Sidebar Sidebar
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

Apple Acquires SnappyLabs, Maker of SnappyCam Burst Photo App [Confirmed]

djgamble said:
Those who use iOS 7 have known for ~6 months that Apple has ALREADY implemented the code into their camera app. My iPhone 4S magically takes 20 photos per second (trust me, you feel it as soon as you upgrade to iOS 7 and try taking a photo). There's no 'saving' wait time after taking multiple photos. Up until this acquisition was known about, this was just iOS 7 'magic' that nobody said much about. We just knew that iOS 7 significantly spend up the camera app...

This explains how Apple achieved these speed gains. They bought out a guy who made a clever new algorithm and used his algorithm in their own app.

Grab your brand new quad-core Android and compare the speed of taking burst photos to a 3 year old iPhone 4S which has been upgraded to iOS 7 (for free, promptly... rather than when/if the carrier allows the upgrade like some competing distribution platforms). You'll notice that the 4S is now faster at taking full-res burst photos...

Now THAT is customer friendly!! Free seamless upgrade (even for old phones) 6 months before people even knew that Apple had purchased the company.
Click to expand...

Sorry, but I'm afraid you're absolutely wrong.

Albeit I don't have the 4S to test it, I seriously doubt it would have received this tech, while the iPhone 5 wouldn't. I've just tested action shooting under the latest iOS version, 7.1b2 (I'm a reg'ed dev so it's a legal installation) with my standard 60 fps counter video I've developed to easily benchmark video players and cameras.

This means a camera that shoots 60 images (frames) a second would not have any dropped (missing) frames. With a camera recording 30 fps, about every second would be missing and so on.

Now, this is what the stock Camera is capable of (with disabled action sound):

ip5actionshootStockCam.jpg


As you can see, the minimal difference between the counters is 17 (see the 313-330 transition), and the maximal is 24 (see e.g. 386-402). This means the actual fps is between 60/17 = 3.52 and 60/24 = 2.5. The average is around 3. Way less than the 20 you stated...

Feel free to prove I'm wrong. Just post the same Photos screen, preferably when 4 digits are displayed, so that we can easily see the actual framerate. Here's the 60 fps test video:

https://dl.dropbox.com/u/81986513/082012/sublerSubs/1080p60-counter -optimized.mp4

Make sure you play it back with maximized brightness so that the low light conditions don't force the iPhone to lengthen the shutter time and/or autofocus time. (Both are generally (much) higher under low light.)

EDIT: for comparison, here is the export of SnappyCam on exactly the same iPhone 5 running the exact same 7.1b2, using the default settings minus geolocation tagging & sound effects:

snappy-settings.jpeg


The results:

snappy-1.jpeg



snappy-2.jpeg


As you can see, most of the time (except for somewhat longer pauses now and then), SnappyCam indeed could record every third frame in the video, resulting in the net framerate of 60/3 = 20 fps. Which is almost an order of magnitude higher than the stock camera app's framerate measured in my previous comment.

All in all, what you've written is wrong. Apple did not add any kind of speedup to the 4S - and it most surely won't. They won't make an "old" model considerably better.
 
Last edited:
Menneisyys2 said:
All in all, what you've written is wrong. Apple did not add any kind of speedup to the 4S - and it most surely won't. They won't make an "old" model considerably better.
Click to expand...

But they did... I don't think you understand what the technology does. Read previous posts in this thread.

Nobody is paying you Androids to spit out BS on Mac forums. Why bother?
 
djgamble said:
But they did... I don't think you understand what the technology does. Read previous posts in this thread.
Click to expand...

Please provide proof of the allegedly enhanced framerate as I've did. A simple screenshot will do. Until then, I can only say you haven't properly measured the actual framerate of the stock Camera app.

Nobody is paying you Androids to spit out BS on Mac forums. Why bother?
Click to expand...

If you read my other posts here at MR (I have thousands of them, most of them helping other people and/or presenting my new iOS apps or JB tweaks), you can see I'm not an Android user. I prefer (JB'n) iOS. However, I'm absolutely disillusioned by Apple's upgrade policy, compared to that of Google or MS / Nokia.
 
djgamble said:
lol quit talking to yourself... seriously why make such an arrogant statement then prove nothing?
Click to expand...

Lolz... was it all? Where's the proof I've asked for?

Look, creating and posting proof only takes some minutes.

1, you download my 60 fps test video and play it with maximal brightness
2, shoot it with the shutter icon depressed
3, post a Photos shot of the recorded frames where the last number is visible.

I don't ask for much.
 
Last edited by a moderator:
Requires iPhone 5S

cRuNcHiE said:
I made a post about this on Monday here https://forums.macrumors.com/threads/1689860/


I've since been able to try out the app and actually, while it's impressive the speed it can take photos.
The actual resulting photos are not very good quality , more like very high ISO photos with lots of noise.

When I compared a burst photo from apples camera app with the exact same scene with this app, this app produced far more noise.
Click to expand...

dbrewer80221 said:
Does anyone else find that the current burst mode sucks? All my photos come out ghosted or blurry.
Click to expand...
You guys both reporting your experience with an iPhone 5S?
 
djgamble said:
lol quit talking to yourself... seriously why make such an arrogant statement then prove nothing?
Click to expand...

Actually my money is on Menneisyys2. I just tried on my 5S with built-in camera app and it only manages 10fps.

I very much doubt that my 2-year-old 4s is twice as fast as that. (and I'll have to dig it out of retirement and charge it to check - which I'm not about to do.)

IMG_0794.PNG
 
Last edited by a moderator:
sigsegv said:
Actually my money is on Menneisyys2. I just tried on my 5S with built-in camera app and it only manages 10fps.

I very much doubt that my 2-year-old 4s is twice as fast as that. (and I'll have to dig it out of retirement and charge it to check - which I'm not about to do.)
Click to expand...

Thanks for the test!
 
MacRumors said:
UpdateUpdate: Apple has confirmed the acquisition to Kara Swisher at Re/code, the new home of the AllThingsD team.
Click to expand...
Thank you. For once you understand what confirmed (within the Apple news/rumour world) means. Please continue only using the word confirmed (regarding Apple directly) only when Apple speaks. MacRumors has gotten this wrong so so many times in the past, but here MacRumors got it right.

Whether Apple actually confirmed this to Re/code or not is anyone's best guess. Lots of people incorrectly use the word confirmed. But that is Re/code's issue and not MacRumor's.

Hopefully MacRumors will keep up this good and correct write style.
 
Am I the only one who actually read the entire SnappyLabs blog post?

This paragraph is very interesting:

"Similar innovations were put into a custom JPEG decoder, powering the unique SnappyCam thumb-to-interact living photo viewer. When dealing with massive 8 Mpx (32 MByte BGRX uncompressed) images, decoder performance became critical to a great user experience."

In simpler terms, this algorithm not only enables saving photos faster, but displaying them faster. That's something that can contribute to a better user experience for everyone, not just those who shoot in burst mode.

Forest for the trees, gang...the fact that this algorithm works both ways is what makes it truly valuable to Apple. As the number of megapixels we're carrying on our phones increases, the difficulty of providing speedy photo browsing and editing increases. This technology will help Apple keep pace with that and continue to provide a (dare I say) snappy user experience. :D
 
msephton said:
I disagree that there are visible differences. So would JPEGmini as they've scientifically proved it whilst creating their software. It's not Save For Web they're doing, it's something much more intelligent.

See an image of the differences between to photos here: http://apertureexpert.squarespace.c...-be-used-to-reduce-your-aperture-library.html

Hint: it's the totally black image
Click to expand...
I looked at the pictures, there were pretty obvious differences. The "totally black image" is of course not really totally black. You pretty much get the same result with any Picture saved in two different compression rates. It's a cheap trick and says nothing about the quality of compression.
And I don't say there is no improvement. I just say they're not impressive at all. Here's an comparison picture of the examples on the website, btw. Don't tell me there is no difference. The JPEGmini version is MUCH blockier.
 

Attachments

  • jpeg-jpegmini-comp.png
    jpeg-jpegmini-comp.png
    103.8 KB · Views: 105
djgamble said:
lol quit talking to yourself... seriously why make such an arrogant statement then prove nothing?
Click to expand...

He proved everything, whereas you make insults with no facts and no guts to test it out.

But no need as I've just tested it on my 4s (numbers going backwards)
2862 2836 2812 2787 2767 2745 2718 2698 2659

Differences 26 24 25 20 etc

That's about 3 frames a second at best
 
Last edited by a moderator:
Tumbleweed666 said:
He proved everything, whereas you make insults with no facts and no guts to test it out.

But no need as I've just tested it on my 4s (numbers going backwards)
2862 2836 2812 2787 2767 2745 2718 2698 2659

Differences 26 24 25 20 etc

That's about 3 frames a second at best
Click to expand...

Thanks for the figures! Indeed what I've expected.
 
Last edited by a moderator:
SgtPepper12 said:
I looked at the pictures, there were pretty obvious differences. The "totally black image" is of course not really totally black. You pretty much get the same result with any Picture saved in two different compression rates. It's a cheap trick and says nothing about the quality of compression.
And I don't say there is no improvement. I just say they're not impressive at all. Here's an comparison picture of the examples on the website, btw. Don't tell me there is no difference. The JPEGmini version is MUCH blockier.
Click to expand...
You're going to have to highlight the differences because I don't see them. Sorry

http://www.jpegmini.com/main/technology
 
Guys,

if you've missed SnappyCam but would still welcome a hack that makes shooting significantly faster and don't mind jailbreaking, there's a fairly new and free tweak "Burst mode" that raises the shooting framerate on "old" models like the iPhone 5 by about 32%. Definitely not as much as SnappyCam, though.

I've posted a detailed review of it, along with a completely new and absolutely great hack, "Slo-mo Mod”, at https://forums.macrumors.com/threads/1698879/
 
You must log in or register to reply here.
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.
Top Bottom
Back