Under the microscope with Nokia's full focus camera phones
LONDON, England – As someone who spends a portion of almost every day reading comments on twitter, discussion forums, tech blogs etc, following the recent introductions of the Nokia E7, C7 and C6, it’s become evident that we need to take time out to explain the technology behind the Nokia Full Focus cameras used in these products.
The big misconception goes something like this. Low tech = cheap = poor performance. Well, let’s explain more and see if that’s the view you have in a few minutes…
Fixed focus cameras have no moving components and have simple cost-driven optics. That’s true. So based on this background, I can understand why people come to such conclusions. After all, fixed focus cameras have been around for years, right? And they’ve always been used in cheap, uninspiring products, right?
However, the only similarity between fixed focus cameras of days gone by and Nokia’s Full Focus cameras is that they have no moving lens elements. That’s where the similarity ends.
What about autofocus cameras? Let’s first start by saying, yes, mechanical autofocus modules are incredibly complex with moving lens elements and lens driving mechanisms which are so incredibly compact with such high precision engineering that they have to be analysed during development using electron microscopes. Putting the mechanics to one side, it could be easily argued that Full Focus cameras are in fact more complex than autofocus cameras. But before we get to that, let’s explain why we developed these Full Focus cameras.
Necessity is the mother of Full Focus
Some years ago, we became very much aware that there would be a need for smaller and smaller camera modules. As mobile devices packed in more and more components, each component over time would need to be progressively miniaturised. In addition to this trend, we could see the frustration for casual users with mobile devices where there usage is increasingly spontaneous.
All the details: Nokia Lumia 800
All about our stunning new smartphone.The specs; the price; the facts
We therefore needed to develop cameras which could be used faster and with less hassle than the autofocus cameras of the time – and were smaller, too. For these principle reasons, Full Focus cameras have been developed and utilised widely across our portfolio to provide the best balance available today between performance and convenience of operation for those people these devices are developed for.
Here are the unique benefits of these cameras:
- With a Nokia Full Focus camera you can simply grab moments without the fear of images or videos ever being out of focus for subjects from around 50cm through to infinity, regardless of lighting conditions or subject position in the frame etc.
- No shooting lag waiting for autofocus. Despite the Nokia N8 featuring the fastest autofocus of any mobile at just 350ms (average), it’s still a delay.
- Everything from approximately 50cm to infinity in the same picture/video appears in focus. Hence the name ‘Full Focus’ cameras. Conventional cameras have less ‘depth of field’ than Full Focus cameras and therefore, distant objects will appear blurred. Those more experienced photographers may – of course – prefer the reduced depth of field look. However, those less involved (the vast majority of those who take pictures) find the “everything in focus” look very appealing.
- Slimmer camera modules – leading to slimmer product designs.
So what makes this possible?
There are two principle elements which make Nokia Full Focus cameras possible. First, very complex lens design and second, incredibly sophisticated image processing. Let’s look at both of these in more detail.
Complex lens design
It’s fair to say that at Nokia, we have been pushing lens design way beyond what people just a few years ago thought was impossible. The latest generation Full Focus camera used in the Nokia E7, C7 and C6 uses a four-element plastic lens design. Some will be saying at this point, “plastic, yuk! Told you so, cheap lenses.” The truth is that the lens designs we are now using are so complex that you simply can’t mass produce them using glass. Referring to these simply as ‘plastic’ would be doing a huge injustice to the lens designers.
In the case of this latest generation Full Focus module, for the four lens elements, we use three different polymers. Each polymer is selected based on the design of each lens element. For example, the way the polymer flows into the incredibly high precision lens moulds will influence the optical characteristics. For this reason, the lens designer will use different polymers based on the flow characteristics of the material’s journey into the mould, amongst many other aspects, such as how the material cools, the surface precision, internal stresses etc. etc. All these things need to be balanced as part of the lens design. And that’s not even considering the optical performance characteristics such as optical index, light dispersion properties, as well as possible aberrations such as colour fringing.
Blurred for clarity
This may sound crazy, but a principle aspect of this complex lens design to is to blur the image. Yes – you heard right, to blur the image, but… this isn’t like smearing Vaseline over the lens like some diffused effect: this is strictly controlled blurring of the image.
Conventional lenses have a spherical surface. This means that the curvature of the lens across the surface is of an even radius. Such lens forms today are easy to produce. The next step was aspherical lenses. Here the radius of the curvature changes as you move from the centre of the lens to the outside. These require high-precision manufacturing and have been instrumental in allowing much smaller and more complex lens designs. For example, all of the lenses used in the Nokia N8 are aspherical.
This Full Focus lens design goes way beyond this though. Lens surfaces have a form which is far more intricate than ever before. It’s surface curvature changes many times as you move from the centre to the edge of the lens element. Conventional lens manufacturing simply can’t be used for such a lens design. The only way you can produce such a lens is by moulding. Our lens tolerances are normally measured to less than a few thousandths of a millimetre. But in the case of these special lenses, the tolerances have to be even higher!
Three or four years ago, it simply wasn’t possible to mass produce lenses with such high precision. It’s this breakthrough that fundamentally make Nokia Full Focus cameras possible. But that’s only one part of the story. And if you know anything about lens manufacturing you’ll know that greater precision increases the cost, rather than reduces it. But that’s a necessary cost to create these cameras.
Add secret sauce
What we’re left with now is a subtly blurred image. So the next stage is the highly complex image processing which takes our slightly blurred image and then processes it so that we end up with an image/video which is sharp from around 50cm through to infinity and yet fits in super slim devices like the new E7 and C7.
Unfortunately, I can’t go into too many details behind the processing as they are classified as secret. However, what I can say is that, despite no moving lens elements and an image which is slightly blurred before processing, these algorithms digitally and dynamically focus the image. So objects at different distances have a dynamic influence on the image processing carried out in each individual image. So each image is processed according to the content of the image rather than some generic image processing algorithm.
The data now leaves the camera module – having been blurred and then processed so it’s sharp again – and is now passed to the image processor. Here, the more conventional image processing is applied to the image such as colour correction, etc. However, this means that we have image optimisation in the camera module as well as in the image processor.
Just handling image processing in the image processor is complex enough but now we need to ensure that the combined data processing is the right mix. Often, we may have to make further optimisations in the camera module processing which will then have an additional knock-on effect to the conventional optimisation in the image processor.
But after all that the result is in focus images or videos from around 50cm to infinity with point-and-click simplicity. Simple on the surface, but massively complex underneath.