kye

No worries! A little bit of excess enthusiasm is the least of our issues!! What are you planning on using the RAW for? or is it just a recent discovery?

Nice! They definitely look like movie stills and not photographs. What was the show LUT? Interesting range of tones/hues there.

My understanding of it was that a colour timer would take the negative and make a positive print using a special machine where each frame of the film was exposed via a separate light for Red, Green and Blue, and the machine allowed the exposure time for each to be adjusted. Thus the phrase "colour timing". Adjusting all of them would raise/lower the overall exposure and adjusting them in relation to each other would adjust the WB. The controls from that operation live on as the "printer lights" controls in Resolve and other software, as they literally adjusted the lights of the printer. https://en.wikipedia.org/wiki/Color_grading#Color_timing

It's definitely very different shooting conditions... I laughed so hard years ago when I saw the review of the P4K by CVP (IIRC) and the guy took it out and took some random shots in public including some shots of his companion while riding on the train. His comment was that he had to work "incredibly fast"... HAHAHAHAHA. It was a person sitting motionless for several minutes!! I can go from zero to rolling in something like 5 seconds, and there are situations where I'll still miss a quarter of the shots I see.

@Brian Williams On the contrary.. Lots of interest, as expressed in this thread! Smartphone sensors have enormous potential if we could just get all the ridiculous over-processing out of the way to get to the image. I've developed a powergrade that matches my iPhone to my GX85 to suit the work I do, but RAW would be the far easier option, that's for sure!

Vintage lenses tend to be 10x the price for the "best" version compared to 2nd place, which is great if you don't happen to need the fastest/sharpest one. Often, if you can find a lens test that does compare the "best" one with the "losers" then you'll find there's not much of a difference between the best one and the next best one (and maybe the one after that..). A year or two ago I thought that Minolta glass was probably the next one that would become super popular, but I haven't kept up with it so I don't know if those have rocketed up yet or not. They're spectacular lenses though. Realistically, if you watch enough of those "OMG I found this incredible $30 vintage lens" then you realise that somewhere between half and three-quarters of all vintage lenses are actually great performers. Even more than that if you're interested in lenses that have a more degraded set of optics, or are mechanically only so-so.

Definitely not mid-season! I suggest: Learn the interface and how to get around Learn where the basic tools are and what they do Learn Colour Management and how to set everything up correctly At that point you'll have the knowledge to start getting into all the fun stuff like playing with other colour spaces and splitting channels and using blending modes etc. The journey in Resolve has three phases: You can't yet swim, so you wade out a bit, but can't go far You learn to swim and have fun swimming around a bit further from shore and grow your skills and confidence You start to explore what's happening underneath the surface and you dive down and suddenly realise that you've only been exploring the surface and that there is no bottom..... I've read posts on the colourist forum where someone makes a one sentence post and it took me 8+ hours over multiple days to work out what they said, how to do it myself, what it meant, and how I might use it.

Good points and they strike me as mirroring the workflows of Cullen Kelly and Walter Volpatto, the only two professional colourists who I've heard break down their workflow. The broad process for both of them seems to be: Import footage and get things setup (or, in our cases, edit the footage!) Setup the colour management correctly so everything is well behaved Setup the global 'look' If required, setup any specific looks for groups of shots Then start reviewing the shots individually (in passes) to even things out, troubleshoot, and then to really polish things up Of course, I tend to bounce between 1 and 2, because for me the colour and visual appeal of the shots matters in terms of which ones I choose. This is in complete contrast to how all the people that play 'colourist' on YT do it - they spend 10 minutes on one shot, whereas pros often only get 1 minute per shot, or less, so would be screwed if they didn't start broad and narrow down. I also particularly like Cullens approach, or at least the approach he's taken to his more professional LUT pack, which is that it's modular, so there are separate LUTs for contrast, saturations, hue rotations, split-toning, and other look adjustments, with there being several options in each of these categories. He recommends mixing and matching and applying them and adjusting the opacity of each to taste. For me, considering I tend to shoot and grade the same sort of material, I'm developing my own default node tree with everything all setup and ready to adjust as required. On many projects it's just a matter of applying the overall look and then just going into the Lightbox mode which shows all your shots at once and then just adjusting any that stick-out and then exporting it and doing a final watch-through. The BM grading panels are a bit of a clue as well, having Next Node and Previous Node buttons, but not buttons to create new nodes, on the smallest one at least, which implies to me a default node structure already applied to each shot. I've been meaning to go back and re-watch the Walter Volpatto masterclass now that I've actually gotten my head around colour management etc. Colour Management was the biggest breakthrough for me. Shooting on cameras that didn't have profiles in ACES/RCM meant that I had a really hard time adjusting levels or WB without the mids and highlights/shadows moving in different ways, but convert to DWG and grade and convert to 709/2.4 and then grade in DWG and all the controls magically do what you'd want them to do.

I thought that the standard focal lengths were normally designed to be spaced out relatively evenly in increments where you could move the camera closer/further to fine-tune. Obviously there are variations within that progression, and also variations in lens line-ups (like 28mm, 56/58mm, 90/100mm lenses etc) but that when you see matched sets of cine lenses, those were the main ones. There were quite a lot of other bundles with other spacings, which was interesting, although I did sort from highest price so started with the lens sets with every lens they made!!

Indeed I do! But seeing as I live in Australia, I tend to keep local hours 😉

There are definitely things that are best done in production, but for those projects where they can't be done in prod (or weren't) then it's nice to understand what can be done in post. I would expect that the field of colour psychology would be just as deep as any other, and worth studying regardless of if you're involved in pre, prod, or post.

Interesting. The only ones I've compared with any level of attention are the GH5 and GX85, which seemed so similar that after doing my initial comparisons I just wrote them off as being "the same", which in my context is probably better defined as "similar enough that any differences between cameras are small compared to the normal differences between different scenes, so there's no reason to develop a transformation to match them in my projects". The testing I did was to shoot a real-world scene and also a colour chart, then I matched the GH5 to the GX85 by adding contrast and saturation and looking at the results on the waveform monitor and vector scope, paying attention to the curve, the rotation and saturations of the patches on the chart. They were close enough that I assumed all Panasonic cameras would match, as that seems the most likely assumption you'd make. So I just moved onto matching the iPhone with the GX85, and (referring to the above definition) that was definitely NOT "the same" 🙂 I agree that the RAW->Profile is the only place they could match them, and I would have assumed that they would have done this. It's odd to hear they haven't! I mean, all you'd have to do is profile each camera, which would involve the same kind of setups that you'd need for testing them during the design phase, and then just install the transformation as a LUT. They're developing a bunch of them anyway, for each of the 709 colour profiles, so why they wouldn't do custom ones each time seems to be a false economy - especially considering how many of their users would own multiple models and shoot in multi-cam setups or at least multi-camera projects. Also, if you profile each camera, you don't have to re-do the colour profiles.

Cool - MF and the IBIS settings are the only downsides I can think of to going this direction, and it seems they're both in hand. I'll be curious to hear your impressions of the setup once you've put it through its paces - weddings are at the serious end of being able to work fast in challenging conditions so it's definitely a rigorous real-world test! Have you settled on which lens? I don't know anything about it specifically, but the CZ 40-80/3.5 should be a stellar performer so I would imagine that you'll be reporting good things...

I disagree. Panasonic is in the business of selling cameras to people, the vast majority of whom either couldn't colour grade their images to save their lives, or can't even spell LUT. However, they have managed to include a couple of features that cater to the more informed, the first of which is the inclusion of V-Log which caters to those who can spell LUT but aren't the most discerning control freaks on the planet. The second, and most refined option is to use the external RAW capability, which is the most neutral, and requires the most effort in post to get a good image. This three-tiered system allows Panasonic to be the most pleasing for people that can't troubleshoot the images, and the most neutral to those with the most capability, with the V-Log tier being the middle-one - it's not the most accurate and it's not the least accurate. It's like a political dog-whistle, only in camera marketing terms. You can cater to multiple markets without them suffering from the limitations of the other markets.

How is your muscle memory for manually focusing? As someone who has shot exclusively for several years using MF lenses handheld in fast-moving real-world situations and trying to get a bit of background blur, my advice is... get lots of practice beforehand! For what I do, I shoot a lot, I get what I get, I miss the shots I miss, and I have an infinite time in the edit to make it work. If I was shooting a wedding I'd be very nervous about using MF!

Panasonic will be getting exactly the results they want to be getting - they're just aiming at a different target. That target might be a different colour science, or it might be to cover up any image problems that creep in when shooting in odd situations like low-light, coloured lighting, poor CRI lighting, etc etc. These are just guesses of course. IIRC there was a test some years ago on the most accurate camera, and Sony won by a good margin, and this was in the middle of their time as being regarded as having terrible colour science. The name of the game is providing something that has the nicest and most flattering colour science across the widest range of images. In my analyses (of which I've done several) this typically involves making skin-tones more flattering, and giving subtle tweaks to the rest of the image by emphasising the warm/cool axis and de-emphasising the green/magenta axis. The standard Panasonic 709 colour reproduction @hyalinejim shows in the video is rather odd because it's not really following the normal patterns that I've seen, even in other Panasonic cameras.

I have the GH5 and GX85 but not the GH6, but I absolutely agree that although the GX85 is a great 4K 8-bit camera, the 1080p on the GH5 is in another league entirely. Prior to the GH5 I was using the XC10 and shooting C-Log, which in theory is a better colour science and a better codec, but the XC10 is only 8-bit and is way noisier, and you can absolutely tell the difference in post. Watching colour grading tutorials with ARRI footage they seem to be able to push and pull it anywhere in basically any direction and it just goes there flawlessly - the GH5 feels like that in post. I've (finally) worked out colour management in Resolve and graded a bunch of clips from different cameras and the files from the GH5 were the easiest to grade by a large margin. The fact the GH6 has more modes / frame-rates / codec options would just be better again than the GH5. I have heard it is noisier though.

There's a technique called "back button focus" which I have enabled on my GX85 ( @mercer told me how to set it up but google should also know how) and would side-step your issue. Basically the camera is in MF the whole time except when you hold down the button you configure for it. So the button is like having a hold-to-focus option. Choosing one you can hit with your thumb seems to work quite well ergonomically. I use it all the time because it means that you can focus on something and then use the shutter button to capture things without the camera having to refocus every time you go through the half-pressed position, and making the shutter button far more responsive too.

Well done on getting such a great match... and the video too. Your career as a camera influencer has begun!!

Great to hear! Did you used to have a GH5 and shoot 1080p with it? The impression I had from those who tested it online was that it was virtually flawless, so if the GH6 is even better then that would be a real achievement 🙂

To expand on the above, here's a bit of colour psychology in which Cullen demonstrates that cooler colours are perceived as further away and warmer colours as closer, so it's a way of adding depth to the image: In the context of building a colour grade by making a number of small adjustments, this individual adjustment may not even be perceivable on its own, but would be in the mix adding to the overall look. Obviously Cullens example is stronger for demonstration purposes (and his LUTs are designed to be used by adjusting the amount they're applied) but the principle would scale.

I've never been a fan of the teal/orange look, which involves a lot of hue shifting, although I've consistently been surprised how much of it a movie can have and you still get used to it while watching, but I've never liked it for my own work. There's a school of thought that you should never know that a colourist did anything with the images, and I've heard more than one colourist say that cooling the shadows just makes them muddy, so I'm not so sure it's an unpopular opinion. What is interesting to me is how much of it is occurring in virtually every camera LUT and every camera profile. It seems to be a different story when it's applied in subtle ways, being more like adding salt to a meal where below a certain threshold you can't taste it directly but just experience everything else heightened. But the teal/orange look isn't going away, although it may get less extreme (it's hard to imagine it getting more extreme at this point!), because it exists in nature and is part of how the world works. It's built into many classic film-stocks so has been with us for literally decades. It is probably hard-wired in our brains to some extent too, although I don't have proof of this. I feel my next stage in colour grading is learning how much of these things is the optimal amount where it adds to the image but doesn't call attention to itself.

This makes sense. The average camera reviewer knows next to nothing about colour grading, and these are the people that create the word on the street. Better to annoy the people that have more expertise, and besides, it might encourage them to buy further up your product line. Perhaps the most significant thing that Yedlin has said in all his essays is where he tries to convince people to look beyond the manufacturers: What is interesting about this sentiment, which seems pretty normal and par-for-the-course on these forums, is that the audience for his articles is the entire feature-film and TV industry and all the professionals in it. It seems that the same "colour comes from the camera" mentality is shared by the pros as well! I don't envy the job of the colour scientists who are trying to make transformations that will be used on billions of images and has to suit more situations than any one person could even imagine. Obviously these folks would have built large libraries of test images and would (I imagine) be able to apply the transformation to all those images and also to other datasets like the RGB cube and then analyse them for artefacts and discontinuities etc, but you're still grading billions of photos blindly. It is interesting that the warmer hues (red orange yellow and browns) are darker than the reference circles, and the cooler hues are lighter than the references, which I would interpret as a degree of making things flattering, but it's pretty limited. Yeah, looks come and go, like fashion trends, but what is old will be new again someday. As I've mentioned elsewhere, I've been binging Cullen Kelly colour grading videos and he's taught me a lot about the taste aspects of colour grading and also look design. I found some things really interesting: He spoke about wanting an accurate film emulation, and asked if people wanted the limited DR of the negative? what about the grain? gate weave? inconsistency in output when processing? etc. He said that when these film stocks were the latest in technology the people working with them were very aware and frustrated to some degree about their limitations, so now that we have digital why would we want to accurately recreate all those limitations, other than to make something look period-accurate? More likely is that we want to capture what is desirable about film without the limitations it had, and in this case the question moves from "how do I accurately emulate film?" to "what are the desirable attributes of film and how can I recreate those?" His approach to this seems to be focused around a few specific operations: Applying a film-like contrast curve Applying a subtractive colour model Applying a split-tone adjustment (typically to cool the shadows and warm the highlights, but keep black / white / mid-grey neutral, although there are other variants of course) Applying some hue modifiers (which I assume are hue rotations) and if a film-look is desired, then adding grain, halation and gate-weave He mentioned that he applies a split-tone to basically every project he grades, but also demonstrated that when applied subtly it is not visible as an orange-teal look, but simply adds colour contrast, which was very surprising as I thought this would always be noticeable Going back to Adobe and their conservative approach to colours, I was surprised to learn that the colour model in Resolve is HSL (which is used basically by all the standard grading tools) which seems to be one of the least useful models when you consider that it's not a good fit for how we perceive colour, how colour works in the physical world, or what we perceive as being desirable or flattering. The kicker is, though, that you can't change it! If you want to perform operations in more desirable colour spaces then you're left to do various complicated work-arounds (such as setting a node to HSV, disabling channels 1 and 3, then adjusting saturation with the Gain control) but this isn't optimal either, and more sophisticated solutions require coding custom scripts like DCTLs etc. I don't know what Baselight does, but Resolve is one of the top-tier colouring systems that has been used from the very beginning to adjust colour reproduction of the physical world in an artistic industry and yet it uses HSL as its colour model. I'm not even saying that it should be some other colour space, just that it should be configurable, but it's not, and all the other options essentially render your control surface and most of the tools irrelevant or inaccessible. How well do you think they work? Fuji has some expensive and technically sophisticated cameras, so they're not a budget option. I ask because I'm wondering if its not just the transformations, but perhaps the sensor itself. CCD vs CMOS seems to come up in these discussions once you dig deep enough.

There appears to be a broader context to all this. I think that people fall into one of two situations.. People that are happy with the images they can achieve This could be that: they can't / don't colour grade (maybe beyond adding a simple transform) but are happy with the images they get don't colour grade at all and just shoot in a 709 profile but are happy with it they can colour grade, potentially at a very high level, and their skill is able to match their expectations This appears to include people like our friend @markr041 who seems to enjoy the results of his many camera tests, and the vast majority of consumers, but it also includes quite a number of professional colourists who have high expectations but also high skill levels to be able to craft images. This is also likely to include the cohort of people who have high expectations but are able to meet them by using tools such as FilmConvert or Dehancer and have had enough practice that the heavy-lifting of these very sophisticated tools is able to get them over the line with their images. @BTM_Pix appears to be in this category, although he seems to be suggesting that the camera is adequate because any shortcomings are due to the lack of the other 499 people typically involved in creating nice images. They do say that great skin-tones begin with the makeup department so this definitely has merit. @hyalinejim might be here too, I'm not sure, but has certainly demonstrated significant skill in film emulations in other threads. This category of people rely on the colour science of the manufacturer to different degrees, but it might be quite significantly, and potentially spent many thousands of dollars buying that colour science. I include even many deeply experienced colourists in this equation, as typically, even though they have significant levels of skill at their disposal, they're also mostly spending time grading footage from the best cameras around like ARRI/RED/VENICE and would absolutely not be able to get such satisfying results from lesser cameras. People that are NOT happy with the images they can achieve This could be that: they can't / don't colour grade beyond adding a simple transform don't colour grade at all and just shoot in a 709 profile they can colour grade, potentially at a very high level, but their skill is NOT able to match their expectations (This list is almost identical to the previous one) These people rely on the colour science of the manufacturer and are unhappy at the manufacturers because the manufacturer doesn't deliver sufficient image quality to get the desired results. Where this is discussed on EOSHD is from people that are familiar with the colour that has been shown to be possible by cameras like the OG BMPCC/BMMCC, Digital Bolex, Canon 5D with ML hack, etc. I think that no-one sensible expects a 3K modern MILC to have the colour reproduction of an Alexa 65, but the fact that a S16 sensor from a decade ago could do a MUCH better job than todays cameras is extremely frustrating. This is where my comment about "more pixels instead of better pixels" comes from, and definitely represents my position, and explains why I have been focusing on colour grading - I have basically accepted that the manufacturers have abandoned us and am learning to do what I can on my own. Cameras seem to be in one of several categories: cameras that have a baked-in look that is quite "video" (e.g. iPhone, DSLRs, MILCs, etc) cameras that have no baked-in look (RAW) and require colour grading for all the colour work cameras that have a baked-in look that is very filmic (e.g. OG BMPCC, Digital Bolex) Unfortunately the last category is now extinct, despite there not being any real reason why this was necessary or desirable. There are thousands of people online who share this view - this thread has 3700 posts and is still going strong: https://www.pentaxforums.com/forums/6-pentax-dslr-discussion/409881-ccd-sensor-cult-i-mean-club.html What is interesting is that the vast majority of colourists are in this second category, despite having the highest skill levels of the bunch and even when grading ARRI/RED/VENICE footage. This is mostly to do with film emulation, combined with their incredible visual acuity, which has to be seen to be believed. It's widely acknowledged that Steve Yedlin has implemented the best film emulation seen in the wild in recent times, which was used on his films such as Knives Out / Glass Onion and especially Star Wars The Last Jedi which was shot on digital and film with the digital processed to match seamlessly with the film. This article gives an excellent overview of the many technical essays that Yedlin has posted to his website Yedlin.net: https://www.polygon.com/2020/2/6/21125680/film-vs-digital-debate-movies-cinematography The reason that I raise this is that Yedlins work is perplexing, because it shows that emulating film basically perfectly is possible, but also that it is not possible for most people, even most high-end colourists. In the Display Prep Demo FAQ Yedlin mentions that he had to create his own tools to do this manipulation: He doesn't say it specifically in a nice quotable format, but the impression I was left with was that he built these things for himself because they're beyond the capabilities of even Resolve / Baselight, which are deeper and more capable than almost anyone on the planet (let alone these forums) could even comprehend, but are not adequate. This is where we get into the difference between a colourist and a colour scientist. Accurate film emulation requires a colour scientist, not a colourist, and the two are quite distinct skillsets. Summary When it comes to getting the look you want, it's either a case of: camera colour science + production design + your grading skills > your expectations camera colour science + production design + your grading skills < your expectations The problem with the second one, especially for film-like images, is that we have seen it's possible for camera manufacturers to create starkly better images than they currently do, so this is enormously frustrating. Obviously production design matters, but if you're just looking at people outside in natural light, this aspect is far less significant than the camera and your colour grading skills. When it comes to film emulation specifically, the cruel joke is that: It is possible, because Steve Yedlin did it, but simultaneously it is not practically possible, because only a handful of colour scientists in the world are capable of doing what Steve Yedlin did to achieve those results Most of us could get better grades with the great examples of older cameras (OG BMPCC, Digital Bolex, etc) than we can with modern cameras, so in a world where the only people that can afford to get access to the colour scientists is the manufacturers, they have deliberately not done this, basically cutting us off from that potential

Reminds me of that Modern Family episode where the whole thing was from the perspectives of their various devices as they video chatted etc with each other all the time. I read that they actually tried shooting it like how it was portrayed but ended up using real Apple device cameras but operated by camera ops and the actors just resting their hands on the camera ops wrist.