kye

This is a topic I'll be visiting when the BM Pocket 4K comes out as if I buy the Pocket 4K I'll probably also be looking for a gimbal for it to make it my single camera setup. What kind of interest do we anticipate in the Ronin from DoPs / professional steadicam operators? I would imagine that Zhiyun wouldn't have a great name in industry, but my (outside) impression is that Ronin is a different story?

In my head I think of the Pocket 4K is a kind of "Official ML". What I mean by this is that it will provide results in the same league as ML (eg, a 4K version of the 3.5K RAW in 5D3) but will be fully supported with documentation etc, will be fully functional (eg, having full realtime monitoring while recording) and will be running the hardware well within its spec (unlike ML which is draining every last drop out of the hardware by pushing it to its limits, or past them in terms of overheating etc). I don't think it's better than ML, just different because it's for a different purpose. I have full respect for the wizards behind ML.

My favourite travel vlogger / film-maker just posted this gear video that might be useful for some people. Every time I go somewhere and shoot it I reflect on it afterwards and try to adapt my approach and equipment and I've found that over time my setup looks more and more like his setup. I wish I had been a bit more of a fanboi in this regard because I've bought a lot of gear that didn't end up working for me and eventually replaced it with things similar or identical to this setup and it's worked really well for me. I can second the Gorillapod, Rode VMP+, use of USB charging for as much stuff as possible, and use of clear bags for cables and whatnot. In watching the above he also mentions the 16-35 can turn into a 56mm with the A7SII crop mode switched on, which I'd forgotten. He's spoken in other videos about his lenses (he used to be a wedding film-maker so there's lots of gear videos on his other channel Wedding Film School) and IIRC he would shoot weddings with only the 16-35 and a 50mm prime because those lenses combined with the 1.6x crop mode gave him enough focal length options. He also shoots in 4K and delivers in 1080 so he can punch in digitally as well. It's interesting that he mentioned wanting to try his wireless lav mic for travel, which I think @IronFilm suggested at some point. It makes sense if you do lots of talking to camera stuff like Kraig does.

Ok, figured it out. Here's how to key out the edges and apply any Resolve adjustments to just the non-edges. This will allow NR without blurring edges (or any other adjustments you want to make). In the Colour panel create two nodes Add the EdgeDetect OFX plugin to one of them to make it the "key" node (Node 1 below) Connect the video signal like this: Make the adjustments you want to in Node 2 (eg, chroma NR) Adjust the settings in Node 1 to refine the mask that gets applied in Node 2 (I recommend adjusting Brightness to maximum to get a strong mask) The above setup excludes edges from being processed in the node, but it seems like you can also exclude other areas as well by adjusting the masks in Node 2 like qualifiers or power windows too, so you could (for instance) use the above to de-noise non-edge shadow areas by having the Qualifier in Node 2 exclude brighter areas of the frame. Enjoy! I wish I had worked this out before so I could do heavy NR processing without heavily blurring the video!!

Hang on a minute - I never agreed to that!! Show me evidence in writing!! **jumps up and down demanding evidence in true EOSHD style**

I'd be interested to hear your impressions of the native noise that comes out of the camera vs the additional noise settings that you have developed. To my eyes it looked nice (once you remove the chroma noise of course). Considering your style it might even be something you could use for creative effect, using full manual settings at the ISO that gives you the right noise level and then control exposure with a variable ND. @mercer I'm probably stating the obvious but Resolve seems to have pretty good NR, which if you put it in a separate node and use masks then you can get it pretty specific. On the video I shot on the GoPro in the club (with tonnes of ISO noise) I tried to run a NR node with a mask eliminating the edges (as NR is basically blur) but I couldn't find a way to generate an edge mask. The Sharpen Edges OFX plugin detects them, but the mask channel out doesn't seem to work for some reason. I haven't looked into this fully but it would be a great way to do heavier NR without blurring the main edges of objects. Have you compared it to the "Force sizing to highest quality" option in the render settings? (it's under Video -> Advanced Settings in the Resolve Render tab). I typically just turn this on when doing a final render but never tested if it made any difference. I'm reminded of early photoshop days when rescaling images often gave you options about what algorithm to use (nearest pixel, linear, bicubic, etc). My memory was that there were lots of different ways to scale images and they made a real difference in both quality and CPU time required.

Your logic makes sense to me. If it's not the LS300 mk2 then perhaps another camera, but definitely something related.

What's that saying? Never let the facts get in the way of a good story!!

Thanks! The ISO noise in the RAW files is a COMPLETELY different beast from the ISO noise when using the same ISOs in Canon stock video mode. This noise is filmic and soft and, perhaps most importantly, has a fine texture. That same noise once scaled and compressed in the Canon stock mode is just horrible. In the above video I did do NR and added some noise before export (on the recommendation from @kidzrevil) but even without that treatment the noise isn't a bad thing. One thing you might want to do though is to remove the chroma noise (colour noise) which doesn't have that nice feel.

@IronFilm I have a couple of beliefs (theories perhaps) that apply here. The first is that everything can be thought of like a signal-path.. obviously audio, where I think the concept comes from, but everything else as well. In the end, everything matters because everything has the potential to go horribly wrong. The second is that it's not enough to just not stuff something up in each stage of the signal-path, but every technical aspect of film-making has associated aesthetic attributes, which may or may not be aligned with what you're trying to do in the film. For example a horror film shot with high-key lighting and bright colours isn't as good as gloomy lighting and dull gritty colours. In the end we're trying to manipulate things (set, costumes, lighting, camera, sound capture, script, acting, colouring, editing, VFX, etc) so that they're all suited to and adding to the overall aesthetic of the film.

I have 700D with ML.. it's a decent setup! I believe the limitation of the 700D in non-crop-mode is 1728 because non-crop-mode takes every third pixel across the sensor and therefore runs out of pixels. It's not a limitation on the SD card. 1728 RAW upscaled to 1080 isn't bad at all, but you'll need to play with the sharpening to keep a bit of detail. I believe the low quality of the non-ML video from the 700D is the combination of 1728 upscaled to 1080 and then heavily compressed, making everything blurry. With ML RAW in non-crop-mode you'll still face the same challenge as you'll likely be compressing the final output for delivery. My card can't do 1728 continuously, so I've used 1600 for this video: This was also a low light test (ISOs at up to 3200 / 6400 for some shots) so not the perfect conditions. With 1728 you'll be able to do better in terms of sharpness and detail. I'm looking forward to the SD card hack being stabilised and re-released so I can get higher resolutions!

Just In Time For Christmas!!!! This space is really hotting up.. A7III, Fuji XH1, BMPCC4K, GH5, etc - so many great options in the compact-4K space! and more to come I'm sure with A7SIII probably only a release cycle away.

What mode/codec do you anticipate would be used on the higher-end productions using this camera? I'm curious because when I look at those bit-rates and combine them with the size of typical media everything above 1080 seems to need large capacity recording For example, a 64Gb card recording RAW 4:1 at 24fps only gives about 14 minutes per card. Or 18 minutes in UHD ProRes 422. These seem like quite low card capacities to me, but maybe you're rotating cards pretty quickly? When I'm out shooting I only take a couple of cards out for a day and I have pretty large shooting ratios, but I'm shooting at much lower bitrates (~30MB/s).

My understanding is that it's somewhere between "in a while" and "never". I certainly wouldn't be making any plans on it!

Agreed. I was referring to a situation where you are putting a noisy restaurant or nightclub ambience over the top. Many years ago when I was helping out student films I was asked to 'fix' dialog tracks with hiss in them and the only way I found to hide the hiss and make them usable was to use a mixture of background noise over the top and a fairly aggressive noise gate on the offending channel. My point in listing it first was that if self-noise on a piece of equipment is high then the rest of its attributes are basically irrelevant.. it could be very very low distortion, have huge bandwidth, large headroom, if it's digital it can be 24/192, etc etc etc, but if it puts unacceptable amounts of hiss into the dialog of a quiet scene then it's game over.

I shoot exclusively hand-held and rely on OIS to help me stabilise shots. The way I think about it is that OIS and IBIS 'absorb' a small amount of movement, with good OIS and IBIS absorbing a greater range of movement, but if you move more than they can handle then the movement shows in the footage. The way to improve this is to use some kind of camera stabiliser or mount. You can start at the small end by using a camera strap, adding a handle to your camera, all the way up through using something like a Gorillapod (a small bendy tripod used as a kind of handle) to full rigs like shoulder-mounts or steadicams. I mention the Gorillapod as it's what I use and I find it to be quite effective up to about 100mm or so in combination with OIS. Happy to give you more detail about how I use it if you're interested.

Maybe - it's a matter of what you're doing. I don't want to pay for, carry, or worry about a fast Zeiss. I want to get the shot, which means that I don't have time to change lenses. A fast Zeiss would likely be a prime, which actually means a set of big, heavy, and expensive lenses. They'd require a pelican case to cart them around too. A fast lens has shallow depth-of-field (unless you use it stopped-down all the time, in which case just buy a slower lens) which means missing focus all the time. I don't want to miss focus all the time - I want to get the shot - having the background blurry in some shots isn't worth the risk of having the subject blurry in other shots. The moral of the story is that everyone has different requirements. It can be quite difficult to hide a noisy audio track though, depending on the situation. If you're recording audio for a scene set in a nightclub or busy restaurant where there will be lots of ambient sound in the scene then you can cover it up, but if you're recording audio for a scene that takes place somewhere quiet then having hiss can ruin the content, or force you to use an ambience track that will compromise the aesthetic of the scene. For me personally I don't mind as my videos are for family and friends and are normally set to music, but if you only had budget for one setup then I'd make sure it didn't have too much hiss. We've covered this before. In your productions there might always be room for a dedicated audio person with a boom mic, but for many of us that isn't the case. Scroll up and read the parts where we're discussing the size of an on-camera microphone attracting too much attention and potentially ruining shots or a whole shoot. Don't confuse your situation for how others are making their films. When we're talking about equipment, the ability for something to record safety tracks is a consideration. Everything else you listed above is either skill (which cannot be purchased) or can be applied digitally in post in virtually any editing software (and therefore a separate topic). What aspect of sound quality do you think it lacks? I heard differences in the sound, but nothing that made me think I needed to spend hundreds of dollars. If I was recording something where the stereo audio was absolutely critical then I'd spend the money, but for most productions (mine included) stereo ambience is something added to the film for an almost subliminal effect. I know a lot about audio so don't dumb-down your answer for me. Feel free to talk about THD+N, IMD, imaging, phase response, macro-dynamic performance, micro-dynamic performance, stereo imaging / width / depth / focus / etc. For any film that is going to have big expensive audio treatments in post they can probably afford to get permission to shoot in locations, etc. Guerilla film-making is more about getting something done with a minimum of budget and resources. Depending on the type of film being made the results might be compromised in quality, however this is part of this type of film-making. If you could get perfection by shooting without getting noticed then why would anyone spend $300M on a film? The words Guerilla and Gorilla have the same pronunciation, so it's similar to saying "hunting gorillas". It's a joke based on this similarity! One thing I see on this forum a lot is that people assume that everyone makes the same types of films they do, or lives in the same type of environment, or has the same quality requirements as they do. Obviously this is false, there are as many different types of film-makers as there are types of films, or types of people for that matter. That is why I put my comments into this thread instead of the one where @IronFilm and others are providing excellent advice. This thread is specifically about films that are being shot on location without permission, which at least implies that they're small productions with limited funding, and I included my comments as part of a discussion about on-camera microphones, which indicates that the production is very small indeed, and therefore willing to compromise quality to a certain degree. If this isn't obvious to someone reading then I apologise, but this is the internet, and everyone should always question and verify everything they read or see on here.

In addition to what @John Brawley said, the current write speed performance of SD and CFast cards might be why they put the USB-C on there. Apart from making it super convenient by turning all the small editing SSDs we've got lying around into storage we don't have to re-buy, they might be the only things that can write 4K60 RAW until some of the tech catches up. You'll find that it's hugely easier to design chips that can transfer data at a given speed than those that can read/write that data. It took years for HDD speeds to out-grow the USB v1 and v2 spec, and apparently USB 3.2 is rated at 2500MBps - almost enough to do 4K RAW at 300fps! I wonder if you can buy small USB 3 SSD all-in-one RAID drives? They'd have crazy performance.

and the number one thing.... we definitely can't tell if it has MOJO!

There is definitely some strange colour splotchiness going on, but my question is what the signal path of the image has been to get it from the camera to my monitor. I agree with @Cinegain that it could be from some wayward processing. The ISO performance on that image is terrible!!! And it's a daylight landscape too!

That's a great idea! Velcro is potentially underrated. I do a lot of hifi stuff and vibrations are a source of distortion so isolating components is a thing. During my research I found that a company that makes scanning electron microscopes uses something to isolate from vibrations that is basically the same principle as velcro (which holds the weight but has a bit of give and therefore doesn't transfer all the vibrations).

That's very impressive! Well done!!

If this camera doesn't make sense then this camera probably isn't for you. I don't understand why people think they're the only people in the world to design cameras for. In fact, why would you even think that it was designed for someone who spoke English? The largest Chinese social network has more than half as many active users as the entire population of the US.

Not everyone is as lucky as you... The US: https://petapixel.com/2016/02/25/ive-stopped-20-times-police-camera-tripod/ The UK: https://petapixel.com/2016/07/16/video-photographer-tests-rights-streets-london/ https://petapixel.com/2011/07/20/six-photographers-test-their-right-to-shoot-in-london/ Australia: https://petapixel.com/2015/11/30/top-australian-photographer-fighting-for-rights-after-near-arrest-on-public-land/ Terrorism video warning about photography in public: https://petapixel.com/2012/08/09/terrorism-prevention-video-asks-public-to-report-photographers-to-police/ And if those are too tame, then this one is just great.. https://petapixel.com/2015/04/21/woman-pointing-smartphone-camera-at-police-has-it-snatched-and-smashed/ That was just a few quick searches of one photography website. The moral of the story is don't attract attention.

Just listened to it.. I'm assuming it was the 12fps clip? Audio sounded good, but there was nothing in front of the camera making noise so not way to judge isolation / directionality Hmm.. unless they're recording 4 channels of audio it is possibly to control pickup pattern. Promising!