I've been looking further at Plasma calibration, and it really is very, very difficult due to the operation of ABL, which can't be turned off.
(ABL is Automated Brightness Level/Control - all plasmas have this, and it can't be turned off).



The above image shows the RGB Separation for a Panasonic plasma profiled yesterday. Profiling was with a very small patch size - approx a 20th screen area, in an attempt to limit the ABL effect.

What you can see is that the ABL kicks in immediately, not after a given level.

The 45deg line shows what is expected - the RGB plots show the variation of RGB values for the grey scale compared to the expected.

So, this means that a patch that is just R, G or B has a great saturation than the equivalent grey patch, meaning that saturation will always be greater for a given value of input compared to the expected value based on the display's grey scale (which is a simple mathematical combination of RGB for a given colour space).

The end result?

A plasma can only ever be close to accurately calibrated by a 3D LUT that can deal with the poor RGB separation - but even that cannot totally override the ABL operation.

Hope this helps with some level of understanding of Plasma calibration issues.

Steve

Yes I see this with my plasma. How would this affect the calibration results. Does this mean we can not use the Quick Profile version.

Tony Manolikakis

Hi Tony, that really depends on your specific plasma...

If the RGB Separation is as bad as above, the calibration via Quick Profiling will not be ideal.
But, because of the ABL the truth is that NO calibration can actually ever work correctly, so a Quick Profile will be better than nothing, and probably not hugely different than a full profile, to be honest.

What you may find is that after making a LUT with a Quick Profile the Primary saturation will still be slightly too large, so a manual adjustment of the display saturation will probably be all that is needed.

But, remember that as soon as you use a different size patch the calibration will change... so what is really a 'calibration' for a plasma???

Steve

So are we to conclude that a plasma calibration will be inaccurate (perhaps only slightly but still...) regardless....?

Tony

Hi Tony,

Yep, the reality with plasmas is that because of the ABL operation they are actually impossible to calibrate accurately... a 3D LUT will be 'better' but not by much in all honesty.

There is no way around this.

But, a Quick Profile should get you 'close enough' for most real footage. Where you will see problems is when you are working on very bright footage.

A client we calibrated recently then went on to work on desert and Arctic wildlife footage, which was next to impossible to grade correctly due to the ABL - and there is no way around that!

Now that we (Light Illusion) have a lot of experience with plasmas we no longer suggest them for grading... but that is us being super critical.

Steve

Steve;

have you had experience with the Pana BT300 series "broadcast" plasma's? they cost a fair bit more than the consumer version, i was wondering if they actualy deliver a closer to useful tool for the coin?

This fence is getting a bit old really... must be all the sitting it's had to endure...

Hi Dermot, yep, I have 'calibrated' them too - no different at all as they still have ABL.

The only difference is HDSDI input capability, and 'supposedly' being closer to Rec709 by default.

I really can't recommend any plasma for grading work - the ABL operation just negates them from being suitable IMHO.

One more question if you have the patience... i was told to look at a range of Samsung Plasma's that run to about 150Cdm, so about 50% brighter than the Panasonic at full throttle.. the theory presented is that by running a LUT to pull the screen back to callibration the ABL is largely negated as that is happening above where the screen would be set to display...

this from a local "expert"

<sigh>

It does not seem to be all that logical really, but before i dismiss the thought completly, i thought i'd ask...

All a bunch-0-bunk?

d

Hi Dermot, sorry to say who ever told you that has no idea what they are talking about...

ABL doesn't cut in above a certain level, it is active from the off, as you can see in the piccy in the first post (click on the piccy to see a larger version). You can test any plasma by doing a Quick Profile and looking at the RGB Separation graph.

If ABL cut in at a certain level the effect would be very obvious. Making it active from the off reduces the impact the ABL has on the viewer.

So If You are building a New Suite. What would be the Ideal monitor setup at this moment....I was about to pull the trigger on the BT300 but after reading this I'm not to sure now...I shoot a lot of F65 , Epic & C300 and output is almost always are for Broadcast or Web (REC709). I have a few feature projects coming up and want the ability to work with & monitor for theatrical release. So if you are building the ideal suite, using Davinci 9.....What monitor set up would you go with at this moment..
Thanks for any help & guidance.
Bobby Holbrook
The F65 Project

Hi Bobby, as is said in this thread, we do not recommend any Plasmas - the ABL function juts makes them unusable for grading.

A good LCD is still the best choice, and at the moment the Penta HD2 Line is probably the best all-round performer, especially at the price.

See: http://www.hd2line.com/

We have tested many similar displays, such as TVlogic, Eizo, etc, and the Penta's do come out on top.

Cheers,

Steve

Thanks Steve!!! I have put in a request for a quote, but i wanted to see what your thoughts were on the FSI 24 w/ new 12bit firmware update?
Thanks Bobby
PS. Still using those EOS Gamma Curves i bought years ago from you..!!

Since this post was first made FSI have adopted LightSpace CMS for all their calibration - as a result their displays are now very accurate, and we do now recommend them!

************************************************************************************* *****************************************

Unfortunately we have not had good experience with FSI displays. The bit depth is not an issue - many 8 bit displays are perfect - and most home systems are 8 bit anyway, so it is important to remember that and make sure images look good on such TVs...

That is NOT the same as plasmas and ABL though!

What is important is colour, and the FSI we have tested and calibrated have not been that impressive at all.
But, others do seem to like them - maybe we have just seen bad examples?
I doubt it.

Steve

hi steve , do you think that a 3D lut is kind of waist on a plasma and it could be manually put ""as close"" as a Lut could do (knwoing the limitations because of the ABL) ?

Did you ever post about how manually calibrate a plasma using LS ? i know about the cal images first to set contrast etc... but how about RGB ?

ps: off course i mean the panasonic pro line (personally i have 12th gen (58PF 12)

thanks

Hi Gabriele, I would use a Quick Profile with a small patch to limit ABL, and then make a 3D LUT from that.

The results will be as good as you can get...

Steve