Actual OIS performance: Lumia 920 against iPhone 5, Galaxy S III, HTC One X | Hacker News Reader

Actual OIS performance: Lumia 920 against iPhone 5, Galaxy S III, HTC One X

engadget.com

54 pointsAnchor54 comments

sigh So the OIS actually does work as well as claimed, which boggles the mind why they felt they had to fake it in the ad.

It probably wasn't ready to run on an actual phone at the time.

Unless you're comparing it directly with another phone, it's hard to imagine any phone image stabilization looking good in a commercial.

Because the phone is not ready yet. btw it's a test controlled by Nokia. I think we should wait until it's finally released before giving the final judgement.

So maybe The Verge's whole freakout over it was completely unwarranted? Maybe Nokia knew what the effect of OIS would be when they made the ad -- after having done the R&D to develop the camera in the first place?

I disagree: I think it was somewhat warranted. The fact the end-result looks similar to the faked result is beside the point.

Many countries, the US included, have decided that if you cannot fake product demos in advertising unless it is clear that's what's going on. Hence the "Sequences Shortened" text on the iPhone ads.

It is rather unethical to produce a marketing video that fakes functionality and doesn't make it clear that's what's going on. You may well not be over-exaggerating anything, but personally I don't think that's relevant.

I would argue that the "Sequences Shorted" text doesn't make anything clear at all. It's a barely noticeable cop-out.

Why would one express sorrow for a huge innovation in technology?

I doubt I'd call it faked. Would you call siri ads faked? Everybody who has used it can vouch that it just doesn't work let alone in noisy downtown.

It was a promo video used when revealing a product (not an ad broadcast on TV) that was trying to prove a point that OIS is important and can make pictures and videos look beautiful. The verge and all other sites that did the detective work was purely for the sake of page views. No substance there...

Yes, I would call that Siri ad faked.

If we're honest, this is hardly surprising. Before Nokia made it's commitment to Microsoft, they were a company at the forefront of phone technology, including their cameras. However, what this does actually do is beg the question of why they opted for Microsoft. I'm without a doubt in my mind that if they were making Androids right now, they'd be dominating.

I know this is going to be controversial.. but they actually made the right move. MS is supporting them and without that help, Samsung and HTC wouldn't give them a chance. Also, in my particular point of view, WP is way more aligned to Nokia's style than Android.

Honestly, as an user of the four OS (iOS, Android, WP and BB) my favourite is by far windows phone. Not only it feels different but also I found it way more stable. Everything is subjective anyway

However, what this does actually do is beg the question of why they opted for Microsoft.

1) It differentiates them from the sea of Android phones.

2) Nokia's CEO, Stephen Elop, was a former head of Microsoft's business division. So there was already a (presumably) favorable relationship in place.

I'm not sure about that. The Android market is the mobile-cost leader market, and has already experienced the race to the bottom.

To be profitable in the low cost game you need to have market share. Being a late comer to the Android game, Nokia would be hard pressed to make a profit with Android phones.

Nokia instead is gambling with MS to capture the "buy-in" market using Microsoft's enterprise. This market was previously known as Blackberry's.

I remember when HTC and Samsung were starting to make some really good Android phones around 2009-2010, and I was thinking that there was no doubt in my mind that if Nokia adopted Android tomorrow, they would easily become the #1 Android manufacturer. Really missed opportunity there. Back then they were still too "proud" to adopt "someone else's operating system"...which now reminds me of RIM.

Apparently, they approached Google, but they failed to provide the same incentives and special treatment to tide over the tough times that Microsoft did(Nokia gets at least a billion a year from MS for platform support).

Many more details here: http://www.businessweek.com/magazine/content/11_24/b42320567...

Regardless of zoom levels or other disadvantages, this is a honest comparison of true results you would get from each phone. It's a moot point to argue technicals you have no control over.

Overall, the 920 just seems way clearer, and objects have much sharper edges.

I noticed that too - the colors seemed deeper and richer as well. The colors on the iphone seemed a bit washed out.

Just another reason I will most likely get a WP when I'm due for another upgrade this winter.

Others have mentioned the different "zoom" levels already. Please correct me if I am wrong but doesn't the zoom level have a direct impact on the differential caused by movement of the camera?

Imagine the following: You have two microscopes mounted on a similar device like seen in the video. The first microscope shows everything two times larger than the second one. If you move "shake" both microscopes the same way you will see more "shakes" when looking through the first one. Isn't that correct and can't this be applied to what is shown in the video?

Engadget has been suspiciously friendly to Nokia lately (I haven't seen them take so many interviews with so many employees from one company before), but I still find it hard to believe that they would increase the zoom level themselves for the other smartphones to make them look worse on purpose. Surely there has to be another explanation for why all the other smartphones look zoomed in.

re: engadget being suspiciously friendly to Nokia.. i think that if you're willing to actually send a reporter to Espoo, Finland (of all places), Nokia would likely be happy to provide access. especially so, given what now appears to have been a significantly overblown reaction to their introduction of the technology (as the actual usage results appear to validate what was demoed, even if not shot with the Lumia at the time)..

and the Lumia 920 camera apparently has a narrower focal length (which results in a wider field of view), so it can capture more on-screen than than the iPhone; hence it looking particularly 'zoomed-in' when compared side-by-side.

additionally, and i'm not entirely sure about this, but the digital image stabilization software (in contrast to Nokia's actual/mechanical OIS) found on the other phones may have some cropping involved to help smooth-out shaking which would be typically most noticible around the edges..

Depending where the camera is located on each phone, that foam holder will "amplify" movement on the outer sides. I hope they put the phones with the cameras near to the center inside it and switched them too. Otherwise this might be very wrong.

I hope they positioned the phone how someone would actually hold it when they were taking the video. For example, on the iPhone I'll always hold the phone with the camera at the top (portrait mode) or the camera to the right (landscape mode) since the volume buttons will then be up.

I don't understand why Nokia doesn't just make a commercial instead of some home movie blog post? Their tech is better so advertise that; Part of Apples success is controlling the 5W's of presenting their information to customers. Another thing is that if engineers with DSLR's are Nokias marketing department, they are in big trouble.

They did but they did it with a professional camera and implied it's the Lumia 920, when it wasn't. I'd be fine with them doing a commercial but actually using the device to show its capabilities. Filming with a much more advanced camera and then saying "that's how Lumia 920 does it, too" is just false advertising, and that's what they did initially.

so you are convinced that Apple used siri in all those commercials they show?

Because the phones are going to be launching in about a month. There is no point airing commercials now.

Yikes, that is quite a large phone. Lumias look like great phones but must they be so large?

It is large. It's thicker than my HTC Titan, but I suspect it will feel really good in your hand. As it is now, the only surface contact I have when holding my phone are the edges.

I wouldn't mind giving up a little thickness for comfort holding the device. I think in the race to make things thinner we might have gotten carried away. Imagine a phone as thin as box board...now imaging trying to hold it.

920 has rounded edges that make it comfortable to hold. the sharp (jewel like!!) edges of iphone make it uncomfortable. Lumia 900 is very similar in how it feels in hand. Give it a shot and decide for yourself.

Not really: http://www.youtube.com/watch?feature=player_embedded&v=m...

Engadget has also done a low-light comparison today: http://www.engadget.com/2012/09/26/lumia-920-low-light-shoot...

920 seems to live up to the promise also in this regard.

I wish they shot something with motion in low light... From pictures my friends post online, my biggest complaint about most point and shoot (and phone) cameras in low light is not digital noise, but motion blur. I would prefer manufactures to bump the gain and take faster exposures.

forget the image stabilization comparison. the 920 image quality is incredibly better than the others.

On one hand - the image stabilization on the Lumia appears to be great. However, the image quality on both the iPhone 5 & Galaxy S3 were much worse than I've experienced in day to day use. I've used both phones personally - and the video quality on the iPhone and GS3 is much better than what was shown (even if it's not "stable"). Any idea on why the quality on those was much lower than in real life?

Were you walking while filming? He wouldn't be able to showcase the differences in image stabilization for each phone if he tried to stay as still as possible.

Yes, in my day-to-day use of each camera I've been moving. I guess my point is - there seems to be a massive quality degradation in what I've seen from personal use.

Actually it is possible to show the benefits of OIS without video. It allows you to use longer shutter speeds when taking pictures in low light. Take a look at this one:

http://www.youtube.com/watch?feature=player_embedded&v=j...

While this 'actual performance' is far from rigorous science, it's still great to see the Lumia performing so well.

Not to give in to conspiracies of any sort, but these videos prove nothing.

1. Why are the zoom levels different?

2. I'm certain the HTC One X can shoot more decent videos than what we saw.

3. Are we dealing with equivalent bitrates?

4. Isn't Image Stabilization a post-processing solved problem? I mean, it's great to have it in real-time (after a google search, it seems all three do have the feature), but there's nothing wrong with post-processing when you REALLY care about the final quality of your work.

1. Lumia have wider field of view.

4. Today commercially available video post processing algorithms can only solve relative stabilisation between frames of video, because inframe stabilisation is still very slow for modern hardware.

*Edited. 4 was "No, because post processing can only solve relative stabilisation between frames of video, it does not work within one frame."

>4. No, because post processing can only solve relative stabilisation between frames of video, it does not work within one frame.

Apparently you haven't seen this: http://www.youtube.com/watch?v=XLLJBfRzXIQ#t=0m57s

I actually read this paper last year... it's pretty impressive but still is not ready for mainstream. The speaker said that his friend was "lucky", but he was the lucky one to find a perfect example to show this off.

try doing it in a mobile device with 30 full hd frames per second... Still, it's pretty amazing, pity it isn't actually in CS7

Who says post-processing has to be in real time?

Thank you for clarifying :)

More demos here:

http://www.youtube.com/watch?feature=player_embedded&v=c...

http://www.youtube.com/watch?feature=player_embedded&v=6...

http://www.youtube.com/watch?v=IlT8qk2ZoGM&feature=playe...

http://wmpoweruser.com/see-nokia-lumia-920s-video-stabilizat...

All of the others seem "zoomed in". I've heard someone say it's because the others use digital stabilization, which means they have to crop the middle of the image to stabilize it against the edges, or something like that. Is that true? Because then OIS seems to significantly improve image quality in video recording as well.

It makes sense 920 has 26mm focal length, S3 16mm, iPhone 5 33mm (equivalents). Smaller focal length - wider field of view so it makes sense that iPhone feels zoomed in. I can't see the difference for S3, though. Magic?

But even the iPhone 5 looks more zoomed in then Lumia 920, though. Look at the initial filming, when he films the other people filming him. They are bigger on the iPhone. The quality is poorer as well.

read the message above you. The lenses have different focal length. 920 gives you broader field of view so you can get wider picture and more light. iPhone doesn't. That's why iPhone looks zoomed in.

To me it looks like the "others" have far lower bitrate than the Nokia. It'd be interesting to see the originals.

OIS has the added advantage that it makes the video encoder's job easier, since there's less motion to compensate for. A stabilized video will show fewer artifacts when encoded at the same bit rate.

I think if you were trying to hold a still shot the image stabilization will work great. When moving around, though, both cameras still look pretty bad to me. The rolling shutter I think is still the weakest link. To be fair, even high-end DSLRs don't have that solved, but it would be great to see somebody come up with a solution.

Reply on news.ycombinator.com