KnightsFan

First of all, a lot of the video/editing jobs aren't art. Analyzing a billion ads and creating something similar for a new product is EXACTLY what machine learning does best. And it's not like it's just a black box--an AI can spit out a dozen, a hundred, or a thousand samples, let a human pick what they like best, and refine, and with each iteration, the machine creates a slightly better algorithm. Instead of hiring a motion graphics artist, a business owner who wants a commercial can just sit down with an AI and pick which ads they like out of a never-ending stream. Second, I disagree entirely. How do human artists work? They build a knowledge of art history, change a few things, and build off of feedback. That is exactly what machine learning does. Instead of C-3PO wandering about shooting a wedding, picture this: A robot scouts the venue ahead of time and sets up a few dozen small cameras to film the wedding from all angles, and then uses those cameras to reconstruct the entire ceremony in 3D. It then picks the best angles based on the knowledge of every single wedding video ever shot, taking into account the satisfaction ratings of the couples (using videos of the couples' faces when they see their video). With each video, it experiments slightly by changing a few things. It composes music for the wedding based on knowing what songs the couple plays, and knowledge of all music ever written. It does all of this by the next morning. No one sees the robots at any stage--completely discrete. With each wedding it shoots, this system improves slightly. And since it's a machine, it can shoot virtually unlimited weddings every day, thus quickly becoming the best wedding videographer on the planet. Obviously this isn't going to happen tomorrow, but there is no way to stop it from becoming a reality in the near future.

@kye Oh, okay, I must have misunderstood what you were getting at when you said: ...since the F6's 32 bit mode should be useful in pretty much any scenario that dual channel is--which, as you say, is "all the time." But yeah, the F6's dynamic range doesn't increase the microphone's dynamic range. You still can't get leaves rustling and a jet engine 50m away in the same file with the same mic, unfortunately.

@kye Have you ever used dual channel recording? How is what you are explaining different from dual channel recording and using the peaks from the lower track to replace the clipped portions of the higher track? Because I do it all the time and it works, especially when mixed in (surprise) 32 bit space, because then you can match the relative levels of the two tracks without distortion. I don't know all the maths, but I know from experience that dual channel recording is a life saver at times. I assume the F6 basically does the same thing, combining two input gains based on peaks, but automatically and internally, saving time and file space (1x 32 bit file is smaller than 2x 24 bit files).

@SR It's absolutely a big deal, but primarily for non-pros. I believe what @kye is saying is that Zoom didn't do anything particularly difficult--it's not like they completely redesigned how the circuitry works. It's similar to the dual channel recording feature many recorders have had for years, except that it merges the two files automatically into a 32 bit file, instead of giving you two 24 bit files that you can manually splice if you so desire. But it's absolutely a useful feature, especially for one man bands who don't have enough eyes to watch the camera and the audio meters at the same time, or for ultra low budget projects (like mine) who employ non-pros without much experience. The dynamic range of the audio file should increase dramatically. 32 bit 48kHz is exactly twice the file size of 16 bit 48kHz, not including the negligibly small amount for metadata.

It has two ADCs. If you overpower the higher gain one, it still has the lower gain one. If you overpower the lower gain one, then your microphone is already distorting (according to Zoom) and no amount of gain reduction will fix it. It doesn't look like you have any control over either of the gain settings on the ADCs.

The difference with photography is that you are never in a situation where your lens clips yoyr highlights, whereas a microphone has a dynamic range. If the 32 bit recording has more dynamic range than the microphone, then yeah, you would never need to adjust gain. Basically all they are saying is the digital recording format is no longer the limiting factor in dynanic range, it's the microphone itself.

I'm glad someone finally went to 32 bit audio recording. A lot of software processes audio in 32 bit for essentially infinite dynamic range, it was only a matter of time before people started recording in that format. It would certainly help on a lot of sets I'm on, where we have students/less experienced people running sound who are more likely to clip audio. Or any time I'm the solo audio guy, I can concentrate 100% on the boom without worrying about gain. Currently, I use dual track recording, but then I've got to patch things together in post manually. However, without having used it obviously, I think the F6 looks a lot less ergonomic than the F4/8. One thing I like about the F4 is having an immediate button to turn tracks on or off, and to solo a track. It looks to me a though you probably have to menu dive to do those actions on the F6, which would really suck. I also like the headphone knob on front on the F4.

6k60, 4k100 "at least" and "more than 200" in HD. Mentions dual native ISO sensor as well with 14+ stops of dynamic range. There will be MFT, EF, and PL mounts. Nice!

Ah yes! I totally forgot about that, but now I remember reading it before. Hmm, I actually might really look into that URayTech one now that I have some time. Ideally I'd like to integrate the video stream into a custom app, and I have a feeling the AccSoon has some closed source pieces since it seems to do intelligent switching and stuff, whereas the URayTech one actually mentions having an open SDK.

I haven't seen anyone talk about the Acsoon CineEye HDMI transmitter yet. It looks like a good, very cheap way to get wireless video. For $219, you get a single device that wirelessly sends HD video directly to your phone. Currently, the only way I know of to do this on the same budget is with consumer wireless HDMI transmitters, which means two devices that need to be rigged and supplied power, and an external monitor at the receiving end (also requiring power!). With this, it's a single device with a builtin battery that can send video to a director, producer, or any 4 people in the area who have phones or tablets. I really think that sending video over Wifi should be a standard feature in all decent cameras these days, although Z Cam is the only one doing it.

After reading up a bit, the whole ZaxNet concept is brilliant. Really a phenomenal feature. Adding an npf battery plate for the f6 is a nice addition. It also says it can be powered by USB c, so I assume that you can use it as an audio interface with a single cable for power and data. Speaking of which, my biggest annoyance with the F4 is that it doesn't boot back into interface mode. It lives on my desk as an audio intetface between shoots, and it's an extra 10 button pushes every time I turn it on.

I don't know much about the more expensive audio toolset and most of those better features would be wasted on the small sets I've been on anyway, to be honest. I'll have to look more into ZaxNet and thr way the Nova has builtin slots for wireless receivers. The zoom F6 is an odd looking thing for sure. Almost cube shaped. Any idea why they would depart from the F4/F8 shape?

It's not theoretical. Check out 4k footage from the NX1 vs. the 1080p footage from the same camera. Downscale the 4k footage to 1080, and compare them at that HD resolution. It's a night and day difference, from the same sensor, processor, and codec. You can do a similar comparison on most other cameras as well. Whether it's "useful in practice" depends on whether you're talking useful to the general public's enjoyment of a compressed YouTube file, or useful to a cinemtographer's critical eye looking at a high bitrate master. And just to be clear, downscaling in post doesn't gain quality. It loses quality. The original 8k 4:2:0 file has better information fidelity and more spatial resolution than a 4k 4:4:4 file downscaled from the 8k original. The only claim is that downscaling to 4k 4:4:4 will retain more information than downscaling to 4k 4:2:0.

@majoraxis 8k 4:2:0 has quarter resolution in the chroma channel, so it is 4k 4:4:4 with oversampled luma. That oversampling provides pseudo-12 bit, but only for luma channel. So it is 4k 4:4:4 10 bit with a nice oversampling on luma. Similarly, you can get 4k 4:4:4 12 bit, but the chroma channel will only have 10 bits worth of information at the maximum, while luma would retain a bit more information.

It's a diminishing returns thing for sure, but the e2 series cameras have an ergonomic advantage, more frame rate options, and don't require a 5" monitor bolted to the rig to record raw. Up until the e2 series, you needed to pay tens of thousands for a kinefinity or red for those diminished returns, and now its down to $5k. Many ultra low budget people will still opt for a z6, s1, a73, gh5, etc., but if you have a little more money to spend and were already considering more expensive, dedicated cinema cameras, the z cam offerings are amazing value.

You know, I almost posted in that topic about Race to the Bottom about how the economic value of our work as filmmakers is falling because of advanced software, and will be completely obsoleted by machine learning within a few decades. The truth is that once an algorithm is developed--whether by traditional coding, or machine learning--it costs $0 to duplicate it globally. (Patents and lawsuits make it cost money, but those are legal fees, not the actual cost of the algorithm the way that you must continue to pay a human worker because they will stop working if they aren't paid.) Every single person who makes money rotoscoping, motion tracking, or really ANY job can be put out of work by an algorithm overnight. Those jobs will exist until the day it's lights out, and we really need to have the foresight to do something about changing the way we think about the value of labor, before the disaster point. The scary part is the vast majority of political leaders seem to have no concept of the massive change to our economic structure that machine learning will bring in the next few years.

"Timeline level resolution, frame rate, scaling and monitoring settings within the same project" FINALLY. It's one of those things that wasn't a dealbreaker, but was annoying for a few specific projects. "Adjustment Clips to apply filters, effects and grades on top of a range of timeline clips" Nice. One of the main things I miss from Premiere. "Improved Fusion playback performance" That's been my main complaint against Fusion's integration vs. Fusion standalone--glad to see improvements in performance. "ResolveFX Object Removal in DaVinci Resolve Studio" Really looking forward to trying this out. "ResolveFX Vignette" I could have used this on one or two projects in the past, glad to see it. "Playback speed indicators in Resolve viewers" Cool! This will help with creating speed ramps. "Support for using DaVinci Resolve Studio dongles to enable Fusion Studio" WHAT? This is AWESOME! I'm so glad I never bought a Fusion dongle now haha. It doesn't look like they have added support for FLog in the color space transform. That would be a nice addition for Fuji users. Though, to be fair, I've found HLG to be better than FLog on the XT3 so I probably wouldn't use it that often.

Looks like Blackmagic skipped a few versions and is going to Fusion 16 now. I'm glad to see that they are keeping standalone Fusion alive, at least for now.

I don't know if I'll be able to support BRaw, since right now I'm using ffmpeg for decoding--though hopefully they will add support eventually. I will look into it though, since technically all I need is an audio stream to sync with, perhaps BRaw has just a normal WAV stream and I don't need to actually decode the video at all? I'll look into it.

Yeah, and that's today. By 2020, I expect there to be FF 4k60 without binning! My point is that the bar of "exceeding expectations" will be higher then than it is now.

The panasonic cameras are today's models. 4k60 with good af might be class leading if it were released now, but in april 2020 it will be competing against the next generation. Sony will need much more than today's specs to top next year's competition. And the s1r isn't cropped, its full frame 4k60. ...Which is not to say that the a7s3 definitely won't be class-leading. Maybe it will be an 8k global shutter camera with triple base ISO, etc. etc. Maybe not. And in the end, I'll get it if it's "only" 4k30, but ticks all the other boxes for me (price, ergonomics, codec, etc). I'm very interested to see what Sony makes, but at this point it does seem like they're playing catch up.

Sony's new HDC-5500 camera has a 3-CMOS design with global shutter. Considering how good current cameras are with sensitivity and dynamic range, I'd love to see rolling shutter reduced/eliminated in the next generation of cameras. There have been a number of global shutter cameras announced in the past year, and I'm glad to see that technology becoming more common.

If Sony ships the A7s3 in April 2020 with 4k 422 10 bit downscaled from a 6k sensor... they'll be about one year behind Panasonic, depending on when that firmware update drops.

Except z cam! But yeah that would be cool.

I am currently making an external tool to do just this and I'd love to hear about what kind of process you are looking for. What kind if workflow do you have? Timecode? Is there exactly 1 audio clip for each video clip? What camera and audio recorder do you use and do they have relatively accurate time stamps on the files themselves?