Night Sight Quality
I don't own an iPhone. But, when I read a title which implied that Apple had beaten Google at night photography I had to see the pictures to see the conclusion myself.
The article is here.
What I'm shocked at is two things; the lack of pictures and the serious flaws in the ones provided.
The article itself admits the framing in the first shot is bad. But, it neglects that it is ALSO bad in the second shot. And those are both the 2 shots where the iPhone is declared the winner.
In fact, on closer examination, I would say even the pumpkin shot suffers from framing issues.
And, next, virtually none of this matters as these pictures appear to be taken without a tripod. We can verify for sure in comic picture as we can see a reflection of a person holding the phones.
The first thing to understand about night photography is that the 2 most important factors are light and stability.
You may say that light is ALWAYS important. And you're not wrong. But, with night photography, we're talking about incredibly small differences in framing resulting in HUGE variations. If you have a Pixel 2, 3 or 4 go out at night and take a few pairs of pictures back to back from just slightly different angles. You'll be blown away at just how much a difference there can be.
And, even when you keep the framing reasonably close, the Pixel at least and probably true of the iPhone, subtle differences in your ability to hold the phone still will affect the exposure length. Which is INSANELY important to night photography as well.
Taking just 3 sets of images, with mild framing differences, without a tripod is not a test. And, who, writing an article comparing photography takes just 3 pictures? I can't say with any certainty, but this really feels like a struggle to stack the deck.
Perhaps the author took looks of picture and simply included those that were reasonably close. Anyway... let's look at my issues with these pictures individually.
The first picture is a street shot. And as the article admits, the framing is bad. Specifically, in the lower left corner there is a bright light. In the Pixel shot, this is MUCH more in frame than in the iPhone shot. This picture should be thrown away. This much variation in light would compromise a non-night photo. This much added light and contrast is more than enough to disqualify the comparison entirely. And it isn't the only issue with the framing.
If you look at the two, you'll notice a flare of light on the building on the left furthest from the lens. When I noticed that, I took stock of something else... the iPhone shot includes a lot more of the lower end of the buildings, which includes a lot of extra ambient light which wouldn't be enough to wash out the shot. In this case, this wouldn't make a difference in a day shot. But, again, these are night shots and every little bit of light matters, and that is adding a LOT of extra evenly lit lighting to the iPhone shot, and reducing the level of contrast, where the Pixel shot has more sky.
While I have no doubts that this was unintentional. The truth is, the fact that the Pixel did so well against the iPhone at such a massive disadvantage is actually damning to the iPhone.
The second image, the framing problem is more subtle. But, the iPhone shot is angled more, and angled towards the light source. If you look at the Superman and Wonder Woman comics you can see a reflective square which is clearly the brightest source of light in the shot. The iPhone shot has more of that light source, and the Pixel has more of the right side of the frame which isn't directing light into the sensor.
AND despite this, text is more legible and lines look clearer in the Pixel shot.
But, this like the shot before is not a valid point of comparison. Unfortunately, that slight shift has directed substantially more of what little light is available into the iPhone sensor.
Lastly is the pumpkin. You'll notice on the Pixel, under the left tooth on the top, there is what looks like an LED which isn't present on the iPhone. What is interesting here is that the light inside of the pumpkin looks more overblown in the iPhone shot, despite it being claimed that Google has a greater tendency to overexpose light sources.
But, again, I'm not even saying that the Pixel wins this shot. It is actually a perfect example of how important framing is when attempting to compare night shots. The Pixel may have been exposed that internal glow less because of that extra light coming from the mouth. It may also may explain why it picked up the fog better.
Another problem of course would be the stated aim. Google never claimed night mode was about accurate image reproduction. In fact, I once read a Google article which (paraphrasing) called it HDR on steroids. Basically, Night mode wants to make dark shots visible, but also bring in a lot of contrast. Not a natural photo.
And this article is comparing them on the basis of what looks more natural. If Apple's algorithms are trained more to preserve color accuracy, then it isn't a shock when they win if the determination is accuracy.
Personally, I think both take amazing pictures at night. If the Pixel 2 screen controversy taught me anything though, it is that people aren't as a big a fan of accurate color reproduction as they might think.
The article is here.
What I'm shocked at is two things; the lack of pictures and the serious flaws in the ones provided.
The article itself admits the framing in the first shot is bad. But, it neglects that it is ALSO bad in the second shot. And those are both the 2 shots where the iPhone is declared the winner.
In fact, on closer examination, I would say even the pumpkin shot suffers from framing issues.
And, next, virtually none of this matters as these pictures appear to be taken without a tripod. We can verify for sure in comic picture as we can see a reflection of a person holding the phones.
The first thing to understand about night photography is that the 2 most important factors are light and stability.
You may say that light is ALWAYS important. And you're not wrong. But, with night photography, we're talking about incredibly small differences in framing resulting in HUGE variations. If you have a Pixel 2, 3 or 4 go out at night and take a few pairs of pictures back to back from just slightly different angles. You'll be blown away at just how much a difference there can be.
And, even when you keep the framing reasonably close, the Pixel at least and probably true of the iPhone, subtle differences in your ability to hold the phone still will affect the exposure length. Which is INSANELY important to night photography as well.
Taking just 3 sets of images, with mild framing differences, without a tripod is not a test. And, who, writing an article comparing photography takes just 3 pictures? I can't say with any certainty, but this really feels like a struggle to stack the deck.
Perhaps the author took looks of picture and simply included those that were reasonably close. Anyway... let's look at my issues with these pictures individually.
The first picture is a street shot. And as the article admits, the framing is bad. Specifically, in the lower left corner there is a bright light. In the Pixel shot, this is MUCH more in frame than in the iPhone shot. This picture should be thrown away. This much variation in light would compromise a non-night photo. This much added light and contrast is more than enough to disqualify the comparison entirely. And it isn't the only issue with the framing.
If you look at the two, you'll notice a flare of light on the building on the left furthest from the lens. When I noticed that, I took stock of something else... the iPhone shot includes a lot more of the lower end of the buildings, which includes a lot of extra ambient light which wouldn't be enough to wash out the shot. In this case, this wouldn't make a difference in a day shot. But, again, these are night shots and every little bit of light matters, and that is adding a LOT of extra evenly lit lighting to the iPhone shot, and reducing the level of contrast, where the Pixel shot has more sky.
While I have no doubts that this was unintentional. The truth is, the fact that the Pixel did so well against the iPhone at such a massive disadvantage is actually damning to the iPhone.
The second image, the framing problem is more subtle. But, the iPhone shot is angled more, and angled towards the light source. If you look at the Superman and Wonder Woman comics you can see a reflective square which is clearly the brightest source of light in the shot. The iPhone shot has more of that light source, and the Pixel has more of the right side of the frame which isn't directing light into the sensor.
AND despite this, text is more legible and lines look clearer in the Pixel shot.
But, this like the shot before is not a valid point of comparison. Unfortunately, that slight shift has directed substantially more of what little light is available into the iPhone sensor.
Lastly is the pumpkin. You'll notice on the Pixel, under the left tooth on the top, there is what looks like an LED which isn't present on the iPhone. What is interesting here is that the light inside of the pumpkin looks more overblown in the iPhone shot, despite it being claimed that Google has a greater tendency to overexpose light sources.
But, again, I'm not even saying that the Pixel wins this shot. It is actually a perfect example of how important framing is when attempting to compare night shots. The Pixel may have been exposed that internal glow less because of that extra light coming from the mouth. It may also may explain why it picked up the fog better.
Another problem of course would be the stated aim. Google never claimed night mode was about accurate image reproduction. In fact, I once read a Google article which (paraphrasing) called it HDR on steroids. Basically, Night mode wants to make dark shots visible, but also bring in a lot of contrast. Not a natural photo.
And this article is comparing them on the basis of what looks more natural. If Apple's algorithms are trained more to preserve color accuracy, then it isn't a shock when they win if the determination is accuracy.
Personally, I think both take amazing pictures at night. If the Pixel 2 screen controversy taught me anything though, it is that people aren't as a big a fan of accurate color reproduction as they might think.
Comments
Post a Comment