Matching Red Epic x Sony A7SII

[Kris Tait]

Hello
I have a larger commercial/documentary production coming in July where we'll be shooting predominantly on Red (Epic X + Dragon), though in some longer dialogue scenes we will be mixing in the Sony a7sii. I've shot extensively on the Sony Fs700 and a7s for a few years, mostly in sLog2, not a fan of 3 on the 8bit cameras. I'm curious to hear if anyone has managed to 'successfully' match Sony footage with Red, and what their profile setups were on both cameras. Although sLog2 will perhaps give me more of a blank canvas in post to
match it to the Epic I'm still apprehensive about going that route. Also if people have matched the Sony to the Red, or if they found it easier the other way around, matching the Red to the Sony (which would obviously be a shame due to Sonys notoriously poor color science), or if they were both matched to a middle common denominator. I'm a fairly seasoned grader due to all the Sony shooting over the years so I'm not shy of putting a bit of grading on my plate in post, but I'm also planning on finally using an x-rite color checker this time around, and if people have found that to be the missing link when matching cameras.
If anyone has any profile recipes they're interested in sharing which lend themselves well to
matching the two in post we would be extremely interested in hearing your findings and much appreciated. Unfortunately the budget doesn't have room for a complete Red fleet, nor will there be storage for shooting so much dialogue using only R3Ds.
Thanks for reading and looking forward to hear your findings.
Thanks
Kris

 

[Kris Tait]

Links to resources, tutorials, blogs, videos etc are also extremely welcome if you happen to have stumbled upon anything relevant in the recent past. All info is of interest. I have a fairly tried and true method of attaining exceptionally Average results. But never anywhere in the 'AaaHaaaa!' range, which is what we are striving for here finally. Have a special recipe or secret sauce?
Many thanks,
Kris Tait

 

[Marcus Friedlander]

Have you considered using ACES?

I've done a good amount of research on the subject, but I have yet to run a project through an ACES pipeline yet so I can't really offer you any specific pointers. But, these type of problems is why ACES was invented.

 

[Hrvoje Simic]

S-Log won't give you a blank canvas for color response, S-Log is just a gamma curve. Color science differs greatly. Typical post work is not sufficient. S-Gamut is significantly messed up, S-Gamut3.Cine in newer cameras is best so far.

There are some useful transforms on my web, designed per particular camera's color response, correcting the results of some Sony cameras deep within it's signal. I haven't analysed A7S2 yet.

S-Log 3 is very limited in 8 bit compressed recording and causes problems in upper signal range when gamma corrected. S-Log2 works better, with varying results in higher luminosity gradients such as blue sky, so don't expect miracles there.

You'll have to find a compromise between Red and 8 bit results from Sony because latter gives less maneuverability and will fall apart sooner.

 

[Zeb B]

S-Log won't give you a blank canvas for color response, S-Log is just a gamma curve. Color science differs greatly. Typical post work is not sufficient. S-Gamut is significantly messed up, S-Gamut3.Cine in newer cameras is best so far. You have some useful transforms on my web, designed per particular camera's color response, correcting the results of some Sony cameras deep within it's signal. I haven't analysed A7S2 yet. S-Log 3 is very limited in 8 bit compressed recording and causes problems in upper signal range when gamma corrected. S-Log2 works better, with varying results in higher luminosity gradients such as blue sky, so don't expect miracles there. You'll have to find a compromise between Red and 8 bit results from Sony because latter will fall apart easier.

Curious to see the S-L2 v S-L3 test results if you know anywhere on the interweb. Would also love to see how others grade RED and A7 together since that is a cam combo that is not going away for me soon

 

[Ted Parkes]

I have the Sony a7rII that I have used as my 2nd camera on a couple low budget projects. It works. I keep it simple, white balance both cameras, shoot a colour chart ( x-Rite) and then use the x-Rite colour checker tool in Resolve to match them up. Works extremely well in some cases but pretty well in most. Anything going to broadcast I send out to a proper colourist for final grading.

 

[Kris Tait]

Exactly Zeb. If there's a recipe for matching these two cameras it's a great camera combo. We're really interested in seeing what's out there and what people have managed, especially which gamma profile they are using as a base starting point on the Sony. We've had some fairly good results so far, but felt it involved dumbing down the Red to match the Sony too much, rather than finding a starting point with the Sony which allows us to match closer to the Red.
Ill be posting some results over the next few weeks which people can analyze. Would be great to finally crack this nut.
Cheers

 

[Hrvoje Simic]

Curious to see the S-L2 v S-L3 test results if you know anywhere on the interweb. Would also love to see how others grade RED and A7 together since that is a cam combo that is not going away for me soon

I did comparative analysis. S-L3 has a bit more DR in highs and smoother roll off to highs and shadows. Looks overall "finer" in the context of contrast distribution. S-L2 is a bit more harsh and burns highs sooner. S-L3 requires a 10 bit codec and falls apart soon in highs on compact cameras due to limited color sampling affecting gamma correction range of a compressed slope.

I keep it simple, white balance both cameras, shoot a colour chart ( x-Rite) and then use the x-Rite colour checker tool in Resolve to match them up. Works extremely well in some cases but pretty well in most.

It works (in a general way and to a limited extent) when a) your scene has less tonalities in camera-specific critical range and more in more stable range, and b) when scene tonalities are closer to the ones on the chart, while everything else is a gamble based on what an algorithm "assumes" through that data of particular scene's light properties. It would be great if it was that simple.

 

[Kris Tait]

I don't touch sLog3 on the a7sii. But some of the closest matches we've made were using sLog2 and RedLog as a starting point. Though we never made it back to the nice color range that REDgamma2-4 have straight out of camera. Something tells me Cine4 with Sgamut.cine on the A7Sii will be the starting point we go with on the Sony. But even there, we've still haven't reached fantastic results. The Red always gets dumbed down to match the Sony in the end. We're hoping to 'lift' the Sony up to the Red color space instead. If at all possible. Hopefully someone has the secret sauce.

 

[Hrvoje Simic]

Something tells me Cine4 with Sgamut.cine on the A7Sii will be the starting point we go with on the Sony. But even there, we've still haven't reached fantastic results.

Cine4 with S-Gamut3.Cine. But needs more tweaks for more optimized response and has less DR than S-L2, about a stop.
Works fine in most cases.

We're hoping to 'lift' the Sony up to the Red color space instead. If at all possible. Hopefully someone has the secret sauce.

There is no secret sauce for matching, too many variables. That needs manual work.

 

[Kris Tait]

There's always a secret sauce. No just kidding. But we mean rather which CineGamma profile, under which conditions will keep us in a color space that allows us to work the Sonys 8bit footage into an area that resembles the Epic. We've tried a myriad of Arri LUT's that are impressive, but resemble a very pleasing Rec 709 palette, not so much the Red 'look' though.

 

[Hrvoje Simic]

But we mean rather which CineGamma profile, under which conditions will keep us in a color space that allows us to work the Sonys 8bit footage into an area that resembles the Epic.

Cinegamma affects the gamma. Color space is just a space. It defines the saturation range and primaries. Camera's spectral response and processing defining its color performance is something else.
Gammas and color spaces do not define spectral response, spectral response defines the effect of gammas and color spaces. It is much more complex than it seems.

Cine4 is "okay" gamma, and useful for capturing more DR than typical camera settings and finer roll-off to whites. S-Log2 is better, S-Log3 even better but they come at a cost of usability with 8 bit recordings. S-Gamut3.Cine gives best color performance so far but it's also not perfect and it still depends on what the camera and its sensor and processing give. Same color profile will give different results on different Sony cameras.

There are many factors defining a "look" and all of this is not something transferable through a simple recipe.

 

[Jason beaumont]

I recently shot a short where to used the sony A7S as the be camera.
First pic is RED Scarlet MX
Second Pic IS A 7S

My last short also used this combination and it worked pretty well. The main thing I noticed is the A7S has a lot more sharpness than the Scarlet already baked in, but its pretty easy to match.

 

[Kris Tait]

Think you missed the point of the post. It was to hear what peoples experiences were with matching the two cameras, and which results they found best with given profiles. Not a platform to showcase how much you think people need schooling in the technicalities of sony's color science and everything else in the universe. Stick to the topic please.

Cinegamma affects the gamma. Color space is just a space. It defines the saturation range and primaries. Camera's spectral response and processing defining its color performance is something else.
Gammas and color spaces do not define spectral response, spectral response defines the effect of gammas and color spaces. It is much more complex than it seems.

Cine4 is "okay" gamma, and useful for capturing more DR than typical camera settings and finer roll-off to whites. S-Log2 is better, S-Log3 even better but they come at a cost of usability with 8 bit recordings. S-Gamut3.Cine gives best color performance so far but it's also not perfect and it still depends on what the camera and its sensor and processing give. Same color profile will give different results on different Sony cameras.

There are many factors defining a "look" and all of this is not something transferable through a simple recipe.

 

[Kris Tait]

Hey Jason, thanks for sharing your results. Don't suppose you remember which profile combo you were shooting the sony with...? Smells like sLog2... Looks very exceptable for matching, I would certainly be able to work with these results. you're definitely right about the sharpening in the sony. The main crossroads for me is whether I should shoot both in Log so as to have the most neutral starting point, or to tweak a cinegamma (probably cine4) to try and bring it close to the epic. Though this is ultimately what has eluded us so far in grade.

Thanks for posting, much appreciated!

I recently shot a short where to used the sony A7S as the be camera.
First pic is RED Scarlet MX
Second Pic IS A 7S

My last short also used this combination and it worked pretty well. The main thing I noticed is the A7S has a lot more sharpness than the Scarlet already baked in, but its pretty easy to match.

 

[Ryan Lightbourn]

I shot a feature on Scarlet + A7SII. The A7SII was yuck compared to the Scarlet...not gonna lie. With perfect lighting, things were great, but with shots that were rushed, the Scarlet still pulled through with a cinematic look while the A7SII looked like low budget ass. I've used Sony for years, so it wasn't user error. Definitely don't use SLOG3 (go with 2).

 

[Kris Tait]

Thanks for your input Ryan. Also my experience with sLog3 on anything lower than the fs7. No go on the 8bit. I'm tempted to leave sLog entirely and try to build a look in camera with Cine4 and some tweaking to bring it as close to REDGamma2 or 3 as possible. But haven't had luck bringing the Sony up to that flavor yet. Hoping the x-rite color checker might help that last 10% of color nuance to match them eventually. Looking for someone who's tried that route...

I shot a feature on Scarlet + A7SII. The A7SII was yuck compared to the Scarlet...not gonna lie. With perfect lighting, things were great, but with shots that were rushed, the Scarlet still pulled through with a cinematic look while the A7SII looked like low budget ass. I've used Sony for years, so it wasn't user error. Definitely don't use SLOG3 (go with 2).

 

[Jason beaumont]

This was shot with the red scarlet and sony a7sII . it was a quick 4 hour shoot and we didn't use both in the same shot. You can tell where the sony was used.

 

[Derek Carlson]

Whatever you figure out, would love to hear what you work out. I'll be using my a7sII in conjunction with the S-W as well. I shoot almost completely cine4 and s-gamut3. If I want to push the DR on the a7s i'll shoot occasionally on slog-2, but won't touch 3.