Luminous Landscape Forum
Raw & Post Processing, Printing => Colour Management => Topic started by: kevs on April 26, 2016, 02:35:13 pm
-
I just bought a new Mac 27" and did the Spyder 3 calibration which I had not done in awhile anyway. It's looks fine, but I thought maybe it looks like a warming filter thrown over a cool blue scene. i.e. Seeing it with the with vs. without calibration at the end button.
Here is the question: How do you know it's accurate? Sure I love Datacolor, but there is no consensus tool and says, yes they did a great job, and your monitor is 100 accurate/ calibrated, correct?
Also, isn't it subjective or not? Honestly, the non calibrated looks nice. It's more neutral tone, but blue, kind of like how I remember shooting at the beach. The "calibrated" is much warmer. They both are fine to look at, and over time my eyes will adjust to either. What is the truth?
-
Here is the question: How do you know it's accurate?
with Spyder of that old version ? most probably not... https://www.drycreekphoto.com/Learn/Calibration/MonitorCalibrationHardware.html
so buy i1DisplayPro
-
Ugh, hoping not to change the topic about equipment wars.
-
What is the truth?
if you print then the truth is the match between monitor and what you print, under the same illumination... now if you don't print at all then it gets tricky
-
Ugh, hoping not to change the topic about equipment wars.
this is not war this is just the fact - DataColor Spyder 3 was/is junk... we hear that with each iteration DataColor makes better and better colorimeters ( 1 -> 2 -> 3 -> 4 -> 5 )... may be so, but Spyder 3 was properly tested by DryCreek and nowhere else their results were disproven since then.
-
Good point Alter. I used to print a lot, but sold my Epson and am not printing for myself currently. As for Spyder, yeah, I'm open to new device, do you have a link to what you would recommend?
-
As for Spyder, yeah, I'm open to new device, do you have a link to what you would recommend?
not me - but it is quite a consensus that i1DisplayPro is the best option (BasiCColor Discus costs ~5-6 times more)... now if you want to save money, OK with slower process and sure that you are not going to buy/use monitors that are can be hardware calibrated (Eizo, Nec, top models from Dell, HP, LG, Samsung, Benq, etc) you can buy ColorMunki __Display__ colorimeter - precision wise it is the same as i1DisplayPro colorimeter, but the speed is crippled by X-Rite and most (if not all) software from the likes of NEC or rebranded X-Rite will refuse to use it with such monitors...
-
with Spyder of that old version ? most probably not... https://www.drycreekphoto.com/Learn/Calibration/MonitorCalibrationHardware.html
so buy i1DisplayPro
I just bought a new Mac 27" and did the Spyder 3 calibration which I had not done in awhile anyway. It's looks fine, but I thought maybe it looks like a warming filter thrown over a cool blue scene. i.e. Seeing it with the with vs. without calibration at the end button.
Here is the question: How do you know it's accurate? Sure I love Datacolor, but there is no consensus tool and says, yes they did a great job, and your monitor is 100 accurate/ calibrated, correct?
Also, isn't it subjective or not? Honestly, the non calibrated looks nice. It's more neutral tone, but blue, kind of like how I remember shooting at the beach. The "calibrated" is much warmer. They both are fine to look at, and over time my eyes will adjust to either. What is the truth?
You dont know if it's accurate. That ultimately requires either multiple different units or a traceable precision unit. There are any number of semi-lab and lab grade units out there if you want (like the Jeti or Konica). If you assemble a grouping of >2 units then you start understanding the differences and errors involved.
No, it's NOT subjective at all. Your biases, vision, and experience may define what you like but that doesn't say anything about accuracy.
-
What's real?
This really is about two different things in photography. The monitor and the printer.
When I make a new printer profile I often check the profile by printing one of the synthetic images of a Colorchecker. I use Photoshop and print it using Absolute Colorimetric. I expect it to match an actual Colorchecker with no perceivable differences when viewed in sunlight. It's best to use a paper that doesn't have optical brighteners or use a profile that was made using M2 measurements. I trim the white edges and compare the two.
Checking a monitor and profile combination really has to be done with an instrument and software that reads out Lab values. Create a monitor profile that is 100 cd/m^2 and D50. You can use the same Colorchecker image, zoom on patches, and read out their Lab values. Except for the cyan patch which is outside sRGB profiled monitors, they should match within a few DeltaEs. Some instruments have software that will do a patch verify so if yours' does you can skip this. Also, the Argyll CMS has some utilities you can use to do this.
-
Create a monitor profile that is 100 cd/m^2 and D50. You can use the same Colorchecker image, zoom on patches, and read out their Lab values.
that does not guarantee that your display (physical device) actually display your Lab values properly...
-
I just bought a new Mac 27" and did the Spyder 3 calibration which I had not done in awhile anyway. It's looks fine, but I thought maybe it looks like a warming filter thrown over a cool blue scene. i.e. Seeing it with the with vs. without calibration at the end button.
Here is the question: How do you know it's accurate? Sure I love Datacolor, but there is no consensus tool and says, yes they did a great job, and your monitor is 100 accurate/ calibrated, correct?
Also, isn't it subjective or not? Honestly, the non calibrated looks nice. It's more neutral tone, but blue, kind of like how I remember shooting at the beach. The "calibrated" is much warmer. They both are fine to look at, and over time my eyes will adjust to either. What is the truth?
It could be that there is a mismatch between the measurement capability of the Spyder 3 in terms of its gamut coverage and your new iMac Display, the latter perhaps being wider gamut than the former. Regardless of whether or not the Spyder 3 is a piece of junk (never used one, but never heard anything good about it either) this factor alone could explain an inadequate profile.
Furthermore, the software should include an internal consistency verification test of the profile it creates. This is called "validation" in some software. Whether it validates anything or not has been debated ad nauseum, but at the least, once the profile is created it has the merit of testing whether with that profile the measurements of the patches the test projects on the display more or less equal the reference values for those patches. If it does, at least you know there is internal consistency in terms of the profile correctly returning the references values of its own construction target.
Because your display is your window into the rest of your workflow on which you will probably be expending a lot on paper and ink, it is likely economic to invest in a quality instrument and software.
-
It could be that there is a mismatch between the measurement capability of the Spyder 3 in terms of its gamut coverage
colorimeter is a like a photo camera (sensor + filters), input device - it does not have gamut in a colorimetric sense... and then because it can't record the emitted spectrum (colorimeter is not spectrometer) it needs calibration data for the software behind it to work properly for a particular LCD panel
-
Thanks guys, great info. Ok so for now, being that I am no longer printing, then in a way it does not matter nearly as much? So this is not subjective, there is ' calibrated monitor', but it takes the best equipment. I just bought a brand new imac 27", so I'm staying with that display.
-
then in a way it does not matter nearly as much?
if you are not printing then what is your target ? if your target is you using the same display then may be - but if you share with smb then think what they use to view your pictures ? for example if you family are all using "iPads" then consider "iPad" as a "print" ???
-
I will have others print. And I make pdfs for clients.
-
I will have others print. And I make pdfs for clients.
so you earn money from photo/processing and one can assume you have a business then no brainer to buy a decent device, will be a business expense, no ? ...
-
I found that the Spyder 3 did not calibrate correctly with the iMac. The calibration did not adjust the screen brightness correctly. I bought the Spyder 4 and it was fine. I have since learnt that the Spyder 3 also works with other calibration software, but not Datacolor.
However the Spyder 3 is definitely a lesser product.
What is real? Get 10 people lined up and ask them the colour of a dress and you get 11 answers. That is subjective.
You can only compare the colour of the print side by side with the original.
Does it matter? If you are doing Art Reproduction or product then probably yes. Otherwise usually not.
-
colorimeter is a like a photo camera (sensor + filters), input device - it does not have gamut in a colorimetric sense... and then because it can't record the emitted spectrum (colorimeter is not spectrometer) it needs calibration data for the software behind it to work properly for a particular LCD panel
I've adopted a policy of not getting into discussions or arguments with screen names, so this will be the end of my participation in this thread. All I'm saying is that some colorimeters are capable of measuring wide gamut displays properly and others are not. Perhaps the Spyder 3 is not, as it is several generations earlier than the O/P's iMac. That's all.
-
I've adopted a policy of not getting into discussions or arguments with screen names, so this will be the end of my participation in this thread.
how is that related to the fact that colorimeter is input device ?
All I'm saying is that some colorimeters are capable of measuring wide gamut displays properly and others are not.
they are all capable - the issue with that Spyder is that it really needs (software that is used w/ it) supplied with proper panel correction based on spectrometer measurements and the OP does not apparently have spectrophotometer to do this... albeit calibration files are available to download (see DisplayCAL database), but then you really need your monitor + your colorimeter for the proper outcome when you are using spectrometer
Perhaps the Spyder 3 is not, as it is several generations earlier than the O/P's iMac. That's all.
wide gamut panels predate Spyder 3 launch... for example NEC 2690WUXi... and then it failed even on standard gamut (see DryCreek test) test miserably... see unit to unit variations in that test
-
now if you want to save money, OK with slower process and sure that you are not going to buy/use monitors that are can be hardware calibrated (Eizo, Nec, top models from Dell, HP, LG, Samsung, Benq, etc) you can buy ColorMunki __Display__ colorimeter
For those on a really small budget, I've found that the ColorMunki Smile is surprisingly good. It seems much better than the preceding units that it physically resembled (i1d1, i1d2 etc.). It's disadvantages compared to the ColorMunki Display are that it is not as good at low light levels (no light capture optics), and doesn't have the display type calibration capabilities (no per-unit spectral curves in it, just two factory matrices). But I have a suspicion that the filters it uses are almost the same as the i1d3.
-
Alter, I earn nothing from processing. I was just doing my own portfolios mostly, and it was a PITA, inks, paper, the time, the space... And with the advent of digital the need is much less. I have tons of expenses to write off including the new ipad pro I got for the 'new' in person portfolio. I am making a box of a loose print portfolio and will have someone print those.
Thanks Bob, so if I'm not printing myself here is the Spyder 3 ok?
BTW when I was printing, I used Colorbyte software and the Spyder 3 and was happy with the monitor to print replication. And I think that was through the G5 to the old imac 27". Now after a day my monitor looks fine, but it does have that warm tone vs non calibrated.
-
For those on a really small budget, I've found that the ColorMunki Smile is surprisingly good. It seems much better than the preceding units that it physically resembled (i1d1, i1d2 etc.). It's disadvantages compared to the ColorMunki Display are that it is not as good at low light levels (no light capture optics), and doesn't have the display type calibration capabilities (no per-unit spectral curves in it, just two factory matrices). But I have a suspicion that the filters it uses are almost the same as the i1d3.
Thanks Graeme, very interesting information. By any chance did you check inter instrument agreement? Considering that the filters are the same as in i1D3 - would it be possible to use random spectral curves of i1D3 and spectral calibration for specific display types in ArgyllCMS?
-
I just bought a new Mac 27" and did the Spyder 3 calibration which I had not done in awhile anyway. It's looks fine, but I thought maybe it looks like a warming filter thrown over a cool blue scene. i.e. Seeing it with the with vs. without calibration at the end button.
Here is the question: How do you know it's accurate? Sure I love Datacolor, but there is no consensus tool and says, yes they did a great job, and your monitor is 100 accurate/ calibrated, correct?
Also, isn't it subjective or not? Honestly, the non calibrated looks nice. It's more neutral tone, but blue, kind of like how I remember shooting at the beach. The "calibrated" is much warmer. They both are fine to look at, and over time my eyes will adjust to either. What is the truth?
A good question, I think.
The reality is that you can't know how accurate your screen is unless you measure the result with a known accurate probe.
And even with that, the appearance of displays with different technologies look different to us, even when they measure the same. Ie. a plasma display has a different feel than an LCD (LED) or an LCD(florescent back light) or an OLED screen or a DLP projector.
The best you might be able to do is to compare your image on your display to the same image on a known accurate display set to the same white point and brightness level.
All I can add is that I use the iOne Display Pro calibrator and the results compare quite well to other displays known, to me, to be quite accurate. (I have used my display, calibrated with the iOne, to color correct movies for theatrical distribution and all looked great in the theater, though each theater looks slightly different, even when they a near accurate)
If you are not printing your photos, then they'll be viewed by others, on their own uncalibrated displays, and everyone will see something different. And there's not anything you can do about that :)
Get the iOne Display and assume everything is close enough.
-
Thanks Bruce, good info. The pucks range from $80, then $150, to $250 for the one you use. I guess I should buy all three and test them out. one day..! ?
-
Thanks Bruce, good info. The pucks range from $80, then $150, to $250 for the one you use. I guess I should buy all three and test them out. one day..! ?
X-Rite has sales on a regular basis recently (shops like B&H)... so you might as well wait and I bet eventually i1DisplayPro will appear again @ <= $160 a piece..
-
Thanks Bruce, good info. The pucks range from $80, then $150, to $250 for the one you use. I guess I should buy all three and test them out. one day..! ?
My test results, with 10k$ Jeti Specbos 1211 spectroradiometer measurement as reference:
(https://dl.dropboxusercontent.com/u/19059944/i1d3vsS52.png)
(https://dl.dropboxusercontent.com/u/19059944/i1d3vsS5.png)
-
thanks Alter, do you think X rite is that much better than the new Spyper, or Munki. those get good reviews too. Have you tested them all?
Honestly, when I was printing with the Spyder 3, the prints looked fine.
-
thanks Alter, do you think X rite is that much better than the new Spyper, or Munki. those get good reviews too. Have you tested them all?
Honestly, when I was printing with the Spyder 3, the prints looked fine.
see... your personal happiness is the most important part... no matter what we say here about Spyder 3 - if you were happy w/ it and if you can get back to that state, forgetting this thread, then please do by all means... otherwise - save $ and buy i1DisplayPro... many years ago when I got my first proper digital camera, I purchased Spyder 2 (sic!) w/o much thoughts and I was happy (ignorance is a bliss) and then I start reading these pesky forums ;D ... yes, I have i1DisplayPro and I am itching to buy Discus once in a while (GAS), even logic tells me that I do not need it really... the gain (if ever) will be miniscule for me, I will not be able see it ever...
-
see... your personal happiness is the most important part... no matter what we say here about Spyder 3 - if you were happy w/ it and if you can get back to that state, forgetting this thread, then please do by all means... otherwise - save $ and buy i1DisplayPro... many years ago when I got my first proper digital camera, I purchased Spyder 2 (sic!) w/o much thoughts and I was happy (ignorance is a bliss) and then I start reading these pesky forums ;D ... yes, I have i1DisplayPro and I am itching to buy Discus once in a while (GAS), even logic tells me that I do not need it really... the gain (if ever) will be miniscule for me, I will not be able see it ever...
I had the Discus, it's not worth it.
-
Alter, thanks. What did you discover between your Spyder and the Display 1? Ignorance was not bliss?
-
By any chance did you check inter instrument agreement?
I only have one ColorMunki Smile, so no.
The thing I noticed about the Smile was that it's measurements were much more independent of the display type or calibration matrix chosen than the i1d2/Display LT/Huey etc., as if the basic filter shape was much closer to the standard observer. This makes sense in a world of more diverse display technologies, and where X-Rite have made the investment in developing and manufacturing accurate filters for the i1d3. (If I were them, I'd be doing something like batch sorting the filter sheets, and using the lesser ones for the Smile instruments.)
Considering that the filters are the same as in i1D3 - would it be possible to use random spectral curves of i1D3 and spectral calibration for specific display types in ArgyllCMS?
I doubt that using any i1d3 calibration would help. The i1d3 calibrations are to correct individual units differences from the perfect, standard observer filter. So unless there was a systematic deviation from the standard observer in the manufactured filter, the Smile factory matrices are likely to be much better than assuming anything from an i1d3.
-
I had the Discus, it's not worth it.
right, but still you have some itch ;D ... it is said that it has that feel of a solid all metal gear ::)
-
That's a pretty chart, czornj. What does it mean? What was being measured and what are the instruments? i1 doesn't define anything. I take it that Spyders are way off warmer, confirming what others have said. Please explain further.
-
Lund: and others
"I take it that Spyders are way off warmer"
This is not just me? This is a known thing with Sypder, even the new 5?
-
I only have one ColorMunki Smile, so no.
The thing I noticed about the Smile was that it's measurements were much more independent of the display type or calibration matrix chosen than the i1d2/Display LT/Huey etc., as if the basic filter shape was much closer to the standard observer. This makes sense in a world of more diverse display technologies, and where X-Rite have made the investment in developing and manufacturing accurate filters for the i1d3. (If I were them, I'd be doing something like batch sorting the filter sheets, and using the lesser ones for the Smile instruments.) I doubt that using any i1d3 calibration would help. The i1d3 calibrations are to correct individual units differences from the perfect, standard observer filter. So unless there was a systematic deviation from the standard observer in the manufactured filter, the Smile factory matrices are likely to be much better than assuming anything from an i1d3.
Thanks a lot Graeme, that makes perfect sense! I'll try to get a couple of Smiles and see if there's good repeatability among them.
right, but still you have some itch ;D ... it is said that it has that feel of a solid all metal gear ::)
Solid, but display type calibration dependent. If the calibration matched the measured monitor it was very precise, otherwise it was deep in the forrest. I suppose it has filters that doesn't match standard observer as good as i1D3 filters.
(https://dl.dropboxusercontent.com/u/19059944/PA271a.jpg)
That's a pretty chart, czornj. What does it mean? What was being measured and what are the instruments? i1 doesn't define anything. I take it that Spyders are way off warmer, confirming what others have said. Please explain further.
I calibrated the wtpt of PA242W to D65 using lab grade spectroradiometer, and then measured it with a couple of popular colorimeters. The i1D3 was spot on, while all Spyders were more or less off, and their inter-instrumental agreement wasn't impressive (to put it mildly)
-
Solid, but display type calibration dependent. If the calibration matched the measured monitor it was very precise, otherwise it was deep in the forrest. I suppose it has filters that doesn't match standard observer as good as i1D3 filters.
now that is the statement ! I am even more curious now !
-
Thanks, that is enough clarification.
-
My test results, with 10k$ Jeti Specbos 1211 spectroradiometer measurement as reference:
(https://dl.dropboxusercontent.com/u/19059944/i1d3vsS52.png)
(https://dl.dropboxusercontent.com/u/19059944/i1d3vsS5.png)
HI Czornyj,
Which monitor model was measured?
Thanks.
-
HI Czornyj,
Which monitor model was measured?
Thanks.
Hi, it was NEC PA242W with BGr LED backlight.
-
Hi,
But the Spyders do not have correction for BGr LED, do they?
BTW, for anyone interest a nice article about monitor backlights
The Evolution of LED Backlights https://pcmonitors.info/articles/the-evolution-of-led-backlights/
-
But the Spyders do not have correction for BGr LED, do they?
but you can supply correction data using spectrometer
-
Hi,
But the Spyders do not have correction for BGr LED, do they?
BTW, for anyone interest a nice article about monitor backlights
The Evolution of LED Backlights https://pcmonitors.info/articles/the-evolution-of-led-backlights/
They have corrections for different backlight types, but they are of poor quality and simply don't match - and even if they would, there's still poor inter instrumental agreement. X-Rite i1Display Pro corrections are often updated, and the sensor has very good match to standard observer CMF, so even if the correction doesn't exactly match backlight spectra the potential error is relatively small.
-
They have corrections for different backlight types, but they are of poor quality and simply don't match - and even if they would, there's still poor inter instrumental agreement. X-Rite i1Display Pro corrections are often updated, and the sensor has very good match to standard observer CMF, so even if the correction doesn't exactly match backlight spectra the potential error is relatively small.
but if you have Spyder 5 (a specific puck) and spectrometer (even a consumer level one - i1Pro/2) you can supply a specific calibration for that specific puck with a specific monitor, no ?
-
but if you have Spyder 5 (a specific puck) and spectrometer (even a consumer level one - i1Pro/2) you can supply a specific calibration for that specific puck with a specific monitor, no ?
Only when using ArgyllCMS, and such correction also introduces some errors (especially in case of LED backlit displays), due to low optical resolution of i1Pro. For optimal results such correction should be done with 5nm (or less) FWHM spectroradiometer.
-
but if you have Spyder 5 (a specific puck) and spectrometer (even a consumer level one - i1Pro/2) you can supply a specific calibration for that specific puck with a specific monitor, no ?
Yes (if the software accepts the custom settings).
-
especially in case of LED backlit displays
and what is so special about LED vs CCFL (or whatever) in this particular scenario
-
and what is so special about LED vs CCFL (or whatever) in this particular scenario
LED/OLED SPD curves have narrow spikes that are difficult to measure for low resolution spectroradiometers. As you can see on my diagram the i1Pro i tested is slightly less accurate than both i1D3. You can also read Graeme's excellent description of this issue here:
http://argyllcms.com/doc/i1proHiRes.html
-
In the case of the OP, who does not print, why bother with monitor calibration at all? The 27" iMac Retina Display is beautiful right out of the box. And in this case, the issue is purely subjective. Whether his monitor shows the proper color of an orange or apple when viewed under controlled lighting situations is irrelevant right? There is no product line requiring color correctness, etc.
My advice is to do the basic software calibration provided in OS X and leave it at that. No need to spend any money at all.
-
In the case of the OP, who does not print, why bother with monitor calibration at all?
Consistently! Displays are not stable devices and should be calibrated and recalibrated on a regular basis. I do mine once a month.
The RGB values I see today must visually match in a year.
-
I understand the need to recalibrate if you need a calibrated monitor. But for someone who likes what he sees on his monitor, does not print and has no real need for serious color accuracy or consistency I don't see why he needs to do anything at all. He mentioned a couple of times not wanting to spend more money than he needs to. I don't think he needs to spend any at all. Not saying he shouldn't calibrate his monitor, just don't see the need.
-
I understand the need to recalibrate if you need a calibrated monitor. But for someone who likes what he sees on his monitor, does not print and has no real need for serious color accuracy or consistency I don't see why he needs to do anything at all.
Again, it's about consistency. The same set of RGB values you view today should appear the same in a year. Unless you edit them.
Your images are a big pile of RGB or CMYK numbers. The only reality to the numbers is the display. That's why calibration is necessary and more than just one time:
http://tinyurl.com/kdgutmz