GH5 How interested are you in a GH6 anymore?

Uh oh..no phase detect AF. That's the miss that will cause me to look harder at OM-1.

According to DPReview even with phase detect the AF on the OM-1 is not stunning.

Phase detect is not without its flaws and DFD has the potential to be better someday. We shall see if this is that day.
 
If the notion of the camera cartel is legitimate, it seems Panasonic has done the best it can with DFD. While it may not be enough for some, it seems it would be as good as it possibly could be. Likely better than any of us would have expected from DFD right now, anyway. The samples in the non private video were not perfect nor as good as the best AF out there, but they were not bad. This one has been beaten to death and no doubt, it will be a make it or break it feature for many. It is what it is. All we can do is wait for more sample footage and reviews in 5 days.
 
To be fair, the OM-1 is missing 422, but the 4K 60P 10 bit h.265 looked very good as did the footage in low light looked pretty decent. I'm not sure either camera GH6 or OM-1 floats my boat over the GH5S on a gimbal and praying the DFD holds focus on the bride. The GH6 has high bit rate 10b 422 modes which should look good but I've been through at least 3 iterations of DFD and I'm done with it. The OM-1 video doesn't seem to suffer too badly from lack of 422 though and the long GOP modes would be media friendly. There's 12 bit 444 raw to an optional recorder through a mini-HDMI that doesn't excite me, but I guess for someone with the patience for Rube Goldberg, there's that. Both cameras have gone under the cripple knife in my opinion, but you're right; I haven't been fair to the GH6 but DFD has me gaggling with hate.
 
So, it sounds like they are doing a bit of digital magic like what is being used on smartphones. Suppressing more noise via even better processing, and extending dynamic range via digital trickery. If it works really well I am for it, but I am a little skeptical... and I would like to know what the dynamic range of the sensor is without any such trickery. If it actually resolved 13 stops and looks indistinguishable from other cameras resolving that, it's good enough for me to be an A-camera.

Yikes, this could be inconvenient. But I suppose no more inconvenient than shooting on a full frame sensor, it may just indicate that higher stopping power ND filters will be needed if outdoors in very bright sunlight and wanting to shoot at F/2.8 or wider.

Doesn't Alexa do something similar with two gain circuits to help extend dynamic range?

It isn't exactly magic and a common way photography creates HDR. Totally valid way to extend dynamic range in a camera. Its just not always possible with video because of howe fast the sensor has to readout for just normal video. There was nothing left to be able to pull off something like this until now.

Its almost always been a given that log will likely need ND filters. Even without shooting log I'm not sure how anybody gets away with shooting video without some level of ND filtering. 2000 ISO is a good sign we may be finally getting full V log on the GH6 and can kiss V-log L goodbye forever. We hope.
 
The idea of shooting ProRes 422 HQ internally to CFExpress is pretty cool. You can get a 4TB card for $1800 which sounds like a lot, but compared against an external atomos setup and SSD, not really - especially since the camera stays super small light and nimble.
 
Doesn't Alexa do something similar with two gain circuits to help extend dynamic range? ...

Yes but Alexa had a lot more processing power for the frame interpolation and even that didn't always work (according to the old JB posts). Presumably, a decade later, one could squeeze enough of it inside a relatively tiny hybrid body.

PS. Side note - this looks to be a totally new, previously unannounced, sensor. Which may explain the delay in the GH-6 introduction. Whoever made the sensor had to do it to Panasonic's specs. It's also likely that Panasonic revised the requirements somewhere along the way. It's also very likely that the company wasn't sure when the sensor was going to be available and was forced into the GH5MKII as the interim solution. Ah, the best laid plans.

The 100 MPX is likely off a pixel shift rather than off a Quad Bayer singe frame but the 4X in-camera composite seems to be possible these days. Once again, one needs to interpolate the various frames to avoid digital "smudges" and, for that, one needs some beefy processing but it can be done. OM-1, as we have discussed earlier, does it as well. Maybe GH-6 will do it even better.
 
And we’ve got a full spec 8 page PDF leak on scribD:

https://photorumors.com/2022/02/16/...camera-full-specifications-leaked-online/amp/

looks like the OLED EVF is same as GH5 at 3.68 million dots (so not as good as S1H 5.7 million) and the LCD screen seems it is slightly smaller than GH5, coming in at 3.0” instead of 3.2”. However it’s an increase in pixels up from 1.6 million to 1.84 million. So perhaps they did this to make the screen sharper and maybe brighter?

anyway, lots more specs there… have at it…
 
Also looks like the battery did change after all, it’s now using BLK22 from the S5 instead of BLF19 from previous GH models.

Battery life also appears to be around a 33% reduction against GH5 battery life times.
 
Doesn't Alexa do something similar with two gain circuits to help extend dynamic range?

Right, same as DGO on canon, I believe. But I believe that those implementations are combinations of sequential readouts. I'm not sure, but I got the impression that the new Olympus and gh6 might finally be doing what was rumored for the a7siii (and did not come to pass). They might take advantage of the quad Bayer pattern to utilize alternating high-gain and low-gain photosites within the same 2x2 patch that becomes a single pixel. So, the 2 different exposures would be registered simultaneously and not sequentially. But I could be wrong. Somebody tell me
 
I'm not sure, but I got the impression that the new Olympus and gh6 might finally be doing what was rumored for the a7siii (and did not come to pass). They might take advantage of the quad Bayer pattern to utilize alternating high-gain and low-gain photosites within the same 2x2 patch that becomes a single pixel. So, the 2 different exposures would be registered simultaneously and not sequentially. But I could be wrong. Somebody tell me

I’m guessing too, but it’s my understanding that (a) the GH6 doesn’t have a quad-Bayer sensor, and (b) Panasonic is reading out the whole sensor at one ISO, say, 800, to preserve the highlights, and then doing another different exposure in quick succession at, say ISO 4000, to preserve the shadows. Then it’s doing an in-camera composite in real-time and moving on to the next frame.

This is only possible because of the phenomenal readout speed of the sensor (and stacks of processing power, of course), and even then it’s limited to 60p. Who knows what it does to the appearance of motion blur when the subject moves between the two exposures. Probably bad things to the critical observer.

Arri does something quite different (and less ambitious but with no chance of compositing errors). Those cameras read out the same exposure twice at different gain, for the minimal benefit of operating the ADCs in their highest performance window for both shadows and highlights. In the high-gain readout the highlights would be blown, but that doesn’t matter because they’re preserved in the second low-gain readout.
 
Well, yeah...because then it's not AF anymore. lol

The camera isn't looking for focus thus resulting in no pulsing. You only have push AF after that which is technically focusing automatically but it's just once and you have to manually operate it.

And get used to that zooming test...don't be happy with something just because you can't see it with your brain/particular set of eyes without zooming in.

You deserve better AF in a Panasonic camera.

When you half way push down the shutter or tap the screen it is AF. It is not continuous AF. So you can shoot that way or manual to stop background pulsing. I normally don't use continuous AF. I hope the GH6 fixes or improves that. But the whole package on the gH6 looks really good. Continuous AF will not be a deal breaker.
 
Bunch more leaks on the GH6 now. Looks like a monster with a fan for cooling. Internal ProRes 422 HQ to card at 1.9 Gb/s. Also physically quite a bit larger than the GH5. Nothing about this looks cheap. I’m guessing three grand?

Under $2500? Maybe with the 12-60 lens around $2700 or less. . Maybe they will match the OM-1 price.
This will be a camera that will last for years.
 
This will be a camera that will last for years.

Sure, though I feel the same way about my GH5S with its internal DCI 4K 4:2:2 10-bit, true 24p, codec and frame-rate flexibility, Raw output to Blackmagic or Atomos recorder, reasonable timecode support and audio generally, good low-light noise, solid exposure- and focus-assist tools, no overheating, fast readout, good battery life by mirrorless standards (though still annoying; by the way, the GH6 looks worse for battery life although maybe they’ve got another powering option in the works?), etc.

The GH5S still strikes me as a camera that concentrates on getting the nitty-gritty details right instead of just doing flashy headline specs. That’s the sort of product I like. Can’t stand flakey half-baked products like nearly every Canon camera. Clearly the market at large thinks otherwise.

Of course it’s of no use to the full-frame fetishists, but that explains its low price, which is what makes it possible for me to own such a capable camera. I’m not a pro and can’t drop four-grand here there and everywhere.

(I see you’re selling your GH5S so evidently you disagree on some level. That package you’re selling is a true bargain for someone, by the way. Good luck with the sale.)
 
Back
Top