2019 Crop Sensor Bodies For Video

[VidThreeNorth]

I will try to accumulate notes regarding video use of recent crop sensor bodies in this topic. So far we have two Sonys, two Canons, a Nikon and an Olympus. I don't think I have much new to say about the Sony a6100 or a6600, since they have been fairly thoroughly looked over for video. Similarly the Fuji X-T30 and Sony a6400 are well documented. The Panasonic G95 is sort of on the borderline. It has not really been so well covered, but then again, the reviews don't point to any unexpected problems, and the upgrades seem straight forward -- nothing that Panasonic has never done before, so I don't expect to hear anything really surprising.

Nikon Z50:

The thing to look for here is whether the sound (pre-amp) is clean. Neither the Z6 nor the Z7 has a clean pre-amp. If you are recording with sound and auto level, and everything goes quiet, then you will hear the hiss. Thanks to Jordon Drake (DPreviewTV) for demonstrating this in their video review. This means you cannot rely on the sound in such situations, such as recording speeches (where people are actually paying attention). You can always add an external recorder, but every piece you add increases the chances of something going wrong, and sometimes you might simply want to work with a "light" or small setup.

If it has clean audio, and you want to build out a Nikon video system (probably because you already have Nikon lenses), then this might be the better "base camera" to build around.

Canon M6 II and EOS 90D

At this point, the only thing I know for sure is that the 4K video of these cameras is not as good as similar cameras by their competitors. Apparently someone has posted that the camera is skipping 2 out of 3 sensors horizontally, for a real resolution about 2.8K wide, but I do now know how reliable that claim is. It would, however, fit with the common analysis that these cameras are producing "better video than their Full HD", but visibly not as good as competing companies' 4K video. In fact, it opens the company up to saying that among the current competing Japanese cameras, for UHD, "the only thing as bad or worse than a Canon is another Canon".
[2019-1022 13:47 corrected "2.8K"]

If this claim that the company is skipping 2 of 3 sensors is correct, then this is not a new situation. Some of the low cost Chinese video camcorders have been doing this sort of thing for some time now, and I have couple of years experience with my Git2 which does this sort of thing. Again, Jordon Drake has put it well that if you are "serious" about recording 4K video then you will want to consider something else, but if you are more "casual", then it could be ok. However, some people are calling out Canon on the grounds that if this is true, then they should not be claiming "4K" capability, and I agree. I do think that it might be justifiable to call this "UHD", though even this is stretching the accepted terminology, but not "4K".
[2019-1022 13:50 slight rewrite for clarity.]

I have also posted a question about what kind of bit-rates are the cameras giving, which is also relevant.

[2019-1022 13:50]
Here are a couple of links that are relevant:

"Canon lying about their video specs?", posted by "Dave McKeegan", Sep 2, 2019,
""
The above video is where I have heard about the "pixel skipping". I have not looked for his source yet.

"Canon M6 Mark II Cinematic 4K Video Test", posted by "DSI Pictures", Sep 4, 2019,
""
In this video you can see the results for yourself. Moreover, Canon has announced that early next year a 24 fps. record mode will be added. It is possible that this will allow better sensor use.

[2019-10-25 16:13]
Follow-up Report on Above Issue:
Are the EOS M6 II and 90D Recording "4K"?

I have pushed my search a bit further and back tracked to:

"Canon EOS M6 Mark II 4K is “pixel binned 3K” and Sony A6600 old sensor debacle “due to internal politics”" published Aug 29, 2019,

on EOSHD.com

"Canon EOS M6 Mark II 4K is "pixel binned 3K" and Sony A6600 old sensor debacle "due to internal politics" - EOSHD"

At this point, I have not found an explanation for the "how do you know?" / "why do you think so?" question (really the same question asked two different ways). Nothing in this artical says anything like "this is a test we performed that proves ..." or "I have seen a document in Canon engineering that says ...", or equivalent. I don't know if I can push my research further. I really don't have the time.

Olympus OM-D EM-5 Mk III

These are just being tested now and the camera looks pretty good, though it lacks a good "high dynamic range" recording mode. There are two things that I think were mistakes for video. First, there is no headphone jack either built-in or attachable. The previous model two had an attachable headphone jack. Also, the battery is smaller. While the circuitry is more efficient, nobody has said that battery life on the older camera was "wonderful". It was only "adequate" or "reasonable". Keeping the old battery would have made users happier than saving such a small amount of weight and size.

 

[VidThreeNorth]

Fuji X-Pro3

For video work, the X-Pro3 is essentially like the X-T30. The difference in view finding systems is mainly a disadvantage because using the flip-down screen on a tripod becomes limited. I do think there might be a case for using the optical viewfinder for still photography burst mode situations, where an electronic view finder (back screen or eye-level) could be inaccurate. But really, attaching a view scope on a "shoe" on other cameras would probably work as well. But realistically, an X-T3 (NOTE "T3", not the "T30") is a better buy for 4K video in the Fuji lineup.

"FUJIFILM X-Pro3: 7 Things to Love About This Camera",
by "Gerald Undone", Oct 22, 2019
""