XH-A1 gamma setting?

[Raneath Sokha]

What gamma should I use with the XH A1 for the film look while trying to get the most latitude?

1) Use gamma-2
or 2) use normal gamma, then do the contrast in post (and can the contrast in the post look the same as the native contrast that gamma-2 gives you?)

raneath

[Kevin Amundson]

Let’s talk first about what Cinegamma is not. It is not designed to generate
images that look more like film. It is in fact designed to create a nice, flat image that is minimally processed and eminently color-correctable. In other words, it creates an image that behaves more like film does in a postproduction environment

The best way to describe what Cinegamma does is that it does as little as possible. In other words, it presents the scene with a very simple gamma curve, without trying to fancy it up with black stretch or a highlight
rolloff “shoulder.

Quotes from Stu Maschwitz's "DV Rebel's Guide" referring to Panasonic's CINE-LIKE modes, which is the same as Canons Cine mode.

I am still toying with my Preset that gets the most Dynamic Range possible.

My current Preset is:
Gamma: Cine2
Color Matrix: Cine2
Knee: Low
Black: Stretch
Detail:-9

This Preset has given me the best results for Color Correcting in Post. Notice the detail level of -9. This gets rid of most compression artifacts and eliminates noise from gain. I have found best results with the gain level set at -3, 0 is OK, but and I wouldn't recommend going above +3 unless you have to. At the very end of post production, right before I render my master I add detail back in with After Effects.

[Raneath Sokha]

Kevin,

Cine-2 compressed some black, doesn't it? and Cine-normal doesn't compress anything at all, right? Because Cine-2 compresses some black, it will lose some latitude in the black area. That's why I was wondering using Cine-normal, which has the most latitude because it doesn't compress black or white, and just compress the black later in the post. That way, I can have original source with the most latitude and adjust it more variably with Cine-normal where Cine-2 already compresses some black native. But the problem is that people and Canon says that Cine-normal is video look and Cine-2 has the curve of the gamma of film. Is that statement regarding the gamma curves of the Cine mode just a matter of how much black is pressed? And the other question is "Would that process produce the same image as what the Cine-2 would natively?" You get what I'm saying?


Raneath

[Peter Jefferson]

"referring to Panasonic's CINE-LIKE modes, which is the same as Canons Cine mode."

I'm sorry, but no way will the A1 match the DVX/HVX in colour rendition. The curves are not the same, they may SEEM to be shaped in a similar fashion, but the CCD's, processors, lens and more importantly, the codecs all play a part in how colour is rendered and archived.

It will come close with A LOT of tweaking (Ive just spent the last 4 weeks tweaking) but skin detail and gradations in underexposed areas (i.e., a side lit face) won't ever come close to what the DVX/HVX can pull off in the same environment/conditions. I'm specifically referring to colour rendition and accuracy straight out of the camera, not luminance or "brightness"

The beauty of the A1 however is that you can set your knees and b-press levels to offer a clear stable ramp within a histogram and/or an RGB histogram. Add a slight S curve in post to pop the image and you will literally be blown away with picture... it's that good.

With proper settings, you can literally blow out a shot and have the camera salvage a lot of information before its even encoded to tape.

In any case, for a wider latitude, keeping your black levels stable by using a midrange setting is always smart. This will keep noise to a minimum and blacks remain black. A higher knee will allow whiter whites and a cleaner rendition of skin tones.

By tweaking the camera presets as a subtractive synthesizer as an example, the idea is to refine your image using the preset Matrices.
By tweaking your presets into a negative value, then boosting the entire colour gain globally once that colour set is created, allows you a lot more control later when refining that preset, as you then have enough "headroom" to be able to boost each joint colour Matrix individually, without killing the codec's response to the signal its compressing. If you boost the in-cam colour too much, you WILL lose dynamic range. In addition, you will be pushing the codec alot more, whereby it may not be able to handle certain environments or colour levels/peaks (think red jumper against a yellow wall with a subject who has fair skin).

It's a little difficult to explain, but basically what you're doing is setting the camera to suck in as much as you can give it, then tweaking or filtering OUT the range of colour and luminance you don't want (much like the subtractive synth I was referring to earlier)

[Kevin Amundson]

To clear one thing up, when you say "latitude" do you mean "dynamic range"?

If so please read this thread:
http://www.dvinfo.net/conf/showthread.php?t=106271

In this thread Kris says:

Peter said- "in many circumstances it's fine to blow out the whites and crush the blacks - and that is where the cine settings help"- the cine gammas don't do this in my tests, they expand the highlight detail and compress the lows (without necessary affecting DR at all of course). They don't clip the blacks.

My tests have confirmed this.

[Raneath Sokha]

so the dynamic range of the three Cine mode are the same? With Cine-2 having less difference between the black in black and the white in white, while the normal Cine has much bigger difference between the black in black and white in white? Am I right? So for conclusion, shoot using Cine-2 for the best possible result?

[Bruce S. Yarock]

Peter,
Great info. Can you share a couple of your presets?
Thanks
Bruce S. Yarock
www.yarock.com

[Jeffrey Liou]

Originally Posted by Kevin Amundson

....... At the very end of post production, right before I render my master I add detail back in with After Effects.

Would you kindly describe a little bit about how to add detail back in AE

thanks!

[Kevin Amundson]

Originally Posted by Jeffrey Liou

Would you kindly describe a little bit about how to add detail back in AE

Sure, I'll shorten Stu Maschwitz explenation. He outlines how to do this in "The DV Rebel's Guide".

When I make a short film this is what I do after I have my final edit locked and all effects and color correction finished; I export my master as a lossless TIFF sequence, this is so I can split it up on to multiple CD's or DVD's when I'm ready to back it up. One large quicktime or avi file can't be backed up onto a DVD.

Once I have exported my master I import my image sequence into a new AE project. I create a composition for each format I need such as a 720x480 Widescreen DVD and/or a 320x240 for my iPod. You will have to fit your movie to the Comps width.

You now add your detail to each composition

In your output comp, with your freshly scaled-up or scaled-down master, create an Adjustment Layer and move it to the top. Choose Effect > Blur & Sharpen > Sharpen.

This effect has one blissfully simple slider. Increase it slowly until you see your video image sharpen up. You will need to test this process by rendering out a section of your film and watching it in its intended format. You may be surprised by how far you can go with this effect. For DVD, you may use values as high as 20 or more. For a Web movie, you might use a bit less. There’s no “correct” setting—it’s totally up to you and your eye. Experiment and pick the sharpening value that works best for your output format.

-DV Rebel's Guide by Stu Maschwitz

[Rumi Geiger]

to Peter Jefferson.
you said this

"but basically what you're doing is setting the camera to suck in as much as you can give it"

So Im curios as to what presets you are using, and what to you recomend to get the most out of the A1 as for as Dynamic rang, best color rendition.

thanks.

[Jeffrey Liou]

Originally Posted by Kevin Amundson

You now add your detail to each composition

Quote:
In your output comp, with your freshly scaled-up or scaled-down master, create an Adjustment Layer and move it to the top. Choose Effect > Blur & Sharpen > Sharpen.

thanks for quick reply!

add detail = sharpen
How about Unsharp Mask instead of simple sharpen
it seemed that USM do job better in still image while using Photoshop