Calibration meter shootout - Page 2

Tom>> Great shootout!

However, I would like to fill in on the question that jdbimmer asked about calibration mode...

I'm not very familiar with Colorfacts (only used it once), but in the HCFR application one should select calibration mode before starting the measurements.
I believe CRT mode is recommended for plasma and for CRT-monitors and LCD mode is recommended for all else.
The measured x,y will be very different depending on measuring mode on many displays, I think it can differ more than 10 dE.

I'm therefore wondering, is it possible to select what type of display is to be calibrated in Colorfacts?
Because IF this calibration mode (which probably induces some kind of offset) should be set when using colorimeters on different displays, then there will be quite large errors in Colorfacts depending on display type if one is using a colorimeter.
The alternative could of course be that HCFR is wrong in asking for this, but I guess not since for example the Spyder2Pro automation software asks what type of display is being calibrated also.

Is there such an option in Colorfacts and have You used it when testing?

Hmmm... I searched the User Guide for Colorfacts now and it seems devoid of this type of setting.
In fact the words "CRT", "LCD" and "DLP" are only mentioned once each - and not in this context. Plasma isn't mentioned at all...

So I'm wondering if readings in Colorfacts with a colorimeter are even close to being reliable. In fact, if calibration mode SHOULD be selected as in HCFR, the dE for the the Spyder2 and the hard to manage LCD could be 20 or 30 or even more and the plasma and CRT readings could just as well be dE=14 as dE=4 with the D2.

Can You please comment on this, Tom?

I don not see why an offset should be activiated at all - this has probably more to do with how averaging of measurements are being treated or activated, which in turn has to do e.g. with different polarization of light.

A measurement should be an accurate representation of light, nothing else, nothing less - IMO if an offset is required the wrong meter is used for the job.

I have compared values reported by Colorfacts using a Minolta CS-250 with the direct reports of the CS-250, and there was no differrence. This was done with several different display types (e.g. frontprojectors, plasma and LCD flatpanales).

cheers

/Par

perra1>> So Your conclusion is that HCFR treats this wrong? That all colorimeters measure all light in the same manner, no matter what type of display is being measured?
Because the difference in RGB levels measured can differ quite a lot depending on the selected mode in HCFR.

My whole question of course is just applicable to colorimeters, not spectrophotomers, there is no mode selection for Eye One Pro in HCFR, just for all the different colorimeters supported. Just to clarify.
The Minolta isn't a colorimeter is it?

I.e. the essence of my question is this:
Is the mode setting missing for colorimeters in Colorfacts, or is it unnecessary (and thus destructive if used) in HCFR?

Ahhh... No, then it's naturally not a factor in this test. Sorry.

Strange though that one application has different modes for different displays but another doesn't, but I guess that's another thread.

Depending on the display type, I think the mode would affect the results somewhat, but the conclusion regarding the S2's accuracy would be the same. My guess is that the S2 was in LCD mode on the Plasma as it measured Y at +23% (Zoyd measured +24%). In CRT mode, most likely the S2 Y reading would be much closer, but with an overall higher deltaE. At least on the plasma, D2 was probably in LCD mode, as it tends to lock up ColorHCFR in CRT mode.

Yes, sorry for the typo - CS-200 of course...

cheers

/Par

Still I'm currious of how the new Spyder3 with the verry large aperture light sensetive improvement, whould have come out this shootout.

The mode selection between CRT and LCD for the Display 2/LT is to access two separate calibrations which are 3X3 matrices stored within the eeprom of the device. Each of these calibrations are made on different display types and allow the unit to provide improved performance when measuring a display using that technology. The "mode" selection has nothing whatsoever to do with an offset file as offsets would be different for each and every display made.

Spectroradiometers do not need offset files as they are not filter based instruments and in most cases for video display analysis they measure the wavelength range from 380-780 nanometers. As long as the field of vision of the instrument is small enough it will not have any difficulty measuring LCD flat panels or any other display technology accurately assuming everything else is properly designed in the instrument.

Ok, no offset used. Doesn't really influence the core in my question:

Why does one software use modes while another doesn't? Which software does it right?

The test in it self should show each meters consistency/linearity well even if no mode or if the wrong mode was selected. The linearity of a meter wouldn't be affected by mode would it?
But if a mode needs to be selected, then the correct mode surely also must be set to get correct results - right?
And a plasma should be measured using the CRT-mode - right?
So, was LCD-mode selected when measuring the plasma with the D2, Tom?

Again, only talking about colorimeters, not spectrophotometers.

IIRC when you use CRT mode with the D2 it configures the probe to try and lock onto the refresh rate of a traditional CRT.

Since a plasma doesn't work the same way it will get "confused" in CRT mode.

I'm pretty sure that the HCFR-team states that colorimeters should be in CRT-mode for measuring plasma, but maybe that's not valid for D2, only other colorimeters?

How about neither package gets it right for all meters, kind of hard to do a meter shoot out with software packages that don’t support the hardware properly.

ColorFacts uses the Spyder2 only in LCD mode with the diffuser on which on a true CRT will be off. HCFR uses the Display2 in LCD mode for Plasmas because of the scan rate sync issue which again will be off.

Here is how each meter should be supported and is in CalMAN v3.

x-rite DTP94 has support for LCD and CRT modes and does not need a scan rate sync for CRT. So you use CRT for all phosphor based devices and LCD for all the rest.

x-rite Display2 has support for LCD, CRT and diffuser modes but does need a scan rate sync to work in CRT mode unless you know how to disable it. So you use CRT for all phosphor based devices including Plasma with the sync disabled and LCD or diffuser for all the rest. The Display2 can also have issues with fast DLP colorwheels or the new fast scan LCDs unless you know how to set it up.

x-rite i1Pro has support for all display types with two modes with and without a diffuser. But it lacks good low light capabilities unless you have a system for dealing with it and I'm not just taking about averaging multiple readings it needs to be much more than just a simple average. You use it without the diffuser for most display types and with the diffuser for a direct source reading like from a front projector.

x-rite Chroma5 is the big brother to the Display2. Has support for LCD, CRT and user. So you use CRT for all phosphor based and LCD or user for all the rest.

x-rite Hubble or also known as the Sencore OTC1000 has support for Projectors and all others. It is really sensitive to aiming I guess that's why they built in a laser pointer. It also has to deal with ambient light since it is not a contact meter.

ColorVision Spyder2 has support for LCD with the diffuser on and CRT with it off. It does need a scan rate sync to work with CRT's. So you use CRT for all phosphor based devices except Plasma and LCD for all the rest. The Spyder2 can have low light issues unless you extend it's reading time to over 8 seconds.

DataColor Spyder3 only has a single mode for all display types since the diffuser is not removable.

Progressive Labs CA-6X software supports all of the instruments mentioned above and provides support for all modes offered on each of the instruments so that they work correctly with each display type. We have been working with the Chroma 5 for close to two years and with the OTC-1000/Hubble for close to 3 years. We currently support more then eighteen instruments including those from X-Rite (Sequel/GretagMacBeth), Minolta, and PhotoResearch and are happy to sell the software by itself to those that wish to have a feature rich software package for display calibration.


Regards,

Cliff

Progressive Labs
212-254-3541

You are right software should not have an effect on accuracy all we do is report what the meter told us it read but it can if the software does not setup the meter correctly in the first place. To get the best accuracy from any meter it comes down to being able to match the internal calibration tables, sync modes and external filters if it has them to the display type.

An example is with the Spyder2 it has 2 tables one for CRT and the other for all other display types. With ColorFacts they only use the second table with the diffuser attached on the Spyder2 for all display types. So if you are calibrating a direct view CRT it won't be as accurate as it would if they allowed you to select the CRT calibration table without the diffuser. Which this is an odd point for ColorVision/DataColor if you are using any of the Spyder2 software packages they do use both calibration tables and have you remove the diffuser when needed. But with SpyderTV, SpyderTVpro or ColorFacts they always leave it in place and only use the second table. It's the same hardware so one of them must be wrong.

Same goes for the Display2 it's a tricky device to support. It also has four modes CRT, all others, with and without the diffuser. But if it is not setup right it can produce inaccurate results on some display types like Plasma, fast colorwheel DLPs and fast scan LCDs.

The i1Pro is some work because you need to have low light handlers for it.

Tom since you are a V2 customer you can upgrade to V3 for free just send me an email and we will get you sorted.

Hi Derek,

After i have my new cinema fitted, which will include a 50-60" plasma display probably 1080P KURO, i want to get the best calibration tools for the job.

I currently have a DTP-94 and used HCFR to calibrate my old 50PHD8 and as a starter kit to learn display calibration and got excellent results from this.

Now i want to move up to more accurate/advanced calibration, i have decided i will be buying your Calman software but i'm not 100% sure if the i1pro you sell be the best bet for my system and accuracy or should i be looking at some other sensor?

What would you recommend as the best sensor for what i need?

TIA.

Sorry Tom, thank you indeed!

I thought to ask Derek as i will be using his software but again thanks for the reply!

The i1pro is within my budget but wanted to know what is available around the i1's price range and could i get something more accurate for my needs as the i1 is quite old now and don't know of anything else or prices too besides spider offerings but i think spiders are a step backwards to what i have/want.

I think with the DTP-94 you already have adding the i1Pro to it would allow for you to profile your DTP-94 to it and get the best of both worlds. We do also have the Chroma5 which is the big brother to the Display2 and the DTP-94.

Thanks Derek!

What benefits would i get from profiling the DTP to the i1, why not just use the i1 only or would this profiling give a more accurate reading at lower IRE levels (EG i1 for 0-30ire and DTP for 30-100 ire) can you explain and what is mean't by "best of both worlds" (is this lower and higher IRE readings?)?

How is the Chroma5 with regards to the other 2 sensors?

TIA.

The i1Pro needs some work at low light levels to get it to the published accuracy so it can be slow up to 10-15 seconds for a single reading were it higher levels it's less than a second. So with the DTP-94 it is fast all the time but not as accurate as the i1Pro unless it has been profiled. Profiling is basicly a way to run a transformation matrix on the readings to make the DTP-94 return results as the i1Pro would but for all light levels. The Chroma5 will outperform both the DTP-94 and Display2.

Thanks again Derek,

i knew it was somewhere along those lines even if i did get it a bit back to front, thought the i1 was good at low end IRE.


With regard to the C5 would i be better getting this or better to stick with 1i/DTP with profiling setup?

How much is the C5?

We don't have a price set on the Chroma5 yet still working it out with x-rite on our discounts. Since you already have a DTP-94 I would recommend just adding the i1Pro.