If you were on Fstoppers at all last week you probably read at least one of my articles about the Panasonic GH5 and the debate over its autofocus. Although we are no closer to a fix, this story keeps getting more entertaining.
If you haven't read my other articles about this and you want the full story you can read them here and here. For those of you who want the condensed version, let me summarize what's been going on.
GH5 users have discovered that continuous autofocus becomes extremely unreliable once recording begins on the camera. Max Yuryev has created multiple videos that have done a great job of highlighting this issue. Last week, YouTuber and sponsored Panasonic shooter "PhotoJoseph" created his own video and basically claimed that Max didn't know what he was doing. Max then offered to pay for Joseph to fly to his hometown so that they could create a video together, fully and fairly testing the AF on the GH5. Surprisingly, Joseph agreed.
As I expected, nothing new was really discovered. Although both men were able to agree on "optimal settings" for certain types of shooting, I still don't think the camera's performance was ever reliable enough to actually count on it.
If you have the patience you can check out Max's 46 minute version of this video below.
Now, check out Joseph's edit of the exact same footage.
To be honest, it doesn't really feel like either of them is being totally genuine. Max's video is at least fair because he shows how bad the AF actually can be but in my opinion he should have been much harsher in his wrap up at the end of his video.
Joseph's video only shows the footage of the camera working at 100 percent which is completely unrealistic.
To complicate matters further, Panasonic finally made an official statement about AF through PhotoJoseph and their "recommended settings" seem to conflict with everything Max and Joseph worked so hard to figure out.
Panasonic states: "We recommend to select 1-area AF in 30p or 60p frame rate for more comfortable Auto Focus speed. When you select 24p, 1-area AF is recommended."
They wrap up their statement with: "Panasonic pledges to further improve the AF quality."
Being that the camera can focus much more reliably at 60fps and that it can track faces better than it can focus on them, I'm still hoping that this is just a software issue and that it can be fixed with a firmware update. It does seem strange that Panasonic wouldn't just come out and say "This will be fixed in a later update," but I'm sure they don't want to make any more promises they know they can't keep.
I agree...Yuryev took it very easy on that guy at the conclusion, which is fantastic. No need to whip a dead horse it won't move a bit faster.
As for the camera, I would certainly expect more for $2K and the argument that "real videographers manual focus" doesn't fly with me.
Seems like a big difference in Max and PhotoJoseph's settings is one was at 60p (Joseph) and the Max was at 24p. 60p looks tolerable with the tweaked settings but 24p looks unreliable no matter what.
I'm fairly certain this is because for cinematic video, you shoot 24p to get some motion blur so the video doesn't look jittery but this blurriness is also confusing the CDAF focusing. I don't see how they can fix 24p without having a sensor with PDAF sensors but that doesn't exist on m4/3 because the circuitry takes up too much space... 😔
Olympus E-M1 has had PDAF since it was announced in 2013. I'm not sure if it works in video, but it is certainly possible to have PDAF and CDAF on the same sensor.
Didn't realize the only em1 had PDAF. Not sure why Panasonic didn't use it then. Only thing I can think of is they thought their DFD/CDAF was comparable and used the extra sensor site space to maximize low light capability...
Correct...though I can't say for sure how effective it is in daily use. Frankly I don't have a problem with my E-MI focusing in the manner I use it.
It was designed for use with the Zuiko DSLR lenses and it works very well with those. According to some reports I have read online, the E-M1 decides on its own when to use PDAF and CDAF on non-DSLR lenses (ie the native m43 lenses), but Olympus aren't saying when this happens.
Once again you're mischaracterizing. In the last article you said I was going to "show him how to use the camera", which was completely misrepresented (did you watch our discussion the day after the "call out"?), and now you're neglecting to mention that:
A) In my video I clearly state that I'm only showing the good stuff, and point out that you should watch Max's video for a more complete view.
B) I clearly label the 24p footage and even call out part way through the video that some was shot that way, and that it is labeled as such.
C) Our results ARE different from what Panasonic suggested (which obviously came after) which should prove the point that we actually tested and reported back the best findings (which was the entire purpose of our joint venture). Our primary surprise finding was that turning custom AF off gave us the best results. Most of what I showed was shot that way. That's pretty consistent.
What's the point in showing failed footage? I've already done that and everyone knows there are settings that don't work. I'm sure it won't surprise you to learn that there are aperture, shutter speed and ISO combinations that don't work in certain situations either, but you don't need a video showing a bunch of over/under exposed images to prove that.
Next time you're going to write an article and say what I did or didn't do, try reaching out. This is at least the second time you've written about me without so much as a courtesy call to ensure you have your facts straight. I'm pretty easy to find.
The "teach him how to focus his camera" comment in my last article was suppose to be a joke because I felt Max had done the most comprehensive af test possible and I didn't believe you were going to discover anything new together.
Everything else you've written is true and a fair critique but I summed up my opinion of both videos in 2 sentences: Max should have been harsher and you showed the best footage. I certainly could have written more but I felt like I was being equally critical of both of you and I didn't want to dwell on that. I also posted every video that I referenced so everyone is welcome to make up their own mind.
I could have reached out to you, and maybe I should have, but I'm not sure I had anything to ask you. I tried to comment only on info that was in your videos.
Anyway I'm not trying to be overly critical of either of you guys but I have been entertained by this entire saga. As a new Panasonic user I'm very thankful for people like you and max who have spent so much time trying to figure this stuff out but at this point I think we can all agree, it's in Panasonics court, or it's time for us to forget about continuous AF and use manual focus like the majority of professional videographers always have.
Thank you Lee. Nice response. If you do have questions (especially as a new LUMIX user) please don't hesitate to reach out.
I gotta be honest, uh, suckkit … I felt the same way but have been told repeatedly that many more pros than I realized use AF. It's certainly more important than I realized!
I've gotta disagree with you "suckk it," Although I personally have never used AF the only reason I haven't is because none of my cameras could do it well enough to rely on it. Since going down this crazy GH5 AF hole, I've learned that many people do rely on the AF on Sony and Canon cameras and they have proven to me that not only is it reliable, but that it is also far more accurate than a human trying to pull focus.
I'm personally not used to these features but I've seen it being used and it is quite impressive.
For wedding, With better af, you could open up possibility for more complex shots. Autofocus on a beholder is awesome. I've tried it.
Maybe you don't want to rely on the autofocus during critical ceremony. But during portraiture session, af is awesome. Just because the "PRO" around you doesn't use it, doesn't mean others don't. Just like when people said "if you want to shoot a video, buy a video camera" when first video dslr came out.
Contrast detect autofocus is pretty much unreliable at focus tracking compared to on sensor PDAF. This is simply the case of asking a product to do something it wasnt design for.
A few seconds with google would show that this is complete BS. I have no idea how well it works, but Panasonic do indeed make very strong claims for the CDAF. The DFD technology is supposed to give exactly CDAF exactly what PDAF had - information on the correct direction to move lens elements to achieve focus:
http://eng.faq.panasonic.com/app/answers/detail/a_id/26478/~/what-is-dfd...
Perhaps would be better not to make sweeping statements about technology you don't understand and can't be bothered to research?
Yes its a sweeping statement. Whats wrong with it??
You said you have no idea how well the DFD works, then how do you know my "sweeping statement" is complete BS?
Again, without giving a damn about the GH5, but in the name of honest journalism, WHY is it "official"? that the GH5 has an AF problem? That makes it sound like they've admitted this. Whether the camera has a problem of not, there is no evidence for this in your article. They've recommended best settings and said that in the future they'll improve the AF - these are very different things to an admission of a problem.
They are, of course typical evasive PR speak, but distorting what's been said doesn't make matters better.
That's fair. I didn't mean official from Panasonic. I meant official after the Joseph and Max series.
Don't apologize. You were right the first time.
That "PR" message is an admission that what has been hyped and advertised is actually falling far shorter than expectations. Very few major companies will ever admit to making a mistake and they will twist and turn to say everything but "we lied or we messed up". The only way that get's done is when pressure becomes so high that they have no choice in order to stem the flow of lost corporate value. Remember Sony's A7 light leaks..etc etc.
This isn't a political site. We can actually say what we know is true despite PR-Speak...sheesh.
The message I delivered is official from Panasonic.
Isn't this trilogy of GH5 "articles" a kind of badly formulated crusade to applaud the author and toddlerishly abuse the Fstoppers position to prove his point (effectively downgrading the value of this website)? The childish clickbait tone reminds me to text-eruptions unleashed by these hopeless e-Don Quichot's trying to convince all the other forum-members *they are right, and you certainly are not*.
I suggest to invest time in useful articles on photography instead of cheering yourself... *yawn*
Why is everyone so freaking opinionated on this Black and White case of "GH5 Auto focus in video is not very good".
Canon has made their current DSLR with incredible video autofocus that is the most reliable ever seen, let alone at the price point, and they get utterly castrated in the public arena (5div 1dxii) with regards to bit rate, file size and crop factor in 4k (none of which mind you would every surprise you after setting up a shot and prevent you from getting said shot) meanwhile the GH5 video auto focus is basically crap compared to the competition and the public opinion is split 50/50 with so many people defending them. Its an incredible camera and at 2,000$ it is groundbreaking, the video autofocus appears to be flawed, and maybe Panasonic can fix it, why on earth this is so polarizing is beyond me.
Initially I defended the GH5 but the more I use it, the more I'm discovering that in backlit situations, not only am I getting horrible green reflections of the sensor in certain situations but also the camera has horrible problems focusing on objects in the foreground in these backlit instances. I'm talking about shutter or button focus not manual. I'm going to consider in the next few days what I'm going to do. The stabilization is great and the 4k footage is amazing and of course the no limit on record time. The camera also doesn't overheat. Sigh..... I really do not want to go back to sony.