Panasonic GH Firmware Updates Are Great, But Still Struggles With AF | ‘Wedding Dance’ Test

Nothing makes me feel like a fangirl like receiving a good firmware update, but it was not always golden. The original Canon 5D, for example, only received one bug-fixing firmware update during its product cycle but was then followed by the Mark ii which received six, some of which included new features.

Manufacturers have realized that delayed product upgrades and firmware alternatives can be more profitable, yield higher customer satisfaction and brand loyalty. Now all the the major ones are releasing interesting upgrades staggered along the product cycle, and Fujifilm is the paragon of this. But Panasonic has also set the bar high.

[REVIEW: Hands On With Canon’s 2 New 70-200mm Lenses | First Look & Shoot]

Last year the Lumix GH5 received firmware with an all-new 10bit codec to continue attracting and satisfying serious cinematographers; the new HEVC standard codec that signals to filmmakers how future-proof the camera is meant to be, and improved IBIS to signal how committed Panasonic is to the ‘run and gun’ filming style. And yet, Lumix owes   their customers more.

It’s been widely reported, that while Sony and Canon have had high success with autofocus for video, the GH5 was released with subpar video autofocus tracking for a flagship camera.

Sure, very serious filmmaking circles they don’t rely on autofocus, they pull focus manually.  In literally no high-end production will you ever see a camera hunting around for focus. Moving focus with grace and intentionality is a hallmark of great cinema, and yet there are countless use-cases where autofocus could help the YouTube community, indy filmmakers, and wedding videographers.

Lumix released this firmware update in three different forms for the GH5, GH5s, and G9.  All of them share a new maximum 20x enlarged preview to confirm focus during manual focus assist. The thinner the depth of field, the more valuable this new 20x enlarged preview will be to help confirm focus.  In addition, Lumix is trying to offer all-in-one solutions in their LCD monitor.

One of the reasons filmmakers invest in an external monitor is for advanced focus confirmation features and waveforms. In this newest update waveforms can be displayed while adjusting white balance.  

Also, when I tested the newest firmware on the GH5, I noticed about a 10% increase in autofocus accuracy as the camera  pulses less frequently to reconfirm, and pulls focus a little more ‘naturally.’

The other fascinating thing about this firmware update is that it primarily unifies the GH5 and G9 cameras around the newest software released with the GH5s launch.  So the upgrade in autofocus accuracy is simply the GH5 and G9 inheriting the latest version of the GH5s. The reason this is interesting is that now Lumix can gather larger pools of user data about what works and what doesn’t work in their software.

For years, camera manufactures have produced different autofocus algorithms and different focusing clusters on their sensors across different camera bodies in their product line.  The idea was to offer product and price differentiation for the marketplace, but that meant each user group returned isolated feedback.  This is dumb.  The success of Apple’s iOS has taught all software developers that a more unified approach simplifies bug fixing solutions.

The GH5 has become a rising star for wedding videographers, and accurately auto-tracking less predictable subject movement (wedding day clients) would help shore-up sales for filmmakers who are also watching the Sony alternatives closely. That said, there’s no getting around the fact that the GH5 has no on-chip PDAF and the Sonys (like a6500 and A7iii) do, and that’s not something a firmware update can change. That said, however, of course there are reasons for not having phase-detect pixels such as the somewhat possible flickering with tracking slow moving subjects – but that’s a topic for a deeper conversation.

Autofocus movement for video is the closest thing camera bodies have to artificial intelligence, because they are trying to replicate human practices.  So this may be one of the big software obstacle to camera automation.  Here are the list of specific updates: