GH5 Game Over, Panasonic

I think something we see a lot from early adopters and people with cams from different brands is super unhealthy. It reminds me of the "console wars" teenagers still argue about. PS4 vs Xbox vs PC. It's pretty silly as an adult... yet here so many people are (not referring to anyone here, just a very general observation on the market), feeling the need to either be all in and defend Panasonic or BMD for whatever reason. Part of that is probably the way other people jump on any perceived faults or admissions to bolster their side, as if there is actually a side to take here instead of just objective realities. None of us are defined by the cameras we choose, they are just tools, all of which have imperfections.

Ok, so this is interesting. The pocket 4k may have issues with wide angle lenses, which may be because the differing sensor stack size is very subtly warping the way the light from the wide angle glass is being pulled in. Is this only happening on adapted lenses? If so, a dedicated speedbooster would "fix" this issue. Not because it's a flaw but perhaps just because it's a different design than the sensor stacks by Panny/Oly. And depending on the user, this may be a non issue to them as it might be subtle enough or reserved for those extreme wides that some people might not use as much, thus meaning they could continue using the "unoptimzed" speedbooster for most things?

Am I understanding that correctly?

If so, the way I see it is we are waiting on an official speedbooster for the Pocket 4K sensor stack size and the bonus is, the current one works well for most things.

Now, I am also confused as some mentioned native MFT exhibiting these issues as well. Well, I guess this makes sense - the huge market share is Panny/Oly, so Laowa for example would have designed their UWA lens to perform optimally on these cameras and their associated sensor stack size. And I suppose you would notice the difference here much more so than standard focal lengths. Well, if so, the way I see it is - that's ok.

Perhaps one compromise with the BMD Pocket 4K is we don't use native UWA lenses on that camera unless we are willing to purchase and use a yet developed lens correction plug in that someone will probably make and live with the results. Or, we just use UWA on adapted lenses only ONCE the official speedbooster dialed to the proper sensor stack width is released.

Unfortunate, sure, but really this all makes sense. It is too bad BMD didn't factor this in, if it truly is resultant of just the sensor stack, and try to match what the existing mass market share currently is (sure, there's no official standard - but unofficially...).

BMD doesn't use OLPF's. This accounts for the majority of the difference in stack thickness. The distortion can be seen with both native and adapted lenses, but it's only present on some very wide lenses at very wide apertures. A SpeedBooster could be made to correct, but there's not much that can be done for native glass, other than possibly the addition of an aftermarket OLPF, which are available for other BMD cameras.
 
Ultra wide angle lenses tend to have noticeable geometric distortion, approaching that of fish-eye lenses. It's not surprising to hear of focus field distortion as well, particularly at the fastest apertures. But with their inherently broad depth of field, shouldn't it be easy to find a sweet spot where both center and edges of the cropped lens circle are in focus?
If you step the aperture down a tad, yes. My 14mm has slight distortion at T/3.1 and is completely gone by T/4.
 
BMD doesn't use OLPF's. This accounts for the majority of the difference in stack thickness. The distortion can be seen with both native and adapted lenses, but it's only present on some very wide lenses at very wide apertures. A SpeedBooster could be made to correct, but there's not much that can be done for native glass, other than possibly the addition of an aftermarket OLPF, which are available for other BMD cameras.

The GH5 doesn’t have an OLPF either.
 
Couldn't they just move the filter away from the sensor to achive this ?

Actually that’s exactly how I understood the D60’s sensor was designed. They were not extra thick, rather spaced further from the sensor to minimize the appearance of dust

Was told this by a Canon representative while beta testing the 1Ds, but not by an engineer, so take with salt.
 
There's actually someone who has tested several lenses on the P4K. His results:

Have issues
Laowa 7.5mm f/2
Samyang 12mm f/2
Tokina 11-16mm f/2.8

Do not have issues
Panasonic 20mm f/1.7
Panasonic 25mm f1/7
Mitakon 25mm f/1.4
7Artisans 25mm f/1.8
Panasonic 42.5mm f/1.7
Sigma 18-35mm f/1.8

And some like 11-16 were used with the speed booster. As brawley already mentioned but it’s worth mentioning again you’re adding more glass, why isn’t the speed booster getting heat. The samyang and Laowa lenses are cheap optic lenses I wouldn’t use as an example for anything as they are so soft wide open its part of their natural characteristic.
 
For anyone else who is curious, thread is here: http://www.bmcuser.com/showthread.php?22425-Soft-edges-with-wide-angle-lenses-on-the-BMP4K/page10

Some of the images are so extreme I'm inclined to think it's flawed testing... but really puts a damper on my intended use if these things can't shoot wide!

After everything is said and forgotten, I think it will be the red channel clipping issue that is the undoing of P4K. You can't have the exceptional dynamic range promised if a color channel is clipping prematurely, although that flaw has the better chance of being fixed through a firmware update. What I think has happened, is BMD which was accepting orders 6 months ahead of the release, generating all the buzz, was so intent on beating the delivery deadline which they did, but the first batch were the beta testers, limited quantities. I bet it will be a long wait for the rest of the pre-orders to get fulfilled, making time to correct problems that crop up.
 
I suspect that the red channel clipping is not actually happening in camera, but rather in the v4 color science. The reason I suspect this is that if I grade with RCM, vector scope red is significantly hot. If I grade manually, it is within 3% of target.

Now, I've yet to shoot night time traffic (funny how rarely paying customers want crappy YouTube videos of sodium lights and red taillights) so I can't verify my theory until i go out of my way to shoot something ridiculously saturated at night.
 
Now, I've yet to shoot night time traffic (funny how rarely paying customers want crappy YouTube videos of sodium lights and red taillights) so I can't verify my theory until i go out of my way to shoot something ridiculously saturated at night.

Agreed but the image below points to nuclear reds in the sipping straws. So rather than using RGB dye filtration to achieve the desired color science, BMD have simply lutted the output; compromised DR. Some comments suggest color that shifts with ISO as well.

http://www.bmcuser.com/showthread.php?22445-Pocket-4K-red-clipping&p=256104&viewfull=1#post256104
 
And this is the part that I really wait for and why I haven't sprung yet - the post hype reality checks. While Panasonic S1 will be a pre-order / day 1 purchase, BMD even at the price point is a few month waiting game for my tastes. The camera puts out such great images but it is not without flaws, and I find it essential for me personally to get them vetted via the community and early adopters for the first few months to make an informed decision.

BTW not being negative, I'm still really keen on this piece of gear, I just think I'll be a hold out until January after the dust settles and I know exactly what I would be getting myself into. I'm mostly in editing mode the next couple months, anyway, and I have plenty of IQ to keep me going for the remaining ~7-10 shoot days this calendar year.
 
...Some comments suggest color that shifts with ISO as well.
This one I can address directly as I've tested. Color shifts across the ISO range on ALL cameras, some more than others. The P4K shifts far less than the GH5, and less than any camera I've personally owned. (Less than 5% average.) I remember the GH5S as also being very good in this regard, but I don't think I methodically tested as I had only rented it for short time.
 
The S1 might be the really exiting release, all Panasonic needs to achieve is all GH5 functionalities with the automatic rise in dynamic range and light gathering coming from a FF sensor and its already a winner. Anything more is great but as is would already tilt the scale price it under 4K and it will sell like hotcakes.
I know a lot of people (I am one of them) who got the GH5 as a B or C Cam that are more or less shooting exclusively on them whilE the "real cameras"they own only come out for special occasions. Specifically people who shoot docs. If anything the smaller sensor has been looked at as a limitation- once there is a FF solution then it becomes a no brainer. Long zooms on the GH5/s and FF Primes on the S1 both with IBIS 10Bit 422 and similar color science will be hard to argue with.
The Pocket as always with blackmagic will become a useful camera for a specific type of user maybe sometime after January or so.


QUOTE=filmguy123;1986771465]And this is the part that I really wait for and why I haven't sprung yet - the post hype reality checks. While Panasonic S1 will be a pre-order / day 1 purchase, BMD even at the price point is a few month waiting game for my tastes. The camera puts out such great images but it is not without flaws, and I find it essential for me personally to get them vetted via the community and early adopters for the first few months to make an informed decision.

BTW not being negative, I'm still really keen on this piece of gear, I just think I'll be a hold out until January after the dust settles and I know exactly what I would be getting myself into. I'm mostly in editing mode the next couple months, anyway, and I have plenty of IQ to keep me going for the remaining ~7-10 shoot days this calendar year.[/QUOTE]
 
This one I can address directly as I've tested. Color shifts across the ISO range on ALL cameras, some more than others. The P4K shifts far less than the GH5, and less than any camera I've personally owned. (Less than 5% average.) I remember the GH5S as also being very good in this regard, but I don't think I methodically tested as I had only rented it for short time.

That has not been my observation with either GH5 or GH5s but for the former I'm usually working in the range of ISO 400-1600 and the latter, ISO 320-6400. If the ISO's are higher than that color shifts would not be the top worry.
 
the s1 might be the really exiting release, all panasonic needs to achieve is all gh5 functionalities with the automatic rise in dynamic range and light gathering coming from a ff sensor and its already a winner. Anything more is great but as is would already tilt the scale price it under 4k and it will sell like hotcakes.

This is all true but it also has to be remembered GH5s low light is already the equal of A7SII at equivalent DOF up to about ISO 12800. And that's particularly significant for me because my wedding videography needs more DOF than would be useful for cinema. In other words, to keep the same DOF with FF would mean stopped down 1 f-stop from MFT.
 
Absolutely I also enjoy the benefits of a smaller sensor for specific typeof tasks. I am excited about using a GH5s with a 12-100mm in Tandem with a S1 with an Otus 55mm 1.4 prime.
Panasonic has done a great job of not just hitting specific technical Benchmarks they are also aware of user experience its how the camera handles which is equally important ir every day shooting.
The pocket has great features BUT unless its on sticks or some kind of rig its basically inoperable at least for my needs.


This is all true but it also has to be remembered GH5s low light is already the equal of A7SII at equivalent DOF up to about ISO 12800. And that's particularly significant for me because my wedding videography needs more DOF than would be useful for cinema. In other words, to keep the same DOF with FF would mean stopped down 1 f-stop from MFT.
 
This is all true but it also has to be remembered GH5s low light is already the equal of A7SII at equivalent DOF up to about ISO 12800. And that's particularly significant for me because my wedding videography needs more DOF than would be useful for cinema. In other words, to keep the same DOF with FF would mean stopped down 1 f-stop from MFT.

Yeah the entire point of the Panasonic FF is to have the sensitivity of the GH5S without sacrificing the stills resolution. It’s about the best of both worlds. Have your cake and eat it too. Very good video quality at higher ISO levels and very good large resolution stills. Pretty sure it was designed to appeal more to stills shooters who actually have a need for larger than 20mp stills and don’t want crappy HD video like those Canon and Nikon things called cameras.
 
and it should get more dynamic range

Yeah the entire point of the Panasonic FF is to have the sensitivity of the GH5S without sacrificing the stills resolution. It’s about the best of both worlds. Have your cake and eat it too. Very good video quality at higher ISO levels and very good large resolution stills. Pretty sure it was designed to appeal more to stills shooters who actually have a need for larger than 20mp stills and don’t want crappy HD video like those Canon and Nikon things called cameras.
 
Back
Top