What do you think about colour grading?

[kye]

I hear very mixed things about colour grading and I'm curious to get a bit more info on how people think about it.

My theory is that it's actually simpler than people think, but maybe I'm not getting the full picture.

[webrunner5]

I think Editing, not coloring, grading is more important, and Audio might as well be like trying to learn Chinese. Look how much stuff is just a music track and really no natural audio at all. Kind of depressing after awhile. What about lighting. I mean sure if all of it is great that is what is great. That is what a Pro looking thing is. But what really is the better progressions. In what order do you concentrate. Hard to be good at all of it. I think color grading maybe the hind end thing to master. Video is hard, real hard to master.

Not trying to derail your thread. But yeah will be interesting to see your results. I am glad to see you pushing to get better at it. It is the difference between good and great stuff.

[HockeyFan12]

2 hours ago, kye said:

My theory is that it's actually simpler than people think, but maybe I'm not getting the full picture.

I think like most things, it's as simple as you make it. I've worked with really good artists who do complicated things, but when I try to make it complicated, I usually screw it up. . Imo Gall's Law applies more obviously to color grading than it does to most anything else, but that doesn't mean there aren't complex systems that work.

[kye]

Interesting results - thanks all who took the poll so far, if you haven't, please do

Currently, these seem to be the trends:

• People seem to think that colour grading is an important part of the process of making a film (it's useful or magic)
• The majority of people go beyond using a LUT into more custom adjustments, and also want to get better results
• The minority who basically only use a LUT are more happy with their results than wanting better results
• The vast majority believe that you don't need Resolve level grading software

My theory (that caused me to create this poll) was that getting a great grade is more about using the simple controls well, rather than having all the tools in the world.  I wondered if I should try and 'prove' that by exploring some grades using only basic tools and sharing them here.

It looks like lots of people want to get better results, but there are also lots of grading tutorials out there and I'm not sure if people are watching them or not.  I can imagine that we're all wanting to get better results from every aspect of our film-making, but it's a matter of time and energy, rather than availability of information or resources.

Is there interest in seeing before-and-after grading examples that also show the adjustments made to create the grade?  Would that be useful?

[heart0less]

1 hour ago, kye said:

Is there interest in seeing before-and-after grading examples that also show the adjustments made to create the grade?  Would that be useful?

Yes, yes, yes! 

[Ehetyz]

I've seen good on-set camerawork totally ruined by terrible color grading, and mediocre work absolutely salvaged by a good, high effort grade. I've seen both happen to my footage and to the footage of other people. 

Nowadays I grade most of my work myself after a series of cases where the final graded footage looked way worse than what I had on my simple single LUT monitoring (due to being handed over to editors not experienced in grading). I would still like to do way better myself - after years with Resolve there are tons of things I'm still learning.

Good color grading is absolutely essential in this day and age, at least on cinematic footage shot on LOG or RAW.

[Django]

I think we have to distinct making tweaks and applying a LUT which is what most of us (myself included) are doing and advanced pro color grading which is a complete artform imo.

I got into shooting log & grading mostly out of a necessity because i'm generally unsatisfied with most picture profile in hybrid cameras (aside from Fuji Eterna, Nikon flat) which are often way too sharp/saturated/crushed blacks, need the extra DR and need to fix exposure, wrong camera AWB, color science issues..etc. 

So I'm usually just trying to get things looking correct and then i might apply a LUT for certain look. I might go full manual if i have the time.

I've been experimenting with ACES recently and it looks like it might be a game changer as far as workflow.

On a side note, I do wish hyrbids would either allow LUT import in camera or offer more filmic picture profiles for fast turnover projects.

Pretty sure Sony has the most complete picture profile settings (a little convoluted though). I also like how in Fuji's you can adjust shadows/highlights.

My personal favorite PP is WideDR on Canon's C line.

[kye]

1 hour ago, Django said:

I think we have to distinct making tweaks and applying a LUT which is what most of us (myself included) are doing and advanced pro color grading which is a complete artform imo.

I got into shooting log & grading mostly out of a necessity because i'm generally unsatisfied with most picture profile in hybrid cameras (aside from Fuji Eterna, Nikon flat) which are often way too sharp/saturated/crushed blacks, need the extra DR and need to fix exposure, wrong camera AWB, color science issues..etc. 

So I'm usually just trying to get things looking correct and then i might apply a LUT for certain look. I might go full manual if i have the time.

I've been experimenting with ACES recently and it looks like it might be a game changer as far as workflow.

On a side note, I do wish hyrbids would either allow LUT import in camera or offer more filmic picture profiles for fast turnover projects.

Pretty sure Sony has the most complete picture profile settings (a little convoluted though). I also like how in Fuji's you can adjust shadows/highlights.

My personal favorite PP is WideDR on Canon's C line.

In a sense I disagree with you.  I would say that getting the colours you want by making changes in post is grading, regardless of how you do it.  

There was a discussion on LiftGammaGain forums about colour grading vs colour correction and their opinion was that they are the same thing, because colour work is just doing what is necessary.  They talk a lot about just adjusting contrast and the colour primaries from the colour chart and that if it was shot and lit properly then this is enough to get great results.

After watching a bunch of YT wannabe colourists taking log footage and screwing with it via all sorts of manual methods I then found the pros and they talk about using ACES, Resolve Colour Management, Colour Space Transforms in software or LUT form, and then making simple adjustments to correct for shot-to-shot variance and they're done.  They don't care about being fancy - they care about how efficient they can be with their workflows.  Getting the job done quickly means a higher hourly rate or extra time to really lift the project and deliver a higher quality result.  Of course, it's different if you're colouring a low budget documentary or a high budget Hollywood blockbuster, but the 'leg work' of the process is the same in terms of matching shots, removing anything distracting (like strong colours in the background or whatever).

[webrunner5]

Yeah but on most of these newer advanced cameras if you bust your ass and set Everything in it from Knee, to Black Point, Sat, Contrast, etc., you can get a look You like if you can save it to a custom setting, that works a lot of the time. All those built in profiles are just a general grade, one size fits all thing.

You can eliminate a lot of time and trouble by trying to get it right in camera more than depending on grading for hours, to fix something, that could have been avoided more times than not.

[thebrothersthre3]

I definitely think it is important, also somewhat depending on the camera you shoot on. People who don't know what they are doing shooting in log and then trying to grade in post can definitely be a mess. 

I am color blind and really dislike color grading. I usually just make minor adjustments with color curves, sometimes color wheels, try to get the scopes and waveform looking where they should. Always use a lut with Log. 

[mercer]

Interesting poll and discussion. I struggle with color... both in theory and practice. I’ve always been a tinkerer and with most things in life, I can figure it out but color correction/grading has eluded my tinkering ways. It is an art, and a craft. And the folks who are really great at it are either naturals or well trained.

There are so many facets to the craft that out of respect for it, and the people that take the time to perfect their craft, I hate even referring to myself as a hobbyist colorist.

I can only speak for myself, but I find the most difficult part of color work to be correction. That portion of the craft can make or break the acquisition phase and the grading phase. There is such a symbiotic relationship that it does make sense that shooters are also coloring their own work but traditionally and professionally that is not the case.

I’m guilty of it myself, but I think mistakes in the acquisition phase can make color correction/grading more difficult than it needs to be... 

For instance, most shooters, in most situations, do not need Log footage, or 10bit video, but they use it anyway because they want more latitude... because somewhere, someone told them they needed it.

Proper exposure for 8 or 10 stops of dynamic range is difficult enough, when you start adding 3 or 5 more stops of DR with a Log curve you are entering a different level of skill set and understanding of Log Curves that could be more detrimental than just shooting with a flat Rec709 profile and keeping your color work more basic.

I have not shot in every scenario possible but with my basic understanding of shooting, most situations can be properly shot with 8-10 stops of DR. So if that’s the case, why shoot with a Log curve designed for 12 or more? I’ve personally found 10-12 stops to be a good sweet spot for the type of image I am looking for. If I was a Hollywood DP, I’m sure I could justify 15-18 stops.

I guess what I’m trying to say is... like most things in life, less is more in most instances.

So in summation, this is a learning process and as I learn I realize what I don’t know and more importantly what I don’t need. But color grading is fun and I wish I were better at it. 

[DBounce]

I believe it best to capture the footage as close to the final output as possible. Granted I’ve done some pretty heavy handed stuff. But usually this is so I can understand the limitations of the footage. When I find looks that work I’ll save them as presets for my edits. That way I can have consistency as I switch from angle to angle in my multi-cam shoots. In order to make certain this works you need to have the cameras setup so that they are easy to match in post. I control lighting and manually set everything.

I just about have everything to where I want it now. But honestly it’s been a long time getting here. 

Grading is an important part of the project to me. But with thoughtful planning it does not need to be painful.

[BTM_Pix]

 

1 hour ago, mercer said:

I’m guilty of it myself, but I think mistakes in the acquisition phase can make color correction/grading more difficult than it needs to be... 

I guess what I’m trying to say is... like most things in life, less is more in most instances.

I'm going to make an analogy to recording music because it has already been through the democratisation (for want of a better word) process that film making is going through.

This is a local band from my town making a half decent album in 1967.

Great room, great mics, great mic placement.

Because they'd got those aspects so right, it didn't matter anywhere near as much that they had such limited capabilities of a console like this to fix it in the mix.

My phone has more recording and mixing capability than they had available there but unless I take that same approach to acquisition then I can tweak away forever without really being able to get anywhere near it.

That isn't necessarily about having to have a world class room and world class mics but it does involve approaching the craft of it in the same way but also making an artistic decision about what it is you are trying to capture in the first place.

The same is true for image making.

I think it is great that we have the colour grading tools available to transform and even potentially rescue footage that you would ideally have shot under more controlled circumstances but I certainly hold with it being something to enhance or finesse a solid base.

I actually find one of the Pocket4K's most interesting yet hardly ever discussed features is the ability to record with baked in LUTs as I think creating a virtual film stock and taking the time to learn its response inside out and making those commitments to the final vision when shooting will ultimately yield better and more consistent results over time than the time draining options of noodling around with RAW all day.

To offer a parallel to this, for work I  have to shoot jpeg with very limited latitude time or manipulation wise to make anything other than minor tweaks. 

Due to this, I have the same 'film stock' on the multiple cameras I'm using and I'm completely familiar and comfortable with how it will react to the multiple types of lighting that I have to shoot in in terms of colour balance and dynamic range, how far I can underexpose it to boost shutter speed etc so no image that comes off any of them is a surprise or a challenge to get it where it needs to be.

As a consequence, images fly in and fly out with no delay, angst, analysis paralysis or regret.

By contrast, if you give me a single RAW file, I can fuck about all day with it.

Often with inversely proportional benefits.

[mercer]

@BTM_Pix ... interesting analogy. That band sounds pretty cool... what do they call themselves?

I always forget that you can bake LUTS with the P4K. If I was a smarter person I’d probably buy one but I still think 1080p video is going to make a comeback.

It’s funny, my best results with color has been using the guided color correction tool in Colorista IV. With some saturation and an s-curve, I’ve received my biggest compliments. Anytime I try to sit down and watch tutorials and learn the proper way... eg. left to my own devices... I spend hours and get worse results.

So I agree that it’s great that the tools are available but to master even one discipline in filmmaking is a life long endeavor let alone every single craft.

[HockeyFan12]

10 minutes ago, mercer said:

@BTM_Pix ... interesting analogy. That band sounds pretty cool... what do they call themselves?

I always forget that you can bake LUTS with the P4K. If I was a smarter person I’d probably buy one but I still think 1080p video is going to make a comeback.

It’s funny, my best results with color has been using the guided color correction tool in Colorista IV. With some saturation and an s-curve, I’ve received my biggest compliments. Anytime I try to sit down and watch tutorials and learn the proper way... eg. left to my own devices... I spend hours and get worse results.

So I agree that it’s great that the tools are available but to master even one discipline in filmmaking is a life long endeavor let alone every single craft.

This is the most basic and oldest color grading textbook I can think of. I must have read it six years ago. But I found it more useful than online tutorials, because of its broader focus:

https://www.amazon.com/Color-Correction-Handbook-Professional-Techniques/dp/0321929667/ref=asc_df_0321929667/?tag=hyprod-20&linkCode=df0&hvadid=312091457223&hvpos=1o2&hvnetw=g&hvrand=14846167971322217554&hvpone=&hvptwo=&hvqmt=&hvdev=c&hvdvcmdl=&hvlocint=&hvlocphy=9001876&hvtargid=pla-524043500801&psc=1&tag=&ref=&adgrpid=62820903995&hvpone=&hvptwo=&hvadid=312091457223&hvpos=1o2&hvnetw=g&hvrand=14846167971322217554&hvqmt=&hvdev=c&hvdvcmdl=&hvlocint=&hvlocphy=9001876&hvtargid=pla-524043500801

I know many people here claim camera doesn't matter much if you can grade really well, but I readily admit I can't. So for me I think camera matters more.

And I think there are things you can do with LUTs that are difficult to do without them. Light Iron and CO3 both have proprietary LUTs. That's the smaller part of what makes those places as good as they are, and they work they do really is fantastic. But I think it's still a factor.

Anyhow, I highly recommend the book.

 

[BTM_Pix]

20 minutes ago, mercer said:

 

I always forget that you can bake LUTS with the P4K. If I was a smarter person I’d probably buy one but I still think 1080p video is going to make a comeback.

The other interesting yet seldom mentioned function it does is having the option to shoot in 1080p

20 minutes ago, mercer said:

To master even one discipline in filmmaking is a life long endeavor let alone every single craft.

I was dragged along to watch Aquaman last week (pro tip, always go and see the 4dX versions of films you'll hate as at least you can enjoy the ride).

I watched the credits to the bitter end because I was fascinated by how many people were involved in making it. I've tried without joy to find the figure online but I doubt a guess of 600-700 would be too far off.

Sobering in one way but inspirational in another as just getting a 2 minute piece together as a one man band is a big achievement !

[mercer]

17 minutes ago, HockeyFan12 said:

This is the most basic and oldest color grading textbook I can think of. I must have read it six years ago. But I found it more useful than online tutorials, because of its broader focus:

https://www.amazon.com/Color-Correction-Handbook-Professional-Techniques/dp/0321929667/ref=asc_df_0321929667/?tag=hyprod-20&linkCode=df0&hvadid=312091457223&hvpos=1o2&hvnetw=g&hvrand=14846167971322217554&hvpone=&hvptwo=&hvqmt=&hvdev=c&hvdvcmdl=&hvlocint=&hvlocphy=9001876&hvtargid=pla-524043500801&psc=1&tag=&ref=&adgrpid=62820903995&hvpone=&hvptwo=&hvadid=312091457223&hvpos=1o2&hvnetw=g&hvrand=14846167971322217554&hvqmt=&hvdev=c&hvdvcmdl=&hvlocint=&hvlocphy=9001876&hvtargid=pla-524043500801

I know many people here claim camera doesn't matter much if you can grade really well, but I readily admit I can't. So for me I think camera matters more.

And I think there are things you can do with LUTs that are difficult to do without them. Light Iron and CO3 both have proprietary LUTs. That's the smaller part of what makes those places as good as they are, and they work they do really is fantastic. But I think it's still a factor.

Anyhow, I highly recommend the book.

 

I couldn’t agree more about the importance of cameras. When I first bought my 5D3, I happened upon a cheap Pocket as well. The Raw footage from the 5D was soo much easier to get to look good than the Pocket. I could get a nice image with half of the steps. It’s funny how certain cameras or Log curves are so much easier to navigate than others. You’d think LogC would be difficult but I find it to be one of the easier Log curves to bring to life... also Canon Log and Nikon Flat are very good picture profiles/Log curves in my opinion.

@hyalinejim recommended that book to me a while ago and I still haven’t bought it. I wish I had. I’m going to put it in my cart and order it. Thanks. 

[HockeyFan12]

1 hour ago, mercer said:

I couldn’t agree more about the importance of cameras. When I first bought my 5D3, I happened upon a cheap Pocket as well. The Raw footage from the 5D was soo much easier to get to look good than the Pocket. I could get a nice image with half of the steps. It’s funny how certain cameras or Log curves are so much easier to navigate than others. You’d think LogC would be difficult but I find it to be one of the easier Log curves to bring to life... also Canon Log and Nikon Flat are very good picture profiles/Log curves in my opinion.

@hyalinejim recommended that book to me a while ago and I still haven’t bought it. I wish I had. I’m going to put it in my cart and order it. Thanks. 

I would agree completely, especially about Log C, but I don't want to write cameras off until I have used them more. 

For instance, I thought the F5 was bad when I first shot with it in SLOG 2, but then I worked with an experienced DP who exposed better and shot a Kodak emulation LUT with it and the footage looked good.

Likewise, I dislike Canon Log 2, which is a surprise since you'd think it's a copy of Log C from looking at it. But it's not as good. It's flatter and thinner. But I generated a Canon Log 2 to Alexa conversion LUT and used a Log C to rec709 LUT on top of that and suddenly it looks fine to me.

I think this kind of thing leads to a bias where people just like what they're used to or what appeals to their personality. I am sort of an Arri/Canon fanboy and maybe it's just because they focus on ease of use. On the opposite end are people who seem to enjoy seeing how far they can push things. I'll never understand some of the more arcane Resolve workflows. It's like a complicated solution in search of a problem. But it's just not for me.

[mercer]

@HockeyFan12 obviously you have a lot more experience with different cameras than I have. Even my experience with LogC is based on Log conversions in different MLV apps but since LogC to Rec709 LUTS respond properly to the converted files, I must assume they’re fairly accurate.

Have you ever used the FS5? The small form factor, internal 10bit 1080p and external Raw seems very appealing. The used prices seem pretty nice as well.

OT - I just bought a Nikkor 85mm 1.4 to go along with my 35mm 1.4. Did you say you have the 85mm?

[kye]

44 minutes ago, HockeyFan12 said:

I would agree completely, especially about Log C, but I don't want to write cameras off until I have used them more. 

IIRC Juan Melara said that he likes LogC because it's closest to the cineon log curve.  Having codecs that have knees in their luminance response makes them very difficult to work with unless you convert them and iron those knees out, otherwise by changing anything you're essentially compressing on one side of the knee and expanding on the other with every adjustment.

49 minutes ago, HockeyFan12 said:

I'll never understand some of the more arcane Resolve workflows. It's like a complicated solution in search of a problem.

Absolutely.  This is why I recommend against YT wannabe colourists.

I think of grading a bit like golf.  You start a long way away from where you want to be, so the first adjustment is large, but crude.  Each further adjustment should be progressively more refined and get you closer to where you want to be.  Ultimately, you want each adjustment to get you much closer to the hole.  In golf, you see people hitting the ball and sometimes it's further away from the hole after they hit it than it was before they hit it.  This is the same for the amateurs on YT - you see them make adjustment after adjustment and each one improves something but creates almost as many problems as it solves.
One of the guys at LiftGammaGain said that you know someone is clueless when they adjust something in a node that they have already adjusted in a previous node.  I really believe that - if you later on adjust something you already adjusted, then it means you didn't do it right in the first place.  This is a broad statement and there are exceptions where technically it's not true, but the principle still stands.

I think this is why when we spend ages doing complicated grades we often just make matters worse, instead of approaching the end result - it's because we're just not good enough at each adjustment.