Month: September 2019

We all expect Canon to announce its EOS 1D X Mark III before the Tokyo olympics, but the rumor miss has been surprisingly quiet on what exactly Canon is planning to do with their next flagship DSLR… until now.

According to Canon Rumors, a “very solid source” has confirmed that the 1D X Mark III will arrive in 2020, and shared a few very exciting specs that promise to compete with, and possibly outperform, Nikon’s upcoming D6.

CR’s source, who seems to have hands-on experience with a 1D X Mark III prototype, says the current iterations have in-body image stabilization (no surprise there), a “significant jump in resolution for an EOS-1 series camera,” and boasts a new DIGIC processor, Dual CFExpress card slots, and the ability to shoot 6K video without a crop. That last tidbit should delight video shooters, and implies that 4K/60p video without a crop is also highly likely.

The source didn’t say when exactly Canon plans to announce their latest, and possibly last, full-frame DSLR flagship, but “early” 2020 is a good bet given the timing of the summer olympics.

It’s no secret that camera quality is near the top of the list when it comes to requirements for a smartphone. And that’s the case no matter how much you’re spending. Thankfully there are plenty of phones out there that offer solid camera experiences without breaking the bank. The best of the best is the Google Pixel 3a. It gives you flagship camera quality, front and back, for a fraction of the cost — and stands out particularly well at night and for selfies.

Best Overall: Google Pixel 3a

The Pixel 3a has almost a perfect copy of the much more expensive Pixel 3’s camera, and that’s a great thing for everyone. The Pixel 3a’s rear camera is super-capable in any lighting and is particularly great in low light thanks to the Night Sight shooting mode. Flip over to the front camera, and you get crystal-clear selfies with a good field of view, auto focus and a good portrait mode.

The rest of the phone rounds out well, too. You get Google’s clean and always up-to-date take on software, along with good-enough specs and hardware that’s befitting the price point. Really, there are no notable downsides considering where this phone’s competing.

Pros:

Flagship-level photo quality, front and back

Best-in-class low-light photos

Simple Google software

Guaranteed software updates

Cons:

Plasticky build

Average battery life

Best for Less: Moto G7

The standards for camera quality are certainly lower when you’re talking about reallyinexpensive phones, but among a weaker segment in the grand scheme of smartphones, the Moto G7 stands out. It keeps things simple with a solid main camera that has above-average quality in daylight and still manages to make things happen in low light. Camera speed is good, too, but don’t get too excited by its extra depth-sensing camera — it’s not worth much.

Elsewhere the Moto G7 makes a compelling case for being the best phone in its price bracket. You get a big display, solid battery life, and nice little software features that separate it from the crowd. The camera isn’t the driving factor in your buying decision here, but it’s a big part.

Pros:

Solid main camera quality

Double-twist camera launch gesture

Incredible value for money

Big 6.2-inch display

Cons:

Questionable software update future

Secondary depth camera not useful

No NFC

Best Value: Samsung Galaxy A50

Samsung is best known for its high-end phones, but the A50 has stepped into the mid-range with a winner. The 25MP main camera is capable in good and mixed lighting, though as expected for this price point comes up short in lower light. But what’s surprising for this money is the addition of both a dedicated wide-angle camera, which takes fun shots with a new perspective, and a dedicated depth camera for portrait shots.

The 6.4-inch display is both bigger and much nicer than you’d ever expect for something that competes in the mid-range, as is the beautiful hardware. The spec sheet is strong, with a 4000mAh battery that gives you great longevity. You’ll have to face some Samsung-induced bloatware in the software, but the A50 is a great overall phone that isn’t at all let down by its cameras.

Pros:

Flagship-style triple camera

Top-tier display

Strong battery life

Great specs for the money

Cons:

Considerable bloatware in the software

Portrait mode shots are weak

No MST Samsung Pay support

Best for Selfies: ASUS ZenFone 6

The ZenFone 6’s rear camera is solidly above average and is really strong for its price. The trick here is that it can flip over and become the front-facing camera. That instantly makes the front-facing camera experience miles ahead of anything else in terms of resolution and lens quality. You get rather exceptional selfies from this camera, without any of the typical shortcomings of a tiny front-facing camera.

The rest of the phone is great for the money, as ASUS has completely refreshed its software experience and settled into some really nice-looking and nice-feeling hardware. All of the specs, including a big battery, exceed expectations for the price point.

Pros:

Excellent selfies from flip-up rear camera

Clean software experience

Above-average rear camera quality

Great battery life

Cons:

Weak display for the money

Camera tends to over-sharpen

More expensive than Pixel 3a XL

Best Internationally: Xiaomi Mi 9T Pro

If you’re outside the U.S. (or looking for something a little different), the Xiaomi Mi 9T Pro could be up your alley. You’ll find a 48MP main camera on the back that can be found in tonsof higher-end phones this year, which produces great photos that can go toe-to-toe with the competition. Plus it has dedicated ultra-wide and telephoto cameras — that’s not very common at this price range — and a neat pop-up selfie camera that saves you from a display notch.

Like most Xiaomi phones it has some over-built specs for its price, like a Snapdragon 855 and 6GB of RAM, plus a huge 6.4-inch display and an equally large 4000mAh battery. That’s a great package with a great set of cameras to match.

Pros:

Very capable triple camera

Near-flagship specs

Big, beautiful display

Excellent price

Cons:

Slow face unlock from pop-up camera

MIUI has software issues

Mixed international availability.

Bottom line

If camera quality is the top consideration when buying a phone, but you’re on a budget, the Google Pixel 3a is absolutely the best choice. Its camera is lifted right from the high-end Pixel 3, and it shows — it’s dramatically better than anything else at the same price. Also, if you want a bigger version, you can pick up a Pixel 3a XL for just a little more.

The Pixel 3a’s cameras may not seem that impressive on paper, and it doesn’t have the variety of shooting methods or lenses to choose from, but it’s all about the results here. Front or back, the 3a’s cameras take exceptionally sharp, balanced photos — and its Night Sight will produce better low-light shots than just about any other phone, regardless of price.

While there are numerous HP Chromebook X360 models across a $1,000 price range, consumers and students are probably most familiar with the $299 HP Chromebook X360 11. That entry-level device uses an 11.6-inch display with rather large bezels. And it appears to be getting an upgrade, losing some bezel size and gaining a barely larger 12-inch display in the new HP Chromebook X360 12.

This model hasn’t been officially announced yet, but it does currently appear on Amazon’s site, complete with all of the specifications. Roland Quandt spied both the product listing and shared it on Twitter (via Brandon Lall), as well as a bevy of official HP Support videos for the HP Chromebook X360 12.

Here’s a full view look at the newly designed HP Chromebook X360 12 with the older model underneath it. The look is far more up to date and you can easily see the diminished bezel size:

The 12-inch model appears sleeker, has a new hinge mechanism for what looks to be a 3:2 display and the B&O logo, indicating tuned speakers if you’re into that sort of thing. I’d say that the older Chromebook might be better suited for students as it appears more ruggedized to me. It’s possible that HP will sell both models instead of replacing the 11.6-inch Chromebook with one that has a barely bigger display.

The internal specifications that Amazon provides for the HP Chromebook X360 12 don’t quite make sense to me. Here’s the listing, which I’ll follow up with my reasoning for that statement:

CPU: Dual-Core Intel Celeron N with 2.6 GHz clock speed

GPU: Intel UHD Graphics 620

Memory: 4 GB LPDDR4 RAM

Storage: 32 GB eMMC

Screen: 12-inches with HD+ resolution and touch support

The processor and GPU combination is what stands out to me as being odd: According to Intel, the Celeron N4000 would be the dual-core CPU with that maximum clock speed. But the N4000 uses the Intel UHD Graphics 600 GPU, so that’s not lining up. I’m guessing the display is actually 12.3- or 12.5-inches as well.

And frankly, if the display is HD+ as Amazon says, it really doesn’t make sense to have a dual-core Celeron from 2017 powering a Chromebook with a higher-end design like this.

My gut says while there may be a low-end configuration with Celeron or Pentium options, as well as HD or FHD display, we’ll see models with Intel Core chips, more storage, and RAM as well.

Note that HP does sell a similar looking HP Chromebook X360 14 model (shown above) with an MSRP of $599. That device is powered by an Intel Core i3-8130U and the previously mentioned Intel UHD Graphics 620, paired with 8 GB of RAM and 64 GB of eMMC storage as well as the B&O tuned speakers.

If the HP Chromebook X360 12 is offered with similar internals, I’d guess an MSRP between $400 and $450 in such a configuration for those that want a nice-looking, solid-performing but more portable 2-in-1 Chromebook.

The Samsung Galaxy Note10 and Note10+ are the first smartphones to get the “Wi-Fi Certified 6” designation from the Wi-Fi Alliance. This signifies that the Notes fully support the new standard and will allow networks to function optimally.

The Notes aren’t alone, the Galaxy S10 trio also features Wi-Fi 6 connectivity. However, it seems they are not getting the “Certified” accreditation. You can find the full list of devices here. Note that the new iPhone 11 trio also features Wi-Fi 6 connectivity, but isn’t on the Certified list yet.

In case you missed it, Wi-Fi adopted a user-friendly naming scheme last year. Wi-Fi 6 is the new name of the 802.11ax standard, Wi-Fi 5 is 802.11ac and so on.

As for what’s new in Wi-Fi 6, it helps mobile devices save battery with the Target-Wakup Time. TWT tells the phone when it needs to wake up to transfer data and lets it sleep the rest of the time. There are also enhancements to MIMO and coding efficiency, which allow more data to be pushed through the same bandwidth. TWT helps here too, by ensuring that only a few devices are awake at the same time, reducing congestion.

There are already several hotspots and routers supporting the new standard like the Huawei 5G CPE Pro and the Asus RT-AX88U.

Modern artificial intelligence is often lauded for its growing sophistication, but mostly in doomer terms. If you’re on the apocalyptic end of the spectrum, the AI revolution will automate millions of jobs, eliminate the barrier between reality and artifice, and, eventually, force humanity to the brink of extinction. Along the way, maybe we get robot butlers, maybe we’re stuffed into embryonic pods and harvested for energy. Who knows.

But it’s easy to forget that most AI right now is terribly stupid and only useful in narrow, niche domains for which its underlying software has been specifically trained, like playing an ancient Chinese board game or translating text in one language into another.

Ask your standard recognition bot to do something novel, like analyze and label a photograph using only its acquired knowledge, and you’ll get some comically nonsensical results. That’s the fun behind ImageNet Roulette, a nifty web tool built as part of an ongoing art exhibition on the history of image recognition systems.

As explained by artist and researcher Trevor Paglen, who created the exhibit Training Humans with AI researcher Kate Crawford, the point is not to make a judgement about AI, but to engage with its current form and its complicated academic and commercial history, as grotesque as it might be.

“When we first started conceptualizing this exhibition over two years ago, we wanted to tell a story about the history of images used to ‘recognize’ humans in computer vision and AI systems. We weren’t interested in either the hyped, marketing version of AI nor the tales of dystopian robot futures,” Crawford old the Fondazione Prada museum in Milan, where Training Humans is featured. “We wanted to engage with the materiality of AI, and to take those everyday images seriously as a part of a rapidly evolving machinic visual culture. That required us to open up the black boxes and look at how these ‘engines of seeing’ currently operate.”

It’s a worthy pursuit and a fascinating project, even if ImageNet Roulette represents the goofier side of it. That’s mostly because ImageNet, a renown training data set AI researchers have relied on for the last decade, is generally bad at recognizing people. It’s mostly an object recognition set, but it has a category for “People” that contains thousands of subcategories, each valiantly trying to help software do the seemingly impossible task of classifying a human being.

And guess what? ImageNet Roulette is super bad at it.

I don’t even smoke! But for some reason, ImageNet Roulette thinks I do. It also appears to believe that I am located in an airplane, although to its credit, open office layouts are only slightly less suffocating than narrow metal tubes suspended tens of thousands of feet in the air.

ImageNet Roulette was put together by developer Leif Ryge working under Paglen, as a way to let the public engage with the art exhibition’s abstract concepts about the inscrutable nature of machine learning systems.

Here’s the behind-the-scenes magic that makes it tick:

ImageNet Roulette uses an open source Caffe deep learning framework (produced at UC Berkeley) trained on the images and labels in the “person” categories (which are currently ‘down for maintenance’). Proper nouns and categories with less than 100 pictures were removed.

When a user uploads a picture, the application first runs a face detector to locate any faces. If it finds any, it sends them to the Caffe model for classification. The application then returns the original images with a bounding box showing the detected face and the label the classifier has assigned to the image. If no faces are detected, the application sends the entire scene to the Caffe model and returns an image with a label in the upper left corner.

Part of the project is also to highlight the fundamentally flawed, and therefore human, ways that ImageNet classifies people in “problematic” and “offensive” ways. (One interest example popping up on Twitter is that some men uploading photos appear to be randomly tagged as “rape suspect” for reasons unexplained.) Paglen says this is crucial to one of the themes the project is highlighting, which is the fallibility of AI systems and the prevalence of machine learning bias as a result of its compromised human creators:

ImageNet contains a number of problematic, offensive and bizarre categories – all drawn from WordNet. Some use misogynistic or racist terminology. Hence, the results ImageNet Roulette returns will also draw upon those categories. That is by design: we want to shed light on what happens when technical systems are trained on problematic training data. AI classifications of people are rarely made visible to the people being classified. ImageNet Roulette provides a glimpse into that process – and to show the ways things can go wrong.

Although ImageNet Roulette is a fun distraction, the underlying message of Training Humans is a dark, but vital, one.

“Training Humans explores two fundamental issues in particular: how humans are represented, interpreted and codified through training datasets, and how technological systems harvest, label and use this material,” reads the exhibition description “As the classifications of humans by AI systems becomes more invasive and complex, their biases and politics become apparent. Within computer vision and AI systems, forms of measurement easily — but surreptitiously — turn into moral judgments.”