GH4 Panasonic GH5 coming in 2017

[Stab]

I didn't get the V-LOG upgrade yet. I'm shooting mostly weddings, all day at 4k 50fps, so that means I'm stuck with 8-bit.

Is V-log worth it for 8-bit footage? And with worht it I don't mean the $100 but all the hassle that comes with it? Is it, in 8-bit mode, that much better than say CineLike-D? I'll happily take a 2-stop Dynamic range advantage if it doesn't come with ghosting, artifacts, worse skin-tones, etc.

Anyone has more info on this or did some tests?
Thanks!

 

[visceralpsyche]

I didn't get the V-LOG upgrade yet. I'm shooting mostly weddings, all day at 4k 50fps, so that means I'm stuck with 8-bit.

Is V-log worth it for 8-bit footage? And with worht it I don't mean the $100 but all the hassle that comes with it? Is it, in 8-bit mode, that much better than say CineLike-D? I'll happily take a 2-stop Dynamic range advantage if it doesn't come with ghosting, artifacts, worse skin-tones, etc.

Anyone has more info on this or did some tests?
Thanks!

I shall test this afternoon and let you know! But don't forget there will be a 10bit HD firmware update coming this month which will give you 1080p50 in 10bit 4:2:2

Cheers,

Paul

 

[Stab]

I shall test this afternoon and let you know! But don't forget there will be a 10bit HD firmware update coming this month which will give you 1080p50 in 10bit 4:2:2

Cheers,

Paul

Wauw thanks Paul. I'm very curious to your test results!

I'm not that interested tbh in 1080p50 even though it is 10 bit. Well, if you can shoot in VFR with that, then I will definetely use it! But otherwise it's hard to go back from 4k I think. Also, doesn't 4k 8-bit provide better colors just because it's 4K and there is more information to begin with? Or at least when you scale it down to 1080p?

Now if only there was 4k50p at 10-bit, then I was all set for the future

 

[flinty]

Does anyone know for certain what the 4K ETC crop is is supposed to be, i have not done a 100% test but i think in 4k with ETC i am getting 960mm with my 100-400 lens at max zoom which is a 1.2 ETC crop,still less than the GH4 gave with no ETC.

 

[visceralpsyche]

Wauw thanks Paul. I'm very curious to your test results!

I'm not that interested tbh in 1080p50 even though it is 10 bit. Well, if you can shoot in VFR with that, then I will definetely use it! But otherwise it's hard to go back from 4k I think. Also, doesn't 4k 8-bit provide better colors just because it's 4K and there is more information to begin with? Or at least when you scale it down to 1080p?

Now if only there was 4k50p at 10-bit, then I was all set for the future

Ok, so I just tested both 1080p and 4K at 8bit 4:2:0 at 60p using V-Log.

Unfortunately, the banding is quite bad, just like it used to be on the GH4.

So you're left with two options for best image quality at high framerates like 50p:

1. Shoot 1080p 10bit 4:2:2 once it is released (hopefully this month), at up to whatever framerate maximum it supports, and retain the dynamic range advantage of V-Log; or

2. Shoot 4K at 8bit 4:2:0 using Cine-D or another profile if you wish, and keep the better tonal range of the linear profile.

4K50p 10bit 4:2:2 is possible, but only by recording to an external recorder. That's option number 3 I guess, but it's a significant change in cost and added complexity/batteries etc.

HTH!

Cheers,

Paul

 

[Stab]

Paul, thanks!

I expected those results. V-Log on 8-bit just isn't cutting it then.
Have you ever measured how many DR this camera has in CineLike-D as compared to V-Log? I've read you stating that V-log provides a 2 stop advantage as opposed to the other profiles. What do you think is the total (usable) DR of this camera in CineLike-D and V-log?

 

[Peter Berry]

Does anyone know for certain what the 4K ETC crop is is supposed to be, i have not done a 100% test but i think in 4k with ETC i am getting 960mm with my 100-400 lens at max zoom which is a 1.2 ETC crop,still less than the GH4 gave with no ETC.

The horiz. FOV math (5184px full sensor/3840px UHD/ETC 1:1 crop) = 1.35X, confirmed by images with and without ETC, and more cropped-in by maybe 12-15% than GH4 UHD with same lens.

 

[Thomas Smet]

Paul, thanks!

I expected those results. V-Log on 8-bit just isn't cutting it then.
Have you ever measured how many DR this camera has in CineLike-D as compared to V-Log? I've read you stating that V-log provides a 2 stop advantage as opposed to the other profiles. What do you think is the total (usable) DR of this camera in CineLike-D and V-log?

To be fair I don't really consider any log format to be acceptable with 8bits. While some flavors of C-log and S-log are better than V-log at 8bits I find too much is sacrificed and 8bits too thin to be worth the hassle. They can be used with varying levels of ok success but I consider 10bits a bare minimum for a log format if quality and grading are at any level of importance.

 

[visceralpsyche]

Paul, thanks!

I expected those results. V-Log on 8-bit just isn't cutting it then.
Have you ever measured how many DR this camera has in CineLike-D as compared to V-Log? I've read you stating that V-log provides a 2 stop advantage as opposed to the other profiles. What do you think is the total (usable) DR of this camera in CineLike-D and V-log?

I don't have the ability to accurately measure dynamic range unfortunately (such as a Stouffer Wedge chart) so I can only make comparative judgements.

Once I'm done building my Cine-D LUT I'm going to start playing with the camera for real, and see how it holds up I'm sick of staring at 100% test shots at the smallest of colour changes I'd rather be out there making lovely images!!

Cheers!

Paul

 

[visceralpsyche]

To be fair I don't really consider any log format to be acceptable with 8bits. While some flavors of C-log and S-log are better than V-log at 8bits I find too much is sacrificed and 8bits too thin to be worth the hassle. They can be used with varying levels of ok success but I consider 10bits a bare minimum for a log format if quality and grading are at any level of importance.

Agreed

 

[flinty]

The horiz. FOV math (5184px full sensor/3840px UHD/ETC 1:1 crop) = 1.35X, confirmed by images with and without ETC, and more cropped-in by maybe 12-15% than GH4 UHD with same lens.

Mm i am unclear what you are saying,but with ETC on my GH5 has less crop than my previous GH4.

 

[Peter Berry]

Mm i am unclear what you are saying,but with ETC on my GH5 has less crop than my previous GH4.

OK, the GH5 20MP sensor is 5184px wide, and the downsampled UHD remains full sensor width. The GH5's UHD ETC is a straight 1:1 pixel central crop to its 3840x2160px frame size. Divide the full sensor pixel width by the ETC frame width and you get an additional crop factor of 1.35X in ETC mode.

For the GH4 16MP sensor pixel width is 4608 and UHD is a 1:1 central crop for a 1.2X additional factor. So GH5 in UHD ETC mode should be a 0.15X tighter crop - not huge, but very noticeable.

The GH5 in ETC mode clearly has a tighter crop than the native GH4 UHD mode - by an estimated 12-15% or so in comparison clips I shot with same lens and FL, which correlates nicely with the calculations.

If you get the reverse on repeat testing, something is definitely out of sorts. Even if GH5 in C4K-ETC, and GH4 UHD, GH5 would still be tighter. Are you sure the lens zoom was locked in your comparison?

 

[flinty]

My GH4 with the 100-400mm lens gave a fraction under 1000mm in 35mm terms a lot more than 1.2x additional just uder 1.5x.THE GH4 with that lens had more reach than my SONY RX10 MK111 with clear zoom and my HC-X1000 with a 1.5x extender.

 

[Thomas Smet]

OK, the GH5 20MP sensor is 5184px wide, and the downsampled UHD remains full sensor width. The GH5's UHD ETC is a straight 1:1 pixel central crop to its 3840x2160px frame size. Divide the full sensor pixel width by the ETC frame width and you get an additional crop factor of 1.35X in ETC mode.

For the GH4 16MP sensor pixel width is 4608 and UHD is a 1:1 central crop for a 1.2X additional factor. So GH5 in UHD ETC mode should be a 0.15X tighter crop - not huge, but very noticeable.

The GH5 in ETC mode clearly has a tighter crop than the native GH4 UHD mode - by an estimated 12-15% or so in comparison clips I shot with same lens and FL, which correlates nicely with the calculations.

If you get the reverse on repeat testing, something is definitely out of sorts. Even if GH5 in C4K-ETC, and GH4 UHD, GH5 would still be tighter. Are you sure the lens zoom was locked in your comparison?

Correct. What is also interesting is if you use lenses with a SpeedBooster Ultra or the Lens Turbo 2 you get back to almost the same focal length as the GH5 without ETC. Could be an interesting move to gain a stop from the GH5 but keep the same focal length of a second GH5 without a focal reducer. The Lens Turbo gets you back to a 0.98x crop factor compared to a native GH5. Not sure why anybody would want to do this but hey it is possible if you really need to match two GH5s and want to gain a stop on one of them.

Also an advantage if you want to use a focal reducer but want as much reach as possible. Using ETC for UHD means you can gain a stop from a focal reducer but keep the lens tighter when you need the extra reach.

 

[Peter Berry]

My GH4 with the 100-400mm lens gave a fraction under 1000mm in 35mm terms a lot more than 1.2x additional just uder 1.5x.THE GH4 with that lens had more reach than my SONY RX10 MK111 with clear zoom and my HC-X1000 with a 1.5x extender.

Forget about your calculated EFL's and comparisons with other cameras than the GH5, as stated and actual lens FL's may vary greatly.

Have you actually compared film clips from the GH4 native UHD to the GH5 UHD+ETC in your hands, head-to-head with the same lens at a locked FL as I have? The maths and the practical test are in total concurrence...

 

[Peter Berry]

Correct. What is also interesting is if you use lenses with a SpeedBooster Ultra or the Lens Turbo 2 you get back to almost the same focal length as the GH5 without ETC. Could be an interesting move to gain a stop from the GH5 but keep the same focal length of a second GH5 without a focal reducer. The Lens Turbo gets you back to a 0.98x crop factor compared to a native GH5. Not sure why anybody would want to do this but hey it is possible if you really need to match two GH5s and want to gain a stop on one of them.

Also an advantage if you want to use a focal reducer but want as much reach as possible. Using ETC for UHD means you can gain a stop from a focal reducer but keep the lens tighter when you need the extra reach.

What a great idea - a full stop at near original FL is not to be sneered at! But I have only the Canon T-Smart adapter now, which needed a work-around to function on the 100-400-II, which works fine except lacking continuous OIS when filming - have to keep release button half-pressed to maintain IS. OK on the GH4.

 

[FrancisT]

Im a bit confused here by the last posts about 8bit - can you record 10bit internal in HD or 4k? or 10bit recording is exclusively external recorder?

 

[Thomas Smet]

Im a bit confused here by the last posts about 8bit - can you record 10bit internal in HD or 4k? or 10bit recording is exclusively external recorder?

4k up to 30p can be internal at 10bit. 50p and 60p can only be 8bit internal due to processing power or whatever. Externally you can do 10bit in any frame rate. So if you shoot 24p, 25p or 30p 4k you can record internally at 10bit. If you need to shoot 4k at 50p or 60p it is 8bit internal only.

 

[flinty]

Forget about your calculated EFL's and comparisons with other cameras than the GH5, as stated and actual lens FL's may vary greatly.

Have you actually compared film clips from the GH4 native UHD to the GH5 UHD+ETC in your hands, head-to-head with the same lens at a locked FL as I have? The maths and the practical test are in total concurrence...

Well i dont have a GH4 anymore but i am wrong ETC has more crop than i thought,but i am afraid using my 100-400mm lens it does not fair very well in a resolution test,a lot less than the Sony RX10 MK 111 with clear zoom on similar to my HC-X1000 WITH A 1.5X extender,even with ETC off with this lens the resolution is poor past 300mm,it is a shame as i am very happy with the colour contrast and overall look of the picture.

 

[Peter Berry]

Well i dont have a GH4 anymore but i am wrong ETC has more crop than i thought,but i am afraid using my 100-400mm lens it does not fair very well in a resolution test,a lot less than the Sony RX10 MK 111 with clear zoom on and even less than my HC-X1000 WITH A 1.5X extender,even with ETC off with this lens the resolution is poor,it is a shame as i am very happy with the colour contrast and overall look of the picture.

That's one of the reasons I got the Canon 100-400-II for a couple of hundred bucks more than the Panny - and the wicked-sharp Canon does beautifully even with a 2X TC (none available for the Panny). Here's a little 4K video of some flashy Allen's hummingbirds migrating through the Monterey Bay area recently. GH4/100-400, Standard 'style" with minor offsets and no grading, handheld in a breeze, mainly @ 400mm. AF-S and MF w/peaking. The 4-stop OIS is fantastic and almost tames my 77 year old somewhat shakey body! The GH5 4K 60p on a 30p timeline would have been perfect if I'd had it then.

https://vimeo.com/208871044