In the tech and camera world, there are a few things that arrived much later than when they were announced and at a much higher price. Remember RED’s original announcement of “3K for $3K” from 2008? The promise was a 3K resolution camera for 3000 dollars. We never saw that camera, especially not for that price. Or the Tesla Cybertruck? That pickup truck was delayed several times over years since its announcement and it only recently started shipping. The price of the Cybertruck is also 50% higher and it offers less range than initially promised.

DJI’s initial 8K promise

In 2021, DJI announced the Ronin 4D, which featured a 6K camera unit, the Zenmuse X9-6K. I reviewed it when it came out and was absolutely amazed by the amount of innovation that DJI packed into this unique camera.

While the Ronin 4D 6K started shipping shortly after its announcement, DJI also announced an 8K version of their camera module, the X9-8K, which was supposed to start “shipping soon”. Over two years have since passed, and of course people started making jokes that the 8K camera would never appear. 

Finally Introducing Ronin 4D 8K

Except for the two-year delay, DJI actually delivered at the end of 2023, and opposed to the other products I mentioned above, DJI are exceeding our expectations and keeping the price as promised: Here is the Ronin 4D 8K, featuring the Zenmuse X9-8K camera unit that can also be purchased separately for previous owners of the Ronin 4D 6K.

The obvious upgrade from the 6K version of the X9 is, that you can shoot full-frame 8K at up to 60 frames per second, uncropped, and 4K with a full-sensor readout at up to 120 frames per second with the X9-8K. It’s really rare to see an 8K sensor to be able to deliver such high frame-rates, and it opens up lots of possibilities for cropping in post.

Physical limitations and 8K 120p

The Ronin 4D is perfect for filming fast-paced sports scenes, but of course the gimbal unit has limitations on how long your lenses can be (physically and from a focal length perspective). The longest you can usually go with a prime that will physically fit is around 85mm, which is not very telephoto on a full-frame sensor. But let’s say you shoot at 8K and only finish in HD, then you’ll be able to crop into the image in post production up to 4 times, which makes it very usable again. 

Of course, shooting slow-motion in 4K using the full sensor at 120 frames per second is a great thing to have! It’s something that users of Sony FX3, FX6 or an a7S III have become used to over the last few years, albeit with a slight crop. It’s great to see that DJI Ronin 4D users can also enjoy this feature with full sensor readout.

Ronin 4D 8K Lab Test

We performed an extensive Lab Test for the Ronin 4D 6K when it was announced, testing Dynamic Range, rolling shutter and exposure latitude. I highly encourage you to read the Ronin 4D 6K Lab Test article, in case you missed it! The results for the 6K camera were already quite impressive. 

For the X9-8K camera, we also ran a preliminary Lab Test, and I say preliminary because we don’t have a production firmware for the 8K camera yet. For this reason, we didn’t do all of the tests yet, but we of course wanted to find out how the new 8K camera stacks up against the 6K version of the X9. 

Dynamic Range Expansion Mode

One thing to highlight is the dual native ISOs of the X9-8K, which are ISO 320 and 1600. This only changes when you enable a menu item called “Dynamic Range Expansion”. This new mode is only available for frame rates of up to 30 frames per second and the Ronin 4D 8K switches the signal processing from 12-bit to 14-bit. The dual native ISO also changes to 800 and 4000 ISO, which is more similar to the 6K camera, which is 800 and 5000 ISO.

It’s important to point out that a 14-bit sensor readout still seems to be quite a rarity in the camera world. Only the FUJIFILM X-H2S also offers 14-bit color depth in a consumer camera, which we have also performed a Lab Test for.

The preliminary Lab Test results for the X9-8K camera are impressive, to say the least. Switching on “Dynamic Range Expansion Mode” will result in up to one additional stop of dynamic range, depending on the recording mode. For example, at ISO 800, we measured 13.8 stops of dynamic range at a signal-to-noise-ratio of 1, and still 12.7 stops of dynamic range at a signal-to-noise-ratio of 2.

Switching from 8K to 4K with a full-sensor readout, the results became even more impressive. At ISO 800 in 4K, we measured 14.2 stops at SNR=1, and 13.1 stops at SNR=2.

We will run these tests again with production firmware installed, once it’s out. If confirmed, this would put the Ronin 4D 8K in 4K mode almost on par with the ARRI ALEXA Mini LF, one of the best performing cameras we ever tested.

How much and when?

DJI is selling the Ronin 4D-8K as a combo, including their 17-28mm T3 lens, a ProSSD 1TB and the ProSSD mount, and much more, for just below $13,000 US dollars.

If you already bought a Ronin 4D-6K, you can purchase the Zenmuse X9-8K gimbal camera separately for around $3,600 dollars. This is a very fair price for the incredible sensor performance you are getting with the new X9-8K.

Are you as excited about this new camera sensor as we are? Do you own a DJI Ronin 4D and are planning to also purchase the X9-8K for it? What kind of projects do you use the Ronin 4D for mostly? Let us know in the comment section below! We love to hear what you have to say!

Even in post-pandemic times, live streaming is here to stay and only destined to grow. Whether you’re holding a corporate meeting, covering a sports event, or teaching an online class, you might be looking for a simple solution that transcends the capabilities of your smartphone or laptop’s webcam.

OBSBOT is a Chinese company that focuses on developing camera tools aimed at facilitating your life when streaming content. Their new Tail Air PTZ 4K streaming camera recently landed on my desk and I was able to test it for a few days. Let’s take a look!

OBSBOT Tail Air – camera and image settings

The OBSBOT Tail Air essentially combines a tiny 4K UHD camera with a 2-axis non-removable gimbal that can be controlled remotely in many different ways. The device comes in a nice carrying case and, while it’s not the tiniest gimbal-mounted camera on the market, it’s still relatively light and portable. Indeed, the whole camera system weighs 350g and measures approximately 7×7.3×13.3cm.

The camera module packs a 1/1.8” 8.4MP CMOS sensor that is paired with a 23mm (equivalent) f/1.8 lens. A fixed lens with a fixed aperture equals two things: a) the “up to 4x” zooming function is performed digitally, and this has a strong impact on image quality; b) if you want to use the camera outdoors without compromising on your shutter speed, then you have to purchase OBSBOT’s optional ND filters, which I didn’t have for this review.

The camera offers both automatic and manual exposure modes, and you can manually adjust your settings using the OBSBOT Start mobile app (available for both iOS and Android), even while recording/streaming – which is nice. Tunable parameters include essential settings such as shutter speed, ISO (100-6400), EV, and white balance. You can also customize your picture style, although I think it’s just better to leave it on Standard.

Additionally, you can toggle between manual (MF) or autofocus (AF-S or AF-C) modes. Focus distance can be adjusted manually using a virtual control wheel inside the app if needed, but I have to say that continuous AF works reasonably well in most situations, although it sometimes struggles a bit in dimly lit environments. The camera can also focus on pretty close objects, as you can tell from the picture below.

Gimbal, power options, and build quality

The camera gimbal can pan ±150° and tilt ±90°. Unfortunately, 360° horizontal range of motion is not natively supported, although it can be achieved using an optional 360° Rotation Charging Base.

Talking about power, the back of the unit has a USB-C power port that can be used to connect the camera to mains power or a portable battery. On top of that, the Tail Air has a built-in 1500mAh Li-Po battery. According to OBSBOT, the battery takes about 90 minutes to fully charge and can keep the camera running for about 2.5 hours when shooting in 1080p30. I found these figures to be pretty accurate in my own tests.

This means that you can stream from pretty much any location on Earth, as long as you have access to the internet and… well, your event is shorter than 2.5 hours. As a workaround, you could use the optional 360° Rotation Charging Base, which is claimed to extend battery life to around 7 hours.

The camera is decently built and has a 1/4”-20 threaded hole at the bottom that makes the Tail Air easy to mount on a small tripod or a light stand. Moreover, its front-facing LED status indicator proved to be extremely useful, especially when I was trying to film myself using AI-based tracking and gesture control (continue reading for more details).

Recording and streaming content

Although the Tail Air is mainly designed to be a live-streaming device, it can actually behave as a traditional, standalone camera, if needed. Indeed, it can shoot 4K UHD videos at up to 30fps or 1080p up to 60fps using the H.264/H.265 codecs (up to 80Mbps) and store them internally on a Micro SD card (up to 512GB in size). One note: these recording settings are slightly hidden inside the menu system of the app, and I wish they were more immediately accessible from the main UI.

On top of that, the camera also has a robust set of connectivity features that allow you to include it in many different types of live-streaming workflows. For example, you can:

• use the Micro HDMI port on the back of the camera to connect it to an encoder or live production switcher;

• connect the camera to a laptop using the USB-C port on the back of the body – which supports the UVC (USB Video Class) protocol – to connect the Tail Air to a switcher, or use it as an external, remotely controllable webcam;

• stream content directly to YouTube, Facebook, Twitch, and other popular platforms via the OBSBOT Start App;

• use the optional USB-C to Ethernet Adapter to incorporate the camera in an NDI-based environment.

While this is quite a lot for a consumer camera, I soon found out that you can only use one of these modes at a time. So, for example, enabling UVC over USB-C will disable NDI and the HDMI output, and vice versa.

Controlling the PTZ camera

When it comes to remotely controlling the Tail Air PTZ camera, you have two main options. The first method is pairing the camera to the OBSBOT Start smartphone app via Wi-Fi or Bluetooth. I found the app to be well-made, stable, and pretty straightforward to connect and use.

From within the app, you can monitor the live image feed, manually adjust camera settings, playback clips and save them to your phone’s gallery, manually control the gimbal using a virtual joystick, or enable/disable the AI tracking and framing features (more on this later).

The second method for controlling the camera is by using a dedicated remote controller. However, connecting the two devices requires you to follow a very specific pairing process outlined exclusively in the user manual. Once that’s out of the way, you can remotely control the gimbal via the controller’s physical joystick, use its side zoom rocker to digitally zoom in/out, trigger record, activate auto tracking on humans and pets, or turn the camera on/off. The remote charges via USB-C, works in the Bluetooth range, integrates a laser pointer for presentations, and can be used to control up to 3 Tail Air units.

AI-powered tools – Auto Tracking, Director Grids, and Gesture Control 2.0

Ok, it’s finally time to talk about the most interesting features of the OBSBOT Tail Air, which are also the ones that impressed me the most: its AI-driven tools. At first, I was afraid that the brand relied on Artificial Intelligence as a buzzword to steal their audience’s attention. But after playing around with the camera for a while, I was really surprised by the performance and utility, especially for a solo shooter.

• Auto Tracking: The camera can automatically detect and track humans and pets, but you can also manually select an object by drawing a box on the screen. Unfortunately, my dogs weren’t very collaborative, so I ended up being the only test subject. At first, I wasn’t really happy with the results I was getting – especially in terms of image composition – but then I found out I could adjust tracking speed (slow/standard/fast) and even tell the camera where I wanted to be positioned inside the frame.
  
  After making these changes, things got much better. Especially, since the maximum controllable speed of the gimbal is 120°/s, I was impressed by how quickly I could move in space. However, this also revealed the other side of the coin, which is a pretty bad rolling shutter.
  
  There are still situations in which the tracking system can get a bit confused. That happens, for example, when a large obstacle gets in the way, or your shot includes multiple human subjects. Still, the performance of a product in this price range is quite remarkable and should provide satisfying results for most users, even in dimly lit rooms.

• Directors Grid: Using AI, the camera automatically creates multiple angles of your subject. These can be monitored from a dedicated tab within the app, and you can quickly select which angle goes “on air”. Essentially, it’s like having a virtual live switcher at your fingertips.
  
  The biggest limitation of this tool is that some of the angles are so cropped that they become pretty much useless unless you’re OK with low image quality. Sometimes “content is king”, as they say.

• Gesture Control 2.0: Now, this is where things get interesting. Do you remember I said there are two ways of controlling the gimbal – app or remote? Well, “good things come in threes” and, in truth, you can also control the gimbal remotely using only… your hands.
  
  Hence, by performing specific gestures in front of the camera, you can start/stop recording, engage Auto Tracking (this can also be done by double-tapping on your subject in the app), or zoom in/out on yourself. Depending on the command, the front LED indicator changes color and the camera emits a specific sound.
  
  This worked very well in my experience, and it can be very useful for streamers and presenters who need to shoot content on their own, but don’t want to have a smartphone or remote controller on them.

USB-C to Ethernet Adapter for NDI workflows

The NDI protocol allows you to share video sources over a local area network (LAN) via Wi-Fi or Ethernet, and the OBSBOT Tail Air is fully ready to support it. However, this comes at a price. Indeed, you will need a paid activation key to enable NDI on your device. Additionally, if you want to rely on a more secure wired Ethernet connection, you also need to purchase the optional USB-C to Ethernet Adapter.

Once you’re all set, you can share a clean, low-latency 4K video feed over your network – the 4K image looks pretty crisp. As usual, activating NDI mode will disable the HDMI output. The NDI video source can then be detected by live streaming software like OBS Studio, for example. This type of workflow is particularly useful if you’re planning on including the Tail Air in a multi-camera live streaming production with other NDI sources.

Although I don’t often rely on NDI technology for my productions, connecting the camera to my LAN using the adapter was fairly easy.

Audio recording

The OBSBOT Tail Air can record audio out of the box thanks to its dual built-in microphone. Nonetheless, I found this solution to be barely usable, since the internal mic also picks up a lot of undesired sounds generated by the camera and gimbal. This is particularly true when tracking subjects.

Luckily, you can easily work around this by connecting an external microphone to the camera via its 3.5mm jack input. For my tests, I used a SYNCO G2 (A2) wireless microphone system and I achieved much cleaner results. I think this is the way to go if you want your live streams to sound professional.

Audio levels can be easily adjusted via the app, and there’s also an option to apply noise reduction to the signal. This can come in handy to improve the sound coming from your internal mics, but in any case, I wouldn’t set it beyond 1 on the 3-step noise reduction range.

Conclusion

So, should you buy the OBSBOT Tail Air? As usual, the answer depends on your specific needs. If you often need to shoot yourself walking around your space during live presentations, and you don’t have high expectations of image quality, then I think the AI Auto Tracking and Gesture Control features can help you tremendously in your job.

Overall, the system can be a valid option if you’re just starting out with live streaming and you’re looking for an end-to-end solution that doesn’t require too much effort to set up and use. The camera produces a nice 4K image (you can see a comparison with the built-in FaceTime HD camera of my MacBook Pro M1 Max – 2021 version – below), although image quality rapidly degrades when you zoom in. Also, by using an external microphone, you can still deliver professional audio.

However, if top-tier image quality is a concern and you often need to zoom in/out on your subjects or stream from a low-light environment, then I’d probably look somewhere else.

Price and availability

The OBSBOT Tail Air is now officially available to order for $499. Pricing for the accessories and NDI activation key is as follows:

• Smart Remote Controller –
• 360° Rotation Charging Base –
• ND Filters –
• NDI Activation Key –

For more information, please visit OBSBOT’s website here.

Do you often produce live-streaming content? Do you think the AI tools of the OBSBOT Tail Air can facilitate your work? Let us know your thoughts in the comment section below!

We already covered the GFX100 II extensively by conducting an exclusive interview with the leaders of the product planning team, and a full camera review was published as well. In my review, I highlighted the diverse recording formats, resolutions, and the ability to record to an external SSD, next to a completely new feature that allows 2x anamorphic recording. (Not only de-squeezing in preview as other cameras do). Well, specifications aside, how does it feel to run & gun with the camera?

Missing in my review

What’s missing in order to complete the task of “knowing the camera” better is a lab test (coming up very soon), engaging footage using a 2x anamorphic lens (to explore the 2x anamorphic recording), and last but not least, a “normal” mini-documentary like we usually do to really “feel the camera” and observe how it behaves in the field.

Back in summer, I met with Yokoe-san who was kind enough to allow me to spend a few hours together with him. Yokoe-san’s goal is to “bring Japanese nature to your living room” by creating small and beautiful terrariums. (More details about his company can be found here).

My idea was to deploy the new FUJIFILM GFX100 II in its most minimalistic way, meaning, mostly handheld (I used a tripod only for the main interview). I set the camera on its highest recording quality (WITHOUT cropping the image) to “Cine 5.8K”, which, in a way, is my favorite because of its given 2.35:1 aspect ratio. My main lens for this shoot was the FUJIFILM GF 35-70mm lens. It’s not a fast one by any means, yet it’s a relatively compact, good all-rounder lens that also happens to be modestly priced.

GFX100 II summary

As I delved deeply into various filming modes, resolutions, and more in my review, I also highlighted the camera’s ergonomics. This aspect is extremely important for me when running & gunning. If I have to summarize it all, I have to say: I truly like what comes out of this camera! That being said, I wish that the internal noise reduction could have been further controlled beyond the available “-3”. Noise in the footage is not necessarily always a bad thing as it can greatly contribute to the picture aesthetic most of us are after. In this regard, I sometimes find the camera to be “too smooth”. In terms of handling, operators with large hands will enjoy using this camera.

All in all, FUJIFILM did a remarkable job in creating a “Swiss Army Knife” camera, one that can be used in so many kinds of productions. (The versatility of this camera is really outstanding). Next, I would love to see FUJIFILM opening up to third-party lens manufacturers by allowing them to license the FG mount protocol. In doing so, they could offer the GFX line of camera users a much broader lens selection, essential for the success of the camera format.

Before closing this article, I would like to thank Ayumi Shimaura-san for connecting me to Yokoe-san and assisting me with translation. I would also like to thank our editor, Jay Choma for doing a marvelous job in editing a small, lovely story from the ordinary footage I shared with her.

Tech info: Filmed with FUJIFILM GFX100 II, 5.8K/24p, ProRes HQ, F-Log 2. LUT by fylm.ai

What do you think about the FUJIFILM GFX100 II? Did you ever consider using a large format camera in your productions? Please share with us your comments in the section below.

Tradition of innovation

Sony is well known for its technological ingenuity and prowess. The first manufacturer to launch a full-frame mirrorless camera (the α7 in 2013), the first to incorporate a stacked CMOS sensor into a hybrid camera (the a9 in 2017), a market leader in the field of autofocus, and, as of yesterday (Nov. 8th, 2023) – the first manufacturer to put a full-frame global shutter sensor in a mass-market camera – the Sony a9 III. The new camera is undoubtedly impressive, but to reign supreme, it has to pass over some serious hurdles among its stablemates.

Power trio – the a1, a7S III and the new a9 III

Sony’s lineup offers many excellent video cameras catering to various use cases. Here we’ll discuss the aforementioned trio, which, in my humble opinion, represents the pinnacle of the Alpha line regarding motion capture. They all have significant features that render them all adequate tools for video. Among those we find:

Similar features

• 4K recording up to 120p
• 10-bit 4:2:2 internal recording, up to 600Mbps
• 16-bit RAW video output via HDMI
• Full-size HDMI-A port, mic/headphones 3.5mm jacks, and smart hot shoe connectivity
• In-body image stabilizer and electronic stabilization options
• Dual CFExpress type A / SD UHS-II card slots

Unique traits

Each of the trio possesses some unique abilities. The Sony a1 has a massive resolution advantage. Its 50mp sensor enables 8K recording at normal speeds. The 4K can be subsampled from the entire width of the sensor, or oversampled from a Super35 5.8K area. This allows for some level of flexibility, but may also degrade the full-frame 4K sharpness.

The Sony a7S III is the most affordable camera. The use of a relatively low-res sensor may render it unusable for many stills applications, but we’re talking video here. Though limited to 4K recording without much oversampling, the enormous pixels on the 12mp sensor create a satisfying 4K image that won’t change much when upping the frame rate. Those large pixels are also responsible for the impressive high-ISO capabilities the a7S III is famous for.

The new Sony a9 III is the master of speed. Its impressive abilities all rise from its core feature – the global shutter-enabled 24.6MP full-frame sensor. Its instantaneous read speed enables 6K oversampled 4K video up to 60p resulting in sharp, clean images. It also goes the extra mile with its uncropped 4K/120p. The autofocus system is Sony’s most recent AI-based algorithm, but I suspect that the faster readout flow will make it even better (is that even possible at this stage?)

Three different sensor architectures

As mentioned – the a1 boasts a 50MP sensor. The more pixels a sensor has, the slower its readout will be (as long as we compare equivalent sensor generation and architecture). This constant poses a challenge to a camera that aspires to rule both stills and video. The solution is a technological one: a Stacked CMOS. This specific architecture, which debuted on the original a9, speeds up the read speed, enabling quite a decent 16.6ms in 8K. While solving this specific problem, the use of such tech did result in one of the most expensive hybrid cameras around.

The a7S III takes the opposite route regarding resolution with its 12MP sensor. Having fewer pixels may seem like a disadvantage at first, but it makes the camera much more affordable and theoretically less prone to overheating. The fewer pixels the camera has to sample the less processing power it requires. Additionally, a low-res sensor reads faster since it has fewer pixels to read. The a7S III achieves very good read speed with a BSI CMOS. It’s actually among the fastest in the industry, all while omitting the higher, more expensive technologies, and significantly reducing the camera’s cost.

The new a9 III takes it one step further with the use of the “holy grail” of fast sensors – a global shutter. As mentioned before – the global shutter reads each and every pixel all at once. This unique chip architecture completely eliminates any strobe synchronization effect, flash sync speed, or the infamous rolling shutter effect. Working at extreme 120fps speed, this loads the camera’s processor with a constant flow of information, optionally improving other data-related abilities such as tracking auto-focus and subject recognition (though we’ll have to wait for an actual test to see if theory translates to practice).

What is the deal with global shutter in real life?

Few cameras use global shutters, and the a9 III is the first stills-oriented camera to use one (excluding Hasselblad’s mechanical global shutter, but that’s another story for another time). Regarding how unique this sensor is among stills cameras, it’s no surprise that Sony emphasized still imagery over video at its launch event.

When it comes to motion capture, there’s one clear advantage: no matter how fast the camera or the subject moves – no distortion will take place. No matter how flickery your light source is – no banding will occur. Global shutter provides some extra peace of mind and lets you focus on the creation itself.

But is it that good?

Well, it is. It’s that good. The advantages are absolute. But are they really necessary? Well, that’s a bigger question. First, let’s talk about the prices of a global shutter. The most obvious is, well, the price. The a9 III will set you back $5999. This kind of money could buy you about two a7S III cameras, or maybe a nice lens, some accessories, etc.

Other potential prices are speculative at this stage, at least until we get a production camera and run it through our rigorous lab tests. One may be the dynamic range. Sometimes faster readout speeds come at the cost of dynamic range. Other times, the results aren’t as decisive (as seen here with the RED KOMODO 6K, another global shutter camera). Another issue that can occasionally affect hybrid cameras with high data rates is the well-known problem of overheating. The amount of data crunched by the a9 III is extraordinary, and it probably uses some sort of passive thermal management. We can’t yet do the math, but I won’t be surprised to see some limitations regarding its high-speed settings.

TL;DR, which is the best?

Well, as always – it depends. The new a9 III brings some impressive features and abilities to the table, but when looking closely, most are quite niche. If your project brief includes extreme action shooting, fast-moving subjects, and chaotic situations, the a9 III may be the best camera. The a1 will provide better all-around features, covering most normal use cases for either stills or video (and then some). The a7S III is such a popular camera for a reason. It caters to most video needs and does so with excellent efficiency.

So are you excited about the new a9 III? Will it turn your creative process upside down? Or maybe another a7S III is a better option because if it ain’t broke don’t fix it? Let us know in the comments.

This new Micro Four Third camera is much anticipated by MFT format lovers, and I’m certain that this new model will satisfy the needs of many. 

Built on the success of its predecessor, the G9II kept its relatively small size and comfortable hand grip, but other than that, it is safe to say that this is a completely new camera! 

So what can you expect from such a flagship model equipped with a newly designed 25-megapixel sensor and the same processor that can be found in the LUMIX S5II? 

Before I continue, I would like to highlight two things: 

– My brief experience and initial impression of the LUMIX G9II is based on a pre-production model and as such, not everything went smoothly. I have no doubt that Panasonic will continue to refine this product and enhance its performance up until it sees the light of day in October.  

– The G9II is Panasonic’s MFT sensor size flagship camera used primarily for taking photos. Panasonic has another MFT sensor-size flagship camera, the GH6, for more demanding video work. (For example, an internal fan and a CFexpress card slot might be valuable for those who are using the GH6 professionally).

Nevertheless, the video capabilities of this new camera have not been neglected, and in fact, depending on your professional needs or type of hobby, perhaps this new LUMIX G9II might be the camera you have been waiting for! 

The original Panasonic G9 was released more than five and a half years ago and proved to be extremely beloved and popular.

Now, here comes the new camera, not only with a fresher look, but you can tell Panasonic was very ambitious to make it even better!

I already mentioned the new sensor and Venus processor. What else? 

For the first time, both photographers and filmmakers will appreciate having a Phase Detection Autofocus System with subject recognition in an MFT camera.

The OLED EVF is also new and larger than the previous version, with a 120fps refresh rate. 

The same goes for the 3-inch, 1.86 million dots LCD screen. 

Filmmakers will appreciate the large HDMI output connector and even more importantly, the ability to record ProRs HQ in up to 5.7K externally to an SSD. 

Personally, I would like to recommend a slightly different solution for external recording. Instead of recording to an SSD, I successfully used a small CFexpress reader made by PAN’S SCHEME (review here) together with CFexpress cards from Angelbird or Nextorage (review here).

However, there is a downside to recording ProRes externally – I’m talking about the camera battery, which runs out faster this way. I couldn’t measure by minutes, but knowing how well Panasonic’s batteries usually hold up, I was surprised to see how relatively fast they drain. 

So to summarize: For recording with the LUMIX G9II in the highest possible resolution and best available codec, one can set the camera to 5.7K in up to 30p in ProRes HQ externally. When it comes to internal recording, 10-bit, 4:2:2, All Intra in up to C4K 60p Is available.  

If you are not so picky about color sampling, 5.8K in up to 30p in 10-bit Long GOP 4.2:0 is also available. 

4K at 120fps can be found too, but I haven’t personally tested it. If you are still seeking the highest possible HD recording quality, internal ProRes is available as a solution.

Other unique features include a completely new LEICA Monochrome preset setting, next to “real-time LUT” recording, which allows importing your favorite look into the camera. By the way, this exact feature is borrowed from the LUMIX S5II. Needless to say, the V-Log option is there for creative work as well. 

As with Panasonic cameras in general, the G9II stabilization works very well, although I advise you to carefully watch the focal length you are using when working with it. Wide shots tend to wobble a bit at the edges. 

One thing that bothers many of us when filming is how well, or not, a camera can handle overheating. Well, I did not face any issues, but further experimentation is definitely warranted.

OK, let me touch on one of the biggest, if not the biggest new feature in the Panasonic LUMIX G9II: I’m talking about the Phase Detection Autofocus. It is the first time that Panasonic has implemented this feature in an MFT camera, and I’m sure you are curious to know how well it works. 

In all honesty, with the pre-production camera that I had, I got mixed results. Let’s say that 90% of the time the continued autofocus worked well, and for the rest, it was a bit of a hit-and-miss up to the point where I had to switch focus to manual mode. 

Again, I know that Panasonic is very dedicated and working hard to improve the new Phase Detection Autofocus in this camera. 

There’s another aspect related to focusing with this camera that I’d like to mention, and that’s the “joystick.” I find that one of the easiest methods of changing focus between objects while running and gunning is to simply use this knob.

The thing is that it is extremely sensitive, and the pace of moving the focusing box between objects cannot be controlled easily, which can result in a few missed focus points. 

While working with the camera, I also had the chance to combine it with the Panasonic Leica 100-400mm lens. At the time of testing both products, the camera and lens did not communicate well, so handheld filming was not possible. But the potential for run-and-gun wildlife photography and filming is absolutely there. 

By the way, if you are a hybrid shooter, make sure to store and assign your favorite settings to a “C” option as unfortunately, the camera will not remember separate settings like picture profiles when changing between video and photo modes. 

So, who is this new camera actually for? From a video perspective, I can easily see small news agencies using it next to independent creators who shoot all kinds of different content but are mostly out in nature. In so many ways, the MFT sensor size is the golden ticket for high resolution, magnification, and camera size. Top this with a zoom, such as the 100-400mm lens, and the “eye of the lion” couldn’t have gotten any closer. 

But here comes the pricing question. The Panasonic LUMIX G9II will set you back 1899 Euro (VAT included), or $1,897.99, and in this price range, it looks as if it was made to compete directly with the OM-1. 

The question remains as to whether this camera is enough to successfully compete against other APS-C sensor cameras that have become smaller and even comparably cheaper over the past years. (The FUJIFILM X-T5 and X-S20 are good examples). 

In the end, I’ll let you guys decide, and I’m very curious to hear your thoughts about the new G9II. Is this the LUMIX flagship camera you have been waiting for? Go ahead and share your comments with us in the section below. 

So, a new day, a new camera, or are we witnessing something special here?

Well, in all honesty, my hat is off to FUJIFILM for emphasizing and enhancing the video capabilities of this new camera. 

True, this is still a hybrid device, which led to certain compromises such as the lack of an articulated LCD screen, but all in all, FUJIFILM did a great job here and I’m extremely curious to see how the market will react in adopting it. 

GFX100 II – what’s new

So let’s start by saying that the GFX100 II has an entirely new sensor, although it kept the same 102 megapixel count as its predecessor. 

Where there is a sensor, there is a processor, and the one found in the GFX100 II is the same X-Processor 5 that drives all the other latest cameras from FUJIFILM.

So let’s start by saying that this camera has a huge selection of formats in up to 8K resolution, with different image circle coverage and recording codecs. 

Out of the box Best Settings

If you are looking for the best internal recording codec, simply use ProRes HQ. Uncompressed 12-bit external recording output via the large HDMI connector is supported as well, of course, but for the first time ever in a FUJIFILM camera, ProRes can also be recorded externally via USB-C to an external SSD drive. 

If you are looking for the fastest sensor readout to minimize the rolling shutter effect, choose 4K/60p, which is thankfully NOT cropped. 

If resolution is your top priority, then choose the 5.8K recording option. Horizontally, the camera will use the full sensor width and will crop vertically a bit so the end result will look like a 2:35:1 CINE aspect ratio. 

Now, if absolute resolution is the thing you are after, you can of course use the 8K resolution that this camera offers, but take into account that the image will be cropped by 1.24x. 

From my experience, the image still looks nice and crispy even with this method of cropping into the sensor. 

By the way, if you take a close look at the menu, you will find a 3×2 recording option, but mind you that this is NOT a true “open gate”, but rather another recording format choice. 

With this camera, FUJIFILM created a new menu option to choose between four filming formats: GF, Premista, 35mm (which is equivalent to full frame), and anamorphic. Depending on the format you choose, the camera will adapt itself to automatically adjust to the sensor portion that can be recorded. 

Speaking of which, it is the first time FUJIFILM has taken anamorphic filmmaking seriously and enabled two crucial features.

The first is simply to de-squeeze the image for viewing purposes like many other cameras do. 

The second is truly a breakthrough, as the camera will allow recording in 8K (cropped) with 2x anamorphic lens, and de-squeeze the image while filming! After checking with FUJIFILM, I can confirm that this option will give the best anamorphic recording quality, and I’m looking forward to trying it out in the near future. 

If different squeeze factors are required, it will have to be manipulated on your favorite NLE. 

OK, let’s tackle other features. Let me start with the new “Autofocus Wide Tracking”, which I find extremely useful for documentary work. Simply place the focus square on the object you want to track and the camera will do the rest, most of the time successfully. 

Another improvement is the ability to connect the FUJIFILM fan to the camera in order to combat the possibility of overheating. Personally, I did not encounter any issues, but for those who are working professionally, I recommend considering this device. 

I also find the camera works very well in low light conditions, so using higher ISO values is not an issue. 

For those who are filming in F-Log-2, please be advised that the minimum ISO can be set to 800. By the way, in the menu, there is one very interesting feature – “F-Log 2 D Range Priority”. Apparently, it is supposed to improve the camera’s Dynamic Range, but at the expense of a greater rolling shutter. This option won’t be available in all recording formats and frame rates. 

We will, of course, test this option in our lab test when we get the camera with the final firmware installed. 

The EVF is large and clear, but what FUJIFILM wisely added is the possibility to control the magnification ratio and reduce it to a manageable size so that the entire frame is visible at once. 

For control freaks, the SUB menu is now very customizable between the photo and video modes. I really appreciate the level of choosing functions for viewing that FUJIFILM allows here. 

If you haven’t noticed, the new GFX100 II is much smaller than the original GFX100 camera as FUJIFILM managed to shave off the built-in battery grip. Not only that but there is no need to buy an additional transmission grip for sending material to the cloud as FUJIFILM managed to squeeze it all into the body, keeping the camera size and weight as low as possible. 

All those high-resolution formats deserve a more advanced recording media, so now one recording slot is dedicated to accommodate CFexpress Type B cards. 

For those who a seeking a Timecode solution, jam sync via ULTRASYNC Blue can be obtained.

Last but not least, a new film simulation setting was added, the REALA ACE. At the time of writing this review, I was not able to test it.

So, no camera is perfect and maybe the two main concerns I have are the following: 

I’ll start with the camera grip. For people with a small hand like mine, the camera is a little uncomfortable to hold. The grip is simply too big. 

But more importantly, I would like to mention the image quality. It is really beautiful, BUT, in my opinion, it is “too clean”. It seems as if the noise reduction is working hard to “smooth things”. True, you can dial down the noise reduction to minus 3, but I wish that the results could be more visible. 

Not all types of noise are harmful, and in my opinion, a bit more “picture texture” is missing. 

In addition to that, regarding future GFX offerings, I hope FUJIFILM will consider the following:

• Implementing a higher frame rate in 4K resolution, as currently it is limited to up to 120 frames per second in full HD 

• Having a fully articulated LCD screen 

• And unlike the modestly priced X-S20, in this camera, subject detection cannot be set to “Auto”. In my opinion, this option is sorely missed

• The in-body stabilization system is generally fine, although I wish it could be even more prominent. 

But maybe the biggest request that I have is to take the inner bit of this camera with all of its functionality and try and implement it in a dedicated cinema camera body. 

I’m sure that FUJIFILM can only gain from finally having an “A camera” on the market. 

Last but not least, the new GFX100 II will set you back $7,499/€6,789 which is way less than the original GFX100 and $1500 USD more than the GFX100S, but at the end of the day, it is a different beast altogether with much higher and more refined video capabilities. 

Now, I’ll be interested to hear what you guys are thinking about the new GFX100 II. Is medium format something that you’ve been considering in the past, and is this new camera piquing your curiosity to finally try one? Please share your thoughts in the comment section below.

The last drone from DJI’s Air series, the Air 2S has been on the market since 2021. Now, the new generation has arrived. I had the chance to test the drone for a few days before its launch and I will share my findings with you in this article and video. If you want to see all the specs and features of the new DJI Air 3, head over here to my other article.

Body design, weight, battery

Apart from the dual-camera head, what strikes you first about the new Air 3 is that it has gained some weight. The body has been redesigned to match the look of the new Mavic 3 series, and it is now noticeably larger than the previous Air series drones. In fact, every new drone in the DJI Air series increased the takeoff weight somewhat.

• DJI Mavic Air (released in 2018) – take-off weight 430g
• DJI Mavic Air 2 (released in 2020) – take-off weight 570g
• DJI Air 2S (released in 2021) – take-off weight 595g
• DJI Air 3 (released in 2023) – take-off weight 720g

Honestly, I think the weight increase is a good thing. First of all, a heavier drone is more resistant to wind. Second of all, more weight means the battery pack can be heavier and this results in more air time – DJI says up to 46 minutes.

DJI Air 3 comes with a C1 label in Europe

Thankfully, the increased weight of the DJI Air 3 still does not put it in a different category when it comes to drone regulations. In Europe, the Air 3 comes with a C1 label, which allows for a maximum take-off weight of 900g and puts the drone in the A1 subcategory. For instance, the recently announced DJI Mavic 3 Pro slightly exceeds the 900g weight and therefore has to have a C2 label in Europe. This puts it into the A2 subcategory that brings further restrictions and requirements. This is all part of the new drone regulation that will be fully effective from January 1, 2024. In the United States, according to the FAA, the rules are the same for all drones between 250g and 25kg (55 lbs).

The DJI Air 3 now has omnidirectional obstacle sensing. There is a sensor on each of the four corners of the drone and a few more sensors along with an LED light at the bottom. Apart from all the sensors on its body, the Air 3 has an internal fan, so it will not overheat during firmware updates or longer idle periods. For my DJI Mini 3 Pro, for instance, I ended up buying a small fan to keep it cool while updating firmware.

Cameras – two usable focal lengths and 4K 100fps

Now, let’s finally talk about the most interesting part of the new Air 3 – the cameras. For the first time in DJI’s Air series, there is a dual camera system. There are two sensors and two lenses in the gimbal-stabilized camera head – 24mm f/1.7 and 70mm f/2.8. The minimum focus distance is 1m for the wide-angle camera and 3m for the telephoto camera.

I think these are both very usable focal lengths and another great thing is that both sensors are the same size, same resolution, and capable of the same framerates. These are both DJI’s stacked 1/1.3″ type 48MP (8064 x 6048) sensors. It is the same sensor that is used in the Mini 3 Pro, and the wide-angle camera of the Air 3 even uses the same lens as the Mini 3 Pro.

Traditionally in the Air series, the aperture on both cameras is fixed, so to keep the shutter speed near the 180° rule, ND filters are a must. DJI is selling a four-piece ND filter set with ND strengths of ND8, ND16, ND32, and ND64. Unfortunately, I didn’t have this set for my review, so the shutter speed in my clips is all over the place.

Compared with the previous DJI Air 2S, which was capable of a 5.4K video with a larger one-inch type sensor, the maximum video resolution with the Air 3 is UHD 4K. In normal mode, 4K is available in 24, 25, 30, 48, 50, and 60 frames per second.

In slow motion mode, 4K UHD is available at 100fps and FHD at 200fps. Here I need to mention that slow-motion clips are slowed down in the drone and saved as 25 fps clips, so keep that in mind if your project has a different base framerate.

The bitrate is limited to 150Mbps in H.265 codec wrapped in an MP4 container. Some modes are also available in the older H.264. From my understanding, in video mode, the drone uses a binned 12MP version of the sensor and there is no oversampling.

D-Log M with 10-bit color, sharpness, and NR settings

Apart from the normal color mode, each camera can record in a D-Log M profile with 10-bit 4:2:0 color. I shot most of my clips in D-Log M. When grading, I tweaked white balance, brightness and contrast, and then used the old D-Log M to REC.709 LUT from DJI, which was created for the Mavic 2 Pro as I could not find any newer version of the LUT. I found this LUT to be too heavy on the saturation, so I invariably had to tune it down a bit. I hope DJI will provide a new, more accurate version of the LUT for each new drone model that supports the D-Log M.

When I was flying the DJI Air 3 for my review, I could not change the sharpness and noise reduction. This is something I like about my Mini 3 Pro as I can tune these settings down and deal with noise and sharpness in post-production if needed. DJI told us these settings will be available with the final firmware, so this is good news.

Comparison with DJI Mini 3 Pro

Talking about the Mini 3 Pro, I tried to compare the D-Cinelike from the ultra-light drone with the D-Log M from the Air 3. I used the same settings and filmed a landscape side-by-side. Comparing the ungraded clips, I find the D-Cinelike from the Mini 3 Pro slightly darker. It is also visible that the Mini had the noise reduction and sharpness turned all the way down. After grading, I can hardly see any difference at all.

Adjustable maximum speeds, upwards tilt

Unfortunately, unlike mobile phones that have multiple cameras, it is not possible to switch cameras while recording a clip on the Air 3. You have to end the clip, switch the cameras, and then start recording again.

Two more things I like about the new DJI drones, and Air 3 is no exception:

• First of all, the camera of the drone can look up. It can be tilted up to 60° upwards, which enables more creative angles.
• Secondly, it is possible to change the maximum flight speed of each mode (cine, normal, and sport). This is useful for achieving very slow movements with the cine mode.

Vertical video, photos

If vertical video is important for you, the Air 3 can film up to 2.7K 60fps vertically. However, unlike the Mini 3 Pro, the Air 3 camera cannot physically rotate by 90° to better capture vertical footage. Instead, it uses the full height of the 4:3 sensor and will therefore give you more vertical resolution and field of view than simply cropping from a horizontal 16:9 video.

To demonstrate this, I made a simple FOV comparison test. You can see a still photo, horizontal video, and vertical video that I made while hovering in one spot. The same mechanism applies to both of the DJI Air 3 cameras.

When it comes to still photos, both cameras can shoot up to 48MP photos in either JPG or DNG raw format. There are also some automatic panoramic modes, but I did not test these as I prefer flying manually for full control.

Night mode

The Air 3 also offers a so-called night mode for video, but it is limited to a maximum of 30fps and is only available with the normal color profile, so only in 8-bit. It allows the ISO to be set all the way up to 12.800. I tried the night mode at ISO 3200, but the image quality was not very good as the drone’s noise reduction heavily processed it.

I would honestly stay in the normal mode and use the D-Log M 10-bit mode. Higher ISO footage is, however, generally not good with this drone. If you need to film in low-light environments often, I would opt for one of the Mavic 3 series drones with their micro four-thirds cameras.

New RC2 controller

The new DJI RC2 controller is very similar to the DJI RC. It features the same button layout and a bright 5.5″ screen. What is new are these two external antennas that can be optimally positioned depending on the drone’s location. The remote has 32GB of internal storage that can be expanded with a microSD card.

The only thing that bothered me was that I could not hide the timer of the screen recording, I could only move it, but it is still visible in the screen recording itself. Also, the DJI Fly app was constantly showing me notifications to update FlySafe, but it was already up to date. I suppose this is only a beta app bug that disappears with the full release.

Charger – power accumulation

The battery charger now has a special function called power accumulation and I think it might be useful in certain scenarios. The charger has the ability to consolidate the remaining power of all three batteries into the one with the highest charge, resulting in extended flight time for that particular battery. It accomplishes this without needing to plug anything in.

O4, ActiveTrack, Waypoint, Price, etc

The Air 3 has a new O4 video transmission system, which extends the maximum reach and makes the signal once again more robust. Very interesting is also the option to plan a route with waypoints, save it, and let the drone fly the route automatically and repeat it whenever you want. The drone again offers a variety of smart features such as ActiveTrack. It should be able to maneuver itself through terrain and follow the selected subject.

Conclusion, price, and availability

To summarize my findings, the DJI Air 3 is a really versatile drone. It offers two very usable focal lengths on two cameras that can easily match each other, thanks to the same image sensors. The flight time is very good. I am a big fan of the 4K 100fps mode. Because of the smaller sensors, however, the low-light performance is not great. Also, the competition is tough – there is the ultra-lightweight sub 250g Mini 3 Pro with the same wide-angle camera as the Air 3 and true vertical mode. Then there is the DJI Mavic 3 classic with a much better main camera. I would say the Air 3 is the king of versatility and I think the C1 label will further increase its popularity in Europe.

The DJI Air 3 is available now and there are three different configurations:

• DJI Air 3 with RC-N2 costs $1,099 (around €1,118.24 with VAT in Europe)
• DJI Air 3 Fly More Combo with RC-N2 sells for $1,349 (around €1,349 with VAT in Europe)
• DJI Air 3 Fly More Combo with RC 2 sells for $1,549 (around €1,549 with VAT in Europe)

As always, there is a wide variety of accessories (such as ND filters, wide-angle lens adapter, car charger, etc.) that can all be purchased separately.

What do you think about the new DJI Air 3 drone? Is it a worthy successor of the Air line? Let us know in the comments section underneath the article.

We have covered trade shows since forever and wanted to believe that we had learned a thing or two about how to bring the latest and greatest filming industry news to our audience in an unconventional way. Throughout the years, “things” of course have changed, but not fundamentally. When I say “fundamentally,” I’m referring to the core aspects of our workflow during show coverage. It starts with capturing footage and then handing over the recorded memory cards to one of our colleagues. They quickly rush to the press room to offload the card. Meanwhile, we have a dedicated editor stationed at the location, working diligently under time constraints to assemble video pieces. They also handle the task of writing corresponding articles and packaging everything for publication on our CineD website and YouTube channel. But then, the opportunity to work in a different way arose, and this is what I would love to share with you today.

Indeed, our team has been traveling to industry-related shows for years. Still, we never took the time to document ourselves on those costly long trips. For those who are interested in knowing how much effort goes into such excursions, here is a glimpse of our reality. After all, it is our duty to publish “polished” finished news pieces where videos need to be in focus, the audio needs to be clear, and maybe some engaging B-roll footage needs to be taken. Sounds easy, no?

Well, not so fast…What started years ago as a “stroll in the (show floor) park” became a demanding, structured task.

FUJIFILM X-H2/S cameras

FUJIFILM’s latest batch of high-end cameras, the X-H2 (review here) and X-H2S (review here) can shoot “directly to the cloud” thanks to the collaboration with Adobe’s Frame.io. In fact, the X-H2 was our camera of the year 2022, and along with the many reasons for liking it back then, being able to execute a Camera to Cloud workflow is another good reason to look at this camera again. Speaking of which, in our databases, you can find a lot more information about how the X-H2 and X-H2S performed in our lab tests.

FUJIFILM Camera to Cloud option for video

There is a FUJIFILM / Frame.io dedicated integrated Camera to Cloud option for photographers, but as the title says, I would like to concentrate on the Camera to Cloud option for video.

Let’s start by describing the basic idea and why it is so intriguing. Imagine the possibility of filming ANYTHING you want, and as soon as you press the stop button, files can be instantly uploaded to a “hub”. From there, your remote editor can actually start importing the footage directly to his/her timeline and start editing.

Well, we are still not entirely there as you will see later in this review, but the direction is certainly promising!

So we have the 2 FUJIFILM cameras, and in addition, one needs need to have a File Transmitter Grip, the FUJIFILM FT-XH to be exact. This device will improve the networking capabilities of the cameras with fast Wi-Fi (speeds up to 600 Mb/s) and wired LAN connectivity. Needless to say that the FT-XH is also serving as a much-needed battery grip, and can accommodate two additional NP-W235 batteries. (Believe me, you will need this extra power when leaving your camera “ON” for a long period of time while uploading the files).

Connecting to the Cloud and workflow

So, the first recommendation is to download the frame.io App. Then, set up your file transmitter (through the camera) to “talk” to your frame.io account and, of course, recognize the wifi network you are riding on. I know that one of the first questions asked is: “Is it secure enough to use”, especially when considering that the original video files are being uploaded? In all honesty, it is beyond my technical knowledge to judge if this connection between the camera and cloud (frame.io) can be hacked, yet both Adobe and frame.io are TPN certified (Trusted Partner Network – more about it, here).

Now that the FUJIFILM camera is connected to the cloud through a process that takes a few minutes, from that point onward, there is not much to worry about if all goes well. The thing is, every time there is an interruption in the connection, you will have to reconnect the camera again, but at least you won’t have to go through the entire setup process.

Seeing the “magic” happening is joyful. Stop recording, and the video will is displayed in your frame.io account almost instantly. This also means that anyone from your team, regardless of where he or she is, can see those files and immediately start working on them (watching, logging, or editing). One important thing to remember is that the uploaded video files are NOT full-resolution, in case you were filming in 4K resolution or above. Those will be recorded into the CFexpress card while what is being uploaded are 1080 proxy files. For some instant work, those will be enough in terms of quality to release on YouTube, but as we tend to upload 4K versions of our show coverage video, Rachael, our editor, re-linked the video files to the original 4K ones.

Like us, I guess that one of the first questions you are asking is: If the transfer of video files is interrupted, then what? Well, as Victor Ha from FUJIFILM describes it, the system was designed to pick up from where the interruption occurred. In reality, it worked out flawlessly, and we gained confidence in the camera and file transmitter knowing that even if we were not uploading any more “live” files, when the connection resumed and the camera was “ON”, the process of uploading would continue without having to worry about it.

Hiccups on the show floor

While FUJIFILM did everything they could to supply us with a dedicated wireless connection (next to a dedicated frame.io account that could handle the amount of footage we were uploading without any limitations), the NAB show floor is simply “hard to bite” when it comes to a smooth “wifi ride”. With so many parallel wifi setups and DMX controllers working in such a small, crowded space, in no time we started to face issues with uploading the videos. The good thing is that in our own “show job description” we were prepared for this, and as a result, we made two fundamental decisions:

• We were not competing with anyone but ourselves. So we took out the “competition” element.

• As this was our first engagement with a Camera to Cloud workflow, we could not afford NOT having an on-site editor, so if it all fell apart, we could always go back to the “running with the recording card to the press room” option. This was a wise decision (of course…:) ), and Rachael, our editor, was still getting the video files from the Cloud (when the connection resumed) and was able to delegate the editing work between herself and two more colleagues of ours in Argentina and Italy.

Any other options but a fixed Wifi network?

FUJIFILM was clever enough to bring to the show floor a few Sclera devices. This LTE “take it anywhere” bonded WiFi device uses multiple cellular carriers and custom high-gain antennas. This would have freed us up completely from the obstacles we encountered and described before, but even though it is relatively small, running around with another device (powered by a V-mount battery that adds weight) seemed like a less preferred option.

So we decided to use the Sclera outdoors. Together with the friendly people at DJI, we took out the newly announced Mavic 3 Pro drone for a test ride at the Seven Magic Mountains attraction on the outskirts of Vegas. And again, I couldn’t stop watching the clips mounting up on my frame.io App. Let’s face it, we were in the middle of “nowhere”, filming, and by the time we got back to the hotel, the news video was ready…

The aftermath

Michael Cioni, the person behind frame.io’s C2C sees a very defined future when in the early 2030s cameras will NOT contain internal recording media anymore, but rather have an internal satellite SIM card that will upload the footage directly to the cloud. Until this happens, we are very much restricted by two things that might be the Achilles heel of that engaging technology.

• Assuming you will not always be able to use fast wifi for uploading footage, the current actual LTE speed might be restrictive (not equal to all providers and countries).

• Data plans. Again, depending on where you are, if wifi is not available and you need to rely on your phone, for example, as the bridge between the camera and the cloud, the cost can be painful.

So in other words, as with many other emerging technologies (electric cars anyone?), it is the infrastructure that limits us from unleashing the full potential of what can be done with Camera to Cloud. Nevertheless, if you are a veteran of this industry, you might remember those big satellite trucks serving Electronic News Gathering crews at some events. Those big trucks got much smaller and were at times obsolete when replaced by LiveU units. I can easily imagine this happening to us filmmakers in the future. It is not for nothing that companies like Sony with their Creators’ Cloud or even Alteon are racing to capture the attention of users with cloud based workflow solutions.

Back to FUJIFILM’s X-H2 or X-H2S cameras and frame.io – I hope the following will be considered:

• The FUJIFILM FT-XH File Transmitter should find itself as a built-in inside of any (future) camera.

• Currently when working with Adobe Premiere and frame.io, one needs to “import” the footage from the cloud to the timeline. Hopefully, editing from “within” the cloud can become a reality soon.

• A better “viewing status” of the connection/upload to frame.io is needed.

• Currently, the white and grey indication on the LCD and EVF is very confusing. One really needs to look for it in order to know what’s going on.

• All related “Status/upload” and such needs to be grouped together and colored for easy viewing.

• There is a list (queue) of the files waiting to be uploaded in the camera menu. There was only an option to cancel (reset) all the uploading jobs, and I would like to be able to choose which ones I want to remove or which groups of files I want to remove from the uploading queue – for example only the RAW stills, etc.

• It would be beneficial if it was possible to choose exactly in which folder the camera uploads to in frame.io. I felt the camera was simply generating way too many subfolders.

Last words

A big thank you to the entire CineD team for making this documentary possible. Since NAB, we successfully covered Cine Gear but truly missed the Camera to Cloud workflow! Once you experience this, it is very hard to go back.

Do you have experience with working on Camera to Cloud workflow? If yes, how was your experience? Please share with us your comment in the section below.

The FUJIFILM X-S20 is here, and in my opinion, it is the best middle-class camera out there. In fact, it is like having 5 cameras in one body: a photo camera, webcam, “normal video camera”, vlogging camera, and what I call a “creative camera”.

Now, in this new age of camera announcements, I bet you see many videos highlighting the vlogging capabilities of a new camera, which is nice, but if you are a filmmaker and looking beyond vlogging or even “normal filming”, there is a unique feature in this small camera which, in my opinion, cannot be found elsewhere in this price category.

New: 6.2K 3:2 open gate recording

While the X-S20 can do a lot, I would like to start my review by mentioning its most attractive feature, especially for those who want to get the most out of the new camera, and I’m talking about the option to film in 6.2K resolution (6240 x 4160) and 3:2 open gate recording (full-sensor). So, why am I so excited about this?

To my knowledge, until now, it was not possible to film using these settings on an APS-C sensor size camera in this price range. This particular option opens the door for several enhancements in your filming and editing workflow.

Filming with anamorphic lenses

Maybe the most noticeable thing is the ability to attach an anamorphic lens and use the entire sensor when de-squeezing the image instead of adding “black bars” on the top and bottom to imitate a “cinematic look”.

Yes, a similar feature can be found on the much more expensive X-H2S. Actually, the only difference is the recording codec. In the new camera, FUJIFILM implemented H.265, ALL-Intra, 4:2:2, 10-bit internal recording at 360Mbps, which is robust enough to get excellent results while doing casual creative work. At this point, it is worth mentioning that on both the X-H2S and X-S20, it is NOT possible to de-squeeze the image for viewing. (By now, I am convinced my brain has a de-squeeze mode…).

In the above video, I teamed up with Peter Koblhirt, who is a professional chef, and we spent a few hours together. We recorded everything with the FUJIFILM X-S20 together with the SIRUI Saturn 35mm 1.6X anamorphic lens, which costs $1,299, the same as the camera.

For the “sake of experimenting”, I also tried working with SIRUI’s 1.25x anamorphic adapter attached to the Saturn 1.6x in order to see what the image looks like when bringing it to 2x anamorphic squeeze (see the interview and outdoor shots), but in reality, if you are looking for a perfect 2.40.1 aspect ratio on a 3:2 sensor, then 1.6x squeeze is the way to go!

SIRUI’s Saturn 35mm T2.9 full-frame anamorphic lens deserves an all-separate conversation (really nice for the size, weight, and price overall). But for now, let me mention that since its minimum focus distance leaves a lot to be desired, I had to use diopters for the food close-up shots.

Filming in 6.2K, 3:2 with a spherical lens

I can hear you asking if recording in 3:2 is the best way to film when working with anamorphic lenses, and the answer is absolutely yes! Depending on your final export, you can use the footage that you shot with your spherical lens on a 16×9 timeline for reframing your shot, or even create a vertical video, which is very common these days.

External RAW recording

For those who are seeking an even higher recording quality beyond what the internal recording offers, with the X-S20 you can output a RAW video signal via HDMI in 6.2K 3:2 open gate in up to 30p, and 5.2K 16:9 up to 60p to Atomos or Blackmagic Design recorders. And no, simultaneous internal and external recording is beyond the capabilities of the camera.

FUJIFILM sensors and processors – mix & match

OK, so we have, let’s call it a “3:2 creative mode” which, by the way, cannot be found in the more expensive FUJIFILM X-H2 and X-T5 cameras. I find it interesting, especially as the new X-H and X-T models from FUJIFILM have the newer X-trans 5 CMOS sensor, while the X-S20 still uses the older X-trans 4. But as we know, the sensor is only part of the story, and the X processor found in the X-S20 is the newest, meaning the processor 5. This might explain the power that this new camera has in order to record 3:2 internally but leaves the question open. Why not the X-H2 and X-T5?

Filming in 4K up to 60p

Next is what I call “normal video mode”, which is pretty much becoming standard on all the latest batches of FUJIFILM cameras. And I’m talking about 4K 16×9 and DCI 17×9 in up to 60p. I would emphasize that having a 4K/60p, ALL-I, 10bit, 4:2:2 in this mid-range camera is a very welcome edition (the image will be slightly cropped in 4K/50/60p). Couple that with impressive low-light capabilities and you have a little “all-rounder” working horse.

It is, of course, possible to film in 1080 up to 60p and in a high frame rate, too (up to 240fps). The recording quality of these high frame rates is like the one found in the X-T4 for example, so it’s fine, but nothing more.

By the way, HDMI external high frame rate recording up to 120p is possible too, but I didn’t have a chance to check the quality. (In general, I would have loved to see a better internal high frame rate recording quality which would make this little camera even more appealing).

New: full-HD LP mode

Extended continuous recording time in Full HD in 50 or 60p is a new setting that somehow creates a nostalgic feeling as FUJIFILM is using the LP (Long Play) mode. If you are a life veteran like me, I’m sure that the combination of those letters can bring sweet analog memories. Anyway, back to the review. If you set the camera to LP mode, it will crop the image a bit (1.29x as in high frame rate recording), and by doing so, less data is processed, allowing longer recording time. If you choose to use it, I do recommend setting the camera from “AUTO POWER OFF TEMP” to “HIGH” to get the most out of this mode.

Overheating

Overheating issues in modern mirrorless cameras are an issue for many, and manufacturers are constantly trying to combat it in different ways. When I first saw the X-S20’s small camera body I was worried about it having very short recording times, especially in 6.2K, but to my surprise, we could shoot for about 1 hour and 55 minutes at room temperature before the battery got exhausted. Yes, the camera got hot, and yes, the overheating sign was there, but nonetheless, the camera kept working. Now, this was OUR experience and I’m truly looking forward to hearing feedback from you guys when the camera hits the market.

Oh, and speaking of which, if you are concerned about overheating, you can always add the FUJIFILM FAN 001 to the back of the camera for greater flexibility.

New: vlog mode

Nowadays, what is a new camera without a vlog mode? Okay, okay… In all seriousness, adding a mode like this might be beneficial for many, especially the “ageless” generation that loves to film, upload and share content. There are several advantages to filming in this mode. You can easily change the resolution, frame rate, film simulation, and other settings as needed (default settings are on FHD 30p and PROVIA film simulation). The second benefit is the reliability of the autofocus system in this mode.

While testing, I was constantly trying to confuse the AF by adding additional faces to my frame, but it stayed “locked” on my face. If you take a closer look at the photo above, you can notice the AF tracking points around my eyes.

I also liked working with the TG-BT1 Bluetooth tripod grip together with the XC 15-45mm f/3.5-5.6 OIS PZ Lens compact zoom lens. With this tripod/lens setup, the camera becomes a complete vlogging kit that can be taken anywhere and used creatively.

New: Automatic Subject Detection settings

Another exciting feature of the new camera is the ability to let it detect subjects and automatically set the focus on them. In the above image, you can see how AF first detects an “animal” and then changes to a “human” when the camera is on Florian. I checked on some of the more expensive FUJIFILM mirrorless cameras that we have at the office, and such a feature does not exist on any of them. Strangely, this helpful feature can be activated ONLY when the camera is in AUTO mode, which is a pity since it would serve well in “vlog” or even “normal” video mode, too! In the little experiment we did at the office, that Auto Subject Detection feature worked well on Florian, my colleague, and Marly, his beloved dog!

Taking photos and using the camera a webcam

Well, this is of course obvious, but as it is beyond our capability to review the photo camera performance of the new X-S20, it is worth mentioning that it has a 26.1MP sensor and the camera can be connected directly to your PC with a USB-C. It can serve as a high-quality 4K/60p webcam without the need for any additional accessories.

Main differences from its predecessor, the X-S10

The new FUJIFILM X-S20 might look similar to the X-S10 from the outside, but the grip is a bit thicker because it accommodates a larger battery inside (the NP-W235).

Of course, the video capabilities of the new camera are enhanced because of the newer engine next to the stronger IBIS. As a reminder, the X-S10 could shoot internally in up to 4K/30p, H.264, 4:2:0, 8bit only, with a 30-min recording time, without the ability to record RAW video externally, and no headphone jack.

In conclusion, the X-S20 is a much more advanced camera, at least for filming video and that alone justifies the price increase from its predecessor, at least in my opinion.

Lab Test summary

We also ran our Lab Test with the final firmware version of the camera and found out that the rolling shutter performance is average. When it comes to dynamic range, it really depends on the frame rate and resolution you decide to work on. We managed to get 13 stops of DR when the camera is set to 4K/25p

X-S20 shortcomings

Where this camera falls short is mostly when looking at the EVF. Its quality is just Fine, making it not so easy to focus manually (0.62x magnification, 2.36 million-dot Electronic vs. 0.8X 5.76Mil in the X-H2/S, for example).

Also, IBIS is okay, but in my opinion, other camera brands are doing a better job when it comes to stabilizing the image. I wish that stabilizing could be enhanced, as the potential is absolutely there.

Who is this camera for?

Last but not least, who is this camera for? Well, if you are a filmmaker who is already using a FUJIFILM camera like the X-H2S, the X-S20 might be a perfect companion, especially if you are shooting a lot with anamorphic lenses.

As for other creators, this camera might be very appealing too. It’s powerful, small, lightweight (491g with battery and card), easy to use, has great video quality, and can simply be taken anywhere. Drone operators might consider using this camera as well, because of its size and open-gate capabilities. Personally, I wish I had had such a camera years ago when I was filming documentaries for broadcasters around the world.

Conclusion

At times, it sure feels like we live in a spoiled (filming) world. What was merely a dream only some years ago has become a reality thanks to “user demands” and the pressure from mobile phones that are performing better and better when it comes to taking photos and videos. That being said, the “mid-range” camera arena surely got upgraded with the introduction of this new filming device. Please correct me if I’m wrong, but such a feature-rich camera with up to 6.2K 3:2 open gate recording in ALL-Intra, 4:2:2 10bit internally, including F-Log2 with an APS-C sensor size at this price point is not so easy to find elsewhere.

If I have to summarize my conclusion in a sentence, it would simply be “a camera that one can grow up with”. You get “A LOT of camera” for $1,299 and, more importantly, it’s simply fun to use!

Pricing and availability

FUJIFILM X-S20’s expected delivery time is the end of June 2023, and next to the camera body only, there will be two available kits (in some territories):

FUJIFILM X-S20 Mirrorless Camera Body in Black with the FUJINON XC15-45mmF3.5-5.6 OIS PZ Lens Kit: $1,399.95 

FUJIFILM X-S20 Mirrorless Camera Body in Black with the FUJINON XF18-55mmF2.8-4 R LM OIS Lens Kit $1,699.95 

What do you think about the FUJIFILM X-S20? Do you see some use for open-gate recording coexisting in a vlogging camera? Please share your thoughts with us in the comments section below.

We had the Sony ZV-E1 camera at our CineD HQ for a relatively short time, but nonetheless, I’m confident I have a comprehensive view of what this camera offers and the key elements that make it a unique filming tool, especially for those who like to be and talk in front of a camera. 

Before I continue, here’s a spoiler: I can say with confidence that the Sony ZV-E1 inherited many of the a7S III filming capabilities, but, took it a step further by enhancing its autofocus performance, and even more importantly, added a dedicated AI processor like the one found in the a7R V that does wonders in many ways. 

Sony ZV-E1 – the heritage

But first things first. What Sony did with the new camera is a bit of a mix-and-match. The 12-megapixel imaging sensor comes from the a7S III, and the Bionz XR imaging processor is the same as the one found in the a7R V, the a7 IV, the a1, and a7S III. The vlogging spirit comes from Sony’s ZV series of cameras, and the form factor is very similar to the A7C, making it an extremely compact and the most lightweight full-frame camera I’ve ever worked with.

Specification highlights

Diving a bit deeper into the specification and features, here is what one can expect from this new camera: 

• Full frame up to 4K/60p with no crop factor
• Cropped 4K/120p and full HD/240p capabilities will be added to the camera around June with a free firmware update
• 10-bit, 4:2:2 All Intra recording
• S-Cinetone next to S-Log picture profile recording 

In fact, there is a new Cinematic Vlog setting that automatically sets the camera to record in 4K 16×9, but with black bars for imitating 2.35.1 cinematic aspect ratio,  24p or 25p depending on the camera region settings, and the default look is S-Cinetone, but this can be changed in the menu to one of the 4 additional “looks” and 4 “moods” presets for creating “instant” “cinematic expression”. 

One thing that I want to highlight about this cinematic Vlog mode and other camera functionalities in general, is how easy it is to access by simply choosing and clicking on it on the LCD screen. Sony did extremely well in taking some of the key features of this camera out of the menu and adding them to the screen for very fast and comfortable access.

When it comes to stabilizing modes, next to its already existing settings, Sony added a new Dynamic active mode which lets you crop the image even further and by that, do an even better-stabilizing job.

By the way, to power it all, Sony is finally using the NP-FZ100 battery, the same one that can be found in its more advanced mirrorless cameras.

Sony ZV-E1 – strengths

OK, let me concentrate on the strengths of this camera and I’ll start with its lowlight capabilities as the sensor is the same as the one found in the a7S III. 

Officially it is not a dual ISO camera, but one can clearly see that like with the a7S III, ISO 640 and ISO 12,800 are the cleanest. Even the rolling shutter performance of both cameras is similar. As for the dynamic range, please head to the newly added ZV-E1 Lab Test in our ever-growing databases section.

Sony added a few new recognized subjects (Bird, Car/Train, Airplane, Insects) and this makes the autofocus even more usable. It is simply working very well, be it when using the new “bokeh switch” to highlight a product, or, even when hiding the face altogether. The camera will simply stay where it was set to last without losing orientation.  

Another thing, that works well is the new focus breathing compensation. Changing focus points between objects is very clean and smooth.  

But the really big news here is the new camera AI functionality and I would like to mention two things: 

• The first is “framing stabilizer”. In this mode, the camera will attempt to keep the subject of filming in the center, regardless of where it is really positioned.
• The second is “auto framing”. When this function is engaged, the camera can crop automatically into the frame making the video more dynamic.

You can actually turn this function “on” inside the camera, and “off” in case you are recording simultaneously to an external recorder. By doing so, you can get a “master clip”, too.

Needless to say that the amount of cropping or the speed of changing framing, can be adjusted in the camera menu.

The same goes for recording an interview. The operator can touch the screen and change the framing between talking heads. 

What can be improved?

No camera is perfect and please allow me to share my concerns with you:

Maybe the biggest one is overheating. I was surprised to see how fast the camera reaches the point that it had to shut off on standard mode, so I definitely recommend using the “high” settings for gaining additional recording time.

Next is audio quality. I have to be careful here as I’m not sure if the camera I have is somehow defective, but in any case, when setting the camera to record audio coming from the front, the amount of hiss made the audio not usable. Changing “Audio Out Timing” to “Lip Sync” made the situation better, but the higher you go up with the recording levels, the more this phenomenon is evident.

Last but not least, when using “M” mode for video recording, please be aware that the “Soft skin” setting is turned “on” by default. Although this feature is popular in some world regions, it might not be for everyone, so I hope that Sony can change this with a firmware update. 

Conclusion

Sony did a good job with this camera by balancing the camera performance, its usability, and features for the targeted vlogger audience. Like with the a7S III, the lowlight capabilities are insane and the AI “tricks” are working very well without the feeling of losing any significant image quality. Personally, I’m missing an EVF for greater filming flexibility, but I guess I am an old fashion operator here…

One question that must be asked is: Where does this new camera put the a7S III which is 1,300 Euro/dollars more expensive? And my answer is, maybe not in such a good place. True, the a7S III might be a more advanced camera in some ways by targeting filmmakers but looking at the complete package I’m not so sure that the 1,300 Euro/dollars price difference between the two cameras is justified and this might put the a7S III in a difficult sales position.

Pricing and availability

The camera will be shipping during April and be sold for € 2,005/$2.198 body only, or, € 2,435/$2.498 with a 28-60mm F/4-5.6 kit lens.

What is your opinion about the new Sony ZV-E1 camera? Would you consider getting it as a companion to any of your other Sony cameras? Please share with us your thoughts in the comment section below