Instagram is reportedly testing a feature with a “small number” of users where its “suggested posts” feature will expand beyond just when you’ve reached the end of your feed and will be mixed throughout a browsing experience, possibly coming ahead of posts from those users follow.
In a confirmation to The Verge, Instagram says that reception to the “suggested posts” feature was so positive that the company decided to try and mix those suggestions in with the average viewing experience, sometimes ahead of photos and videos from people a user explicitly follows.
Additionally, Instagram is testing new controls that will allow users to add a specific topic of interest for suggested posts as well as the ability to “snooze” the recommendations for 30 days or hide them from a feed entirely. “Suggested posts” is a feature that Instagram added last year, but prior to this small test was only ever seen after a user saw everything from all people they followed that was shared.
This shift would make Instagram theoretically function similarly to how YouTube manages its “home” page, which highlights content that is a mix of videos made by those a user is subscribed to as well as videos that YouTube’s algorithm thinks are of interest. YouTube leans so heavily on this analytics-forward approach that subscriptions have fallen in importance over the years.
If Instagram were to adopt placing suggested posts for all users, it may have a dramatic impact on how the social network functions. Instagram has a vested interest in keeping users on the platform for as long as possible, and keep them coming back. As such, the social network should not necessarily need to put as much value on showing a user content from people they follow as opposed to delivering photos and videos that keep them engaged. As YouTube has shown, just because a person subscribes to a Channel doesn’t mean that the user necessarily wants to see all content that Channel produces.
That’s the theory, anyway, and it might have ramifications in the long term for content creators who rely on Instagram to reach an audience, especially if suggested posts become more popular and show higher engagement than that of specifically followed accounts.
Algorithm-focused approaches have generally superseded ones that focus on giving users complete control over what they see. Facebook and Instagram both ditched a timeline-based approach years ago, and Twitter defaults users to the “Home” view instead of “Latest Tweets” which uses an algorithm to determine the most interesting content. At least in Twitter’s case, switching back to Latest Tweets is an option.
Instagram did not specify how many people its “suggested posts” test would affect nor how long it intended to test the feature.
Plans for this year’s Photography Show, which opens its doors between 18 and 21 September, are well underway, according to organisers Future Publishing.
Rankin and fine art photographer, Brooke Shaden, are already confirmed to speak on the Super Stage. Other highlights will include the new Shutter Street stages, plus the Photo Live and Video Live stages, and the Editing & Post-Production, Behind the Lens and The Studio theatres. There will also be a wide range of masterclasses.
Hundreds of brands are signed up, according to the organisers, including Canon, Nikon, Sony, Fujifilm, Olympus and Epson, plus retailers Wex Photo Video, London Camera Exchange and CameraWorld. The exhibition floor is reportedly 90% full, with just over four months to go.
Rankin is confirmed as a Super Stage speaker
“We’re very positive that The Photography Show & The Video Show is on for a cracking year,” said Ruth Folkard, event director. “Photographers and filmmakers have been unable to get hands-on with kit or get that all-important face-to-face time with the brands they want to purchase from for months; it’s a top priority, now more than ever – and our visitors are telling us how much they’re looking forward to the show daily.”
“We’re really excited to be planning Canon’s presence at the The Photography Show in September, and can’t wait to see customers face-to-face at the NEC”, added Tracey Fielden, Canon’s marketing director UK & Ireland.
The show is a great chance to get hands-on with the latest kit – restrictions permitting
Despite the excitement, the effects of the pandemic have not been forgotten. “Though we’re hopeful that restrictions will fall away fully on 21 June, we’re mindful that some elements may still be in place for a while,” said Ruth Folkard.
“Either way, visitors and exhibitors alike will want to feel confident and safe at the event. We’ve planned for a number of measures such as staggered entry and enhanced sanitisation, alongside wider aisles and improved crowd control to keep traffic flowing, as a minimum. We are following guidance carefully and are ready to implement further adjustments as required.”
The Photography Show & The Video Show will run between 18 and 21 September (Saturday to Tuesday) in halls 2 and 3 at the NEC Birmingham. Note that 2020 ticket holders will automatically have tickets transferred to these dates and will receive communications from the event team over the coming weeks. Information and updates will be added regularly here.
We’ve already tested the camera of the Exynos-powered international version of Samsung’s new S-series flagship, the Galaxy S21 Ultra 5G. In this review, we take a close look at the version of the S21 Ultra 5G that uses a Qualcomm Snapdragon 888 chipset and is marketed primarily in North America and China.
Chipsets aside, the Exynos and Snapdragon versions are identical, with both featuring S-Pen stylus support and a 6.8-inch Dynamic AMOLED 2x display with 20:9 aspect ratio, 3200 x 1440-pixel resolution, and flexible frame rates up to 120 Hz.
For the camera, both models combine two dedicated tele-lenses (72- and 240-mm equivalent) with a standard wide and ultra-wide camera. That’s an interesting modification over last year’s S20 Ultra, which came with a single 103 mm-equivalent periscope tele-lens.
Let’s see how the Samsung Galaxy S21 Ultra 5G (Snapdragon) compares to its Exynos cousin and to some major competitors from rival brands.
About DXOMARK Camera tests: For scoring and analysis in our smartphone camera reviews, DXOMARK engineers capture and evaluate over 3000 test images and more than 2.5 hours of video both in controlled lab environments and in natural indoor and outdoor scenes, using the camera’s default settings. This article is designed to highlight the most important results of our testing. For more information about the DXOMARK Camera test protocol, click here. More details on how we score smartphone cameras are available here.
Samsung Galaxy S21 Ultra 5G (Snapdragon)
With a DXOMARK Camera overall score of 123, the Samsung Galaxy S21 Ultra 5G (Snapdragon) achieves two points more than the Exynos model but is still some ways off the current best-in-class models and doesn’t make it into our current top ten ranking.
Its Photo score of 129 is one point higher than for the Exynos device. This is mainly based on better texture and noise as well as a slightly better preview image. Autofocus performance is a little worse with the Snapdragon chipset, however.
The Snapdragon model also scores two points higher for Zoom, with slightly better results in both the tele and wide categories. The score is four points higher for Video, with the Snapdragon-powered camera delivering better results for autofocus, texture, and stabilization. On the downside, noise is a little worse than on the international Exynos version.
The S21 Ultra 5G with Snapdragon chipset produces pleasant images with vivid colors, natural skin tones, and good levels of detail.
Still images captured on the Samsung Galaxy S21 Ultra 5G (Snapdragon) show good exposure in most conditions, and a fairly wide dynamic range means the camera deals well with difficult high-contrast scenes. This said, highlight clipping can often be seen in backlit conditions, and exposure can be unstable between shots, especially when shooting in HDR conditions.
Samsung Galaxy S21 Ultra 5G (Snapdragon), good target exposure on face, highlight clipping in the background
Huawei Mate 40 Pro, good target exposure, better highlight retention in the background
The camera delivers vivid colors and nice skin tones with an overall color rendering that is similar to the Exynos model’s, but there are more white balance casts in low light. In terms of texture/noise, the Snapdragon model is an improvement over the Exynos variant, with better detail rendering and lower noise levels. Fine detail is well preserved and noise is mostly visible only in low light.
Samsung Galaxy S21 Ultra 5G (Exynos), crop: slightly less detail
Apple iPhone 12 Pro Max, indoor texture
Apple iPhone 12 Pro Max, crop: slightly less detail
The difference between the Snapdragon version and the international Exynos-powered version is even more visible in low light. In this scene the Snapdragon device produces noticeably lower noise levels than the Exynos version.
Image artifacts are reasonably well controlled, but strong aliasing as well as color fringing and ghosting can be found in some images that are captured in bright light outdoors.
Autofocus is an area where the performance of the Snapdragon device drops slightly behind the Exynos-powered model, with occasional autofocus failures and focus noticeably slowing down in high-contrast conditions. In this graph you can see that the Samsung Galaxy S21 Ultra 5G (Snapdragon) takes almost a second to capture an image after the shutter has been pressed in daylight conditions. This is slower than the Exynos phone (which also doesn’t offer zero shutter lag and therefore isn’t the quickest either).
Autofocus comparison in daylight conditions (1000 lux) with 2 EV variation
Images captured in bokeh simulation mode show a slight lack of detail, and there is some variance between shots in terms of the strength of the blur applied to the background as well as of the shape of spotlights in the background.
Bokeh simulation, slight lack of detail
Ultra-wide, good exposure, wide dynamic range
The phone’s preview image is very similar to the Exynos version’s, with very good zoom smoothness and similar exposure to the final capture.
Tele-zoom image quality is fairly good across a wide range of zoom factors, but not quite on the same level as the best devices in this category, despite the Samsung’s dual-tele approach. In this close range studio shot (approximately 2x), the Snapdragon device does better at preserving fine detail than its Exynos counterpart. However, both Samsung phones cannot quite keep up with the Huawei Mate 40 Pro.
Samsung Galaxy S21 Ultra 5G (Exynos), crop: good detail, noise
Huawei Mate 40 Pro, medium-range tele
Huawei Mate 40 Pro, crop, good detail, low noise
At long range (approximately 12x), things change and the S21 Ultra 5G (Snapdragon) delivers the best result, making the Samsung a good option for those who take a lot of images at maximum tele settings.
Samsung Galaxy S21 Ultra 5G (Exynos), crop: lack of detail
Huawei Mate 40 Pro, long-range tele
Huawei Mate 40 Pro, crop: strong lack of detail
Like the Exynos version, we tested the Samsung Galaxy S21 Ultra 5G (Snapdragon) at 4K resolution and 60 frames per second. Video clips recorded on the Samsung in decent light conditions show good exposure but can be underexposed in low light. Dynamic range is good but not on the same level as the best in class.
Like for still images, video color rendering is pretty close to the Exynos version. However, our testers noticed some unnatural skin tones on the Snapdragon device, caused by a blueish white balance under indoor lighting, and some slight color casts when shooting outdoors in bright light.
On the other hand, the video autofocus performs better than on the Exynos device and is one of the camera’s strong points, thanks to good tracking and stability. The level of detail could be better for 4K recording, though, and is some way off the best devices in this class. Excessive ringing, most visible along high-contrast edges, also reduced the quality of texture tendering.
Noise remains a weakness, as it is visible in all light conditions. In low light, luminance noise becomes more intrusive and coarse, and some chroma noise is thrown into the mix as well.
On the plus side, the 60 fps recording means smooth motion and video stabilization is pretty efficient and counteracts camera motion well. Jello effects and frame shifts are better controlled than on the Exynos version as well, but a sharpness difference between frames can be quite noticeable, especially when recording under indoor conditions.
Samsung Galaxy S21 Ultra 5G (Snapdragon), outdoor video
Samsung Galaxy S21 Ultra 5G (Exynos), outdoor video
Apple iPhone 12 Pro Max, outdoor video
When we reviewed the Exynos version of the Galaxy S21 Ultra 5G, we said that given the top-end specifications of the device, we would have expected a higher camera score. The same is true for the Snapdragon version sold in North America, China and a few other markets.
The Samsung Galaxy S21 Ultra 5G (Snapdragon) is capable of capturing nice photos and videos in most situations. It is also very slightly better than its Exynos counterpart across pretty much all test areas, but still lags behind most direct competitors from rival brands.
Accurate exposure and fairly wide dynamic range
Good detail in outdoor and indoor shots and videos
Vivid color and nice skin tones
Good image quality with the ultra-wide camera
Fast autofocus with good tracking in videos
Effective video stabilization
High frame rate results in smooth videos.
Occasional autofocus failures, slow AF in HDR conditions
Noise in low light
Exposure instabilities in HDR conditions
Noise in videos at all light levels, including chroma and coarse luminance noise in low light
Occasional underexposure in video, especially in low light or backlit portrait scenes
White balance casts in videos, with occasional skin tone inaccuracies
In an attempt to prevent the “real-time spread of potentially harmful content” through its platform, photo sharing app Instagram has temporarily removed the “Recent” tab from hashtag pages ahead of the US Presidential election.
The change was announced through the Instagram Comms Twitter account (the irony…), where the company said that it will be making some changes that “make it harder for people to come across possible misinformation on Instagram […] as we near the U.S. elections.” First up: removing the “Recent” tab from hashtag pages.
Starting today, for people in the U.S. we will temporarily remove the “Recent” tab from hashtag pages. We’re doing this to reduce the real-time spread of potentially harmful content that could pop up around the election.
This is cited as a “temporary” change that went into effect yesterday, although Instagram does not specify when the Recent tab will be turned back on. Ostensibly this will happen after the election on November 3rd, but given the contentious political climate in the US and widespread fear that election results could be challenged in court, it’s possible the feature won’t be back for some time.
Facebook CEO and Instagram’s owner Mark Zuckerberg voiced this exact concern during an earnings call yesterday, saying:
I’m worried that with our nation so divided and election results potentially taking days or weeks to be finalized, there is a risk of civil unrest across the country. Given this, companies like ours need to go well beyond what we’ve done before.
Translation: this is probably one of many changes coming to the world’s most popular photo sharing app as we near election day.
Cuddle up with TikTok. What else is there to do? (Stan Horaczek /)
TikTok is a fun, silly place. To scroll through it is to take in a sensational amount of people dancing and lip-syncing. There’s a video of Cameron Diaz’s wine drinking challenge, a very polite kid named Grey, and a clip from Taylor Swift’s “Love Story” that you’ll hear way too much. There are even frogs. It’s addictive and ridiculous.
But the platform has been in the news lately for reasons relating to privacy and security, an issue that came to a head when President Trump issued an executive order on August 6 that outlaws transactions with TikTok’s parent company. (Read TikTok’s response here and an analysis of the head-scratching presidential decision here.) The drama stems from a fact that separates TikTok from other similar social media apps like Instagram: ByteDance, the company’s parent, is Chinese.
A path forward may be found through American ownership for the app, and on August 2 Microsoft announced they might buy TikTok. Conversations about the potential acquisition are already taking place, but we won’t know more until September 15.
What you’re hearing in the news may compel you to wonder—is the app safe to use? And if you are already using it, is there anything you should keep in mind?
The short answer is that it’s probably completely fine and harmless for most people to be on TikTok, as long as they keep in mind that, just as with other social media apps, it hoovers up data.
Platforms like TikTok and Instagram collect personal information from users with different levels of transparency, points out Shuman Ghosemajumder, the global head of artificial intelligence at F5, an internet infrastructure and security company. The first level is information that you’re clearly aware you’re sharing with them. This includes the email or account you used to sign up for the app, and of course the content you actively share on the platform.
“When you’re taking a video of yourself, and uploading that to TikTok, everybody knows that TikTok is taking that data and storing it on their servers, and performing various types of analysis on it,” Ghosemajumder says.
Finally, the third category of data collection is the nefarious, criminal kind which experts scrutinize apps to find. But Ghosemajumder says it’s extremely difficult for an app that operates at the scale of TikTok to be able to hide from the forensics researchers who would love to expose that kind of behavior. Has TikTok done that? Ghosemajumder says he knows of no evidence that it has.
There was a blip in that department, though—in June, it surfaced that TikTok had access to the clipboards of users running the next version of Apple’s operating system, iOS 14. That means that if you had recently copied and pasted anything sensitive, TikTok could have seen that. “The reason that TikTok claimed to be doing that was to detect users who were using the clipboard to spam comments,” Ghosemajumder says. Cutting and pasting is a common way to distribute spam. In fact, anytime you’ve noticed that the paste function isn’t working in a field on a website, that’s because the merchant is trying to fight fraud. It would make sense for TikTok to do this too, but since the issue was exposed, the platform stopped the practice.
Let common sense be your guide if you’re using an app like TikTok—and of course, start by not sharing anything you don’t want people to see. If you don’t want thousands of people to watch you dancing in your living room with your family, then don’t upload a video that shows just that. Ultimately, Ghosemajumder doesn’t see much daylight between TikTok and its competitors. “There’s no fundamental difference in using TikTok versus using [apps like] Facebook or Instagram,” Ghosemajumder reflects.
James Andrew Lewis, who directs the technology policy program at the Center for Strategic and International Studies, agrees. “Right now there’s no risk in using TikTok—it’s pretty harmless,” Lewis says. “The information on it is not valuable to an intelligence agency, the PII [personally identifying information] is nothing special, and there’s no evidence that TikTok has been used as a vehicle for delivering malicious code.”
As for the fact that the parent company is Chinese, Andrés Arrieta, director of consumer privacy engineering at the Electronic Frontier Foundation, points out that “a lot of the political discourse is more about xenophobia than actually privacy or security concerns.” Although, he adds, “if the Chinese government is your worry, then yes, it’s a worry.” For context, Human Rights Watch’s 2020 report on China has this to say about the country’s repressive tactics and global reach: “Government censorship now extends far beyond its borders; its mix of typically financial incentives and intimidation are manipulating discourse about China around the world.” An important reminder: The best way for anyone to communicate who does not want a government or company potentially snooping on what they say is through a platform that offers end-to-end encryption, such as WhatsApp, Signal, or iMessage.
“The concern here is, the Chinese could censor [content on TikTok]—right now they haven’t,” Lewis says. “Or they could put short propaganda videos on TikTok—right now they haven’t.”
Another common concern is that Beijing could lean on ByteDance to try to get American users’ data on the platform. But there is “no evidence that that has happened,” Lewis says. Interestingly, TikTok’s servers aren’t in China—they’re in Singapore and Virginia. And TikTok does not exist as an app within China itself.
“No one trusts China, and for good reason—China is engaged in a huge espionage campaign,” Lewis adds. And even though he says it’s wise for the United States government not to trust China (thus, the Pentagon doesn’t want members of the military to have the app, especially on their official work devices) individual users need not worry about using TikTok: “There is zero risk,” he says.
So, feel free to scroll through frog videos and people lip-syncing all you want. “If the Chinese can get intelligence advantage out of that,” Lewis reflects, “it would be an amazement to me.”
Update on August 13:
On Tuesday, The Wall Street Journalreported that TikTok had been collecting a unique numerical device identifier, known as the MAC address, with its Android users, but stopped doing so in November, 2019. It did so for over a year, the Journal stated.
The MAC addresses can be used not only for advertising purposes but also for fraud detection, Ghosemajumder notes.
“The current TikTok app does not collect MAC addresses,” a TikTok spokesperson said in an email to Popular Science. “We encourage our users to download the most current version of TikTok.”
Privacy & Cookies Policy
Necessary cookies are absolutely essential for the website to function properly. This category only includes cookies that ensures basic functionalities and security features of the website. These cookies do not store any personal information.
Any cookies that may not be particularly necessary for the website to function and is used specifically to collect user personal data via analytics, ads, other embedded contents are termed as non-necessary cookies. It is mandatory to procure user consent prior to running these cookies on your website.