3D Calibration

[Anonymous]

Hi Gang,
I was wondering if anyone has done a 3D calibration with a meter such as the i1Pro. It seems hanging the glasses in front of the meter would yield some inconsistent results because, from what I understand, viewing angle is important. Also, what did you use for patterns?

Maybe, Gregg or Michael or some other ace calibrator could post a quick tutorial.

Thanks.

Topic

[Gregg Loewen]

yup….totally true.
you also have to consider left eye and right eye. I am finding up to a 4-6 dE variance between left and right eyes…not to mention between different pairs of glasses.
So basically…calibrate the black level while wearing glasses and shoot the WB point through the glasses.

[Anonymous]

Do you think there is a chance that someone may develop a fixture or appliance with the 3d lens for the i1Pro?

[Gregg Loewen]

doubtful.
For which manufacturer of glasses? All the mounts would be different.

[Gregg Loewen]

what I could imagine for the future….the eye glass manufacturer might supply a standard “optical glass” that we could use instead of the glasses. This is what the D cinema guys are currently doing.

[Anonymous]

Wouldn’t it be better if we just did away with 3D?

[Anonymous]

What about patterns? Are the 2D patterns from say, a Sencore VP-403 suitable to use for 3D cal?

[Gregg Loewen]

We really need 3D patterns, which we currently dont have. the 780 will have them very soon (so if you bought a 780, we can get you updated for this).

You can use 2D patterns if you can force the display into 3D mode.

[Anonymous]

Something else to keep in mind when shooting a meter through the glasses, I discovered readings were affected by the angle of the glasses in relation to the meter. As I pitched the glasses my readings changed. Red (x-axis)seemed to change the most. I now make it a point to make sure the lens of the glasses is on the same plane as the lens of my meter.

[Anonymous]

Actually, the QD780s currently have 3D patterns on them. My MP500 that was magically 🙂 transformed into a QD780 came back with the 3D patterns. Not sure if they’re available for download yet – I haven’t looked.

Best,
Greg

[Anonymous]

That is what I thought might happen. I think we need a more professional way to go about this.

[Anonymous]

I too am looking for a THX standard of calibration for the 3D displays. Since the display manufactures are not using a 3D glasses standard this makes things even more difficult. Right now my only option would be the glasses for that display with my I1pro as most glasses don’t fully cover my K10 meter. Also how does the shutter rate affect the accuracy of the meter?
Most displays default to dynamic mode when switched to 3D those that offer the selection of Movie mode seem to carry the grey scale settings from the 2D to 3D mode but require resetting of brightness and contrast to compensate for the ND effect of the glasses. I know we would like to get the most accurate reproduction of the 3D picture as possible but it seems the 3D effects become stronger the higher the contrast ratio and more saturated the colors. What is the process used on the displays used in creating the master videos?

FYI
Need to have more hands on with the AVFoundry video forge pattern generator. But it seems that you can set it for various 3d outputs. Do not know yet if it applies to all the test patterns.

[Anonymous]

Has anyone been using actual 3D patterns for calibration and, if so, is there any advantage to them over calibrating with 2D patterns (using 2D -> 3D mode, when available)? (I see Tom Huffman’s AccuPel generator now has 3D patterns.)

I realize for some TVs that don’t have a simulated 3D mode, e.g., VT25, they would be advantageous. But, I’m wondering if there are any real benefits for using the 3D patterns insofar as the results of the 3D calibration itself.

Best,
Greg

[Anonymous]

I use the VF to give me the 3D patterns.

I have done one samsung that upconverts but its 3D was so awful looking that nothing would have helped.

[Anonymous]

You can also use the 3D Lumagens (mini , xs, xd) for 3D test pattens…. You just need to feed them a 3d source to get them into 3D output mode, then you can turn on the test patterns which are then output as a 3D signal.

[Anonymous]

Read the Joe Kane article in the new issue of ‘Widescreen Review.’ He raises many unique points and details his approach with the 3D patterns designed for the Video Forge.

[Author]

Replies