Eno

1 hour ago, Michael Ma said:

You missed the part people have been using 8bit C-Log for years with very little issues with banding.  Are you saying GH5 has better LOG because it is 10bits? 

Yes, I'm saying any LOG recorded in 8 bit is useless IMO. You can disagree of course but I wont change my mind cause my repulsion for very flat profiles and 8 bit comes from experience (problems that needed solving, headaches etc).

The 5D mk4 is a nice photo camera... and it's better to remain that way.

53 minutes ago, Django said:

Have you ever shot with the 5D4? because it sounds like baseless erroneous assumptions you are making here.. fyi FHD has improved since the 5D3.

The images speak from themselves. I owned the 5D mk2 and it was a great filming tool in 2008...Unfortunately the Mark 4 looks way to similar in 2017.

You should check the video on minute 1:18 seconds for some "nice banding" and all sorts of nasty artifacts:

1 hour ago, Django said:

DPAF isn't the only advantage (it's a big one though!). The 5DIV now has C-Log with view assist. It can shoot DCI 4K. MJPEG is a high bitrate 422 codec (for some that's a plus). Wi-fi remote capability is superior to Nikons.

As for the 4K image quality, I disagree. Sharpness isn't everything (and I wouldn't call the 5D4 4K very soft). As far as color sampling (422), motion, lack of artefacts etc.. the 5D4/1DC/1DX2 DCI 4K is superior to most internal 4K from the competition imo..

C-LOG is close to useless in a 8 bit only IMO, + it exhibits banding in the shadows (even Canon themselves caution this). MJPG is an dinosaur of a codec, it's superiority ended in 90's. + if you dare to think about the "luxury" of recording 4K externally from the 5D mk4, you don't have it...nada, niente, no...you're stuck with the very old, extremely inefficient Motion JPOG encoding...That's it, end of story!

I never said anything about sharpness, I was referring to detail or the lack of it. The so called Canon 4K is more like a 2,5K resolution wise, the FHD is as pour as in 2008 (this year Canon celebrates 10 years of unchanged "1080" quality).

Did I missed anything?

9 hours ago, mercer said:

What exactly does a horribly "non efficient codec" entail? I mean video gets recorded, right? And it's a 4:22 codec that's easily editable on most computers, and usually the motion cadence looks great.

Okay so a 1.7 crop isn't exactly S35 but it's closer than 2x or 2.4x or 2.88x. And with the crop, a bunch of aps-c lens options open up.

I don't care about the this camera is better than that camera game. Each camera is unique to itself and none of them are perfect, but I do think the 5D4 gets a little more hatred than deserved.

The crop factor is actually 1,75x in UHD (16:9 aspect ratio) - just horrible (an nod, the APS-C lenses do not work on the EF mount, you're stuck with FF glass on a huge crop)! The codec is very inefficient, the "4K" image (more like 2,5K actually) is very soft, high ISO is worst than on the Gh5 (in video mode), FHD is like any other Canon DSLR a pour 720p resolution full of moire and aliasing (the only exception was the 5D mk3 which was even softer). It has 0 advanced video controls but hey, Canon "must be praised" for it's dual pixel AF, to do exactly what with it if the image quality is so pour?!

The 5D mk4 deserves all the hatred, it is an expensive embarrassment video wise!

15 minutes ago, Andrew Reid said:

We are a long way past line skipping now, it died long ago.

Unfortunately not for Canon.

"I hope the D850 looks as clean and doesn’t suffer too much from the pixel binning, which can hurt low light performance and increase moire & aliasing."

Please @Andrew Reid, don't confuse "Pixel Binning" with "Line Skipping". They are two different things with totally different results! 

The D90, D3s, D300s, D5100, D7000 where all line skipping cameras, like all the Canon DSLR's in FHD. But the D5200, D5300, D5600, D5600, D7100, D7200, D810, D750, D5, D500, D7500 (I hope I didn't missed any) are all using pixel binning in FHD with minimal moire and aliasing but with detail not as high as an image obtained from full sensor readout with resizing (like on the GH5, A6300, A6500, A9 etc).

By the way,thanks for the video feature summary, the other sites are focusing to much on the still capabilities of the D850.

A very interesting product, thanks for sharing it!

Thank you very much for the slow motion comparison test @Andrew Reid!

I don't think Nikon is struggling with their sensors, just DXO mark is very unreliable!

For example, any sample image you compare between the 1Dx mk2 and the D5, the D5 starts with 1/3 advantage in noise at medium ISO values and improves with almost one stop at very high high ISO. If you do an average between those values the D5 is around half a better in noise department compared with the Canon, but DXO scores the low light performance of the D5 with almost 1/3 stops worst than the Canon (that's more than 2/3 stops worst than the camera really is). So are Nikon sensors bad, or just DXO measurements poorly made?

The same thing appliers to the Nikon D500 sensor, Panasonic GH5 and a couple of other cameras that are underrated by DXO. Wen directly confronted them about the flawed I got a permanent ban on their site from the computer I was writing the observations (very curios).

@Andrew Reid

Thank you for the pertinent article!

I don't think the new Sony 24 MP stacked sensor is fast enough for 60 fps full sensor readout, even for the 30 fps they had to reduce the read out surface area (because of that it has a crop factor compared to 24 and 25 fps).

My bet is on a new and faster 12 MP sensor, just ready for 4K 60p. Sony will do everything in it's power to differentiate the models as much as possible. If you want speed, get the A9, if you want resolution get the A9R, if you are interested in video get the A7sIII or A9s. For them it wold be ideal if you buy them all.

@Lintelfilm

The V-Log profile most probably is the same but the information in it will definitely be another.

11 hours ago, Lintelfilm said:

Not if it's the same Vlog that shipped for the GH4. My grading skills are perfectly adequate and never had a problem with the hundreds of hours of Blackmagic film or C-log I've worked on, but the externally recorded 10 bit v-log I got from my GH4 was truly abysmal. Infinite skill and patience can't do anything about the weird banding and colour artefacts. Great if it's different on the GH5 but the Cinema5D analysis didn't fill me with hope.

"The recent article featured on the Nikkei regarding Panasonic’s Imaging business was not announced by Panasonic and refers to a change in our internal organizational structure.
Integrating all consumer electronics divisions, our consumer Digital Imaging business will move under Panasonic Appliances Company and is not being dismantled.
The aim of this change is to further deepen our relationships with customers, strengthen our product capabilities, and continue to firmly develop and promote our business."

https://***URL removed***/news/4882489366/panasonic-likely-to-scale-back-camera-division

@Andrew Reid

I've enjoyed your article and think you are right. The above (just made public) Panasonic official response to those "dismantling" news say practical the same thing.

@zerocool22

I've read "clean 1600 like the GH5 is not enough for me."  Sorry.

@zerocool22

GH5 has quite good low light capabilities with ISO 3200 very usable, heck even ISO 6400 is somewhat usable IMO. I don't know what more were you expecting from this camera?

@Andrew Reid

A very interesting and invigorating article, thanks for sharing it!

27 minutes ago, Lintelfilm said:

Panasonic can shuffle around their company as much as they like, but until they offer colour science on par with Canon and Nikon, a log profile that is useable and good video AF, i stand by my switch from the GH line to an 8bit 1080 camera (C100Mkii) and for small body 4K an XC10. 

@Lintelfilm

I find the GH5 colors to the best I've seen so far in any ILC camera to date.

The V-Log is very nice if you have the skills or desire to grade it.

The AF is usable but not as good as Canon's dual AF. Anyway, for shallow dept of field applications I prefer a very good MF assist options to any current AF tech cause the camera doesn't know where I want the focus to be, how and when.

@hyalinejim, from what I saw, Fuji by default applies some stronger NR cause the detail is lower than the GH5 and both of them are suppose to be resizing from a 5,5K image to 4K.

@TheRenaissanceM, you are very correct about their lack knowledge how to properly measure DR. It is as Driftwood said:

"Cineon adaptation \ log curve interpretation generally looks different from toe to shoulder between different cameras. Its like Slog 2 is good for highlight detail, slog3 for bringing out shadow detail and vlog sits somewhere in between if not too unlike slog3! C log curve differs too. The point is you have to expose\grade correctly with Log & for the camera and DR youve got."

@jonpais

I'm not taking sides either, I've just pointed out (what I consider) flaws in their testing methodology.

@Andrew Reid

Totally agree with your points Andrew!

And if I sum it up:

They've disable the NR on the GH5 (at least that's what I understood) than complained they see noise in the shadows :)), so "the DR must be low". Heck I downloaded their test, brightened the shadows and applied a little bit of NR and whoops the DR is over 12 stops.

They complain about bad color on Vlog but they edit the log profile with LUTS. :))

They complain about lack of detail but they've dialed the sharpness to a minimum cause it's to much...seriously?

They see no advantage in 10 bits vs 8 bits but never show any example, like on a plain sky with a blue gradient + heavy grading.

They complain about soft 180 fps on the GH5 but never complain about the utterly crap Canon 1080p in DLSM's, which is nowhere near at 30fps compare the the GH5 180 fps.

Etc, etc.

@Andrew Reid,

"What does God have to do with it you lunatic?"

I think only Him can have an "absolute proof" about anything really.

It is wise no to start with personal insults cause you'll lose credibility.

@Andrew Reid

"I just wondered if you had absolute proof or not."

I think only God may have that prerogative.

@Andrew Reid,

The info is from the internet:

The sensor and processing units are definitely made by Panasonic: http://leicarumors.com/2015/08/13/the-sensor-inside-the-leica-q-typ-116-is-most-likely-produced-by-towerjazz-an-israeli-semiconductor-manufacturer-with-tight-relations-to-panasonic.aspx/

hence the similarities with the GH4 like the V-LOG format, codec, AF, bit rates etc.

The zoom lenses are made by Panasonic: http://leicarumors.com/2016/08/17/leica-apo-vario-elmarit-sl-90-280mm-f2-8-4-lens-designed-by-panasonic.aspx/

The 4.4 M dot LCD is made by Epson: https://***URL removed***/articles/7963916751/epson-mass-produces-4-41m-dot-lcd

Of course I may be wrong about origin of the SL1... But the viewfinder is definitely NOT 4.4 MP but 4.4 M dots or 1,46 MP and the crop factors in 4K and DCI are as a wrote.

Some corrections @Andrew Reid,

The SL1 EVF is NOT 4,4 MP but 4,4 million dots or 1,46 MP and it's made by Epson (not in Germany that's for sure).

"How does the 4K do vs the 1D X Mark II? It’s very good. However just like the 1D X II this is approximately a 1.4x crop of the full frame sensor." The crop factor is NOT 1,4x as you said but something like 1,56x in 4K UHD and 1,46x in 4K DCI.

And by the way, the whole camera is made by Panasonic in Japan not physically in Germany by Leica.

@Andrew Reid, sorry I was referring about Rode stereo video mic pro vs stereo video mic X. I own two pro versions and contemplating the upgrade but the cost difference it's huge. I was wandering if the difference in sound quality is worth the price or not...

I wander how does the new Rode stereo video mic X sound compared to the pro version, any insights?

Right on spot with the article Andrew!