Custom presets menu

[Lauri Kettunen]

Is there anybody, who is able to define precisely the contents of the custom presets? What are the algorithms behind the settings? Chris, A.J., do you know?

To give an example of what I mean: I suspect the color matrices adjust the off-diagonal entries of the matrix mapping input RGB-vector to output RBG-vector, and that the R,G,B-gains are the diagonal entries of the same matrix. But, when they say "RG matrix", is the input R or G, and to what do the values -9 - 9 refer to? What is the algorithm behind master ped., setup level, coring, etc.

Or is it, every retailer advertizes these menu options of the camcorder, but eventually, and literally, nobody outside the engineering team of Canon know what they are all about? Is the situation this ridiculous, sounds amazing.

[Nick Hiltgen]

IN the F900 the R-G matrix serves as a way of controlling the color taken into the camera. So Adjusting the R-G would either add or subtract green from the red channel. (the matrix can then be tweaked for the remaining 5 matricies) Giving you close to complete control over the colors that are being processed by the camera. The RGB gains I believe are more for a saturation control.

To get precise numbers you would have to go to a waveform, but even then there are no absolute values, because each tweak of the matrix menu is going to be relative to whatever setting was originally in the camera. (the F900 it goes from -99 to 99 but still no quantifiable value). Does that make sense?

[Shannon Rawls]

Originally Posted by Nick Hiltgen

Does that make sense?

Hell no.
*scratching my head*

Nick your a genius.

- ShannonRawls.com

[Lauri Kettunen]

Originally Posted by Nick Hiltgen

So Adjusting the R-G would either add or subtract green from the red channel.

So, you guess the same way as I do. In terms of linear algebra you say the RG-matrices refer to the off-diagonal entries of the linear operator. But now, let's ask the very first thing: In case of RG, is R the input and G the output, or vice versa?

[Nick Hiltgen]

Lauri, not sure, but I got my camera today and will begin testing so hopefully later on tonight I should be able to let you know. Off hand my guess would be that RG means adding or subtracting green from the red channel, but I'll let you know after I test it.

That doesn't answer your question though, so let me make sure I'm understanding it correctly, you believe that the RG matricies read the information such that the information being taken into the red channel can then be manipulated by the Green channal for outputting to the new (and laid to tape/sent to monitor RGB) In which case Red is the base channel (input?) and green is the variable channel (output?) or am I still confusing the question?

Shannon: Hardly, but thanks for the compliment. *smile*

[Lauri Kettunen]

Originally Posted by Nick Hiltgen

you believe that the RG matricies read the information such that the information being taken into the red channel can then be manipulated by the Green channal for outputting to the new (and laid to tape/sent to monitor RGB) In which case Red is the base channel (input?) and green is the variable channel (output?) or am I still confusing the question?

Yes, in more understandable words my guess is: Let's call R the red channel which goes out to the tape (or to HD-SDI, firewire output, etc.) and r is the red channel from the CCD. Correspondingly let's call the other output and input channels G,g,B, and b.

In simple words my guess is: with the color matrices RG and RB you can fine tune the amount of input CCD green and blue signal mixed to the output red channel. More formally,

The value of R = (Gain of red) times r
+ ("value of RG-matrix") times g
+ ("value of RB-matrix") times b.

Here, "value of RG-matrix" does not refer to the number (-9 to 9) you see in the menu, but to some real number which is related to the integer varying from -9 to 9 in the menu.

In my camera the image is tinted towards magenta. To get rid of it, I have to find out whether the input red is biased towards the blues, input blue towards reds, input green towards blues, or etc., and then correct this by adjusting the values of color matrices. The Console program is very helpful in this, for it has "variations", making it easy to compare different settings. The same applies, in fact, to all custom presets.

But, my real problem is, when I'm filming in the middle of nowhere for several days, it's not possible to carry a laptop along. However, the camera needed to be fine tuned constantly according to the changing light situations (for the relation between the input and output is nonlinear). Now, the EVF is not precise enough for fine tuning, but that's all I have somwhere there in the wilderness. So, I needed to know what the custom presets do precisely. Some vague idea of them is not enough, for the need is to reliably estimate what kind of fine tuning is needed.

The idea behind Canon's policy seems to be: the users can test and then they remember what is needed in different conditions. Maybe they don't want to confuse their customers. But the result is like creating a kind of "list of actions" what the user should do in various conditions. But now, in my case (of wildlife filming) there would be an endless list of occasions to memorize. So, I rather seek for the definitions, for only then one is able to predict what kind of fine tuning is needed.

For example, say one is filming the sun set and expecting the rare green flash, which occurs for a moment when the sun hits the horizon. There's no way you can test the custom preset settings for this purpose, for one has a chance to see the green flash say less than 0 to 5 times during your life (at least if you lived here). So, you need to know on before hand, how one has the best chance to pick the bright green flash to the tape and to avoid the overwhelming red tones.

The engineering team of Canon knows what these custom presets do precisely, and there should be nothing really special behind them. If they only gave the technical specs, then in a forum like this there's enough people who were happy to translate the signal processing into common words. Accordingly, I hope that some publicity around the issue made Canon to reconsider the situation and enabled us to take the most out of their products.

[Levan Bakhia]

Why don't someone (much clever guy than myself) make custom presets with different looks and feels and post them somewhere. Of course with samples of what that full preset is going to look like. and then less clever guys like me would be so happy to use them.

Or why don't canon make it. It is so hard to understand all the combinations of these presets. It will take whole lot of time to find the correct ones. So I wish we could all share it somewhere. Like for example, once I achieve the look I like for one specific job, I will post it and all of the rest could use it, and once someone else achieves an interesting look also post it and so on.

Easier?

Oh, and please let me understand. When I capture the content to FCP. Is it 25F or is it 50i. What can I do to make sure it is 25F. Because as I understand, even thou the camera records (F - progressive) it records to tape as interlaced. But then, when I transfer it, I guess it is interlaced. So how can I make it progressive? Or is this a stupid question?

[Martin Costa]

Shannon, are you ready to share any of your custom presets yet? I haven't got the hang of all the settings yet, and I rememberd you saying you'd saved a few film settings already. ;-)

[Shannon Rawls]

No, not yet. The more I look at them the more I think I am off. LOL

I watch the movie i am trying to emulate
I watch the HD monitor
watch the movie
watch the monitor
look at movie
look at monitor

(sum'n ain't right)

I don't know what the hell I am doing. LOL NATE WEAVER, WHERE ARE YOU???????? *smile*

- ShannonRawls.com