My recommended picture style and settings for the 5D Mark III

[Andrew Reid]

[html]

[img]http://www.eoshd.com/wp-content/uploads/2012/03/5d-mark-iii-patch-vault1.jpg[/img]

[url="http://www.eoshd.com/5d-mark-iii-settings-vault"]Get ‘Cinema Mark III’ by EOSHD in the 5D Mark III Settings Vault[/url]

I’ve created a section of the site for recommended 5D Mark III settings, including the all important picture profile settings that I personally shoot with. These are general purpose settings – the ones I use most often.

[url="http://www.eoshd.com/?p=7677/"]Read full article[/url]

[/html]

[Samuel H]

thanks a lot for the recommendation!!

I'll pay back saying the following here first:
if the 5D3 gets Magic Lantern, this will probably unlock the extra DR that's available in stills mode but not in video mode
so, without ML, all I can get on the 5D3 in video mode is 11.5 stops of DR, basically the same as with a T2i
but with ML, all the DR captured by the sensor will (probably) be available in video mode too; that means 11.5 stops on the T2i, but clearly more with the 5D3 (the dxomark review is not out yet, but from my quick tests there's clearly more than the 11.5 I can get in video mode!!)

[Andrew Reid]

Thanks for the comment ;D I remember Gale Tattersall (House) saying 11 stops was generous for the Canon DSLRs. I'd say it was more like 9 or 10? 5D Mark III should be better given the new sensor... Looking forward to seeing how this plays out...

[Samuel H]

11 stops on the old cameras (Flaat_11 and Flaat_12) is noisy in the heavily-pushed-up shadows; some people say it's useable, some say it isn't

this is the light response of Flaat_11:
[img]http://www.similaar.com/foto/flaat-picture-styles/chart-waveform/flaat_3.jpg[/img]
I count 32 unclipped wedges, which means 11 stops of DR (my synthetic chart works in 1/3 stops)
the bottom 9 wedges are very noisy, not easy to distinguish from each other, but those are 1/3 stop marks, if you jump every two then they're clearly distinct; but that noise is nasty, so for many they're unuseable anyway; it's subjective, in any case

the cleaner image of the 5D3 should lend itself better to these expanded-DR picture styles

[Andrew Reid]

Certainly the sensor will. Not sure about the codec. It does seem to have more in the lows than before, I am able to get a lot back from an underexposed shot. But there is often a bit of a fizz from the codec, not ISO noise but something else, even in the mids, mainly in the green channel.

[MattH]

I believe you are correct about not turning down the saturation.  I have heard elsewhere that it does throw away information if you reduce it.  Yet a lot of people make the mistake of doing so. so you are right when you say:

"It is easier to desaturate something than to amplify something which isn’t really there in the first place."

You then say "This goes for contrast as well."  but I have to say in my experience I wouldn't agree with this.  I find you can easily add contrast to a dull image by raising the black level or applying curves, whilst it is difficult to do anything with a really contrasty image with crushed blacks and blown highlights.

I find that reducing the contrast on a canon dslr picture style actually expands the dynamic range closer to raw level.  Try looking at the histogram on a static scene.  Make a mental note of where the information starts and ends, then reduce the contrast and look again.  The information shrinks within the histogram giving more room on the left and right.

[Samuel H]

[quote author=MattH link=topic=481.msg3119#msg3119 date=1332938713]
You then say "This goes for contrast as well."  but I have to say in my experience I wouldn't agree with this.  I find you can easily add contrast to a dull image by raising the black level or applying curves, whilst it is difficult to do anything with a really contrasty image with crushed blacks and blown highlights.[/quote]

You're absolutely right.

What I mean is this:
* if you only use chrominance values in the range [0,128] and push them back to [0,256] in post, you're using 7 bits instead of 8 for the color channels, and your final image will look worse
* same way, if you only use luminance values [16-199] and push them back to [0,256] in post, you're again throwing away another big chunk of the codec's color space, and the effect on final IQ will be similar; and by doing this you're not expanding dynamic range at all (clipping points in highlights and shadows are the same in CineStyle and Flaat_10, but Flaat_10 uses all of the codec's color space, not just a small portion of it)
(actually CineStyle uses luminance values [16-256], but reserves [200-256] for the brightest stop of light: 23% of the used color space is dedicated to 10% of the recorded DR, without affecting the point at which the highlights clip, which is basically the same in all picture styles I've tested, including "portrait with contrast=-1")

[jcs]

Hey Andrew- I did further deep analysis due to this thread: [url=http://cinema5d.com/viewtopic.php?f=30&t=39999]http://cinema5d.com/viewtopic.php?f=30&t=39999[/url]
Log in to see clear pictures showing significant problems with ALL-I vs. IPB (pics also attached below: ALL-I is first). ALL-I has significant noise issues, and the bigger problem is excessive high-detail loss due to a macroblock quantization issue (even when the camera is putting out 80+Mbps; ~30Mbps IPB is significantly better (see 400% crops)). IPB looks beautiful and uses ~3x less disk space. The quality reduction in ALL-I is even visible after youtube compression (watch fullscreen, 1080p): [url=http://www.youtube.com/watch?v=eTEV6_os4gY#ws]Canon 5D Mark III - I-Only compared to IPB[/url] (even more visible with original footage).

The problem with in-camera sharpening is that it uses an Unsharp Mask technique, which creates halos that cannot be removed in post. There is no mathematical advantage and only a disadvantage to using sharpen in camera. With sharpening turned off, compression should be more efficient, and sharpening can be very extreme in post before artifacts appear (using either a convolution sharpen "Sharpen" in PPro, or Unsharp Mask with a radius of < 1.0).

Thus my recommendation based on these tests and analysis is:

[size=12pt][b]IPB, Faithful, 0,0,0,0, Manual White Balance[/b] Sharpen in post with "Sharpen" (convolution, real-time in PPro), or Unsharp Mask with radius < 1.0 (typically slower and not real-time). Larger values of Unsharp Mask can enhance mid frequencies, and really large values (50+) can create pleasing "Local Contrast Enhancement": [url=http://www.luminous-landscape.com/tutorials/contrast-enhancement.shtml.]http://www.luminous-landscape.com/tutorials/contrast-enhancement.shtml.[/url]
[/size]
Haven't tested any custom profiles yet (should work as expected).