Why does the color in this A7s video look so fake?

Lpowell

Veteran
From the YouTube description:

Join Rotolight Worldwide Ambassador and Sony Artisan of Imagery Jason Lanier for a pin up shoot with the iconic Golden Gate Bridge as a backdrop for the world's first recorded shoot with the amazing Rotolight Neo video lights.

Jason uses both the Sony A7 and A7s for the shoot using three different lenses. The Sony FE 16-35 Zeiss by itself with no adapter. He then uses Sony A mount glass with the 135mm f/1.8 Zeiss and the Sony 70-400 G lens.

Sorry, but to my eyes the skin tones and dress colors look plastic and over-saturated and the background looks rotoscoped. I'm not claiming the shoot was faked, the video is obviously documenting a live shoot. What I'm wondering is whether the results were due to the camera, the Rotolight Neo, or post-processing?

 
Agreed.
I love the look of the A7s for video.....until you put it on people. Is that kinda important?

I have looked at a ton of A7s video and I have disliked virtually ALL of it when people were in the frame. There's just something wrong with the skin tones.

And "plasticy" is as good as any other term...It looks like it's been "airbrushed" or is suffering the plastic look as if it had too much noise reduction.

Either way I just am at the point where I don't even want to see anything shot on the A7s if people are in it.
It's disorienting to the eye to see the environment looking realistic and then skin looking so.....not.

I think this brings us back to the "inherent look" of each camera discussion.
Some people will PREFER the look of the A7s. But it is sure not for me.
 
Last edited:
Grading and use Chart match for Video
The SLOG2 skintone grades in this video are so unflattering I actually find it disturbing. It's as if the scenes were lit by a half dozen different fluorescent lights, each casting a slightly different greenish or magenta hue, combining into something that is technically balanced, but visually deficient. It's a real mystery to me, as we've seen Sony image sensors produce gorgeous results in other camera brands.
 
With the sample video you linked to, I think part of the color problem is caused by the model's make-up as the color of her face does not match the color of her arms, which to me is an indication that somebody screwed-up with the make-up. This may also contribute to the "plasticy" look of the model's face, where too much make-up will remove texture from the skin and give it that "putty" like appearance.
 
There are two issues being asked that are disparate. One is why this particular video has the funky or stylized skin and the second is why are the fleshtones on a7S footage so flat and plastic like.

To the first. He is intentionally trying to get this look. You don't have to like it or his manner but it is intentional as he is trying to get this nineteen fourties - fifties glam glow thing happening (BTW I didn't/couldn't watch the whole video so if the photographer took a left turn half way through in how he does things someone will have to tell me).

Second is the issue of plastic skin with the a7S. It's this over exposing by two stops (sometimes more) that people are doing to maximize the dynamic range. If you go back 5 years in this and other forums on how to expose Sony Cinegammas and even Slog on the F3, you will see reports of people getting plastic skin without much dimension or tonal range. There were vigorous pleas, even from some people who now are telling people to over expose, to not push the fleshtones in the +70% tonal range because you are going to hit either the knee or the portion of the gamma curve where it sharply comes down to compress the highlights. You push the skintones too high in the tonal range they go all plastic-like. That is the nature of how the Sony gamma curves work... on any Sony camera.

If your scene is properly lit, you expose properly for your subject matter and choose the right gamma curve for your lighting conditions, your footage will look great and not unlike Sony footage from other cameras in the recent past.

EDIT:

To correct something I said above. The overexposing is not to maximize dynamic range per se but to alleviate some of the visually seen noise in the shadow regions by over exposing and then reducing the exposure in post. If you poke around here you will see many posts and threads going on about this technique.
 
Last edited:
There are two issues being asked that are disparate. One is why this particular video has the funky or stylized skin and the second is why are the fleshtones on a7S footage so flat and plastic like.

To the first. He is intentionally trying to get this look. You don't have to like it or his manner but it is intentional as he is trying to get this nineteen fourties - fifties glam glow thing happening (BTW I didn't/couldn't watch the whole video so if the photographer took a left turn half way through in how he does things someone will have to tell me).

Second is the issue of plastic skin with the a7S. It's this over exposing by two stops (sometimes more) that people are doing to maximize the dynamic range. If you go back 5 years in this and other forums on how to expose Sony Cinegammas and even Slog on the F3, you will see reports of people getting plastic skin without much dimension or tonal range. There were vigorous pleas, even from some people who now are telling people to over expose, to not push the fleshtones in the +70% tonal range because you are going to hit either the knee or the portion of the gamma curve where it sharply comes down to compress the highlights. You push the skintones too high in the tonal range they go all plastic-like. That is the nature of how the Sony gamma curves work... on any Sony camera.

If your scene is properly lit, you expose properly for your subject matter and choose the right gamma curve for your lighting conditions, your footage will look great and not unlike Sony footage from other cameras in the recent past.

EDIT:

To correct something I said above. The overexposing is not to maximize dynamic range per se but to alleviate some of the visually seen noise in the shadow regions by over exposing and then reducing the exposure in post. If you poke around here you will see many posts and threads going on about this technique.

^This.

Yes, ETTR. A few folks here have shown very powerfully that it reduces noise in the whole image a lot. WIN.
But then a few other folks explained exactly where those mid-tones are hitting the compression knee like you said. LOSE.

I think the take away from all this is that the A7s is NOT an easy camera to color correct. Not impossible of course, but by no means easy.
I personally think the problem is worse that just improper exposure because most of the video I've seen that was done with the other picture profiles, and that was done quickly with all "by the book" methods or preset settings also shows these skintones that just....that just ain't right.
 
Second is the issue of plastic skin with the a7S. It's this over exposing by two stops (sometimes more) that people are doing to maximize the dynamic range. If you go back 5 years in this and other forums on how to expose Sony Cinegammas and even Slog on the F3, you will see reports of people getting plastic skin without much dimension or tonal range. There were vigorous pleas, even from some people who now are telling people to over expose, to not push the fleshtones in the +70% tonal range because you are going to hit either the knee or the portion of the gamma curve where it sharply comes down to compress the highlights. You push the skintones too high in the tonal range they go all plastic-like. That is the nature of how the Sony gamma curves work... on any Sony camera.
So the plastic skintones are due to a knee in the Sony tone curves? I'm confused, I thought the fundamental property of curves like SLOG2 was to convert the linear response of the image sensor into a monotonic logarithmic curve that accurately modeled the luminance response of human eyesight. With a normalized log curve, you can then use a calibrated LUT to emulate any type of tone curve or film emulsion, regardless of the original camera model. Adding a highlight knee to a logarithmic curve would sabotage its value as an independent standard, making it just as proprietary as any non-log tone curve. Do SLOG2 videos require Sony-specific LUT's calibrated to deal properly with the knee?
 
I don't understand this thread - the guy is shooting stills surely in raw - how he processes them later is, well with this style of photography probably using all sorts of dodgy plug ins, chuck in some LED sources and you have a recipe for horror.

---

Slog2 compresses the highlights and will go plastic 'up top' this stops one doing ETTR or getting exposure wrong - which if it is like Slog1 makes it completely horrid, Slog3 however is basically linear to clip and therefore grades like a baby.

S
 
Careful. You will note I said knee and gamma curve. Once you get above 70% in the tonal range all the gamma curve start to compress the tonal range left in the higher end. The SLOG gammas just do it A BIT more gradually that the Cinegammas and Hypergammas. You can invoke a knee on the cinegammas and hypergammas and some of them have a knee set by default or others using the camera will play around with them unwittingly. Here's a chart that does some illustration of what is going on.

s-log-hg-curve.jpg

Remember the knee can be moved around and lots of people set the knee to kick in at 85% or even lower. The graph above shows it much higher which a lot of people DO NOT do. Over expose, you have a knee at 85 and you are asking for trouble unless you know exactly what you are doing and expose the faces in your shot appropriately. You can mess up the skin in SLOG, you just have to work a bit harder to make it look bad.
 
Last edited:
I don't understand this thread - the guy is shooting stills surely in raw - how he processes them later is, well with this style of photography probably using all sorts of dodgy plug ins, chuck in some LED sources and you have a recipe for horror.

You are right, of course, on the RAW and him shooting stills. That is why I separated the original post into two issues with the real issue for discussion being the plastic skin issue with some a7S video footage.
 
The graph you posted basically shows why all sony stuff looks horrid with that compression of the highlights, SLog was designed for 444 where even though the compression is there it still untangles nicely due to the bit depth oversampling compared to the 8bit delivery space.

This is the graph you need - we see Slog3 is nearly linear and free of the dreaded kne, any knee will have that Sony look which I dislike.. (slog 1 and 2 have knee - or rather highlight roll - and are 'unusable' in 8 bit and 'marginal' in 10bit)

http://community.sony.com/t5/image/serverpage/image-id/9114i82ED2F040EF5641B?v=mpbl-1
 
The graph you posted basically shows why all sony stuff looks horrid with that compression of the highlights, SLog was designed for 444 where even though the compression is there it still untangles nicely due to the bit depth oversampling compared to the 8bit delivery space.

This is the graph you need - we see Slog3 is nearly linear and free of the dreaded kne, any knee will have that Sony look which I dislike.. (slog 1 and 2 have knee - or rather highlight roll - and are 'unusable' in 8 bit and 'marginal' in 10bit)

http://community.sony.com/t5/image/serverpage/image-id/9114i82ED2F040EF5641B?v=mpbl-1

According to Sony charts S-log2 doesn't have a knee, it is just steeper.
http://community.sony.com/sony/atta...ry_for_S-Gamut3Cine_S-Gamut3_S-Log3_V1_00.pdf
(see page 3)
 
I didn't read the entire thread, but I'm confused by the topic. I always thought skin tone and color issues was something common among all DSLRs, not specific to any in particular. So, why this topic about the A7s in specific?

Skin tone wise I've seen ugly results with the GH4 as well. Am I missing something?
I'm not being negative here, I'm just confused...
 
I didn't read the entire thread, but I'm confused by the topic. I always thought skin tone and color issues was something common among all DSLRs, not specific to any in particular. So, why this topic about the A7s in specific?

Skin tone wise I've seen ugly results with the GH4 as well. Am I missing something?
I'm not being negative here, I'm just confused...

Youve just picked the two DSLR type cameras that seem to have the most problems delivering good skin tones. Right out of the camera Canon and Nikon render much better and more pleasing skin tones - hence their dominance in the wedding video market.

All cameras have their own unique look - sometimes called "color science".
I have a Panasonic, but I like its natural colors far less than what comes out of a 5DmkIII or a Nikon D750/810.

Everyone will have their own preferences in camera character.

The A7s is being discussed here because it seems to frequently give its users real difficulty in achieving good skin tones.
 
The other problem is those little led hand held lights . Not sure of the colour temp. they put out, but a recipe for crap and all shot in great light , magic hour .( low light yes but doesn't this camera excel in it? )

ps. @ joejitsu thanks .
 
This is all post processing of RAW files. It's not S-log or any other setting. This is a photo shoot by a pro photographer. Dont get video settings and RAW stills mixed up. Very different.
 
Back
Top