kye

Possibly, but you know Sony - only the latest releases are updated and the rest are dead to them!

Yeah, I think I'm just a bit frustrated overall TBH. I'm also fighting with Resolve over half-a-dozen small issues, to the point where I'm genuinely considering writing my own colour grading plugin where I can get what I want. I'm also rapidly becoming frustrated with the colourist forums too, as recently I randomly heard a snippet in a podcast that was a simple and beautiful answer to a question I asked several years ago and got zero helpful replies to, but that everyone involved in colour for the last 30 years would have done hundreds of times on the average project, so now I'm wondering what the quality of the rest of the info on there is. I also watched something a number of months ago that was right at the cutting edge of colour science tools which was what I was thinking about but no-one is talking about. I know it's a sign that I'm progressing beyond a certain level of knowledge/skill, but I just find it frustrating to be seemingly going against the current in yet another area of this stuff. What does this all have to do with MFT cameras? (apart from being frustration in general..). Well, the better I can capture things in camera the less work I have to do in post.

Yeah, but remember, our situations have almost nothing in common. You're shooting dreamy images for someone else to pay you money and doing so in a situation where you're expected to be taking up space and using as impressive a camera setup as possible. I am shooting environmental images for myself in situations where I'm discouraged from taking up space and where having the smallest possible setup is to my advantage. As far as my understanding goes, if I was doing what you do then I'd also have the same thoughts as you. I imagine that if you were doing what I do then you'd also understand why I have the priorities that I do. It's easy to look in at a situation from a distance and see enough of it to have a good overview, but to be far enough from the details that the subtleties aren't apparent. The FF user telling the MFT user who wants small cameras that FF is just as good as MFT is like the cinematographer telling the wedding shooter that an FX9 is just as good for shooting wedding videos as a mirrorless, and completely missing that the price, size, weight, workflow, and a dozen small details make the proposition impractical at best. The comparison probably isn't bulletproof and you could probably find weaknesses in it, but I've had enough issues in the field with the size of my XC10 and GH5 that I can pretty confidently say that the considerations involved are real, even if they aren't obvious to outsiders. I mean, the other recent thread from @John Matthews was about cameras substantially smaller than the GX85!

Not really. All I want is a GX85 with a modern sensor, so Dual ISO and higher DR. I don't need / want LOG, 10-bit, 422, ALL-I, RAW, PDAF, a flippy screen, dual card slots, 6K or more, mic-in, etc, etc, although if they came with no penalties then I wouldn't hate having them. In the real world though, they do come with penalties - huge ones (see rant at the end of this post). Absolutely. Yeah, this thread was more of a temperature check, just to make sure there wasn't something I'd forgotten, and to have another thorough think-through of it. I'm not sure I'd swap to another system right now even if it was free. The GX85 with 12-35mm still feels on the large side, so casually stretching to something 50% larger isn't a casual proposition at all. That's my feeling too. Panasonic has said there won't be more compact cameras, but has also said over and over that they're fully behind MFT, and having spent a long time developing the dual-gain PDAF sensor in the GH6 it would be pretty easy for them to capitalise on that investment by doing refreshes of the rest of the lineup. Hell, even if things trickled down and they released a GX camera with the sensor and processors from the GH5, that would be a good outcome for me. I'm actually deeply suspicious about the current size comparison between MFT, S35, and FF cameras, and how they're all a similar size and weight. The FF bodies with IBIS do tend to be heavier, but are the same size, and something smells fishy to me. If a sensor is smaller then it takes less power to run it, and it takes FAR less power to move it around for IBIS. The sensor is also a lot lighter, and the IBIS motors will be dramatically lighter. This means that if all else was equal the batteries should be smaller or last longer, and the internals should be smaller and lighter. I understand that the screen is independent of sensor size, as are lots of other things in the camera, but when combined with a smaller sensor with a lower resolution the processors should be smaller. The camera module from the iPhone is practically microscopic. BUT, that's not what we're seeing. I suspect it's a combination of lowered IBIS performance (the motors not being able to move as fast or as far, but no technician out there will get their shit together and actually make a test setup for this), combined with the manufacturers just not trying, which I think is quite sad. What this means in practice though is that it should be easily possible for a GX sized camera with modern components and design to offer a serious challenge to the specs of the larger S35 and FF cameras. Now I've moved to zooms, I seem to be swimming against the prevailing winds of the camera market once again. I want MFT zooms to be F2.8 or slower, a standard 3x (~24-70mm) or 10x zoom ratio, and as small as possible. I'd be happy if the attempts to make them smaller meant that they were 12MP or less in resolution. Instead, the market has put making the fastest and sharpest lenses known to man as the first priority, relegating size to third priority (there is no second priority), and because everyone is shooting wide open they've managed to get the whole industry to basically stop discussing anything else except 150fps animal-eye-detect focus performance. When I think about a FF camera with an F2.8 lens I think of it having too shallow DoF for my compositions, and by the time I stop it down I now need the second base ISO just to give it a normal exposure after the sun has gone down. It's been a long journey for me, but now MFT seems to be the sweet spot of DoF and exposure to create the right aesthetic in a sensible amount of light, which makes the AF performance far less critical, and when combined with a 4K sensor the data rates aren't stratospheric, the lenses can be budget friendly, you can do it on affordable SD cards internally, making the whole camera package smaller. ....not to mention you don't need a supercomputer to edit the footage. The "better" cameras on the market seem to have "improved" every aspect, but the challenge is that when you "improve" one aspect you make all the others far worse, in a never-ending game of "one step forward, six steps back, and how will you be paying for this today sir?".

Ha! What a great word! 🙂

CineD says so. I've kept a spreadsheet of all their DR tests and random mentions, and the following image appeared in one of their early reviews before they got serious about publishing the lab tests - it includes both the P2K and GH5S: They haven't tested the external RAW update, so if you want to assimilate your GH5S into the collective then that might give a bump? DR is a tricky beast though, depending heavily on how you test it and the NR applied, etc, so I'd suggest the above is a rough statement rather than a definitive result without any caveats, because we don't know exactly what they did and why in the test.

It does seem to be a case of the missing Panasonic GX-series camera with modern sensor. Revisiting the lineup, each with their smallest equivalent of a 24-70mm and a 24-240mm, here's the GX85, the A6700, the FX3, and the FP: The A6700 seems to have good options in the 18-55 and 18-135mm variable aperture lenses, which aren't too big. The FX3 would be the choice for Sony FF due to its high-ISO performance, and the tiny 28-60 really is tiny. The 24-240 is a lot bigger than the 14-140 on the GX85, but it's pretty small for FF. Lastly, the FP seems to only have OIS options for 24-105 and 28-200, but they're not in the camerasize database, so pictured above is the 24-70/2.8 which is a tank. Canon and Nikon are notably absent. By the time the FX3 and FP are in the mix, we've deviated so far from the GX85 that we may as well be having an open house. I guess the summary is that there isn't anything in MFT, but if I wanted to completely re-buy everything in my entire camera bag, there are options that are larger, incredibly more expensive, and better, but no-where near better enough to make me consider any of them for a single second. The more I think about it, the better the AI options in post seem.

I suspect it's a symptom of how the camera processes the image rather than a limitation of 8-bit images. I know this because the images you posted to the forum will be 8-bit images, so if it was a limitation with all 8-bit images then it would be visible on all images posted. I can see it, when zoomed radically and saturated and being pointed out, but suspect that it wouldn't be visible under normal circumstances. If that was the price to avoid overheating and storage anxiety while recording then it would have to be a pretty specific set of circumstances for that to win out over the increased creativity.

Well, that rules them out, which is unfortunate. Yeah, it has to be MFT. Not only to share my existing lenses with the GX85, but also because the FF lenses are gargantuan in comparison and pretty sparse really. Not only would it be too large (and "Smallest" is literally the first word in the title of this thread) but compatibility is another whole thing that just isn't a factor when sticking to a single lens system. Yeah, it's just the frustrations of when I hit its limitations, which when I'm travelling is a multiple-times-per-day sort of thing, and is often in a that-would-have-been-the-killer-shot-from-this-location-but-the-camera-couldn't-do-it sort of way. I'm fully aware that I want a camera that can do a wide range of things very very well, but you can't fault someone for trying to improve things. The thing I find fascinating is that even if I had a million-dollar budget for a setup, I still couldn't get what I want.

10-bit is always better than 8-bit, and 4:2:2 is always better than 4:2:0, but the question is if that difference is actually visible / meaningful. If you're doing very little grading in post then 8-bit vs 10-bit and 4:2:0 vs 4:2:2 probably doesn't have any visible or meaningful difference. But I suggest doing a test. Just find a bunch of scenes around the house (or not out of the way) and shoot the same composition in each mode and then just load them up side-by-side and look at them. If the 8-bit 4:2:0 shot doesn't make your stomach turn then just go with it. In the end, if you can lessen the impact or eliminate overheating, that you can lessen or eliminate the impact of running out of SD card space, then that means you can relax more while recording, you could shoot more, you can have extra time and energy and headspace that was previously devoted to worrying about or managing these things. If you can have more headspace and be more relaxed while shooting then the way you use the camera, and the way that you behave while recording will be better. If you behave better then what is in front of the camera might also be better. So, realistically, the option is potentially of having a very slightly worse recording of potentially much better material.

It's in that camp with the GH5 as being on the large side.. Realistically, I like the GX85 but just want some upgrades in DR and low-light, and in the absence of that existing I've worked out that I could have two setups, one for running around and one for slower situations where I don't have exactly the same requirements. So the GH5S is better than the GX85, but is larger and has worse DR than the P2K, so it's sort of wins and losses in comparison. Considering I already own a GH5 and a P2K, it's a tough sell to buy a whole new camera for essentially no net benefit. I know very little about Olympus lineup and was secretly hoping that someone would go "there's always the Olympus XYZ - it's perfect!".... what models might suit?

I should say, that video is NSFW...

Well, you fooled me into thinking you didn't struggle. In my travels I've found there are only a few types of editors: People that love it and push themselves and put in the hours to create great work People that don't love it, but have somehow found a way (or get pushed by someone else) to grind through the process to get great edits People that don't put in the work and create edits that are good enough / functional / uninspired / plain / blah People that create shit edits People that don't edit and either do 25 minute rambling time wasters, livestreams, or simply don't finish projects at all Cool to hear that you're editing women skaters like men, and especially with someone as skilled as Chynna. I skated for many years in my youth, both skateboarding and later roller-blading, and I have just enough skill to know how crazy the tricks in that video are. Especially considering they were all shot in one day, I assume from the title? If you're someone who likes analysing the work of others and can take tricks from one genre to another, this episode of Parts Unknown is the best editing I've ever seen: It has every effect I've ever seen in those trendy 'transition-fest' travel videos from the hey-day of Vimeo, but done way better, has transitions I've never seen elsewhere, has editing styles from across the board, and makes all other non-narrative editing look plain and pedestrian, and makes the best of YT creators look like amateur hour at the local craft fair. Anthony Bourdain was famous for giving editing notes worse than anything the YT comments section contains, and doing it for version after version after version, and I think this might have been the high-water-mark for this type of highly-stylised editing.

What are the options for a small MFT camera with usable AF-S, decent DR, and dual-ISO? I shoot family and travel videos, which are fast-moving, completely uncontrolled situations, and in available light. The cameras on my radar (and their associated issues) are: GX85 Great all around camera and currently my main camera, but is weak in low-light and DR. The low-light is ok with F2.8 lenses in well-lit places at night, but darker areas require primes (either expensive AF lenses or MF) and this quickly gets into the territory of lenses being a bit soft, DOF being too shallow for the compositions or making focusing difficult. The DR is good enough when the sun isn't low in the sky, but when it is low then the lack of DR starts making me have to choose between faces or not having a digitally white sky. P2K (OG BMPCC) Low light has same issues as above. DR is excellent and is enough to not have difficult shooting when the sun gets low. The AF is the weak-point though. It works, most of the time, but it's so slow that if there's any movement in the frame it can't go fast enough and gives up. It's also so slow that I tend to miss half the moments involving people, so it's really a camera for filming stationary objects. Plus the screen is ..... problematic. It's black with polarised sunglasses, the little AF box is very difficult to see in many compositions, it's not bright enough for shooting in direct sun, and it doesn't articulate. M4K (BMMSC BM Micro Studio Camera) Dual ISO, so low light should be excellent. DR should be really good. All the relevant specs from BM point to this being the same sensor as the P4K so the image should be excellent. AF is unknown, I'm waiting for tests. The screen is external, but I could use it with my M2K (BMMCC) setups, which make it almost too large for consideration but maybe not quite. It needs an SSD in addition to the M2K setup, so that's a whole other thing too. GH5 / GH5ii / GH6 / G9ii / etc Isn't there something smaller??? I'm not in any hurry for this, so perhaps it's a waiting game for Panasonic to release something smaller with their new sensor lineup?

That's a cool edit too @Al Dolega. Nice pacing, good rhythms, nice use of the non-trick shots to break things up, etc. These more 'percussive' styles of editing remind me of programming break-beats in music production. You need the right amount of things happening on the rhythm to keep it going, but enough things that are off to keep it interesting and fresh, and enough variety in the whole thing to separate it from everyone else's work.

Nice images @gt3rs!

It might be testable. Record equivalent clips of motion in 120p and in 24p, then put them both on a low-resolution timeline like SD (which will smooth over any codec / bitrate advantage of the 24p) and see how they compare. Obviously you'll need to use a short shutter speed so it won't be the nicest images ever, but you should get something comparable with RS as the only significant difference. I've done extensive tests on both 60p OS monitors and on dedicated controlled monitors and 24p wins over 30p in all situations by a significant margin. The fact that 24p has jerky pans isn't ideal, but I only find it visible in the odd shot here and there, whereas the slippery sensation from 30p completely saturates the images and looks awful the whole time. I raised the idea of setting the display to 48 or 120Hz because I'd like to eliminate the jitter of 24p on my 60Hz OS display, but I'm a million miles away from switching to 30p, even if my display was fixed at 60p forever. Perhaps the whole point of the 24p thread was that one person shouldn't tell another person what is good or not good to see or prefer in an image. I suspect that some codecs contribute to this as well. For example if you have an IPB codec with a low bitrate then in moments with high motion the codec won't have enough bits to describe all the parts of the image that have changed, so there will still be remnants of the previous frame that are included in the current one. I've seen this in real examples and things like low-contrast textures on a wall can remain in place while the camera pans, making it look like the wall is moving but the surface of it is staying in place, like the wall is a portal to another dimension, etc. I'm not sure where the motion cadence magic comes from, but my suspicions agree with you that it's likely a combination of a number of factors. One thing that I suspect plays a part is the tripling of some of these codec/compression effects: you shoot on a camera that processes the image (potentially emphasising these with sharpening, NR etc) the camera then compresses in a codec which creates a number of false motion artefacts in the footage you edit and process the image (potentially emphasising the artefacts with sharpening, NR etc) the NLE then compresses in a delivery codec which creates a number of false motion artefacts in the footage then this footage is uploaded to a streaming service where it processes the image (potentially emphasising these with sharpening, NR etc) the streaming service then compresses in a very low-bitrate streaming codec which creates a number of false motion artefacts in the footage There are likely to be artefacts that get 'seen' by the processing / compression algorithms and further emphasised. For example in a panning shot anything that doesn't move might be deliberately left alone as a compression optimisation, so by the time that you're watching it on YT the image has deviated significantly from the source.

Since that 24p argument discussion I've been thinking about frame rates and motion a lot and shot a few tests etc, and eventually got around to thinking about how to view content on the computer properly. I finally realised that the solution is to set your display to 48Hz (or 120Hz / 240Hz if you can do it) and that will give you proper 24p playback. I do all my streaming and YT viewing on my computer via a display that the OS controls, so currently with a 60Hz refresh rate I'm perfectly recreating the 30p (yuk) and 60p videos (double yuk!) but the 24p is jittery. My current MBP can't do 48Hz but the newest ones from Apple can do 48Hz and 120Hz / 240Hz, so it's on my radar. I currently edit in Resolve through a UltraStudio 3G box that drives the display directly with whatever the timeline resolution and frame rate is, but that's only for my videos, not for watching anything else unfortunately.

With all the digital standards out there it's unfortunate that no consumer camera manufacturer developed the hot shoe as a modular communication standard. We could have had add-on EVFs, microphones and audio adapters, monitors, timecode boxes, etc. The argument that they wouldn't have sold becomes quite different if you think of the potential sales of accessories over 10 or 20 years of cameras that supported them. Imagine if you could buy an adapter that mounted into the hot-shoe and gave two or three hot-shoes across the top of the camera (maybe only to the left to avoid the dials) and you could put whatever you wanted in there. USB was released in 1996, almost 30 years ago with 12Mbps, and by 2008 it had 5Gbps, so it's not like the technology in, say, 2004 wasn't there yet.

I've been watching the TV show M*A*S*H (1972-1983) and I'm finding it really interesting: The storytelling trumps the image quality, and it only takes a few seconds to stop looking at the image and start paying attention to the content It was shot on 35mm film, but was before colour grading, so the image is about as pure as you get The image is rather drab in comparison to any image that has been colour graded Whenever they cut to a shot that fades out or does a freeze frame the image degrades significantly in resolution and the colours and contrast shift significantly, presumably because it had to go through another generation to add the effect, so the quality and resolution limits are no joke However, having said that, the skin-tones and the high-DR external shots filmed in full direct sun are world class, even compared to the flagship cameras of today from ARRI. It makes it obvious that ARRI are in the business of making cameras with as many film-quality pixels as possible, and everyone else is in the business of making cameras with as many pixels as possible and letting quality fall where it may.

Meanwhile, wedding videographers are getting pinged for having audio that's too clean due to AI processing.. linked to timestamp: How on earth did we get here....

AI generated cartoon aired in China. It's a 26 episode series of 7 minute episodes about classical Chinese poems, so is sort of a cultural education thing. It took 6 months to do using AI text-to-video and the studio that developed it released the series to coincide with the opening of their new AI studio.

Just thinking about this more, and I recall that testing the GoPro on that little outing was actually really illuminating. I have always shot video in uncontrolled conditions, so never seriously studied coverage or blocking or sequences etc. What I found was: I shot everything I could see (which was really fun because I didn't have to wait for the camera) Not being able to see (the GoPro I had didn't have a screen) eliminated all aspects of chimping Pulling everything into the edit, I realised that I had so many different options, and make it a more whimsical edit with little sequences of fast-cuts like when getting seated at the cafe, and then again for standing up and walking out The success of those fast-cut transitions made me realise the value in having lots of shots to cut up, but also made me realise things like you can cut faster if you keep the compositions similar between shots so that the subject doesn't move, etc I made several videos with it, essentially going through a learning process each time, understanding more about how you can push the edit pretty hard without it seeming odd, realising where the barriers were in the edit and how I might solve them when shooting, shooting with that in mind, pushing further in the edit, etc I highly recommend shooting a few test videos like this, literally of anything, just to get a sense of the possibilities. Also, it's good to specifically test: How long it takes from hitting the button to the first frame, and how much it chops the ends off clips (the clip may end before you hit the stop button) How far off-centre you can put a subject before the wide-angle lens starts becoming unflattering or straight-up odd How active you can be in shooting and how much motion that creates in the frame (knowing you can run and get usable footage for example might open up possibilities) How far you can go in low-light before it's no good (for example, it might be ok indoors during the day in rooms with windows but not ok if there aren't any, etc) You probably know all this, but in case you don't, or the lurkers don't, this stuff is invaluable. Unfortunately, despite me shooting little tests quite frequently, I always learn things on trips that I wish I'd learned before and come away basically thinking I've screwed up each one.

The only reason you might not be able to is if the optics in the speed booster protrude too much and will interfere with the lenses you're planning on using. I have a vague recollection this was the case for some of those BM-specific speed boosters, but couldn't tell you which ones.

Definitely agree about the bulk of an external monitor. I've been bouncing between my OG BMPCC (P2K) and BMMCC (M2K!) for years now, and one of the biggest challenges of the M2K is how large the screen makes the rig, even though I have the Ikan 3.5" monitor: it requires a separate battery the mounting isn't slimline at all and so the monitor is quite far from a camera (compared to a flip or tilt screen) the HDMI cable takes it a very large step into "we are the Borg" territory I've used camcorders in the past, and also have the XC10 which is a camcorder in all but body-design, but my GX85 + 12-35/2.8 setup is what I use as my main camera now and I use it like a camcorder, so I get most of the benefits: It looks like a little P&S (because it is) so I invoke the dorky Dad vibe I'm planning to also get the 14-140mm F3.5-5.6 lens to pair with it which will give it a similar all-in-one lens The Dual IS gives it equivalent levels of stabilisation It's limited to 8-bit 709 style images, but these look pretty good in reality, so has a similar limitation but also strength in that regard too It lets me stop thinking about the camera and start thinking about what is in front of it