Advice for a spectrophotometer

Hi all!

I am in search of a spectrophotometer. It will be used to calibrate displays, printer, scanner or whatnot, and all the usual. Especially it is also to step up in color management, soft-proofing code and other color accuracy work in GIMP development (as it’s one of the next big work I planned before GIMP 3.0 release).

Of course I need it to be usable on Linux too with the usual color management “suspects” (DisplayCal/ArgyllCMS in particular, I use GNOME Color too sometimes, which is also ArgyllCMS based anyway AFAIK).

This won’t be paid personally (but on GIMP funds), nevertheless as I see the prices of spectrophotometers can be quite expensive (from a few hundreds to thousands), and since I don’t want this money to go to waste, I’d welcome suggestions from people who know better. What should I get?


P.S.: as a side good point, as I was searching if such request had been done recently on this forum, I realize it could even make the colorhug I bought years ago actually usable! That’s nice as right now this object is mostly an expensive paperweight (though I don’t regret paying it as a support of an OpenHardware project; it’s just sad that this project stopped before improving further :cry:).

I have the color munki design and it works well for this purpose.

I could swear that @ggbutcher built one at home, saving a lot of money in the process.

1 Like

I swore building it… :crazy_face:

Actually, wasn’t so bad a project, used a Raspberry Pi and a monochrome camera. Described here, under “Power Measurement”:

The problem that stopped me was the IR cutoff, eliminated measurements above ~660nm.

Since, I’ve messed around a bit with this:

18 channels covering the range 410nm - 940nm. A bit coarse, but you can plot useful power distributions with it. You’d be hard-pressed to read patches with it, however, getting the patch simultaneously to all three sensors. Arduino interface can take advantage of readily available sketches to read it; I did it the hard way on a Raspberry Pi, still haven’t worked out the channel ordering.

This is what I’d like to use:

340nm ~ 850nm at 15nm intervals, through a single aperture. Pricey, but well-under the lab-grade devices.

Okay, all that said, probably the easiest-to-use, bang-for-the-buck device is this:

The old i1Studio spectrophotometer, originally offered by X-Rite but spun off to a company called Calibrite. In doing so, the price went from ~$470US to ~$569US, serves me right for not getting one earlier. You might find the X-Rite version on Ebay for ~$400US. This thing will read ambient light, patches, 380nm-730nm at 10nm intervals. After this, you’ll be shelling out north of $1000US for a lab-grade device.


I bought a spyder x pro for my windows computer. Previously I had an x-rite, but when Windows version changed they refused to supply a new driver that would work. That really irked me and I would not buy or recoomend their gear if that is how they treat customers. I have not yet tried to calibrate my Linux desktop which I have started using two weeks ago. Any suggestions on the best way to run a spyder or x-rite on a linux machine?

I’m surprised it even needs a driver. It doesn’t with DisplayCAL, since it’s a hid device (unlike the Spyder).

DisplayCAL. Same for Windows, actually. As I understand it, the software that comes with the device is generally not recommended if you care about highly accurate results. Always use third party if you do care, and DisplayCAL/ArgyllCMS is about as good as it gets.

Another thing to note is that the Spyders are pretty much universally derided as being junk, by those that know about these things. Something about the colour filters make them unreliable, especially as they get older, but even when new. I posted some links with further details in this thread:

1 Like

Thanks for your informative reply. I am now dowloading displaycal and I may try a comparison of the early x-rite device and the Spyder device I have. But my nose was really put out of joint when x-rite would not supply new software for windows, but expected me to buy a whole new device.

The ColorHug and Spyder X Pro aren’t Photospectrometers, they’re tristimulus colorimeters.

I seem to remember X-rite partnering with another company to produce their consumer PSMs and colorimeters. Jeti, Colorimeter Research and Klein are also worth a look.

1 Like

Building your own spectrometer is a romantic idea, and a very worthwhile leaning exercise, but the Achilles heal of such projects is calibrating them. Traceable calibration standards and sources are much more expensive than a reasonable quality spectrometer. So unless you happen to know someone with a NIST certified calibration lab, the best way to get an instrument that you want use, is to buy one.


does that make them good or bad?

@Donatzsky I have installed displaycal and successfully used my old xrite i1 to calibrate two different monitors. Using darktables two screen preview mode the screens are very close to each other. Better than I have ever achieved before. Another win for FOSS saving hardware that the OEM will not support.

1 Like

They are less capable devices, and irrelevant to the OP.

Just in case it was not clear, I know that (for the colorhug). :slight_smile:

I only made a remark about this colorimeter as I made a research on the forum (before posting my request for advices, to check if someone else had recently asked the same thing) and I found this other thread where someone was saying that by buying a spectrophotometer, they could finally get their colorhug to calibrate a bit more accurately (as I recall, that is because commercial colorimeters usually have some per-display matrix data to fix calibration, and the colorhug doesn’t have these data by default; so people were sharing their homemade ones on a community database; it required that someone with the same screen than you had a spectrophotometer then shared the resulting correction file).
A spectrophotometer colorhug was in fact planned eventually IIRC, but it seems that it never happened. :cry:

Anyway it was just a fun remark. And yes, now I search a spectrophotometer, not a colorimeter. So I’ll ignore the discussion on the SpyderX as it seems irrelevant to my case.

It looks like it’s not sold anymore. Apparently the ColorChecker Studio (which @ggbutcher also advises) would be the replacement for this too? My web search seem to indicate that ColorMunki became i1 Studio which in turn became ColorChecker Studio?

So bottom line, it looks like ColorChecker Studio is what is recommended so far?

Forgive my ignorance. I didn’t realise the difference. But anyway I got helpful advice despite my stupidity. Thanks for your patience and good responses

Once upon a time I bought at work and used a device and was quite happy for the price. Don’t know about Linux support though…

Hello, I bought the X-Rite iStudio three years ago. It works fine with ArgyllCMS and DisplayCal (flatpak) on Linux. I use it for profiling/calibrating my screen and for printer/ink/paper profiles.

About the price, I bought it for 375€, today I see the average price in France is 550-600€!! Mega inflation… :frowning:

Yup. Looks like marketing :slight_smile:

Take a look at the left-most column on this page:
where you can compare accuracy and speed for some devices.

Have fun!
Claes in Lund, Sweden

Here’s a good DisplayCAL tutorial, if you haven’t seen it already:

1 Like

Do you mean it’s actually not that good? In your earlier comment, you were saying it was. I’m lost!