DP Review

- Digital Photography Review
Sony a7 III dynamic range and high ISO improve over its predecessor

Sony recently announced the a7 III, a comparatively affordable full frame mirrorless camera that incorporates a host of advanced features derived from the a9 and a7R III. The combination of price point and feature set makes it attractive to both enthusiasts and pros, particularly those looking to get into full frame or perhaps even make the switch to mirrorless. While we've already shot quite a bit with it and offered our thoughts on the camera as a whole, we hadn't had a chance to take a deep dive into its image quality performance.

And we know many of you are wondering: what's the dynamic range like? The high ISO performance?

Let's take a look.

Low light (high ISO) performance

a7 III
ISO 25,600
ISO 25,600
a7 II
ISO 25,600

Low light performance has improved markedly over the a7 II, putting it more or less in-line with the a7R III (and therefore a9) when images are viewed at the same size (we've downsized the a7R III shot to 24MP). These are 100% crops here (if you're viewing on a smartphone or Retina / 4K display, see this footnote* below). Roll over the captions, or click on any of the images to view our full studio scene images for each camera.

This is a great result, but also comes as no surprise: noise performance is broadly determined by a combination of sensor size and technology, and we've recently seen some significant improvements to sensor technology made by Sony. In particular, the backside-illuminated (BSI) and dual gain architecture of most recent Sony sensors helps squeeze every last bit of performance out of these already low noise imaging chips. Furthermore, the original a7 and a7 II lagged in high ISO performance, often failing to surpass the best APS-C sensors.

Dynamic range vs. the a7R III

The a7 III more or less matches the base ISO dynamic range of the a7R III, when both are viewed at common size (we've normalized all our graphs to 8MP). That means both cameras will give you similar ability to make use of (brighten) shadows in Raw files if you want to show a wider dynamic range than shown with the default tone curve. And as long as you're shooting uncompressed Raw, performance is no different whether you're shooting Single or Continuous drive.

In numbers, that's 14.6 EV and 14.8 EV for the a7 III and a7R III, respectively, which falls within our margin of error. You might see a difference in extreme pushes or exposure adjustments, but it's not likely to be photographically relevant.

a7 III (orange) vs. a7R III (blue). There's a slight chance you might notice the 0.2 EV advantage of the a7R III at base ISO or the 0.3 EV advantage of the a7 III at higher ISOs, but we doubt it. As our test scene images show, the two cameras look very similar when viewed at the same output size.

Note the jump in dynamic range at ISO 640 for both cameras. That's essentially the camera's second 'base' ISO, where the second stage of the dual-gain architecture kicks in. At ISOs 640 and above, most recent Sony sensors use a higher gain mode that essentially amplifies the signal at the pixel-level to get it above the (already pretty low) noise floor.** In laymen's terms, that just means 'more picture, less noise', particularly in shadows – hence the increase in dynamic range.

Our analysis shows the a7 III to just edge out the a7R III at these higher ISOs, albeit only by about 0.3 EV (which happens to be right around our margin of error). You might see this in the deepest shadows – in fact, if you look very closely at the darkest patch in our ISO 25,600 rollover above, you can kind of see a tad bit less noise in the a7 III, but is that photographically relevant? Up to you.

... but it shows a marked improvement over its predecessor

While base ISO dynamic range remains the same as its predecessor, the dual-gain design brings a marked improvement at high ISO. Shadows at high ISO will be notably cleaner on the a7 III, and that's before you consider the better overall high ISO performance – even in brighter tones – likely due to either a more efficient sensor or lower upstream read noise.

Compared with the a7 II (green), the a7 III (orange) shows much better dynamic range (at least 1.6 EV) at higher ISOs. Also, whereas you can see noise reduction being applied to the a7 II's Raw at 25,600, it doesn't kick in until ISO 64,000 (beyond the graph) on the Mark III.

Compressed continuous drive performance

If you shoot compressed Raw, the camera drops to 12-bit sensor readout in continuous drive modes. This negatively impacts dynamic range, dropping 1.4 EV at base ISO and roughly 1 EV at ISO 640. Dynamic range catches up at higher ISOs, though never quite matches the performance of 14-bit readout. Even at ISO 6400, 12-bit files are roughly 0.4 EV behind - though this is unlikely to significantly impact your photography. The differences at lower ISOs and at ISO 640, on the other hand, you might notice in more extreme pushes.

a7 III Uncompressed (orange) vs. Compressed 12-bit (light orange) performance. We're not sure about the jumps at ISO 160 and 800, but for the most part there's a drop in dynamic range at lower ISOs that more or less evens out at the higher ISOs.

In Single drive mode, compressed Raw continues to use 14-bit sensor readout, so measured roughly the same dynamic range as Uncompressed (it dropped 0.1 EV, but that's within our margin of error).

And if you're confused about when the camera drops to 12-bit – which is the only time you'd see these drops in DR – the only combination that diverges from 14-bit is when you shoot compressed Raw in (any) continuous drive mode. All other combinations of Mechanical or Electronic shutter, drive mode or Raw type are 14-bit.

vs. a7R II

We threw this one in here because the a7 III and a7R II are currently being sold for roughly similar price (the latter is $400 more expensive), so we're aware of some discussion about choosing between the two. You're unlikely to notice our measured 0.2 EV higher base ISO dynamic range of the a7 III, but you might notice the 0.5 EV advantage at ISO 640. At higher ISOs the cameras even out.

Realistically though, there's not much difference between these cameras.

a7 III (orange) vs a7R II (red) dynamic range. You might notice the 0.5 EV advantage of the a7 III at ISO 640, but for the most part performance is similar.


Due to the dual-gain architecture, there are two 'ISO-invariant' ranges: ISO 100-500, and ISO 640-51,200. This means that if your midtone exposure demands ISO 400 but you're worried about clipping highlights, you're better off keeping your exposure settings the same but dialing the camera back to ISO 100 and then selectively brightening the Raw later. This affords you 2 EV extra highlight headroom, with no extra noise in shadows or midtones. If on the other hand your midtone exposure demands ISO 6400, you're better off keeping the same shutter speed and aperture and dialing the ISO down to ISO 640, affording you 3.3 EV extra highlight headroom at no noise cost.

Wait, does this mean I should shoot ISO 640 instead of 320?

No. Not necessarily.

If you have enough light to expose ISOs 200-500 correctly, you should use those ISOs. For example, say you can set a shutter speed and aperture to expose ISO 320 properly. You should not rather choose ISO 640 and shorten your exposure (to preserve highlights that the higher amplification of ISO 640 might clip). That would mean lower overall signal:noise ratio due to increased photon shot noise contribution, and would essentially have the same overall effect of shooting with a smaller (in this example: APS-C) sized sensor.

Recall that dynamic range is not everything, and generally the more light you collect, the better your image. Bill Claff's 'Photographic Dynamic Range' data for the a7R III, which uses a higher threshold for 'acceptable noise in shadows' and therefore considers total light captured more than our measurements, shows that ISO 100-400 outperform ISO 640 and higher. Dual-gain boosts low light performance, and shouldn't affect your exposure decisions any differently, other than perhaps biasing toward ISO 640 rather than 500 in low light.


We've summarized our results in numbers in the table below.

ISO 100 (24MP) ISO 100 (8MP) ISO 640 (24MP) ISO 640 (8MP)
a7 III 13.8 EV 14.6 EV 13.4 EV 14.2 EV
a7 III (compressed 12-bit) 12.4 EV 13.2 EV 12.3 EV 13.2 EV
a7 II 13.9 EV 14.7 EV 11.8 EV 12.6 EV
a7R III 14 EV 14.8 EV 13.1 EV 13.9 EV
a7R II 13.6 EV 14.4 EV 12.9 EV 13.7 EV
a9 12.6 EV 13.4 EV 12.4 EV 13.2 EV

So what's the take-away? The a7 III's image quality more or less matches what we've come to expect from modern, well-performing full-frame sensors. There's really not much difference between the a7 III, the a7R III, the a7R II, or the Nikon D850 for that matter.

But if you're coming from one of the original a7 cameras, you'll notice the dramatic increase in low light performance. The a7 III bests its predecessors both in dynamic range and general noise performance at higher ISOs, thanks to a number of sensor improvements (efficiency, BSI, dual-gain). Interestingly, the a7 III, which we'd imagine shares a similar sensor to the a9 minus the stacked design, offers roughly 1 EV more dynamic range than that camera at ISOs 100 and 640 (the cameras even out at the highest ISOs). General noise performance of the a9 - if you're not pushing your files - is similar though.

The a7 III's image quality more or less matches what we've come to expect from modern, well-performing full-frame sensors

The a7 III offers great image quality performance at an affordable price point. That said, it's not image quality that sets this camera apart from its contemporaries but, rather, its significant other capabilities like autofocus, silent shooting, video and a number of other things we'll be delving into in our full review.

* Retina & smartphone optimized 100% crops:

a7 III
ISO 25,600
ISO 25,600
a7 II
ISO 25,600

** Technically speaking, it's not exactly more amplification. Rather, the sensor switches to a different circuit within the pixel that has different capacitance at the floating diffusion node. This essentially generates a larger voltage swing (signal) per photoelectron captured, which means the signal - your picture - is less affected by the noise floor of the sensor and electronics.

Photo story of the week: Flowing under a solar storm
A night of stunning Northern Lights dancing above Haukland Beach, the Lofoten Islands, Arctic Norway, on a moonless evening.

The serene stream that flows from the surrounding mountains and pours into the Norwegian Sea curved into a beautiful shape, paralleling the curves of the Auroral display. Haukland is a very good location for shooting Aurora, since it has numerous interesting features (such as the mountain and the stream), and since any water left stationary frequently freezes over and supplies more variety and interest. It's also relatively shielded from artificial lights.

This image was taken in the winter of 2016 during my Lofoten workshop. I used a Sony A7R and a Samyang 14mm F2.8 with a Metabones adapter. The photograph was taken at F2.8, ISO 3200, and 8 sec exposure. The high ISO, wide aperture and long exposure were used to counter the darkness and produce a balanced exposure.

Erez Marom is a professional nature photographer, photography guide and traveler based in Israel. You can follow Erez's work on Instagram, Facebook and 500px, and subscribe to his mailing list for updates. Erez offers photo workshops worldwide.

AI-powered Google Lens visual search tool is now available on iOS devices

The AI-powered Google Lens feature uses visual recognition to provide information about whatever your smartphone's camera is pointed at. For example, it can identify landmarks, a type of flower, or provide information about a restaurant or other businesses you're photographing.

Google first showed of this feature at the I/O 2017 event, then integrated it into the company's Pixel phones, and later made available for all Android devices. Now, the final step of the natural Google Lens evolution is complete: the company has announced that Google Lens is coming to Apple's iOS operation system:

iOS users should see a preview of Google Lens appear in the latest version of the Google Photos app over the next week. So, look out for the update and, if you haven't got the Google Photos app already, you can download and install it from the iOS App Store.

Adobe posts record revenue yet again, earning $2.08 billion in Q1 of 2018
Photo by Kevin

It's starting to feel old hat by now: another quarter, another record-breaking earnings report coming out of Adobe. No matter how much people—present company certainly not excluded—gripe about Adobe's move to the Creative Cloud subscription model, the software company is absolutely raking in the dough as a result.

The last record-breaking revenue report we shared out of Adobe came in Q3 of 2017, when Adobe announced that it had earned $1.84 billion that quarter. Now, in Q1 of 2018, the company is staring at that figure in the rearview mirror.

This quarter, Adobe is posting record quarterly revenue of $2.08 billion, $1.23 billion of which came straight from Creative Cloud in the Digital Media Segment. That $2.08B figure represents a jump of 24 percent year-over-year, and contributes to the 43 percent growth in YoY operating income and 64 percent growth in YoY net income that Adobe also revealed.

You can dive into the nitty gritty details in the release below, and see the full number-by-number breakdown in this PDF.

Press Release

Adobe Achieves Record Revenue

Creative ARR Exceeds $5 Billion in Q1 FY2018

Thursday, March 15, 2018 4:05 pm EDT | San Jose, California – Adobe (Nasdaq:ADBE) today reported strong financial results for its first quarter fiscal year 2018 ended March 2, 2018.

Financial Highlights

  • Adobe achieved record quarterly revenue of $2.08 billion in its first quarter of fiscal year 2018, which represents 24 percent year-over-year revenue growth.
  • Diluted earnings per share was $1.17 on a GAAP-basis, and $1.55 on a non-GAAP basis.
  • Digital Media segment revenue was $1.46 billion, with Creative revenue growing to $1.23 billion and Document Cloud achieving revenue of $231 million.
  • Digital Media Annualized Recurring Revenue (“ARR”) grew to $5.72 billion exiting the quarter, a quarter-over-quarter increase of $336 million. Creative ARR grew to $5.07 billion, and Document Cloud ARR grew to $647 million.
  • Digital Experience segment revenue was $554 million, which represents 16 percent year-over-year growth.
  • Operating income grew 50 percent and net income grew 46 percent year-over-year on a GAAP-basis; operating income grew 43 percent and net income grew 64 percent year-over-year on a non-GAAP basis.
  • Cash flow from operations was $990 million, and deferred revenue grew 25 percent year-over-year to approximately $2.57 billion.
  • Adobe repurchased approximately 1.6 million shares during the quarter, returning $301 million of cash to stockholders.

A reconciliation between GAAP and non-GAAP results is provided at the end of this press release and on Adobe’s website.

Executive Quotes

“Adobe’s outstanding growth is driven by enabling our customers to be more creative, work smarter and transform their businesses through our relentless focus on delivering innovation and intelligence across our solutions,” said Shantanu Narayen, president and CEO, Adobe.

“Our leadership in the large addressable markets we created, combined with Adobe’s leveraged operating model, contributed to another record quarter in Q1," said Mark Garrett, executive vice president and CFO, Adobe.

Adobe to Webcast Earnings Conference Call

Adobe will webcast its first quarter fiscal year 2018 earnings conference call today at 2:00 p.m. Pacific Time from its investor relations website: www.adobe.com/ADBE. Earnings documents, including Adobe management’s prepared conference call remarks with slides, financial targets and an investor datasheet are posted to Adobe’s investor relations website in advance of the conference call for reference. A reconciliation between GAAP and non-GAAP earnings results and financial targets is also provided on the website.

Forward-Looking Statements Disclosure

This press release contains forward-looking statements, including those related to customer success, product innovation, business momentum, our addressable market, revenue, annualized recurring revenue, non-operating other expense, tax rate on a GAAP and non-GAAP basis, earnings per share on a GAAP and non-GAAP basis, and share count, all of which involve risks and uncertainties that could cause actual results to differ materially. Factors that might cause or contribute to such differences include, but are not limited to: failure to develop, acquire, market and offer products and services that meet customer requirements, failure to compete effectively, introduction of new technology, complex sales cycles, risks related to the timing of revenue recognition from our subscription offerings, fluctuations in subscription renewal rates, potential interruptions or delays in hosted services provided by us or third parties, risks associated with cyber-attacks, information security and privacy, failure to realize the anticipated benefits of past or future acquisitions, changes in accounting principles and tax regulations, uncertainty in the financial markets and economic conditions in the countries where we operate, and other various risks associated with being a multinational corporation. For a discussion of these and other risks and uncertainties, please refer to Adobe’s Annual Report on Form 10-K for our fiscal year 2017 ended Dec. 1, 2017, and Adobe's Quarterly Reports on Form 10-Q issued in fiscal year 2018.

The financial information set forth in this press release reflects estimates based on information available at this time. These amounts could differ from actual reported amounts stated in Adobe’s Quarterly Report on Form 10-Q for our quarter ended March 2, 2018, which Adobe expects to file in March 2018.

Adobe assumes no obligation to, and does not currently intend to, update these forward-looking statements.

The New York Times is looking to hire a Photo Director
The New York Times Building by wsifrancis | CC BY-NC-ND 2.0

The New York Times has posted a job opening for the position of Photo Director. If you're looking for a high-profile job in the world of photojournalism, and you live in (or don't mind moving to) New York City, you could do a lot worse than working for The Gray Lady.

The opening was listed seven days ago, and it goes to great lengths to emphasize the importance of photography to the Times' mission. "Photography is a central part of our identity," reads the posting. "It’s how we bear witness to events that matter, and our Photo department is one of the treasures of our newsroom."

As for the job of Photo Director itself, the posting reads:

Now we’re looking for someone to lead this talented and diverse team and to become part of the visual leadership of the organization. We want to continue integrating photography and other forms of visual journalism into the fabric of our report — as closely as our words.

This role is one of the most important and high-profile jobs in visual journalism, and we’re seeking candidates with a rare combination of journalistic experience, organizational expertise and extraordinary visual talent.

Some of the listed qualifications include:

  • Daily leadership of a large staff of photo editors and photographers who work across the globe, covering all subjects.

  • Candidates should be able to maintain high journalistic standards and sustain a level of excellence that makes photography a core component of The Times’s identity.

  • Sophisticated news judgment and a compelling vision for how The Times can produce world-class journalism and innovative storytelling. We’re looking for a strong digital sensibility, including the ability to recognize emerging techniques and platforms and a clear understanding of how to define a modern photo desk.

  • Strong grasp of feature and portrait photography and the ability to improvise visual solutions for news coverage that may not be obviously visual.

  • Sharp eye for talent and ability to recruit a diverse, first-rate team of photo editors and photographers.

If you think you have what it takes to be the new Photo Director at the New York Times, click here to read the full job opening and apply.

Samsung wants to dethrone Sony, reach #1 in the global image sensor market

According to a new report out of South Korea, Samsung is increasing production of its ISOCELL image sensors at its Hwasung, South Korea location in a bid to clinch the #1 spot in image sensors worldwide.

However, this feat is definitely easier said than done. Current market leader Sony has a comfortable advantage over its South Korean rival, and certainly won't go down without a fight. In the lucrative smartphone segment alone, Sony currently has a 46 percent market share versus Samsung's much smaller 19 percent.

That said, technologically at least, Samsung is well-placed to take on the challenge. Its latest Galaxy Note 8 and Galaxy S9/S9 Plus devices all come with innovative imaging technologies and offer excellent camera performance built on Samsung's own sensor technology.

The company rebranded its image sensor range as Isocell in June 2017. Since then, Samsung has not only expanded its high-end sensor offerings, it also designed low-cost image sensor modules that are easy to implement into devices by other manufacturers. Several of those, for example Xiaomi and Meizu, are already using Samsung image sensors.

However, market leadership cannot be achieved with smartphones alone. Samsung is also planning to grow in the automotive space where CMOS sensors are increasingly used in the autonomous vehicle space and for other applications. In this segment, Samsung will face stiff competition from the likes of Bosch and Continental.

Whoever ends up ruling the image sensor market, a large company like Samsung challenging Sony's quasi-monopoly for image sensors can only be good news for consumers.

Moar Megapixels! Pixel peeping a 709MP drum scan of 8x10 slide film

Large format wilderness photographer Ben Horne recently embarked on a little experiment with some help from his friend, Michael Strickland. Horne shoots large format 8x10 slide film, and Strickland has a drum scanner that can scan that film at insanely high resolution. How high? Using a little bit of trickery, Strickland was able to provide Horne with a 709.6-megapixel file to pixel peep in this video.

Take that, 100MP medium format sensors!

To give you an idea of just how high resolution this file is, printed at 300ppi, the resulting print would measure 79.3 x 99.4 inches. As we mentioned, this took a bit of 'trickery'—namely: Strickland actually had to drum scan the print twice. He first scanned the top half, then the bottom half, and then merged the two scans together in post.

In the video, Horne zooms in to 100% and makes his way around the file. He explains how he shot the image, what sacrifices he had to make regarding sharpness in the closest foreground and furthest background, and shows off just how sharp this thing is in the parts of the image he's most concerned with.

Check out the full explanation for yourself up top, and then head over to Horne's YouTube channel for more videos like this one.

The Android-only LyfieEye200 is 'the world's smallest VR/AR camera'

eCapture Technologies has launched a new version of the LyfieEye mobile camera on crowdfunding website Indiegogo. Called LyfieEye200, this model is being dubbed "the world's smallest 360° VR/AR camera," and offers a bunch of neat AR/VR features for Android users who want to get more mileage out of their smartphone photography adventures.

The LyfieEye200 was designed for Android smartphones, and adds 1440p support in addition to the original model's 1080p resolution. The removable camera plugs directly into a smartphone's USB-C port, where a pair of greater-than-180° FOV fisheye lenses work together to enable both 360° image/video capture and 360° livestreaming.

To make the magic happen, the camera works in conjunction with the LyfieView200 Android companion app on devices running Android 5.0 or newer. And if you want even more creative possibilities, eCapture offers both the LyfieStroll and LyfieRoam apps for creating simple VR and AR content, respectively. Finally, the camera is also compatible with PCs running Windows 7 or higher, but it does not support iOS.

The LyfieEye200 is available now on Indiegogo, where backers can 'reserve' theirs by pledging at least $90 USD. Shipments to backers are expected to start in June, assuming the campaign reaches full funding and doesn't pull a KitSentry.

Syrp launches Magic Carpet Pro slider with 'infinitely extendable' track

Filmmakers who need really long sliders are the target of the new Magic Carpet Pro, the latest in a series of sliders announced by accessories manufacturer Syrp. The key selling point of the Magic Carpet Pro is its 'limiteless' range, a feature made possible by a new track lever design that allow users to lengthen their slide by simply clicking additional lengths of track into place—no tools required.

But that's not all the Magic Carpet Pro can do.

A newly designed flywheel sits inside the camera carriage, eliminating the need for belts and pulleys while creating smooth motion during manually controlled camera movements. And as with all of Syrp's other slider systems, automated motion control can be added via a Genie system; in fact, the Genie ll is able to slot directly into the new quick release holder in the carriage itself.

The tracks are made from aluminum, and are expected to manage a payload of up to 70lbs when no extension tracks are in use, and up to 50lb when they are. Here's a quick video intro from Syrp itself:

The Syrp Magic Carpet Pro can be pre-ordered now for May/June delivery, with kit prices ranging from $990 to $1470. Three kits will be available, including one with a two foot ‘Short Track’, one with a 'Medium Track' of three feet, and a third that includes both the short and the medium tracks to create a combined five foot slider.

For more information, visit the Syrp website.

Sony interview: 'If cameras are going to develop, manufacturers have to develop mirrorless technologies'
Kenji Tanaka, Senior General Manager of Sony's Digital Imaging Business Group, pictured in Yokohama for the 2018 CP+ show.

At the recent CP+ show in Yokohama, we sat down with executives from several major camera and lens manufacturers. Among them was Kenji Tanaka, of Sony. In our interview we discussed the new a7 III, as well as Sony's plans to attract more professional users, without ignoring entry-level and APS-C customers.

The following interview has been edited slightly for clarity and flow.

What is your target customer for the new Alpha a7 III?

We describe it as a basic model but maybe our definition is different [to other manufacturers]. What we mean is that any customer can use this model. Many professionals could use the a7 III, I think. I hope that many kinds of customer will be happy with this model, so we’re not strictly defining a target customer for the a7R III.

The new Sony Alpha a7 III is ostensibly an entry-level model in Sony's a7-series lineup but despite its relative affordability, it's packed with features.

Sony had a very busy year last year – what are your priorities going into 2018 and beyond?

We’re displaying the 400mm F2.8 [at CP+] – of course many articles are written about the a7 III, but as well as the camera bodies, the lenses are very important. Especially lenses like the 400mmm F2.8 – sports photographers are a new category for us.

One of the most important lenses for sports photographers is the 400mm F2.8

Last year we launched the a9 and some sports photographers are already using the a9, for example at the Olympics, but one of the most important lenses for sports photographers is the 400mm F2.8. The weight is very light. Usually sports photographers use monopods because the lenses are very heavy, but the weight of our 400mm F2.8 is very light, and you can use it handheld, which makes it easy to create different kinds of photographs. We already announced the development of this lens, and the launch is scheduled for this summer.

How important is feedback from sports photographers?

It’s very important. Not only when it comes to quality, but also durability. The winter Olympics for example, with the low temperatures, whether a product works in those tough conditions is very important. Whether or not we will launch a new product, the proof of concept is very important.

For a product aimed at a hobbyist, maybe it's less important but for the 400mm F2.8 we’re really dedicated to create a ‘monster’ lens.

Sony was showing a prototype of its forthcoming 400mm F2.8 at CP+, which Mr. Tanaka sees as an essential weapon in Sony's arsenal of lenses if the company is going to attract professional sports photographers to the brand.

There have been questions about the weather sealing of a7-series and a9 cameras. Is this something you want to address?

In really bad conditions, in really heavy rain, will photographers keep on taking pictures [for long periods of time?] I don’t think so. In those conditions, most photographers will use some kind of rain cover. But of course durability is very important. Photographers should be able to shoot [in poor weather]. We have an internal ‘weather test’ and for each kind of customer we will aim to produce products with adequate durability.

For a professional camera, the requirement for durability is higher

Is that something that your professional users are asking for?

Yes. But we need a balance between durability, and size and weight. For a professional camera, the requirement for durability is higher, but for hobbyist kinds of camera, the priority is smaller size and lower weight.

Tamron and Sigma are now creating lenses for Sony FE - is this a good thing for Sony?

Yes, it’s a good thing. The E-mount is an open standard – anyone can create a lens for the E-mount system. Of course there are criteria for compatibility, but because we think that the E-mount is a good technology, we think that the open format is good for the market and good for customers.

Tamron's first lens for full-frame Sony mirrorless cameras will be the upcoming 28-75mm F2.8 zoom. Sigma is planning its own range of native E-Mount primes and zooms, too.

How long will it be before Canon and Nikon join Sony and mirrorless full-frame becomes the norm?

This is just my personal opinion, but I think that maybe by next year’s CP+ you’ll see full-frame mirrorless cameras from Canon and Nikon. I think [by then] they will be participating in this market.

If cameras are going to develop, manufacturers have to develop mirrorless technologies

Just look at our technologies, like eye focus. All of those are made possible because of data from image sensors. In DSLRs, the data comes from separate sensors. The main imaging sensor is blacked out, 90% of the time by the mirror. The sensor is turned off. But the imaging sensor is very important. So if cameras are going to develop, and be able to capture the moment [more effectively], manufacturers have to develop mirrorless technologies. So within one year, I think.

Do you think we’ll see mostly mirrorless cameras at the 2020 Olympic Games in Tokyo?


How many of those cameras do you want to be made by Sony?

Many! But the professional market is very conservative, so we’re taking it step by step. We saw some photographers using the Alpha 9 at this year’s winter Olympics but of course the majority was Canon and Nikon. But the number of Sony photographers is increasing.

The Sony Alpha a9's innovative wide-coverage autofocus system makes it a powerful tool in the hands of an experienced sports photographer.

A lot of our readers want more Sony APS-C lenses…

We know that some people think we’ve neglected the APS-C market, but it’s just an issue of prioritization. A couple of years ago we introduced the a6500. Then the next year we introduced the a9, and the a7R III. But we think that the APS-C market, and APS-C customers are both very important, because the majority of the market is APS-C, and we’re developing many kinds of APS-C products, so please be patient – we will never ignore APS-C.

Some of your competitors keep flagship products up to date over time with firmware updates. This seems like the opposite strategy to Sony. Is this something that might change?

We’re still in the early stages of challenging the market with our products, and the new model cycle is relatively rapid, compared to our competitors. But the next step is to increase our market share. And if we want to reach new customers, we need [to make] new types of cameras.

We’re still in the early stages of challenging the market with our products, and the new model cycle is relatively rapid

Sony makes a lot of key devices, for example image sensors and processors. I’m originally an engineer. Engineers always want to provide the latest sensor, the latest processor, and so on. Maybe this is one of the reasons our product release cycle is faster than some of our competitors. [But] user-upgradable software is very important. Our new model cycle is speedy, however I think that firmware updates are something we should offer.

Editor's note:

Our conversation with Mr. Tanaka was candid and interesting, coming in the middle of a very busy period for Sony. The company has released a lot of high-end products over the past 18 months, and shows no signs of slowing down. We don't know how far out the new 400mm F2.8 sports lens is, but given recent sightings of at least one working prototype 'in the wild' at the winter Olympics, it could be pretty imminent.

Mr. Tanaka knows that Sony won't have the full-frame mirrorless field to itself for much longer, and welcomes the competition

Mostly I came away from this interview with the strong sense that Sony isn't planning on resting on its laurels. Mr. Tanaka knows that his company won't have the full-frame mirrorless field to itself for much longer, and welcomes the inevitable competition from established DSLR manufacturers like Canon and Nikon, as well as third-party lens manufacturers like Sigma and Tamron. As he correctly points out, some of the most useful features to emerge in the photography market in recent years could only have been possible thanks to mirrorless technologies, and Sony deserves enormous credit for developing and perfecting many of these technologies faster than any other manufacturer.

Sony will not ignore either APS-C users, or entry-level full-frame customers

It was very reassuring to hear Mr. Tanaka stress the importance of durability, as well as technology in Sony's high-end cameras. Concerns have been raised about the ability of some of its products to withstand use in wet conditions, but clearly this is something that the company is mindful of – especially in cameras and lenses designed for professional use.

That's not to say that Sony is focused entirely on breaking into the professional market. Mr. Tanaka was at pains to reassure us that Sony will not ignore either APS-C users, or entry-level full-frame customers. The new a7 III is proof of the latter point – a 'basic' model in Sony's terminology, but one that I suspect will satisfy the needs of many enthusiasts and even professionals.

Previous Sony interviews:

CP+ 2017 - Sony interview: 'We need to offer new imaging experiences'

'We want to make lenses that can be used forever': Sony engineer discusses G Master lenses

Interview: Kimio Maki of Sony - 'the customer’s voice is the most important data for me'

Large format macro: a crazy wet plate photography experiment

Austrian wet plate photographer Markus Hofstaetter is back with another crazy large format photography experiment. This time, he decided to find a way to shoot macro photos on a large format wet plate camera. To do this, he actually had to stack two wet plate cameras front to back, bellows fully extended.

Markus documented the whole experiment on his blog, and shares a behind-the-scenes 'Making Of' look at the shoot in the video above. His subject was a little snowdrop from his garden, with a simple tin-foil background for some pretty bokeh. But getting any sort of magnification with a large format camera is no easy feat. He needed a lot of distance between his film plane and the little flower.

That's why he decided to 'connect' two large format wet plate cameras together, giving him enough extension to magnify the flower onto an 18x24cm plate.

Here's a diagram that shows the difference between your standard "full-frame" size, a 10x12cm plate, and an 18x24cm plate (left) and that same diagram overlaid on the final plate:

The next problem he faced was getting enough light. The farther the plate is from the subject, the more light he needs—the plate has an ISO value of about 0.5—and he was pretty far away from his subject. The trick to solving this problem, says Markus, is using fresh chemicals and a LOT of artificial light.

"Freshly mixed chemicals are more sensitive to light," he tells DPReview. "If I had used older chemicals, I’m not sure if this macro shot would have been possible." Add to that two flashes of 7,000W of light, and you've got JUST enough exposure to make this work.

Mix all of this together and here's what you get. Scroll to the very end to see the final image:

Definitely check out the whole Making Of video at the top if you want to see how this shot came together. Markus goes into more detail, revealing interesting tidbits about the lens he used, his lighting setup, and lets you tag along for the entire developing process as well.

And if you like his work, don't forget to visit his website, check out his blog, and give him a follow on Facebook and Instagram. His quirky experiments—several of which we've featured on DPReview—might just get your creative juices flowing, too.

Fujifilm X-H1 Review

The Fujifilm X-H1 is the company's range-topping APS-C camera and its most video-capable camera to date. It's based around the same 24MP sensor as the X-T2 but adds in-body image stabilization as well as a more comprehensive set of video options.

The X-H1 looks like a fractionally larger X-T2 but with the sloped viewfinder 'prism' and top-panel LCD that hint at the styling of the GFX 50S. Fujifilm has also clearly been listening to critics of the X-T series and have made the camera's grip and buttons significantly larger, particularly the AE-L and newly-added AF-On buttons.

Key specifications

  • 24MP X-Trans APS-C sensor
  • 5-axis in-body image stabilization (rated at 5EV)*
  • 3.69M-dot OLED viewfinder
  • Touch sensitive rear LCD with two-axis tilt
  • DCI and UHD 4K capture at up to 200 Mbps
  • Slow motion 1080 (from 120 and 100 fps)
  • Internal F-Log capture
  • 24-bit audio capture
  • Eterna/Cinema Film Simulation mode
  • Timecode
  • Reduced blackout in continuous shooting
  • Twin UHS-II-compatible card slots
  • Anti-flicker shooting mode
  • Wi-Fi with Bluetooth for constant connection

The company says it's made further improvements to its AF system and says the new camera will be able to focus in lower light and with smaller apertures.

Despite being based around the same sensor and processor, the X-H1 promises significantly improved video performance, with the range of shooting options extended to include DCI as well as UHD 4K shooting, bitrates up to 200 Mbps and the ability to record F-Log footage internally.

Other additions include the movie style 'Eterna' Film Simulation and an anti-flicker option for shooting under artificial lights.

Interestingly, although rated at 5EV, Fujifilm says the stabilization can hit 5.5EV of effectiveness if paired with non-IS lenses. The explanation for this is that the unstabilized lenses tend to be primes and are generally relatively wide focal lengths, both of which mean they're more likely to project a larger image circle than the sensor requires. This gives the sensor more room to move around, providing greater stabilization.

Enhanced video

The X-T2 is already a very credible video performer: offering good levels of detail capture and Log output over HDMI if needed. The X-H1 takes this a step further. In addition to being able to shoot UHD 4K at up to 30p it can also shoot the wider aspect ratio DCI 4K format at 23.98 and 24p. Enhanced compression options allow capture at up to 200 Mbps and it can also capture F-Log footage internally.

Like the X-T2, the H1 uses a 1.17x crop region of its sensor to capture its UHD and DCI 4K video. This means using roughly 1.4x more pixels than necessary, in each dimension, to produce its UHD footage. This oversampling leads to higher levels of detail capture than would be possible by simply using a 3840 x 2160 region. If the X-T2 is anything to go by, it should look good and have pretty well-controlled rolling shutter.

It seems most of the camera's additional size relates to the addition of the stabilization unit, but thermal management has also been improved, allowing the camera to shoot 4K for 15 minutes, rather than the 10 of the X-T2. However, as with the X-T2, there's an optional battery grip that lets the camera cycle between drawing power from each of three batteries. Presumably this avoids too much heat building up in the same place, since it extends the camera's 4K shooting duration out to the traditional 29 minutes, 59 seconds stipulated by import duty regulations.

On top of this comes the ability for the camera to retain a raft of settings separately for stills and video. This means you don't have to significantly reconfigure the camera every time you switch from stills to video shooting or back.

Parameters treated independently for movie shooting
  • Film Simulation
  • Dynamic Range mode
  • White Balance
  • Highlight Tone
  • Shadow Tone
  • Color (saturation)
  • Sharpness (sharpening)
  • Noise reduction
  • Peripheral light correction (vignetting )
  • Focus area
  • Focus mode
  • AF-C Custom Settings
  • Pre-AF
  • Face/Eye Detection
  • MF Assist
  • Focus Check

The obvious things that can't be set independently for stills and movie shooting are the exposure settings, since these are primarily defined by dedicated control dials. If you plan to swap back and forth between stills and video shooting, the camera's new 'Movie Silent Control' mode is one way around this.

Movie Silent Control disables the aperture ring, shutter speed dial and ISO dial, passing control to a touchscreen, joystick and four-way controller-based interface. This means discrete stills and video settings can be maintained, since the dedicated control points no longer have any affect in video mode.

However you choose to control exposure in movie mode, you'll quickly find that the X-H1 offers shutter speeds equivalent to 360, 180 and 90 degree shutter angles for 24, 30 and 60p video capture, with the options for 1/24th, 1/48th, 1/96th, 120th and 1/240th becoming available.

Like its sibling, the X-H1 offers a series of focus peaking options (color and intensity) but no zebra warnings for setting exposure, beyond the 'Live View Highlight Warning' option that indicates an unspecified and unspecifiable brightness.

The X-H1 also brings Fujifilm's DR modes to movie capture for the first time, allowing you to capture more highlight information, if you can tolerate higher ISO settings. Meanwhile the 'Eterna/Cinema' Film simulation is designed to give 'soft,' low-saturation footage with low contrast but distinct shadows. Fujifilm says it can be used as an end-point in itself or to give yourself a degree of latitude for color grading.

Users of Fujifilm's MK lenses (launched in X-mount alongside the X-H1) will appreciate the ability to view aperture as T-stops, rather than F-numbers. It's unclear at this point whether this option will be available with adapted and third-party lenses identified this way.

Dynamic Range Priority

Fujifilm was one of the first brands to exploit the ISO-invariant properties of the sensors it uses through its Dynamic Range modes (The DR modes offer multiple ways of delivering ISO settings using different amounts of hardware amplification to capture additional highlight information).

The X-H1 takes this further with a 'Dynamic Range Priority' mode. This uses the existing DR modes in combination with the camera's ability to adjust the Highlight and Shadow aspects of its tone curves. There are four settings: Weak, Strong, Auto and Off. The 'Weak' setting is DR200% mode with highlights and shadows softened by 1 step (since it's based on DR200%, is only available from ISO 400 upwards), while 'Strong' is DR400% with Highlights and Shadows set to -2. Strong is only available from ISO 800 or higher.

New shutter mechanism

Along with in-body stabilization, the X-H1 gains a new, quieter shutter mechanism. In addition to being quieter, it also allows the camera to offer Electronic First Curtain (EFC) shutter mode. In this mode the sensor being activated starts the exposure but a physical shutter is still used to end it, so that you significantly reduce the risk of shutter shock without increasing the risk of rolling shutter.

Various combinations of EFC, mechanical and fully electronic shutter are available, to allow the use of each mode for the shutter speeds where it gives its greatest advantage.

Compared with its peers

The X-H1 is the latest high-end crop sensor camera to offer both stills and video shooting but each one provides a different set of features:

Fujifilm X-H1 Fujifilm X-T2 Sony a6500 Panasonic GH5
(body only)
$1900 $1600 $1400 $2000
Pixel count 24MP 24MP 24MP 20MP
Sensor size APS-C APS-C APS-C Four Thirds
Image Stablization 5-axis, 5.5EV Lens only 5-axis, 5EV 5-axis, 5EV
Maximum shooting rate 14 fps with e-shutter, 8 fps mechanical (11 with grip)

14 fps with e-shutter, 8 fps mechanical (11 with grip)

11 fps 9 fps (11 with S-AF)
AF Joystick? 8-way 8-way No 4-way
Touchscreen Yes No Yes Yes
Screen articulation Two-axis tilt Two-axis tilt Tilt Fully articulated
EVF 3.69M dots 2.36M dots 2.36M dots 3.69M dots
Viewfinder magnification 0.75x 0.77x 0.70x 0.76x
Video Bit depth 8 8 8 10
Max bitrate
200 100 100 400 (150 in 8-bit mode
Mic / Headphone sockets? Yes / On VPB-XH1 accessory grip Yes / On VPB-XT2 accessory grip Yes / No Yes / Yes
Log capture? Yes HDMI out only Yes HLG (V-Log L Via paid upgrade)
HDMI Micro Micro Micro Full size
USB 3.0 Micro Type B 3.0 Micro Type B 2.0 Micro Type B 3.1 Type C
Shots per charge (CIPA rating) 310 340 310 410
Weight (with card and battery) 673g 507g 453g 725g

Pricing and availability

The X-H1 is available with an MSRP of $1899 body only and $2199 bundled with the VPB-XH1 vertical grip.

Review Publication History
February 15 Introduction, body and handling, first impressions and samples
March 15 In Use..., Autofocus, Image Stabilization, Image Quality,
Dynamic Range, Video and Conclusion added

*Fujifilm says the camera will give up to 5.5EV of stabilization when paired with non-stabilized XF lenses. As with all CIPA ratings, the performance is likely to be lower with very wide or long lenses.

Behind the scenes: Shooting Lamborghinis in the snow

Every year, Lamborghini puts on something called the "Winter Accademia." It's a week-long event wherein professional racing drivers teach clients how to drive the latest Lamborghini supercars on snow and ice—in short: one of the coolest events a motorsport photographer could cover. And this year, motorsport photojournalist Jamey Price was there to document the whole thing.

I stumbled across Jamey's story earlier this week on Facebook, where he shared the Behind the Scenes video above. But rather than just post the video without further context, I reached out and asked Jamey for some details about the Winter Accademia and what it's like to shoot it.

What he ended up sending us is worth publishing in full, so read on to find out what it's like to shoot the Lamborghini Winter Accademia 2018, and then scroll through the gallery at the bottom to see some Jamey's favorite shots from the event.

Shooting the Lamborghini Winter Accademia 2018

Every winter, Lamborghini North America organizes a driving event called Winter Accademia for existing and prospective clients to drive the latest Lamborghini models on snow and ice. Most people see a Lamborghini as an exotic car meant for driving in “normal” conditions, but in reality, with a little help from Pirelli winter tires with small metal studs, they are more than capable as performance cars on the ice too.

I’ve been fortunate to work with Lamborghini since 2013 covering races, as well as private driving events, concours and auto shows around the United States. This is simply one of the many events that is organized for the brand during the year. But as you can guess, it's one of my favorites.

Lamborghini descended on Hotel Sacacombie on the lake that bears the same name in Quebec for a week of driving in late February and early March. The cold temperatures during the winter freeze the lake with over 30 inches of ice, which is more than enough to drive on. The track is simply a plowed section of snow on the frozen surface of the lake where we can run a drift circle, figure eights and the full track where a team of 7 professional racing drivers can teach clients the finer points of drifting a Lamborghini.

As a photographer, it doesn’t get much better than this. Yeah, it’s long days in the cold which is brutal on both the gear and myself, but it’s amazing photography chasing these high performance cars in an environment that is very contained, safe and, most of all, fun.

Especially with the professional drivers, I can put myself in places I would never consider standing under normal conditions at a track. Because everything happens in slow motion on the ice, you can get pretty close and capture lots of photos of the cars completely sideways and throwing snow and ice into the sky.

Credits: Lamborghini, Lamborghini Squadra Corse, IRIS Worldwide and the amazing team of professional drivers I'm fortunate to work with every week.

A big thank you to Jamey for sharing his experience and photos with us. To see more of his work or follow along as he shoots some of the most adrenaline (and gasoline) fueled events around the world, visit his website or give him a follow on Instagram and Facebook.

Google explains the tech behind the Pixel 2's Motion Photos feature

Apple was the first mobile manufacturer to popularize still/video hybrid files with its Live Photos that were introduced on the iPhone 6s. Google then launched the Motion Stills app to improve and stabilize Apple's Live Photos, and ported the system to the Android world soon after.

For the new Motion Photos feature on its latest Pixel 2 devices Google built on Motion Stills, improving the technology by using advanced stabilization that combines the devices' soft and hardware capabilities. As before, Motion Photos captures a full-res JPEG with an embedded 3 second video clip every time you hit the shutter.

However, on the Pixel 2, the video clip also contains motion metadata that is derived from the gyroscope and optical image stabilization sensors.

This data is used to optimize trimming and stabilization of the motion photo and, combined with software based visual tracking, the new approach approach aligns the background more precisely than we've seen in the previous Motion Stills system (which was purely software-based). As before, the final results can be shared with friends or on the web as video files or GIFs.

If you are interested in more technical details of the Motion Photos feature, head over to the Google Research Blog. A gallery of Motion Photo files is available here.

Novoflex's new TrioPod PRO75 modular tripod can handle 'immense load'

Novoflex has announced an expansion of its TrioPod tripod product line with the launch of the TrioPod PRO75 model. This is a high-capacity advanced spider tripod, according to Novoflex, which claims that the PRO75 model can be used "for even the heaviest photo and video equipment."

Novoflex explains that its new PRO75 tripod is constructed with 8-layered carbon fiber legs alongside reinforced components. Buyers have a choice between three or four-section carbon fiber legs, as well as a pair of mini legs. The PRO75 model also supports existing TrioPod legs.

The PRO base by itself will run you over $880 USD.

The TrioPod PRO75 model can support a maximum load of 65kg / 143lbs and has a total of seven leg lock positions at 23-, 43-, 45-, 65-, 87-, 155-, and 180-degrees. The tripod is made with a high level of versatility in mind, offering support for most 75mm accessories and both 1/4" and 3/8" connections, a removable base plate for using a leveling ball, and a self-locking geared center column.

Of course, all of these capabilities are going to cost you a pretty penny. The TrioPod PRO75 base alone, without legs, costs a whopping $885.28. Add 3-section carbon fibre legs (plus 2 mini legs) to the kit and you're looking at $1,476.64, while a set of 4-section legs by themselves will run you $663.52.

To find out more about the TrioPod PRO75 lineup or order some bits and pieces for yourself, head over to the Novoflex website.

Press Release

Introducing the NOVOFLEX TrioPod PRO75

The new Modular Tripod of choice for discerning professionals looking for the sturdiest, modular support solution.

For almost four years, the NOVOFLEX TrioPod tripod collection has been lauded as the most innovative tripod system on the market. The Product Innovation Team at NOVOFLEX is proud to announce the expansion of this tripod series with a modular, high-capacity tripod system, the TrioPod PRO75.

Exceptionally stable and modular with immense load capacity.

The basis for the all-new TrioPod PRO75 is the advanced tripod spider, which offers exceptional stability and immense load capacity. Thanks to the reinforced design of the individual components and new 8-layered carbon fiber legs, this modular tripod can be used for even the heaviest photo and video equipment.

The system can be purchased with 3 or 4 section carbon fiber legs plus two mini legs for maximum versatility.

In addition to the recommended new tripod legs C3930 and C3940, existing TrioPod legs can also be used with the new TrioPod PRO75 as well. With the optionally available carbon leg extensions, a total height of 79 in./2.0 m is possible. Short 2-segment carbon fiber legs C2820 are also available allowing you to achieve completely new perspectives and are even compatible with the existing TrioPod and QuadroPod Systems.


Maximum Load Capacity – 143 lbs/65 kg – Maximum holding capacity for maximum protection.

Removable Base Plate – A new feature of the tripod spider is the removable base plate that easily be detached and replaced with the optional leveling ball (MBAL-PRO75) or a geared center column (TRIO-CC-PRO75).

Unmatched Versatility – To build your tripod to meet the needs of any job the PRO75 can fit most 75mm accessories, even from other manufacturers. The optional leveling ball (MBAL-PRO75) can also be used with any other tripod you may have with a ¼” & 3/8” connection.

Self-Locking Geared Center Column – The innovative 3/8” studs on both ends of the geared center column provide an extra measure of security and compatibility by taking the load off the ¼” threads. The column itself can be adjusted by 19 in./48 cm.

7 Leg Lock Positions for Any Situation – For shoots on level surfaces, the TrioPod PRO75 spider allows the legs to be set at 23°, 45°, 65°, and 87°. Should you find yourself in a more confined space, a further locking position of 155° enables you to support the tripod against vertical surfaces. Particularly innovative is a new 43° angle that allows the tripod to flip upside down quickly for ground-level work, eliminating the need to reverse the center column and/or tripod head. For easy packaging, the legs can also be folded up at 180°.

Ground Level Work

Easy Packaging

The TrioPod PRO75 is available in two kits:

  • TRIOPROC3930 – consisting of TrioPod PRO75 tripod spider, (3x) C3930 3-segment carbon fiber legs, (2x) A1010 mini-legs, a tripod bag and multi-tool.

  • TRIOPROC3940 – consisting of TrioPod PRO75 tripod spider, (3x) C3940 4-segment carbon fiber legs, (2x) A1010 mini-legs, a tripod bag and multi-tool.

The two QLEG-A1010 mini-legs that are part of the kits can be screwed into the accessory thread creating limitless possibilities: Add one leg to smooth your workflow by attaching your camera bag, keeping your tripod weighted and giving you quick access to your equipment. Alternatively, replace two legs to create a “leaning pod” and achieve a new variety of creative perspectives for your boundless imagination.

Video: 360° 8K timelapse of the aurora borealis during a lunar eclipse

Photographer William Briscoe captured this spectacular 360° 8K timelapse on January 31st near Fairbanks, Alaska. He shared the video on his YouTube channel and Facebook page a few weeks ago, alongside this description:

Here is a 360 video of the Lunar Eclipse, Alaska style, which I filmed on January 31st near Fairbanks. Lady Aurora, being the Diva she is, just couldn't let the moon have all the attention that night, so she made a nice showing as well.

If you have VR glasses of some sort, by all means slap them on! If not, simply drag the video around until you spot the moon, then watch as it disappears and becomes a black disk by about 30 seconds in, allowing the aurora to cover it entirely for the duration of the eclipse.

Unfortunately, capturing both aurora and eclipse in a single exposure is not possible, something Briscoe revealed in a comment on YouTube:

It is a composite video. The difference between the settings required to properly expose the moon and the Aurora is too great to do it in a single shot. The 360 camera was set to expose the Aurora and Landscape, while a second camera attached to a telephoto lens was used to time lapse the lunar eclipse itself. I combined them in post.

Without the second timelapse, the moon, even during eclipse totality, was bright enough to cause a white dot in the sky at the settings I had to use.

Of course, the fact that it was composited doesn't mean it was easy to shoot. Responding to a comment on YouTube, Briscoe revealed that it was -31°F (-35°C) out that night, so just getting his 360° rig (a custom-built array of numerous 35mm DSLRs) to work and staying warm for the three hours it took to shoot this was a challenge!

Vivo's AI-powered 'Super-HDR' tech takes on Google's HDR+

Google's HDR+ mode is widely regarded as the current benchmark for computational imaging on smartphones, but Chinese manufacturer Vivo wants to unseat the champion. Earlier today, Vivo announced its AI-powered Super HDR feature—a direct competitor to the Google system found in Pixel devices.

Super HDR is designed to improve HDR performance while keeping a natural and "unprocessed" look. To achieve this, the system captures 12 exposures (Google uses 9) and merges them into a composite image, allowing for a fine control over image processing.

Additionally, AI-powered scene detection algorithms identify different elements of a scene—for example: people, the sky, the clouds, rocks, trees, etc.—and adjust exposure for each of them individually. According to Vivo, the end result looks more natural than most images that use the simpler tone-mapping technique.

Looking at the provided sample images, the system appears to be doing an impressive job. That said, these kind of marketing images have to be swallowed with a pinch of salt; we'll see what the system is really capable of when it's available in a production device we can test.

Speaking of which, as of now, we don't know which device Super HDR will be shipping on first, but there is a chance it might be implemented on the upcoming Vivo V9, which is expected to be announced on March 22nd. The V9 is currently rumored to feature a Snapdragon 660 chipset and 12+8MP dual-camera.

Teardown video shows how the Galaxy S9 variable aperture works

The YouTube channel JerryRigEverything recently tore down (or rather tore apart...) the new Samsung Galaxy S9, giving us the closest look at yet at the new smartphone's camera hardware. So if the still images in the iFixit teardown weren't quite interesting enough for you, this might just do the trick.

The camera teardown is about one minute and a half long, running from the 3:30 mark until about 5:00. In that time, we get to see the Optical Image Stabilization system demoed and torn open to reveal the magnets inside:

Then, we get really close look at one of the phone's most intriguing features: the variable aperture. It turns out the system works using a little lever on the side of the housing. So when the phone senses that there is enough light to justify it, it'll flip this switch electronically and switch from it's world's-brightest F1.5 setting to F2.4.

Here's a very close look at that switch in action:

You can check out the full teardown in the video at the top. And stay tuned, because we'll be bringing you a full smartphone camera review of the Samsung Galaxy S9 just as soon as we can put a unit through its paces.

This may be Amelia Earhart’s Leica, and you can buy it for $69,000 on eBay

A working-condition Leica camera that may have belonged to Amelia Earhart—the first female pilot to fly solo across the Atlantic and an iconic historical figure—has popped up for auction on eBay.

The Leica l Model A, dating from between 1926 and 1927, is set at a 'Buy it Now' price of £50,000/$69,490, and comes with a card signed by Earhart herself; unfortunately, this is the only proof we have that the camera really did belong to her. The seller acknowledges that the connection is a little tenuous, but claims the camera was given to his grandfather by Earhart in 1933 when she decided that she preferred a more user-friendly Kodak folder.

Earhart became famous for her pioneering flying and the records that she broke during her career, receiving the US Distinguished Flying Cross for her solo Atlantic flight. In 1937, she went missing during an attempt to fly around the world, and is presumed to have crashed somewhere in the Pacific Ocean (although the uncertainty surrounding her death has led to numerous theories that she didn’t crash at all).

Photo: eBay Auction

The camera, which is said to be in good condition, has a non-interchangeable Leitz Elmar 50mm F3.5 lens and comes with the tall vertical rangefinder paired with the camera at the time, but which was available before the camera was made. The seller is also including a pair of metal film cassettes, and a ‘rare’ Leica purse to hold the lens cap.

The seller’s family collected cameras, and a part of that collection went to auction last year in Glasgow, Scotland, but this model failed to reach its reserve of £15,000/$21,000 and remained unsold. The seller believed the auction house gathered the wrong audience for the camera, which is why it is now on sale for a somewhat higher price.

Photo: eBay Auction Photo: eBay Auction
Photo: eBay Auction Photo: eBay Auction

At the time of writing, the eBay item has 26 days to run, and if you feel the 'Buy It Now' price is a fraction high for a camera with questionable provenance, you can still make an offer. Examples owned by less famous people can be had on the same site for as little as $1,500... or a bit more from a reputable dealer.

Seller's Description

Im selling Amelia Earharts camera which was gifted by her to a family memeber in 1933 after returning back from a trip to Chicago with her Husband.

The camera has been in my family possession since that time and has been in long term storage, the camera appears to be working correctly.

The hand signed card was personally signed by Amelia and given to my Grandfather along with the camera by Amelia Earhart back in 1933 in Rye New York

Everything is authentic, I've known this camera all my life the signed card is almost like new as it has been stored carefully will post world wide. I would like the camera to go to a museum if possible.

Please note I have absolutely nothing to prove that this was in fact Miss Earharts Camera and research would need to be done to confirm such, I have absolutely no idea how to do that myself. From memory, over 40 years ago my Father told me that she found it fidly to load, Miss Earhart may have studied Photography, my Grandfather had said as much and described her as a keen photographer, she preferred a Kodak folding camera as I recall being told, she was also described as very nice and down to earth.

Amelia's camera was at Mctears Auction house in Glasgow in March 2017, it was part of a rather large collection of cameras that I sold through Mctears. Unfortunately, the auction house could not find enough interest in the UK for Amelia's camera, and as such the camera remained unsold. I can say this as I want to be totally transparent. Auction estimate was 10- £15,000. The last picture is from Mctears Auction house, I was there on the day that picture was taken, Mctears had used a trade gazette to advertise the collection and as such I considered some items were sold at less than their true value and then re-sold on by dealers at a profit later.

I do understand that Provenance is an issue, If I had that the camera would be worth Millions, not thousands. I had Bonhams Auctions out in 2016 who said as much when they inspected the camera.

Rokinon AF 35mm F2.8 FE sample gallery and impressions

The Rokinon / Samyang AF 35mm F2.8 FE is an absolutely tiny full-frame lens built for Sony's E-Mount. Similar in size and specification to Sony's own Zeiss 35mm F2.8, the Rokinon is listed at an MSRP of $349 whereas the Sony is listed at $699 at the time of this writing.

So, by virtue of costing half as much as the Sony, does the Rokinon offer merely half the performance? Not exactly. The AF 35mm F2.8 FE may not knock your socks off, but it's still a solid performer and a great option for budget-oriented E-mount users.

The Rokinon AF 35mm F2.8 FE mounted on a Sony a7R III.

Hold the Rokinon 35mm F2.8 in your hand, and you have to wonder if there's any glass in it at all. It weighs over an ounce less than the Sony 35mm F2.8, which wasn't exactly a heavyweight to begin with. It's also very slightly shorter than its Sony equivalent, though both lenses take 49mm filters and can focus down to 0.35m. Unfortunately, the Rokinon omits the Sony's claimed weather-sealing, and the lens mount lacks any sort of gasket.

Subjectively, the Rokinon's plasticky build doesn't have the premium feel of the Sony (nor would we expect it to at the price), but the construction feels solid. The mount is metal, and the included bayonet mount hood offers a bit of extra protection for the front element. The focus ring is damped enough to prevent accidental turning, but it too suffers an overly plasticky feel.

The Rokinon comes with a compact, bayonet-style hood.

Optical performance is pretty good, even on a 42MP a7R III. It isn't eye-searingly sharp wide-open, but it's more than adequate. Bokeh is neither buttery nor overly busy, but out-of-focus highlights take on a cats-eye shape near the edges of the frame, which may or may not be to your taste. Sunstars are possible if you close your aperture far enough, but they're of average quality.

Autofocus performance is solid. It isn't as instantaneous as lenses with floating focus elements, but is about on par with Sony's own 85mm F1.8 and 50mm F1.4 Zeiss – certainly, swift enough for general use.

Though the Rokinon generally exhibits good out-of-focus renderings, there is some green and purple fringing noticeable in the upper-middle of this frame.

Longitudinal chromatic aberration hasn't been too much of an issue for me, but you can see some green and purple fringing on the high-contrast edges in the upper-middle of the above image. Keep in mind that for our sample gallery, Adobe Camera Raw has a built-in profile for distortion and vignetting corrections. Lateral CA corrections were left off, and the lens seems to control for them fairly well.

It must be said, the Rokinon AF 35mm F2.8 FE is just a fun lens to use. On any a7-series camera, the Rokinon is so small, light and unobtrusive that it basically disappears on the camera body.

Though its F2.8 maximum aperture won't isolate subjects as well as Rokinon's or Sony's F1.4 options, it strikes a great compromise between size, performance, and perhaps most crucially, price. If you've been eyeing the Sony Zeiss 35mm F2.8, we think this Rokinon AF 35mm F2.8 FE is also worth a look.

Google just made the tech behind its 'portrait mode' open source

Semantic image segmentation is the task of categorizing every pixel in an image and assigning it a semantic label, such as “road”, “sky”, “person” or “dog”. And now, Google has released its latest image segmentation model as open source, making it available to any developers whose apps could benefit from the technology.

This kind of tech can be used in many ways. One recent application in the world of smartphones is the portrait mode on Google's latest Pixel 2 devices. Here, semantic image segmentation is used to help separate objects in the foreground from the image background. However, you could also imagine applications for optimizing auto exposure or color settings.

This kind of pixel-precise labeling requires a higher localization accuracy than other object recognition technologies, but can also deliver higher-quality results. The good news is that Google has now released its latest image segmentation model, DeepLab-v3+, as open source, making it available to any developers who might want to bake it into their own applications.

Modern semantic image segmentation systems built on top of convolutional neural networks (CNNs) have reached accuracy levels that were hard to imagine even five years ago, thanks to advances in methods, hardware, and datasets. We hope that publicly sharing our system with the community will make it easier for other groups in academia and industry to reproduce and further improve upon state-of-art systems, train models on new datasets, and envision new applications for this technology.

If you are interested in finding out more about DeepLab-v3+, head over to the Google Research Blog for more details.

Huawei shares P20 triple-cam teaser videos

We are getting close to the launch of Huawei's upcoming flagship P20 smartphone on the 27th of March, and thanks to a number of leaks we're already fairly certain that the new device will come with triple cam setup, offering a total resolution of 40MP and a 5x optical/digital hybrid zoom.

Now Huawai itself has started teasing the new model with a couple of videos on its YouTube channel.

The first one, called "See Brighter" (above), teases the smartphone's low-light capabilities. It features a "pro" photographers with a DSLR and huge assortment of lighting gear next to a P20 user who is happily snapping away in the same lighting conditions without any additional equipment.

The second video (below) is titled "See Closer", hinting at the P20's zoom capabilities. It follows the same scheme as the first clip, showing a DSLR-photographer shooting with a selection of heavy prime and zoom lenses, next to a smartphone user with triple-cam-equipped Huawei P20.

If the teasers are anything to go by, the P20 could combine a "regular" RGB image sensor with a high-resolution monochrome chip for improved low-light performance and digital zooming, and another RGB sensor with a longer lens for optical zooming. And if Huawei's engineers can manage to merge image data captured by all three sensors in an efficient way, the P20 could be the mobile camera to beat in 2018.

Fujifilm GFX 50S firmware update adds Focus Bracketing and 35mm Format Mode

Earlier today, Fujifilm released firmware version 3.0 for its GFX 50S medium-format camera. And with the update come two new features: Focus Bracketing and 35mm Format Mode.

Focus Bracketing enables focus distance bracketing for up to 999 frames, shifting the focal location with each shutter activation at a scale from 1 to 10. The new 35mm Format Mode, meanwhile, results in the camera using the central 36 x 24mm portion of the sensor, producing 30.5MP Raw and JPEG images.

"When using 35mm format adapters," the company explains in its changelog, "it will be easier to adjust the image size to 35mm format image circle lenses by changing the setting."

The GFX 50S v3.0 firmware update can be downloaded now from Fujifilm's website. Additionally, Fujifilm has also released a compatibility update for the H Mount Adapter G accessory.

Update Notes

GFX 50S Camera Body Firmware Update Ver.3.00

Addition of “Focus Bracketing”

The update will enable the photographer to shoot focus distance bracketing up to 999 frames. When the photographer start shooting, the focal location is shifted with each activation of the shutter by the step of focus shift set from 1 to 10.

Addition of “35mm Format Mode”

The update will enable the photographer to shoot central 36.0mm x 24.0mm (30.5M) cropped images as both JPEG and RAW files. When using 35mm format adapters, it will be easier to adjust the image size to 35mm format image circle lenses by changing the setting.

H MOUNT ADAPTER G” new firmware

The firmware update will expand the compatibility for the “H MOUNT ADAPTER G” accessory. Regarding the compatibility of lenses and accessories with H MOUNT ADAPTER G, refer to this URL.