Critical Looks at Demo Reels

[webrunner5]

3 hours ago, IronFilm said:

Nikon D750 is an interesting oddball, as if stills is a high priority to a person then Nikon can make a compelling case for any of the Low-low / Mid-Low / High-low categories with their D5x00 / D7500 / D750 / D500 / D850 cameras. But stills capabilities in my ranking was one of the lowest factors for ranking.
 

I am not talking stills at all, I am talking Video! Hell I could give a rats ass about specs. I want to see beautiful Color Science. There is nothing much better looking than D750 colors wise. I like it better than the Canon 5D mk IV colors wise.

That is why the BMPCC is in my list. It is a total turd to shoot. But it is about as good as it gets output wise. Calling all those Panasonic cameras you had listed Cine looking, beautiful outputs, ehh I don't think so. Maybe the GH5s, GH2 hacked. And the AF on a EVA1 is even worse than the GH5. Same AF that was on my AF100A,  WTH. A Canon 5D mk III using ML not on a list? Wow, Glenn ain't liking that, me either.

OK I will give you most of the Panny cameras have better video Specs on your list, doesn't mean I am going to use one. I have No Panny cameras left right now, no m4/3 at all, and other than Maybe a GH5s, none in my sights. I gave up on specs for what I need. I want to see stuff that makes the hair on the back of my neck stand up. Canon 1DC 4k stuff. Canon C300 1080p, that kind of stuff.

Now if I was going to try and make a living doing it, that can change. Probably would be a Canon C300 mk II in there someplace just because of the DPAF.  Yeah I do like the EVA1 other than the AF stuff. Not that sold on the C200 enough without knowing the middle Codec yet. And knowing Canon probably will be disappointed. Can't afford one anyways, or a C300 mk II LoL. 

[IronFilm]

On 1/26/2018 at 7:04 PM, ntblowz said:

(so as some nz tv series, funded with millions of nz tax money ? )

Which ones? I know Shortland St is using the Varicam S35 (they don't get funding I am pretty sure? Well, many many many many years ago Shortland St got some from NZ On Air. Back in the 90's, for the first four seasons), but I can't think of any other kiwi TV series at the moment using a Varicam. 

[Aussie Ash]

2 hours ago, IronFilm said:

Which ones? I know Shortland St is using the Varicam S35 (they don't get funding I am pretty sure? Well, many many many many years ago Shortland St got some from NZ On Air. Back in the 90's, for the first four seasons), but I can't think of any other kiwi TV series at the moment using a Varicam. 

Not even "Country Calendar"

https://teara.govt.nz/en/video/45718/country-calendar

[IronFilm]

They use a Varicam? (S35 or LT?) 
Awesome!

[ntblowz]

Filthy Rich 2nd season use Varicam LT

[IronFilm]

Ah cool! Didn't know that. Just had seen them during the 1st season when they were using Arri cameras for it. 

[IronFilm]

A little off topic for the thread, but I didn't want to bother with messing up the forum with a whole new thread, however if you know German this is worth a read:

https://agdok.de/de_DE/kameratest20172

(and perhaps even if you don't!)

[kye]

Can we talk about the new ARRI FF 4k demo reel?

After reading the two articles in the OP of this thread (very educational to me) I watched this one with great interest.

Some observations and questions..

Everything seems quite desaturated, or at least 'modest'.  Down to the choice of clothing, etc.  The animals are also not screaming with saturation (which some would be in real life).

All the skin tones seem very yellow, without a huge variance between the yellower and pinker skin tones.  Is this an ARRI 'look'?  Or is this the grade?

[jonpais]

Not trying to be facetious, but the muted color is an aesthetic choice, and the yellow skin tones you are seeing are the fault of your monitor. 

[IronFilm]

3 hours ago, kye said:

Can we talk about the new ARRI FF 4k demo reel?

To me it was obvious what ARRI's #1 focus of their demo reel for their new camera: skin tones. 

[kye]

Here are 8 screen grabs - the top row are the people from the ARRI reel likely to have the most pink in their skin tones, and the bottom row are people from the "Shot on the C200" thread over at DVX (http://www.dvxuser.com/V6/showthread.php?357491-SHOT-ON-THE-C200-Completed-released-commercial-projects-only).

It was easy to find people with significantly more pink in their skin tones than anyone in the ARRI video.

In regards to being facetious, maybe it's best to try avoiding it rather than just not aiming for it.

[jonpais]

@kye may I ask what monitor you’re viewing these on? Because watching the Arri reel on my 55” OLED, the image quality looks spectacular. Do you prefer the Canon colors to Arri’s?

[webrunner5]

I loved the Arri shots also but in these screen grabs above I think all the Arri stuff look too pale and all the Canon stuff too reddish. But Canon seems to always lean toward red in faces.

[jonpais]

They all look dreadful on my iPhone 7, which is why I’m interested in learning what monitor @kye is watching the video on. I’m wondering whether he has a preference for saturated color, or if he believes the Arri only shoots muted color. It would appear as though he prefers Canon colors, which is fine.

[kye]

Hi Jonpais.  

I think you misunderstand me.  I am saying that the ARRI skin tones all appear to be on the yellow-side of the acceptable range of skin tones, not that they're out of the acceptable range.  As someone who has quite a lot of colour variation in my own skin tones (a model I am definitely not!) and I lack willing volunteers for my camera tests I end up grading myself - both in still images and video - so I'm familiar with the problem of people who are pink in some locations (lips, around the nose, in the ears, etc) and yellow in others (temple areas, shadows under the chin, etc).

I shoot Canon which I know has a red tinted bias but I thought was still acceptable, especially if you compress the skin hues with the Hue vs Hue curve in Resolve.

The reason I didn't mention what I am viewing these on (MacBook Pro 2016 calibrated with Datacolor Spyder 4 PRO in my studio with all controlled lighting) is that I am an amateur and new to the colour game and my monitor is far more calibrated than my eyes.  That being said, while I wouldn't trust my own grading, I do trust the professionally released C200 videos I linked above.  The C200 videos showed a range of skin tones, some on the yellow side similar to ARRI and some on the redder side which I chose for the screen captures above.  All of the screen grabs I took are pinker than the pinkest person in the ARRI video.

The gentleman on the left with the tie from the ARRI video has quite a lot of hue variation in his skin - typical of an older person, however the hue differences *as shown in the ARRI video* are pretty minimal - look at the colour of all of their lips for example.  I would suggest the model probably has pinker skin tones than all of the people from the C200 videos (with the possible exception of the guy in the cap who has a touch of sunburn on his nose) yet the ARRI video doesn't show this - it shows barely any variation in hue.

My original question was a genuine question and not a veiled criticism, and it still stands - is this preference for yellow skin tones (which are on the yellow side but still quite acceptable) an ARRI 'look'?  In the way that Canon prefers redder ones?  Or is this part of the grade?

The author of the two articles in the original post of this thread talked at length about camera manufacturers making choices that determined their looks, both from an aesthetic perspective and also from a technical perspective to work around the limitations of their hardware, so this is the context of my question.

[jonpais]

@kye My 2016 MacBook Pro isn’t calibrated, but it definitely shows a bias toward red/yellow. Behind the Mac is an LG C7.

[webrunner5]

Well I can tell you from living in the United States I know no one that looks like the bottom Canon grabs, unless they have a sunburn, and I know more than half of people that I know look like the people in the Arri shots, well at least in the winter time LoL. Maybe they are all from England, and that might be true.

To be honest neither one of the screen grabs look right to me. Ergo Canon is not that bad, and Arri sure as hell is not that bad in real life shots..

[jonpais]

? I might add that my room has yellow walls and curtains, and that the MacBook screen is like a mirror, whereas the LG reflects less than 1% of the light in the room. But Mac Performance Guide confirms the tendency toward red/yellow. Also, to the best of my knowledge, 3rd party calibration can only make adjustments in 8 bits, which is vastly inferior to calibrating the monitor itself in 12 bits. Correct me if I’m mistaken. 

The 27” 2017 iMac is by far better for grading, not only for the greater screen real estate, but also for its truer rendering of reds/yellows, so crucial to skin tones; though it still has problems of its own. Still no substitute for a reference monitor.

[kye]

The ARRI footage is lovely, the demo is hugely impressive, and if someone was to offer me a free camera I would fly anywhere in the world to come collect it.  But I think you guys are missing my point, I'm trying to understand where within the realm of acceptable skin-tones it sits.

If we start with the concept that the C200 clips show skin tones that are redder than the reddest person in the ARRI footage, then one of the following statements must be true:

1) If the C200 colours are acceptable as skin tones, that would mean that the range of acceptable skin tones contains skin as yellow as the ARRI footage and as red as the C200 clips, which means that the ARRI skin tones are all on the yellow side of the acceptable skin tone range.

or

2) The ARRI footage is not at the yellow side of the acceptable skin tone range, but is more neutrally presented, which means that as the C200 skin tones are considerably redder then this they must be unacceptably red and outside the realm of acceptable skin tones.

I believe it's the first one, and that it's just a style thing, or an ARRI 'look'.  The C200 footage thread contained positive feedback about some of the grades from the films I screen-grabbed, so I'm guessing they're within the acceptable range of skin tones.

Please note that at no point did I mention anything subjective in the above statements, it's quite simple - either Canon is fine and the ARRI is yellow-biased or the ARRI isn't yellow-biased and the Canon grades are out of bounds.  I'm just trying to understand which is which.  Maybe it's too obvious a question and you guys are reading into it some kind of subtext?

[jonpais]

@kye Your monitor is not accurate, so why do you persist in calling the Arri shots yellow? Secondly, those images are graded by someone referred to as a colorist. If you bothered to have a look at the other Alexa LF footage, you would see different skin tones. And of course, there is a wide range of acceptable skin tones.