androidlad

3 minutes ago, David Bowgett said:

Hey now, 300nm used to be a cutting-edge process... all the way back in 1996 or so, but it was certainly cutting-edge back then!

(To be fair, 90nm processes aren't exactly new either - they've been around since roughly 2004 - but still, quite the difference)

For a 36x24mm FF sensor, 300nm vs 90nm process really make very little if any difference to image quality. There are more important contributors.

CMOS sensors are not CPU, there's no constant need to shrink the die.

This 63MP FF sensor from Canon (35MM63MXSCD, FSI with 300nm process) is a response to Sony's 61MP IMX455 (BSI & DTI with 90nm process), which recently entered mass production and will be sold to Nikon and Pentax.

All readout drive modes detailed here:

DJI Mavic 2 Zoom:

D-Cinelike profile -> GH5 Cine-D PRE -> GHa Daylight Linear -> Contrast adjustments

1 hour ago, Desk said:

Hi,

Thanks for this. I'll take that support non-native, as an add-on, but is this only on the Lab?  The motors wouldn't work with the Crane 2?

Desk

Crane 2 has its own focus motor, while Crane 3 Lab/Weebill Lab have their own motors for both focus and zoom, at the same time.

4 hours ago, Desk said:

Does anyone know if there's a gimbal which allows follow focus and zoom control support for the XT3?

Desk

Zhiyun Weebill Lab/Crane 3 Lab support it, by providing two motors for zoom and focus control.

Tilta is showing a cage designed for X-T3 (the one on the right).

41 minutes ago, KnightsFan said:

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!

Higher frame rate involves a lot of vertical cropping and/or dropping readout bitdepth.

It's the same "Movie Digital IS" on EOS R.

The Japanese text says:

"When shooting movie, the built-in gyro sensor provides 5-axis image stabilisation with a compatible lens."

5 hours ago, DanielVranic said:

Please post photo examples. 

And the reason I ask is I just shot an entire short at an open field/horse barn and graded the entire project using GHALEX and there is not one trace of banding. Not even in the outdoor shots going directly into a blue sky, all the way to a way underexposed inside the barn office and even then it was just fine. 

Even before on the old thread, I was unable to make my Fuji's footage band at all. I am coming from shooting on Sony's.. so I know what banding is! Is it entirely possible that you have a bad copy? I have googled and googled and you are the only person I have found that has issues with banding on a Fuji system. Every search I have done for it has brought up issues with photo and the ES, but nothing for the 10bit video. 

I've seen a couple complaints regarding F-log banding/blocking on DPReview forum. Most people don't care though, not everybody is an image quality purist.

Doesn't look like a hardware problem so bad copy is very unlikely.

11 minutes ago, John Roque said:

I don't know if this has been addressed here but I'm finding some issues with flog with the xt3. Has anyone experienced these "pink splotches" before? These were shot in 10bit H.265. I'm finding that the HLG doesn't suffer from this issue. I placed a lut to emphasize the issue. Any help would be appreciated!

Yes this issue was discussed extensively on this forum, see my post comparing F-log and HLG. There's something wrong with Fuji's F-log implementation, could be flaws in the encoding parameter. The solution is to shoot HLG instead, it's completely clean and in my opinion also offers better colourimetry than F-Gamut.

Please report this issue to Fujifilm by sending these photos along with a description http://www.fujifilm.com/contact/

We have to get Fujifilm's attention for a possible fix.

Observation from preliminary testing GHa on X-T3:

With F-log, even at 10bit 400Mbps, applying two LUTs (F-log to V-log then GHa linear) produces noticeable banding in the footage. The F-log banding issue was discussed in other threads but using two LUTs really accentuate these artifacts.

HLG works amazingly well, the GH5 HLG preparation LUT is a perfect match (BT.2100 HLG is a standardised colour space so shouldn't come as a surprise). Again as discussed in another thread, HLG on X-T3 removes the magenta hue. Shades of blue have that Alexa signature cyan tint.

https://www.lindy.co.uk/

Lindy makes some of the best cables and they are reasonably priced.

5 minutes ago, tweak said:

Exactly what I'm saying...

"probably going to be", i.e. you actually have no idea.

It is the A7 III/Mavo LF sensor IMX410, without PDAF mask of course.

The 8K one is IMX455.

18 minutes ago, zerocool22 said:

Hoping on BRAW or something similar instead. I dont have the budgets to store 6K or 8K RAW files on my disks.  

They made it very clear that it's Cineform RAW, a wavelet compressed highly efficient RAW format.

Great video. I submitted a ticket to Adobe to address this issue, maybe you could also submit a ticket with a link to this video?

Your English is great and your accent is hot.

5 minutes ago, thebrothersthre3 said:

Yeah I am hoping they go the 4k sensor low MP route. 

No they won't.

A7S III is still primarily a stills camera, with cutting edge video features.

https://lilliputdirect.com/lilliput-4k-ultra-hd-monitors?product_id=781

Now the firmwares for their cameras will never come out of beta...

"Fix"?

It's a nice colour contrast, I don't know what the mood of the scene or the story is, but to me it's not really a big problem.

If it's too blue, just do a secondary grade and drop blue saturation.

14 minutes ago, Andrew Reid said:

Similar architecture to the Sony-Fuji 100MP medium format sensor?

Personally I don't need 100MP stills but the technology is interesting as is all forward progress, it leads to an interesting future and is better than staying the same, otherwise we'd all be shooting 720p and 8MP stills

Almost completely different architecture with this one being more advanced. 3.76um vs 2.91um, slim DTI, way more ADCs and they are now colour-aware (meaning significant readout noise reduction and QE boost).

There are various 2x2, 3x3, 6x6 binning modes optimised for speed, quality or a compromise between the two.

One of the 2x2 binning modes produces a 25MP image that even surpasses A7 III.

3 minutes ago, Eric Calabros said:

Good luck demosaicing this puzzle mess and get a clean 4k as full read out. 

Ha, this IS the demosaicing, all done on-chip at hardware level. You get super clean and sharp 4K 4:4:4 straight away.

4K RGB 4:4:4 readout mode:

Until we have variable curvature sensor, curved sensor at this size will largely be limited to industrial/scientific use as one curvature has to match a specific focal length.

X-H1 F-Log -> F-log/F-Gamut to V-log/V-Gamut LUT from LUTCalc -> GHa Daylight: