AI enthusiasts who like the Raspberry Pi range of products can rejoice, as the company is now announcing its new Raspberry Pi AI Camera. This product is the result of the company’s collaboration with Sony Semiconductor Solutions (SSS), which began in 2023. The AI Camera is compatible with all of Raspberry Pi’s single-board computers.
The approximately 12.3-megapixel AI Camera is intended for vision-based AI projects, and it’s based on SSS’ IMX500 image sensor. The integrated RP2040 microcontroller manages the neural network firmware, allowing the camera to perform onboard AI image processing and freeing up the Raspberry Pi for other processes. Thus, users who want to integrate AI into their Raspberry Pi projects are no longer limited to the Raspberry Pi AI Kit.
The AI Camera isn’t a total replacement for Raspberry Pi’s Camera Module 3, which is still available. For those interested in the new AI Camera, it’s available right now from Raspberry Pi’s approved resellers for $70.
This article originally appeared on Engadget at https://www.engadget.com/computing/accessories/raspberry-pi-built-an-ai-camera-with-sony-165049998.html?src=rss
Fujifilm’s original GFX100S was a popular alternative to the GFX100, as it offered the same picture quality in a slimmer body at a much lower price. It did have some flaws, though, most notably its autofocus that didn’t always keep up with the subject. In a similar vein, the company recently released the $5,000 GFX100S II as a smaller and cheaper alternative to the $7,500 flagship GFX100 II. It looks much the same as before, but has an updated sensor, processor, autofocus and more.
Both cameras are clearly aimed at pros, but is the GFX100 II worth the extra $2,500, or is the S model fine for most?
Body
At 1.9 pounds, the GFX100S II weighs an ounce less than its predecessor and its body's design and controls are nearly the same. As before, it comes with a mode dial and focus selector switch, dual control dials, drive select buttons and a quick menu. On top is a small, always-on display that shows settings, simulated mode dials and a histogram. That way, you can see all your key settings when shooting at hip level.
It feels a lot more like a mirrorless than a classic medium format camera. It handles well, but after spending time with both models, I prefer the GFX100 II’s larger body and extra controls, even though it's five ounces heavier.
Steve Dent for Engadget
The rear 3.2-inch, 2.36-million-dot-display is unchanged, tilting up, down and to the right, but it doesn't flip out. However, Fujifilm boosted the viewfinder’s resolution to 5.76 million dots, so it’s noticeably sharper than the last model’s 3.69-million-dot EVF.
The GFX100S II includes dual SD UHS-II card slots but lacks a high-speed CFexpress port like the GFX100 II. Though it uses the same battery as before, it can now shoot 540 photos on a charge, up from 530. Other features include USB-C, mic and headphone ports, along with a micro HDMI connection. A new feature is USB-SSD connectivity that allows for faster burst shooting and ProRes video capture.
Performance
Steve Dent for Engadget
The GFX100S II does shoot faster than its predecessor at up to 7 fps compared to 5 fps in mechanical shutter mode. Without a fast CFexpress card, though, you can only capture about 30 compressed RAW frames before the buffer fills. There’s also a slight image quality penalty as the camera downshifts from 16- to 14-bit RAW images when shooting bursts. Fortunately, most of the frames are likely to be sharp thanks to some big autofocus improvements. Even with the faster burst speeds, I noticed a large boost in the percentage of sharp photos when shooting quick moving subjects.
The AF improvements don’t end there. AI face and eye tracking is smarter and more accurate, meaning your subject's eye and not their eyelash is more likely to be in focus — something that’s critical with 100 megapixels of resolution. Fujifilm also introduced new subject tracking modes for animals, birds, automobiles, motorcycles, bicycles, airplanes and trains.
In-body stabilization also gets an upgrade, with eight stops in shake reduction compared to six before. That allows you to capture dimly lit scenes or blur movement at shutter speeds as low as a half second, without the need for a tripod. As with the previous model, rolling shutter distortion is a real problem with this camera in electronic mode, so neither you nor your subject can move much when using it. Otherwise, you’ll get some nasty skewing that will ruin your photo.
Image quality
For pro photographers who can afford it, the GFX100S II offers unmatched image quality. The 100-megapixel sensor delivers pin-sharp photos and 16-bit RAW pixel depth that delivers 14 stops of dynamic range, beating most other mirrorless cameras. That means you can shoot in very dark, bright or contrasty conditions and retain detail in all parts of the photo.
Fujifilm’s color science is ideal for portrait work, with accurate and pleasing skin tones. The level of detail means prints can be blown up to very large sizes, nearly 40 x 30 inches at 300 pixels per inch (PPI). The resolution and precision color rendering also makes it ideal for landscape photography, art archival work and more.
The outstanding RAW quality lets photographers perform detailed post-production work without hurting image quality, even on shots not exposed correctly. In low-light, you’ll start to see a fair amount of noise above ISO 8000 or so. However, a lot of that grain disappears when you shrink those 100-megapixel images down to a more reasonable size.
Nathanael Charpentier for Engadget
At the same time, JPEG images are outstanding straight out of the camera. Like other Fuji models, the GFX100S II comes with an array of film simulations, from the new Reala Ace to Velvia, Acros and others. Most pro photographers like to tweak their images, but many may shoot both RAW and JPEG with film simulations so they can show their clients something right away and make detailed adjustments later on.
Video
Unlike the more expensive GFX100 II, video is a weak point with this camera. Resolution is limited to 4K 10-bit, and while there’s no cropping, there’s considerable pixel binning that reduces sharpness. You can capture 10-bit 4:2:2 ProRes video at up to 30fps, though that entails hooking up an SSD drive. There aren’t really any slo-mo modes, as both 4K and 1080p recording are limited to 60fps.
While video lacks the sharpness seen on other mirrorless cameras like the Sony A1, color rendering is excellent. You can also improve dynamic range by shooting in Fujifilm’s FLog or FLog2 modes and applying a LUT later on in post.
Rolling shutter in video mode is reduced compared to photos because of the pixel binning, but you still need to be careful not to move the camera too much to avoid excessive wobble and skewing. The in-body stabilization is only effective for handheld shooting without much movement. If you try to walk while vlogging, you’ll see some severe jolts. All told, I’d say the Fujifilm GFX100S II is mostly a photo camera, with video available for emergency use only.
Wrap-up
Steve Dent for Engadget
The $5,000 Fujifilm GFX100S II gives professionals outstanding picture quality at a significantly reduced price over the $7,500 flagship GFX100 II. It even goes beyond what buyers might expect by offering fairly quick burst shooting with autofocus that’s much more reliable than on the original GFX100S. Video isn’t a strong point, but that won’t be important for its target market.
The GFX100S II is also the cheapest camera in the mainstream medium format space, with its primary competition being either the GFX100 II or Hasselblad’s $8,200 X2D100C. I’d highly recommend it for portraits and other types of shooting where detail is key. However, burst speeds and autofocus still aren’t quite good enough in fast moving scenarios.
Those who need more speed and reliability for things like weddings or live events would be better off with high-resolution full-frame models like the $3,700 Nikon Z8, $6,500 Sony A1 or Canon’s new $4,300 EOS R5 II.
This article originally appeared on Engadget at https://www.engadget.com/cameras/fujifilm-gfx100s-ii-review-outstanding-performance-at-a-reduced-price-153020497.html?src=rss
DJI has been chipping away at GoPro's action cam dominance and just took another big step with the launch of the Action 5 Pro. With a new "next-gen" sensor, it promises improved video quality, along with extended battery life, bigger displays and a new feature, subject tracking and centering.
The 1/1.3-inch sensor is the same size as the one on the Action 4, but DJI says it now boasts 13.5 stops of dynamic range, "making it the first action camera with specs that rival professional cameras." Maximum photo resolution is 40 megapixels (MP), up from 10MP on the Action 4. The lens has the same 155 degree field of view and f/2.8 aperture as the last model.
Steve Dent for Engadget
You can shoot 4K at up to 120p in 16:9 as before, but it now supports 4K 120p capture at 4:3 (3,840 x 2,880 pixels) — so slow ultra-slow-mo can be captured to the entire sensor if you want to crop the image later. Resolution is still capped at 4K UltraHD, though, where the GoPro 13 supports up to 5.3K 8:7 (5,312 x 4,648) capture, giving users more cropping options.
Moreover, it supports 1080p capture at up to 240fps, and can accelerate that to an eye-ripping 960fps by using frame extrapolation (making up fake frames out of real ones).
As before, images can be recorded in 10-bit D-Log M and HLG to boost dynamic range. DJI says you can shoot "high-dynamic, low-light video" at 4K/60fps to retain shadow details without overexposing highlights. It also comes with a SuperNight mode that uses AI noise reduction to yield relatively clean video in low light.
Steve Dent for Engadget
The new 1950mAh has a higher capacity than the previous 1770mAh model, which allows for up to 4 hours of continuous use or around 50 percent more than before (the Action 5 Pro can still use the previous model's batteries, too). Also helping that is a new 4nm chip and algorithm that optimizes for energy savings. "Even in temperatures as low as -20° C (-4°F), the camera can reliably record video for up to 3.6 hours," DJI adds.
A key new feature is subject centering and tracking that lets creators shoot without a gimbal. As with other DJI products (ActiveTrack on drones), it intelligently detects a subject's position and dynamically adjusts the composition to keep the subject centered in the frame. This is done digitally, obviously, and resolution is reduced to 2.7K — but it could still be a great benefit for extreme sports and other creators. As before, it offers "robust stabilization" and HorizonSteady that keeps the camera level through 360 degrees of motion.
Steve Dent for Engadget
The Action 5 Pro is EN13319 certified to be waterproof without a case down to 20m (66 feet), 2m more than before. It can start recording automatically as soon as you're underwater and comes with an interesting new feature: a pressure gauge to record underwater depth, duration and altitude data to ensure user safety.
Though the body is the same size as before, both OLED screens are larger thanks to smaller bezels, up 16 percent. The larger rear screen is now 2.5-inches diagonally, up from 2.25 inches and the largest on any mainstream action cam, according to DJI. As before, the Action 4 works with the DJI Mic 2 for high-quality sound recording and you can simultaneously use the built-in mic for ambient sound.
Steve Dent for Engadget
Another nice addition is 47GB of built-in storage and 80MB/s transfer rates via Wi-Fi 6.0 or USB 3.0. Other features include an updated Mimo app, Wi-Fi livestreaming, timecode for multiple camera syncing and a pre-record feature.
I've had the Action 5 for a short time and it largely resembles the previous model, other than larger power and record buttons, a different internal mic opening and a hole at the bottom for the pressure sensor. Other features like the magnetic mount are carried over from the Action 3. The improvements look promising, so stay tuned for a more detailed review (where I hopefully won't wreck my face).
Steve Dent for Engadget
Perhaps the best news about the Action 5 is that it's cheaper than the Action 4 at $349 in the Pro Combo package that includes the camera and one battery, a protective frame, quick-release adapter mount and more. The Action 5 Pro Adventure Combo adds two more batteries and a charger along with a 1.5m extension rod for $449. Both are now available at DJI's store and most authorized retail partners.
This article originally appeared on Engadget at https://www.engadget.com/cameras/dji-launches-the-action-5-pro-camera-with-subject-tracking-and-improved-video-130006706.html?src=rss
GoPro’s latest action cam has officially broken cover, after a few leaks spoiled some of the surprises. With the Hero 13 Black, the company is bolstering its position as the go-to action cam with more versatility and features aimed at high-level users who’ve probably owned a GoPro or two for years. GPS is back, the battery is bigger and there are upgrades to slow-motion capture and more, but the big news might be outside of the camera unit itself.
The biggest upgrade may be the new family of modular lenses it’s calling the HB Series. In addition to ultra-wide, anamorphic (unfortunately not ready for testing) and macro, GoPro is introducing an ND (neutral density) filter four-pack, aimed at amping up motion blur.
Photo by Mat Smith/Engadget
If you’re a GoPro fan, you’re probably aware that the company has already dabbled in filters with the Max Lens Mods for the Hero 12, but the company has taken it a lot further here. All the new HB-Series lenses are waterproof and compatible with GoPro’s HyperSmooth video processing. Crucially for non-professionals, the Hero 13 Black will automatically adjust camera settings, including resolution, frame rate and even your Protune settings to match what's attached. This means you may be locked out of certain capture modes, frame rates or resolutions if the lens isn’t compatible. But that’s preferable to recording something special with a lens, only to discover afterward that it looks like a mess or cropped the best parts out.
Depending on how you plan to use your new GoPro, some lenses will offer more utility than others. While I’ve only had the camera for a few days, the Ultra Wide Lens ($100) seems like the most tempting addition. With a new 1:1 aspect ratio, it means you can decide whether your video is vertical or horizontal after capture. The lens expands the field of view to 177 degrees and can still capture 4K at 60fps. During my testing, I liked the flexibility to use footage in both orientations, and not having to fuss so much about framing.
If you’re looking at the Hero 13 Black as an additional camera for interesting cutaway shots and close-ups, the Macro Lens ($130) might tempt you. With a minimal focus distance of 11cm (4.3 inches), that’s four times closer than the GoPro’s original lens. This lens also includes its own rotating ring to hone in on the right focal length. (There’s even a focus peaking option, hidden away in Protune settings).
During GoPro’s presentation, it was the Anamorphic Lens Mod ($130) I was most intrigued by, but I’ll have to wait to test that out. Able to capture video with a 21:9 aspect ratio, it’s pitched as a lens for “professional-level artistic filmmaking.”
The footage is de-squeezed while being captured, making for easier editing. (De-squeezing is the process of correcting any oval-shaped image distortion caused by anamorphic lenses.) I like the notion of an action-cam cropping those mushy stretched edges for punchier footage. Plus, it’ll capture lens flares, too.
When it came to ND filters, I found them challenging to work with in the past, even when they were built into traditional cameras. This is where GoPro’s auto-detection makes a lot of sense. Alongside adjustments to shutter speed to suit the filter and available light, the Hero 13 Black will even indicate arrows if it thinks a stronger (or weaker) ND filter would be better within the four-pack ($70). However, you can’t combine these HD-series accessories, so you can’t use an ND filter with the Macro lens, for example. There’s a new “AutoCinematic” shutter option to optimize motion blur while holding onto image stabilization, while a Motion Blur Protune option lets you tweak exactly how much blur you want.
So what about the camera itself? The Hero 13 Black has a 10-percent bigger battery than its predecessor and can capture around one and a half hours of 4K video at 30fps. There’s also improved thermal performance both when it’s static or moving with airflow. GoPro says the new camera has longer runtimes in both hot and cold environments, claiming that some of the biggest improvements will come if you’re used to capturing in 10-bit video, another nod to pro users.
Photo by Mat Smith/Engadget
GoPro has boosted ultra-slow-motion capture to 400 frames per second at 720p resolution. However, this is burst capture, limited to 15 seconds of real-time recording. At its maximum resolution of 5.3K, the Hero 13 Black can capture 120fps for up to five seconds. The Hero 13 Black also supports Hybrid Log Gamma (HLG), an industry-standard HDR format that should yield the highest dynamic range yet on GoPro video.
As I teased earlier, GPS is back, too. This means you’ll be able to search for footage based on location, but the camera can capture telemetry data like speed, path and altitude, pulling that data into your video file. You can then bring this stuff to life with overlay graphics added through GoPro’s Quik app.
There are a lot of smaller, useful UI improvements and upgrades. I won’t touch on all of them, but the most notable may be compatibility with AirPods and other wireless headphones over Bluetooth. The Hero 13 black can capture sounds in a Standard Audio Mode, what GoPro calls “balanced, true-to-life sound,” or there’s a Voice Audio Mode to prioritize speech. With WiFi 6 compatibility, you can also transfer your videos and shots up to 40 percent faster to your smartphone. That’s a true quality-of-life upgrade you’ll notice immediately if you’ve ever had to wait around for files to transfer from action cams (or any camera) over WiFi.
If you’re used to extended shooting (or time-lapse capture), the Hero 13 Black’s new Contacto magnetic door and power cable could be very useful, too. With a pogo-pin style latch, the cable can charge or power the camera from external USB-C battery packs or an adapter plugged into the wall. Notably, it will power the camera even if there’s no battery inside.
GoPro has also decided to add a third mount, a magnetic latch system that works with several new mounts (sold separately). If you’re not here for the peripheral upsell, however, there are still fold-down mounting fingers for extra-secure installation and the standard 1/4-20 mounting threads for your existing tripods and mounts. I love the new magnetic mounts, but now I want one for my tripod, one with a clamp, one for everything. (In the company’s defense, its newest camera is compatible with over 60 GoPro mounts and accessories. It has identical proportions to the Hero 12 Black.)
Photo by Mat Smith/Engadget
If you were hoping for a 1-inch sensor you’ll have to wait a little longer. Current GoPro users may be slightly disappointed by what seems to be the same camera sensor. Many devotees online hoped for a larger sensor to boost low-light performance and picture quality. A lot of the upgrades and additions here seem aimed at high-level users, but it’s those same people hoping for a substantial hardware upgrade beneath all the UI improvements and peripheral additions.
The Hero 13 Black is priced at $400 for the standalone camera, while a Creator Edition ($600) bundles the camera with a Volta Power Grip, Media Mod and Light Mod. The company has teased more bundles that will combine its new camera with lenses and other accessories.
All are available for preorder today at GoPro.com and will ship on September 10.
This article originally appeared on Engadget at https://www.engadget.com/cameras/gopro-hero-13-black-release-date-price-hands-on-impressions-130027978.html?src=rss
Instant cameras have become a popular category for one simple reason: they’re fun. You can whip one out at a party or event and take a snap, then a couple of minutes later you have a real, physical photo that everyone can touch, feel and enjoy instead of staring at a screen. It’s also a lot cooler than a smartphone shot thanks to the retro, nostalgic vibes.
Given the large number of models on the market, choosing one can be tricky. That’s why we’ve put together this buying guide that will help you choose the best instant camera for you based on factors like price, film size, photo quality and more.
What to consider before buying an instant camera
The main factor for most people is the size of the film. Fujifilm’s Instax alone offers three formats: Mini, Square and Wide, with film sizes of 3.4 x 2.1 inches, 3.4 x 2.8 inches and 3.4 x 4.3 inches. The corresponding image sizes are 2.44 x 1.81 inches, 2.44 x 2.44 inches and 2.44 x 3.9 inches.
Leica’s Sofort 2 and Lomography models also uses Fujifilm’s Mini format, while Polaroid’s smaller Go format is 2.64 x 2.13 inches (1.81 x 1.83-inch image) and its i-Type size is 3.46 x 4.21 inches, with a square image size of 3.11 x 3.11 inches.
When most people think of retro instant cameras, they think of a square image size. That’s why Instagram, which used to have a Polaroid-like app icon, only supported square images for years. If that’s what you want, then Fuji’s Square or either of Polaroid’s formats are best. However, if you want as large an image as possible, Fujifilm’s Instax wide is the way to go.
As for prices, Fujifilm’s Instax films are generally around $1 per shot for square or wide color film and $.60 - $.75 for mini color film. Those prices go up a bit if you choose custom models with colorful or patterned borders, and down if you buy in bulk. Polaroid’s i-Type film is a bit more at about $1.50 per shot or $1.00 per shot for Go color film.
Image quality isn’t really the point with instant cameras of course, as they’re always going to be fuzzy compared to digital photos. However, some models (notably Fujifilm and Polaroid models) are a bit better than others.
And in terms of pricing, cameras with larger film are generally more expensive. It also increases if you add hybrid features like USB-C connectivity, smartphone sharing, the ability to print out photos and more, along with niceties like a selfie mirror. The best, most feature-laden cameras can cost upwards of $200 and basic versions can be well under $100.
This article originally appeared on Engadget at https://www.engadget.com/cameras/best-instant-camera-120034778.html?src=rss
Halide is one of the best camera alternatives for the iPhone, a lovingly-crafted app that gives users plenty of manual control over the images they take. And today they just pushed out a major update with an intriguing proposition: an “anti-intelligent” camera.
Of course, more advanced digital photographers will already be familiar with the RAW format, something Halide (and basically all digital camera makers) have offered for years now. RAW simply means you’re getting every bit of data straight from the camera’s sensor, with few modifications or enhancements. The iPhone (and, again, most other digital cameras) typically store photos in a compressed format to save space and with a variety of enhancements applied to make the image look great as soon as you click the shutter.
So Halide’s new format, which they’re referring to as “Process Zero,” is the app’s attempt to make RAW a little more user-friendly. When you open the app, you can choose from three settings: ProRAW (Apple’s custom RAW format that does apply some of the company’s image modifications), Apple Processed (which applies the same computational photography tricks that you’ll get when shooting with Apple’s default camera app) and Process Zero.
The left image is captured in Apple's ProRAW format, while the right is captured with Halide Process Zero.
Halide
Process Zero gives you a RAW file that you can then apply a quick image brightness adjustment to. One of the big benefits from shooting RAW is that you have wide latitude in brightening up a dark image, or toning down one that is blown out. After you make this adjustment, Halide saves the RAW plus brightness adjustment in a new JPEG file that you can then easily export to other apps like Instagram, VSCO, Lightroom, or whatever image editing tool you choose.
The idea here is to let photographers capture RAW images without the computational and algorithmic changes that Apple makes and then easily do something with those photos. So, when you shoot in Process Zero mode, the phone is taking just one image — unlike the Apple camera, which shoots multiple photos and combines them to make a more balanced result. So while that might lead to an image with more noise and with some darker or lighter areas, it can also be significantly sharper and capture more detail than Apple’s process. Halide posted a detailed blog with tons of info on how this all works, complete with examples, and I highly recommend you check it out if you’re curious.
The Halide team also mentioned that the company is working on a Mark III of their app. But unlike Mark II, which arrived with a ton of new features, they’re planning to early-launch some Mark III features to gather feedback; Process Zero is just the first of those. And if you’re curious to check out these RAW capture updates alongside whatever else is in the works, you can get a yearly subscription to Halide for $12 right now, down from the usual $20 price. (If you hate subscriptions, you can also buy Halide Mark II and the eventually III release outright for $60.)
This article originally appeared on Engadget at https://www.engadget.com/apps/iphone-camera-app-halide-has-an-anti-intelligent-mode-to-make-shooting-with-raw-easier-211802575.html?src=rss
Fujifilm’s X-T30 II is one of my favorite cameras, as it offers many of the capabilities of the higher-end models for a lot less money. So I was particularly excited when the company launched the X-T50, which is like a miniature version of the $1,700 X-T5.
It shares many features, including a 40.2-megapixel sensor, 6K 30P video support and, in a first for the series, in-body stabilization. At the same time, it retains the small size and retro-cute styling of past models.
At $1,400 it’s not nearly as affordable as the $900 X-T30, though. To find out if it’s worth an extra half grand, I tested the X-T50 out in London with some professional photographer friends.
Body and handling
The X-T50 is all about portability and style, so it’s nice and light at 438 grams (0.96 pounds), just 50 grams more than the previous model. If you’re traveling with a compact lens, it can easily be slipped into a bag.
At the same time, the grip has been enlarged enough to make the camera easier to hang on to. Operation is largely unchanged, with two control dials, a shutter speed dial and an exposure compensation dial, along with a joystick and eight buttons. However, Fujifilm did introduce one new trick: a film simulation dial in place of the previous drive mode — more on that shortly.
All of that makes it nearly as easy to control as much larger cameras. The dedicated shutter dial is an unusual feature these days, so it does take some getting used to. The advantage is that you can see that setting just by looking down at the camera, a must for discreet street or travel shooting.
Like past models in this series, the X-T50 has a tilt-only display that doesn’t flip out, which is likely a bummer for the vloggers it’s tantalizing with its video capabilities. And rival models like the Sony A6700 and Canon R7 do have flip-out displays. The 2.36-million-dot electronic viewfinder isn’t super sharp, but is par for the course in this price range.
With a smallish battery like the one on the X100 VI, the X-T50 can manage just 305 shots on a charge or 390 in economy mode. That’s near the bottom in its category. By comparison, Canon’s like-priced R7 can take up to 660 photos before the battery dies.
The X-T50 has a decent array of ports, including microHDMI, USB-C and a mic jack. As before, there’s no headphone input, but you can use the USB-C slot with a 3.5mm adapter to monitor audio. It comes with a single, but fast UHS-II card slot.
Performance
If it’s pure performance you’re after, the X-T50 probably isn’t for you. Burst speeds are decent at up to 13 fps in electronic shutter mode or 8 fps with the mechanical shutter — far from the 30 fps of the Canon R7 in electronic mode. Also, the electronic shutter is only good if your subject doesn’t move too quickly, as rolling shutter distortion is very noticeable in this model.
Steve Dent for Engadget
Autofocus still isn’t a Fujifilm strong point, either. When I was shooting quick-moving subjects, it was a bit hit or miss in continuous mode, particularly when they were close to the camera. However, performance depends greatly on the lens being used.
On top of face and eye detection, Fujifilm has added subject-tracking autofocus modes for animals, birds, cars, motorcycles, airplanes and trains. This feature is completely separate from the regular eye and face detection modes, so you have to dive into the menus to switch between them, or program two separate buttons.
For the other types of photography it's designed for, the X-T50's autofocus works well, though it isn’t quite up to Sony, Canon or even Nikon’s standards.
A key new feature is in-body stabilization. With up to seven stops of shake reduction, you can get sharp photos down to about a quarter second. That’s the same as the 557-gram X-T5, which is quite an accomplishment given the size difference. It makes the X-T50 far more versatile, letting you blur motion to create interesting shots or shoot in dark environments.
Image quality
With the same 40MP sensor as the X-H2 and X-T5, the X-T50 offers a major boost in resolution over the 26MP X-30 II. On top of boosting sharpness, the extra pixels make it possible to crop in, something that’s handy for travel photography, wildlife shooting or anything else where you might not be able to get close to the subject.
In-camera treatment of JPEG files is handled well, with pleasant, accurate colors and just the right amount of sharpening and noise reduction. You can often share photos straight out of the camera, with no need for any post-processing — an advantage for street and other photographers who don’t like to mess around in post.
The 14-bit RAW photos offer plenty of room for fine-tuning, as they retain high amounts of detail in highlights and shadows. As such, I saw low levels of banding, noise and other artifacts at low ISOs in all image areas.
Steve Dent for Engadget
The higher resolution doesn’t hurt image quality much in low light, either. Noise is controlled well up to ISO 6,400, and you can go up to ISO 12,800 in a pinch. Considering that the X-T50 is an APS-C and not a full-frame camera, I was impressed with the lack of noise when shooting in bars and other dim environments.
And of course, the X100 VI offers Fujifilm’s full array of film simulation modes, and they’re easily accessible on the new dial. You can experiment with popular looks like Velvia, Eterna, Acros black and white and others. It also comes with Fujifilm’s new Reala Ace, based on one of the company’s old negative films. At the same time, you get a RAW backup in full color, so you’re not locked into a specific look.
Video
With excellent video specs, the X-T50 is a solid hybrid camera, as long as you keep in mind some caveats — particularly around cropping and sharpness.
Much like the X-T5, you can shoot 6.2K at 30 fps with a 1.23x crop, or 4K at up to 60 fps with line-skipping and a 1.14x crop. The cropping occurs because 6K video only requires 24 megapixels, and the sensor/processor combination isn’t up to the job of supersampling the entire sensor width.
Steve Dent for Engadget
The camera also offers sub-sampled 4K at up to 30p using the full sensor width, or high-quality 4K 30p with a 1.23x crop. For super slow-motion, Fujifilm also introduced 10-bit and F-Log2 settings for extra dynamic range.
At 6.2K and 4K HQ, rolling shutter skewing is pronounced so you’ll want to avoid excessive or sharp camera movements. Sub-sampled 4K is better in that area, but video is noticeably less sharp.
Video autofocus matches what I saw with photos, meaning it was decent but not ideal for moving subjects. The AI-powered AF did lock onto subjects, but again, couldn’t always keep up with birds, animals or vehicles.
Handheld video is now a realistic option with in-body stabilization. It works well if you don’t make abrupt movements, and the “boost” mode offers near tripod-like smoothness. Digital stabilization is also available, but isn’t supported with the 6K or HQ modes, and doesn’t reduce jolts when walking.
Video quality is solid for such a small camera, offering the same accurate colors I saw in photo modes. Shooting in 10-bit F-Log made it possible to adjust footage considerably in post or get creative. You can also shoot video using the film simulation modes if you want a specific look straight out of the camera.
Wrap-up
Steve Dent for Engadget
All told, the X-T50 is an awesome upgrade to the X-T30 II, with improvements in nearly every area — particularly resolution, stabilization and video. The film simulation button is a bit gimmicky, but it should appeal to fans of the brand, particularly those who love the X100 VI.
The problem I have with this camera is the price. The $900 X-T30 II was easy to recommend to buyers on a mid-level budget. However, at $1,400, the X-T50 is only $300 less than the superior X-T5, and at that price it has lots of stiff competition from the likes of the Canon R7 and Sony A6700, along with full-frame models like the Canon R8, Nikon Z6 II/Z5/Zfc and Panasonic S9.
Fujifilm likely figured that the X-T50 was nearly as capable as the X-T5 and that the 40-megapixel resolution beats out rivals — but it has priced a lot of potential buyers out of the market. If you’re looking for something that costs around the same as the X-T30 II, Nikon’s $1,100 Zfc is just as pretty and a lot less expensive.
This article originally appeared on Engadget at https://www.engadget.com/cameras/fujifilm-x-t50-review-a-big-improvement-for-a-lot-more-money-180046809.html?src=rss
You can's say Fujifilm is boring. It stuck to APS-C sensors instead of going full-frame like everyone else, while releasing cool and weird models like the X100 VI. That strategy has been refreshing in a conservative industry and undeniably successful.
It also went big by introducing its first medium-format camera seven years ago, the GFX50S. After eight models, they’ve proven to be popular among pro portrait and scenic photographers, a market Fujifilm never really had before. Each has become increasingly more sophisticated, with better image quality, faster shooting speeds and improved video.
Now that the company’s flagship $7,500 100-megapixel GFX 100 II, has been out for awhile and had several firmware updates, I was keen to test the new AF speeds and more. So I went to London to try it out alongside two pro photographer friends who are thinking of buying one.
Body
The original GFX 100 is a gigantic camera, weighing over three pounds with the viewfinder. The GFX 100 II is more manageable at 2.27 pounds, the same as Panasonic’s full-frame S1. Photographers are still likely to be carrying a heavy bag, though, as medium-format GFX lenses are generally bigger and heavier than full-frame glass.
The GFX 100 II also feels more like a full-frame camera than an old-school top-down viewfinder medium format model. It has an updated, modern control layout, with a pair of control dials, a mode dial, a joystick, 14 buttons and a movie/photo switch.
The rear display tilts up, down and to the side, but doesn’t flip out — not a huge deal, as this will never be a vlogging camera. It shines where it counts, though, with a high 2.36 million dot resolution and enough brightness to use in sunlight. The viewfinder, meanwhile, is one of the best on any camera, with an extremely sharp 9.36-million dot resolution and 100 percent magnification.
It’s easy to handle, thanks to the well-placed controls and large grip. The top display, which stays on even when the camera is switched off, shows all the main settings at a glance. I’m not a huge fan of Fujifilm’s overly complicated menu system, but it’s fine once you get used to it.
As with other recent high-end cameras, you get both an SD UHS II card slot and a much faster CFexpress B option. The latter is required for fast burst shooting, as I’ll discuss soon. Battery life is solid, with up to 540 shots on a charge, or about an hour of 8K or 4K 60p recording.
Performance
Steve Dent for Engadget
The GFX 100 II is the fastest medium-format camera to date. You can fire bursts at up to 8 fps with the mechanical shutter enabled and capture about 300 lossless RAW frames before the buffer fills. That’s about 36GB of data, so it requires a fast CFexpress card.
Autofocus wasn’t a strong point on the GFX, but it’s a big step up on this model. The majority of shots in our burst testing were in focus, though it becomes less accurate when the subject is close to the camera. This isn’t a sports camera, obviously, but it still has the best AF I’ve seen on any medium format camera.
Face and eye detection have also improved, usually locking onto the eye and not, say the eyebrow as the older model did. Fujifilm also introduced AI subject detection from recent models, so it now has settings for animals, birds, automobiles, motorcycles, bikes, airplanes and trains.
Nathanael Charpentier for Engadget
The GFX 100 II has a new 5-axis stabilization system with up to eight stops of shake reduction, compared to 5.5 stops before. This is useful for portraits and scenics, letting you shoot down to a quarter second or slower and blur water or people, while keeping the background sharp.
Rolling shutter was pretty abysmal on the original model, and isn’t a lot better here. If you’re taking street photos and want to remain silent, it’s fine if the subject doesn’t move much. For anything else, use the mechanical shutter to avoid some bad skewing.
Image quality
Image quality is this camera’s forte. Naturally, photos are pin sharp thanks to the 102-megapixel sensor. And with 16 bits of color depth in RAW mode, dynamic range is outstanding, right up there with Sony and Nikon. All of that makes it ideal for portraits and landscapes, on top of tasks that benefit from high-resolution, like art preservation.
The GFX 100 II now goes down to ISO 80 instead of 100 to further boost dynamic range. All of that allows photographers to get creative with RAW photos, or tease detail out of highlights and shadows.
It’s not bad at high ISOs either, thanks to the sensor’s backside illumination and dual-gain design. There’s very little noise visible at ISO 6400, and photos are usable up to ISO 12800 if exposure is correct.
The medium format sensor offers incredibly shallow depth of field if you need that for portrait shooting. Combined with a fast lens like the 80 mm f/1.7, it allows for incredible bokeh and subject separation.
For those who prefer to use JPEGs straight out of the camera, it delivers color-accurate images with the perfect amount of in-camera sharpening. That’s ideal for previews or for folks who want to use Fujifilm’s impressive film simulation modes. For the GFX 100 II, Fujifilm introduced a new one called Reala Ace that’s based directly on one of its old negative films. With a punchy, saturated and slightly nostalgic feel, it has become one of my new favorites.
There is one quality issue — the GFX 100 II drops from 16- to effectively less than 14-bits when shooting 8fps bursts in order to reduce throughput. That in itself isn’t a huge problem, but Fujifilm has been cagey about how it markets this, which has rubbed a lot of pro photographers the wrong way.
Video
Steve Dent for Engadget
I’m starting to sound like a broken record, but the X100 II is also Fuji’s best medium format camera for video. It has a host of new modes, most notably 8K. It also offers 6K, 4K/60p and 1080p at 240fps. All those formats can be captured in 12-bit ProRes, along with 10-bit H.265 formats. You also get access to Fujifilm’s excellent F-Log2 capture that boosts dynamic range.
There are some considerable compromises, though. 8K is captured with a 1.53 times crop, reducing the effective sensor size to less than full frame — which negates one of the main medium-format advantages: shallow depth of field. Other resolutions use the full sensor width, but pixel binning reduces sharpness.
Rolling shutter is also an issue at 8K, so be sure not to move the camera much at that resolution. It’s less bothersome at 4K resolutions, likely due to the pixel binning.
All that aside, video from the GFX100 II has a different quality than I’ve seen from most mirrorless cameras. The larger sensor makes it cinematic, especially with some of Fujifilm’s prime lenses. And the 8K video is extremely sharp when downsampled to 4K in DaVinci Resolve.
Realistically though, video is more of a nice-to-have feature for occasional use, as the majority of buyers will certainly be using it for photography.
Wrap-up
Nathanael Charpentier for Engadget
The $7,500 GFX100 II is an impressive medium format camera with improvements in every area compared to the previous model. More importantly, what did my pro photographer friends think and will they buy one? “What’s most noticeable is the evolution of the autofocus compared to the GFX100,” said Nathanael Charpentier. “In our studio we usually work with Sony, and the GFX100 II autofocus is still far from Sony’s level, but it’s a big improvement.
“It’s not a sports camera, it doesn’t have super-fast burst speeds. It’s more for studio portrait work. For certain types of ‘reportage’ like candid wedding shoots, if we really need the extra dynamic range offered by a medium-format camera, I could see using it.” At this point, they’re not planning on buying one due to the high price (and the fact that they just laid down 6,000 euros for an A9 III), but it’s high on their list of future equipment purchases.
Its main competitor is the $8,200 Hasselblad X2D 100C, which has perhaps slightly better color science and image quality — while also bringing a certain prestige with the Hasselblad name. However, the GFX100 II is superior in most other ways, including speeds, autofocus and video. If you really need to nail autofocus in busy or difficult situations, though, full-frame is still best: Sony’s 45-megapixel $6,500 A1 or Nikon’s $3,800 Z8 or $5,500 Z9 (both 45MP as well) are better choices.
This article originally appeared on Engadget at https://www.engadget.com/fujifilm-gfx-100-ii-the-king-of-medium-format-mirrorless-cameras-143009929.html?src=rss
Canon has unveiled its most important camera in years — the EOS R1 mirrorless. Launched alongside the 45-megapixel R5 II, it’s the company's new flagship designed to replace the 1DX Mark III DSLR and help Canon maintain its leadership in the pro sports photography field.
The R1 is all about speed, with the stacked sensor allowing 40 fps RAW bursts with continuous autofocus. Other features are designed to help nail crucial shots, including pre-capture, eye-tracking AF and sports-specific settings. At the same time, it should be great for video, thanks to its support for 6K RAW capture.
One potential downside is the 24-megapixel resolution, which is lower than rivals like Sony’s A1 and the Nikon Z9. To see how that affected image quality and gather some impressions on performance and new features, I was able to get a short amount of hands-on time with the R1 in Phoenix, Arizona.
Though massive for a mirrorless camera, the R1 is lighter than it looks — it only weighs 2.12 pounds. The magnesium alloy body is dust- and water-resistant and the large, comfortable grip uses new rubber to reduce slippage. It’s a stable platform, even with telephoto lenses, and was clearly designed for working photographers.
Controls are laid out much like the EOS R3, with two control dials, a rear dial, joystick, mode switch and many customizable buttons. It has matching vertical controls, with the same layout in portrait and landscape modes.
Canon did add a touch-sensitive IR button called the smart controller, which offers three ways to adjust autofocus. Overall, the R1 handles like an F1 car — fast and reactive, but there's a steep learning curve.
The viewfinder is Canon’s best yet with 9.44 million dots, 0.9 times magnification and 3x brightness, all tops in the EOS series. The rear display fully articulates for high- and low-angle shooting.
Canon decided to ditch SD cards in favor of two CFexpress type B slots. That allows for maximum RAW burst speeds, as well as fast and reliable backups, but you'll end up throwing your SD card collection out the window. By contrast, Sony's R1 has dual card slots for both SD UHS II and CFexpress type A.
The R1 uses a variant of the battery from the EOS R3 that supports up to 650 shots and 70 minutes of 6K video on a charge. Connectivity is the best I’ve seen on any Canon camera, with Wi-Fi 6E, ethernet, USB-C PD and more.
The R1 is built for action and suffice to say, it’s fast. In blackout free electronic shutter mode, you can shoot JPEG and RAW frames at up to 40 fps for a long while before the massive buffer fills — with Canon promising "1,000 or more" shots when shooting uncompressed RAW and JPEG files. It hits 12 fps with the mechanical shutter, a feature missing on Nikon’s rival Z9.
Steve Dent for Engadget
The latter isn’t usually necessary, though, as the stacked shutter largely eliminates rolling shutter distortion.
Canon is bringing numerous autofocus improvements as well. With regular single-point tracking, you can change the target with either the joystick or smart controller. In addition, the R1 has a new eye-tracking system called Eye AF II, a more sophisticated version of its Eye Control AF from the R3.
As with the R3, you must calibrate Eye AF II first. And it won’t necessarily function well for all eye types, but it worked great for me, as the tracking box followed my gaze with less drift than before. That didn’t always translate to sharp photos, but again, I was using a pre-production unit.
Of course the R1 offers face- and eye-detection for both humans and animals. But what happens if you’re trying to single out someone in a crowd of players?
Steve Dent for Engadget
First off, there are now AI modes for three specific sports, volleyball, basketball and soccer. The idea is that the system will track action around the ball to help you nail key shots. It’s also supposed to stay locked on a player even in traffic. Finally, you can register up to 10 faces to keep track of a whole team of athletes.
How well does this work? It’s hard to say in the short time I had the camera, but I didn’t notice a dramatic improvement with chaotic action or specific sports. But these features require time to master, so a full review will tell me more.
Another new feature to help pros avoid missing shots is pre-continuous shooting, something that's seen on many high-end cameras these days. The R1 starts taking photos when you half-press the shutter button, recording up to the last 20 frames once it’s fully pressed.
Canon has led the way in image stabilization since the R5, and the R1 offers a solid 8.5 stops of shake reduction with supported lenses. This isn’t necessarily a great benefit for action shooting, but it does help with landscapes or low-light shooting.
Canon likely anticipated complaints about the relatively low 24-megapixel resolution. In response, it unveiled a new feature called in-camera upscaling. It uses deep learning to quadruple image size to 96 megapixels in-camera after shooting, supposedly while maintaining apparent resolution.
This is different from Sony's pixel-shift scheme that takes multiple separate photos and merges them into one. However, it was hard to get a straight answer from Canon as to how exactly the upscaling works. Obviously, it can’t be a true resolution increase without actually capturing extra pixels.
Image quality is difficult to judge in pre-production cameras, and Canon specifically said that in-camera RAW processing was a work in progress. Nonetheless, JPEGs looked sharp out of the camera and had the company’s typically excellent skin tones.
Low-light image quality is a selling point of this camera, and it did appear to perform well at high ISOs, with decent image quality at ISO 12,800 and beyond. However, it’s no Sony A7S III at very high ISOs, so the max of ISO 102,400 is for emergency use only.
Steve Dent for Engadget
Stacked sensor cameras tend to sacrifice some dynamic range in favor of speed, but there’s no sign of that with the R1. However, I'll need more time to get a full idea of RAW image quality, particularly at high ISOs.
Though mainly aimed at photographers, the R1 is potentially a monster video camera on its own. It lacks the R5 II’s 8K feature, but does offer 6K Canon RAW video at up to 60 fps, using the full width of the sensor.
You also get fully oversampled full-sensor 4K at up to 60fps, or 4K pixel-binned or cropped at 120 fps. Full HD can be recorded at up to 240 fps, either cropped or with pixel-binning. On top of Log 3, Canon’s Log 2 is now supported with 10-bit capture.
Canon has significantly reduced its previous overheating issues, though they're not completely eliminated. You can record more than 120 minutes of 6K 60p RAW with proxy recording, as well as 4K 60p, with no restriction for 4K 30p. The primary limitation is 15 minutes for 4K 120p, but that shouldn’t be an issue for most videographers or content creators.
As you’d expect, the oversampled 4K 60p and 6K 60p RAW video is sharp. Canon’s RAW video is very easy to work with in post and with the Log 2 setting delivers extra dynamic range and latitude in post.
Steve Dent for Engadget
It’ll be interesting to see how pro photographers react to Canon’s new flagship. They do gain an extra 10 fps in RAW shooting speeds over Sony’s A1 and the Nikon Z9, and the sensor is solid in low light. The downside, though, is the fact that it offers half the resolution of its rivals.
The Eye AF II is an interesting gimmick that may give Canon an edge. Other features like face-memorization, sports-specific AF and cross action seem good on paper, but I wasn’t able to get the most out of them in the short time I had the camera — so further testing is required. The EOS R1 is set to arrive in the fall of 2024 with an estimated price of $6,300.
Canon paid for this contributor's travel and accommodation for the launch of its new cameras in Phoenix, Arizona. This article was produced independently for Engadget, with no oversight from Canon.
This article originally appeared on Engadget at https://www.engadget.com/canon-r1-hands-on-incredible-speed-but-24mp-resolution-may-disappoint-100044085.html?src=rss
As it teased earlier, Canon has launched the R5 II, a successor to the powerful but imperfect EOS R5. With a new 45-megapixel backside-illuminated (BSI) stacked sensor, it not only has superior specs for video, shooting speeds and more, but also adds advanced features like eye-controlled AF. The R5 II was launched alongside Canon's new flagship, the EOS R1, which I've covered in a separate post.
With the new R5, Canon has mostly dealt with the original's primary problem: overheating while shooting video. To see what’s different and try out some of the new features, I spent some time with an R5 II pre-production camera in Phoenix, Arizona.
The R5 II’s body is largely the same as before, but there are a couple of key changes. First off, Canon has moved the power switch over to the top right, next to the mode dial. On the left, where the on/off switch used to be, the company added a dedicated video and photo switch. That should make it easier for hybrid shooters to switch modes and keep settings separate.
Other controls are mostly where they were. The original R5 was already a sweet-handling camera, with a comfortable grip and logical control layout, so Canon wisely left well enough alone.
Also unchanged is the flip-out rear touch display, but the R5 II does have a new OLED electronic viewfinder. It has the same sharp 5.76-million dot resolution — on par with other models in this price range — but is brighter than before.
It supports both UHS-II SD cards and CFexpress Type B, with a slot for each. If you want to shoot 8K RAW or sustain long shooting bursts, you’ll need to use CFexpress cards, which are much faster but considerably more expensive.
Canon introduced three new battery grips for the R5 II, with one containing a cooling fan. That largely reduces heating issues, as I’ll discuss shortly. Another grip is designed to improve vertical operability, making the R5 II handle more like the R1. The third grip adds 2.5G base-T ethernet for studio shooters.
With the new stacked sensor, the EOS R5 II is faster than ever. You can shoot blackout-free bursts at up to 30 fps in electronic shutter mode, or 12 fps using the mechanical shutter. Rolling shutter is also substantially reduced. The R5 II has a large buffer as well, supporting up to 93 RAW and 20 compressed RAW shots before it fills in electronic shutter mode.
Steve Dent for Engadget
Autofocus is improved in nearly every way, too. Like the R1, the R5 II comes with Canon’s new EyeAF II eye-tracking autofocus that sets focus where you look. Canon touts it as the fastest way to to move the AF tracking point.
It needs to be calibrated, and won't work with all eye types, but in my brief experience it worked for me. It can be a bit twitchy and isn’t 100 percent reliable, but I didn’t have enough time to really shake it out — so stay tuned for a full review.
Canon also added sports-related features like an action priority mode, subject crossing and the ability to register multiple faces. Face- and eye-tracking is improved, and it supports animals, birds and vehicles, too. Of course, it has 5-axis in-body stabilization with 8.5 stops of shake reduction, up from 8 before.
Shooting in the basketball scenario Canon set up, I found the AF to be generally reliable. The new pre-continuous shooting feature is also a nice bonus, helping ensure you capture shots even before fully pressing the shutter. However, some new things like action crossing require practice to get used to and I didn’t have enough time with the camera to adapt.
It’s hard to evaluate image quality in pre-production cameras, and Canon warned us that it was still working on RAW processing. Still, the benefit of a 45-megapixel sensor is seriously sharp photos.
JPEG images appeared color accurate with pleasing skin tones, as you’d expect from Canon. Using the company’s software, I was able to process the 14-bit RAW images and found that they offer plenty of latitude for pushing and pulling in post.
Like the R1, the R5 also supports Canon's new in-camera upscaling, which quadruples resolution to around 179 megapixels. This uses deep learning to reduce noise, but Canon didn't explain precisely how it works.
A lot of my shooting was done in a gym with relatively low light, but the R5 II handled the high ISOs without generating excessive noise up to ISO 6400 or even beyond. It’s not in the same league as the R1, though, so you’ll need fast lenses if shooting action in dark spaces.
Steve Dent for Engadget
The R5 II is a seriously powerful video camera, right up there with Nikon’s Z8 or Z9. You can shoot 8K DCI RAW video at up to 60 fps, or 4K RAW at 60 fps. It also offers 8K MP4 support at up to 30 fps, or 4K DCI at 120 fps with pixel binning.
4K profiles at up to 30 fps use the full sensor width in “fine” mode, or you can capture higher frame rates with pixel binning. Full HD can be recorded at 240 fps with pixel binning as well. Most 4K modes can also be captured cropped to an APS-C frame size.
The addition of RAW recording and C-Log2 boosts dynamic range, and all the oversampled video modes are very sharp. Canon’s RAW is also easy to work with in Premiere Pro or DaVinci Resolve.
That takes us to the most glaring issue: overheating. The original R5 certainly struggled in this respect, with temperature limiting recording times in some modes.
Steve Dent for Engadget
With the new camera, the only modes affected are 8K 60p RAW, which is capped at 18 minutes, 8K 30p limited to 37 minutes and 4K 120p, which shuts down after 22 minutes. Adding the fan accessory removes most of those limitations and increases operating times at 8K 60p RAW and 4K 120P.
With incredible video and photos specs, Canon’s R5 II could retake its old crown as the best hybrid camera. The only model that can keep up with it for photos is the $6,500 Sony A1, but it's over $2,000 more expensive. Nikon’s $3,500 Z8 can match it video-wise with 8K RAW at up to 60P. However, it isn’t as good for photos as the R5 II, because speeds are limited to 20 fps and it lacks a mechanical shutter.
The R5 II will be more expensive than the Z8, however, at an estimated $4,300. It will arrive in stores sometime in August.
Canon paid for this contributor's travel and accommodation for the launch of its new cameras in Phoenix, Arizona. This article was produced independently for Engadget, with no oversight from Canon.
This article originally appeared on Engadget at https://www.engadget.com/canon-eos-r5-ii-hands-on-nifty-eye-tracking-autofocus-and-reduced-overheating-problems-100002588.html?src=rss