aldolega

Yea, it's frustrating to see the camera companies have such a blind spot for this part of the market. Seems to me that a small video/cine cam that had the IBIS and AF tech we're seeing in mirrorless, but still with the IQ and codec pluses of a "real" camera, would be a big seller.

I've only shot with an A7SII once, but HD resolution is important to me too, and I was also initially skeptical of the A7III, so hopefully I can help. My benchmark for good HD quality was my GH4, which looked pretty great IMO in both its 200mbps All-I and 100mbps IPB modes. I film stuff with lots of camera & subject motion and detail, so I would use the All-I mode if I was moving the camera a lot, or the IPB if it was a simpler pan or tilt (IPB is sharper at lower motion levels but degrades as motion increases; All-I is less sharp than IPB's max, but stays consistent despite motion). After the GH4 I went to the G85, and the lower 28mbps HD was definitely softer than the GH4's HD, unless it was an almost static shot. One of my only disappointments with that camera though- price, IBIS, and the slight color and ISO improvements over the GH4 outweighed the HD quality dip for me. I was all set to go to the GH5, which by all accounts has pretty stunning HD resolution, when the A7III was announced, at 50mbps in 1080p60, and 100mbps in 1080p120. Bitrate isn't really an absolute measuring tool though: there are plenty of cams that can pack tons of detail into even less, and Sony says its encoding is two-pass vs. everyone else doing one pass, so it seemed totally possible that it could be fine. And bitrate is only part of the equation anyways: how the cameras are reading the sensor in HD mode is the real key here. I wanted FF for photos, and AF for video, and I had been tolerating the G85's HD pretty well, so I went for it, and I'm mostly happy. Its HD is better than the G85 to my eye, but not quite up to the GH4, although the 96p was garbage on the GH4 so having 120 that's at least decent (and not cropped) kinda helps balance things out I guess. I haven't shot with an a6300/6500 but from what I've seen online, their HD is pretty awful, and the A7III HD is definitely better. These two videos from Max Yuryev helped put my mind mostly at ease, hopefully they will play at the relevant bits: So, a little worse than the A7SII, but better than all the other Sonys, especially the a6500. Personally the AF and better rolling shutter alone outweigh that small difference with the A7SII, not to mention all the other improvements, but it's still kind of a personal call to make. Motion cadence, I find it fine, but I shoot mostly at a (gasp!) 30p base, not 24p, so I am probably not the best judge of this. The 24p stretch thing was like 1% I think? Not even noticeable unless you're A/B'ing identical shots back and forth. And I believe one of the firmware updates has fixed it anyways? Oh, and here's a handy tool over at DPReview, for comparing video res between cameras. Doesn't tell you how it performs with motion, of course, but it's something. https://***URL removed***/reviews/image-comparison/fullscreen?attr29_0=sony_a7iii&attr29_1=sony_a7sii&attr29_2=sony_a6500&attr29_3=sony_a7riii&attr72_0=1080&attr72_1=1080&attr72_2=1080&attr72_3=1080&normalization=full&widget=602&x=-0.48927038626609437&y=-0.17768028516588977

Seems to me that a typical mirrorless body has no space for a filter to swing/slide out of the way of the sensor. So you either make the the body much bigger/thicker, or use a fixed filter, which then eats up at least a stop or two of your ISO performance (whatever the minimum power of the eND panel is). Either way you're affecting aspects (size/weight, ISO performance) of the camera that the camera makers consider much more important than a request from niche video-centric users. And who knows if an ND and a physical shutter will both fit behind most modern mirrorless mounts. Sure, FS5 etc have NDs, but they don't have shutters. Global shutters that will work for photography (i.e. truly global, and don't affect sensor performance in DR/ISO) are still years off, so I wouldn't hold your breath about it.

No spreader, the legs are single-tube, photo style construction. They're just the Manfrotto 535 with a Sachtler badge on them (Sachtler and Manfrotto are owned by the same company).

I've been very satisfied with my Ace L head and Sachtler carbon legs (single-tube, can't remember the model right now. Same thing as the Manfrotto 535). Very smooth, easy to balance. I do wish it had one more step of drag, but that's a bit of a nitpick. I've used it with my GH4, G85, and A7III, all with a cage/rods, a small monitor, small mic, vintage full-frame glass, and occasionally a follow focus. So maybe 7-8lbs with one of my bigger teles? Performed well with everything, especially if I have time/space to hang my bag on the legs for some more stability. Keep in mind I have the L, not the M, which has a lower weight capacity (4kg/8.8lbs I believe). The L is rated for 6kg/13.2lbs, and the old rule of thumb I've been told is that for fluid heads you don't want to go much over half the rated weight capacity for best performance. So that makes sense with the weight of my setups. Build is a tiiiiny bit plasticky, compared to the FSBs, but still very sturdy. Definitely a step up from Manfrotto, Benro etc.

CDAF is software-based, nothing to do with the sensor physically.

I agree that the A7III is the reference point for pricing, and probably will be for another 6-9 months, but it's not a perfect camera and comparisons can never be absolute. If the XT3 had IBIS I feel it could and would comfortably sell for almost $2k, even with the A7III and Z6 right there with bigger sensors. The A7SIII being $3k would be a pleasant surprise, at launch the A7S was $3200 and the A7SII was $3400, if I remember correctly. The only way I see the A7SIII being less than $3400-3500 is if they skimp on a feature or two (say, no 10-bit anything, and no new codec. Or if they stay at 12MP). The A7000 costing the same or more than the A7III makes total sense, assuming it's got the high-end specs that outclass the A7III. The 7DII cost the same or a bit more than the 6D, same with the D500 and D610. I expect the A7000 and A7SIII will be pretty similar specswise (new sensors, 4K60 8 bit, possibly 10bit out but I doubt it, probably the same old h264 100mbps codec, third-gen body/battery, nice EVF and screen, new AF from the A6400), with mostly the sensor size differentiating them. $2000-2200 and $3400-3600. I don't see the new HEVC codec making it into the Alpha cameras anytime soon, the broadcast/cinema divisions will demand to have a year or two with it to themselves. I think 10-bit HDMI is the most we can reasonably expect in the next generation.

"My $1500 camera's screen isn't as good as my $6,000 camera's, or even my $3,400 camera's! What gives?"

I went from a GH4 to a G85 and loved pretty much everything about the G85, except the HD bitrate. The build was still nice (the G85 is half metal, half plastic, whereas the G7 is all plastic), IBIS was a game-changer for me, and the lowlight and color were a bit better too. But the 25mbps HD wasn't quite enough to stay sharp during fast action and camera movement, which is a lot of what I shoot. The codec has to default to a softer mode to keep up, and it's noticeably softer to my eye than shots with less motion. The GH4's higher bitrates didn't have to do this as they had enough data to work with to maintain a higher level of detail. But, so long as there isn't a ton of motion in the frame, and you don't try to grade/CC it too much, the 25mbps looks pretty great. Which is how 95% of people will be using it.

This is almost more useful for guessing what the camera will be like than the sensor specs, IMO (assuming these are indeed the codecs in the SIII). The max bitrate is clearly intended to work with U3 SD cards (which do a minimum 30MB/s continuous write, aka 240mbps). So Sony is probably shy to switch to a new media format, or even a more specific SD class/rating. So no new media format, which doesn't necessarily mean the same third-gen body style, but it doesn't rule it out either. They'll have to have dedicated processors that this codec is designed around: I would assume they'll be at a modern level of power/heat performance, which may also point to not needing a bigger body (while also not ruling it out).

These sensors seem more suited to an A9II, or A9R. Mayybe an A7RIV? They need a higher tier of processing power, and thus will come in at a higher price, and thus a different market, than the A7xxxx series. IMO all the A7SIII really needs to be, to be a huge sales success right now, is just an A7III with a 18-20MPish sensor (so APS-C 4K is 1:1 or slightly oversampled), 4K60p, and the better EVF and screen of the A7RIII. That's it. Everything else about the A7III is good enough right now. Release this cam now and every workaday videographer and sub-Hollywood production house buys it, simple as that. 8K, insane framerates, 10bit, the other nuances we wish for here just aren't as important as we think they are, not to the bulk of this camera's market. The market that really cares about these things will have zero problem paying another $1500 for an A9II.

Not if outputting HDMI restricts AF, like it does on my A7III (no face detect when using a monitor). And I believe the GH5 AF is worse with HDMI being used, too.

I hope they continue with both, and don't handicap or pigeonhole one line in favor of the other (like only making ultra-compact bodies for m4/3, or keeping video features out of the FF bodies because they're "for photographers" ). It would be really nice if they made an adapter to use m4/3 lenses on the FF mount, so people can ease into the new system, and to make both systems more appealing. The SL mount is a big enough diameter that I think this could work.

It can be done on a few of the Minoltas, but not all, FYI.

The Lens Turbo II's were fine in terms of resolution/sharpness and color for me... but they went downhill fast when shooting into bright lights or sun, or even just most moderately backlit subjects. Flaring and hazing galore. And the "blue dot" problem, while much better in the second gen models, would still pop up pretty frequently. At first I blamed it on the MDs being old and lacking modern coatings, but the Nikon version had all the same problems with modern Nikon glass. 10bit is nice, but at the time I was on a G85, and my choice was either go GH5 and stay frustrated on the photo side of things, or go A7III and be happy there and stay in roughly the same territory videowise, and be able to actually use 100% of my lenses without hassles.

I went through basically the same thing with m4/3... the Minoltas are great glass, but unfortunately just a bit too short of a flange to adapt to any modern SLR mount. I've never found the cheap boosters to be up to snuff- flaring, soft corners, low contrast, backfocus going out because the adjustment screw goes loose, sloppy fits... just a lot of frustration. I went through a few and eventually just had to plunk down on two used Metabones boosters, one for the MDs and one for my Nikon glass. Then the A7III came out, and I realized that it was actually way cheaper than a GH5 with the two Metabones, with less fuss. Plus now I can use an ND Throttle adapter for outdoors stuff, and have even less fuss with external NDs.

Sounds like the ND adapter will basically be an OEM version of the FotoDiox ND Throttle. Which I've actually been pretty pleased with so far actually. If the Canon version is electronic, and/or can move out of the way, that would be pretty awesome.

Panasonic's peaking never let me down, so long as I fed it a scene with at least decent lighting and contrast. Which it sounds like you will definitely be doing, I imagine surfers against waves will give more than enough contrasty edges for the camera to detect and display peaking on. So maybe look for a G7, G85, or GH4. Alternatively, I can recommend the Lilliput A5 as a cheap-and-cheerful little 5" monitor with good peaking. It's not the brightest thing out there though, so to use in bright beach sun you would probably need some sort of extended hood. Upgrading the camera is probably smarter. If you get a 4K model, you'll be able to crop a bit (or use ETC mode), which will give you even more reach. Plop a cheap magnetic loupe on the LCD, switch to monochrome display mode with blue or red peaking, and off you go.

I'm excited for the specs that Vitaliy posted, but I have a sneaking feeling that while this will be a good camera, it will still be far from hitting the potential this camera size/niche of the market has for an incredibly popular milestone-type camera, that's good for vloggers, adventure/nature types, travelers, street photogs, etc. With the market shrinking it only seems logical to try to sell to as many of these little niches as possible. Give us something that's just a little bit bigger than the LX100, but has: -a flip-out screen, OR a tilt-only screen that flips up over the top BUT with a relocated hotshoe so an appropriate small mic, like a VideoMicro, doesn't block it. -a slightly ruggedized body with some rubber coating/bumpers. Not so much as to make it bulky, but just enough to protect against small bumps, and make it secure to hold. Like the "Active" Samsung Galaxy models, which basically have a mid-size protective case built into them (but not to the level of the huge massive Otterbox cases). -the lens needs to go just a little wider. 21 or 22mm equivalent would make the camera so much more appealing to vloggers, as apparently 24mm is just a little uncomfortable to selfie-film with. If it could do this while staying fast and still going out to 70mm, great, otherwise I would prioritize speed over tele reach. -make the filter threads and material around the front lens element from steel, to allow for magnetic filters that easily click onto the front of the lens. I'm assuming that this body size would not have space for a moveable internal ND, so this allows for a quick and relatively painless filtering option. -make a matching dedicated hotshoe mic, like the MS2 you came out with for the GH3, but that connects via the hotshoe- no cords. It needs to be small- VideoMicro size or smaller- and it needs to flip over to face backwards so that vloggers can narrate during shots (picture how the SmallHD Focus flips over). -a bit bigger battery would be nice. USB charging/powering is a must. I know normal USB won't supply enough current to charge and run the camera at once, so make the camera compatible with the QuickCharge protocol, which lets a QC charger send more than 5V (9V or 12V, I believe). I believe USB-C allows for.more current as well, so maybe that's a more modern option. -mic jack- put it somewhere where it will never interfere with the screen or the mic itself. Maybe on top of the body? Or on the front? -stick with 12MP, or lower- keep those pixels big. Keep the oversized sensor and the MAR cropping. -steal a page from BlackMagic and design the grip and button layout to also function well when held "backwards" in selfie grip. If there's a joystick on the back it should be positioned so you can use your index finger on it to move the focus point when in selfie grip. -use the position sensor on the screen (that tells the camera to flip the image when the screen is turned/flipped forward) to trigger a selfie mode, where buttons reconfigure for selfie grip, maybe focusing distance is limited to less than 3ft unless a button is pushed, etc. And definitely allow the user to configure what this triggers. -don't be afraid to offer a non-standard video resolutions, to allow for higher frame rates while keeping over-HD res, especially if 4K60 isn't doable. A 2560x1440 60p mode would be just as good for most people. Or if 4K60 works, give us 2.5K100.

I don't think the size/mass of the sensor is that limiting in itself, as it's still a very light assembly. Just would need a bit larger magnets than a m43 sensor... which would draw more power I guess. I do agree that it's probably the mount diameter and body size that are the limiting factors. The body size can change, I can't see how Sony could do a new mount though. I wonder if perhaps there's generally less "extra" image circle to utilize with most full-frame lens designs? I remember when the LS300 came out it and people started playing with the variable crop, many m4/3 lenses had a ton of "extra" image circle to use, with some even covering the whole s35/APS-C sensor.

The more appropriate question is: why would you want to?

The optical design was updated for the Ultra, and I believe it uses a better quality of glass as well. The designer is actually on this board, Brian Caldwell. Maybe search for his username and you can find his older posts where he explains the exact differences.

It's... adequate. Everything is fit together just fine, my NP-Fs engage and disengage nicely, and the HDMI jacks seem sturdy. But I doubt it'd survive a drop onto cement, and the 1/4-20 mounting sockets wiggle a bit in the casing so it's a bit wobbly when you grab it to reposition it, and the Velcro on the sunhood is not all that velcro-ey... but come on, it's $150, none of this surprises me. I haven't handled an A6/F6 or an F570 so I can't compare directly unfortunately. Whichever one of those has the metal casing is almost certainly better built than the A5, but they're also larger and heavier, which were minuses for me. I shot with it again tonight and it's doing all that I need from it for my A7iii, meaning mostly just a higher-res display with better peaking. LUTs, a touchscreen with a nice UI, and a brighter screen would all be real nice, but even if those were available in a 5" 1080p monitor right now, I don't think I'd want to spend for it anyways (assuming it would be at least $300-400), as it would depreciate so much once the NinjaV comes out. Even if I can only get $50 for the A5 at that time, it's worth it. And for $50 I'll actually just keep it as a backup.

I got my Lilliput A5 last week, used it on a quick gig the next day with my A7iii. I'm satisfied with it, happy even, considering the price. I like the size, resolution and super low weight. The buttons and menu are a little fiddly but not too bad. Can't figure out how to re-assign the single function button. The brightness was better than I was expecting. It'll do just fine for me until the NinjaV comes out.

I really hope Panasonic has an incredible LX200 in the works, it seems there are multiple market niches it could fill if they get it right- vlogger cam, travel cam, small put-it-in-weird/tight-spots cam (like a GoPro but better lens and IQ), walkaround cam, etc.