- This topic has 88 replies, 17 voices, and was last updated 3 years, 11 months ago by
.
-
Posts
-
…
Right now we do not know if:
-it is going to be supported soon by HW vendors (mostly NEC & Eizo, since other vendor like Dell or Viewsonic or HP is unlikely to support it)
-it is going to be supported by ArgyllCMS, but we can speculate that it is going to be supported soon
-it can or cannot use CCSS spectral corrections, since they are more or less portable for i1d3, so most people do not need an spectrophotometer
-it is accurate or not (out of the box or CCSS corrected)
so until these uncertainties are known… it is not wise to buy (or advice to buy) a Spyder X “right now” .
I absolutely agree that with the “box” the calibration will be more accurate only for the instrument that has the spectral correction matrix for this monitor, for this reason, different software with different instruments shows different results.
SpyderX Pro on Amazon
Disclosure: As an Amazon Associate I earn from qualifying purchases.Hi,
I just bought the new SpyderX to replace my old i1d2.
I only followed the recommandation of the most famous french guy that reviews calibrators:https://www.guide-gestion-des-couleurs.com/test-spyderX-pro-datacolor.html
For a very long time the i1d3 was his best buy recommandation until the review of the new SpyderX.
Can’t wait to try it with DisplayCal.
Let me summarize:
-If you use a matrix correction to correct a colorimeter to some “reference” device (be it truly a reference or not), then all corrected devices should agree in white (your results) but an i1DisplayPro is going to have the upper hand because in low light measurements it is better than the others or at least better than Spyder4.
-colorimeters that have its own sensivity curves stored in firmware could be spectrally corrected to some “known sample”. Such correction is accurate if that sensivity curves are close to your desired observer and accurate toactual device response (and stable after a few years) and spectral sample is a good representative of a familly of backlights.
-“out of the box” means out of the box, no correction. “Spectrally corrected” is to use vendor or community spectral samples.
PA242W is a GBLED, and its owners have spectral samples for it (from NEC and from community). Devices in i1d3 could measure that backlight with just a generic GB-LED sample without issue, without the owner needing to buy or rent an i1pro2, and they can do it even better than i1Pro-family.
This means that Czornyj test is valid at least for those units (small population sample).
Uncertainties arise with unknown backlights but community helps in this task… here an i1Studio-i1Pro2 is going to help, no doubt of it… but community spectral samples too.-Right now we have a family of well tested devices (i1d3) which supports: argyllcms, vendor software HW calibration (i1displaypro), spectral corrections from vendor & community and very good performance once spectrally corrected without needing to buy/rent an i1Studio-i1Pro2-JETI for most common monitors which owners asked themselves “Do I need a colorimeter” (which is a small set of all monitors owners, but with HUGE relative populations of certain displays).
Let me re-translate my own sencence: I DO NOT need an i1pro2 for measuring with DisplayCAL a current graphic arts widegamut if I own an i1d3 (I cant even outperfom that i1Pro2!), although it is very recomendable to have spectro at hand for other displays like most TVs & laptops.-If Spyder X wants to be an alternative (and maybe it can), it needs to offer the same as i1d3 or more.
Right now this is unknown (and this is a fact), so until it is supported by vendor HW calibration solutions and Argyllcms and we can use vendor & community spectral corrections (3 conditions) it is absolutely pointless and even un-ethical to advise to buy a Spyder X colorimeter.
In a few months, maybe earlier, when vendors update their HW calibration packages (NEC, Eizo, and maybe Benq and LG, also Spectracal, Lightspace, Basiccolor and all those players out there) and ArgyllCMS fully supports Spyder X we can re evaluate this situation and say “this device is awesome” but right now is a big “NO” with bold letters.You are now talking only about the fact that the “boxed” version 1d3 has a good correction matrix to the PA242W monitor with a specific type of panel and backlight, but completely forget that there are a number of other monitors with which the device cannot cope with the matrices that are in the “boxed” version. It is for this reason that there is a whole series of devices, stitched with additional matrices adapted to specific monitors or their huge range, but in the latter case the cost of such a set of tools (soft + device) increases several times in relation to the basic “boxed package”.
Others “boxed package”: BasICColor display SQUID 3, HP DreamColor (DCS), CalMAN for FSI with i1 Display Pro OEM Bundle, ChromaPure with Display 3 PRO (Rev B) Colorimeter, … see https://chromapure.com/newgear-new.asp“There are a few takeaways in this data.
- It is clear that the i1d3’s color accuracy exhibits considerable unit-to-unit variation. Look at meters 13 and 14, which are incredibly accurate. Then compare that with meters 3 and 9, which are not so much. It shows a dE variation of about 3.0, which is not trivial. One of the consequences of this is that meters 3 and 9 would greatly benefit from a spectroradiometer correction, even from an i1Pro 2. Any benefit for meters 13 and 14 would be negligible.
- The i1d3 struggles when measuring Samsung Quantum Dot, and to a lesser extent Samsung LED, plasma, and CRT.
- The i1d3 is the most accurate reading green and blue and the least accurate reading white and red.
- Although the average i1d3 is less accurate than the i1Pro 2 and the worst considerably less accurate (not surprising), the best i1d3s are actually a little more accurate (very surprising).“
Wait for the complete testing of the device discussed in this post, and then discuss it! In the meantime, this is a conversation about the “skin of a dead bear.”
I think that you do not understand how i1d3s work, or what is “out of the box”.
i1d3s do not store several boxed matrices for several displays, they are not K10As. They just store an individual (or batch sample) colorimeter sensivity curves (after some factory correction).
You provide a GENERIC spectral sample, then you read INDIVIDUAL i1d3 observer and compute a correction ON THE FLY (at colorimeter initialization time by measuring application) for THAT colorimeter.
This is why i1d3 cracked the market several years ago: a non fading filters, reasonably accurate 2 degree observer colorimeter that can be software corrected with GENERIC spectral samples.
And since its a HID device, we do not even need Xrite support at all (although if you want HW calibration you’ll have to play the game vendor wants you to play with limited free options like some Dells or HPs).
They do not need that specific model “fine tuned” corrections you imply, you can supply an i1d3 with NEC’s GBLED GENERIC sample, with Xrite GENERIC sample (1nm and reference devices) or with community provided 3nm-10nm GB-LED samples (munki-i1pro accuracy) and all they will provive ~1dE between them on the same colorimeter. That is why a lot of people find amazing that cheap device called i1d3.
They can be used too with partially wrong corrections and keep some degree of accuracy, like you can see if you switch GBLED or WLED PFS corrections with RGB LED spectral generic samples… as long as these partially wrong samples are not signficatively different in wavelengths were i1d3 observer performs worse than reference 2 degree observer.
Of course we can discuss if actual colorimeter observer is close to what it is stored in firmware, but all test imply that is reasonably close to that.All i1d3s are equal (“pro” variant, munki is HW/firmware limited), they just have different unlock codes to serve Xrite i1d3SDK licensing policies (which are not customer friendly).
Take a look in ArgyllCMS code, all are the same. That ID / unlock code differentiation allows some applications to offer or do not offer some spectral samples in software (again, take a look on Calman folders).Now to Chomapure test.
-That is an “uncorrected test”. A true “out of the box” test with no correction applied.
We know that because <~4dE is the typical error you get in a calibrated widegamut display when you run a WP measurement with an uncorrecetd i1d3. As easy as each one reading this lines running dispcal -v -r with and without -X option.
Since it’s an uncorrected test it can show all kind of inter instrument variation and its usefulness is limited to evaluate i1d3 performance in unknown TVs or laptops… nothing more. This is useful for TV calibrators that do not whish to spend money in a K10A ( aka Chromapure customers) … not so useful to most people that need to calibrate their monitor (because they are a subset of all monitor owners/users).
If it were an individually correcetd to some reference device test then WP error should be close to zero, no matter what device you corrected: i1d3, spyder4/5, i1display2, Spyder3… etc, that’s the point of individally corrected devices!
Devil is in the detatils, or in the missing parts of an explanation/test made by a vendor.-Czornyj’s test was a GENERIC corrected test. He used vendor supplied GENERIC corrections for i1d3 and Spyder4/5 (Measure with SpectraviewII “colorimeter measure” sub window, copy coordinates, thet set them as preset in Multiprofiler for plotting).
After reading i1d3 observer, and with the same vendor generic spectral samples A SPECIFIC correction what computed FOR EACH of the two i1d3 in the test… and inter instrument variation was gone. In i1d3s GENERIC spectral sample means SPECIFIC INDIVIUAL correction.
Spyder 4/5 do not behave as good as i1d3…that means that HW is a poor performer or that vendor supplied corrections are bad… or both.-A fairer test would be to use the same PA242W calibrated with a JETI, like in Czornyj’s test… but using DisplayCAL/ArgyllCMS to measure the display.
That way Spyders4/5 could use other sources for corrections. We know that those Spyders there are worse than i1d3s, extremely poor performers, but with just Czornyj’s we do not know “why”: HW or SW.-That fairer test DO NOT help you if you own a display with HW calibration, because you have to use certain applications (…and that leads us to current issues with PA271Q and the lack of W-LED PFS corrections in SVII v1.1.39, but this is another story for another thread), so older Spyder models (4/5) are not an option for those users.
Your last paragraphs means that you did not carefully read what I wrote specifically about “Spyder X”.
We do not need some testing to be completed (although all 3rd party independent test are welcome), we need application support.
Without application support SpyderX is just an expensive paperweight. Without application support SpyderX is not even an option for customers so if you really *need* to buy a colorimeter *right now* (March 14, 2019), avoid SpyderX go for an i1d3 (the one you need because of Xrite licensing policy). If you *do not need* a colorimeter *right now*, you can wait and see what happens. That is what I wrote in my previous messages.K10-A on Amazon
Disclosure: As an Amazon Associate I earn from qualifying purchases.…
If it were an individually correcetd to some reference device test then WP error should be close to zero, no matter what device you corrected: i1d3, spyder4/5, i1display2, Spyder3… etc, that’s the point of individally corrected devices! …Yes! Did I write something else? How the amendment is formed is just understandable to me. I did not give other, best in terms of measurement devices intentionally. The fact that i1D3 more accurately measures the shadow than Spyder4 know. I just noticed that the Spyder makes measurements with the accuracy declared by the manufacturer and is sufficient to meet the ISO 12646: 2015 standard. Anything better can remain so for perfectionists.
…
If it were an individually correcetd to some reference device test then WP error should be close to zero, no matter what device you corrected: i1d3, spyder4/5, i1display2, Spyder3… etc, that’s the point of individally corrected devices! …Yes! Did I write something else? How the amendment is formed is just understandable to me. I did not give other, best in terms of measurement devices intentionally. The fact that i1D3 more accurately measures the shadow than Spyder4 know. I just noticed that the Spyder makes measurements with the accuracy declared by the manufacturer and is sufficient to meet the ISO 12646: 2015 standard. Anything better can remain so for perfectionists.
How? Matrix correlated to a reference device? Yes, like all other devices out there when measuring white point.
The key point that needs to be understood is that i1d3s do not need that to be very accurate. They just need an spectral sample for a backlight technology close to the one your display has. A custom correction will be computed on the fly for them.
Here is where Spyder4 & 5 fails miserably when using vendor provided spectral samples (4-5dE)… hence they are inaccurate devices if you want to use then with vendor HW calibration solutions like Spectraview II (Czornyj’s test). The population of such displays (HW cal) is small compared to all displays out there but limited to the population of displays owned by people that need to calibrate their displays (the population that matters for the subject of this thread), then their number became statistically significant, even huge. Hence these severe innacuracies from Spyder4/5 are not a minor issue for a huge % of their past potential buyers.
Would Spyders perform better with a bigger and better catalog of spectral samples like the ones from DisplayCAL’s community? That is what Czornyj’s test did not cover as I wrote. Even if they improve their poor performace using DisplayCAL instead of vendor software it does not help people with displays that support HW calibration.
I think that this is easy to understand. That is what I wrote previously.
Preliminary evaluation of the SpyderX by Graeme Gill, developer of ArgyllCMS:
https://www.avsforum.com/forum/139-display-calibration/3050976-datacolor-spyderx.html#post57785616TL;DR: It doesn’t seem like the i1 Display Pro is at all threatened in a technical sense.
On this occasion, my five cents on the SpyderX:
Funnily enough, I find the device extremely hard to recommend at the current price point (around $ 170 for the Pro), specifically because the ColorMunki Display exists and seems to be a superior device in every way at an actually lower price point (around $ 150).
I’m not even sure yet if I would outright recommend the SpyderX over its predecessor, the Spyder5 – while the latter holds no candle to the SpyderX in terms of speed (Spyder5 and its predecessors are just so painfully slow), it seems to have better field upgradeability due to the possibility to use spectral calibration like the i1D3 line of instruments (even if it’s not quite as accurate). My final verdict on the SpyderX is still pending though, because I have not yet received mine.
If the price point were closer to the Spyder5, it would be easier to see why someone would choose it over the Spyder5 (speed), although I really wonder if people make sound financial decisions by buying Spyder devices – the i1D3 exists since 2011, the prices have been relatively stable, and if you have bought every generation of Spyder released since then (i.e. 4 and 5) in the hopes it would match the i1D3’s performance, you have spent already about 1 1/2 as much as the cost of the ColorMunki Display at the time, and two 1/2 times as much if you now also purchase the SpyderX, and you’re still not there.So, if anyone was looking for a reason (actually, two reasons, price and performance) to buy a ColorMunki Display over a SpyderX, there you have it.
Funny you bring this up, because I recently bought one. Simply put, because I am new to the whole thing and not many reviews exist. My initial thoughts were based on basic google searches. The Colormunki didn’t really come up, so it was spyder5 vs i1D. Digital art is part of my work (just a part, and nothing print related) and so far I have always used iMacs and not calibrated them. I got an external display now though and I wanted to calibrate it because it was so visibly different (I expected this). When I got the Spyder5 I started reading these forums and others like luminous-landscape to uncove the unfortunate reputation the spyder5 had. I settled on the Colormunki after more reading and then foud out about the SpyderX. There is not much info about it, but it is much newer and appears to work with a glass filter like the Colormunki. I found one review here, which was very positive, also compared to the Colormunki.
I have returned the Spyder5 and I will compare the SpyderX to the Colormunki (which I will get later this week). If you had written this post a couple of days ago I probably wouldn’t have bothered with the SpyderX. But as ‘new hardware’ it seemed more appealing and I couldn’t find any bad reviews, other than ‘the spyder5 was inferior to the x-rites, so this probably as well”.
When will you receive one? Will it be short term? I’d love to read more critical reviews / experiences about it.
edit: How do you know it is less accurate if you don’t have one?
I found one review here, which was very positive, also compared to the Colormunki.
Problem with many colorimeter “reviews” floating around online is that they often do not actually review the colorimeter – they care more about user experience (which naturally tends to make the vendor software a main focus as it’s the interface through which to interact with the device) and “pleasantness” of the calibration/profiling result, and while I understand that these are points people may deem important because they are more easily accessible than a more technical in-depth look at a colorimeter, they also do little to address the main point of owning a good colorimeter over a lesser one, which is to ensure a level of colorimetric accuracy so you can employ color management to match (within limits of the used hard- and software) what is shown across different (display) devices, or at the very least to reproduce the intended color on-screen that is described by the colorimetric definition of (say) an image’s color space, or to get a close look at a device’s color reproduction.
An actual colorimeter test would not only compare it to other colorimeters (ideally using the same software, to have feature parity as far as possible, so this would almost by design exclude the respective vendor software), but also to at least one (ideally high-end) spectrometer on various display types, to assess the achievable absolute colorimetric accuracy. Even my own testing will not be a full-blown review, because I simply lack the time and resources to do it properly (I’m expecting to confirm Graeme’s preliminary findings, which I have linked to above).So, TL;DR: If a colorimeter review doesn’t (at least) compare the actual colorimetric performance to that of other devices in the same class (and ideally with different display technologies), and ideally to some reference (like, a high-end spectrometer), it lacks a lot of the information required to make a reasonably sound purchase decision, especially when “aftermarket” software is then going to be used, which acts as an equalizer for purely software features.
(Note that I’m not saying the SpyderX, or any Spyder, is a “bad” device, just that there seem to be better, and in case of the SpyderX, at least currently, even cheaper choices available).
When will you receive one? Will it be short term?
I’m hoping this week, but I have yet to receive shipping confirmation (which actually surprises me a bit). We’ll see.
edit: How do you know it is less accurate if you don’t have one?
Because Graeme Gill, the developer of ArgyllCMS, has already looked at it and posted his preliminary findings (see link I posted above).
