The Galaxy S7/Edge can take photos with a few seconds of video before the photo. iOS and Windows Phone devices can do the same, as I’m sure many other phones can. I was curious about how well it would work on the street, as it might be useful to show how a shot came about. Surprisingly, it worked well! I could even take two photos in quick succession, and the phone would record overlapping video for each. I didn’t notice any lag.
The phone records the motion photo as a regular jpg file. It’s much bigger though, as there’s a mp4 video files appended to the Jpeg image. You can find some technical info about the file format here.
Yes you can manually do this is 10 seconds with a hex editor by opening your original Photo and searching for MotionPhoto_Data then select everything above MotionPhoto_Data and copy and paste it as a new file and save as a JPG. Do the same for the MPG BUT this time select every thing BELOW MotionPhoto_Data make sure for either one you are doing not to copy the MotionPhoto_Data text. Also the hex for the MotionPhoto_Data is 4D 6F 74 69 6F 6E 50 68 6F 74 6F 5F 44 61 74 61.
Technically the JPG ends at ÿÙ or FF D9 so the ÿÙ SHOULD be included too.
Technically You really want to delete all this “……..Image_UTC_Data1458170015363SEFHe……… ..#…#…….SEFT..0…..MotionPhoto_Data”
But if you leave all this “……..Image_UTC_Data1458170015363SEFHe……… ..#…#…….SEFT..0…..MotionPhoto_Data” in the end of the JPG you can easily merge them back together and have the display and play on your phone
For this first test of motion photos I just used the phone to extract the images. Here’s what I did:
Sometimes ImageOptim messes up and Lightroom shows the image as a sort of double-image with interlaced lines like this:
If you don’t have any backups, I found I could fix the problem by loading the image in GIMP and saving it again. Then I’d have to update the EXIF data or overwrite it in Lightroom.
Using the Ruby script above is even easier. Point it at your import directory and it’ll extract any images and videos it finds, appending “_Extracted” to the filename. It’s probably worth renaming them in some deterministic way to avoid the stacking problem I talked about above. Since it extracts everything, you can delete the original jpg file. I should modify the Ruby script so it does the rename step..
I doubt I’ll use this feature much but I’m sure I’ll use it from time to time.
The Samsung Galaxy S7 and S7 Edge have a 12MP camera, which is fairly mediocre as far as megapixel counting goes these days. The Galaxy S5 produced 16MP photos, and that device is two years old now! It’s not the megapixel count that matters of course but what kind of pixels they are and what software processes the data they create.
Apparently because the camera has fewer pixels, those pixels can be larger, and therefore collect more light, like buckets collecting water. The camera has image stabilisation built in too which helps keep photos sharp when your hand isn’t quite steady enough.
I shoot with the camera held at chest height in both hands and I used the volume buttons to take the photo without looking at the screen to compose. This worked fairly well, except when I accidentally obscured part of the lens with a finger. I need more practise.
I found I sometimes activated image review, probably by brushing against the screen with a finger. That was really annoying, especially when I thought I had captured the moment.
It also allows you to shoot RAW photos, capturing in DNG files that Lightroom or other RAW processor can read. You’ll have to enable PRO mode which disables other features like panorama, but it’s worth it. The advantage over Jpeg is that you get more leeway when conditions aren’t perfect for a photo. Say, you’re shooting a photo of someone against a bright sky. You’ll have a decent chance of recovering detail in the shadows.
I tried manually setting the exposure time, hoping the auto ISO would kick in when needed, but when exposure is set, auto ISO is disabled so I went back to auto-everything which worked very well anyway.
The Jpeg files the camera creates are very sharpened. You probably won’t need to apply much sharpening at all to them in development. This is something to keep in mind when switching between Jpeg and RAW.
The DNG files the S7 creates aren’t quite up to the standard you’ll get from a DSLR camera however. Jpeg files record 8 bits of data, a DSLR will commonly create RAW files containing 14 or 16 bit data. The Galaxy S7 outputs 10 bit data according to Lightroom. While that isn’t much different to Jpeg, it’s still enough to make a difference when working in taxing conditions. Laura Shoe has a good article explaining the differences between the different bit depths.
There’s something odd about those DNG files too. The files are generally 24-25MB in size, which is similar to the RAW files created by my 20MP DSLR. I thought that was strange enough that I decided to run Lightroom’s “Convert to DNG” on them. Without using lossy compression it converted those large files into much smaller 7-8MB files! Using a new DNG file I tested to see how well it would compress. A 24MB file was zipped down to 12MB. That means most of the file is empty space, not even random data which wouldn’t compress well. For comparison, a RAW file from my DSLR will compress from 20MB to just over 19MB.
It’s worth saving the space by converting the Samsung DNG files to Adobe DNG files.
There’s something odd about the metadata in the Samsung DNG files too. Mac OS X Finder won’t show previews of these files. When I looked at the metadata I saw the preview dimensions were set to 0x0. The converted DNG files were fine however.
The metadata problem extends to panoramic photos shot with the camera too. Here’s a lengthy thread on the matter. I fixed the problem by importing the photos into Apple Photos and then exporting the panoramic images. As a side bonus, the size of the file went from 44MB to around 17MB without discernible loss of quality.
The camera will record Jpeg and RAW images when in RAW mode, and it saves them to the internal memory, probably for performance reasons. I use the following chunk of code to delete those Jpeg files:
for i in *dng; do rm `basename $i .dng`.jpg; done
Snapseed and Lightroom mobile both allow you to edit DNG files. Snapseed will display a Develop interface where you can “develop” the raw file by adjusting much the same controls as “Tune image” has, but you can stretch them a little more. Here’s a Jpeg I edited in Snapseed.
I hate using a smart phone without a case. The S7/Edge is so slippery I’d be afraid to carry it outside without one. I opted for the Galaxy S7 Edge Case – Exact [TANK Series] case. I already have the S5 version and the S7 one is just as good.
I really liked shooting with the Galaxy S7, and I’ll use it again. It’s certainly a pleasure to use, even if there are some problems with the files it makes. A minor problem, given that it’s easy to fix them.
Edit: The S7 is fairly waterproof, able to operate while submerged. I’m not going to test that capability but a little rain doesn’t bother it at all!
A number of Cork street photographers travelled to London on Saturday to photograph the populace of this sprawling metropolis. If you don’t know, Cork is a city in Ireland, home to roughly 400,000 people and served by an international airport where flights to London in the UK take a little over an hour or so.
From left to right, at the back are Rory, me (Donncha), Colm, Pat and Robert. In front is Shahriar and Stela. This group photo was taken in Stansted Airport at the end of a long day when we were all exhausted. Except Pat. He could keep walking for ever!
The idea was originally Pat Falvey’s. Cork Street Photographers went to London for the day the year before (or was it two?) and Pat had been with them. So, early in March I asked some of the Cork Street Photography Meetup Group if any of them were interested in going to London on April 9th. We’d get the first flight out of Cork at 6:25am, and fly home on the last one back at 8:35pm. There was immediate interest!
A few hours later I created a Facebook event, posted a few links there and invited people to go. We used the event page to talk about logistics and planning. I posted some links to street photography in London. We researched our transport options. Stansted Express has a discount if buying four tickets at once which we took advantage of. The Ryanair flights were cheap, from 25 Euro to 37 Euro depending on when they were bought. (Paradoxically, they went down in price after most of us bought our tickets!)
It was somewhat confusing looking up information about Oyster Cards. The visitor ones offered discounts but it turns out they’re more expensive if you’re only going for one day. We eventually bought regular Oyster cards in Liverpool Street Station and topped them up. According to the TFL website our charges would be capped at £7.70 for the day no matter how many journeys we took.
I made out an optimistic schedule two days before leaving, first in a Facebook update, then in a Google spreadsheet. I started a Facebook messenger group chat to get feedback.
We wanted to visit the Victoria & Albert museum to see the Paul Strand exhibition there but by the time we got there it was sold out. Half the group had stayed outside making photos around the area so those of us inside the museum wandered over to the cafe and were awed at the beautiful room. It was then a quick hop over to the Science Museum where I got a photo of their chunk of Moon rock, Stela took a photo of me with it and we ran back to meet the others.
We left the museums around 5:15pm and found the underground station at East Kensington packed because of a delay. Thankfully it didn’t take long and we soon made it back to Liverpool Street Station, and then back to Stansted Airport.
The schedule was overly optimistic as I knew it would be. We never got to Trafalgar Square, Westminister, or Tower Bridge. In planning I had used the TFL website to plan our journeys and inserted short links to them in the spreadsheet which I could load in the Google Drive app on my phone.
You won’t need much roaming data on the day, so if you can buy a 100MB or 200MB upgrade for a few Euro before leaving that will take care of checking Facebook, or uploading a few (resized) photos.
Trying to keep track of seven photographers was problematic. We have a tendency to wander off when something catches the eye. Trying to get through the crowd near the London Eye is almost impossible, taking a good half hour! 🙂
It was an eventful day. I thought I had shot more than 880 photos, but Lightroom says it was 859. The vast majority of those will be consigned to the trash can, but I’ve already worked on 25 images I’m happy with and there are lots more to look at yet. It was great being with a group of enthusiastic and friendly photographers too. They really made the trip worth while.
I didn’t spend much. I brought £100 with me and came home with £40. That covered buying small gifts for my family, lunch, Oyster card, tips for street performers and an (overpriced) sandwich and bottles of water at Stansted Airport.
I’ll be posting photos from the trip over the next few weeks. You’ll eventually find them all here.
Going over to London from Cork and back in one day was exhausting, but well worth it. You should try it, go visit a city you can reach by a short one hour flight. You get bonus points if it’s in another country!
|Camera||Canon EOS 6D|
I had fun rescuing this really underexposed landscape photo in Lightroom.
It’s a photo of Kilcrea Friary in Co Cork and I went out there one September evening in 2015 with a few others from Blarney Photography Club to shoot the sunset. I remember the day was somewhat cloudy so I hoped there would be a magnificent backdrop of yellows and oranges to shoot the friary against. It was not to be. It was a solid mass of grey and darker grey. We took a number of photos from this side of the building, but for this shot I was experimenting and completely messed up the settings. It was a 61 second exposure but should have been almost twice that. Luckily I shoot in RAW so there was plenty of data for Lightroom to work with. This short video shows how I developed and rescued the photo.
I posted the photo of Kilcrea Friary last December. Here’s what the finished photo looks like.
This video was shot at the Automattic grand meetup of 2015. This is a gathering of everyone (or nearly everyone) who works at Automattic. All attendees have to give a short talk. Some people love public speaking, but most of us don’t, so this is the hardest part of the meetup for many. The talks are called “flash talks” because they have to be completed within a certain length of time. A few years ago it was five minutes and the limit wasn’t enforced, but because the company is so much bigger now the talk must be four minutes or less, with someone holding up a sign warning when you hit the three minute mark!
This one was filmed in Park City, Utah in the United States at an altitude of around 1,000m so the air is thinner and I hoped I wouldn’t run out of breath like I did the previous year. This time I had forgotten I was due to talk until about 5 minutes before and had to run halfway across the hotel to get my notes, running down stairs, up other stairs, sprinting along corridors and then back, doing the same. Finally seated in the conference room my bottle of water spilled it’s contents on to the carpet, my stress levels were rising and then it was my turn to give a flash talk. That was practically a relief after the hectic activity of the previous few minutes!
A few days later I gave an hour long workshop on Lightroom, sharing tips, tricks and techniques that improve the look and feel of photos. That was fun as I could go more slowly and into more detail explaining the various tools in the app.
A few months ago I experimented with Imagemagick by using it to merge very similar photos of flowing water to give the impression of a longer exposure.
Here are a few more examples.
By merging a series of thirty photos taken two seconds apart I created a pleasing image that looks like a sixty second exposure. The day I took these photo was overcast and dull, but not dark enough to do an actual long exposure like that without the help of some fairly dark filters. I used the intervalometer in Magic Lantern to shoot this so I didn’t even need a remote release. My camera did all the work! This was created using the following Imagemagick command line:
convert *.jpg -average average.jpg
I combined 60 long exposure shots of the night sky in the mountains of Utah (during the Automattic Grand Meetup a few weeks ago) to create a single long exposure of the stars moving through the sky. Thanks mkaz for publishing this post on interval shooting where I got this command line:
convert *.jpg -evaluate-sequence max combined.jpg
Finally, another series of sixty shots taken in Utah. The same convert command line was used to process these.
In the image above I should have cloned out the wires in the bottom left of the image. Lightroom makes it easy to make the same modifications to every image. Work on one image, then select all the ones you need and click “Sync Settings”.
It can be frustrating taking these types of photos as your camera is shooting a long series of very ordinary shots, and the final result can’t be seen until the images are processed correctly but it’s certainly worth it.
Update on December 1st: here’s one that didn’t work out so well. It was so windy the camera shook the whole time.
|Camera||Canon EOS 6D|
At the recent Automattic Grand Meetup in Utah I presented a Lightroom tutorial. During the course of the tutorial I went through the various Develop panels explaining what (most of) the sliders did, offering some advice sometimes on how to use them.
Here’s a few tips from the night and links you’ll find useful:
When editing a photo it’s useful to begin by setting highlights to -100, shadows to +100, and adjust the whites/blacks using one of the techniques above. Pull back the highlights and shadows to suit your taste, and increase the contrast.
Serge Ramelli has a great Youtube channel. He has an over abundance of “subscribe to my newsletter” and “buy my course” notifications but his videos are still worth watching. If you want to download the RAW files he uses you’ll find them here. I enjoy his sunset tutorials and videos:
And I love what he did with this:
I think I discovered his channel first by searching for black and white tutorial videos. This one got me hooked:
Also check out Anthony Morganti’s channel. His video on sharpening and noise reduction is excellent.
Sometimes it’s useful if the shutter button doesn’t focus. Instead you press another button to focus. I’ve used it in the past when shooting street photography (focus on the ground and I know anything a metre away will be in focus), and at night on a tripod when taking long exposure shots of moving lights.
Back button focus is when you don’t use your shutter button to focus the camera. Instead you’ll use a button on the back of your camera. Not every camera can do it but check your manual or use Google to search for your camera name and “back button focus”. Here are two good videos describing why it’s good for sports photography:
Here’s a video I found to get it working on the Canon 6D using the AF-On button. I usually use the Q menu now to disable focusing on the shutter button.
The beauty of back button focusing is that you can focus your camera before the proper shoot and then take as many photographs as you want of the scene, often when the lighting has changed and focusing is impossible. This is especially important at night, but it also lets you shoot faster as the lens is already focused. This may just mean the difference between a great street photo and a missed opportunity.
Finally, here’s a photo I shot last night using back button focus to set up the shot. I could as easily have manually focused the lens but this worked just as well.
|Camera||Canon EOS 6D|
One of the horrors of owning a DSLR and using multiple lenses is the dread of getting dust on the camera sensor. In small amounts it’s visible whenever you shoot using a small aperture like f/22 but if you leave it go too long you’ll see round smudges show up on your every day snaps at f/8 or f/4 too.
This video describes part of the process I go through when I clean my camera lens. A Nikon body is used in the video but the method is similar for Canon and other manufacturers. I only have a blower and soft brush but they have served me well over the course of the last decade.
Cleaning the sensor is risky. It’s a sensitive piece of electronics but there’s an infrared filter in front of it and that’s what you’re really cleaning. It’s sturdy and tough but if you pushed too hard on it you will scratch it. I’ve never scratched the sensors of any of my cameras though so I don’t worry too much about it as long as I’m careful.
After blowing and wiping the dust away you should do the same with the inside glass of your lens. It’s likely that’s where the dust came from in the first place.
Once I’ve rebooted my camera I’ll test the sensor by doing the following:
|Camera||Canon EOS 6D|
The usual way of taking long exposure photos is by leaving the shutter open for a long time. This requires either a dark location or ND filters of some sort on the end of your lens. There is another way and that’s called image stacking.
In a nutshell, you take a series of well exposed photos of short exposure for the same duration you’d use for a long exposure shot. Then in development you create a number of images files which are then merged together to create a final “averaged” image. You can use Photoshop, GIMP, Imagemagick or I presume many other apps to create this image. I used Imagemagick here as I don’t have Photoshop and didn’t want to load all those images into GIMP and I had read that “convert” would do a good job.
Here’s one I created earlier today in the slightly shady but bright gardens of Blarney Castle:
The images that went into creating that image all look like this. They were shot at f/4, mostly shot at 1/80 of a second and ISO 100. I wanted fast images with minimal noise and the scene is mostly flat so I wasn’t too worried about depth of field.
To make this image I took 27 photos of the scene. I imported them into Lightroom (didn’t do much apart from applying my default settings) and exported Jpeg files. In the directory where I exported them I ran the convert programme from ImageMagick like this:
convert *.jpg -average average.jpg
That command was used by Luis Argerich, the author of this post who created a simulated 4 hour exposure of the sky and went on to say:
Averaging can be used in static scenes to create a new photo with less noise. Noise is reduced in the function of the square root of the number of images. So if you average 25 shots you have 5 times less noise than in a single image.
Averaging can also be used in non-static scenes to simulate a long exposure.
Patrick David in this post suggests a slightly different command which I tried and it created a similar photo.
If your exposure somehow went wildly different with some images bright, some dark, use the “Match Total Exposures” feature in Lightroom to pull them back.
ImageMagick is fairly easy to install. If you are using Linux you probably already have it installed. You can download it from this site or if you use Mac OS X, install Homebrew and then it’s as simple as this to install:
brew install imagemagick
Here are a few more examples:
This one didn’t work out. I was trying to capture the swirling of the bubbles on the water.
This one works better. The source photos are all around 1/25th of a second which is certainly not fast enough to freeze the water.
Almost happy with this one. The source photos go from 1/60th to 1/100th of a second.
I like this. It was created from ten photos shot at f/4, 1/60th sec, ISO 100.
I didn’t try very hard to make the images here pretty. They’re purely experimental and done with minimal effort but I’m very happy with most of the results. In shaded daylight I was able to take the equivalent of five to ten second exposures without ND filters or reducing the size of the aperture. It would have been easier to make the aperture smaller and add an ND filter but that’s not the point of this exercise.
One issue that may bite you is the speed of your media. My camera quite often flashed a “busy” notice because it was writing to the SD card. Then again, I was shooting 20MP images and my card says it’s rated at 30MB/second which is a problem as each image is around 20MB. As they’ll only ever be viewed online that’s probably not really needed. I might try the half-size option next time.
One nice bonus to this method is that Google Photos will generate nice animated GIFs from your image stack. I posted mine here. I won’t embed them as the files are huge!
I’m definitely trying this out again!
|Camera||Canon EOS 6D|