The Great Lilliput Calibration Adventure

[Josh Bass]

Gentlemen, I, The Bass, am going to boldly attempt a very unscientific experiment. I will attempt to make my Lilliput 669HB monitor SOMEWHAT ACCURATE for judging color/contrast!

How will I do this, you ask?

UNSCIENTIFICALLY!

I have a lovely SD CRT Sony monitor that I trust to be accurate, so what I will do is film this color chart (which will be nicely lit with a 300w fresnel):

http://www.bealecorner.com/trv900/re...olor-chart.png

with my Sony properly calibrated, feed the signal to both monitors, eyeballing back and forth between them 'til the Lilliput matches the Sony.

I will settle for BALLPARK. I used the Lilliput in the field for the first time the other day and noticed it looks awfully blue compared to real life. . .I was shooting in an office and it was trying to tell me I hadn't white balanced correctly (when I had). Finally got to see the footage at home, and it was fine.

So we will see.

I will report back with results.

I will not do this tonight because I am tired as hell, have to get up early, and now have a belly full of oatmeal.

Good day sirs.

[Josh Bass]

Alright sirs, I have been through the process and taken it as far as I can go.

Here's what I did: hooked up XL2 to the lilliput and my sony PVM14m2u simultaneously via the RCA out on XL2, to a y-cable, then rca out to lilliput into the DVI/RCA input, and also into Sony.

Sony calibrated to XL2 color bars.

tacked the above color chart, printed out, onto the wall, lit it with 300w fresnel.

XL2 set to 3200/indoor WB preset.

Eyeballed color chart to adjust everything, then shot myself, to judge skin tones, then shot a few things around the room to see other real world colors/contrast/brightness.

It's now decently close, though I still wouldn't say perfectly accurate. If I shot and lit something using only the Lilliput to judge my work, I would be decently happy when I finally saw it on a "real" monitor.

I suppose this is going to be somewhat camera specific (again, this test was with an XL2), but here are my menu settings on the lilliput, using the VIDEO 1 input

Brightness - 50
Tint - 85
Saturation 40
Contrast 30

These are the only relevant adjustable parameters on the VIDEO 1 input on the Liliput.

Later I hooked up my blue ray player via HDMI (no camera available right now) to the lilliput, while using the BR's rca out to go the sony monitor, and played through various scenes of "Big Love" on DVD to see how the monitors matched.

HDMI and DVI inputs have additional controls in the form of monitor color temperature (6500K, etc.), and separate Red, Green, and Blue controls. Cycling through the color temps makes 0 visible difference. . .I guess I don't understand this feature.

I played with the RGB settings for quite a while, going through different scenes on the DVD.

It seems you just cannot get consistency with this monitor. . .if your skin tones are perfect, greens and blues are off, if greens and blues are on, skin tones are too red, etc. I have found a decent compromise for everything with these settings:

red 50
green 60
blue 40

Skin tones look pretty dead on, though blues have quite a bit of green in them making them more cyan, but it can't be corrected without throwing everything off.

Also, on the HDMI/DVI, the contrast is different from the VIDEO1 for some reason. . .it was 30 on VIDEO 1 but reads at 73% on HDMI.

Again, this is all using a DVD source/blue ray player to judge. I would like to look at this through a camera and see what happens, but can't til tonight.

[Wayne Reimer]

THANK YOU!!!

I've been struggling since I got the monitor last week with settings, trying to get a reasonable balance and accurate color rendition.

I just adjusted to your numbers and Voila! MUCH closer to accurate color rendition. I'm going to play around with white balance today, but my initial sense is the same as yours; I don't know that they really do anything on the monitor. changing WB on the camera initiates change, as you would expect it to, but on the monitor, not so much.

For those of you with Lilliputs that are struggling with correct settings READ THIS! It may have been unscientifically derived, but they work!

[Josh Bass]

Glad someone else is finding these useful and somewhat accurate.


I still want to hook up the 5DM2 to HDMI and see how it looks compared to the sony/composite, just to make sure what's coming from the blue ray player isn't totally weird.

[Josh Bass]

Alright folks, finally got a chance to play with the 5DM2 hooked up to both the Sony CRT and the Lilliput via HDMI.

I did this in the dead of night/morning, blasted myself with a 300w fresnel and looked at the result on camera, and eyeballed between the two monitors.

This calibration is harder because the 5D won't output to both sources at once, so I have to unplug the HDMI cable watch the CRT, and replug it to see the lilliput. . .can't just swivel my head to compare. Sad face.

Anyway, here's what I came up with looking at myself (skin tone), and what little background was in the shot (my computer/speakers/wall).

Again, these settings are the ones you will tweak via the HDMI input. Why should the be any different from before? I am not an engineer, but perhaps there are differences between the signal a camera outputs via HDMI vs a blu ray player? I'm just going by what my eyes tell me.

bright 50
contrast 15
sat 50
tint 60

red 53
green 58
blue 40

Will try something during the day with the cam to look at other colors, hopefully.

[Josh Bass]

Okay, I think it's time to give up and call it a day.

I settled tonight on


tint 60
sat 50
bright 50
cont 50
r 53
g 60
b 40

It's very difficult to do this when you can't watch both monitors at the same time, the way I could with the blu ray player. I have to try to "memorize" it when I unplug one to view the other.

You'll see that many of the above settings are the defaults (50), so who knows? Feel free to try the different sets of settings and see if one appears more accurate to you. I will say having the contrast way up the way I did on the initial set of numbers makes things appear blown out way before they are. The above appear more accurate.

I also think the monitor simply can't handle the spectrum of colors and luminance the camera can actually see. . .so having shadows not disappear and highlights not blow out at the same time may not be possible.

I will say the color in these settings appears overall in the neighborhood of accurate, even if the luminance can't be trusted.

I wish you all luck.