Marc Wielage

Whenever a client asks me this, I always laugh and say, "this is like the 'how much does it cost to build a boat?' conversation." I've done entire features in one day, and I've done features in four months. (Dances with Wolves was 4 months, and each Star Wars film I did took two months, and those were often 12-hour days, probably 250 hours each.) I think Jamie's estimate of a day per 20-minute reel is reasonable. I generally divide it into shots: if we have 400 shots in a one-reel timeline, and we have to get it done in 8 hours, then we have about 1.5 minutes per shot (maximum) to get it done in about 8 hours. Any time we have to stop and do intricate keys, layers, window tracking, beauty fixes, drastic time-of-day changes, or relighting will throw off the schedule. In episodic TV, you're lucky to get 2 x 10-hour days, and I'd say 2 x 8-hour days is more typical. There are always exceptions: I'm told that Game of Thrones takes 5 days, but that's a very VFX-heavy show with a lot of very challenging work. If I can set my own pace, I think 1 day per reel and then another day for a client-supervised pass with fixes is reasonable. If this is a 6-reel film, then that would be 12 days, plus a day at the beginning for the conform and another day at the end for final deliverables. But I also try to build in provisions for overages, changes, redos, fixes, and so on. Legitimate QC fixes I'll do for free provided it's a shot here or there (maybe 6 total).

I'm a fan of using no LUT at all and just doing everything manually. The advantage is that this way, I have complete control of the image pipeline and know how and where everything is going to clip or crush. Once I have a given look (like a Powergrade) for a specific show or a specific scene, I'll use that throughout the scene and then make changes where necessary for style and to make it match. I don't have a problem with specific color science for Red, like RedWideRGB for color and Log3G10 for gamma (or RedLogFilm for gamma), but after that it's all me. I also have no problem with creating LUTs so that the VFX department can use those while creating composites with log footage, so they know the material will look correct by the time it round-trips back to me. When this is done right, the original correction drops right on and it all looks fantastic. But as a look... I think you're better off just experimenting and creating your own Powergrades (or stored looks in other systems) and then try those with new projects. If they don't work, throw them out and come up with something new. To me, that's what a colorist does. I don't have a problem at all with LUTs thrown on to material on set or as a temporary look for editing, which makes total sense. I also use the temp version as a reference and will routinely match to it with the controls I have, and then use that look as a starting point. The Reference is good to drag along during the session so that you can say, "here's the dailies look -- now here's where we are in the final." They can be identical or as different as you want them to be.

Naw, Steve had a much larger charisma. I am very close to Steve Jobs' age, though, and I tend to wear only black T-shirts and blue jeans (no reflections in the monitor or bounce to DI screens).

I'm not convinced the film LUTs really do anything than provided one person's idea of what the emulsion was. Bear in mind that what the stock looks like heavily depends on how it's developed and how it's scanned. I've seen specific pieces of film (even from the same project) look as much as 25% different because of developing and scanning problems. Even different batches of the same negative stock can look different. This is a moving target.

The only way the Mini Panel is usable to me is with macro keypads on the side. I'm currently using an XKeys24 on the right and a G13 Logitech on the left, and that's been working so far. I have a longer review of the pros and cons for free over on MixingLight.com: https://mixinglight.com/color-tutorial/color-correction-gear-head-thoughts-on-the-resolve-mini-control-panel/

You might ask Juan Salvo at The Colour Space/NY: http://thecolourspace.com/

Log as shot on what camera? Log is different on just about every camera, because each manufacturer kind of has their own idea of what that is. I find it useful to try out the built-in LUTs within programs like Baselight and Resolve, then find a way to reverse-engineer what they do only using curves and primaries. That way, I find you can have more control than just with a LUT. The stock Alexa LogC -> Rec709 can be pretty harsh, but Arri has a free service on their website so that if you wanted to, you could make custom adjustments to the toe or shoulder to get different results. Another thing that's useful is to have the DP shoot standardized charts (like the DSC Chroma DuMonde), and see how those react with your own software. Once you can get a predictable result from a chart, the rest is just customizing the look for a DP's particular technique and style. A LUT per se isn't necessary, but I have no problem using a technical LUT for something like dailies, where quick turnaround is necessary. Custom LUTs are also useful in some situations, like turning over a look to the VFX team.

Highlight control is very valuable in situations like this, and if you use it as part of a tracked window (and/or a key), you can pin point problems like this fairly surgically.

Often, I think you have to track different parts of the face, treating (say) the forehead separately from the area under the eyes, and the neck separately from the face. I find SNR in Resolve can be helpful (with a qualified key), and sometimes negative midtone detail. It depends on the nature of the original material.

I'm using an 8-bay G-Speed Studio XL in RAID5, and it's doing almost 800MBps with Thunderbolt 2, which is more than fast enough for 4K. It's rare I get any performance issues from the drive. The Trashcan has its own problems.

Short answer: No, there are no inexpensive HDR grading monitors. (Not yet.) As far as I know, Netflix and Amazon are the only companies demanding HDR delivery, and I believe they're asking for Dolby Vision masters. It's hard for a consumer set to even hit Rec709 specs, and HDR is literally 10 times harder. You can make a good case that nobody is going to see precisely the same HDR pictures on different consumer displays, but this will get better over time. Here's a link to some miscellaneous HDR tech documents if you'd like to read them: https://spaces.hightail.com/space/nEaXy

If they're shooting on a digital camera, yes, I would rather they shoot in color so I could manipulate the RGB channels in monochrome mode. I did several B&W music videos from film in the 1980s, and in one major case, they wish they had shot in color because they wanted to do a color "effect" at one point. But back then, we didn't have the power of the RGB mixer.

Anybody who wins an award and has an audience has the right to express an opinion -- whether or not you agree with it. If the person has a public forum, more power to them. If you don't like it or don't agree with it, don't watch.

The glass is half-empty, but it's also half-full. If you pick up a copy of the Resolve 8 user manual (which is out there on the web), you'll go into shock at how simplistic and awful the user interface was back then. Be aware that the software was also about $250,000+ and required another $100K in hardware and the control surface. And it was about 1/10th the speed of what we have today. I think a lot of user complaints boil down to, "I'm uncomfortable that this thing is over there, and that thing is over here." At some point, you have to get with the program and stop wailing at the four winds and just buckle down and learn to use what you have. A lot hinges on changing your working philosophy to taking advantage of how Resolve works, rather than complaining how it doesn't do something in a certain way done in another program. It helps to have a control surface, to memorize (and customize) keyboard shortcuts, to use a 3rd-party macro panel (like the XKeys), and to know the manual pretty well. Do that, and you can be very productive and work very, very fast. I've used 12 different color correction programs over the last 30 years (Topsy, Amigo, Dubner, the Image system, Arcas, daVinci Renaissance, daVinci 888, daVinci 2K, Pandora Pogle Platinum, Lustre, Baselight, and Resolve). All of them have pros and cons. All of them had flaws. Each of them could do fast, billable, reliable work. Coming from my perspective, Resolve is far more practical and productive than any of them, save possibly for Baselight (which is also an excellent system with some brilliant features). Given the cost difference between Resolve and Baselight, I'll gladly take Resolve's limitations for 1/10th the cost. As for editing, there are highly-personal choices in terms of user interface and features. I can think of a hundred pros and cons with Avid vs. Premiere vs. FCP7 vs. FCPX vs. Resolve (and also dark horses like Lightworks). I think given enough macros and a macro keyboard, you could do damned nearly anything with Resolve, assuming very fast hardware and fast drives. I wouldn't necessarily want to edit a 20-camera concert with Resolve, nor would I want to edit a feature-length documentary with 1000 hours of source material. But for certain things, I think it's fine. It's certainly good for finishing and conforming, and I have done some light duty editing projects that work well. Read Paul Saccone's recent book on editing on Resolve: https://www.amazon.com/Definitive-Editing-DaVinci-Blackmagic-Learning/dp/0996152830/ref=sr_1_1?s=books&ie=UTF8&qid=1483522353&sr=1-1 Alexis Van Hurkman also has some good tutorials just on editing on Resolve, and I think he has some good things to say (plus warnings on certain traps and also revealing some tricks along the way): http://www.rippletraining.com/products/davinci-resolve/editing-effects-in-davinci-resolve-12-5/

I'm not convinced that YouTube can handle normal HD, let alone HDR. Unless and until we have high-bandwidth, low-compression 10-bit images, this is all bogus. And it will happen someday.

Yes, the RGB mixer can do some very interesting things. I have had situations where the camera clipped one color channel but the others were OK, so I was able to "steal" some detail from the two other channels and get the channel a little more detail. This is a huge help for faces when those are clipped red. I have worked on a few features where camera clipping made us nuts, and this was in the days prior to Resolve (very early 2000s). I would love the chance to redo some of those shots, because I think now we could salvage them in a way that wasn't possible 10-12 years ago. It's also dynamite for B&W correction for projects shot in color, because it allows you vary the grayscale response so dramatically. It's a huge, huge help in cases like that.

I hate the curve because it affects the "flatness" of horizontal lines in the interface. I have done exactly two features in a 21x9 display and my neck was broken at the end of the day because I was getting whiplash having to turn off to the right to see the keyframe display (which I use heavily in Resolve). I don't think it's worth it -- I think a nice big 27" 2550x1440 display is huge and shows me everything I need to see, and then a smaller 19" display off to the right for scopes. I may go wild and get a 16x9 display for scopes at some point, but at the moment, I'm using a crap 4x3 display because that's what we used for about 20 years (actually, 17 years in HD). The third display for me is a 40" plasma, probably to be updated with a 4K OLED or something later on in the year when things settle down. I honestly think this works fine. If I was in a big room, I'd go for two 27" displays, split the GUI, and then drop the scopes down and in front of the console, with a projector head on in front.

I wouldn't try to push it back into log. I would start with what the filmmakers shot with and take it from there, and make sure they understand they shot themselves in the foot by shooting in Rec709 in the first place. It limits the available choices and they just have to live with that. Be aware that hundreds, maybe thousands, of some of the most memorable videotape videos of all time were shot in Rec601 (standard def) or Rec709 (HD). It's not rocket science, it just takes time and effort to make it work. I would also tell the client to expect less. Larry David's philosophy is good: "Curb Your Enthusiasm."

Walter Volpatto from Fotokem has a good method that I think makes sense: when you start a new project, create a new folder just for the Gallery stills and scratch disk, with the name of that project. Later on, when you move on to a new project, you can choose to delete that folder or archive it or do whatever you need to do.

It would be very interesting to compare the total price of equivalent systems from Baselight, daVinci Resolve, Lustre, Mistika, Nucoda, and Pablo Rio, and see how they compare. There was a time when they were vastly different. I honestly think today, an experienced colorist can achieve virtually-identical results, but there's always certain operations that are faster or simpler in one system than another. Resolve 12 is not a perfect system, but it does what I need it to do. Compare it to what was available even 6-7 years ago, it's lightyears better than anything. I think the competition between all the various systems ultimately benefits the users, but Blackmagic has been a formidable competitor and changed a lot of aspects of the business.

I won't comment on the HDR question, but I will say that if your profession depends on your eyesight, get a good ophthalmologist and get your eyes checked at least once a year or two. I also find using preservative-free eyedrops helps avoid me drying out my eyes, and I do that about 2-3 times a day. I just had a 45-minute exam a couple of weeks ago, and I asked the doctor how my eyes were, and he said, "just about perfect... but you need to blink more." Which I thought was funny, because basically, I don't blink much when I work because I'm always looking for that glitch I missed or that shot element that's way in the background.