Why a 16MP camera is better than 12MP

[D13H4RD2L1V3]

Barely readable text. With 16MP, small text should be more readable when you zoom in. Note 8 camera is excellent but heavily dependant on software processing.

It's worth noting that AC seems to apply some compression on top of what the app did. It's more detailed before uploading

It's not as detailed on the stock app due to the additional NR.

 

[Aquila]

Barely readable text. With 16MP, small text should be more readable when you zoom in. Note 8 camera is excellent but heavily dependant on software processing.

All smartphone cameras are heavily dependant on software processing, including LG's. The MP wouldn't have been the difference here. The "magic" of the V30's camera has nothing to do with hardware, it is their software processing that reduces distortion in wide angle shots. The hardware solution, the main shooter, the IMX351, isn't that good and is notably small for a smartphone sensor, which is why it's overall performance on still shots lacks so much compared to other flagships which use larger sensors to admit a boatload more light into their shots.

 

[vasekvi]

Took pictures of the moon with my V20 and V30 last night.

Stock software manual settings. The best I could get with each phone.

 

[muzzy996]

Smartphone camera megapixels do count, but it doesn't mean the image quality will be great if it's higher, just more pixels in an image so when you blow it up (zoom or crop), you won't have issues with detail. Why do you think everyone is complaining about LG's front 5MP camera? Images look too grainy because low megapixels. Same with the V20 8MP wide angle.. it doesn't look bad just a little pixelated from low pixel count. Give me Sony Xperia's 19MP camera anyday over Samsung's 12MP... I can just photo edit the image later, higher means more detail.

I'm not sure what the point of this thread is really - even if the argument was that all other things being equal megapixels matter there'd still be room for debate. Think about it, if all other things are equal meaning lens, physical sensor size and software then the primary difference would be pixel density on the SENSOR. It's been a known fact that digital cameras with larger sensor sizes tend to perform better in low light and that's in large part because of their pixel density. I've read reviews on the V30 that lament about the pairing of a good large aperture lens with a tiny sensor with high density which severely impacts its low light performance. The trouble with discussions on low light is that people have different ideas of what the definition of that is so their expectations vary.

I dunno . . at the end of the day, for me I believe in using the right tool for the job so I'll always reach for a larger sensor camera in a device dedicated to the task of taking good pictures rather than a smartphone with some hyped up camera whos sensor is smaller than the nail on my pinky toe. If i'm just out and about I'll surely shoot shots with my phone but if I know i'm going to shoot something of importance I'll consider a good point and shoot or grab my DSLR. So that would be my answer to your original question in post 1 on what I think about taking shots of the moon . . i'd grab a DSLR and the best lens I had to put on it to reduce cropping yet let in as much light as possible.

 

[flyingkytez]

Took pictures of the moon with my V20 and V30 last night.

Stock software manual settings. The best I could get with each phone.

//uploads.tapatalk-cdn.com/20180201/c41e45e8d1fec6f2d83dc73c6c014619.jpg

That looks very good actually! You can see the details of the moon pretty well. A 12MP would look blurry.

 

[flyingkytez]

Uh, far away text? S7 Edge, original pic vs cropped:

//uploads.tapatalk-cdn.com/20180201/3d85efe0c24db93fffd7c7ec231bcdcb.jpg
//uploads.tapatalk-cdn.com/20180201/907c4e6ca908d30eae7f857ff171828c.jpg

Sony lettering maybe a foot tall at like 100m away.

Very low quality when zoomed in. Smaller letters will not be readable.

 

[chanchan05]

Very low quality when zoomed in. Smaller letters will not be readable.

Maybe it's just me but why would you even be taking pictures of letters smaller than 1ft tall from over 100ft away? That's barely at the edge of normal vision.

 

[flyingkytez]

I'm not sure what the point of this thread is really - even if the argument was that all other things being equal megapixels matter there'd still be room for debate. Think about it, if all other things are equal meaning lens, physical sensor size and software then the primary difference would be pixel density on the SENSOR. It's been a known fact that digital cameras with larger sensor sizes tend to perform better in low light and that's in large part because of their pixel density. I've read reviews on the V30 that lament about the pairing of a good large aperture lens with a tiny sensor with high density which severely impacts its low light performance. The trouble with discussions on low light is that people have different ideas of what the definition of that is so their expectations vary.

I dunno . . at the end of the day, for me I believe in using the right tool for the job so I'll always reach for a larger sensor camera in a device dedicated to the task of taking good pictures rather than a smartphone with some hyped up camera whos sensor is smaller than the nail on my pinky toe. If i'm just out and about I'll surely shoot shots with my phone but if I know i'm going to shoot something of importance I'll consider a good point and shoot or grab my DSLR. So that would be my answer to your original question in post 1 on what I think about taking shots of the moon . . i'd grab a DSLR and the best lens I had to put on it to reduce cropping yet let in as much light as possible.

The argument is, yes larger megapixels do count to a certain extent. And what I'm trying to mention is taking pictures of distant objects I.E. licence plates or birds in the sky will be much better with a higher megapixel phone camera vs lower when you zoom/crop. I actually hated the Galaxy S7 camera.. megapixels were too low.. landscape photos sucked. However, the dual pixel technical is good.. extremely fast focusing and better low-light.. There's always a sacrifice. I'm satisfied with the V20's 16MP camera though.

 

[D13H4RD2L1V3]

The argument is, yes larger megapixels do count to a certain extent. And what I'm trying to mention is taking pictures of distant objects I.E. licence plates or birds in the sky will be much better with a higher megapixel phone camera vs lower when you zoom/crop. I actually hated the Galaxy S7 camera.. megapixels were too low.. landscape photos sucked. However, the dual pixel technical is good.. extremely fast focusing and better low-light.. There's always a sacrifice. I'm satisfied with the V20's 16MP camera though.

Again, it's not due to the MP. It's due to the heavy handed noise reduction.

With HDR+, the 12MP shooter on my Note8 and the Pixel 2 XL can resolve more detail. Heck, I used the camera on the Mate 9 even despite being "only" 12MP, I can make out all the lines and fingerprint on my hand.

Another point you're missing is diffraction. Smartphones cameras suffer from it due to them having small sensors and smaller pixels compared to an actual camera. Simply adding the resolution won't help, especially when the photodiode size is compromised as a result since that makes the diffraction problem worse. Multi-frame processing helps counter it to a degree.

 

[LeoRex]

V30 is even more processing dependant... They have to apply heavy noise reduction and sharpening to clean it up. It does ok with static images with defined borders and even coloring.. but textures and subtle details get absolutely clobbered.

I spent a good three weeks obsessively testing the V30's camera and in the situation you describe, it did well. Defining text like you say, is a function of the processing rebuilding those hard areas of contrast. But there's a flip side to all that.

One of my test pictures I often used is a printed calendar pinned to a wall by my desk. It does an excellent job of testing a cameras detail and resolving ability. When I pointed the V30 at it, I got will defined text and a nice, noise free image.

But I also lost reality. I take that same picture with my 2XL and I do, indeed, get sightly less crisp definition on the letters. But what I do see is texture... The subtle mottled look of the printer paper, the slight variation in ink of the letters... The depth of the threads on the fabric covering the cube walls. These are all details that are all but gone with the V30. The paper is a solid white, uniform text, digital artifacts where I saw weave in the fabric.

Taking pictures in my house I can see similar differences when looking at walls... Imperfections in the paint on the walls where I didn't bother to sand after priming are nowhere to be seen on the V30, but still persevered on my Pixel.

So yes, the V30 can be sharper, but what you are looking at in that image isn't what you shot.. all the warmth is lost. That little, solitary freckle near the corner of my son's mouth... Gone, wiped away when the noise reduction hammer fell.

That's why the Pixels keep scoring so high... Even though they may be a little softer, maybe with a little more noise, but they look real... The human eye is a finely tuned image processor that's evolved over eons to pick out the most minute details and patterns... Over-processed pictures, even though they are crisp and clear, trip a fuse in our brains telling us there is something not quite right about what we see.

 

[flyingkytez]

V30 is even more processing dependant... They have to apply heavy noise reduction and sharpening to clean it up. It does ok with static images with defined borders and even coloring.. but textures and subtle details get absolutely clobbered.

I spent a good three weeks obsessively testing the V30's camera and in the situation you describe, it did well. Defining text like you say, is a function of the processing rebuilding those hard areas of contrast. But there's a flip side to all that.

One of my test pictures I often used is a printed calendar pinned to a wall by my desk. It does an excellent job of testing a cameras detail and resolving ability. When I pointed the V30 at it, I got will defined text and a nice, noise free image.

But I also lost reality. I take that same picture with my 2XL and I do, indeed, get sightly less crisp definition on the letters. But what I do see is texture... The subtle mottled look of the printer paper, the slight variation in ink of the letters... The depth of the threads on the fabric covering the cube walls. These are all details that are all but gone with the V30. The paper is a solid white, uniform text, digital artifacts where I saw weave in the fabric.

Taking pictures in my house I can see similar differences when looking at walls... Imperfections in the paint on the walls where I didn't bother to sand after priming are nowhere to be seen on the V30, but still persevered on my Pixel.

So yes, the V30 can be sharper, but what you are looking at in that image isn't what you shot.. all the warmth is lost. That little, solitary freckle near the corner of my son's mouth... Gone, wiped away when the noise reduction hammer fell.

That's why the Pixels keep scoring so high... Even though they may be a little softer, maybe with a little more noise, but they look real... The human eye is a finely tuned image processor that's evolved over eons to pick out the most minute details and patterns... Over-processed pictures, even though they are crisp and clear, trip a fuse in our brains telling us there is something not quite right about what we see.

Google Pixel and iPhone X over process the photos. Honestly the iPhone camera looks ridiculously fake and too much HDR. Colors are boosted and shadow levels are raised, easily achievable through editing apps. Would rather have sharpness and higher resolution photos.

 

[flyingkytez]

Again, it's not due to the MP. It's due to the heavy handed noise reduction.

With HDR+, the 12MP shooter on my Note8 and the Pixel 2 XL can resolve more detail. Heck, I used the camera on the Mate 9 even despite being "only" 12MP, I can make out all the lines and fingerprint on my hand.

Another point you're missing is diffraction. Smartphones cameras suffer from it due to them having small sensors and smaller pixels compared to an actual camera. Simply adding the resolution won't help, especially when the photodiode size is compromised as a result since that makes the diffraction problem worse. Multi-frame processing helps counter it to a degree.

Try chasing someone down the highway and capture their license plate with a 12MP camera.. much more difficult than you think. Blownup pictures will not come out sharp enough. Same if you plan to print a big poster of the image or view on a 4k screen, not going to look great.

 

[D13H4RD2L1V3]

Try chasing someone down the highway and capture their license plate with a 12MP camera.. much more difficult than you think. Blownup pictures will not come out sharp enough. Same if you plan to print a big poster of the image or view on a 4k screen, not going to look great.

I did. Multiple times.

They came out sharp enough to have the plate be legible.

Also, 12MP is plenty for a decently sized poster. Unless you're framing art on a mega huge canvas, 12MP is ample for a standard print photo.

I will just say this. Photography enthusiasts are probably going to laugh at anyone who says megapixels matter more than anything else.

 

[chanchan05]

Why don't we just agree that picture quality is in the eye of the beholder.
Some people like their pictures sharp enough to cut, never mind some loss in detail and lighting, as long as they can zoom in and count somebody's nose hairs.
Some people like their photos to capture as much light play detail that they can tell who's shirt wasn't ironed that morning.

 

[D13H4RD2L1V3]

Why don't we just agree that picture quality is in the eye of the beholder.
Some people like their pictures sharp enough to cut, never mind some loss in detail and lighting, as long as they can zoom in and count somebody's nose hairs.
Some people like their photos to capture as much light play detail that they can tell who's shirt wasn't ironed that morning.

Now, that would be fine, but the problem is that one shouldn't treat a singular metric of objective camera performance over all else. Cameras work as a system and every factor and metric need to be taken into account.

I've said this before. I would be totally fine with 16MP or even 23MP if the sensor size was increased to compensate. Problem is that even with the added resolution, the V30 uses a seriously tiny sensor, which is smaller than even that used on the G6.

 

[Mooncatt]

V30 is even more processing dependant... They have to apply heavy noise reduction and sharpening to clean it up. It does ok with static images with defined borders and even coloring.. but textures and subtle details get absolutely clobbered.

I spent a good three weeks obsessively testing the V30's camera and in the situation you describe, it did well. Defining text like you say, is a function of the processing rebuilding those hard areas of contrast. But there's a flip side to all that.

One of my test pictures I often used is a printed calendar pinned to a wall by my desk. It does an excellent job of testing a cameras detail and resolving ability. When I pointed the V30 at it, I got will defined text and a nice, noise free image.

But I also lost reality. I take that same picture with my 2XL and I do, indeed, get sightly less crisp definition on the letters. But what I do see is texture... The subtle mottled look of the printer paper, the slight variation in ink of the letters... The depth of the threads on the fabric covering the cube walls. These are all details that are all but gone with the V30. The paper is a solid white, uniform text, digital artifacts where I saw weave in the fabric.

Taking pictures in my house I can see similar differences when looking at walls... Imperfections in the paint on the walls where I didn't bother to sand after priming are nowhere to be seen on the V30, but still persevered on my Pixel.

So yes, the V30 can be sharper, but what you are looking at in that image isn't what you shot.. all the warmth is lost. That little, solitary freckle near the corner of my son's mouth... Gone, wiped away when the noise reduction hammer fell.

That's why the Pixels keep scoring so high... Even though they may be a little softer, maybe with a little more noise, but they look real... The human eye is a finely tuned image processor that's evolved over eons to pick out the most minute details and patterns... Over-processed pictures, even though they are crisp and clear, trip a fuse in our brains telling us there is something not quite right about what we see.

Have you compared the V30 when saving as a RAW file? That would for sure not apply any processing. Of course this means you now have to edit the photo to be usable, but I'm guessing you'd get most of that detail back. On my V20, I'm almost thinking shooting in manual bypasses the noise reduction even when saving as jpeg+ RAW, but I'm not 100% sure.

 

[NationalTrust]

Can someone list, or point to a link that shows the various sensor sizes. (Pixel 2, V30, G6, latest IPhone, ETC)

 

[Mooncatt]

This chart gives a size comparison across the major sensor sizes. The V20 is 1/2.6" in size, so just a hair smaller than the smallest one in this chart. The 35mm "full frame" is the size of traditional 35mm film for reference

https://commons.m.wikimedia.org/wiki/File:Sensor_sizes_overlaid_inside.svg#mw-jump-to-license

 

[D13H4RD2L1V3]

This chart gives a size comparison across the major sensor sizes. The V20 is 1/2.6" in size, so just a hair smaller than the smallest one in this chart. The 35mm "full frame" is the size of traditional 35mm film for reference

https://commons.m.wikimedia.org/wiki/File:Sensor_sizes_overlaid_inside.svg#mw-jump-to-license

V20 is actually 1/2.8”, so slightly smaller than the V10.

But it being natively 4:3 on the IMX298 instead of 16:9 on the IMX234 means the pixel size is roughly the same

 

[Mooncatt]

V20 is actually 1/2.8”, so slightly smaller than the V10.

But it being natively 4:3 on the IMX298 instead of 16:9 on the IMX234 means the pixel size is roughly the same

My mistake. I had checked several articles before posting that, and they all were saying 1/2.6". I just found it on LG's site and you are correct, so someone put out the wrong info somewhere along the line.