kye

Also, a question... The Panasonic quotes the GH5 EVF as having "3,680k dots" but does this mean 3.6MPixels, or is it counting the RGB lights in a pixel separately (which would mean it's about 1.2MPixels)?

I don't know about "affordable" but this might be helpful?

Shooting 4K and downscaling in post is a pretty good way of getting high quality 1080p, and it may record in a higher bitrate too. I have done image quality comparisons of different resolutions given the same file size, and you're normally better off with the higher resolution, so even if the camera shoots the same bitrate for 1080 and 4K, you'd still be better with the 4K file.

Great question.. I assumed @User was just using Resolve to generate media for use in some other software, but maybe that's not true?

Cool! Let us know how you go when you've had a chance to test it all

True, but only for longer focal lengths. If you're looking for a lens in the 100mm+ equivalent range then 50mm f1.8 lenses are great and cheap, and longer and slightly slower lenses are (almost literally) a dime a dozen. If you're looking for lenses in the range of, say, 50-100mm equivalent then you can get things like 28mm or 35mm FF lenses, but they're either cheap and slow or expensive and fast. If you're looking for lenses around 35mm equivalent then there are lots of 18mm APSC lenses around but you have the same cheap and slow or expensive and fast problem. And if you're looking for lenses under 35mm equivalent then you're basically screwed with a 2X crop on adapted lenses as 8mm or 14mm lenses are more expensive, and if you want 8mm and non-fisheye then it's time to sell a kidney!

Nice looking image. I'm doing the same thing with my GH5 and adapting lenses - using the GH5 10-bit mode to get the colour depth (not quite like RAW, but better than 8-bit) and the lenses to render the scene in a non-clinical way. I've got a couple of Helios lenses, and I have both SB and non-SB adapters for it. One thing to note is that modern lenses can be used as "semi-vintage" lenses too, if you use them wide-open or completely stopped-down, as this will normally soften the image significantly and is one of the things that people like about some vintage lenses. You can also 'cheat' a bit with them and use a bokeh modifier on the front of the lens to change the shape of the bokeh from the normal shape to something a bit more interesting. I suspect that if you have a 3D rather than 2D bokeh modifier then you can get different shaped bokeh in different parts of the frame, the way vintage lenses do, but I'll have to test this. It depends on what you're interested in.

I agree, and have said so previously. It's already available (the Sony FS5, FS7ii, etc) but will be great when it trickles down the product lines into our hands. Being able to set shutter angle and aperture (which are creative controls) and then control exposure with ND and ISO (which are exposure controls and not creative controls) will be a huge step forward. Also, setting auto-ND and auto-ISO will allow those of us who shoot in faster run-n-gun situations to keep away from very short shutter speeds. This is the kind of feature that will be significant enough for people to change systems. It's not. There are many native m43 lenses available. However, I think that adapting lenses with a SB is a valid and popular choice for economic reasons: FF lenses are often cheaper than their native counterparts (when you remember to convert the aperture!), and APSC/FF offers fast zooms that aren't available natively either People often already own lenses, so there's a convenience factor There is also the question about how much value your investment in lenses will retain over time, as at the moment it seems like everyone is going FF and it remains to be seen if this is a fad or if m43 will die, or if it will survive but get left behind as a lesser format Of course, adapting a $100 nifty-fifty by buying a $650 speed booster is a false-economy so you'd have to have quite a few cheaper FF lenses to recover the cost of the speed booster. There is also the option to adapt vintage lenses to get a desirable aesthetic to potentially offset the "lack of soul" that some people perceive in todays nearly-perfect lenses. This is art, after all.

Let's be clear about our terminology. When we talk about panning, tilting, or rolling shots, we are talking about shots where the camera stays in the same place but rotates. Eg, a panning shot is where the camera starts by looking left, and then rotates to point to the right. These shots are best accomplished with a tripod where the camera will be held still and the fluid head will provide a smooth rotation. When we talk about dolly shots or crane shots, we are talking about the camera physically moving, and it may or may not be rotating at the same time. These shots have that great parallax effect where the foreground moves faster than the background and you can do reveals and create nice depth. These are created by sliders, camera cranes, and dollys. If you want to up your production value, then the typical setup is a tripod and a slider. The tripod gives you flexibility to position the camera and get the right angles, and then by attaching the slider to the top of the tripod you can get movement that will be steady without it bouncing around or whatever. If you mount the slider left-right then you get sideways sliding shots, if you mount it forwards-backwards you can get push-in or pull-out shots, and if you mount it so it's got some up-down travel then you can get some crane-style shots. Some sliders have a wheel to control the movement, some have a flywheel, and some are motorised, and these are all mechanisms to try and smooth the speed of travel. What this setup will not give you is stabilised rotation. If you mount a fluid-head between the slider and the camera then you can move the camera on the slider and also pan/tilt the camera at the same time, but this requires skill and a steady hand. Monopods can offer panning and tilting shots, and can also do push-ins or pull-outs if you have a fluid-head and a steady hand. Also worth mentioning is table-top devices that give you either a sliding action, or a combination of sliding and panning, so you can go around a product. These are covered in the video previously posted by @BTM_Pix which I've quoted below. Your next step is to be clear about what you want: is it to move the camera? is it to rotate the camera? if you want a combination of those moves, then which combinations do you want? Only then can you think about what options are available and what you should get. These things are typical of film-making in the sense that: they add production value, you get the quality level you pay for, and the more flexible the setup the bigger and heavier it is and the longer it takes to setup and pack-down.

Wow - cool image. People over-use shallow DoF and then everyone gets critical of anyone who uses shallow DoF, but when used in an artistically relevant way it is a valuable technique, as this image shows. Both the fog and the shallow DoF really contribute to the beauty and kind of suffocating feel this has.

Cool you found it. I have lots of those moments! Also, if you don't want to edit the clips but just want to convert all of them to Prores, then the Media Management tool under the File menu (when the Media page is selected) is a great tool. It can also export the things only on the timeline too, either the whole clips that appear on the timeline, or the same but also trimming the clips on the timeline and optionally adding extra frames to the start/end of each clip for flexibility in editing later on.

DaVinci Resolve has heaps of functionality for converting between different colour profiles / gammas and is free, but you'll have to work out what the names of the colour / gamma spaces that you used are, not just the model numbers of the cameras. Alternatively, there is a LUT calculator that might have the profiles you're interested in: https://cameramanben.github.io/LUTCalc/LUTCalc/index.html Of course, the best approach is to not use a LUT at all, and use some kind of proper transformation that doesn't do a destructive transformation. You don't mention what software you're using, and that might help.

And can we afford the disk space and memory card requirements of the All-I codec??

I think I'm moving in that direction too. Maybe I should change my preset configuration to High focus peaking and see how I like it

Good article Andrew. I would like to see you also do an article that is your current top cameras, regardless of release date. I'm sure you'd provide some interesting commentary and some inclusions from previous years that are still holding their own against the newest releases. It would be of real value to those who are in the market too.

I know, but with your history with ENG cameras, you'd recommend this lens as a lightweight run'n'gun option!!

Nice to hear you're having a good experience! In case you're not aware of it, the second level of caching to keep that performance even if you've got a million effects applied is to render the relevant parts of the timeline. To do this, go to Playback -> Render Cache -> set to User. Then on the timeline, right-click on the relevant clips and select Render Cache Colour Output and it will render those clips with the grading you've applied to that clip, like the Render Timeline in days of old. You can also set the Render Cache to Smart and it tries to work out which bits of the timeline to render for you, sometimes it gets it right and other times it doesn't, but it can be useful. There's also the Playback -> Proxy Mode settings, which I'm not that clear on, but I think they reduce the resolution of the preview window, requiring less work to play things in the preview window. Also of note is that Resolve has two playback modes, toggled by the Show All Video Frames option in the context menu on the viewer window (on the top-right corner of the viewer there is a button with three dots which opens the menu). If that option is enabled it will show all frames and if it can't keep up then the sound will cut in and out, and if that option is disabled and it can't keep up then it will play the timeline at normal speed with continuous audio but with a jerky video component. If Resolve can keep up then that option doesn't make any difference. Performance in Resolve is managed by many different methods and settings. This flowchart is old but might still be useful for some.

@mercer @HockeyFan12 Interesting question about if a 50mm is needed on top of a 35mm. I'd say that in the context of a controlled shoot, it's probably not. I recently switched to a new setup of GH5, 8mm, 17.5mm and 58mm, which combined with the ETC crop mode give equivalent focal lengths of 16mm/22.4mm, 35mm/49mm, and 116mm/162.4mm. In practice the difference between the 35mm and 49mm is a lot less than I thought it would be before I started using it. It's handy for my work, but I certainly wouldn't add an additional lens to my kit with the additional cost, weight, and extra work to colour correct it! I've found that the jump from 49mm to 116mm is a big jump and often you want something a little shorter. These two observations lead me to think that something around the 80mm mark would be the perfect next size up from 35mm. By the time you "zoom in" with your 35mm by just getting closer, and can "zoom out" by swapping to the ~80mm and moving further back, I'd say that you wouldn't find much gap between those two focal lengths. The gap between my 16mm and 35mm feels about right, and is a ~2x change. A ~2x change from 35mm is also about 80mm so that checks out as well. The idea of having one lens is pretty cool, and not changing lenses that much simplifies the shooting process quite a bit I've found. You'll also be able to train your eye to 'see' in one focal length and really get into that headspace which might suit your creative process.

In addition to the excellent comments from @KnightsFan above, proxies have cons for colour grading and VFX. The pros warn against doing colour correction and grading on proxies because they're not an exact colour match to the original footage. Also, if you're doing any tracking then you'll want to do that on the original footage so that you get the best movement accuracy possible. If you're tracking a grading window with a large soft edge then it might not be that important, but the harder the edge on a grading window or the stronger the adjustment the more chance it will be visible to the viewer. For VFX, tracking accuracy is an absolute must, as if your compositing doesn't track perfectly with the scene then it can be quite obvious - human perception is a lot better than you'd think. This is why for VFX work and green screening it's best to shoot RAW as it eliminates the pixel-level errors of compression. In a practical sense, and if you're not doing huge budget work or VFX stuff, you can use lower resolution proxies to edit and do rough colour work, switching to the source media for final grading and if you're tracking any windows. For my own projects, I will render out the final project and watch it through for any tweaks I want to do, then tweak and re-export. This works if you have time to do so, but it depends on your schedule and level of attention to detail that your budget covers

Yeah, if you were starting from scratch and building a system that didn't need to be portable then a laptop in general isn't a good choice from a value for money perspective. My points are more around the "what computer do I need?" questions, which people think of in terms of Can/Cannot instead of Can up to a point and Cannot after that point Depending on what you are doing. Here's a video showing Resolve getting smooth playback of 4K h264 footage on a 2013 laptop... and here's the Blackmagic Fusion 3D promotional video where the guy renders a 3D animated title sequence where the iMac (and potentially their external GPU which is sitting conspicuously on the desk) looks like it's getting barely 1fps with all the modules he's loaded up... I don't know about how seriously other people use their editing software but I always have parts of my project where it plays smoothly without proxies and parts where no consumer setup could play smoothly, so the wrong answer is always "more" and the solution is always to works within your systems limitations.

I use Resolve on my MBP laptop and I don't think it's a bad combination actually. But I stand by my original statement about there never being enough hardware performance and so you just need to learn to work around it. I understand that @Snowbro has a machine that can edit 10-bit h265, but what about after adding transitions, colour grading, titles and effects, and image processing? The OP has specified that they're shooting RAW and using a proxy, so it's not like there's a magical Yes/No barrier to performance. If you have to render proxies anyway, then just render proxies that are OK for your system. Having a MBP laptop isn't really a problem - I render 720p Prores proxies and it plays 60p like butter, even with some effects applied, and this is in Resolve It might be far more powerful than that, but I only render proxies at 720 because it's enough quality to edit with and takes less disk space. Then when I disable the proxies and colour grade the original footage it's fine too because if I want to play it in real-time I can set it to cache the timeline, but for normal colour grading tasks like colour you don't really need to do that. In fact, if it plays back at 10fps or whatever it's actually easier to see grading issues because the footage isn't playing as fast! The only time when hardware performance really matters is if you have to be able to play the original footage in real-time with the processing applied because you're doing it in front of the director and producer who are paying by the hour for you and your grading suite, which I doubt is the case with the OP who started this by asking about spinning hard-disks. Of course, Resolve can swap between the RAW and proxy files with a single button press, and you can edit, mix, add VFX, and grade your project, and then pop back to the RAW tab and adjust the de-bayering settings on a clip and see the results instantly, so maybe this is a Resolve thing and with the others their limitations force you to buy much more expensive hardware?

It might be worthwhile doing a side-by-side comparison of the noise and the film grain, and trying to emulate it with effects. I have no idea how PP works, but perhaps adding a semi-transparent blur to the grain (or some other simple effect) might get it close enough to do the trick when applied to real footage?

@mercer nice shots, and cool location!

I didn't see this thread the first time around, but I sympathise with @Thpriest about focussing on the GH5. I'm new to the GH5 world, but I was disappointed with the focus peaking modes as the Low mode wrongly highlights blurred contrasty things like edges in the BG, and the High mode doesn't highlight anything in lower contrast parts of the image (like someones face, which is exactly what you want it to be useful for!). I've also worked out that the focus peaking doesn't operate on the full-resolution image - only on the resolution of the preview (screen or viewfinder) so if there's fine detail there that is lost in the downscaling then it doesn't get detected. The system I've taken to using is a combination of a few things. I like the viewfinder over the screen, partly because it's more visible in bright conditions, and partly because your face helps to stabilise when you're shooting hand-held like I do. In terms of focussing, I set my aperture to be quite large so the DoF is shallower, and I have the digital zoom set to a button so I enable that to 4X and then set my focus like that. Then I disable that function, adjust my aperture to the desired setting, and then take the shot. In my case I actually tend to do things in a slightly different order sometimes if I have to shoot fast to capture a fleeting moment, like if one of my kids is about to do something cool, or there's a bird and it's about to take off, or whatever. That approach is to hit record, then adjust focus as best as I can, then open the aperture, fine-tune focus, then adjust the aperture again. Once you're recording you can't use the digital zoom, so you have to focus without it. I work with fully manual lenses (the Voigtlander 17.5mm 0.95 is my main lens) and have auto-SS and auto-ISO so when I'm adjusting aperture it's only a temporary dimming or brightening of the shot before the camera adjusts. I'm probably still adjusting things by the time that the moment happens (so much of candid photography is about seeing what is about to happen) but sometimes you manage to nail the shot and it's good. It sounds like a lot but it can be pretty quick if you practice a lot. This is an example from a recent trip - my kids don't pose for me at all but they do take a lot of selfies so I find that shooting them shooting themselves is both a good moment and also a representation of our trip, so catching the 2s shot is pretty difficult, but still achievable. In terms of using the screen when you shoot from the hip, I'd use the same technique as above, with the focus-peaking just a little less helpful.

I shoot hand-held with my GH5 and it does a great job but it's not perfect. It's important to realise that having steady hands is more important than the IBIS or OIS. There are many videos giving techniques for how to get steady hand-held shots - three points of contact, use the camera strap, control your breathing, if you're walking then learning how to do the ninja walk, etc etc etc. The better way to get a steady camera is to use some kind of rig. A shoulder-rig, a monopod, a slider or a tripod. I am perhaps the most ardent hand-held shooter on this forum because I shoot in situations where I have no control over what is going on, it situations where tripods are banned, professional shooting is banned, and I have to carry a camera all day and so I can't physically carry a gimbal as it's too heavy. All that said, if you're shooting product shots then I can't imagine how you couldn't just use a tripod or a slider. Maybe I'm missing something. If it's a matter of budget then there are DIY sliders and things that you can make literally for free, and give 100% results. It's also worth saying that gimbals don't give completely stabilised recording. They don't stop the camera from moving up/down/left/right/forwards/backwards, so if you've got a shot with any foreground/background separation then the best gimbal in the world will still have shaky camera movement visible. Just look at people walking with a gimbal and watch the camera bob up and down as they walk...