|
|||||||||
|
Thread Tools | Search this Thread |
February 21st, 2007, 05:25 PM | #1 |
Regular Crew
Join Date: Dec 2005
Location: Santa Cruz, CA
Posts: 133
|
Color Calibration
Hi all.
I am hoping that Chris might have a response to this question, but if anyone else does too, please help! I am working desperately to color match the H1 to the XL2. We use them together for multicam projects and correcting in post is a big time waster. I bought a DSC labs frontbox pro chart and have been fiddling with it and both cameras. I am looking for a tutorial on using the chart and how everything should be set up. I have put bars to the vectorscope with both cameras and the bars are good, so no problem there. I switch to shooting the chart with the cameras and from what I read on DSC's site they have lower saturation and the vectorscope should be amped up 1.85x Having done so. The two are so different, and I am killing myself trying to put the dots in the boxes. Also, I noticed that the blackburst line is far too long. Any reason for this? Thanks all. -Brian |
February 21st, 2007, 05:52 PM | #2 |
Obstreperous Rex
|
Michael Wiegand of DSC Labs is a member here. I'll ask him to provide some input if he's able.
|
February 21st, 2007, 07:23 PM | #3 |
Trustee
Join Date: Jan 2006
Location: New York, NY
Posts: 1,267
|
Brian,
A procedure for matching the XLh1 and XL2 (two different cameras) with charts goes something like this. Be warned the different cameras architectures will give you different results under different lighting conditions and white balances. Post has much better tools to deal with some issues than the cameras in the field. Ie the settings on the XLh1 are not sensitive enough to always match another camera. Sometimes you need a 3.5 instead of a 4 or a 3. Console software might be a help with this. Light the chart evenly with a good quality light usually tungsten without filters or gels to change the color. Make sure there is no extraneous light hitting the charts. Set both cameras up side by side with as similar an angle to the chart and light avoiding flares and glares of light which can occur. Frame the charts as identically as possible.Put the camera in manual iris and gain. Try and match the exposure so the whites are as fully exposed without clipping as possible. With different gammas and matrixs this may not be 100 IRE. First look at the cameras in preset to see how they respond without white balance adjustments. Use the XLh1 Custom preset mode to try and match the XL2 Custom preset Using the gray scale on the chart set the master pedestal,pedestals gamma,knee, matrix, detail and gains so the wavefrom and vectorscope show similar if not identical looks and levels on the the chart.To be properly white balanced in Preset the vector scope should have just a small dot in the center of the XY meaning no color when looking at the gray. Start with the lens closed and look at black. That is what you want your white and gamma to look like as you open the iris. In truth the dot will be bigger but it should stay in the center. (this is not creating a particular look but the preset for 3200K for instance) Switch back and forth between the two cameras on the same monitor and scope. Use the red and blue adjustments for pedestal, gamma and gain to balance out the vector scope adjustments. Some settings change others so try and choose matrix settings b efore making other adjustments. If the grayscale chart looks the the same on both the monitor and scope then you can start looking at the colors. If the camera has a color saturation ( the XLh1) setting this can help you match the quantity of the color outputs of the two cameras. By the way Charts shot by the camera will almost never get to the boxs on the vectorscope that the bars do. What you want to do is make it so they have similar if not identical response. Save your settings and then try and shoot other images to see how you did. It is unlikely the two cameras will have the same sensitivity and other issues may crop up in different lighting situations but you should be able to get them to look pretty close to each other. If the presets look similar there is no guarantee that Auto white will be similar. HTH |
February 22nd, 2007, 10:58 AM | #4 |
Regular Crew
Join Date: Dec 2005
Location: Santa Cruz, CA
Posts: 133
|
Thanks for all the information. I will follow it step by step.
This looks like a process that is dying for a step by step photo or video tutorial. Maybe something that DSC could post on their site. I will report back with my results Cheers! |
February 22nd, 2007, 11:14 AM | #5 |
Obstreperous Rex
|
Actually we could post it here. Looking forward to your report, Brian!
|
| ||||||
|
|