Yesterday I released my first set of wallpapers, and the response has been quite gratifying. I want to say thanks to those of you who’ve been downloading them and passing along the new images.
But that said, the release of the images was about a week later than I’d originally hoped, and to be honest, I’m not entirely happy with the wallpapers for the IOS devices, especially the tablet. Apple’s made generating wallpapers for IOS 7 a fair pain in the parallax.
The way they implement the new parallax effect on their backgrounds is to scale an image to a larger size, and then as you swing your device up or down or side to side, shift the image around the screen a bit. It is, in a way, a mini “ken burns” effect.
If you do what Apple’s done and build abstract backgrounds, that’s fine. There’s a long tradition of subject-oriented wallpapers, though, and the way Apple’s built things in IOS7 ignores that use case completely, and creates a big case of heartburn for people trying to build it.
As far as I can tell (I’m still experimenting), the image they use for the backgrounds is extended about 200 pixels off each edge of the screen. That means a 2048×1536 image you might have used on an IOS6 device is stretched out to 2448×1936, and then the edges are all stuff out of view. No tall of the image hidden off screen becomes visible when you activate the parallax. My estimate is about 100px is lost around the edge completely.
If you are a photographer building wallpapers out of your images, and one that’s trying to take some care about image quality, just having your images stretched by some unknown algorithm before display is going to cause you to reach for the Maalox. But it gets worse.
Those lost edges can be significant to the image. If you look at this image on your computer screen, and then install it on the IOS device and look at it, suddenly a non-trivial part of the image is gone.
The framing and composition that a photographer probably did in creating that image is damaged. It makes a number of images completely unusable as wallpapers.
I spent a few days experimenting with ways around this, and with work-arounds, and finally decided to release it and explore options later. Right now, three options seem possible:
First, release the image as is and not worry about it. As someone who takes some pride in my images and compositions, I’m frankly not happy with what this change in IOS7 is doing to my images. I would likely just not release for the IOS devices and only for desktops instead (which, since I’m doing this primarily because I want them, I find terribly annoying; that others can enjoy them as well is a bonus).
The second option: wrap everything in a simple frame that fills that extra space, so the original composition shows up on screen and the hidden areas are filled with non-image pixels. So far, my experimentations have shown this is more difficult than it sounds, because the parallax effect makes it hard to find a set of numbers that remove or limit how much image is lost to this. Either that, or you see the frame peeking out at various times as you play with the device, which I also find looks sloppy and unprofessional.
Third option: simply recrop the image to a wider composition and use those “extra bits” that were cropped out during processing as that framing area. That, of course, assumes those bits existed and were cropped off during processing. Not a safe assumption in most cases.
What really makes this painful is that it breaks any wallpapers built and formatted for IOS6 or before. It also implies heavily a need to process and publish separately for IOS 7 and for IOS6 and non-IOS devices. You end up, if you want to do this right, needing to consider building out two or three versions of an image for different devices and releases, and then having to live with the complexity of trying to explain to non-technical types which image to use in what circumstances.
This is a big pain in the parallax. And Apple broke an existing setup and a known way to produce wallpapers because as far as I can tell, they only considered their own use case, and not the existing use cases for their devices by their users. This si the kind of decision we see out of Apple at times that makes my forehead get all scrunched up and wrinkled, because it’s poor product management. there’s no reason they had to break the existing setups to create this parallax effect for their cute backgrounds.
So count me annoyed because Apple and the product managers didn’t think through the use cases, or care about them, and built something that worked for a really neat demo, but didn’t take into consideration existing uses and workflows.
The sound of silence you hear in the distance is how much Apple cares what I think about this…
No, this isn’t a “damn, things have gone to hell since Steve died” grump; this kind of “only thinking about what matters to Apple” product design has been creeping into Apple’s products for years. Ask your favorite friendly Mac or IOS dev about some of the sandboxing Apple sprung on them, for instance.
Users of my wallpapers should stay tuned for updates if and when I decide how I want to work around this. I’m still leaning towards the simple frame, but I’m not completely convinced and I’m exploring what other options exist I might haven’t of thought of (I’m open to suggestions, big bad internet). Or maybe Apple could realize they’ve broken this and add functionality to support existing wallpapers in a non-parallax mode (this would be technically quite tough, I realize: like, say, recognizing a specific EXIF field in the image, or using a special suffix on the file name… it’d take a decent engineer a couple of hours to implement and a couple of days to build a decent UI onto it…)
My recommendation on my images: if you like them on IOS, use them. If you see one you think looks bad, let me know. I’ll see what I can do. And if you’re Apple, stop being so enthusiastic about blowing away existing functionality to implement stuff that gives great demo. Especially when a bit of thought would allow the system to support both rather easily…
I know a lot of pundits love to worry about Apple “Post-Steve”. I don’t, because I like Tim Cook and the executive team he has. But I will admit that the quality of product management on Apple products has increasingly worried me, because the company doesn’t seem to understand the users as well as they need to, and they increasingly show they aren’t understanding how their products are used, or don’t really care. Almost as if they’re starting to design products for the big release demo more than the day to day usage. And that’s not a good trend… And it’s not in the big features, it’s here in the details of the product, things like how wallpapers work and what breaks when you change how you display them.
That’s a minor thing in itself, but it’s a minor thing that can be really annoying to a group of users when you break it, and as that set of minor things that annoy users starts to pile up, that’s how you erode your user loyalty. Apple used to sweat these kind of details. Now they seem more interested in flashy demos…
Update: Macdailynews picked up on this and published a pointer to it, so thanks to them for doing that. Their response is Chuq’s all worried about “framing and composition” of images that, he seems to forget, as wallpaper will have text and/or icons splattered all over them which is both correct and misses the points, one of which is that if you’re a creative the quality of the thing you create matters, and the other is that this was something that worked perfectly in IOS6 and before and was broken for no good reason. If there’s a reason for the rant, it’s the latter, that this is a symptom of Apple’s product management problems. That’s something Macdailynews ignored completely looking for an easy response.
More amusing in its way is the comment stream on their article. One member wrote all of this off with a trivial “first world problem” — someone want to think about anything involving owning an IOS device that isn’t?
Still, it’s nice to see people talking about the issue, and maybe at some point about the larger issue that this is a symptom of.