HMC 150 & Canon 7D Quick White Balance test

8string

8string

Well-known member
So here's a little test I did this afternoon. My outcome seems to show that the 7D Auto White Balance in shade/indoor window light, seems to help get closest to the HMC 150 w/F6 stock Cine D setting. Finally, I converted both to AVI in post, and added Frederick Baumann's new FBM Software's white balance filter http://fredericbaumann.free.fr/FBMSoftware
by eye dropping on the white patch in the back of the final clip. I'm going to do some other tests over the weekend. Does anyone see any problems with this?

Hope this helps others. I love both cameras for their various strengths.




Alf
 
Last edited:
Hi Alf,

Btw, thanks for the reference to my plug-in!

Looking at your video, I was wondering one thing: does the conversion to AVI change something to the colors of your clips?
Or is this "only" to improve the Vegas performance? (I quote "only" because such a conversion definitely helps Vegas perform faster).
Any other reason? (might be for using the 4:2:0 -> 4:2:2 feature of Neoscene?)

Thanks in advance,
Frederic
 
Frederick, it is my belief, after running this test, that the Neoscene conversion to 4:2:2 is helping the color balance to some degree. I'd love to have any of the folks who really understand why that would be answer that question. I've read about the differences of 4:2:0 to 4:2:2 so I understand the theory here, but it does seem that the conversion brings out a better color rendering. If we don't get a post on this issue that clears it up, I'll repost on the general forum and see if Barry or someone picks it up.

From the Neoscene website:
Within NeoScene we offer telecine removal (60i –> 24p) and deinterlacing (60i –> 30p or 50i –> 25p), proper chroma interpolation from 4:2:0 to 4:2:2, and fixing of improper luma levels in Canon cameras. (underline is mine).

Also, to be clear, I did not convert from Studio RGB to Computer RGB when I color corrected this footage in Vegas. I might go back and do that to see what difference I get. As stated in a post at Neoscene

Sony Vegas choose to map YUV sources differently such that black is displayed at RGB level 16,16,16 rather than 0,0,0. While Studio RGB is a legitimate colorspace with some advantages even, it should not be displayed on a computer monitor to be used for visual color correction. If you do visually color correct within the Vegas preview area, the resulting files will be more contrasty when displayed elsewhere (DVD, BluRay etc). To properly color correct in Vegas you need to either use an external HDSDI/HDMI card to a YUV display (TV, pro video monitor) or add the color correction filter “Studio RGB to Computer RGB”. You can add this filter to your preview window within Vegas, and all the resulting images will be displayed correctly (without losing any of the processing advantages for Studio RGB.)
Click to expand...
 
Last edited:
Hi Alf,

Thanks for this information.

I have refreshed my memory about 4:2:0 and 4:2:2 by browsing the web, and I understand, to make it short, that 4:2:0 samples chrominance twice as less than 4:2:2. And in addition, 4:2:0 kind of interlaces Cr and Cb, while 4:2:2 brings a new CrCb couple every 2 pixels in the picture, meaning that with no interpolation, each of r and b values only changes every 2x2 pixel block (instead of every 2x1 pixel block for 4:2:2), and that from 1 pixel to the next one, there might be discrepencies if both r and b values change in the original scene (because 4:2:0 only describes one of r or b changes on one line, the other composite is considered as constant on the whole line). So, yes, I can imagine that converting from 4:2:0 to 4:2:2 can improve color accuracy (even though it will still only be an interpolation, and the original values are lost during recording).

Vegas could do this interpolation by itself, but I guess that if Cineform shows better results, we have to conclude that Vegas does not do it.

Regarding improper luma levels of Canon cameras, I wonder where this improper management come from. I have just read this page:

http://www.brisk.org.uk/photog/hiblow1.html

saying that, for pictures at least, the blinking of highlight on the LCD screen is not based on true luminance, but on individual RGB channel values. Do you think this might be the origin of improper luma levels mentioned by Cineform? Would you see any other reason?

Best regards,
Frédéric
 
That's an interesting read, and I took a close look at my camera (I'll try downloading his file and comparing). But you could be right. I'd prefer to let a much smarter person than I am answer the question you pose. Perhaps it's time to post this over on the Canon board. I'll do that and see who shows up.

Just checked over there. There is a technical discussion thread. You might want to start here>
http://www.dvxuser.com/V6/showthrea...fer-4110C-dynamic-range-test&highlight=chroma

I did. I'm going to shift this whole discussion over there after carefully reading through the past threads. See you in the Canon forum for further discussion.
 
Last edited:
You must log in or register to reply here.
Back
Top