Panasonic GH5 - all is revealed!

[Phil A]

https://www.slashcam.de/artikel/Test/Panasonic-GH5---8-Bit--10-Bit--4-2-2--V-Log-L---alles-OK--.html

Next part of the Slashcam review, testing 8bit vs 10bit and V-Log L.

Summary:

• They say there is some issues how NLEs interpret the white point and black point of the footage
  In Premiere Pro CC neither 8bit nor 10bit fit to the whitepaper from Panasonic.
  In Media Player Classic - Home Cinema the files look exactly the same
  In Davinci Resolve under OSX the files are different again, even in comparison to Premiere Pro on Windows
  In FinalCut X the files are practically identical but they can't really judge the white and black point because it gives % instead of values
• Latitude for grading with V Log is, as expected, better with the 10bit material than the 8bit. A lot less posterization when bending the image.
• They think both Adobe and Blackmagic have to fix compatibility within their software to really make the most out of the footage and be able to actually judge the 10bit V Log of the GH5

[hyalinejim]

Martin Wallgren's short ad for a Danish barber, shot in V Log internal, available light:

The focus is off on a lot of shots but the video does show that internal V-Log is very usable indeed.

18 minutes ago, Phil A said:

https://www.slashcam.de/artikel/Test/Panasonic-GH5---8-Bit--10-Bit--4-2-2--V-Log-L---alles-OK--.html

Next part of the Slashcam review, testing 8bit vs 10bit and V-Log L.

Thanks for this. Google translate is funny:

Quote

if too little gradations are stored here, visible color frights occur

I'm using that one!

Here's their comparison of 8bit (left) v 10bit (right), contrast stress test:

[Fritz Pierre]

10 hours ago, webrunner5 said:

I agree it has some pretty good, maybe great resolution, but... Like you sort of hinted at, maybe too good. I think most of us on here are looking for more of a Cine look, I am afraid the GH5 is just going to be a GH4 on Steroids! Still videoish, is that a word? Too much, later Panasonic video camera, looking.

But I have to admit I guess I would rather have too much data to play with then too much less. We have seen rays of hope on it from a few people. I am sure it is going to sell like Hotcakes, but is it our brand, style of Hotcakes on here?

I have to admit, it really is Not on my Radar at all. Nada. I can think of 3 or 4 older semi pro, pro used cameras that I would buy before it in a heartbeat for 2 to 3 grand or less.

And yes, the FS700 IS on that list. Someone got a deal of a lifetime with is auction.  That is a Video camera! http://www.ebay.com/itm/Sony-NEX-FS700U-w-4K-Upgrade-and-Zacuto-Accessories-/322441953597?hash=item4b1309c13d%3Ag%3AzScAAOSw32lYuwn2&nma=true&si=BN8gvAtMb%252BuG%252BtYOSgQcP31agMs%253D&orig_cvip=true&rt=nc&_trksid=p2047675.l2557 T

Perspective is such a curious thing....give me the same operators that make the FS700 sing for you on a rigged GH5, and to my eye and taste, the GH5 will blow it away...I intend to use it along with a Varicam LT, and I would NEVER consider the FS700 as Bcam to the Varicam...oh well...this way we're all happy with our choices

And for those that think these cameras look like "video"...which of corse they are....below is the trailer for a feature shot on the GH2...for anybody who's never seen it...this is what you can do with a GH camera

Off topic I know...but possibly relevant to those actually interested in the GH5 for narrative work.

 

[Borbarad]

1 hour ago, Phil A said:

https://www.slashcam.de/artikel/Test/Panasonic-GH5---8-Bit--10-Bit--4-2-2--V-Log-L---alles-OK--.html

Next part of the Slashcam review, testing 8bit vs 10bit and V-Log L.

Summary:

• They say there is some issues how NLEs interpret the white point and black point of the footage
  In Premiere Pro CC neither 8bit nor 10bit fit to the whitepaper from Panasonic.
  In Media Player Classic - Home Cinema the files look exactly the same
  In Davinci Resolve under OSX the files are different again, even in comparison to Premiere Pro on Windows
  In FinalCut X the files are practically identical but they can't really judge the white and black point because it gives % instead of values
• Latitude for grading with V Log is, as expected, better with the 10bit material than the 8bit. A lot less posterization when bending the image.
• They think both Adobe and Blackmagic have to fix compatibility within their software to really make the most out of the footage and be able to actually judge the 10bit V Log of the GH5

Thanks for the link!!! Excellent read!!!

i think cinema5D is now in real trouble and we should have a separate blog entry here to discuss openly their findings.

 

B

[Ken Ross]

The more you read the more you realize Andrew was correct. 

[Davey]

2 hours ago, Borbarad said:

Thanks for the link!!! Excellent read!!!

i think cinema5D is now in real trouble and we should have a separate blog entry here to discuss openly their findings.

 

B

Trouble as in legal trouble, or trouble as in their reputation taking a nose dive? If it is the former then a precedent might be set (in court) that will make would be reviewers and testers think a thousand times before hitting the publish button.

[Fritz Pierre]

33 minutes ago, Davey said:

Trouble as in legal trouble, or trouble as in their reputation taking a nose dive? If it is the former then a precedent might be set (in court) that will make would be reviewers and testers think a thousand times before hitting the publish button.

Well in my mind that would completely reasonable on Panasonic's part to go after them legally...to publish something as a review and then call a major feature "unusable" in a pre-production camera with, what many have concluded, were flawed tests, certainly would seem to get to the realm of legal liability!...for whatever their motivation was...I think in reality, C5D has hurt themselves more than Panasonic...GH users wait a long time for the next releases of the cameras...for perhaps the sole reason that they know in the next incarnation they get a lot...in a bulletproof platform to boot....not difficult to see from this thread who do and don't like the GH5...becomes obvious rather quickly...even if the "interest" in the camera is initially shown, the drop away rate at any perceived "flaw" in the camera, regardless of operator ignorance or error is instantaneous...for those who actually work with the GH cameras, this may just the most anticipated I can remember for a while....

[Cas1]

Question to me is, did C5D intentionally produce a misleading review. Maybe their conclusions are just wrong* because C5D is not competent or able to judge the product?

* Maybe they should have said "We've tried x, y, z, but could not find a way to see a useful difference in 8 bit versus 10 bit recording" ? - instead of saying something IS.

The Slashcam article shows the editors are possibly not ready to work with the GH5 10 bit files, I think C5D should have taken this possibility into account. Like one would not judge the RAW photos without having an updated RAW interpretation.

From the over processed video comparing 8 vs 10 bit I do think 10 bit does help a lot.

On the other hand I think the 8 bit's limitation acts as a noise reduction filter, whereas 10 bit records more noise/detail in the shadows. This may eat more of the available bitrate leaving less for the well exposed areas? I think with a good noise suppression setting in the editor you probably can have a more pleasing result from the 10 bit recording. BUT we will see!

[Fritz Pierre]

1 hour ago, Cas1 said:

Question to me is, did C5D intentionally produce a misleading review. Maybe their conclusions are just wrong* because C5D is not competent or able to judge the product?

* Maybe they should have said "We've tried x, y, z, but could not find a way to see a useful difference in 8 bit versus 10 bit recording" ? - instead of saying something IS.

The Slashcam article shows the editors are possibly not ready to work with the GH5 10 bit files, I think C5D should have taken this possibility into account. Like one would not judge the RAW photos without having an updated RAW interpretation.

From the over processed video comparing 8 vs 10 bit I do think 10 bit does help a lot.

On the other hand I think the 8 bit's limitation acts as a noise reduction filter, whereas 10 bit records more noise/detail in the shadows. This may eat more of the available bitrate leaving less for the well exposed areas? I think with a good noise suppression setting in the editor you probably can have a more pleasing result from the 10 bit recording. BUT we will see!

Well the issue for Panasonic could be and is what they spent on R&D, manufacture and marketing of this product...for a publication like C5D to come along, and label what to me personally is a huge feature in this camera (Vlog) as "unusable) WTF!...motivated possibly as "click bate" as some members have pointed out...maybe that's a part of C5D's business model...but you cannot make your business model dependent, on possibly slandering an extremely anticipated product!...of course Panny lent them the camera????...but still

[Neumann Films]

39 minutes ago, Fritz Pierre said:

Well the issue for Panasonic could be and is what they spent on R&D, manufacture and marketing of this product...for a publication like C5D to come along, and label what to me personally is a huge feature in this camera (Vlog) as "unusable) WTF!...motivated possibly as "click bate" as some members have pointed out...maybe that's a part of C5D's business model...but you cannot make your business model dependent, on possibly slandering an extremely anticipated product!...of course Panny lent them the camera????...but still

No way Panasonic lent them the camera, they got their hands on one, but no pre production models were meant for "review".  I was told that multiple times.  IF Panasonic did lend them the camera for review, that almost makes it worse.  

[dbp]

7 hours ago, hyalinejim said:

Mart

Here's their comparison of 8bit (left) v 10bit (right), contrast stress test:

Cool, another good example. Can definitely see improvements on the dolls face by the red/pink and yellow-ish spots. Transition is smoother. This makes me happy. I was pretty sure that 10bit ought to show improvements that don't require pixel peeping, based on my experience. Glad to see a few tests are confirming that. 

[noone]

7 minutes ago, dbp said:

Cool, another good example. Can definitely see improvements on the dolls face by the red/pink and yellow-ish spots. Transition is smoother. This makes me happy. I was pretty sure that 10bit ought to show improvements that don't require pixel peeping, based on my experience. Glad to see a few tests are confirming that. 

Maybe but the slightly different angle might make a difference too.

[Oliver Daniel]

The implementation of 10bit is the major benefit as compared to 8bit. 

The codec, the processing, the data rate and sensor all play a factor - 10bit is not going to work at its full potential if these other factors are left behind. 

The FS7 is a great example of well implemented 10bit. Absolutely superior to the FS5 in 10bit internal. 

That said, with most watching video on smart devices - the "final" image quality of 10bit doesn't offer a huge, life-changing leap. 

I'm caring much less about specs and focusing on improving the content in the frame.

Understanding light and composition is far more powerful in improving image quality. We're at that time now. The tools are awesome. Beyond enough. 

[AKED-M]

A brief look at a production model of the GH5:
https://youtu.be/yx95FelLXN0

[Emanuel]

2 hours ago, Oliver Daniel said:

The implementation of 10bit is the major benefit as compared to 8bit. 

The codec, the processing, the data rate and sensor all play a factor - 10bit is not going to work at its full potential if these other factors are left behind. 

The FS7 is a great example of well implemented 10bit. Absolutely superior to the FS5 in 10bit internal. 

That said, with most watching video on smart devices - the "final" image quality of 10bit doesn't offer a huge, life-changing leap. 

I'm caring much less about specs and focusing on improving the content in the frame.

Understanding light and composition is far more powerful in improving image quality. We're at that time now. The tools are awesome. Beyond enough. 

Well said. The benefit can be mostly testified at the big screen.

[Zak Forsman]

Really had my fingers firmly crossed for this, but I'm seeing the same cyan/magenta blocking in this video that lead me to sell my GH4. Guess, I will stick with my Blackmagics until something better comes along.

 

[PannySVHS]

1 hour ago, Zak Forsman said:

Really had my fingers firmly crossed for this, but I'm seeing the same cyan/magenta blocking in this video that lead me to sell my GH4. Guess, I will stick with my Blackmagics until something better comes along.

 

 

Hallo Zak,

Martin shot and shoots nice videos on the Lumix range of cameras I think.

For this one he used  CineLike V 0/-5/-5/-5/0, Master Pedestal +10, Luminance Level 0-255 (copy and paste from his description)

As experience has shown, one should not mess with master pedestal. Also, h264 Inter Frame is not a beautiful thing on even surfaces, additionally

very low light didnt help the 150mbit codec neither.

By the way I really liked your CineV and VLog footage, even if it was a GH4  

[Emanuel]

Actually, high ISOs ain't nothing recommendable at all on GH series, even sensor size apart... BTW, neither Blackmagics.

[Zak Forsman]

49 minutes ago, Emanuel said:

Actually, high ISOs ain't nothing recommendable at all on GH series, even sensor size apart... BTW, neither Blackmagics.

This wasn't an issue exclusive to high ISOs. GH4 was plagued by this issue at any ISO. 

 

3 hours ago, PannySVHS said:

Hallo Zak,

Martin shot and shoots nice videos on the Lumix range of cameras I think.

For this one he used  CineLike V 0/-5/-5/-5/0, Master Pedestal +10, Luminance Level 0-255 (copy and paste from his description)

As experience has shown, one should not mess with master pedestal. Also, h264 Inter Frame is not a beautiful thing on even surfaces, additionally

very low light didnt help the 150mbit codec neither.

By the way I really liked your CineV and VLog footage, even if it was a GH4  

 

1

Thanks! I would love, love, love to be wrong, but this cyan/magenta blocking was present in everything from the GH4. It was even noticeable (albeit tighter) when I would capture in 10 bit externally on my Odyssey 7Q+. internal compression made it look worse, definitely. but I'm pretty confident there was a fundamental issue at work that the GH5 likely shares. Which really bums me out. I was inches from pre-ordering one.

[Fritz Pierre]

9 hours ago, Zak Forsman said:

This wasn't an issue exclusive to high ISOs. GH4 was plagued by this issue at any ISO. 

 

Thanks! I would love, love, love to be wrong, but this cyan/magenta blocking was present in everything from the GH4. It was even noticeable (albeit tighter) when I would capture in 10 bit externally on my Odyssey 7Q+. internal compression made it look worse, definitely. but I'm pretty confident there was a fundamental issue at work that the GH5 likely shares. Which really bums me out. I was inches from pre-ordering one.

I must have been lucky with my GH4's...still very happy with the cameras...though I don't do the stress tests that you do...I'm simply more than happy with the image and what it takes to operate the camera!...Looking forward to the GH5 also!....Saw some lovely footage from you on the BMMCC though....did'nt you have to come up with a major workaround to make that camera work for you?

And for those waiting for their GH5s...some anamorphic footage shot on GH4 I just saw today!....