Eizo Foris FG2421: 120hz VA Panel

Another important thing to note! It seems that for whatever reason, my videocard switches from 1080p PC resultion to 1080p native DisplayPort/HDMI, where only 60Hz can be selected and only 16-235 range can be used. Now, I do use the 0-255 HDMI registry editions/ "hacks" to force 0-255 for HDMI, but I don't think it works for DisplayPort. I am quite certain that THIS is the reason why some people reported a completely washed out picture.

BTW, no matter how hard I look close at ANY pattern, I cannot see any crosshatching what-so-ever. I could not see it on my VG249QE.

I say that this monitor suffers 2 major issues:
1. Slight left and right backlight bleed of 3-5mm in length that occurs only with dark patterns and is exacerbated by VA angles, which makes it seem as if the backlight bleed extends to like 1inch or so.
2. Primary colors at 100% saturation are whacked up, but skin tones are excellent, real-world ColorChecker results are excellent, 25-75% saturation sweeps are acceptable/decent.

Both are over-compensated by excellent contrast ratio, awesome 240Hz turbo mode, and easy to use controls!

MonarchX said:

Another important thing to note! It seems that for whatever reason, my videocard switches from 1080p PC resultion to 1080p native DisplayPort/HDMI, where only 60Hz can be selected and only 16-235 range can be used. Now, I do use the 0-255 HDMI registry editions/ "hacks" to force 0-255 for HDMI, but I don't think it works for DisplayPort. I am quite certain that THIS is the reason why some people reported a completely washed out picture.

BTW, no matter how hard I look close at ANY pattern, I cannot see any crosshatching what-so-ever. I could not see it on my VG249QE.

I say that this monitor suffers 2 major issues:
1. Slight left and right backlight bleed of 3-5mm in length that occurs only with dark patterns and is exacerbated by VA angles, which makes it seem as if the backlight bleed extends to like 1inch or so.
2. Primary colors at 100% saturation are whacked up, but skin tones are excellent, real-world ColorChecker results are excellent, 25-75% saturation sweeps are acceptable/decent.

Both are over-compensated by excellent contrast ratio, awesome 240Hz turbo mode, and easy to use controls!

Click to expand...

Man, I'm pretty relieved you got a good sample, and even more relieved to hear that it didn't fall over completely under fierce scrutiny. Wish I had one as good as yours but I'll live. I'm going to be comparing this guy to a couple of other new monitors I've ordered, the firmware plagued Benq XL2420Z and a matte X-Star OCable 1440p. I'll be renting the colorimeter as you suggested as well.

Monarch, can you respond to this post?

I'm curious about why your contrast was reduced after calibration. Why would calibrating it raise your black level?

i also just bought a Eizo Foris FG2421 from newegg, expecting it here around tuesday.. looking forward to finding out what i got ;p do you think they sell refurbished/returned orders?

Zarosath said:

i also just bought a Eizo Foris FG2421 from newegg, expecting it here around tuesday.. looking forward to finding out what i got ;p do you think they sell refurbished/returned orders?

Click to expand...

No lol. Not on these. I'm guessing they got in trouble for trying to legitimately sell the FG2421's open box; I kept buying them and sending them back and then they stopped coming up for sale. I bought 2 of them for $480 and 1 for $450 but they all restarted randomly and 2 of them also had pixel issues--they payed shipping both ways on every one.

Newegg may have some unscrupulous people working in the returns department that like to try to turn around open box motherboards and graphics cards as new, but it's pretty hard to do that with a monitor. Just to make it clear I've been there I peel off every bit of plastic and set monitors up as if I plan to keep them. I found it odd that all 4 of the used/open box FG2421's I purchased, including the one I kept which had 250hours on it already, still had the protective plastic on them. I guess people are just too lazy too peel it off. Personally, I relish that part.

MonarchX said:

Another important thing to note! It seems that for whatever reason, my videocard switches from 1080p PC resultion to 1080p native DisplayPort/HDMI, where only 60Hz can be selected and only 16-235 range can be used. Now, I do use the 0-255 HDMI registry editions/ "hacks" to force 0-255 for HDMI, but I don't think it works for DisplayPort. I am quite certain that THIS is the reason why some people reported a completely washed out picture.
....

Click to expand...

I think you got the point
I'm now using dual link DVI instead of Display port to avoid that.

spacediver said:

Monarch, can you respond to this post?

I'm curious about why your contrast was reduced after calibration. Why would calibrating it raise your black level?

Click to expand...

I found out that CR in the dead center is only 3500:1 but all around it its much higher - just due to your regular VA uniformity issues.

Calibration reduces CR or can do so if your 100% white point is not well calibrated because the resulting software LUT will cut out of those bits. As I hope you know - decreasing gains can very much decrease contrast ratio, so that was the case here.

On a side note - I think this is a keeper! It did restart one time so far, but then I found out that there was a Power Saver option that was enabled. I disabled it and I am hoping that will prevent the issue. How often was it happening for others?

Games look so damn fabulous on this monitor using Turbo 240Hz. I care less about that PWM dimming that causes headaches and etc. All I perceive is super-smooth motion and no noticeable blur. I got sucked back into games like when I had a CRT.

I think that moving from CRTs to LCDs was part of my declining interest in video games. The other part was moving from crappy monitors to acceptable TVs. Now I remember how monitor privacy and immersion felt like along with an awesome pixel size! Luckily 24" and 1080p go perfectly together. 27" needs 1440p at least.

Witcher 2, using extreme HD textures, LOD distance, flora quality, rebalanced combat, and a few other mods looks and plays so well on this baby.

Soon I may need treatment for video game addiction...again. The only thing that keeps me from truly relapsing is that there aren't many new AAA games, but mostly good indies and average AAA games. That and I will be kicked out of my house on the 28th of this month if I don't find a job until then, which is likely to be the case even after 1000s of applications! Anyone need an over-achieving business analyst with experience in extremely high-paced and stressful environments who can speak 3 languages fluently, 1 moderately, and is capable of at least meeting and likely exceeding any expectations thrown at him??? Nobody needs that these days I guess...

Best buy i1Pro spectrometer or at the very least rent ColorMunki Photo spectrometer from here - http://www.lensrentals.com/rent/calibration/colormunki . i1Display Pro and all other colorimeters are based off display type tables, which were created using manufacturer's reference displays, which are not likely to perform the same way as consumer displays. Spectrometers read the actual light and are not based off tables. Any pro calibrator would tell you that if they were forced to have only one device - they would buy i1Pro or i1Pro2 and would trust it over i1Display Pro any day. In my experience, i1Display Pro is nowhere near good enough to be used by itself to achieve accurate or semi accurate results.

Click to expand...

Yes and no. When only measuring luminous colors modern colorimeters like the X-Rite i1 Display Pro or basICColor Discus are often the better solution. All prosumer spectrophotometers are primarily optimized for measuring of surface colors (the advantages of the i1 Pro 2 compared to the i1 Pro are mostly located in this area). Emissive readings are an additional but important feature. Their optical bandpass (!= sampling intervall) is quite coarse which leads to some uncertainty when measuring the sharp spectra of self-luminous devices. Established recommendations like ISO 3664 or 12646 therefore call for a finer graduation. Moreover the signal-to-noise ratio is considerably more problematic which causes larger fluctuations in the shadows.

The filters of modern colorimeters are are progressing towards the ideal characteristic although they don't fully replicate normative observers. X-Rite provides the i1 Display Pro with a sophisticated correction system that is far more flexible than simply applying a fixed linear transformation at the end of the measurment process. The probes are characterized at factory. Together with the desired CMF (observer function) and EDR (spectral characterization of the display to be measured; generic data is provided) the absolute error is absolutely acceptable for most use-cases.

We should also take into account that an absolute error (to some extend) doesn't lead to a measurement that is inconsistent within itself. Effects of observer metamerism (the color-matching functions are no perfect match for individual sensitivity) can have a much higher influence in corresponding scenarios (for example dual screen setups with different display types - my CCFL and WCG-CCFL screens have to be calibrated with a whitepoint differing by dE ~ 10 to achieve a visual match). Aside from that the eye has a wide adaptation range regarding white - D65 as the one and only choice is a myth.

Wether the i1 Pro nor the i1 Pro 2 are bad instruments and we are using them ourselves - while considering advantages and disadvantages.

MonarchX said:

On a side note - I think this is a keeper! It did restart one time so far, but then I found out that there was a Power Saver option that was enabled. I disabled it and I am hoping that will prevent the issue. How often was it happening for others?

Click to expand...

That is not good news. You are going to need to RMA it or return it. I strongly urge you to RMA it through Eizo so you get the same panel back--they will take care of you. Call them on the phone. This restarting problem on a 600 dollar monitor is Bullshit. It will happen maybe 1-5 times a day or more.

MonarchX said:

Games look so damn fabulous on this monitor using Turbo 240Hz. I care less about that PWM dimming that causes headaches and etc.

Click to expand...

This monitor is PWM free above 25 brightness.

Sailor_Moon said:

The filters of modern colorimeters are are progressing towards the ideal characteristic although they don't fully replicate normative observers. X-Rite provides the i1 Display Pro with a sophisticated correction system that is far more flexible than simply applying a fixed linear transformation at the end of the measurment process. The probes are characterized at factory. Together with the desired CMF (observer function) and EDR (spectral characterization of the display to be measured; generic data is provided) the absolute error is absolutely acceptable for most use-cases.

Click to expand...

Oh. Badass! Thanks for weighing in on this. Now I get to OWN a device to calibrate my displays with instead of renting one. I don't really care if it's not as accurate as something more expensive. It's going to do a better job than I can just fiddling with the controls or copying someone else's values.

MonarchX said:

Calibration reduces CR or can do so if your 100% white point is not well calibrated because the resulting software LUT will cut out of those bits. As I hope you know - decreasing gains can very much decrease contrast ratio, so that was the case here.

Click to expand...

thanks for the reply. Yes, I was just wondering whether the contrast reduction was from raising black or lowering white. With argyll, you have a fair amount of control over how much priority you want for accurate chromaticity at the low end, so you can preserve a good black level. And raising black level even a tiny bit can dramatically reduce contrast.

Sailor_Moon said:

Yes and no. When only measuring luminous colors modern colorimeters like the X-Rite i1 Display Pro or basICColor Discus are often the better solution. All prosumer spectrophotometers are primarily optimized for measuring of surface colors (the advantages of the i1 Pro 2 compared to the i1 Pro are mostly located in this area). Emissive readings are an additional but important feature. Their optical bandpass (!= sampling intervall) is quite coarse which leads to some uncertainty when measuring the sharp spectra of self-luminous devices. Established recommendations like ISO 3664 or 12646 therefore call for a finer graduation. Moreover the signal-to-noise ratio is considerably more problematic which causes larger fluctuations in the shadows.

The filters of modern colorimeters are are progressing towards the ideal characteristic although they don't fully replicate normative observers. X-Rite provides the i1 Display Pro with a sophisticated correction system that is far more flexible than simply applying a fixed linear transformation at the end of the measurment process. The probes are characterized at factory. Together with the desired CMF (observer function) and EDR (spectral characterization of the display to be measured; generic data is provided) the absolute error is absolutely acceptable for most use-cases.

We should also take into account that an absolute error (to some extend) doesn't lead to a measurement that is inconsistent within itself. Effects of observer metamerism (the color-matching functions are no perfect match for individual sensitivity) can have a much higher influence in corresponding scenarios (for example dual screen setups with different display types - my CCFL and WCG-CCFL screens have to be calibrated with a whitepoint differing by dE ~ 10 to achieve a visual match). Aside from that the eye has a wide adaptation range regarding white - D65 as the one and only choice is a myth.

Wether the i1 Pro nor the i1 Pro 2 are bad instruments and we are using them ourselves - while considering advantages and disadvantages.

Click to expand...

If I am to trust my i1Display Pro and my friend's i1Display Pro then I have a severe eye problem because both i1Display Pros read a red-tinted gray level as perfectly neutral. I have no idea where you get your information, but tables created by X-Rite do NOT apply well to consumer displays. Read more information of ChromaPure website - i1Display Pros are off by at least dE 4-9. The use of i1Pro had significantly improved the neutrality of my grays. Are my dE 1s and 0.5s that HCFR reports actually 1s and 0.5s after i1Pro profiling? Nope, not likely, but definitely within dE 1-3 and definitely closer to truth than i1Display Pro measurement results. Any pro calibrator with experience would trust i1Pro over i1Display Pro any day... What you say is just theoretical information.

Sailor_Moon said:

The filters of modern colorimeters are are progressing towards the ideal characteristic although they don't fully replicate normative observers. X-Rite provides the i1 Display Pro with a sophisticated correction system that is far more flexible than simply applying a fixed linear transformation at the end of the measurment process. The probes are characterized at factory. Together with the desired CMF (observer function) and EDR (spectral characterization of the display to be measured; generic data is provided) the absolute error is absolutely acceptable for most use-cases.

Click to expand...

Nice post.

When I learned about the operation of spectroradiometers and colorimeters, one of the first thoughts I had was that, in the context of deriving tristimulus values, an ideal colorimeter would trump a spectroradiometer, because (and correct me if I'm wrong), a colorimeter doesn't quantize the spectrum: its filters are physical objects that integrate energy across the spectrum continuously. A spectroradiometer, on the other hand, even a mechanically scanned one as opposed to a detector array, takes measurements at discrete intervals.

Bluesun311 said:

That is not good news. You are going to need to RMA it or return it. I strongly urge you to RMA it through Eizo so you get the same panel back--they will take care of you. Call them on the phone. This restarting problem on a 600 dollar monitor is Bullshit. It will happen maybe 1-5 times a day or more.


This monitor is PWM free above 25 brightness.


Oh. Badass! Thanks for weighing in on this. Now I get to OWN a device to calibrate my displays with instead of renting one. I don't really care if it's not as accurate as something more expensive. It's going to do a better job than I can just fiddling with the controls or copying someone else's values.

Click to expand...

It restarted once and if it restarted because of the power-saving issue then its 100% normal as that is what the power-saving feature is doing. Have you had yours disabled? Its PWM-free with Turbo 240Hz enabled??? Strobing IS PWM AFAIK.

And no, you really do need to rent a spectrometer if you want to be within acceptable error value of dE 1-3, The information he provided is theoretical. I am providing information based on experience. When your whites are pink instead of neutral even though i1Display Pro reads them as if they are perfectly balanced/neutral - you know there is a problem. Renting a spectrometer is not that expensive. You should BUY either i1Display Pro/3 or cheaper ColorMunki Display, but only RENT ColorMunki Photo.

Also, basICColor is miles worse than ArgyllCMS when it comes to software LUT calibration and of course 3DLUT generation.

I am not going to get into an argument over who is right and wrong about software - wait and see, but I strongly suggest you visit AVS Forums Display Calibration section where you can see the results of ArgyllCMS calibration and see discussions as to whether you should rent a spectrometer or just use a colorimeter...

spacediver said:

Nice post.

When I learned about the operation of spectroradiometers and colorimeters, one of the first thoughts I had was that, in the context of deriving tristimulus values, an ideal colorimeter would trump a spectroradiometer, because (and correct me if I'm wrong), a colorimeter doesn't quantize the spectrum: its filters are physical objects that integrate energy across the spectrum continuously. A spectroradiometer, on the other hand, even a mechanically scanned one as opposed to a detector array, takes measurements at discrete intervals.

Click to expand...

And yet results and experience show the opposite picture. All of my i1Display Pro calibrations were off by and it was obvious, but i1Pro profiling completely fixed that issue. Maybe I am unlucky and my monitors and TVs are way off the tables made for them, but I wouldn't take a chance and think that someone else's fit those tables perfectly.

I have no idea where you get your information

Click to expand...

I'm amongst others developing (inhouse) certification tools (co and working for prad.de as consultant/ editor. Apart from that I'm also end user regarding DTP/ CEPS in color managed workflows.

What you say is just theoretical information.

Click to expand...

No offense - but short and simple: No .

When I learned about the operation of spectroradiometers and colorimeters, one of the first thoughts I had was that, in the context of deriving tristimulus values, an ideal colorimeter would trump a spectroradiometer, because (and correct me if I'm wrong), a colorimeter doesn't quantize the spectrum: its filters are physical objects that integrate energy across the spectrum continuously

Click to expand...

This is correct. Our eye doesn't implement a diffraction grating for evaluating the color stimulus ;-).

he use of i1Pro had significantly improved the neutrality of my grays. Are my dE 1s and 0.5s that HCFR reports actually 1s and 0.5s after i1Pro profiling?

Click to expand...

HCFR doesn't assume visual adaptation to the measured display whitepoint (as well as most calibration solutions in the TV area - in contrast to control tools intended for the graphic industry/ ICC workflow). That leads to a problematic evaluation of the grey balance apart from other mostly unwanted effects*. In addition I have tried to explain the consequences of a modest absolute error and other effects that "overshadow" it (in particular: Observer metamerism and omnipresent chromatic adaptation) - but I'm not saying that there is no meaningful threshold for it. However: Its significance is often overrated, especially in the TV/ home cinema community.

It's going to do a better job than I can just fiddling with the controls or copying someone else's values.

Click to expand...

With the i1 Display Pro you will get a very capable and flexible probe for calibrating/ profilating your screen.

*
edit: I'm quite busy at the moment but will provide an example including i1 Display Pro and i1 Pro focusing on grey balance performance until Sunday.

MonarchX said:

Another important thing to note! It seems that for whatever reason, my videocard switches from 1080p PC resultion to 1080p native DisplayPort/HDMI, where only 60Hz can be selected and only 16-235 range can be used. Now, I do use the 0-255 HDMI registry editions/ "hacks" to force 0-255 for HDMI, but I don't think it works for DisplayPort. I am quite certain that THIS is the reason why some people reported a completely washed out picture.

Click to expand...

This makes a lot of sense. I wrote earlier in the thread that i managed to get one of mine in a bugged bright/washed out/pixelated mode that sounds just like what people with problems report.
This occurred when I was testing dp+dvi and switching between them multiple times.

Been running DVI flawlessly ever since on all three.

Now, I do use the 0-255 HDMI registry editions/ "hacks" to force 0-255 for HDMI, but I don't think it works for DisplayPort.

Click to expand...

Unfortunately this is correct. Dynamic range will still be compressed.

Sailor_Moon said:

I'm amongst others developing (inhouse) certification tools (co and working for prad.de as consultant/ editor. Apart from that I'm also end user regarding DTP/ CEPS in color managed workflows.


No offense - but short and simple: No .


This is correct. Our eye doesn't implement a diffraction grating for evaluating the color stimulus ;-).


HCFR doesn't assume visual adaptation to the measured display whitepoint (as well as most calibration solutions in the TV area - in contrast to control tools intended for the graphic industry/ ICC workflow). That leads to a problematic evaluation of the grey balance apart from other mostly unwanted effects. In addition I have tried to explain the consequences of a modest absolute error and other effects that "overshadow" it (in particular: Observer metamerism and omnipresent chromatic adaptation) - but I'm not saying that there is no meaningful threshold for it. However: Its significance is often overrated, especially in the TV/ home cinema community.


With the i1 Display Pro you will get a very capable and flexible probe for calibrating/ profilating your screen.

Click to expand...

Are you talking about each individual's visual adaptation? How can that be taken into consideration? One way or another you to have precise measurements and using i1Display Pro alone is less likely to provide that than profiling with i1Pro or better spectrometer. Just about every TV or monitor review site uses both - i1Display Pro and i1Pro devices to provide measurement results. Its a very popular combo and for a good reason.

Film mastering and game development takes visual perception into consideration and the only way to witness all that was made is to use properly calibrated displays. That is the only way to judge content. If you eyes need special adjustments - you still need to measure the display device accurately and then somehow measure what one perceives as neutrally white/gray, then take that info and apply it to calibration.

Just about every TV or monitor review site uses both - i1Display Pro and i1Pro devices to provide measurement results. Its a very popular combo and for a good reason.

Click to expand...

Yes, but that doesn't change interaction and effects. And I have already pointed out that we are using spectroradiometric measurements with prosumer equipment too.

Are you talking about each individual's visual adaptation? How can that be taken into consideration?

Click to expand...

Regarding which statement from the quoted text? Generally speaking: Research and development in the area of colorimetry science has not finished with the color-matching functions of the 2 degree standard observer defined in 1931 (based on still valid findings of the 19th century, especially from H. G. Graßmann).

To offer a little teaser: Here ist the profile validation of a DCCS calibration (D65, Gamma 2.2) with the i1 Display Pro, validated with an i1 Pro (screen: IPS-panel, GB-r-LED blu). Have a look at the grey balance (DeltaC first page) which is of course still flawless. Presuming that the i1 Pro is a perfect reference (which isn't the case) there is a deviation of ~ dE = 3 regarding the calibration target. A moderate absolute error (still presuming that the i1 Pro sets the reference). HCFR would report an accordant high error for all achromatic colors because it relies on the fixed D65 reference and not the measured display white. But of course a visual adaptation is still perfectly possible (the measured whitepoint even is very close to the daylight locus, dE <= 0.8 ) and would occur. The achromatic colors don't show any significant color shift regarding white. A comparison with D65 doesn't make sense in this case: Where should the adaptation stimulus originate from?

MonarchX said:

It restarted once and if it restarted because of the power-saving issue then its 100% normal as that is what the power-saving feature is doing. Have you had yours disabled? Its PWM-free with Turbo 240Hz enabled??? Strobing IS PWM AFAIK.

Click to expand...

The Power save feature doesn't cause restarting--it puts the monitor in standby.

I've tested this monitor backwards and forwards, if it's randomly restarting I don't know how to fix it in the menu and I'm pretty sure it needs to be worked on. I've had 4 different FG2421's that did that. Most of them didn't do it for a day or 2 and then started more frequently happening.

I wondered if strobing backlight didn't count as PWM... it's certainly "flickering" so I shoulda known and then yep It's PWM free according to blurbusters only if you turn turbo 240 off.

MonarchX said:

And yet results and experience show the opposite picture. All of my i1Display Pro calibrations were off by and it was obvious, but i1Pro profiling completely fixed that issue. Maybe I am unlucky and my monitors and TVs are way off the tables made for them, but I wouldn't take a chance and think that someone else's fit those tables perfectly.

Click to expand...

I was talking about an ideal colorimeter. The more expensive colorimeters have filters which better approximate the standard observer functions, but they don't match them exactly.

You stated "the eye has a wide adaptation range regarding white - D65 as the one and only choice is a myth." How can we figure out how each person's eye adapts to whichever white? Its not possible and we have to settle for a standard. This is what calibration is about. Deciding that some other standard looks more neutral to you and using personal preferences is NOT a calibration.

What do you mean by "measured display white" ? A default, uncalibrated, 100% gray/white point measurement? Are you simply saying that if you were to pick a different standard, one where 6200K is the accurate temperature then HCFR would not be be able to accommodate it? Yeah, OK, but the standard for D65 is 6500K... Why would you calibrate to anything other than D65 6500K if you wish to see games and films developed for that standard? Yes, you would need a different software, but again - D65 is the standard used.

I am not a color science expert, so I am trying not to argue, but as far TV calibrations go for NA consumers - D65 is the standard for white point. There are other standards for other uses, like 9300K standard for Japan, but they are not relevant here... I am aware of different observer standards too, but 2nd degree 1931 standard is the default one. HCFR actually lets you select different observer standards...

I never said that i1Pro is truly reference, but when making a colorimeter profile i1Pro should be selected as reference device, unless you have a better spectrometer or unless you know for a fact that the display in questions behaves 100% like the display that was used to create whichever EDR correction for your i1Display Pro. i1Pro is not super-accurate for emissive measurements, but if you use a working i1Pro + i1Display Pro and calibrate grayscale to very low dEs of 0.5 and 1, then there is a very high chance that those dE values are close to about dE 3, which is still considered accurate. So, when i1Pro reports dE 1, the real dE is likely to be between dE 1 and 3, but when i1Display Pro reports dE 1, then the real dE is likely to be between dE 4 and 9 (which is NOT acceptable). If you want a true reference device - JETI 1211 or Klein K-10 are the right tools, but I would trust JETI 1211 over Klein K-10.

My findings are my own, but even this specific Eizo Foris FG2421 display produces obviously pink grays as it gets closer to 100% white, and yet, i1Display Pro reads a nearly-perfect balance with White LED EDR correction file! Then I take i1Pro and it reads that B is actually 10% below of where i1Display Pro reads it. Then I profile i1Display Pro with i1Pro and re-calibrate. Result? Neutral grays, not pink grays...

I was e-mailing X-Rite about how consumer displays do not behave like reference displays and they confirmed it, saying consumer displays differ and measurement results can differ widely! For some people there is little to no difference between dE 1 and 7, while others can easily tell a difference between dE 1 and 3. Some people are plain blind and could care less! One way or another, you should strive to have the most accurate calibration. Think of it this way - people detect the slightest motions of body language. If you open your eyes just 1mm wider - your expression may be perceived differently. The same applies to colors. Artists who paint often convey a message or want to display something and for that they use specific colors. The same goes for film mastering and game development. If you are to please as many people as possible with a calibration, (which again, is NOT about personal preferences & white point adaptations) then you should strive to have the utmost accurate calibration you can give them. Period. For those who cannot afford a JETI 1211 - i1Pro and i1Display Pro is a perfect bundle.

I was also wondering why Prad.de does not run saturation sweeps and ColorChecker with Skin Tones measurements? They rated Eizo Foris FG2421 as having very poor color accuracy, but in reality, only the outer limits/edges of colorspace (100% saturation sweeps) are very inaccurate. 0-75% saturation sweeps are very acceptable, while Skin Tones are very accurate and so are ColorChecker measurements. I do see that 3D cube is used to assess accuracy, but it should have shown that color accuracy is not terrible, only the limits of color gamut is. There also seems to be little to no weight placed on contrast ratio, which is arguably the single most important measurable quality in a display device. There is also no weight placed on possible refresh rates, etc. All in all, Eizo Foris FG2421 is THE best LCD-technology-based monitor for games, but was rated "Satisfactory"...

Bluesun311 said:

The Power save feature doesn't cause restarting--it puts the monitor in standby.

I've tested this monitor backwards and forwards, if it's randomly restarting I don't know how to fix it in the menu and I'm pretty sure it needs to be worked on. I've had 4 different FG2421's that did that. Most of them didn't do it for a day or 2 and then started more frequently happening.

I wondered if strobing backlight didn't count as PWM... it's certainly "flickering" so I shoulda known and then yep It's PWM free according to blurbusters only if you turn turbo 240 off.

Click to expand...

Were you using the same exact input like DisplayPort? What about the USB cable?

Is there anything else that is different/better in your 5th unit that was worse/different in your 4 previous units? I am VERY reluctant to start the exchange thing because if the left/right edge bleed is a lot worse on others or some other crap will come up then I may go on exchanging them forever and be regretting it...

Is Eizo willing to send out a whole new unit BEFORE I send them the old one back? How long is the turn-around time for these things? Does Eizo pay for shipping both ways?

MonarchX said:

Were you using the same exact input like DisplayPort? What about the USB cable?

Is there anything else that is different/better in your 5th unit that was worse/different in your 4 previous units? I am VERY reluctant to start the exchange thing because if the left/right edge bleed is a lot worse on others or some other crap will come up then I may go on exchanging them forever and be regretting it...

Is Eizo willing to send out a whole new unit BEFORE I send them the old one back? How long is the turn-around time for these things? Does Eizo pay for shipping both ways?

Click to expand...

I tried everything... The very first unit I received was pretty much flawless except for the restarts. I don't know much about Eizo's warranty policies other than the pixel part isn't very good for the Foris line. For this issue, and they definitely know about it, you might be able to get them to take special care, but there's no way I'd do what I did again. I wish I had sent that first unit to Eizo (I got it for 480 dollars...), the problem was I didn't have a spare gaming monitor--now I do, and I would pay to ship a good fg2421 to Eizo to have it fixed so I could be sure to get the same panel back. They may have advanced RMA available but I'd be surprised. There's no freakin way I'd gamble on exchanging that one you got. You said it calibrated really well.

The only other suggestion I have would be to try to put it on a battery backup, I bought an AVR UPS and had it up and running before I got this last monitor. But the seller had 200+ hours on it and said it had never restarted. He agreed to take the unit back if it did, actually. But I don't think the UPS helped, I think this monitor just doesn't have the issue, whatever it is.

How can we figure out how each person's eye adapts to whichever white? Its not possible and we have to settle for a standard. This is what calibration is about. Deciding that some other standard looks more neutral to you and using personal preferences is NOT a calibration.

Click to expand...

No. You don't automatically get a wrong reproduction if deviating from D65 which is used as reference white for many working color spaces like sRGB or AdobeRGB. That's a wide spread misunderstanding. Colorimetry can even adjust to the chromatic adaptation of the eye (some kind of gain control of the visual system) and calculate corresponding colors by transfering the tristimulus values into a cone response domain and back under consideration of different illuminations. This is for example done when building an ICC profile after display profilation because the reference white of the PCS is D50.

Why would you calibrate to anything other than D65 6500K if you wish to see games and films developed for that standard? Yes, you would need a different software, but again - D65 is the standard used.

Click to expand...

Because there is not the one correct whitepoint. A normative reference white doesn't set a fixed whitepoint for reproduction. As I said the human eye has a wide adaptation range (just look at a white sheet of paper under diffrerent illuminations). And then there are the effects of observer metamerism: I'm currently using two different screens (WCC-CCFL, CCFL). They are adjusted to a normlight illumination. When calibrating both screens to a colorimetric identical whitepoint neutrals differ significantly (!). I have to change the whitepoint target for one screen to achieve a visual match. The whitepoints differs by dE ~ 10. The upcomming ISO 14861 (requirements for colour soft proofing systems) takes such effects into account.

but if you use a working i1Pro + i1Display Pro and calibrate grayscale to very low dEs of 0.5 and 1, then there is a very high chance that those dE values are close to about dE 3, which is still considered accurate.

Click to expand...

The grey balance itself won't be improved. It is refers to the display white itself. Have a look at it on page 1. You lessen the absolute error - although it wasn't very high in my example - with an individual correction (with probes like the DTP94 or i1 Display Pro you don't even have to weight the measurements against white but can simply correct via red, green and blue reference). But if going this way it would be better to use the methods intended by X-Rite - charaterization via EDR - if supported by the software. Usually only the generic X-Rite files are available but Argyll for example allows for individual characterizations instead of directly applying a matrix.

For some people there is little to no difference between dE 1 and 7, while others can easily tell a difference between dE 1 and 3.

Click to expand...

CIELAB isn't perceptual uniform. Even with modern formulas including compensations this target isn't achieved. However: When using color distances it must be very exactly pointed out what these values refer to and how the connected conditions were. A common problem because I have seen many dE graphs that are just colorful but don't represent useful information respectively rely on an inaccurate workflow.

Back to the root of the question: Just look at the provided profile validation which shows a good correlation between captured display characteristic and actual behaviour even when using a different probe for both measurements.

I was also wondering why Prad.de does not run saturation sweeps and ColorChecker with Skin Tones measurements?

Click to expand...

The current test set is sufficient but I will include samples with respect to ISO 12646 (or the new 14861 when finally approved) for the next software revision (launch should have already been carried out months ago but there were some problems regarding probe support).

Eizo Foris FG2421 is THE best LCD-technology-based monitor for games, but was rated "Satisfactory"...

Click to expand...

Can't say much to the FG2421 - never had it on my desk.

Sailor, your data was from a Dell U3014, which I'm assuming has a different set of spectral signatures from the Eizo FG2421. How can you generalize from the U3014 to the FG2421?

Sailor_Moon said:

Because there is not the one correct whitepoint. A normative reference white doesn't set a fixed whitepoint for reproduction. As I said the human eye has a wide adaptation range (just look at a white sheet of paper under diffrerent illuminations). And then there are the effects of observer metamerism: I'm currently using two different screens (WCC-CCFL, CCFL). They are adjusted to a normlight illumination. When calibrating both screens to a colorimetric identical whitepoint neutrals differ significantly (!). I have to change the whitepoint target for one screen to achieve a visual match. The whitepoints differs by dE ~ 10. The upcomming ISO 14861 (requirements for colour soft proofing systems) takes such effects into account.

Click to expand...

Yes there is no one "correct" whitepoint, but sRGB and Rec 709 both specify a particular whitepoint: D65, and the assumption is that the observer will be chromatically adapted to that point (bias lights are D65 and surround is neutral).

Perhaps I'm misunderstanding you, but you seem to be ignoring Monarch's observation that his i1 display pro reported quite a different reading compared to an i1 pro, and that when he calibrated with the i1 pro, a neutral grayscale was achieved. You seem to be blaming this on observer metamerism, or image apperance issues (e.g. he had a green curtain in the background). To me, it seems more likely that (his) i1 display pro simply had inappropriate offsets/EDR relative to his display.

Yes there is no one "correct" whitepoint, but sRGB and Rec 709 both specify a particular whitepoint: D65 and the assumption is that the observer will be chromatically adapted to that point (bias lights are D65 and surround is neutral

Click to expand...

Yes, but that doesn't change the fact that you can reproduce perfectly under different conditions. In a typical home environment the color matching conditions aren't defined anyway.

Just one additional note referring to the bias light: If I would use D65 normlight (instead of my D50 solution), calibrating to D65 display white wouldn't result in a sufficient match.

Perhaps I'm misunderstanding you, but you seem to be ignoring Monarch's observation that his i1 display pro reported quite a different reading compared to an i1 pro, and that when he calibrated with the i1 pro, a neutral grayscale was achieved

Click to expand...

The greybalance of my example was very good - there are no significant variations in the achromatic color regarding display white.

Sailor, your data was from a Dell U3014, which I'm assuming has a different set of spectral signatures from the Eizo FG2421. How can you generalize from the U3014 to the FG2421?

Click to expand...

The sharp spectrum of the GB-r solution already is quite disadvantageous regarding "filter failure" and generic characterization data.

Sailor_Moon said:

Yes, but that doesn't change the fact that you can reproduce perfectly under different conditions. In a typical home environment the environment/ color matching conditions aren't defined anyway.

Click to expand...

Yes, I understand you can apply transforms and the like to create a perceptual match, and in the print world, this is an important consideration. But we're talking specifically about Monarch's measurements here. You seem to be suggesting that he is wrong to calibrate to D65. Would you rather he calibrated to a different standard and go through the trouble of applying these transforms to create a perceptual match to the artists' intent? Wouldn't it be much simpler if he instead attempted to adhere to Rec 709?

As for the viewing environment, many home theatre enthusiasts actually go to significant lengths to achieve a correct environment.

Cinemaquest is an example of this.

Sailor_Moon said:

The greybalance of my example was very good - there are no significant variations in the achromatic color regarding display white.

Click to expand...

But this was on a different display. We're discussing the Eizo FG2421 here.

Sailor_Moon said:

The sharp spectrum of the GB-r solution already is quite disadvantageous regarding "filter failure" and generic characterization data.

Click to expand...

Just because one spiky spectral signature (GB-r on the dell) doesn't significantly challenge the i1display pro and its "offsets" doesn't mean a different signature (those on the Foris) won't challenge it.

And you have yet to explain why the i1 pro provided dramatically different readings on his display, which, when adhered to, provided a perceptually neutral grayscale.

I suppose one possibility is that he chose the wrong display type when choosing which offsets to use. Monarch, when you chose the display type, what options were there, and which one did you choose?

You seem to be suggesting that he is wrong to calibrate to D65.

Click to expand...

No on the contrary. It's a good standard solution.

But this was on a different display. We're discussing the Eizo FG2421 here.

Click to expand...

Even if there is a higher absolute error - the measurement would still be consistent in itself in this case. I'm not saying that that a pure absolute error is irrelevant (it should lie within a useful range) - but in the total context its effects are often a bit overrated.

You have yet to explain why the i1 pro provided dramatically different readings on his display, which, when adhered to, provided a perceptually neutral grayscale.

Click to expand...

Two effects: The measurement data itself leads to some kind of self-fulfilling prophecy. If HCFR indicates "too much red" you are tending to see too much red. But more important: It takes some time to adapt. To the dramatic different readings: Consistency in itself is again the keyword: The grey balance in my own example would of course have been significantly worse when evaluated against D65. But this distorts the real behaviour: The achromatic colors don't deviate much from the actual whitepoint.

I'm sorry, I'm not following much of what you're saying here.

Why do you keep bringing up your own data, when we're talking about a completely different display?

Why do you keep bringing up your own data, when we're talking about a completely different display?

Click to expand...

As I said: Even if the absolute error would be higher (which is very unsure) than expressed by the dE ~ 3 regarding white (i1 Pro as reference) in my validation - I guarantee that the measurement will still be consistent and comparable to the provided data.

Sailor_Moon said:

As I said: Even if the absolute error would be higher (which is very unsure) than expressed by the dE ~ 3 regarding white (i1 Pro as reference) in my validation - I guarantee that the measurement will still be consistent and comparable to the provided data.

Click to expand...

Yet we have the following observation:

MonarchX said:

My findings are my own, but even this specific Eizo Foris FG2421 display produces obviously pink grays as it gets closer to 100% white, and yet, i1Display Pro reads a nearly-perfect balance with White LED EDR correction file! Then I take i1Pro and it reads that B is actually 10% below of where i1Display Pro reads it. Then I profile i1Display Pro with i1Pro and re-calibrate. Result? Neutral grays, not pink grays...

I was e-mailing X-Rite about how consumer displays do not behave like reference displays and they confirmed it, saying consumer displays differ and measurement results can differ widely!

Click to expand...

Sailor, are you implying that the i1 pro was reading less accurately in this case than the i1 display pro?

And that the pinkish whites he saw were either confirmation bias, or due to chromatic "corruption" from his viewing environment?

I have seen the stated quote but a remote diagnosis is very difficult. Are there really a different color tints amongst the achromatic colors regarding the actual whitepoint (technically speaking: Strong deviations in the a*b* plane) while using the i1 Display Pro which are absent when using the i1 Pro? A linear transformation - even when weighted with white - could not solve this kind of problem satisfactorily. At the moment I assume the already stated behaviour - but if desired I will try to get a FG2421. Can take some time though.

Sailor_Moon said:

Even if there is a higher absolute error - the measurement would still be consistent in itself in this case. I'm not saying that that a pure absolute error is irrelevant (it should lie within a useful range) - but in the total context its effects are often a bit overrated.

Click to expand...

I kinda honed in on this statement as being the gist of what Sailor was saying when they first interjected to suggest that the low-cost X-rite i1 Display Pro might still be worth using, as opposed to nothing.

Also, if the readings are leading me towards an obviously pink white point, I can adjust for that. I have zero experience with either spectrometers or colorimeters and I wanted one, so I bought it and that's about what I can imagine spending on something that will really just help me decide on the values for R G B gain. I never intended to start an academic debate by asking for advice on colorimeters, but this is H and someone is always going to suggest that you should ONLY use the best. I have no doubt that following Monarch's advice on LCD display calibrating techniques would both save me some money by renting equipment and also probably get me better results, but in general I just hate renting things. If next year I've become even more obsessed with this technical stuff then perhaps I'll consider investing in a proper spectrometer. For now, I'd like the time and space to leisurely fiddle around with the i1display pro, so I can try to say for myself whether or not it was worth the $200.

This is all way off-topic, though. MOST gamers are not going to employ either a spectrometer or a colorimeter to an fg2421 and simply copying TFTCentrals values is still gonna get them a pretty good looking monitor. How many ways are there really to adjust the color on one of these? I don't see much adjustability there but I suppose RGB values already give a very wide range of possibilities. Still... we already sort of know where they should be ballpark. How far away from the TFTcentral RGB gain values did you end up, if I might ask Monarch? They used X-rite i1 Pro + LaCie: Gamma2.0 Brightness24 BlackLvl50 Red100 Green96 Blue90

? I don't see much adjustability there but I suppose RGB values already give a very wide range of possibilities.

Click to expand...

Adjusting the RGB gains is only one small step in the calibration workflow. Without probe I would just rely on the kelvin presets of the screen and choose a value which is comfortable for you. The linearisation takes place in the videocard LUT (software calibration), is loaded from the vcgt at OS start and ensures the desired gradation and a good grey balance. To keep it active in games you may have to reload it. Important color space transformation are only carried out by color managed applications (some screens offer color space emulations that allow for a defined reproduction during video color grading or just games).

For now, I'd like the time and space to leisurely fiddle around with the i1display pro, so I can try to say for myself whether or not it was worth the $200.

Click to expand...

The i1 Display Pro is a good solution and far away from being a bad compromise - even regarding color critcal work. For 200$ you get a light sensitive probe (=> dedicated lense) with well tuned interference filters and solid correction system. The next logical upgrade is the basICColor Discus (totally oversized for your setup).

Sailor_Moon said:

Adjusting the RGB gains is only one small step in the calibration workflow. Without probe I would just rely on the kelvin presets of the screen and choose a value which is comfortable for you. The linearisation takes place in the videocard LUT (software calibration), is loaded from the vcgt at OS start and ensures the desired gradation and a good grey balance. To keep it active in games you may have to reload it. Important color space transformation are only carried out by color managed applications (some screens offer color space emulations that allow for a defined reproduction during video color grading or just games).

Click to expand...

Well, I have some homework to do about LUTs. I have come across this software calibration terminology in regards to the OP display in particular. I do think, from the perspective of a professional display engineer, this "gaming" monitor deserves a once or twice over. There is something really strange going on with the corners of the screen in particular, which I have never seen before. It's consistent across the 5 units I have tested, but slightly difficult to reproduce on pictures, in my posts on this thread along the way there are photographs that detail it.


I guess I would call it dithering.

It's there whether Turbo 240 feature is on or off, but Eizo's backlight strobing technology here is very destructive on static images. Is this dithering as well?
Turbo240 OFF


Turbo240 ON


I find it less desirable than Lightboost in many ways, because it actually degrades the image... but in practice while gaming the other perks of the monitor (notably the 3000-5000:1 CR) along with the potent blur reduction effect remind you there isn't really any better alternative.

I have to assume implementing an 8bit panel with the 10bit functionality via Frame Rate Control (FRC) that almost no gamer will make use of requires a compromise of software and hardware quality VS cost of production. Thus, there is the Eizo Duravision FDF2405, which uses the same Sharp panel (still marked as having possible pixel issues), but offers an included 3D LUT and functionality for controlling a 3d emitter. This makes the Duravision version a really great 3d map editting display, I would imagine, and that is what it is being marketed for. I'm very curious to look at the corners of an FDF2405 and see the severity of this dithering effect. I suspect you will not find it or find it much reduced.

Sailor_Moon said:

The i1 Display Pro is a good solution and far away from being a bad compromise - even regarding color critcal work. For 200$ you get a light sensitive probe (=> dedicated lense) with well tuned interference filters and solid correction system. The next logical upgrade is the basICColor Discus (totally oversized for your setup).

Click to expand...

I really really appreciate the recommendation, I have a number of displays incoming to test the unit on, and then ill do my wife's laptop and see if she notices. I actually spend a LOT of time NOT gaming on this monitor, so a well tuned LUT or ICC profile or whatever (lol I really need to do some homework, I'm light on the technical side) really will benefit me. I'm assuming any regular applications like browsing in Chrome or using a basic photo editor like the one embedded with Windows will still benefit from calibration. Whether or not games can make use of the software tweaking is more or less irrelevant for my needs. Then again, if it weren't possible to make use of a colorimeter in improving the way a game looks, I don't think Monarch would be so enthusiastic about the OTHER i1, which is pretty much 10x the price.

I don't have a FG2421 but I follow the thread since I am interested in zero blur monitors.

Speaking of individual perception - calibration provides a good baseline but is often calibrated with the hardware right up against the monitor in a dark room. Once you start actually using your monitor in your gaming/design "studio" , the ambient light and lighting design of the room comes into play . Brighter environments tend to make screen contrast, brightness, and saturation paler to our eyes. Darker environments result in the opposite. Unless you keep the "studio" at the same dark levels as when you calibrated you are going to see different results. If you allow varying lighting conditions your perception of your monitor's settings will swing as well.

For my main rig I try to keep the lighting levels similar from day to night using floor lamps at the end of the room where the daylight comes in during the day. I face my desk outward from the wall rather than against the wall like a bookshelf (which would act as a catcher's mitt for light pollution and direct light sources otherwise). I also keep a desk/table lamp in line with my monitor array on each end, but not in front where they would be a direct light source on the panel surfaces.

Once I get my lighting environment fairly normalized I will tweak from there. In my living room home theater setup I prefer to be able to have different lighting conditions. My tv allows for four settings profiles so I tweaked those for daylight, evening/room lamp levels, full dark, and a another mode I use for particularly dark movies in the dark as an alternate option. If I use one of the modes in the wrong lighting environment it looks very bad - most noticeably very overly bright or very pale and unsaturated respectively. The same type of thing happens to monitor's levels even with moderate room lighting variances to varying degrees. I thought I should mention all that especially if you are looking for (pun intended) the best real world perception of a monitor's settings to your eyes as a final point and not only to hardware calibrators or cameras.

Sailor_Moon said:

A linear transformation - even when weighted with white - could not solve this kind of problem satisfactorily. At the moment I assume the already stated behaviour - but if desired I will try to get a FG2421. Can take some time though.

Click to expand...

A 1D LUT should be able to fix it.

A 1D LUT should be able to fix it.

Click to expand...

Of course - the calibration - but I was referring to an inconsistent measurement.