kye

Welcome back to the world of camera buying - nothing has changed... it's still a minefield of pros, cons, gotchas, cripple-hammers, etc! A couple of follow-up questions that I think would help recommendations: It sounds like AF is a major focus - do you need continuous AF? Face detect? Eye detect? How shallow a DoF will you be shooting? What sort of motion are you expecting the AF to be able to follow? No AF is perfect but some might be better for your needs than others. What lenses do you have and what focal lengths do you need / use? and how fast do they need to be? Re-using lenses could mean more budget for body and might help get a more premium performance setup. What low-light requirements do you have? How sensitive will you be to overheating? Long takes? Hot conditions? Cool-down times between shots? Budget for whole rig?

No-one who's seen my videos at 24p have said "Doesn't look like a TV soap opera the frame rate is too low". If that's true then why are you posting to a thread on the internet arguing that 60p is better? Hoping to get Deakins to finally upgrade to a proper framerate?

Congrats on getting the camera, from the reviews it looks like a winner if you don't mind the size. You might be interested in this thread:

Great images... High resolution but not too sharp. Looks like you've had some rain!

Your arguments are all technically correct, however they fail to understand one critical thing: how it actually appears to humans. There are exceptions of course, but a casual glance through the thread will tell you one thing overwhelmingly clearly - 60p looks awful. Considering that the entire purpose of cinema and TV is to be viewed by humans, this is game over.

To change the subject slightly (and with apologies to @John Matthews) after shooting with my GF3 and 15mm F8 pancake lens I've been thinking about tiny cameras further, and am wondering what the best option would be for the smallest and fastest 4K camera setup without a super-wide lens. To be specific, I want it to turn it on, aim it, and hit record and be capturing 4K / 100Mbps video in the lowest possible number of seconds. To this end, auto-everything is preferable, except auto-focus, which has to either be fixed focus or fast enough so I never have to wait for it. Essentially, I want an action camera, but I don't want the super-wide FOV. I've re-read the thread, and here are the options I think are on the table: Zcam E1 with 15mm F8 body cap lens GX850 with 15mm F8 body cap lens Sony RX0 (mk1 or 2) iPhone Things that don't make the cut: Cameras that are too big: GX85, Olympus E-P7, Osmo Pocket series Too wide FOV: action cameras Modifying an action camera with third-party lens Things I'm not sure about: LX10 (is the AF instant and reliable?) Are there other options I missed? Other thoughts?

Interesting example. Images look really good, handled mixed lighting really well, low light very well too, and some shots would be indistinguishable from a Sony flagship camera if the DoF was shallower. Pity it was uploaded only at 1080p and not 4K - the extra bitrate from YT would have shown the image much better.

I was wondering if the conversation would get to discussing this. I was curious some time ago and did some testing and some math. In testing I can see the difference between 24p and 30p easily, on both a 60p display or a display that is set to the native frame rate. The difference is obvious and the look of 30p is quite distasteful to me, regardless of the display frame rate / refresh rate. 24p on a 60p display does indeed introduce jitter in the timing of the frames (where the frames displayed are "nearest" and not synthesised from multiple frames in the source material). When you go to higher frame rates the jitter becomes less, with 120p being an even multiple of 24p, so the jitter of 24p will be eliminated or drastically reduced with higher display frame rates. In the math I did, I was surprised to see that capture frame rates are remarkably preserved even if put through different frame-rate timelines / displays etc. Assuming I didn't screw up the logic, here's what you see when watching 24p source material on 30p display. Timing is all over the place, but for whatever reason both 24p on a 24p display as well as the below are still preferable to 30p for me. What becomes interesting is when we shoot 30p, put it on a 24p timeline, and then display it on a 30p display: Apart from a doubled-up frame every so often (because there are only 24 frames per second to choose from), the 30p is completely resurrected! I have wondered if Netflix etc apps on smart TVs actually change the frame rate based on the source material or if they just run the TV at some fps and pick the nearest frame to display. I have been meaning to test my TV with my phone (recording the screen with 240fps slow motion and then reviewing the footage and counting the frames is pretty straight-forwards). TLDR; 24p is far superior to 30p/60p regardless of display refresh rate (for me anyway) When displays move to faster refresh rates the jitter from 24p sources will be reduced / eliminated Frame rate conversions can involve interesting time-aliasing effects where the time-resolution of some frame rates can pass through almost completely in-tact

In the real world (as always) things are more complicated. Getting things right in camera is normally considered good advice because it's assumed that the results will be higher quality than not getting it right in camera and then adjusting afterwards. It can also be good advice from the perspective that depending on the situation it can take significant time / effort to process in post, and there's a risk that the desired results can't be obtained, and by then a re-shoot might be very difficult. On the other hand, some situations will be made better by getting in wrong in camera, but in some way that provides an advantage. ETTR is getting it wrong in camera, and can improve the image once adjusted in post. There are other situations where this might be the case, depending on the situation. Your shot might have benefited from being exposed normally and then be pulled down in post, assuming nothing was clipped. Your shot might also have benefitted from being shot at a normal WB, and then having the red and green channels pulled down to make the image blue (ETTR but only of those channels). Good old fashioned "movie magic" involves trickery from time to time, sometimes by a huge margin (e.g. shooting day-for-night) and sometimes getting it wrong in camera can be advantageous to aid in the illusions.

The issue at hand is "getting things right in camera". Should I expose for the highlights, the shadows, or for the skintones? The answer depends on what "getting things right" actually means. The only "correct" way to use a camera is to understand the objectives of the project first, then use the camera to "get things right" in that context. Here's a recent interview with ARRI, taken in their technical testing facility in Canada. The place is literally a location for testing equipment. How do ARRI think about the tech? "It starts with the story".

Interesting review, not sponsored so contains criticisms and nitpicks, but in general it's a very positive review.

Today we had an early xmas lunch (to avoid the conflicting invitations on the 25th) and I shot it with the GX85 and 12-35mm F2.8. Shooting was really fast and not under ideal conditions (massive window which is often backlighting people) but it did a great job. Here are a few random stills, SOOC - zero editing. Standard colour profile, with -5 Contrast / Sharpening / NR and 0 Saturation. People often lower saturation in-camera but I push my grades to have lots of saturation so I prefer to have the camera do the 8-bit conversion and compression on stronger saturated colours so that I'm not boosting up a weak colour signal in-post. When I grade it I'll be softening it up, fine-tuning the colours and adding some colour grading secret-sauce but for a someone shooting a family gathering with an 8-bit camera while also being part of the festivities, the images really speak for themselves.

Sure. Perhaps the important part that I didn't mention is that I often miss moments because I was a second or three late to get the camera pointed in the right direction and rolling, so to say I don't have time to do a manual WB before hitting record would be an understatement because often I have a negative amount of time I can devote to such things! Doing what you can to get it right in-camera is definitely preferable, but in how I shoot, getting it into the camera at all is something that is by no means guaranteed. Of course, I am in a very tiny minority of all the people who shoot and know how to change the WB in post, but while my process doesn't apply to the majority of shooters, I still think there can be things learned by sharing 🙂

Absolutely. My situation is that I don't record long clips much, and in the edit I almost never have a shot more than 4s long. Also, I have a lot of time to edit compared to the footage I capture, so having time to fiddle with the footage isn't the problem, having the skill to do it well is!

There's a school of thought that says you just set a fixed WB of (IIRC) 5500K, which will make day seem on the bluer side golden hour / sunset warmer, blue hour cooler, and incandescent lights warmer. The idea is that this is sort of intuitively how we experience the world on some subconscious level, so the final result will seem right intuitively. Personally, I set auto-WB because it's more likely to get closest to correct than I will, because I don't shoot at a lightning pace rather than a snails pace, and gives me the least amount of adjustment required to do. In terms of getting it right in-camera, you should absolutely do it if you have time to. I don't, so adjust in post, but the better you can get it in-camera, the more potential your footage has in post.

Shoot 60p, edit on a 24p timeline, colour grade on a 24p timeline, do sound design on a 24p timeline, do VFX / titles on a 24p timeline, then when you're finished, copy all the objects on the timeline to a 60p timeline and hit export! Despite most education professionals being genuine and having the best of intentions, much of the learning happens despite the educational system rather than because of it. Follow your own path by default, submit to their nonsense where needed, save your energy for the work, and then when you graduate go off and have the best life 🙂

I've seen enough corporate shenanigans up close to know that (unfortunately) it's often the case that the political interests and motivations of the people in charge are completely divorced from the purpose of the business, and sometimes, reality itself. For example, I've seen decisions between two options which each had pros and cons get taken to a committee of senior leaders and emerge with the decision of a magical new third option that is clearly worse than either option originally proposed. I've seen this happen more than once across more than one organisation. Let's hope that some semblance of common sense prevails. You'd imagine that the demand for products in a marketplace would help drive things in the right direction just by sheer economics, but you'd be amazed at how isolated and deluded organisational decision-making can be, especially in committee.

Speaking of lenses, the size of the camera body is only part of the equation. The smallest lenses I have include... Olympus 15mm F8 Body Cap Tiny lens, tack-sharp, has all the advantages of a fixed-focus manual prime. Only downsides are the focusing lever can easily get bumped, and F8 is sloooooow Panasonic 14mm F2.5 Fast, sharp, small. A favourite for a reason! Laowa 7.5mm F2.0 Thin but potentially getting too long for convenience / pocketability. Fast and sharp, wide with zero distortion, but it's expensive. SLR Magic 8mm F4.0 Only one stop slower than the Laowa and not really much smaller, but hugely cheaper. Ergonomics are crap though - it's designed for drones and the focus mechanism isn't damped, but instead has a locking screw that you have to loosen/tighten every time you want to adjust it. But, it's so deep DoF that you could easily use it as a fixed-focus lens. Other interesting small lenses I don't have include: 7artisans Photoelectric 18mm f/6.3 UFO 7artisans Photoelectric 18mm f/6.3 Mark II Olympus M.Zuiko Digital 17mm F2.8 Pancake Panasonic Lumix G 20mm F1.7 Lumix G Vario 12-32mm F3.5-5.6 Olympus M.Zuiko Digital ED 14-42mm F3.5-5.6 EZ A batch of lenses from AstrHori have shown up on B&H, a manufacturer I've never heard of, which also look small.. AstrHori 27mm f/2.8 II AstrHori 10mm f/8 II AstrHori 14mm f/4.5 What are everyone's go-to lenses for these tiny cameras? Or are you giving the middle finger to the lens size and just being a rockstar?

I've looked at the E1 before as well, it really is absolutely tiny - almost action camera sized! What lenses do you use it with? I must admin that by the time you add a lens then the size advantage is potentially diluted somewhat. Does it have AF? I have a vague sense that there was some 'gotcha' that steered me away from it. The 4K really is great though. I definitely agree with @John Matthews about the advantage of 4K on these little cameras, even if the 4K isn't that great in absolute terms the mode tends to have more bitrate and when you put it on a 1080p timeline the compression artefacts on edges etc are often smoothed over in the downsampling. I also looked at the GM1, it's also very small... I think the tilt-screen of the GX85 won me over - my GF3 has a fixed screen and doing high and low angle shots with my GF3 was almost impossible with manual focus lenses. I've since moved away from MF to AF lenses, but even framing is a challenge and no screen can compete with reflections of the sky if you're holding it above your head and tilting it down so you're looking at the screen at almost 90 degrees!

Thanks 🙂 That's an interesting film. The repeated shots are a bit distracting but I forgive them, those who haven't published works don't have the authority to criticise those who have, and those who have published works will likely understand that sometimes compromises must be made, and that compromises might have been more necessary in 1948 shooting a road movie on film than now! It's much harder than people think to keep an audience engaged without dialogue.

DPreview said "The GF5 is a fairly subtle refresh of the GF3" and side-by-side they're almost identical, as well as the specs, so I don't think your revelation really changes things all that much 🙂

Actually, because it's full-auto, I did nothing to protect the highlights - it does what it wants and that's that! I think it exposes to protect some highlights, but I'm not entirely sure. I didn't really look at the SOOC clips with any detail, I just edited them and applied the film treatment. Incidentally, all that was done in Resolve with the built-in plugins - no Dehancer / FilmConvert / Filmbox etc - using the Film Damage / Film Grain OFX plugins and the Kodak 2383 LUT. All the camera shake was in-camera (I'm so talented!) but when I tried to add some shake to the end title it had some sort of bug and I couldn't work out how to get around it, which is why they're disappointingly stable. I should also work out how to add flicker, I don't know if there's a plugin for that or not. As nice as the beach and sunset look, I'm really bored of filming the same locations over and over again, but the GF3 / body cap lens combo and the different style of video made it quite fun and encouraged me to add camera movement and break out of my normal style, so that was really good. I essentially only film at home for practice so that when I go on trips I am familiar with the equipment and can shoot fast, but it's hard to maintain motivation for months or years on end between trips. I can see myself using that setup again, I feel like the small size helps the shooting experience around people, although one person did look at the camera in the footage (in a shot I cut). So it's not invisible.

This was tonights project. Just for fun...

Yes, when I was buying the GF3 I read some raving reviews of the GF1 - I don't remember the details now but there was some sort of button layout that was magic to use apparently, but it changed on the GF2 and subsequent models. I was torn about which to buy, but the GF1 had such a following that GF1s were more expensive to buy than the GF3, even though the kit I got included the GF3, 14-42 kit lens, 14mm f2.5, spare batteries, charger, memory card, screen protector, and a plethora of other bits and pieces. Thanks, I'll have a closer look over the holidays 🙂 Thanks 🙂 I guess I asked because there are very few people using cameras this small, and I have generally had the idea that smaller is better because it attracts less attention, but there's a point where smaller isn't better and I'm not really sure where that point is, so my thinking in this area isn't that clear. I definitely think of the GX85 as being superior to the GF3, but that is for so many reasons that I'm really not sure of which of them matter and which are just part of the package. I'd be stunned if anything being discussed here wouldn't blow the GF3 away in video terms, its a real weakness, which is a pity because I took the GF3 around the world back when I was taking photos and it captures stills that are about 50 levels higher than the video!

Ah, I think I found it. Changing from Light Mode to Dark Mode seems to fix it, even keeping Ambient Mode enabled (which seems to be the thing that gets some other people). Odd. Seems like a bug in YT.