xpad

volcan said:

xpad said:

Since you've added this, look at #2, then look at the trees in the first photo in the article. The bright areas in the trees are bokeh, not blur.

I think the image you refer to resembles Wikipedia's number 3 more than anything else. I and many other people have described bokeh as the background having circles of convolution of which I see none in the iPhone images. So we disagree, no big deal.

Look at the trees in the first image (the others don't have the type of background that amplifies the effect of bokeh). Above the child's head. Those bright spots are circles of confusion visibly spreading into the tree, which the trees aren't visibly doing into the bright areas.

wiggin said:

xpad said:

Follow up to my previous comment, look at the sky showing through the leaves in the first photo. The light pattern is not merely a blur, but is bokeh-like in how the bright area expands into the darker area.

I think I would agree with you that we can call this bokeh. But I would also say that I don't find it as pleasing as the affect you get with a larger lens which is able to achieve a narrow depth-of-field. However, most of the time it's going to be better than nothing (although in a few photos I've seen I actually find it quite distracting, as if my own eyes aren't focusing properly and they are trying to adjust to get the background back into focus).

That's all that I'm saying. It is bokeh, in spite of what snobs are saying. They don't have to like the bokeh, but it's there (and it's not just a Gaussian blur).

sflocal said:

I'm just cracking-up at the back-and-forth ranting about the "bokeh" discussion.  Apple's implementation is not anywhere near what true bokeh is.  They're passing off a simple background blur - whatever "gaussian" or other name you want to call it - and the posters here claiming it to be bokeh.  It is not.  It's nice, but it has nothing to do with it.  The aperture blades found in cameras also contribute to what real bokeh does, in addition to how the DOF is rendered in a way only analog lenses can do.

No, it's not just a simple blur. You are confusing "something I like" with "bokeh". It's the True Scotsman fallacy. "That's not a pizza, it has chicken on it". "That's not bokeh, it doesn't look as nice as my dSLR!"

Bokeh is simply the noticeably out of focus area of a photo. That's it. All iPhones have had bokeh. But due to the wide depth of field, it's very limited. This is a software process that simulates bokeh NOT SIMPLY WITH JUST A BLUR, but it takes into account distance and shows bright areas expanding into darker areas, not just blurring the two into each other.

Are SLRs better? Almost universally yes, you don't even have to look at any test photos to tentatively assume this. No one is saying it's exactly just as good as a nice Nikon or Canon with a fast lens. But it's nice, impressive, not just a Gaussian blur, and is a good simulation of bokeh.

volcan said:

This is how Wikipedia describes it. 1) none, 2) bokeh, 3) Gaussian blur

Since you've added this, look at #2, then look at the trees in the first photo in the article. The bright areas in the trees are bokeh, not blur.

volcan said:

xpad said:
Apple's implementation doesn't use a Gaussian blur...

Whatever you want to call it, it is indistinguishable from Gaussian blur in my opinion and entirely unlike any bokeh examples I have created or seen on the net. Please go search Google images for bokeh and find an image that remotely resembles the photos in this article.

That's not how words work. Your argument is that it's not bokeh, but just Gaussian blur.

The problem is that it's not a Gaussian blur. Just because you can't tell the difference doesn't make it so.

This wouldn't be a problem if you were just ignorant, but you are making a strong claim based on ignorance. Those "examples" you are talking about are just a subset of bokeh. Bokeh itself is just the out of focus area of a photograph. Every normal camera has bokeh, even past iPhones. You will not find that most iPhone photos have the same qualities that those examples you are referring to have.

Those extreme effects you are thinking about are most noticeable when there is a small, bright light source contrasted against a darker area, like with Christmas lights. You don't see that as readily in photos without such light sources.

Another place you find it is in looking through trees where the sunlight or bright sky shines through. In the first photo, you can see this. If it were just a Gaussian blur, the darker tree area would blur into the bright sky area equally to the amount the sky blurs into the tree, creating a simple gradient, not the true bokeh effect that you see in that photo.

Follow up to my previous comment, look at the sky showing through the leaves in the first photo. The light pattern is not merely a blur, but is bokeh-like in how the bright area expands into the darker area.

sflagel said:

Dear author: a blurred background is not bokeh, it is just depth of field. Bokeh is when you have light points turn into perfectly concentric circles, resulting in a "magical" and dreamy background. 

The depth of field looks ok in these pictures (better than none), I have not yet seen a bokeh effect anywhere. 

No, that's not the definition of bokeh, that's just the description of one attribute of bokeh that can vary (for example, you can have most any arbitrary solid shape for the patterns).

During this testing, someone needs to use a scene with some point-like light sources. Apple's implementation doesn't use a Gaussian blur, so there should be proper bokeh-like effects.