccuracies can in fact be calibrated out. The point of the white paper's statement is that the instrument's output will deviate from a perfect color gamut by 5.5 deltaE. Thus if you use it to calibrate your display, and the software says you have a deltaE of < 1, you have achieved a very tight match to the units 5.5 deltaE from a perfect color gamut.

:cool:

Hi everybody

Why two TAOS sensors on HCFR probe ? Now, it's useless. But we internally use it for miscellaneous experiments, and we hope to find a way, using one or two different filters, to get rid of calibration files... Theoretically, it would be feasible, but in actual state of our researches on this subject, accuracy wouldn't be good enough to ensure perfect results on all projectors.

Regards
Georges

Thanks for the reply Georges, you guys have done some really interesting work. Can you comment on what you think the probe's absolute accuracy is, and how it varies by display? The Poynton white paper indicates an average accuracy of 5.5 dE using GretagMacbeth colorchecker patches.

-scott

The biggest thing I see in a quick glance is the hump on the EO curve about 90IRE. (That and you’re color temp is a bit high but you may prefer it that way and your display may be much more efficient and brighter there) You’re color temp also falls way off @ <40IRE but that be a function of your meter not getting enough signal (light). What do your eyes tell you? Are your lower/mid blacks looking orangeish/brownish to you?

You might also see if reducing your saturation will tend to pull your R/G primaries (and tag-along secondaries) in but it looks to me you’ll need to get your colortemp lower though to get the secondaries much closer.

I’m not familiar with your display or what adjustments you have available. If you have the right sized hammer, I’d try pounding on the 90IRE bump first.

Dave
Where should I start to reduce the color temp closer to 6500K? I can access the service menu and have located the RGB gains and bias controls. I am not going to do anything but read and re read around here before I do anything else. I will take another set of readings when I get the Eye One.
Kenal0

Where should I start to reduce the color temp closer to 6500K? I can access the service menu and have located the RGB gains and bias controls. I am not going to do anything but read and re read around here before I do anything else. I will take another set of readings when I get the Eye One.
Kenal0


I live in the chicago area..so let me know if you ever want some help..

I live in the chicago area..so let me know if you ever want some help..

Thanks, I am up in the north suburbs.
Kenal0

Thanks, I am up in the north suburbs.
Kenal0

Im in J-Town but work in the Loop and grewup in Pilsen