Video

DJI Mavic Pro – Initial Thoughts

Long time readers will know that I’ve been playing with a DJI Phantom 3 Advanced for over two years now. I’ve posted a number of videos and photos from that drone in the past and I’ve been very happy with it. However there’s no denying that its size limits where you can take it. You do have to actively choose to take it with you – perhaps at the expense of other things. I’ve had the whole kit fit into a rucksack, but I need other baggage if I’m taking anything else with me. It’s especially limiting for something like a cycle ride.

So when DJI release the Mavic Pro last year I was very tempted. It has more sensors than my Phantom 3 Advanced. Notably my early mistake of crashing the Phantom into a tree wouldn’t be possible with this. And the functionality is greatly improved too with lots more modes making use of the available sensors.

Now I confess that I was actually pretty happy with the 2.7k camera of the Phantom. While 4k video is nice and future proofs myself, I don’t actually have any way to see it back at that resolution. My laptop isn’t that high a resolution and nor is my TV. In any case, there are limits to the bitrate it records 4k at, and I suspect I’ll be using 2.7k to a larger extent with the Mavic.

One other consideration was the ultraportable DJI Spark which has just been released. This only shoots 1080 but is tiny and is a true go-anywhere drone. The arms don’t fold smartly like the Mavic, but it’s very pocketable and is the sort of thing you can easily keep in the bottom of your bag. I suppose my biggest concern was its performance in windy conditions. The UK doesn’t always have perfect drone flying weather, and while I’ll always avoid the rain, bigger beasts tend to be a bit more solid in the air. That said, in some of the YouTube reviews posted, it looks like it performs decently.

Another consideration is the likelihood that DJI will release an updated Mavic at some point soon. The Mavic Pro was released in October last year but it was post-Christmas before it was more widely available. No doubt there will be a new model available ahead of Christmas this year, but it’s like waiting for the next phone or the next computer. There’s never quite the right time to buy.

I still went with the Mavic, and I bought the Fly More pack – as much as anything for the additional two batteries. Other accessories were moderately useful, including some extra props, a car charger and so on, and although the bag is nice it’s not perhaps quite as tough as would be useful for protection. The hub charger is good, although I was slightly disappointed to learn that it charged batteries sequentially. So if you load 3 or 4 batteries onto it, it still only does them one at a time.

You will need to allow time to charge everything up, and importantly, get the most up to date firmware. DJI basically forces users to use the current firmware by making you use an account. If you don’t have the current firmware, your flight distances are highly restrictive. That all said, while it’s still slow, it’s a bit more obvious than the old Phantom 3 way of doing things. I updated using my phone (on WiFi at home), and that gives you a good idea of where you’ve got in the installation and update. Much better than listening for strange sounds from you Phantom!

The little remote is very nice, and DJI include microUSB, Lightning and USB C cables in the box to connect with your phone. I’ve been used to using a Nexus 7 for my phantom – which is basically that device’s single use. And I notice that there are lots of devices around that let you “mount” a tablet to your remote if you want. In use my phone (an HTC 10) worked well, although you’re going to be flattening your phone’s battery using it rather than a bespoke device. I can live with that for the benefits in portability. I’ll probably pack a small Anker battery charger in my DJI bag to recharge my phone if it flattens after a decent flight. I did however notice that the Mavic’s remote loses charge faster than my old Phantom 3 Advanced remote, which could effectively run for weeks on a single charge.

I played a little with some of the modes that track you as move, although I need to learn more about them.

The Mavic has a range of up to 7 km, but UK regulations say that you need to be able to see your drone at all times. There’s no way you can see something as small as a Mavic that far away. And I’m not sure whether a first-person headset would actually be legal here. One of the problems I had with a white Phantom was losing it against a white cloud sky. The dark grey of Mavic makes it easier to spot (and remain legal), and in any case, I probably don’t want to have to trek 7 km to retrieve an errant drone should it decide to land there. (I confess that I’ve never had a drone “fly away” from me. The closest I’ve got is have to use my eyes to fly a drone back when video connectivity caused me problems one time.)

One issue I will think about is bringing some kind of mat or cloth with me take off and land from. The lowness of the Mavic means even quite short grass could get caught by the propellers. I often find myself launching from grass, so a simple and lightweight solution would be useful.

I’m still learning about what the best recording settings are. I use RAW+JPG for photos, but video is another questions. I’ve also never bought any ND filters for a drone, although they do seem to be used a lot by the pros. We’ll have to see.

I did try doing some live streaming. I first tried YouTube, but the app kept crashing repeatedly when I tried it. So I fell back to Facebook which is never anyone’s favourite platform. They limit the video quality a bit for starters. But I confess that it just about worked (I’d embed the video, but Facebook makes that truly fiddly). A few people watched, although playing back, the video quality was at times poor. More annoyingly, there were cutouts continually. I was on 4G so it shouldn’t have been too bad, but it wasn’t great. Still it was good that the phone’s mic could be used for a sort of commentary of what I was shooting.

Otherwise, I’m really happy with my purchase. Expect more drone footage soon. In the meantime, here’s a sample of what I shot today.

Note that I shot on the default settings and have only done a tiny bit of colour correction on a couple of shots. I’m also interested to see what the difference is like between YouTube and Vimeo, so I’ve uploaded it to both, using Premiere Pro CC 2017’s settings.

Hilly Fields – First Mavic Pro Flight from Adam Bowie on Vimeo.

There are a couple more photos on Flickr too.

Problems with News Video

Recently the Reuters Institute for the Study of Journalism published its annual Digital News Report, authored by Nic Newman.

If you’re interested in the media, and particularly journalism in the digital age, then it’s an essential read. The report, which is supported by groups such as Google and the BBC, surveys 50,000 people across 26 countries about their digital news habits.

The report is available to download, with lots of additional resources like data tables and chart packs for deep-diving into.

I’m going to concentrate on one area of the report: video.

If you’ve been paying attention to news sites, and indeed digital media in general, there has been a lot more video in recent years. Social media and news sites more often than not playing videos by default, and spending money to push the platform. Video, the belief is, will grab users’ attention and drive increased readership.

And for the most part, this seemed like a sensible move. More people were watching more video as both home broadband and mobile 4G coverage improved. But with regard to digital news, there’s been a bit of a speed-bump on the road.

“One surprise in this year’s data is that online news video appears to be growing more slowly than might be expected. Across all 26 countries only a quarter (24%) of respondents say they access online news video in a given week. This represents surprisingly weak growth given the explosive growth and prominence on the supply side.”
(Page 19)

The real reason for the growth in video, beyond the perceived demand from users, is the higher advertising yields that can be achieved from video. Those pre-roll adverts, whether skippable or not, are worth much more than other display inventory which has not been the saviour that news organisations or others had hoped it would be. Something to do with infinite inventory I suspect.

News providers were positively driven to increase their volumes of video to meet revenue targets.

“Across our entire sample, the vast majority (78%) say they only read news in text or occasionally watch news video that looks interesting. Just one in twenty (5%) say they mostly watch rather than read news online. “
(Page 20)

And the reasons for this relatively low growth are pretty obvious. This chart is from the report:

newsvideos

I think those reasons – the first four in particular – chime with me, with the fact that I can read text quicker than watch a video being the chief one.

Yet frustratingly, more news seems to be appearing in video-only form. I read much of my news via the feedreader Feedly, and most news site’s RSS feeds limit what Feedly can see. That’s fine – whether coming from a feedreader, or much more likely, social media, news providers want to ensure they have strong branding and potentially monetise me with advertising.

But when I click through to a site and see a story that is only, or mostly, video, then I simply close the tab and click away.

Video really needs to add something to what I can read for it to be of true value. I’m not saying I don’t like video news – I watch TV news bulletins on a daily basis – but in a digital world, video is much more an interruptor.

– If I’m on the train to work looking on my mobile, I may be listening to music. Video puts that on pause so I can hear the soundtrack. Newspapers never forced that on me. I can read text and listen to music simultaneously.

– If I’m at work, then I can quickly scan a story to see if it’s important. With video I have to fumble around for headphones, or risk interrupting colleagues.

And video takes time. From hitting the play button to getting to what I want to see is not usually the best experience. Frankly, there’s nothing worse than a news provider who has built their own video platform (or bought one), and you just know it’s not going to be as fast-starting as, say, YouTube. You’re going to see a swirly “loading” graphic before an advert loads painfully slowly. At the end of the advert, there’ll be another delay as the actual video loads. 30 seconds of that before a video that’s only 45 seconds long itself doesn’t seem like a fair transaction.

Fundamentally, humans can read in their heads faster than someone can read out loud. So all things being equal, I’ll choose the most the most efficient way to get to the story. For the most part, I want to read stories not watch videos. I can quickly gauge how interested I am in a story from the text. Video is a hit or miss affair.

It’s perfectly true that some may prefer video, so by all means offer both video and text. But consider even making the transcript of the video available. As a friend pointed out on a social media, that instantly makes the video more accessible, and increases the search engine optimisation of what you’re producing.

Video is actually much more expensive than text – or text illustrated by photos – yet everyone seems to want to do it.

My suggestion is that unless video is a primary output of your organisation, I would use it sparingly. Produce only videos that really add something to the story. There are various groups who are adding text to videos and making them viewable without sound. Fine as far as it goes, but they tend to be relatively simplistic. You can’t delve deeper into a story that way, yet if I’m spending 2-3 minutes with a story which is what a video is demanding of me, then I expect to come out with a much richer understanding of the issues than I went in with.

Video is not the be-all and end-all, and news providers would do well to remember that.

Cycling in the City

Cycling in the City from Adam Bowie on Vimeo.

Here’s a short film I shot with my Super 8 camera about four years ago that somehow hadn’t previously seen the light of day. The film was basically shot on my commute to work along the cycle paths of Bloomsbury and around Tavistock Square. Unfortunately, I didn’t use the correct filter for my daylight film, so some remedial colour correction has been necessary.

A Timelapse of 25 Great Pulteney Street, London

25 Great Pulteney Street – Timelapse from Adam Bowie on Vimeo.

Do you ever embark on a project that somehow you never quite get around to completing?

I certainly do. And here, finally, is a project that I’ve completed some six years after I started it.

This begins back in 2010 when I was still working for Absolute Radio in One Golden Square, Soho, London. I sat at the back of the fourth floor, windows behind me looking across the Soho alleyway that is Bridle Lane and towards a building on Great Pulteney Street beyond. The fairly ugly building from the 1960s at 25 Great Pulteney Street had once been home to the agency Starcom Motive. But they’d long moved out, and the building had been empty for at least two years. When Google Street View’s team passed it in 2008, the building was boarded up, and that was still the case by 2010.

Now finally the developers were moving in, and it looked like something was happening with the building. I brought in an old Canon A470 digital camera, bought cheaply on eBay, and loaded a memory card with CHDK – the alternative firmware that would provide my camera with timelapse facilties. I also bought an external power supply and suction camera mount.

Over the next year and a half, between April 2010 and October 2011, I set the camera taking photos – first of the demolition of the building, and then of the new building rising in its place. I wasn’t consistent in either the location of the camera, the frequency of it taking photos, or what it was pointing at. When I went on holiday, I made sure to keep the number of photos a day low to ensure the memory card didn’t fill.

In retrospect, there are a lot of things I’d have done differently, including changing the aspect ratio, the photo size and so on. I was left limited in what I could do with zooming or panning across scenes. The camera was also limited in its angle of view from my window, and the camera was often mounted on a slight angle. The buildings were too close and the lens not wide enough to capture everything in one shot. The window meant reflections, and it wasn’t perfect either – neither clean nor unscratched.

Over the course of having the camera in my window, I had to ensure that cleaners at Absolute didn’t unplug it. The suction of my suction mount would invariably fail over time, and I’d come into work to find the camera on the floor. On one occasion the fall was “fatal” and I bought an identical replacement on eBay to continue the project.

In total I ended up with something like 250,000 photos. I wrangled them into something useable with Quicktime Pro, getting MP4 files from my JPGs. On underpowered PCs, this was a slow process.

Finally I had a collection of 102 files, taking up about 10GB. And then I sat on the videos. I couldn’t say way exactly. I suspect that I found it a little daunting. I knew that there was too much video and it needed editing down, although it wasn’t really a big job. I had to find some music – two tracks. One for the demolition, and another for the new building. I’ve sped many of the clips up further, removed nights for the most part, when as already mentioned, reflections of my office were a problem.

Finally it should be said that this is by no means every minute or day of the building being demolished and rebuilt. But it’s lots of it.

And so it is, that some years after capturing this footage over many many months, I’ve pieced together this video! I hope you enjoy it.

Notes and Further reading:

The architects were WilkinsonEyre, and they have a nice project page with some lovely photos of the finished building.

The building is reported to have cost £9.5m to complete, and was for the client F&C Property Asset Management (now BMO Real Estate Partners)

The front of the building has some interactive railings in a piece called Finial Response designed by Cinimod Studio.