Jump to content

Panasonic GH5 - all is revealed!


Recommended Posts

Jesus. Ghosting & bad CAF... I'm never going to buy any camera at this rate ;) I'm kinda kidding, kinda not. I admit I get wrapped up in specs and issues. it leads to paralysis and then angst. I either need to buy something, shoot and forget about it, or go see a therapist.

Funny thing is bought the SmallRig cage for 50% off despite not ordering a GH5.

 

Link to post
Share on other sites
  • Replies 3.6k
  • Created
  • Last Reply

Top Posters In This Topic

Top Posters In This Topic

Popular Posts

"6K/24p Anamorphic Video Mode, while fun, is severely hampered by its 4:3 aspect ratio" UM! That's what an anamorphic mode is - 4:3 Someone let our dear friends at Cinema5D know.

Here are some 1080 JPEGS from a music video that I shot with the GH5 + SLR Magic anamorphic primes.  

A couple of quick screengrabs from a recent Jazz concert I shot. I must say I was super impressed with the GH5 on this one - not only it recorded for 1h30m straight with no issues but it did so on one

Posted Images

4 hours ago, Ki Rin said:

I'm hoping there are some settings he didn't have right for this test, because the AF performance here is pretty disappointing :(
 

 

Also a bit concerned that the camera will not fit correctly on the Zhiyun Crane (he mentioned it a bit earlier in the video). Still usable, but not ideal. 

Very odd. I got my GH5 yesterday, and as usually happens whenever I get a new camera, it rained. So all my testing was indoors, as I went through the house with lights off and on, testing AF with a 14-140 lens. When the lights were off, it was fairly dark given the weather we had been having. I experienced nothing like what he did with the AF.

I'm surprised he never tried a central zone for his AF mode as I did in my tests. Most often I was testing objects and not faces since my wife wasn't home, but I still would have used a center zone if I was only doing faces. I learned from my G85 that that's often more reliable than face tracking. 

I found the AF to be much improved over my G85 as well as my RX10iii and in the same ballpark as my now sold A6300.

On another unrelated note, I watched his entire video and was trying to figure out why his split screen tests with the tele extender mode (as it was labeled) was less zoomed than the side without it. Unless I'm misinterpreting what he did, it looked to be mislabeled. 

Link to post
Share on other sites
55 minutes ago, Hanriverprod said:

This seems better than Max's test but still slow.

He said just what I did, center area is better than face tracking. I'm surprised Max didn't try that. As you can see in this test, he never had outright failures like Max had and nor did I. I think the AF is just fine. As with any camera I've owned, there's a time for AF, a time for AF lock and a time for MF. The GH5 gives you a lot of tools for AF, you just need to learn them.

Link to post
Share on other sites
18 hours ago, AaronChicago said:

Think about the scene in Jaws where they do a dolly out - zoom in. That would look entirely different if you were dollying out while digitally scaling the image.

Wrong. It would look actually 100% the same (excluding the resolution difference as you scale).

As you can see, there is no difference in "look".

 

Link to post
Share on other sites
46 minutes ago, hmcindie said:

Wrong. It would look actually 100% the same (excluding the resolution difference as you scale).

As you can see, there is no difference in "look".

 

That is a pretty good way to replicate it but it's not the same. Even the author admits such. A 24mm lens is a 24mm lens. A 50mm lens is a 50mm lens.

Link to post
Share on other sites
1 hour ago, Ken Ross said:

He said just what I did, center area is better than face tracking. I'm surprised Max didn't try that. As you can see in this test, he never had outright failures like Max had and nor did I. I think the AF is just fine. As with any camera I've owned, there's a time for AF, a time for AF lock and a time for MF. The GH5 gives you a lot of tools for AF, you just need to learn them.

This is much better. Not perfect. But definitely usable. 
The results of AF tests I've seen seem all over the place. Some of them suggest it is terrible, and others seem pretty good.. I wonder what's going on. 

 

Link to post
Share on other sites
1 hour ago, Ki Rin said:

This is much better. Not perfect. But definitely usable. 
The results of AF tests I've seen seem all over the place. Some of them suggest it is terrible, and others seem pretty good.. I wonder what's going on. 

 

I suspect, in most cases, it's simply people using the wrong AF mode or settings. But yeah, no AF, in any camera, is perfect. 

Link to post
Share on other sites
6 hours ago, hyalinejim said:

Yes, it's there. But have you ever spotted ghosting in any of the real world low light tests completed so far?

Yes this is pretty much what I am seeing. But having tested it indoors in the evening I can tell you that with footage involving people it's even worse in V-Log. And it's not only the ghosting. There is some sort of strange banding and other artifacts going on.

I watched two previous low light V-Log tests and I could not see the same ghosting that I am seeing in my production camera. It's either a firmware change, or it's the fact that there is not much panning in the test videos.

 

Link to post
Share on other sites
2 hours ago, AaronChicago said:

That is a pretty good way to replicate it but it's not the same. Even the author admits such. A 24mm lens is a 24mm lens. A 50mm lens is a 50mm lens.

I don't want to derail the discussion about the GH5, but I hate seeing bad information out in the world, and just have to correct this.

For all intents and purposes, as long as the focal plane of the camera is in an identical place in the world between shots with different focal lengths, the effect is mathematically identical to a crop from a larger FOV to a smaller one.  In other words you can animate a crop on an image to recreate the effect of a zoom with a lens.

The only things that could possibly be different would be characteristics of the different lenses, such as pin cushion or barrel distortion throughout the zoom.  Of course there are going to be varying resolving differences depending on the technique used.  For example a digital zoom is going to reduce pixel resolution,  but an optical zoom is going to be affected by any resolving differences of the lens at different focal lengths.   That said, if you took a theoretically "perfect pinhole lens" and performed a zoom it would be identical to to a crop on an image with infinite resolution.

I've run into so many photographers who don't get this, but I think it's an important bit of information to keep in the back of your head while out shooting--  you can crop in on an image and  it will have the same basic effect as putting a longer lens on your camera.  What's neat about the GH5 is that trick doesn't seem to have much effect on the detail in the image, so it really effectively gives you two lenses which is very cool.

Just as a side note, that Jaws effect is achieved by combining a move through the environment combined with a zoom or crop in to keep the subject in the same scale in the frame.  The key to the effect is that the camera is moving and not in the same spot from the beginning to the end of the shot, so you are getting the parallax of moving the camera.  In fact even before digital, they used to do the effect using optical printers in the same way in which we can crop a shot today.

Link to post
Share on other sites
7 hours ago, Herbert Massey said:

Jesus. Ghosting & bad CAF... I'm never going to buy any camera at this rate ;) I'm kinda kidding, kinda not. I admit I get wrapped up in specs and issues. it leads to paralysis and then angst. I either need to buy something, shoot and forget about it, or go see a therapist.

Funny thing is bought the SmallRig cage for 50% off despite not ordering a GH5.

 

I feel badly for people who are torn between this or any other camera for that matter based firstly on seeing great footage and then bad footage...if this is your criteria, then no film would have ever been shot on an Alexa...my advice to @Herbert Massey would be wait for Andrew's footage/review...he knows what he's doing, will give an unvarnished viewpoint of the camera and then you can make an objective decision re wanting one or not...this thread is becoming like a typical disaster movie where all the  hysterical victims are heading for the exits first, only die off before anyone else in the subsequent stampede lol...

Link to post
Share on other sites
2 hours ago, Towd said:

I don't want to derail the discussion about the GH5, but I hate seeing bad information out in the world, and just have to correct this.

For all intents and purposes, as long as the focal plane of the camera is in an identical place in the world between shots with different focal lengths, the effect is mathematically identical to a crop from a larger FOV to a smaller one.  In other words you can animate a crop on an image to recreate the effect of a zoom with a lens.

The only things that could possibly be different would be characteristics of the different lenses, such as pin cushion or barrel distortion throughout the zoom.  Of course there are going to be varying resolving differences depending on the technique used.  For example a digital zoom is going to reduce pixel resolution,  but an optical zoom is going to be affected by any resolving differences of the lens at different focal lengths.   That said, if you took a theoretically "perfect pinhole lens" and performed a zoom it would be identical to to a crop on an image with infinite resolution.

I've run into so many photographers who don't get this, but I think it's an important bit of information to keep in the back of your head while out shooting--  you can crop in on an image and  it will have the same basic effect as putting a longer lens on your camera.  What's neat about the GH5 is that trick doesn't seem to have much effect on the detail in the image, so it really effectively gives you two lenses which is very cool.

Just as a side note, that Jaws effect is achieved by combining a move through the environment combined with a zoom or crop in to keep the subject in the same scale in the frame.  The key to the effect is that the camera is moving and not in the same spot from the beginning to the end of the shot, so you are getting the parallax of moving the camera.  In fact even before digital, they used to do the effect using optical printers in the same way in which we can crop a shot today.

To satisfy my curiosity (it has been a while since I used the variable zoom with my A7s in video), I just mounted an old crappy Canon 28-90 zoom on my A7s, mounted on a tripod and set it to 47mm then used Clearzoom (virtually lossless variable digital zoom to 2x) and zoomed in to 2x then back out, then I used the optical zoom to zoom to 90mm.

The Clearzoom actually looks better as the optical zoom is not parfocal  so is a bit out of focus at the end of the 90mm.    The optical zoom is also quite shaky (even on the tripod) while the digital zoom is smooth.

I am not going to post it as it is pretty lame and not a great lens but anyone with a Sony E mount camera with clearzoom and a zoom lens that covers 2x at least can test it.

For stills I would always prefer an optical zoom but for video in SOME circumstances, I would go with the digital zoom.

Link to post
Share on other sites
13 minutes ago, zmarty said:

The autofocus is a bit bizarre on this camera. It works pretty well for stills, but for video it's inconsistent. I went outside and shot at ISO 400 in V-Log. It seems to be having issues focusing when there are big objects close to the lens.

What lens?

What about tap to focus? Have you tested it for video?

Link to post
Share on other sites
28 minutes ago, Emanuel said:

What lens?

What about tap to focus? Have you tested it for video?

The new Leica 12-60 variable f/2.8. Shooting 4K 60fps 8 bit and 4K 30 fps 10 bit in V-Log.

I did tap on the screen and used the shutter button a few times to coax it into focusing. It was slow. The reason why I am saying it's bizarre is that I don't yet understand what circumstances makes it fail to focus. Sometimes it works fine, especially for wider shots and when the entire object is in the frame.

Am I correct in assuming the contrast based focusing is done before the image gets converted to flat V-Log?

P.S. When the image does get in focus the image quality is fantastic.

Link to post
Share on other sites
1 hour ago, noone said:

To satisfy my curiosity (it has been a while since I used the variable zoom with my A7s in video), I just mounted an old crappy Canon 28-90 zoom on my A7s, mounted on a tripod and set it to 47mm then used Clearzoom (virtually lossless variable digital zoom to 2x) and zoomed in to 2x then back out, then I used the optical zoom to zoom to 90mm.

The Clearzoom actually looks better as the optical zoom is not parfocal  so is a bit out of focus at the end of the 90mm.    The optical zoom is also quite shaky (even on the tripod) while the digital zoom is smooth.

I am not going to post it as it is pretty lame and not a great lens but anyone with a Sony E mount camera with clearzoom and a zoom lens that covers 2x at least can test it.

For stills I would always prefer an optical zoom but for video in SOME circumstances, I would go with the digital zoom.

I used clearzoom few times and most of the time it sucks balls, result is too soft/pixelated when mix the footage with another one with proper zoom on another angle in multiple cam setup during event/theater setup, ETC mode while gives you really good result without degrading sharpness except the increase in noise at lower light environment, though both wont offer you the same bokeh as proper zoom.

Link to post
Share on other sites

4K 60 fps 8-bit 150 Mbps footage shot in V-Log L yields green and magenta (compression?) artifacts when converted to Rec709 by DaVinci Resolve with their color management solution.

While the image below is a 2X magnification, it is also plainly visible in moving images at normal size. This is my dog's hair, btw :)

I wonder if the color processing for V-Log in Resolve is mostly geared towards Varicams.

I will also check 4K 30 fps 10 bit internal and external.

4k60vlog8bit.png

HOWEVER, if I color grade the same footage manually starting from the V-Log, without using automatic color management, then I can make it look better. The artifacts are still visible when pixel peeping, but not obvious during play:

4k60vlog8bit-2.png

Link to post
Share on other sites
17 minutes ago, ntblowz said:

I used clearzoom few times and most of the time it sucks balls, result is too soft/pixelated when mix the footage with another one with proper zoom on another angle in multiple cam setup during event/theater setup, ETC mode while gives you really good result without degrading sharpness except the increase in noise at lower light environment, though both wont offer you the same bokeh as proper zoom.

What camera/s was that with?      My experience with Clearzoom with my A7s and previous A7 is that it really IS virtually lossless.

Above 2x (normal digital zoom), I start seeing degradation which is why I have clearzoom set rather than digital zoom.

ETC with the GX7 is ok too but not quite as good as the Sony clearzoom and again, it is not variable so nowhere near as useful.     

As for bokeh, I think I would prefer the bokeh of a fast prime used to zoom rather than that from most zooms (at least those I have seen or used).     This is very subjective though and very much individually lens dependent. 

Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Restore formatting

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.


×
×
  • Create New...