[stama]

A few days ago the makers of the Spyder colorimeter announced a new product, SpyderX.

Unfortunately, there is not much info on the hardware capabilities in their press release. Or I don't know where to look. Does anyone know more?

Looks like they might have given up on the organic filters in the Spider 5, so maybe it's finally a cheaper viable alternative to i1D3. A test using a spectro would probably reveal how accurate it can be, but I don't have one. Anyone willing to take up the challenge?

 

[gwgill]

Unfortunately, there is not much info on the hardware capabilities in their press release.

From what I've seen so far, it seems to be a completely new design. Roughly the same form factor as the Spyder 5, but the plastic work is completely different in detail. (It has a nice high quality feel to it). They've finally discarded their discrete band cut filters & sensor design and are using an AMS AS7264N integrated sensor and added a light concentration lens. Silicon based interference filters, so it should be relatively stable. Integrated 16 bit A/D. Accuracy will depend on how close to the standard observer the filters are, how consistent AMS's manufacturing and QA is, and how good Datacolors production calibration process is.

 

[stama]

Thank you for the in-depth analysis! I take it you've already poked inside one?

That sounds very promising, indeed! They also advertise the measurement speed to be much improved (admittedly vs. their old Spyder 5), as well as a better low light accuracy. If they manage to make these probes consistent in quality, this could be really great. We'll need to see how they measure different types of displays and what kind of spectral corrections are provided. From perusing the various reviews of their calibration software that appeared on the net it does not seem they have anything for WRGB OLED displays, or any kind of OLED displays, unfortunately. And the laptop OLED displays that got announced by Samsung and will be used in various units starting this April are going to be RGB OLED displays, not WRGB OLED displays like in the TVs.

 

[Vicent]

They've finally discarded their discrete band cut filters & sensor design and are using an AMS AS7264N integrated sensor and added a light concentration lens. Silicon based interference filters, so it should be relatively stable. Integrated 16 bit A/D. Accuracy will depend on how close to the standard observer the filters are, how consistent AMS's manufacturing and QA is, and how good Datacolors production calibration process is.

Spec sheet says under spectral responsivity graphs (page 9, bottom):


"The accuracy of each channel count is ±25%"


Is this common? It looks like it will need a extreme fine tune.

 

[gwgill]

Spec sheet says under spectral responsivity graphs (page 9, bottom):
"The accuracy of each channel count is ±25%"
Is this common? It looks like it will need a extreme fine tune.

I'd imagine it's pretty common. That's why any serious user of this part needs to calibrate them.
Of more interest than the absolute sensitivity calibration, is how good the X, Y & Z filter shapes are, and whether they are smoother than the JENCOLOR filters used on prior devices (i.e. whether the AS7264 is accurate for light sources with peaks < 10nm).

 

[Anger.miki]

I asked DataColor for the SpyderX specifications some days ago but still no answer.

Anyway, is it better of i1 Display Pro? In which particular feature/capability? Thanks.

 

[Vicent]

Anyway, is it better of i1 Display Pro? In which particular feature/capability? Thanks.

Untill it gets ArgyllCMS support, skip it.No ArgyllCMS' CCSS support for vendor or "portable" community corrections... I would skip it too.


Once SpiderX gets that functionality you can start thinking about which one is better.

 

[David Susilo]

Are they going to make a Spyder TV package again with this sensor?

 

[gwgill]

Here's a preliminary critique of the SpyderX perfomance:

Pluses:

Accuracy: Seems OK, but I haven't done any detailed cross checking yet.

Fast: it always takes a little under 1 second per measurement.

Repeatability seems good. I get consistent black measurements (but see below for caveats about black accuracy).

Minuses:

Doesn't seem to have the spectral sensitivity calibration data that the Spyder 4 & 5 have, so can't be updated for new displays by distributing display spectral samples. It instead has 4 calibration matrices for General, Standard LED, Wide LED and GB LED displays.

Low light level resolution seems limited. The Y channel values seem to be in 0.013 cd/m^2 steps. Maximum light level is theoretically 870 cd/m^2, although that could be increased by reducing the integration time.

Black point stability seems fairly poor. When cold I get a Y value of -0.014, but when it warms up to 25 degrees C I get black Y value of 0.054 cd/m^2. The sensors auto zero function doesn't seem very effective in combating this.

Summary:

The sensor they have chosen is struggling for light. They have a concentration lens, have set maximum gain (x64) and are almost using maximum integration time (714 msec, 716.8 is the max.), but the sensor count limits the dark resolution. This would probably be a particular problem with displays like OLED that have very low level blacks.

Conclusion:

Cheap and cheerful.
It doesn't seem like the i1 Display Pro is at all threatened in a technical sense.

 

[Rolls-Royce]

Once SpiderX gets that functionality you can start thinking about which one is better.

I think @gwgill just answered that...

 

[Vicent]

I wonder how bad that Wide LED matrix will be when used against QLED (Quantum Dot, like PG27U/SW2700PT) or W-LED PFS because they are "current" technologies... or if that matrix is computed by Datacolor using their old RGB LED 10nm White sample and it's useless at all. Difficult to know.



The lack of that feature is a "NO" for me but YMMV.

 

[gwgill]

I wonder how bad that Wide LED matrix will be when used against QLED (Quantum Dot, like PG27U/SW2700PT) or W-LED PFS because they are "current" technologies... or if that matrix is computed by Datacolor using their old RGB LED 10nm White sample and it's useless at all. Difficult to know.

Depends on how accurate the underlying spectral sensitivities are. If they were perfect 1931 2 degree observer curves, then their 4 calibration matrices would be identical, and it would be perfectly accurate on any type of display. (Eyeballing their 4 matrices in the unit I have, they are quite similar, but not identical).

Of course if you don't think the 1931 observer is an ideal reference and would prefer something more modern (like the 2012 proposed observer), then you will be out of luck with a matrix calibrated colorimeter, and a spectrally calibrated colorimeter or a spectrometer is what you need.

 

[stama]

That doesn't sound good.

The fact that the spectral transforms are built-in and can not be changed is a big downside. It's as if they don't care the probe can become useless for backlights with spectrums that don't fit those four built-in.

I guess this is really built only with photographers and their PC monitors in mind. In their Elite version they add the ability to deal with video world color spaces, so I was expecting support for WRGB and RGB OLEDs at least, but this doesn't seem to be the case from what you've noticed.

 

[Bachelor]

Has anyone been successful using the spyderx with HCFR? I installed the updated Argyll 2.1.1 and updated the driver but HCFR doesn't see it. I am using a Spyder5 just fine. What am I doing wrong?

 

[Vicent]

Has anyone been successful using the spyderx with HCFR? I installed the updated Argyll 2.1.1 and updated the driver but HCFR doesn't see it. I am using a Spyder5 just fine. What am I doing wrong?

AFAIK HCFR uses ArgyllCMS' code, not ArgyllCMS set of applications (like DisplayCAL does). That means that HCFR needs to be compiled with newer ArgyllCMS functionality.


P.S: Since you own a Spyder5 ¿why would somebody want to buy an SpyderX?

 

[fcb]

Spyderx is now supported by both HCFR and Lightspace. Oddly, I get slightly different readings from each.

 

[gwgill]

Spyderx is now supported by both HCFR and Lightspace. Oddly, I get slightly different readings from each.

Are you sure that you're not just comparing different calibration pre-sets ?

[ The numbers should be very similar unless you are comparing low level values, where the ArgyllCMS black calibration may result in slightly better accuracy. ]

 

[CyberPaul]

Spyderx is now supported by both HCFR and Lightspace. Oddly, I get slightly different readings from each.

which version of HCFR? I just tried 3.5.1.4 and the SpyderX is not detected

 

[CyberPaul]

which version of HCFR? I just tried 3.5.1.4 and the SpyderX is not detected

Found it: 3.5.2! Here: https://sourceforge.net/projects/hcfr-2019/

 

[CyberPaul]

Found it: 3.5.2! Here: https://sourceforge.net/projects/hcfr-2019/

Is it required to install ArgyllCMS SpyderX driver? Any hint? I'm on Win10