Apple Acquires Rapid-Fire Camera App Developer SnappyLabs

http://techcrunch.com/2014/01/04/snappylabs/

by nirvanatikku • 13 years ago

220 89 13 years ago

xenonite• 13 years ago

It is amazing that it is still possible to optimize the compression algorithm for JPEG, a file format that is already over 20 years old.

fiatmoney• 13 years ago

It sounds like he didn't optimize the algorithm per se; he optimized the implementation of the algorithm for the particular processor. Still very impressive.

vonseel2• 13 years ago

"Having developed a blazing fast DCT implementation, Huffman then became a bottleneck. We innovated on that portion with tight hand-tuned assembly code that leverages special features of the ARM processor instruction set to make it as fast as possible.”

I understood he first optimized the algorithm, then tuned it to be even faster on ARM?

zafka• 13 years ago

I was amazed the first several times I saw things i thought where old as dirt have amazing improvements made. One of the first was about 10 years ago, when a friend made improvements on lighting ballasts in conjunction with solar work. Now I try to keep my eyes open for these opportunities.

jmspring• 13 years ago

The actual standard allows for a variety of options around how the compression portion is handled -- there is the baseline mode and then options for custom huffman tables, etc. The best source for the JPEG Standard --

http://www.amazon.com/JPEG-Compression-Standard-Multimedia-S...

My memory is a bit fuzzy here, but from what I recall, a friend tried implementing JPEG compression on a parallel system (back in the mid 90s). One of the trouble spots was the reliance on the relationship between blocks regarding prediction of the DC coefficient.

One startup I worked at, we had an entropy encoding method that could add 20-30% additional compression just by recording coefficients. Patent here -- http://www.google.com/patents/US6236762

raverbashing• 13 years ago

I'm not so sure about it, but lossy compression algorithms are usually "non deterministic" and what I mean by that is that there is a target format to encode to, but you have some variation in how you're turning the original data into that

So there's a lot of room in optimising the process

sergiotapia• 13 years ago

Major congratulations to him! What an achievement!

I wonder though: in these situations if he were to refuse the acquisition would Apple just steal his idea (NOT implementation) and bake it into the iPhone regardless? I think Microsoft did something like this, no? Any thoughts?

jakobe• 13 years ago

I don't think that this is something he would need to worry about. Implementing complex algorithms is extremely difficult, even if you have detailed instructions (eg. from a published paper). I've read his website before and if I remember, he only published high level details. To steal his ideas, you'd need to have a similar background, and you'd need to know just as much about compression etc. as he does. And even if you found someone with the right expertise, it would still take a very long time to "steal" an idea like this.

michaelt2• 13 years ago

When you're coming up with a new idea, you often have to go down a bunch of dead-end alleys and you're not even sure if what you're attempting is possible. When you're reimplementing an existing product you know exactly the right avenue to go down, and you know for sure it's possible.

And Apple has over a hundred billion dollars in cash and eighty thousand employees. They could certainly clone this software if they wanted to.

The reason to acquire is to signal to other people that doing innovative things on apple platforms can make you a millionaire, instead of signalling the same thing is a sucker's game as apple will steal your ideas if they're any good.

qntmfred2• 13 years ago

you think Microsoft did this, huh? show me a large tech company that hasn't done this.

huxley2• 13 years ago

You're probably thinking of Stac Electronics v Microsoft Corporation:

http://en.wikipedia.org/wiki/Stac_Electronics#Microsoft_laws...

miles• 13 years ago

would Apple just steal his idea (NOT implementation) and bake it into the iPhone regardless? I think Microsoft did something like this

Apple has certainly done this before:

Apple Literally Stole My Thunder https://medium.com/wwdc-round-up/253aed27a455

Apple rips off student's rejected iPhone app - iOS 5 lifts idea, name, even logo http://www.theregister.co.uk/2011/06/08/apple_copies_rejecte...

Konfabulator, Dashboard controversy flows out of WWDC http://www.macworld.com/article/1035200/konfabulator.html

Apple stole Karelia Watson http://forums.macworld.com/index.php?/topic/2477-apples-stol...

threeseed• 13 years ago

Three of those four are pretty poor examples.

It was a different time when the images/weather app was rejected by Apple. They were deliberately trying to prevent people creating thin wrappers around websites and calling it an app. And at the start it made a lot of sense.

Syncing over WiFi was a feature that users had demanded long before that app had been around and the name/logo are not distinctive. They are the most common sense choices you would use.

And widgets existed decades before Konfabulator thought to bring them to OSX. They were available as Desk Accessories in the previous versions of Mac OS as well as on Windows/Linux.

ceejayoz2• 13 years ago

Apple Literally Stole My Thunder https://medium.com/wwdc-round-up/253aed27a455

The article states "I’m not naive enough to claim Apple actually took my idea." You apparently are, though.

Apple rips off student's rejected iPhone app - iOS 5 lifts idea, name, even logo

C'mon. "Wi-Fi Sync" is hardly a surprising name for such a feature, and making its logo the bog-standard Wifi logo with a bog-standard Sync logo superimposed is hardly a surprising choice either. It was a long-expected and long-awaited feature.

Konfabulator, Dashboard controversy flows out of WWDC

A controversy about the bog-standard term "widget".

Apple stole Karelia Watson

Well, it made its own version and named it cheekily. This seems to be the one mostly valid case in your post.

salient• 13 years ago

The iBooks interface, too:

http://techcrunch.com/2010/01/27/think-ibooks-looks-familiar...

devinfoley• 13 years ago

Ironically, the "Stole My Thunder" guy ripped off that design/animation from Solar: http://thisissolar.com/

lostlogin3• 13 years ago

I believe Instapaper belongs on that list too.

nedwin• 13 years ago

This risk is always in every conversation like this.

Look at SnapChat turning down $3bn from Facebook and now FB integrating a lot of that functionality into Instagram.

muglug2• 13 years ago

Yeah, but that's a slightly different scenario – they a lot of developers on staff already know how to implement private message functionality, and they have better expertise on staff than SnapChat do for the difficult part - getting the UX right. Here, just one guy was doing this a very sophisticated thing that was clearly out of the reach of Apple's developers (SnappyCam came out long before iOS 7, but the new in-house iPhone burst functionality doesn't work half as well as SnappyCam's).

DenisM• 13 years ago

Apple has earlier ripped off the Swiss Railway watch face, and also pilfered the idea of using volume button as a shutter for the camera. They apologized for the watch face, IIRC.

smackfu3• 13 years ago

For the watch face, they were sued and settled for $21 million.

31reasons• 13 years ago

>would Apple just steal his idea (NOT implementation)

I think the App was more about the implementation than the idea. He created extremely efficient implementation of jpeg compression with 50000 lines of C++ code as far as I remember from the dev's blog. So Apple is obviously more interested in the guy than anything else.

philwelch• 13 years ago

As far as I can tell, Apple's acquisitions are fairly standard build-vs-buy decisions. Lala had the infrastructure for what became iTunes Match and the price was right, so they bought them. Siri was a fairly complete piece of technology that solved some hard technical problems, so they bought them. Konfabulator was a technically trivial application that Apple wanted to rewrite in a totally different way, so they built Dashboard.

mandeepj• 13 years ago

In terms of stealing the idea - how can you forget Xerox? Although there was no acquisition involved.

interpol_p5• 13 years ago

That was licensed from Xerox.

applecore• 13 years ago

This is great for all iPhone users, and a much better outcome than had a company like Facebook or Snapchat acquired them.

buro9• 13 years ago

Why?

If Facebook acquired this then the technology would be accessible to a far greater audience than just the users of Apple products.

Seems to me that we should want advances in technology (software or hardware) to benefit the greatest number of users possible, and with Apple owning this it now restricts this method to only those who use Apple products.

Oh, just noticed your username.

Edit: Being downvoted? Do I have to do the defensive thing of pointing out I own Apple products too?

jhales• 13 years ago

Agree with your general thesis, but it seems that the tech is for the iPhone specifically, not cameras in general.

manmal• 13 years ago

You seem to imply that this kind of performance tuning would carry over to Android, via a Java/NDK bridge? I sincerely doubt that this would yield results as good as SnappyCam.

erichocean• 13 years ago

Simple: Facebook app w/tech on two platforms << All Apple apps and users w/tech on one platform.

If Google or Samsung thinks it matters to their users, they've got more than enough $$$ to fund something for Android specifically.

raganwald• 13 years ago

If Facebook acquired this then the technology would be accessible to a far greater audience than just the users of Apple products.

I'm unclear why we are arguing whether this technology is somehow better off in the hands of one monopolistic, predatory giant instead of the other monopolistic, predatory giant.

The way to make this technology truly accessible to a far greater audience is to make it open. Everything else is quibbling over which King has the softer boot on the neck :-)

nl• 13 years ago

In the previous post about SnappyLabs the author said he wasn't looking at doing an android version at the moment, and gave the impression that there were technical reasons why it was more difficult.

makomk• 13 years ago

Why? The app's been discontinued, and based on previous Apple acquisitions the integrated version will probably only be available for the iPhone 6 even if there's no technical reason to limit it in that way.

lostlogin3• 13 years ago

Partly true - although ios7 on the iphone 4 hasn't won Apple many fans.

kalleboo• 13 years ago

This is great for all iPhone users

Except for those with a phone older than the 5S, who now can't use this feature anymore because Apple wants to sell them a new phone.

themartorana• 13 years ago

Good for him! The app is a bit of technical marvel. I love playing with it, and now that Apple has it, hopefully the UX will improve a bit in whatever way they decide to implement the core tech. Wins all around!

vonseel2• 13 years ago

This is one of those stories that makes me so happy. It's nice to see someone's hard work pay off.

ShaneOG• 13 years ago

Huge congratulations in order. This sounds like a real success story!

sabalaba• 13 years ago

For those interested in the technical details, the original blog post is archived here:

https://web.archive.org/web/20131010012005/http://www.snappy...

I wonder why they took it down.

wmf2• 13 years ago

Apple takes down everything in an acquisition. In general Apple rarely admits that the past or future exists at all.

proksoup• 13 years ago

Apple is a strange mix of buddhism and 1984.

Middle path, present focused, to an extreme, enforced on you whether you want it or not :)

asveikau• 13 years ago

That reminds me of something I read once on Apple's website one day, that seemed to suggest Apple invented CUPS in 1997. They hired the developer and bought the rights in 2007, so it's fair to say it's theirs, but it was worded as if this was always so.

runewell2• 13 years ago

Cool, hopefully they open source this code sooner rather than later. Obviously they are probably going to keep it as a competitive advantage for the next batch of iOS devices but I hope they open it up shortly after that.

qntmfred2• 13 years ago

what took so long?

chenster2• 13 years ago

When I first heard about the SnappyCam that is capable of taking blazing fast photos using a regular photo cam, I immediately purchased a copy and experimented with it. Not only it doesn't everything it claims (essentially a breakthrough in super optimized image compression) but also the UI is well thought. At the time, I was certain that it's just a matter of time to get Apple's attention and very likely Apple would want to acquire the technology. Kudo to Papandriopoulos (How do you exactly pronounce that??)

What I'd like to add that, it apparently SnappyCam would only make sense on iPhone 5 or lower. That is because iPhone 5S already can do what SnappyCam does in its native Cam app by holding down the shoot button. Perhaps Apple got other ideas for its use in the future.

hobs• 13 years ago

I think you may want to do a wee bit of editing, because I assume you meant: "Not only it doesn't everything it claims" -> "Not only does it do everything it claims"

"but also the UI is well thought" -> "but also the UI is well thought out"

pap andria polous

chenster2• 13 years ago

Wish I could, but I don't see the edit button. :(

wyager• 13 years ago

That is because iPhone 5S already can do what SnappyCam does in its native Cam app by holding down the shoot button.

Wow! I had no idea! That's very cool. I was also very impressed with the 5S's slo-mo capture. It's exceptional quality for 120fps.

krashidov• 13 years ago

Can somebody remind me if this was originally a Show HN post? I think I vaguely remember the app creator talking about it.

citruspi• 13 years ago

There was this[0]. The author of SnappyCam made some comments under the username jpap.

[0] https://news.ycombinator.com/item?id=6137979

scanr• 13 years ago

Excellent prediction by Oculus:

I have a feeling that soon SnappyLabs is going to have Apple knocking on their door with a very nice offer.

https://news.ycombinator.com/item?id=6138278

lelandbatey• 13 years ago

I don't remember a Show HN post, but I do remember it being a popular post and the title mentioning "10,000 lines of hand tuned assembly".

runewell2• 13 years ago

Will this tech improve live video streams?

wmf2• 13 years ago

Not directly, but this guy could probably do some good work if you let him loose on a video codec.

ececconi• 13 years ago

That guys work is amazing. I saw snappycam before and didn't bother downloading it or looking into it. Now I'm reading up about it and it seems absolutely amazing. I can't wait to see how this technology is implemented in future iPhones.

semerda• 13 years ago

Aussie Aussie Aussie, Oi Oi Oi!!! Congrats to John Papandriopoulos! Very impressive work.. also the amount of work that went into it is wow!! .. "The final implementation comprises nearly 10,000 lines of hand-tuned assembly code, and over 20,000 lines of low-level C code. (In comparison, the SnappyCam app comprises almost 50,000 lines of Objective C code.)"

gdubs• 13 years ago

Inspiring story on a bunch of levels, but in particular it's great to see the payoff of focus directed at a hard problem.

interpol_p5• 13 years ago

Why is a buyout like this reason for celebration and congratulations? I understand that the developers received money; but now they no longer own their IP, and can no longer decide the future path for their software.

I say this as someone who was contacted by a VC firm recently. They were interested in investing, but their assumption was that we wanted to grow and "exit". When I asked for clarification on what "exit" meant, I was told that it means to sell the company. I was confused as to why someone would want to sell something they spent so many years creating.

(That said, I think in this case their tech will genuinely improve Apple's software. And maybe they get to continue doing passionate work at Apple. It just feels like they are no longer in control.)

dippyskoodlez2• 13 years ago

He may be more interested in furthering his career not by making large piles of money, but by being an engineer working on fantastic new products.

I would certainly call that congratulatory worthy.

yapcguy• 13 years ago

The open source community loses too.

No doubt Apple put a very lucrative offer on the table, but another route could have been for the author to dual-license the code - GPL it so the community benefits, while any improvements he can then push out in his app which he continues to sell.

ZeroGd• 13 years ago

Great for the guy, but this screws customers of his app. SnappyCam made animated GIFs of your snaps. I highly doubt Apple's native camera app will ever do that. Now that the app is gone, this royally sucks and is a fat middle finger to IMGUR and Reddit's communities that produced animated GIFs using SnappyCam.

oh_sigh• 13 years ago

Then write your own version of it. You have no right to demand an app stay for sale when the developer doesn't want to sell it.

alok-g• 13 years ago

I am confused. Is it that the transfer speed was limited by the compression algorithm rather than the speed at which the images can be retrieved from the photo sensor array? I had assumed it is the latter.

ghshephard3• 13 years ago

If we can believe that jpap told us the entire story, It's the former. He never mentioned modifying, in any way, the mechanism by which images were retrieved from the photo sensor array, just how he compressed and saved them.

https://news.ycombinator.com/item?id=7010474

smackfu3• 13 years ago

Since Apple already has this as an advertised feature in the 5S, I see this as one part acquihire, one part locking up tech so it can't be used on other platforms.

muglug2• 13 years ago

I wonder how this ties into Apple's other recent acquisitions - if I remember correctly, they recently purchased a company that was mounting depth sensors on iPads. Fast image processing is central to usable 3d augmented reality interfaces, and Apple might be planning ahead for 2015's iOS innovations.

pervycreeper• 13 years ago

I wonder if any of the money geniuses on here would care to speculate on what a deal like this would be worth.

abozzelli• 13 years ago

Shameless spam: my app push the limit in the opposite direction, I hope Apple acquires it too. http://www.alfonsobozzelli.com/thomas/

Apple Acquires Rapid-Fire Camera App Developer SnappyLabs

Related Stories