TheRenaissanceMan

Converting perfect 24 to 23.976 is much more problematic than doing it the other way around, so that's what most people seem to recommend. There have been some good threads on this over at DVX User and Red User. 

Also, I've worked at 3 local film festivals and have never encountered a projector that couldn't handle 23.976. 

No peaking during 4K. Lame, I know. You can prefocus in video preview mode, but useless for racking focus during the shot. 

0-255 I believe. Either way, your editor should handle it just fine. 

I believe it's a clean 8-bit 4:2:0 4K signal, so there's not a huge benefit to external recording unless you really don't want to transcode. 

The GH4 with V-Log has all of that except the lowlight. Come on, Panasonic!

I didn't say it was better than anything. I referred to a review by a seasoned professional who finds the 5DS to be a useful tool in his work. Value and quality should always be evaluated based on the needs and preferences of the user, not just test chart data.

For web, broadcast, Bluray, or even cinema projection, it doesn't matter one iota. So unless your clients are specifically requesting perfect 24, I wouldn't sweat it.

Hey Andrew, why are you comparing the RX10 II to the A7S and 1DC instead of its natural competition, the Panasonic LX100 and FZ1000?

What is your situation where 23.976 is an actual problem?

They just updated the firmware for "Pro" bitrates in video mode too, which is a nice bump.

Also, you can buy one grey market off Ebay for $350. http://www.ebay.com/itm/Samsung-NX500-Mirrorless-Digital-Camera-Black-/191641873596?hash=item2c9ebeecbc

The main problem with people shooting S-LOG specifically is that they're using S-Gamut in 8-bit, which if you know anything about color spaces is a HORRIBLE idea. You shot your 50mbps 8-bit codec in an enormous twisted-axis color space and tried to color correct it yourself? Without LUTs? Are you INSANE?

I suppose I'm ahead of the curve because I know I'm shit with color correction. Besides the cheap and dirty DrewNetwork workflow for quick jobs, I use Speedlooks on all my footage. It gives me a great range of base looks that I can tweak a little to my personal taste. 

I've sold TVs at Best Buy for 2 years, including during the 4K boom. So I do speak from experience. And in that experience, the perfect contrast of OLED, superior motion handling of Plasma, and the sheer size of a projector screen are far more important than just a resolution bump, especially at normal viewing sizes/distances. The 4K sets in the store, at least in the early days, tended to have all the manufacturers' other bells and whistles too, like local dimming. Those sets looked great, but it was mainly due to panel quality rather than pixel quantity. As more and more come out at lower price points, however, they're including less of those extra picture-improving features and more software gimmickry like motion smoothing and "brilliant color" modes to try and differentiate between the 1080p and 4K sets. Put a 1080p and a 4K set of equal quality next to each other, calibrate them to the same settings, and feed them both a 1080p source. Guess what? From 12 feet away, guess how many employees at my store (who look at these things all day every day) could tell the difference? 4. Out of 80-some. Add to that the dearth of actual 4K content, and I don't know how you can justify that "investment" in a set that'll be completely outdone once OLED lands in earnest.

Most people are much, much better off buying a great 1080p set and investing the rest in a 5.1 system, a stack of Blurays, and some comfy chairs. 

I think people spent a lot of time plumbing the depths of Canon's picture profiles and settings when Canon was our only choice, so we found the best possible settings (or at least a range of good settings) through trial and error. People aren't going through that same process with the newer Sony and Panasonic cams, then they're disappointed when they don't get comparable results. What did you expect? You didn't put in the effort. The only guy I see really going into all the different picture profiles and color settings on all his cameras is Nick Driftwood. Maybe Brandon Li, too. 

The topic is about stills, so I'm not sure where this fits into the conversation...

That is so fucking cool!

Do work. Share results. Learn from them. Repeat. 

This may be a little OT, but it is a stills topic so what the hell. How are you two finding the transition from Adobe to C1? Is it fast and intuitive enough to use? How do you like the default NR/Sharpening/Color/Tonality? (I'm very disappointed with Adobe in that regard.) Anything else you can share about your experience transitioning? 

Not to mention that most people are not only unwilling to learn grading, but downright offended when you suggest that maybe their grade is the problem. A problem that affects 80% of us--not being colorists--is impossible to acknowledge without pissing people off. Isn't that insane? 

Some flat/log profiles, like C-Log (faux-log designed for 8-bit), Nikon's FLAT profile (not true log), and V-LOG L (I don't really know why) respond perfectly to nothing but an S-curve in post, but they're by far the minority cases and you're still thinning out your skin tones.

People realized a while back that 10-bit ProRes (Blackmagic) and 12/14 bit RAW (Blackmagic and ML) held the key to great image quality, then 75% of us promptly forgot the second 4K hit the affordable market. The other 25% are blessed with an amazing gift: the power to remember things! 

Blackmagic did it first, and people jumped off that bandwagon the second they realized it required actual work. 

Where is your area of expertise, then? Not an attack--honestly curious. 

Totally agree with the post, though. Why everyone isn't using the GH4's 10-bit out is beyond me. 

Also, how many times have you read this exact conversation:

"I can't get great colors with X-LOG!"
"Yeah, you'll have to learn more about color grading to get the most out of a LOG profile."
"But that's so much work! I'm a filmmaker, not a colorist. Isn't there an easier way?"
"Well, you could always buy some LUTs and let professional colorists do the work for you."
"LUTs are too expensive! I want something I can grade myself!"
"Um...how about just not using LOG? Your camera has lots of other color profiles that are easy to handle."

"Then I won't get the best DR the sensor can give me! What kind of noob doesn't use LOG!?"
-_-

For understanding, sure, but the more immediate problem is that we're all using different words to talk about the same thing. 
 

For sure, and I think they'll get over that in their next product iteration. As we speak, they're probably designing an unsharpened log-style profile for the NX1 II. I'd like to see a C-LOG style Samsung profile designed for 8-bit, or something more aggressive if their new cameras include 10-bit external (or internal!) recording.

What profile are/were you using on the Canon?

We should have a "terms to know" sticky at the top of the page, like the beginning of a new chapter in high school science. At least then we'd be speaking the same language.

Resolution: how much actual detail is in the image

Sharpness: how prominent or emphasized the details are in an image

Soft: can refer to a lack of either Resolution or Sharpness. 

What we need are separate words for resolution-soft and sharpness-soft. Actually...resolution-soft and sharpness-soft might do it.

Refresh my memory: acutance is local contrast, yes? (I'd look at the link, but it isn't playing nice with my phone)

I think people are misreading the term "soft." The people in favor of "soft 4K" are saying they want 4K of resolution with no local contrast/detail enhancement processing applied, so that they get the extra detail without making it feel harsh. The people against "soft 4K" are saying that if you're paying for 3840x2160 resolution, you should get that and not the blurry mess of the rebels/5D. 

Both of you want the same thing: lots of detail. But just because the detail is more prominent/emphasized by the processing doesn't mean there's more of it.  In fact, adding sharpening can actually obscure smaller details. 

So if the manufacturers are giving us a beefy enough file, they should turn sharpening (detail enhancement) off. That way, the people who want "softer" footage with lots of detail can keep it that way, and the people who want "sharper" footage can sharpen to their heart's desire in their NLE of choice. 

Savvy? 

How close are you sitting? Unless you're relatively close, the benefit over 1080p is minimal. That's why 4K monitors and 4K projectors make sense to me, but not TVs; people will continue sitting about 10-12 feet back from their sets, rendering that increased resolution pretty damn pointless at sizes under 80 inches.

And 4K will in no way eliminate grain. It's likely your set is just eliminating it in software. 

Canon RAW tends to be more pleasing out of the camera because it has more contrast and a less linear tonal response than Sony/Nikon RAW (except the D810). You can read more about the subject here: http://blog.mingthein.com/2015/03/18/understanding-native-tonal-response/

This is also hugely affected by your RAW converter, your choice of color profile, whether or not you profile your cameras, and your PP abilities in general. Given a flat RAW file shot with good technique, "nice contrast and color" is entirely in the hands of the photographer. 

Ming has also noticed a slight difference in color rendering across the tonal range in RAW files after profiling (as noted in his 5DS review), but he considers it too minor to make a difference in everyday shooting. 

Also, something to keep in mind: "better" is a dangerous and subjective word. Better for what? In what way? With what other drawbacks? At what price point? In what form factor? 

Hey Andrew, I have a suggestion for future reviews in regards to comparing color. The real-life scenes are great and give us an idea of what the cameras deliver in actual shooting scenarios; however, if you're going to make definitive statements about color accuracy and precision, it would behoove you to buy/rent a DSC OneShot/Chroma Du Monde chart, shoot it with both cameras in both daylight and tungsten lighting, then put them up on the vectorscope and do some analysis. Otherwise, the whole thing gets very subjective very fast, and there's no way for other people to adequetely chime in without having the original files and having been there at the exact time you shot it (to know the exact color of the objects and how they appeared in that lighting).

A Macbeth chart would be a step in the right direction. It's not as useful as the OneShot/Chroma Du Monde but a hell of a lot less expensive. Displaying one on an iPad proves nothing and provides no useful information whatsoever. 

At the very least, you should enlist a friend or two to sit in the middle of your test scenes instead of using that book, which may or may not be color accurate and doesn't reflect the skin colors most of us regularly shoot. It's easy. Offer them $10 or a home cooked meal or something, then tell them a funny story while they sit in a chair for 15 minutes. Boom. You're done. If you're having trouble finding people, I'm sure there's an EOSHDer in/close to Berlin who'd love to help. 

If you have any questions about the gear/methodology I'm talking about, feel free to PM me or, better yet, check out Art Adams' articles on the DSC OneShot chart (which he helped design) and his analyses of camera color (on which he's something of an expert).

Cheers.