Welcome to our community

Be a part of something great, join today!

Log in
  • Hey all, just changed over the backend after 15 years I figured time to give it a bit of an update, its probably gonna be a bit weird for most of you and i am sure there is a few bugs to work out but it should kinda work the same as before... hopefully :)

Calibration

Christoffer Glans

Christoffer Glans

Well-known member
Joined
Jun 11, 2007
Messages
4,993
Reaction score
36
Points
48
Age
42
Location
Stockholm, Sweden
If I get a Decklink 3D extreme and use Resolve Lite, how can I calibrate the monitor in the best possible way? I got an LG RGB-LED IPS screen I use for my grading and my clients are not the type that can afford post production facilities, so I need to calibrate it in the best possible way with the resources I have.
To this date I've been using a MXO2 mini with it's calibration, but they have issues that aren't resolved and I don't see them fixing it any time soon.
So I'm gonna sell it and by black magic stuff instead.

So, what is the best way to calibrate the monitor for using Resolve (PC when it's out)?
 
AN LED is impossible to calibrate - this is not good enough even for cinephiles, let alone graders.

Get a plasma, or a DLP front projector, or better still a CRT. These are all cheaper than LEDs. LED technology is designed around "enhancing" images, not accurately portraying Rec 709.
 
LED is impossible to calibrate... wow!...

Rob, are you referring to LCD?...

For rec709 I can calibrate any LED monitor. I only need a good processor + good colorimeter. But processor can cost more than the monitor itself.
 
This LED is used for professional still photography, I use it for video instead. Saying it's impossible to use for grading, well... it's wrong.
And even if it's not as good as CRT's, how many people today watch stuff on a CRT or plasma compared to Lcd's? If this screen enables me to see a 10-bit image with 1,07 billion colors, then it's more then enough to use for grading, if it's calibrated correct.
Which is what I'm asking about.

And please, show me a CRT with 1080 25p capabilites under $2000, please do. My primary work is for television beside this type of work and at the station we have a CRT for such work... I did work on it, calibrated and it looked like crap in people's home.
Just saying, there's a difference between reality and the science behind it. Yes, a LED LCD isn't better at colors then a CRT, but then again, CRT's are pretty much dead.

So, how can I calibrate my Resolve output to this screen in the best possible way? I got a Spyder3, but i assume that the HDMI output/input can't take advantage of it?
So to stream out a calibration image, is there a way to do this from Resolve to make it right? Does BlackMagic have some kind of software calibration for their outputs?
 
Chris, I'd recommend having someone calibrate your monitors to Rec709 specs. Might not be as expensive as you'd think to have a professional come do this. I have my TV Logic calibrated professionally. Need to have it done at least once a year. And if they can't get your monitor within specs, at least you'll know where you stand.

Tools like Resolve Lite are great, but you need to have confidence that what you are monitoring is relatively accurate, otherwise a lot of time is wasted.

Where are you located?
 
LED uses dimming technology to "enhance" contrast - which means that an image with some white in it displays differently than the same image once that white object has moved out of frame. This should not be the case.

AN LED IS AN LCD, just with a LED backlight. LED backlights can either be side (actually often less good than traditional fluo backlights in terms of screen uniformity and color fidelity) or locally dimmed. The locally dimmed models create strange halos around bright objects (which is a misrepresentation of your picture) and the side-lit models turn all the lights up and down to "enhance contrast" based on the percentage of dark pixels it finds at any given time (which is also a misrepresentation of your picture). LED suffer the same off-angel color skewing as LCDS, unless they have extremely high pixel density and you are very, very careful to watch them, only head-on. Your typical Best But "LED" TV is a 1080p 50 or 42 inch affair with very low pixel density and very poor screen uniformity that does so many automatic picture enhancements that it is indeed impossible to calibrate. You might get/calibrate a proper reading from a square on screen that covers 30% of screen space, but that same level of RBG, if it were enlarged to 50% of the screen, say, might appear much brighter as the LEDs are cued to brighten up for a bright scene, so unless all you grade is the same perfect squares of the same full screens your software uses to get a calibration read off your puck/sensor/whatever your screen will almost always be lying to you. I also find it highly unlikely that you could even get your display to give you a correct gamma curves as LEDs are not marketed to cinephiles and usually have fewer controls to affect curves, and have totally incorrect primary gamma curves out of the box.

$2000 is pretty damn cheap for a calibration display, but any Panny plasma from 2010 would do a better job - and are less than $1000.

Your standards of professional are obviously much lower than mine. It is not "wrong" to say that an Best Buy LED TV is an inadequate display for color grading or pro photo work. No truly knowledgeable pro would use it for that.

Also, you are not seeing a 10 bit image, you are seeing some feeble attempt at 10 bit processing - no LED monitors are actually able to display 10bit, and most are not even able to fully display 8 bit RBG. Do you actually believe that you TV has 1 000 000 000:1 contrast ratio too?

If people have crap TVs at home that is not your problem. You are grading to a standard, and all programming is graded within that standard, and people are used to watching things graded in that manner.
 
Rob what you say is completely wrong. Like what you think about calibrators...

I think you need a lot to learn
 
Christoffer Glans said:
So, what is the best way to calibrate the monitor for using Resolve (PC when it's out)?
Click to expand...

Perhaps you'd be better off purchasing a separate, dedicated monitor with an SDI connection that has been pre-calibrated to REC709. If you can't afford 10 bits, have a look at something like Panasonic BT-L2150. Monitor calibration and profiling is a tricky and complicated business. You might have to spend as much or even more in calibration tools than in the monitor itself.

Also, if you have contacts with Swedish Television, you might consider getting a broadcast monitor that could be calibrated cheaply or for free in their monitor calibration facility.
 
Jose Lomeña said:
Rob what you say is completely wrong. Like what you think about calibrators...

I think you need a lot to learn
Click to expand...

No offense Jose, but I used to run a post house with over $500 000 of annual billings. What I say on this topic is not completely wrong. Do you see LED TVs at *ANY* top flight post house? No, you don't. Do you see Panny plasmas? Yes, you do. Must be you know much more than the likes of OffHollywood and even Company 3, let alone me. Must be you're some kind of genius.
 
I can't find yor point rob. Are you speaking about led or about cheap tvs?.

The first post is about computer monitor, not tv monitor.

I think near all photographers are workig with led monitors, like eizo, or nec...contrast is not the only thing in is this world. Led monitors are more stable than plasma and smaller. You cant work with a 50" monitor always.
Dinamic contrast is only for tv, and you can disable this feature but you cant disable power management in plasma.

I work In my grading suit with led monitor, lcos prjector, crt and panasonic plasma. The most accurate is the led. Calibrated correctly of course.

Why the most expensive monitors and grade 1 are lcd led?, like autocue, tvlogic, or cheap nec, eizo, dreamcolor...?. Because with led you can set white point natively, and can make gamut of color separately. You cant do this with the old backlight technology
 
Christoffer Glans said:
I got a Spyder3
Click to expand...

Low-end colorimeters are best suited for closed loop environments where color accuracy is verified through subjective analysis, not for matching to broadcast specs. That's why they're marketed toward hobbyists and amateur (and a subset of professional) photographers. The proper gear costs as much as a secondhand car or much more. If that's cost prohibitive for you, hiring a specialist who owns the gear would be a better alternative.
 
Mike Lary said:
Low-end colorimeters are best suited for closed loop environments where color accuracy is verified through subjective analysis, not for matching to broadcast specs. That's why they're marketed toward hobbyists and amateur (and a subset of professional) photographers. The proper gear costs as much as a secondhand car or much more. If that's cost prohibitive for you, hiring a specialist who owns the gear would be a better alternative.
Click to expand...

Completely agree. Dont waste your money on low-end meters, have a specialist with professional equipment do it. Give David Abrams a call at Avical.

Avical - Video System Specialists
Mobile - 877.428.4225, ext. 102
Email - dave@avical.com
Website - www.avical.com
 
I would generally agree, but there is an exception to this rule, recently released, which is the new i1Display Pro. Reports show negligible inaccuracy - the main thing is to be patient and take very long readings which Lightspace will automatically have you do. Also, you can rent higher end ones by the day in many large cities - you can get by calibrating every 6 weeks, and once you know the software, it doesn't take long to perform the calibration.
 
I know still to motion is like apples to oranges, but I as a color management pro in stills used the Eye One products for over 12 years for print. Their devices are the defacto color measurement devices in that industry. My largest client was Spiegel catalog, followed by Bloomingdales, Bergdorf Goodman, Vogue, and some of the most renown photographers in the world, relied on me, and I relied on my Eye One devices. From monitoring on set all the way to the printed page, I could maintain a 95% match.

So why would this not work for motion. My Eizo displays a beautiful neutral image, and each and every client (directors, DP's, editors) are amazed by the images of their work on my screen. Last year I taught at a post production conference in Ireland. The first speaker was a color management expert from the UK. His info was terrific, but he delivered his presentation off a laptop connected to a HD projector. The images displayed he had to apologize for, because what looks proper on his laptop, displayed badly on the LCD projector.

After his presentation, I had 5 minutes, I spent that time profiling the projector (with an Eye One device). I started my program with the RED sizzle reel provided by Ted Schilowitz. My colors were dead on, skin tones correct, colors popped, and I explained that if you spend the time to produce a work of art, why kill it by presenting it in less than accurate color. To me, if these other more expensive devices work better that my Eye One, I bet you could not tell the difference, my opinion. But unless someone shows me a side by side, it's all talk.

Von
 
Von Thomas said:
I know still to motion is like apples to oranges, but I as a color management pro in stills used the Eye One products for over 12 years for print. My colors were dead on, skin tones correct, colors popped, and I explained that if you spend the time to produce a work of art, why kill it by presenting it in less than accurate color.
Click to expand...
I don't doubt it looked good on your projector. The question is how it will look in the real world, in XYZ color space for D-Cinema, or Rec709 color space for broadcast video. Where I see the cheaper devices fall down is in issues like grayscale and the basic gamma curve. I have seen things go terribly wrong in some cases (not with me personally, but at a few LA-based companies). And I do know of cases where set monitors, editing-room monitors, and final-color monitors were so far off, the producers and directors were really thrown for a loop. With only internal standards, this kind of thing can get out of hand very quickly.

I know of a very big, major film where the director was so concerned about the home video "look," the facility went out and bought a dozen different flat screens, just to reassure him that the film would translate well to many different kinds of flat screens. None of them precisely matched the pro monitors and projectors in the main room, but the engineers were able to tweak them so that at least the director's fears were addressed. I have also dealt with a few A-list directors who approved the transfer in our room, then took a check disk or tape home and called to complain that they didn't like the way it looked on their home monitor. It's not fun to confront a director with the idea that maybe their set doesn't tell them the truth. "But I watch lots of movies on my monitor, and they look great! My movie looks like crap!"

I have also worked at facilities that used both the (very expensive) Minolta and Philips Color Analyzers -- and these did not agree. So this is a case where even costly pro devices each adhered to a slightly different standard. Which do you believe? We literally averaged the two and hoped for the best. It worked for 25 years that I know of.
 
Von Thomas said:
From monitoring on set all the way to the printed page, I could maintain a 95% match.
Click to expand...

Hi Von,

95% subjective accuracy is the best you can expect to achieve when simulating CMYK or multi-ink output on an RGB display. Your soft proof is not as color critical as your source-to-print color transform which is happening under the hood without regard to your display. In video production you treat your display as the substrate. You calibrate it to the spec. In the case of broadcast, you have no control over the calibration of the systems used to view the final product. Display accuracy and consistency are critical.

You mentioned using Eye One products. If you're creating printer profiles, you're using a spectro, not a colorimeter. That's an important distinction because spectros are usually more accurate than colorimeters.

In regards to A/B comparisons, mass produced low end colorimeters are more likely to show variation between same model units right out of the box than their more expensive counterparts, they're less durable (more susceptible to handling damage), and could quickly become less of an investment than an expendable when you compare the cost of regular calibration to the cost of replacement. They're simply not the right tool for professionals in color critical environments.
 
i1Display pro works really well, and many units have been extensively tested with less than 0.5 Delta variance between them (Delta 3 is limit of average human vision, so 1/6 of that). It is very new, released June 2011. I agree Spyder 3 is a joke

To the other poster, sorry, I thought you meant LED TV, (not computer monitor). However, Flanders Scientific has avoided LED, and I think for good reason, but there probably are some reasonable LED compu monitors out there. Still, viewing angles are an issue with anything other than front-projection, CRT and plasma. You never know whether you are grading out how you've leaned in your chair, and that idea scares me.

Any great CRT or top-quality plasma can be properly calibarted using a 3D LUT - the underlying screen tech is excellent
 
You must log in or register to reply here.
Back
Top