Can you please post a few examples of work you've done with these settings? Thanks!
Consensus on Sony A7SII Picture Profiles
- Thread starter jordancslee
- Start date
Can you please post a few examples of work you've done with these settings? Thanks!
adaptation
Active member
Unfortunately its all client work so im privacy bound - but i will attempt to go shoot some tests later this month and post them.
I just tested cine4-sgamut3 combo and compare 8-16bpc modes after manual grading (After Effects). I hope it helps to fix bending problem. Please watch it in 1080p @ full screen to see details.
https://vimeo.com/157424898/767721379e
https://vimeo.com/157424898/767721379e
Bram Van Workum
Member
Hi, I've been reading up a bit here on A7s settings. My current headhurt is matching the A7s to GH4 or any other non-sony non s-gamut(3) cam. So far i've tried some basic stuff by having each cam look at an Xrite colorchart and derive a sort of correction LUT of that with my resolve 12.3. The result is that all the images i got from the A7 (slog or otherwise) look fine and the GH4 looks fine but theres still a lot of difference between the A7 clips and the GH4 clips. For one the GH4 seems much more blue at the same kelvin setting. I've recorded the Xrite at every possible setting with each cam but the result isnt really good enough.
So the question is: what setting on the A7s would make it more flexible. sLOG2 on sgamut3 or on cine-3 gamut? The reason im focussing on the sony is that im less familiar with the gh4 and that the panasonic seems less tweakable. I've tweaked it to data settings instead of 16-235 video levels and i've shot my xrite footage at matching settings (iso 200, f5.6 etc etc) the Sony seems to give material thats more flexible so id want to be able to move the sony recordings towards the GH4 recordings for matiching. The place where i work has a lot of different cams... LUTcalc would be a thing to look into i guess.
So the question is: what setting on the A7s would make it more flexible. sLOG2 on sgamut3 or on cine-3 gamut? The reason im focussing on the sony is that im less familiar with the gh4 and that the panasonic seems less tweakable. I've tweaked it to data settings instead of 16-235 video levels and i've shot my xrite footage at matching settings (iso 200, f5.6 etc etc) the Sony seems to give material thats more flexible so id want to be able to move the sony recordings towards the GH4 recordings for matiching. The place where i work has a lot of different cams... LUTcalc would be a thing to look into i guess.
adaptation
Active member
I haven't used the GH4 so i can't speak about it, but definitely use sgamut3.CINE over sgamut3 - cine was specifically made to be easier to grade in exchange for a smaller color spectrum. Unless you very advanced grading in resolve, i would stay away from sgamut and gamut 3. Sgamut3.cine, PRO and Cinema are my personal favorites.
Aside from color spaces - play around with saturation in s-log2 - with a mix of sat + color adjustments, i think you could get close results.
Has anyone tested the dynamic range of Brandon Li's profile: Autumn leaves creative style (-3,0,-3 DRO 4) ? Is it comparable to Cinegammas picture profiles since it has lowered contrast settings ?(vs cine 4 for instance?). After seeing this https://vimeo.com/157995688 and brandon li's work, I think this is the profile I would pick but i just dont want to sacrifice lot of dynamic range. What do you think?
Viscount Omega
Carbonite Member
The problem is the highlights--even if you underexpose and bring it up in post. I really struggled with whether to use the "Light" profile on my film. Really loved the colors, skin tones. The problem is the video-ish looking white hot spots (the sun, light bulbs, reflections from shiny objects--even on people's faces) that strain my eyes after a few minutes. I can tolerate a short music video of it though. Another thing, you can't get that gentle filmic curve of light and shadow across a person's face. s-log all the way.
theprestige
Active member
How much do you guys have detail set to? I listened to Philip Bloom's advice and put it down to -7, but i'm thinking maybe just leaving it at -3 might be better? I don't often apply sharpening in post anyway and when I do it's very little.
Good advice about using different gamma for 100fps shots. The noise in those shots can really be quite hideous and certainly messed me up on my first project with the A7S II.
I really feel like the best footage I have seen with either A7S's is the ones by Casey Wilson so SLOG 2 seems like it's really worth while and i'll be using it first time for a client later this week.
Good advice about using different gamma for 100fps shots. The noise in those shots can really be quite hideous and certainly messed me up on my first project with the A7S II.
I really feel like the best footage I have seen with either A7S's is the ones by Casey Wilson so SLOG 2 seems like it's really worth while and i'll be using it first time for a client later this week.
Does anyone know if Brandon used an external recorder?
Just came across this video: https://vimeo.com/167200822
Can you guys tell me if Cine2/Cine4 with raised Black Level is a good alternative to SLOG?
I'm tired of ISO 1600 in bright daylight and for slow motion.
Thanks!
Can you guys tell me if Cine2/Cine4 with raised Black Level is a good alternative to SLOG?
I'm tired of ISO 1600 in bright daylight and for slow motion.
Thanks!
Viscount Omega
Carbonite Member
I love the Cine gammas as long as I avoid windows, sunlight, light bulbs, bright reflections, city lights, people's faces and film everything in a warmly-lit closet. LOL.
mojo43
Well-known member
No he does not.
paulcurtis
Veteran
FWIW and IMHO and all the other acronyms that mean this is highly subjective and personal comments.
I find the 'advanced' PP settings a bit pointless. Adjusting the colour globally as part of a 'named profile' makes little sense because the light that you're shooting under will impart more colour effects. +1 M would be totally obliterated with a dodgy LED panel. I suspect even a lens choice would make all the difference.
Adjusting the colour matrix is only really useful on a scene by scene basis if you're dealing with challenging light. Otherwise the defaults would be different. But then there are no tools on the camera to actually evaluate the colour properly anyway, making it all a bit moot.
Then black gamma and knee. All this is doing is taking your base profile curve and reshaping it But in ways that aren't all that useful most of the time anyway, it's not changing the absolute black or white points, because it's only modifying the basic profile curve used. If you lower the black point for example, you're just lowering digital numbers, you're not capturing more. If you up the black gamma you're creating a bump in the curve, but you're not capturing different detail, just visually raising the shadows, the exact same way you would in post.
The only settings that would make a difference is the basic gamma curve and the colorspace (and WB)
The Slog2 curve distribution records more range. The cine curves record a tiny bit more tonality in the mids. Can you see a difference? I've struggled to ever see one, i've pushed and pulled at the data but in the tests that i've done i've managed to grade the slog2 image to the same as the cines, down to the level noise as well. (bearing in mind that 1600 iso is pushed 2 stops compare to cine). There's a graph on our blog under exposing for slog that shows the actual code values being used.
I don't think aside from the main setting options there is anything that the PPs offer that can't be graded after the fact from a practical viewpoint.
Also if you get into messing with cine/slog curves then if you use any LUTs you'll be feeding the wrong curve into them. And if you mess with the colour matrix then again any workflow where the colorspace is a known quality would be different. Hence you're better off recording clean.
Now, just to contradict myself. There are a few places where it *might* have a tiny influence:
1. Raising the shadows and the black gamma will affect the way the image is compressed. Compressors throw away shadows, raise them to retain. Perhaps. A tiny tiny bit
2. Increasing the saturation in the colours may record a higher level of colour in an 8 bit container, leading to better colour tonality. A tiny tiny bit.
I find just shooting slog2/sgamut3.cine all of the time the better option. It records a little more, it's a very very established post workflow but you just have to keep an eye on exposure.
cheers
Paul
I find the 'advanced' PP settings a bit pointless. Adjusting the colour globally as part of a 'named profile' makes little sense because the light that you're shooting under will impart more colour effects. +1 M would be totally obliterated with a dodgy LED panel. I suspect even a lens choice would make all the difference.
Adjusting the colour matrix is only really useful on a scene by scene basis if you're dealing with challenging light. Otherwise the defaults would be different. But then there are no tools on the camera to actually evaluate the colour properly anyway, making it all a bit moot.
Then black gamma and knee. All this is doing is taking your base profile curve and reshaping it But in ways that aren't all that useful most of the time anyway, it's not changing the absolute black or white points, because it's only modifying the basic profile curve used. If you lower the black point for example, you're just lowering digital numbers, you're not capturing more. If you up the black gamma you're creating a bump in the curve, but you're not capturing different detail, just visually raising the shadows, the exact same way you would in post.
The only settings that would make a difference is the basic gamma curve and the colorspace (and WB)
The Slog2 curve distribution records more range. The cine curves record a tiny bit more tonality in the mids. Can you see a difference? I've struggled to ever see one, i've pushed and pulled at the data but in the tests that i've done i've managed to grade the slog2 image to the same as the cines, down to the level noise as well. (bearing in mind that 1600 iso is pushed 2 stops compare to cine). There's a graph on our blog under exposing for slog that shows the actual code values being used.
I don't think aside from the main setting options there is anything that the PPs offer that can't be graded after the fact from a practical viewpoint.
Also if you get into messing with cine/slog curves then if you use any LUTs you'll be feeding the wrong curve into them. And if you mess with the colour matrix then again any workflow where the colorspace is a known quality would be different. Hence you're better off recording clean.
Now, just to contradict myself. There are a few places where it *might* have a tiny influence:
1. Raising the shadows and the black gamma will affect the way the image is compressed. Compressors throw away shadows, raise them to retain. Perhaps. A tiny tiny bit
2. Increasing the saturation in the colours may record a higher level of colour in an 8 bit container, leading to better colour tonality. A tiny tiny bit.
I find just shooting slog2/sgamut3.cine all of the time the better option. It records a little more, it's a very very established post workflow but you just have to keep an eye on exposure.
cheers
Paul
Sounds legit. Thanks for your thoughts.
However in my documentary style scenario I don't always have full control over exposure.
Most of the time I'm shooting in bright daylight and slow motion is used a lot as well. I do have a very good colorist.
I think that's why I'm leaning towards one of the Cine curves.
I'm just not sure which of the two.
paulcurtis
Veteran
I think you should stick to Slog2. It will retain more of your scene and allow for post based exposure adjustments much easier than if you recorded under a cine style curve.
The point about Slog2 is that each stop of light has roughly the same code values, so making shot to shot exposure adjustments is mathematically quite easy. But if you're running an arbitrary cine curve, more so if you've made adjustments, then there are no tools to adjust exposure aside from just eyeballing (which is fine too).
I would go out shoot wildly as a test, miss exposure, try to do all the things that typically happen when running and gunning and then sit down with your colourist and see what you can do with it.
If they have a good log based workflow i think they may find that footage easier to massage.
If you were shooting say a scene with *limited* and *controlled* range then perhaps the cines would offer a tiny bit more tonality.
cheers
Paul
Viscount Omega
Carbonite Member
I'm finding that too. I did a test a while ago where I shot in Rec. 709 color with s-log2 gamma, thinking I could skip the Rec. 709 conversion LUT, but it turns out the blue (and sometimes the red) channels can be blown out--even with the saturation set at zero. I don't see this problem yet with sgamut3.cine.
It's an extra step to add the Rec. 709 LUT, and it degrades the footage slightly but so far it's better.
Ben Scott
Veteran
I'm using my A7sII more and more now and still haven't got it dialled in perfetly. What are people's issues with banding when using S-log gammas and is there a quick fix to avoid them? Or isit user error and not exposing properly.
I know with my FS7 I get better results exposing 90% white at 75 instead or Sony's recommended 61IRE. Is it the same here?
I know with my FS7 I get better results exposing 90% white at 75 instead or Sony's recommended 61IRE. Is it the same here?