Single-Lens Google Pixel 2 Camera Takes Top Spot From iPhone 8 Plus in DxO Labs Tests

[symphara]

well, real optical zoom is simply unavailable in phone atm until some digital lens technologies will somehow solve not getting the phone thicker.

I think the zoom offered by computational means using a high-resolution monochrome sensor is a pretty good idea in the meantime, which I'd rather have. As I said, it helps with framing, so you don't need to edit at all.

Again, I have no need for the narrow angle lens on the iPhone or Note 8. You can just crop after, with a loss of resolution but it's usually fine.

I noticed you labelled your normal shot as "wide angle". That's precisely what you don't have, a wide angle lens. That it something else, and there are many shots, for example room interiors, that you cannot properly take with the normal lens.

 

[0129972]

I've been reading this and the other popular Apple enthusiast websites for years and this was the article that finally got me to join and comment. There were so many errors in the DXOMark review that I couldn't let them go unaddressed. We need more critical thinking. If I were in collage and I submitted the results of DXOMark's research (if you can call it that) as a source for any sort of academic research, it would be similar to citing a personal blog. It is unscientific and subjective.

To those with critical minds, there are several obvious inconsistencies:

· Several of the photos are not taken from the same angle and distance to the subject.
· There is no control to compare the results to
· DXOMark claims to be “the reference for image quality”—a misguided fallacy at best

Without every photograph being taken from the exact same position in relationship to the subject, the results are inconclusive and irrelevant. Minor changes in distance and, particularly, angle can cause changes to a smartphone’s camera settings, e.g., the focal point, ISO, exposure, etc. It was evident that several of the photographs, particularly those with the greatest variances, were not taken with the same relationship to the subject.

Next, without a control photograph to compare the smartphone images to, there’s no way for anyone to know exactly how close to “accurate” the images really are. While there’s no perfect way to obtain a control image, a DSLR image taken by a professional, ideally the same person taking the photographs with the smartphone cameras.

Finally, science requires a single variable to be test at a time for reliable results. All other factors need to be controlled for. It’s clear the effort here was minimal, relegating the outcome to junk science or yellow journalism masquerading as truth. This “test” was so sloppy it would not be admissible to any peer reviewed scientific journal or acceptable to any academic critical observer.

After all variables have been controlled for, only then can the results be objectively reviewed. Even then, there’s much that will be subjective. There are objective aspects, such as noise, color saturation, sharpness and definition, detail, etc. that can be compared. Those aspects can only be compared against reality or the control photo, in order for the results to be objective. After that, much of an images appeal and even “quality” will depend on the preferences of the beholder. The primary issue with DXOMark’s review is that there was no good separation of subjective and objective and no control of the variables.

I’m waiting for a scientific review before a real winner can be determined.

 

[symphara]

I've been reading this and the other popular Apple enthusiast websites for years and this was the article that finally got me to join and comment. There were so many errors in the DXOMark review that I couldn't let them go unaddressed. We need more critical thinking. If I were in collage and I submitted the results of DXOMark's research (if you can call it that) as a source for any sort of academic research, it would be similar to citing a personal blog. It is unscientific and subjective.

To those with critical minds, there are several obvious inconsistencies:

· Several of the photos are not taken from the same angle and distance to the subject.
· There is no control to compare the results to
· DXOMark claims to be “the reference for image quality”—a misguided fallacy at best

Without every photograph being taken from the exact same position in relationship to the subject, the results are inconclusive and irrelevant. Minor changes in distance and, particularly, angle can cause changes to a smartphone’s camera settings, e.g., the focal point, ISO, exposure, etc. It was evident that several of the photographs, particularly those with the greatest variances, were not taken with the same relationship to the subject.

Next, without a control photograph to compare the smartphone images to, there’s no way for anyone to know exactly how close to “accurate” the images really are. While there’s no perfect way to obtain a control image, a DSLR image taken by a professional, ideally the same person taking the photographs with the smartphone cameras.

Finally, science requires a single variable to be test at a time for reliable results. All other factors need to be controlled for. It’s clear the effort here was minimal, relegating the outcome to junk science or yellow journalism masquerading as truth. This “test” was so sloppy it would not be admissible to any peer reviewed scientific journal or acceptable to any academic critical observer.

After all variables have been controlled for, only then can the results be objectively reviewed. Even then, there’s much that will be subjective. There are objective aspects, such as noise, color saturation, sharpness and definition, detail, etc. that can be compared. Those aspects can only be compared against reality or the control photo, in order for the results to be objective. After that, much of an images appeal and even “quality” will depend on the preferences of the beholder. The primary issue with DXOMark’s review is that there was no good separation of subjective and objective and no control of the variables.

I’m waiting for a scientific review before a real winner can be determined.

Interesting that DxO proclaiming the iPhone 8 to be the best ever, using the same flawed methodology, didn't make you post anything

My major complaint about the testing they do (others too) is using the default settings to take shots. This, to me, hardly tests the capabilities of a camera.

 

[0129972]

Interesting that DxO proclaiming the iPhone 8 to be the best ever, using the same flawed methodology, didn't make you post anything

My major complaint about the testing they do (others too) is using the default settings to take shots. This, to me, hardly tests the capabilities of a camera.

What's really interesting is that you know nothing about me or what my preferences are, yet you accuse me as though you do. Seek first to understand, then to be understood.

 

[0lf]

I understand you like this iPhone/Note 8 implementation, but I'm sorry, I don't find it that useful. With my DSLR I use this kind of lens (50mm) rarely, when I want to shoot portraits, but my prime is a 2.4 and has beautiful resolution, so it actually has some advantages which you don't have on the iPhone.

Don’t be sorry, there is different camera implementations that give choice, it is a good thing.

I think zoom, even in small amounts, is much more useful (for framing) than a fixed lens that's not optically very good nor very bright. It's not even stabilized as far as I know (only the X?). I also think that wide-angle is much better. The V30 did that one right, if it had anything but a sub-mediocre front camera, it would have tempted me.

iPhones do have zoom, but non purely optical one. It does not necessarily behave like if you jumped from one fov to the other. You can do it, but it is not the only behavior available. That said, I am perfectly fine with step zoom. Even my camera (G7x2) is set to step zoom (but the steps are smaller). On the other hand, IMHO, around 1.3x is too limited to be useful, and it is even more limited where you only can choose between wide and less wide.
[doublepost=1507458149][/doublepost]

I noticed you labelled your normal shot as "wide angle". That's precisely what you don't have, a wide angle lens. That it something else, and there are many shots, for example room interiors, that you cannot properly take with the normal lens.

28mm fov is a wide angle. And 56mm fov is narrower than a normal focal length. A normal lens has a focal length around the diagonal of the sensor. It is between 40mm and 45mm fov.

 

[symphara]

What's really interesting is that you know nothing about me or what my preferences are, yet you accuse me as though you do. Seek first to understand, then to be understood.

Chill out, I didn't accuse you of anything.
[doublepost=1507461361][/doublepost]

Don’t be sorry, there is different camera implementations that give choice, it is a good thing.

Just the choice offered by the iPhone/Note 8 implementation leaves me unimpressed. I find both the V30 solution (with the wide-angle lens) and the Huawei solution (with a monochrome lens) preferable for dual cameras.

I understand you consider your main lens to be wide-angle, and the secondary one a telephoto. I just think they're slightly wide, and respectively slightly tele. Too slightly to actually count as wide or tele without qualifiers. To me the iPhone just has a normal lens and narrower angle lens. As in, if I told my wife before going on a trip "please take the wide angle and the tele" and I'd later discover she took the 28mm and the 56mm, I'd be pretty confused as to why didn't she take the 18mm and the 200-500mm. But she knows better

At this stage it's clear though I won't convince you and you won't convince me of the contrary.

The V30 has an actual wide-angle lens that you can use to take shots that you just cannot do with your iPhone or any other phone. For example, you can shoot well room interiors, which is a problem for phones in general, mine included. Their main lens is not wide enough.

Huaweis have really good camera apps with excellent and easy to use manual controls, the best manual controls for the wide aperture mode, and really interesting in-app camera effects such as desaturated bokeh. Plus, I like the colours they do, I think their collaboration with Leica shows here. In contrast, I think the iPhone photos are over-sharpened and rather un-natural in terms of colours, but that's me. I tried a few times my iPad Pro's camera (which is the same as on the iPhone 7) and it's definitely over-processed, I don't like the look. It reminds me of the photos taken by my Samsung Galaxy Alpha, which superficially look nice and punchy, but fail at close detail inspection and don't actually reflect the scene colours and light conditions that well.

Having said that, it looks like the Pixel 2 is beating everyone right now. At least until the Mate 10 comes in a week or so

PS: my wife ordered the Pixel 2 XL so I'm curious to see its photographic prowess in person!

 

[0lf]

Having said that, it looks like the Pixel 2 is beating everyone right now. At least until the Mate 10 comes in a week or so

For you, sure. For me, not so much.

 

[lordofthereef]

My major complaint about the testing they do (others too) is using the default settings to take shots. This, to me, hardly tests the capabilities of a camera.

I can see both sides of the argument. The average user just points and shoots. They don’t fuss with settings. They don’t edit after the fact. They slam down the button and what comes out is the final product.

So, if you want to compare products for someone looking to upgrade their cellphone and get the best pics no muss no fuss you present the data in a way that makes sense to them.

Now, of course you could present the data multiple ways. There’s no reason not to go more in depth. But I do see value in testing the product as is out of the box.

FWIW there are also some great bargain TVs that peoooencould have ifbthey research and professionally calibrate then to the room. Instead most people walk into a store with a couple dozen TVs on the wall and pick what looks best there. Often those TVs are calibrated differently to fit that atmosphere and don’t look quite the same when they get them home.

 

[LordVic]

I don’t think I’d describe either of them as vibrant. The Pixel 2 looks more maroon than red to me in the carousel image.

Not that it matters, any and all images we see in comparison shots like this. Or indeed for a review of a single device have one inherent problem.
Not a single one of us knows how the subject looked on the day the image was taken.

Even if they were taken on the same day immediately after one another we still have no actual frame of reference as to how the subject looked in real life, other than possibly someone’s subjective opinion.

That’s why I always advise that someone try cameras out for themselves before buying them. There’s a lot you can’t tell from an image on the internet.
Of course that’s less of an issue with phones as we, I would hope at least, tend to buy them for more important reasons than the camera alone.

two inherent problems.

your first one,

and second, we're all viewing them on different displays from eachother, and while many can be calibrated, if you believe that the majority of the world is looking at these images on identically calibrated, perfect colour reproducing displays, you'd be amiss.

I bet MOST of us are looking at these pictures on a display with subtle colour differences. Differences in brightness, or some monitor with some additional settings that might change how accurately they are displayed.

So not only can we not know how close to real life the photos look, we can't even guarantee we're all looking at them the same way.
[doublepost=1507509575][/doublepost]

For the love of God, this fake Bokeh fad needs to go away fast. That shot is even worse than the ones shown from the Note 8 in the N8 vs iPhone Thread

Problem is it' snot even Bokeh. it's artificial blur, and anyone who knows even a small bit of photography is going to see through the fake blur on ANY smartphone in a second.
[doublepost=1507509757][/doublepost]

But to my knowledge, they are made by different company. Pixel 2 was made byHTC and the Pixel 2 XL by LG. There could be differences.

for the PIxels it'll come down to where they sourced the image sensors. if HTC sourced their own and LG sourced their own, yes, this might be the case.

if Google sourced the image sensors for both makers, and set the standard for both, and the same software for both, they'll both produce the same quality of images.
[doublepost=1507510240][/doublepost]

What do you mean narrow angle?
The second lens in the iphone + is real tele lens.

just a snobbish correction. It's not a real telephoto lens. It is a 2x optical zoom lens.

for a Telephoto lens to be a telephoto lens, i'm going to quite wiki cause it's easier than my brain trying to word it differently

"In photography and cinematography, a telephoto lens is a specific type of a long-focus lens in which the physical length of the lens is shorter than the focal length.[1] This is achieved by incorporating a special lens group known as a telephoto group that extends the light path to create a long-focus lens in a much shorter overall design."

 

[TrueBlou]

two inherent problems.

your first one,

and second, we're all viewing them on different displays from eachother, and while many can be calibrated, if you believe that the majority of the world is looking at these images on identically calibrated, perfect colour reproducing displays, you'd be amiss.

I bet MOST of us are looking at these pictures on a display with subtle colour differences. Differences in brightness, or some monitor with some additional settings that might change how accurately they are displayed.

So not only can we not know how close to real life the photos look, we can't even guarantee we're all looking at them the same way.

So very true, for me there are too many variables to compare two cameras of any type, or manufacturer, from images we do not take ourselves.

Maybe I’m nit picky, but I always have chosen my cameras from real world testing. With the exception of a phone camera, because I don’t buy the phone for the camera at all. There’s far more important things to consider for a phone, but it’s nice to have a decent camera on there. Thankfully all of the big players produce good enough phone cameras anyway.

 

[LordVic]

So very true, for me there are too many variables to compare two cameras of any type, or manufacturer, from images we do not take ourselves.

Maybe I’m nit picky, but I always have chosen my cameras from real world testing. With the exception of a phone camera, because I don’t buy the phone for the camera at all. There’s far more important things to consider for a phone, but it’s nice to have a decent camera on there. Thankfully all of the big players produce good enough phone cameras anyway.

you're nit picky. i only pick my cameras based on whatever I can convince my father he no longer needed before I take it

All kidding aside. I'm a Canon colour fan, and have been shooting (purely amateur for fun) with my Canons. Starting with a D300 and now still using a 40D. I've taken a few photography courses (which really helped me learn how all of the tech in cameras work).

THe one thing that these camera shootouts really tell me?

Smartphone cameras are fantastic for capturing the moment. They're almost on you all the time, and they're generally really fast to get a shot taken. And now, they're generally good enough that I don't carry my DSLR in my car with me most places (and only take them out when i feel artsy).

These camera's are all really decent now. They're easily better than any point and shoot i ever had. But still not up to even my older DSLR's quality. But that to me is a given considering that Camera's, of all shapes and sizes, are limited by the physics of light. THe smaller the camera sensor, the less light it can "see"

Buy any flagship phone in 2017 and you will have a camera that will treat you well.

 

[TrueBlou]

you're nit picky. i only pick my cameras based on whatever I can convince my father he no longer needed before I take it

All kidding aside. I'm a Canon colour fan, and have been shooting (purely amateur for fun) with my Canons. Starting with a D300 and now still using a 40D. I've taken a few photography courses (which really helped me learn how all of the tech in cameras work).

THe one thing that these camera shootouts really tell me?

Smartphone cameras are fantastic for capturing the moment. They're almost on you all the time, and they're generally really fast to get a shot taken. And now, they're generally good enough that I don't carry my DSLR in my car with me most places (and only take them out when i feel artsy).

These camera's are all really decent now. They're easily better than any point and shoot i ever had. But still not up to even my older DSLR's quality. But that to me is a given considering that Camera's, of all shapes and sizes, are limited by the physics of light. THe smaller the camera sensor, the less light it can "see"

Buy any flagship phone in 2017 and you will have a camera that will treat you well.

Ohhhh I don’t know that I’d agree on the p&s camera bit, I’ve definitely had better.

But like you say, the big benefit of the phone camera is that it’s almost always with you. And as the saying goes, there are no bad cameras, only bad photographers

 

[symphara]

I can see both sides of the argument. The average user just points and shoots. They don’t fuss with settings. They don’t edit after the fact. They slam down the button and what comes out is the final product.

So, if you want to compare products for someone looking to upgrade their cellphone and get the best pics no muss no fuss you present the data in a way that makes sense to them.

Now, of course you could present the data multiple ways. There’s no reason not to go more in depth. But I do see value in testing the product as is out of the box.

FWIW there are also some great bargain TVs that peoooencould have ifbthey research and professionally calibrate then to the room. Instead most people walk into a store with a couple dozen TVs on the wall and pick what looks best there. Often those TVs are calibrated differently to fit that atmosphere and don’t look quite the same when they get them home.

I see value too in testing the out-of-the-box settings, of course. But you can hardly say you're exploring the capabilities of a camera if all you do is shoot in auto mode. It's not like having your TV calibrated - very few people do that, because it's not worth doing it for most consumer TVs.

Look at dpreview camera reviews. They're about what the camera can do, not about what it does in auto mode. Again, I recognise that a lot of people will never use a manual control. But with my phone camera, metering for example is re-done when you tap the screen, and it can dramatically improve a shot or change its nature. Does that count as manual mode?

I think every serious phone camera test should discuss the pure auto mode, have another section on the "easy" adjustments such as HDR, change of metering and focus by tapping, other modes such as "portrait" etc, and a section on full manual adjustments (ISO, exposure, WB etc) and raw shooting, to show what the limits of the camera are. I'd also like more attention paid to the camera apps, they are very different.
[doublepost=1507537079][/doublepost]

just a snobbish correction. It's not a real telephoto lens. It is a 2x optical zoom lens.

I think it's lazy and misleading to call what the iPhone has a "telephoto" lens. It implies to me capabilities it doesn't actually have.

 

[lordofthereef]

I see value too in testing the out-of-the-box settings, of course. But you can hardly say you're exploring the capabilities of a camera if all you do is shoot in auto mode. It's not like having your TV calibrated - very few people do that, because it's not worth doing it for most consumer TVs.

Look at dpreview camera reviews. They're about what the camera can do, not about what it does in auto mode. Again, I recognise that a lot of people will never use a manual control. But with my phone camera, metering for example is re-done when you tap the screen, and it can dramatically improve a shot or change its nature. Does that count as manual mode?

I think every serious phone camera test should discuss the pure auto mode, have another section on the "easy" adjustments such as HDR, change of metering and focus by tapping, other modes such as "portrait" etc, and a section on full manual adjustments (ISO, exposure, WB etc) and raw shooting, to show what the limits of the camera are. I'd also like more attention paid to the camera apps, they are very different.
[doublepost=1507537079][/doublepost]
I think it's lazy and misleading to call what the iPhone has a "telephoto" lens. It implies to me capabilities it doesn't actually have.

I don’t think you’re wrong. Just saying that most people don’t care about a single setting beyond removing it from their pocket and taking a photo.

Taking it another step further is benchmarks. Even fewer people care about that except to argue in forums lol.

 

[sidewinder3000]

My phone has two lenses, same angle, one color and the other monochrome. That's how it does what you call "portrait mode", on my phone is actually a much more flexible "wide aperture" mode which lets you adjust the aperture (thus the amount of bokeh) and the focus point, even after you take the shot.

It never occurred to me that I'd need a narrow angle lens. If it was a real zoom - definitely, an actual telephoto, maybe. But just a narrower angle? I don't think I need that.

On occasion I would have liked a wide angle lens like the LGs do, but that's about it.

What's interesting to me is that the newer Huaweis (Mate 9 and the P10s) use a 20Mpx monochrome sensor and a 12Mpx color sensor. If you shoot in 12Mpx color shots, it lets you actually zoom without artefacts up to 20Mpx. This is nice for framing. I'm sure the Mate 10 will offer something like this.

Wow. If you don’t understand the benefit of a neutral lens (equivalent of a 50mm on a full frame) over the wide angle that’s the default on most smartphones, then you really don’t understand photography, and nothing I say will change your mind. But in case your mind opens for a few seconds, there’s this: being able to shoot subjects through the longer focal length lens is a huge benefit for any sane photographer who occasionally wants to shoot without the ever-present distortion of a wide angle lens.