Monitor for photo editing


(Isaac Ullah) #21

@paperdigits I have the same series monitors on my work computers. I agree that they are very good for the price!

(Pat David) #22

I’ve actually got the older “Premier Color” version of this same monitor

(John) #23

Dell ultrasharp are a cheapish way of obtaining a decent monitor. Even the factory calibration isn’t too bad but can be a lot better. Set way too bright for photo work though as they all are now to make dynamic range sound better.

I wondered about Dell PremierColor but last time I looked the hardware calibration isn’t supported by open source. So far on 2 Dell monitors I have been able to get well within the better level of colour rendition standards via a LUT. I don’t print so just look for good coverage of sRGB.

4K though - I have wondered but only to obtain more screen at around 100dpi. I don’t have my nose 6" from the screen. Currently I use a 27" U2713HM at 2560x1440. No suitable 4K monitors so far when I looked a month or so ago.

Setting wise my monitor nearly passes this.
It needs a bit of imagination to pass the pairs of extreme black and the white squares. It’s not currently calibrated though just set as close as it can be without a profile. When it was calibrated all was ok except for the highest white.

(Damon Lynch) #24

Whether you print or not is kind of irrelevant. The question is how important it is to you to be able to view on the screen a somewhat faithful approximation of what you can see in the world around you. Colours found in everyday reality – those found in objects likes clothes, for instance – can sometimes look awful when viewed in sRGB. If being able to view those colours is not tremendously important to you and sRGB is fine then no problem.

(John) #25

I have come across a number of people who view aRGB on a monitor and comment about the colours but they all also print. There is a bit of a catch with it. They can see it but when it’s posted on the web it will be shown in sRGB in the vast majority of cases - an understatement. If it’s in any other format and includes a profile it will be transposed to sRGB. If it doesn’t have a profile it will be shown as sRGB and that can have some weird effects. It’s the same for prophoto and others.

Clothes are a particular area. I’m sure fashion people think about sRGB when they design the stuff. I understand that they have moaned about this aspect. There are 2 other factors that have an effect as well. For that sort of work the camera ideally needs profiling. Then comes mixed fabrics that look to be the same colour. That has always been a bit of a pain for wedding photographers even in film days. They may look the same colour to us but wont to a camera. The same can be true of clothes full stop.

Maybe one day sRGB will be updated with something that has a deeper colour depth. It might not be aRGB. The wiki points out some interesting factors about aRGB even that the usual way of showing it exaggerates the difference. It also mentions that it is intended for CMYK printing really not RGB. Note that it also contains a mistake.


(Damon Lynch) #26

Isn’t it more important what you can see before worrying about what others can see? It’s an unfortunate fact that everyday objects from the world around you cannot always be represented sRGB. There is no way around that.

None of this discussion avoids the fact that exporting images as sRGB for the web is necessary. But I don’t see why that necessity should dictate everything else in a workflow.

(John) #27

This whole area is confusing. aRGB’s main purpose in life is higher colour saturation. The reason for that is printing which is using reflected light and has a restrictive dynamic range. On a monitor a similar effect can be obtained in sRGB by punching up the saturation. This is probably what commercial printers do to shots when people have them printed. It’s what my little canon post card printer seems to do too. My laser is calibrated to print sRGB.

There is also an argument that aRGB uses larger colour steps than sRGB. That implies that both can show colours that the other one can’t. That can probably be cured by using pro photo. I hear that editing that for printing on an aRGB monitor is real fun 'cause all of the colours that are produced can’t be viewed.

There is a good page on gamuts here
It also shows the frequency of the light at the various points around the usual full gamut curve and the human eye’s response curves. To much blue light can actually damage eyes in the near UV range and at some wavelengths we need an awful lot of it to even see it. The link also covers Pointer’s gamut that tries to account for the fact that we see and photograph reflected light. It mentions what I think will happen eventually. We will see things like RGGB monitors or maybe even more colours making up the back lighting. GG= 2 shades of green. It’s already happened/happening in architectural lighting.


(Damon Lynch) #28

I think you’re overcomplicating things. If you don’t want to see on your own computer all that your camera can record of the reality around you, then stick to sRGB. If it’s not that important to you then there’s no need to stress out about it. If you do want to see it, implement a workflow that uses a higher gamut. It will cost more money and probably require some new hardware, but if you’re paying attention to colour you’ll notice the difference.


If we are talking about hardware here, my monitor’s sRGB coverage is extremely poor. I don’t mind though because I cannot afford a better one. Then there is profiling, color management, soft proofing, color theory, color perception, industry standards, etc. @Ajohn, it might be better to discuss one aspect at a time and/or create a new topic or series of topics.

(John) #30

@afre It’s too complicated a subject to discuss on a forum. It’s also a difficult subject to get to the bottom of. Mainly because the general view of aRGB is that it’s bigger so must be better. The visual spectrum gamuts are shown against is also misleading. We really don’t want to see intense levels of all of it. Pointer’s gamut is a much better view of what we need to see and nothing currently covers it.

I did mention a clue as to what it’s all about really. Bigger colour steps and more saturated colours meant to be viewed via reflected light via a printed photo. That is less efficient than a monitor - net effect colours are more subdued and contrast reduced hence the bigger steps etc. People still push things to get round that when it’s printed from aRGB. Some go prophoto and buy a printer with a larger than normal gamut to make it worth while. That takes saturation out of the aRGB gamut.

:wink: Generally I don’t bother getting into this subject. There is a rudish expression that goes something like urinating into the wind. That sums it up because there is plenty of info on the fact that the gamut is bigger but virtually zero spelling out why, what it’s intended for and the implications.

Many monitors will calibrate to cover most of sRGB. All of the ones I have owned since I did more photo work have been up in the high 90%'s. Dell is a good bet but many will do it at all sorts of prices. It’s also wise to go for one with an IPS panel. The main problem is the need for a colorimeter. I started with used one. The risk there is that the filters will be aged but it’s still likely to do a better job than can be done by eye. I now use a colormunki. Nothing special. Software to drive them isn’t a problem. DisplayCal will drive the majority of them. It’s a pro package and can even be used beneficially with spectrometer types that cost £1000’s and even more with similar software.That means that a calibration takes longer and using it effectively isn’t just click a button and done.

There is a monitor review site that I find useful but like many funding is a problem so the coverage isn’t complete.
This time I have bought one that he hasn’t reviewed. I also did a Belinea years ago when LCD panels really did have problems. It was a TFT. Much better than what preceded it. It was a lot better when it was calibrated and gave well into the 90’s coverage. When I view shots adjusted on that now I’ll probably see over sharpening. Colours no problem and maybe more punch than I would usually give them. Here’s a really bad example

I don’t think it’s possible to buy a monitor these days that wouldn’t show so much over sharpening providing it’s set up reasonably well. Coverage of that one from memory was very low. I thought it was ok and had been using it for years.

If some one wants 100% exact sRGB coverage I’d guess they would need to buy a hardware calibrated monitor. Say the Dell at >£1000 or a similar NEC, Eizo etc. Really colour temperature, brightness and no serious colour errors are more important than coverage and the odd few % loss on that doesn’t really matter. Often the loss will only be some fraction of 1% or a little more after calibration anyway.

Should mention that I post some junk on flickr. I even take 2 shots at times in case I run out of dynamic range and these may finish up on there.



my two displays are getting old and now I’m searching for something new :grinning:. Currently I own two older Eizo displays (S2431 and S2410). I plan to replace them with a single display with higher resolution (in terms of ppi) and color gamut. Quite an important aspect to me is uniformity in color and brightness.
The latter constraint is what actually keeps me away from dell, which is recommended quite often here and elsewhere. I returned my dell xps15 (4k display) due to a quite visible color shift (reddish on the on side, greenish on the other), which the support considered to be acceptable. In the office we got two dell displays which also show similar color shifts. Several reviews seem to confirm my impression that uniformity is not the strength of dell displays. Maybe I’m a little over-sensitive regarding uniformity, but it drove me nuts to see this color shift on the xps15 :stuck_out_tongue_winking_eye:

A nice display, which also defines the upper bound of what I’m willing to afford, is the Eizo CS2730. It is quite pricy, but offers a lot.
A major question is: Can I actually make use of features like 10bit color and hardware calibration?

10 bit per color seems to be basically supported by nvidia graphics cards and proprietary drivers but what about applications like darktable? There were some hints (from 2016) on the darktable mailing list. My interpretation was that there is some lack of 10bit support on the gtk and/or cairo level. However, such things may change quickly or even did already?

Hardware calibration is another topic. My understanding is that the calibration part of a monitor profile is actually loaded into a lut in the display instead of the lut of the graphics card. The process of hardware calibration seem to involve proprietary software, which is not running on linux or at least in wine.

thanks for comments :grin:


Nothing changed, it’s still not supported. And without a computer supporting 10bit monitors I won’t be able to help fixing that.

(Morgan Hardwood) #33

Probably not. Last time I tried, KDE looked broken. This was 2 year ago now, hopefully things have changed.

Probably yes, but you will probably need to use it through Windows, either a real one or in a virtual machine. When I had an Eizo I used its built-in calibration and profiling but I had to use Windows to get the resulting profile (which I could subsequently use in Linux without Windows).


Does “not supported” mean, it will just not use the 10bit, or will it look awkward?


I have no idea.


Thanks for your replies @houz and @Morgan_Hardwood.

The hardware calibration thing sounds at least quite usable with minimal windows interaction ;).
I wonder if such a hardware calibration process only measures properties of the monitor or does it measure the properties of the combination of monitor and the graphics card the monitor is connected to. If it does not involve properties of the graphics card (or other components), it should be possible to calibrate the monitor from a mac book (i wouldn’t have to purchase a win license :wink: ) and use the profile on the desktop.

Regarding the 10bit stuff… mh. Maybe it is interesting to try out. However, reading a little more on the web is quite consistent to houz’s latest comment :wink: Maybe we should raise some money to fund a 10bpc-enabled computer for houz :grinning:


Thanks but no thanks. I wouldn’t even know where to put it. :smile: