Andrew Reid

They are not STILL doing that are they? FFS

Can confirm this is a bug on mine too. Did you ever find a fix? It happens to me no matter what video mode or settings are used. Reminds me of the similar issue on the Nikon Z6 which has a similar sensor!

The ALL-I codec on this camera is REALLY good. Maintains a LOT more in the shadows than the Long GOP IPB mode. It is 400Mbit so quite big files but a lot smaller than RAW at 1800Mbit. This is one of those occasions where 400Mbit is worth it... (Shadows pushed up approx. 3 stops in Resolve) 4K ALL-I 4K Long GOP IPB

The S1 battery is absolutely enormous. Here's me wondering how the Sigma Fp has no problems with it's GH2 battery from 2011! It even powers an SSD while doing 12bit 4K RAW

Yes the autofocus is pretty, erm, shall we say... Hybrid. Somewhere between contrast detect AF and phase. It hunts. Not sure why. It's extremely unreliable in video mode. 32-64mm zoom has internal stepping motor, silent, fast, but the camera does not make the most of it. The X-T4 is far better for AF. Some lenses like the 63mm f2.8 are incredibly bad on the GFX cameras and it's not much better on the GFX 100 with phase-detect AF on sensor. I am hoping for a firmware update that dramatically improves it as I can see no good reason why it has to be so bad! Reminds me of the X-Pro 1!! Apart from that, it's my favourite all-round camera right now. Love it. You can put a £20 lens on it and it still looks like medium format. So cinematic.

I get better performance with OpenCL than Metal, so it's the other way round. Running Mojave. Hackintosh is working flawlessly. Resolve 16.2 Can't figure it out at all yet.

I seem to remember the NX1 did limit controls when connected to a monitor. The firmware updates addressed it only partially. It is just a video feed via HDMI, it doesn't matter what monitor you use. To be honest I would use the NX1 without one as it has a great codec and very nice OLED on the back.

This is also interesting. Thanks.

OK, I think the best solution is to buy a new battery (at least temporarily and return it to Amazon if it doesn't fix the issue). This will illuminate the possibility that it's just a faulty battery. If it is proven that the problem happens with the new battery too, then we have a situation and I'll notify Panasonic. It may be that the cameras are very marginal in their peak current requirements, so any degradation with the battery as it ages causes an issue. This isn't good either and Panasonic should investigate like Leica are doing with the SL2.

A question for Resolve users... Is it still possible to display the fps in the viewer? I used to have this in the Canon Magic Lantern RAW days but it's vanished and can't find the option. I have an RTX 2080 Ti in my PC and will see what performance is like on that. Nvidia cards no longer work with the latest Mac OS verisions - so that's why there's a Vega 64 in my Hackintosh. It has excellent compute performance even though it's not as powerful for games as an RTX 2080. Will ask Blackmagic support what the problem might be. The crushed blacks in 10bit H.265 is also an issue with my Fuji footage in Resolve, and it badly needs a solution... Will see if this happens in Windows or if it is an Apple thing. By the way I think the reason the 1DX3 doesn't allow autofocus in 4K/60p is the battery can't handle it... Peak power requirements. Driving that sensor in turbo mode, and image processor at 100% utilisation, along with autofocus just doesn't sound like something you'd be able to do on a battery like the 1DX3 has.

I have edited 4K DNG RAW on much lesser systems without a hitch. Back to the store for the 1DX3? Looks like it.

Intel Core i9 9900K 32GB 3000mhz DDR4 Vega 64 8GB Mac OS 10.14 ** Seems to run better in OpenCL mode compared to Metal. Well, it did for about 5 seconds and now it's back to being very choppy. Playback resolution full / half / quarter doesn't make a difference. Project drive is an Nvme SSD with 3500MB/s read speed.

My god the playback performance of 5.5K Canon RAW in Resolve is shoddy. Can barely get above 5fps.

One of the best features of the 1D X Mark III for me on the first day I've had it, has been the full frame 4K/60p 10bit Canon LOG 4:2:2 with DIS and a Z-finder on the back screen for my "EVF". Yes, you lose the big AF selling point, so it's not perfect for shooting everything, but I find it very nice to use for handheld cinema in this mode. Slow-mo when you want it, adds to stability handheld. Loupe adds to stability, extra point of contact with the body. Then DIS rounds off any rough edges - gets rid of the micro jitter. It works well. In fact it works so well, I forgot it didn't have IBIS for a moment. There is a small 10% crop from full frame with DIS but no loss in image quality. Enhanced DIS has a loss of image quality and is more useful as an Ex-Tele mode than an extra stabiliser mode. Good if you need the extra reach. So the AF is there when you need it for 24p. And the DIS is there if you need it. And the slow-mo. If not all at once, then at least close to it. I am warming to it.

Please give us some more details sir! We need specific info... 1. Lens in use when power cut 2. AF mode in use, or was it manual focus? 3. Ambient temperture - roughly 4. Video mode - 4K/60p? or 24p? 5. Age of battery and did the problem only start occuring under same conditions as battery aged? Please help me reproduce the error with my S1. And I will pass the results onto Panasonic in the UK who will feed it back to Japan.

Not quite my point. With the original film negative, you have the proof (of untouched "real" scenes). Doesn't matter if the edit is digital or not. If somebody questions if what they see is real, unaltered, the proof is there in the background, on a reel. All that is a form of manipulation and fantasy yes, sure... But computational photography goes through a phase-change into something else entirely different. One example - deep fakes. In the future we'll have AI driven cameras that can film in real time a completely fake scene, superimposing specific people and faces over anybody in the scene, in a way that looks so real you can't tell. Another example - news footage. In the future, AI driven cameras will be able to film statements and interviews which are completely made up from a script. You will never trust what you see again from a digital camera. Give it 10-20 years to reach this point, and suddenly the analogue film negative becomes incredibly valuable as a source of "truth".

Yes, good plan! No conclusions until the fat lady sings. But it ain't looking great, is it? You spend in my case £5800 and find out the rolling shutter in 24p is worse than Sony's low-end mirrorless camera? These things should not happen. Not at that price. Nope. True, no other camera offers 5.5K 60p RAW. However the image quality must be evaluated to see if the massive file sizes make it worth shooting. Some cameras take a short-cut with the sensor output in RAW. Optimally, RAW is a 1:1 readout - the RAW pixel data, written directly to card. If the Nikon Z6 for example had an optimal ProRes RAW solution with the Ninja V it would be 6K. The sensor is 6K. Ask yourself how they get to 4K RAW. Look closely at the pixel quality vs 10bit 422 V-LOG on the S1H. It ain't a pretty picture. It is very special to have a cinematic, film like image, but then there is the question of getting the shot in the real world. What hinders you, etc. Superb image quality is nothing if you miss a shot, or the rolling shutter messes up the shot. Taking away AF in the high resolution full frame 60P modes really annoys me, but I am not going to unfairly throw the camera under the bus for it... The AF is still in a different league to the S1H, which is this camera's closest competitor for video quality.

S1H is the best, since it's dual native ISO. It is extremely clean. ISO 3200 looks like 200. Goes past 12,800 easy. Put an F0.95 lens on there and there's nothing it can't do. You even get a boost by downsampling the 6K to 2.8K in post. Reduces the noise even further. Z6 and S1 are good but S1H better.

Fear not. The DIS is here to save the day. Very good and only 10% crop in full frame with it turned on. The extended mode is a BIG crop, and softer, but useful as a 2x crop mode in S35 for extra focal lengths from one prime. You can pretty much do a lot with a 35mm on there. DIS looks like IBIS it's that good. Although haven't done anything like an in-depth test with it yet, and just locked down shots. With IBIS you can't really pan anyway, it looks crap. A6500 and NX1 were around 30ms. But these measurements all seem to change depending on who you ask. Slashcam have 28ms for the 1DX3 24p. I am inclined to believe that one. The old 5D2 1080p in 24p was around 24ms.

It has arrived. Found another quirk. Max 1/4000 shutter in video mode. No shooting wide open outdoors without NDs. It is rare for an electronic shutter to have such a low maximum speed. Usually it is 1/8000 or even 1/16000. Ideally I don't want to use NDs for the 180 degree look and prefer to add that in post with the Resolve motion blur tools. Oh well can't have everything I suppose. So far the gut feeling is the GFX 100 and X-T4 are better for video, but I've yet to see the images from the 1DX3. The CFexpress card is coming tomorrow.

The digital IS and careful camera work helps. Otherwise there is nothing you can do but lose AF and shoot 60p, or lose full frame and shoot 60p with AF. It is a clear indication of what the EOS R5 8K will be like. Massive RS. AF quirks. Huge file sizes.

In the future digital will be so far gone... so far manipulated with computational photography... that documentary filmmakers will use film, to prove what they shot was real. Digital cameras are going to be reality distortion fields. The archival of so much digital data is still an ever present challenge for such large file sizes. If you shot everything at 2400Mbit/s you should add a zero to the cost of your camera.

It may be DPAF related yes. But seems odd that DPAF works fine in crop mode at 4K/60. Why not in full frame? Even if they have to disabled it so that it only covers the centre 80% of the chip, so less DPAF data to process, it would have been better than disabling it outright or forcing us into the slow lane.

I am pretty shocked by the rolling shutter behaviour reported by Slashcam. 16ms in 4K/60p 32ms in 4K/24p What a mess Canon! 32ms is completely unacceptable. It's worse than an A6500.

It's not going to be treated like film on a shoot, because with digital you can quickly offload a reel to affordable storage via USB. However it is treated like film in the way you archive 8K RAW material. After a few years you will need a building the size of the Paramount studios lot to put the hard drives