Samsung's next trick in fighting large sensors - Motorola phone with 200MP camera leak

[Imbadatnames]

1 hour ago, That Franc said:

Supersampling. Also, smaller pixels and more pixels per picture mean much smaller, less noticeable noise grain - which should allow to iron it out without losing too much detail before downsizing the image.

Issue is more MP = less light = higher iso = grainy results = aggressive computation = shit image. Samsung is already struggling with this at 108MP

[LAwLz]

51 minutes ago, Imbadatnames said:

Issue is more MP = less light = higher iso = grainy results = aggressive computation = shit image. Samsung is already struggling with this at 108MP

See:  

2 hours ago, LAwLz said:

please note that the HP1 sensor can do 2x2 and 4x4 pixel binning. It will very rarely actually capture 200MP pictures.

 

Also, what makes you say Samsung "struggles" with this at 108MP? The Galaxy S21 Ultra takes some of the best pictures of any smartphone camera. They are hardly "struggling".

 

 

Edit:

Here is a comparison of the S21 Ultra (108MP sensor) vs the iPhone 13 Pro in low light:

 

[Imbadatnames]

15 minutes ago, LAwLz said:

See:  

 

Also, what makes you say Samsung "struggles" with this at 108MP? The Galaxy S21 Ultra takes some of the best pictures of any smartphone camera. They are hardly "struggling".

 

 

Edit:

Here is a comparison of the S21 Ultra (108MP sensor) vs the iPhone 13 Pro in low light:

 

Pixel binning doesn’t increase the amount of light it can collect. 4 pixels catching the same amount of light doesn’t mean you’re catching 4x the light. 
 

They’re using a higher iso and a faster shutter speed on the iPhone… it’s also known if you leave the iPhone on auto for low light to test low light it won’t do it properly. 
 

the Samsung is also full or artificial sharpening. 

[LAwLz]

33 minutes ago, Imbadatnames said:

Pixel binning doesn’t increase the amount of light it can collect. 4 pixels catching the same amount of light doesn’t mean you’re catching 4x the light. 

Ehm... Yes it does?

Not sure why you are saying it doesn't. Four 1µm pixels will capture the same amount of light as one 2µm pixel, minus the gaps between the pixels.

 

 

 

33 minutes ago, Imbadatnames said:

They’re using a higher iso and a faster shutter speed on the iPhone

And the picture from the iPhone looks way worse. Not sure what your point is.

 

33 minutes ago, Imbadatnames said:

it’s also known if you leave the iPhone on auto for low light to test low light it won’t do it properly. 

Not sure what argument you're trying to make. Are you saying the software on the iPhone is bad so we shouldn't look at test photos and just assume it has better hardware than the S21 Ultra?

If we start enabling things like night mode then we move away from evaluating the hardware and go more into the software, which is not what we want to do if we evaulate the camera hardware.

 

Did you ever think that maybe the iPhone takes bad photos in low light when using auto because the night mode uses very good software to make up for the lack of hardware? That if it had the same hardware as Samsung's phone it would take similarly good photos in low light even without night mode enabled.

 

33 minutes ago, Imbadatnames said:

the Samsung is also full or artificial sharpening. 

I'd take some sharpening artefacts over the smeary picture the iPhone 13 Pro took in this test any day of the week.

Also, I am not so sure those are sharpening artefacts. Are you sure you're not just used to really soft and smeary photos and think seeing details must be artificial?

 

 

I can link other comparison pictures too if you want.

Left is the S21 Ultra. Right is the iPhone 12 Pro. The S21 Ultra has waaay more details.

 

 

Can you at least admit that you were speaking in hyperbole and that the 108MP camera in the S21 Ultra does not take shit photos and it does not struggle?

You're having to speak out of both sides of your mouth to keep this opinion.

"High megapixel requires high ISO which means it gets grainy in low light!"

"Oh the S21 Ultra takes better pictures than the iPhone 13 in low light? Well it's just because it can run at a lower ISO!"

[Imbadatnames]

9 minutes ago, LAwLz said:

Ehm... Yes it does?

Not sure why you are saying it doesn't. Four 1µm pixels will capture the same amount of light as one 2µm pixel, minus the gaps between the pixels.

No it doesn’t you’re still capturing the same amount of light per pixel. Mashing the pixels together doesn’t create more data. Just adding them together doesn’t work. Changing my camera to take a photo at half res doesn’t magically make it better in low light. 

9 minutes ago, LAwLz said:

And the picture from the iPhone looks way worse. Not sure what your point is.

 

Not sure what argument you're trying to make. Are you saying the software on the iPhone is bad so we shouldn't look at test photos and just assume it has better hardware than the S21 Ultra?

If we start enabling things like night mode then we move away from evaluating the hardware and go more into the software, which is not what we want to do if we evaulate the camera hardware.

Phone cameras are ALL software. The phones aren’t programmed to take photos in low light without night mode. The hardware across the board, minus the Pro I, isn’t good because they all are using computational photography. 

9 minutes ago, LAwLz said:

 

Did you ever think that maybe the iPhone takes bad photos in low light when using auto because the night mode uses very good software to make up for the lack of hardware? That if it had the same hardware as Samsung's phone it would take similarly good photos in low light even without night mode enabled.

The phone is programmed to take advantage of night mode, something you’d never turn off. Again phones aren’t great hardware wise, kneecapping one in an unrealistic way to make the other look better (Samsungs night mode is abysmal) isn’t a good testing methodology. 

9 minutes ago, LAwLz said:

I'd take some sharpening artefacts over the smeary picture the iPhone 13 Pro took in this test any day of the week.

Also, I am not so sure those are sharpening artefacts. Are you sure you're not just used to really soft and smeary photos and think seeing details must be artificial?

 

 

I can link other comparison pictures too if you want.

Left is the S21 Ultra. Right is the iPhone 12 Pro. The S21 Ultra has waaay more details.

I mean you could just leave night mode on and let the phone function as intended rather than trying to make one look bad by not using the stock config? 

[LAwLz]

52 minutes ago, Imbadatnames said:

No it doesn’t you’re still capturing the same amount of light per pixel. Mashing the pixels together doesn’t create more data. Just adding them together doesn’t work. Changing my camera to take a photo at half res doesn’t magically make it better in low light. 

I think you have a fundamental misunderstand of how this technology works. Pixel binning is not the same as downsampling.

 

Adding them together does work, and it is exactly how it works.

If you got four 1µm pixels and each captures 10 electrons, the sensor will read that as 40 electrons.

If you got one big 2µm pixel that captures 40 electrons, the sensor will read that as 40 electrons.

 

Pixel binning happens in the sensor before the image is captured. That's why it works to just combine the charges of multiple pixels. It wouldn't work (like you said) if it was just software trickery where they just "mashed together" some pixels.

 

 

52 minutes ago, Imbadatnames said:

Phone cameras are ALL software.

No it's not.

If it was then nobody (including Apple) would bother making better sensors, since it's just software. And yet, phones like the Pro models takes better pictures than the non-Pro models even though they run the same software. You're once again talking out of both sides of your mouth. First you comment that the 108 megapixels in Samsung's camera are the reason why they "suffer", and then you turn around and say hardware doesn't matter.

Are you trolling or just doing terrible mental gymnastics?

 

52 minutes ago, Imbadatnames said:

The phones aren’t programmed to take photos in low light without night mode.

Yes, because Apple relies more heavily on software to make up for worse hardware. If they had better hardware (like Samsung) then they wouldn't have to rely so much on their software. 

I'm not saying that iPhones have bad sensors, but if we want to compare the iPhone's sensor vs the S21 Ultra's sensor (which is what I assume we want) then we should try and eliminate as much software variables as possible.

 

You don't compare an Intel CPU vs an AMD CPU by running Cinebench 7-Zip on the Intel processor and WinRAR on the AMD processor.

 

 

52 minutes ago, Imbadatnames said:

Again phones aren’t great hardware wise, kneecapping one in an unrealistic way to make the other look better (Samsungs night mode is abysmal) isn’t a good testing methodology.

I am not "knee capping" the iPhone. I am trying to keep the comparison as hardware focused as possible. If not having night mode enabled is "knee capping" then I am knee capping both phones equally because neither has night mode on.

 

 

 

 

 

I don't even know what your position on this is. Can you elaborate what exactly your thoughts are regarding phone cameras? You are flip flopping so much that it's hard to understand what you even mean. It doesn't help that you don't seem to understand the technologies you are discussing either.

[Spindel]

You all talk about the camera but the most interesting tidbit is the 125 W charger. 

 

Either they have a Li-ion solid-state battery, some other battery chemistry or you will be able to expect a really short life time of that battery if you use that charger. 

[LAwLz]

@Imbadatnames can you evaluate these three images? I'd like to hear your opinion on which one is the best.

They were all taken with night mode enabled, since you are so insistent on making the software as different as possible even though we are trying to evaluate hardware.

 

Spoiler

[Imbadatnames]

33 minutes ago, LAwLz said:

I think you have a fundamental misunderstand of how this technology works. Pixel binning is not the same as downsampling.

I’m not talking about downsampling ffs. You’re not understand how the HARDWARE works.

33 minutes ago, LAwLz said:

 

Adding them together does work, and it is exactly how it works.

If you got four 1µm pixels and each captures 10 electrons, the sensor will read that as 40 electrons.

If you got one big 2µm pixel that captures 40 electrons, the sensor will read that as 40 electrons.

My issue is how to dumb this down for you because you’re clearly not understanding. It doesn’t work like that. 

33 minutes ago, LAwLz said:

 

Pixel binning happens in the sensor before the image is captured. That's why it works to just combine the charges of multiple pixels. It wouldn't work (like you said) if it was just software trickery where they just "mashed together" some pixels.

 

 

No it's not.

If it was then nobody (including Apple) would bother making better sensors, since it's just software. And yet, phones like the Pro models takes better pictures than the non-Pro models even though they run the same software. You're once again talking out of both sides of your mouth. First you comment that the 108 megapixels in Samsung's camera are the reason why they "suffer", and then you turn around and say hardware doesn't matter.

Are you trolling or just doing terrible mental gymnastics?

you do know more pixels =/= a better sensor right? Unless you’re claiming the Samsung sensor is better than say the one in the new A7SIV because it has more pixels? 
 

It does because they have to use a LOT of cleanup on their images which is why they have  to amp up the colours to cover it. 

33 minutes ago, LAwLz said:

 

Yes, because Apple relies more heavily on software to make up for worse hardware. If they had better hardware (like Samsung) then they wouldn't have to rely so much on their software. 

I'm not saying that iPhones have bad sensors, but if we want to compare the iPhone's sensor vs the S21 Ultra's sensor (which is what I assume we want) then we should try and eliminate as much software variables as possible.

Not really because you’re selectively taking out the software to make one look better. 

33 minutes ago, LAwLz said:

 

You don't compare an Intel CPU vs an AMD CPU by running Cinebench 7-Zip on the Intel processor and WinRAR on the AMD processor.

 

 

I am not "knee capping" the iPhone. I am trying to keep the comparison as hardware focused as possible. If not having night mode enabled is "knee capping" then I am knee capping both phones equally because neither has night mode on.

It’s literally changing the default setting of the camera to make it look worse, how is that not kneecapping? 

33 minutes ago, LAwLz said:

 

 

 

 

 

I don't even know what your position on this is. Can you elaborate what exactly your thoughts are regarding phone cameras? You are flip flopping so much that it's hard to understand what you even mean. It doesn't help that you don't seem to understand the technologies you are discussing either.

You have literally no idea how the hardware works. Have you touched an actual camera in your life? 

[Imbadatnames]

16 minutes ago, LAwLz said:

@Imbadatnames can you evaluate these three images? I'd like to hear your opinion on which one is the best.

They were all taken with night mode enabled, since you are so insistent on making the software as different as possible even though we are trying to evaluate hardware.

 

  Reveal hidden contents

Not even the original photos fml

[DildorTheDecent]

>125W charging

 

What a time to be alive.

[TOMPPIX]

https://drive.google.com/file/d/15o4CsM0-pt3iAnOxuyvPAE65kfERbanA/view?usp=drivesdk

Some photos i took with 50 ISO 1/8 Shutter on both pictures

Mi10t pro 108mp camera

Feel free to delete if this is against some rules.

[LAwLz]

52 minutes ago, Imbadatnames said:

I’m not talking about downsampling ffs. You’re not understand how the HARDWARE works.

Yes I do.

 

52 minutes ago, Imbadatnames said:

My issue is how to dumb this down for you because you’re clearly not understanding. It doesn’t work like that. 

Yes it does.

 

52 minutes ago, Imbadatnames said:

you do know more pixels =/= a better sensor right?

Yes I understand that.

Do you understand that more pixels =/= worse sensor?

That there are valid reasons for pushing for more pixels, such as the things I linked to earlier in this thread?

 

54 minutes ago, Imbadatnames said:

Unless you’re claiming the Samsung sensor is better than say the one in the new A7SIV because it has more pixels? 

Of course I am not claiming that. Pixels aren't everything, but there are benefits to it. You seem to think that more pixels = worse sensor which is not the case. It is more nuanced than that.

 

55 minutes ago, Imbadatnames said:

It does because they have to use a LOT of cleanup on their images which is why they have  to amp up the colours to cover it. 

What are you on about?

They don't "amp up the colors", and "amping up the colors" wouldn't do things like reduce noise.

 

56 minutes ago, Imbadatnames said:

Not really because you’re selectively taking out the software to make one look better. 

I am "taking out the software" on both cameras because we want to compare HARDWARE, not SOFTWARE. Do you understand the difference?

Did you not understand my analogy between AMD and Intel, where you run 7-zip on one and WinRAR on the other?

We want as little post-processing as possible when comparing the sensors. What you are saying is "no, let's add a bunch of post processing and software differences to our hardware test".

That's not how you compare hardware, which is what we are comparing. We are not comparing software, so the software should be as equal as possible to take it out of the equation.

 

 

58 minutes ago, Imbadatnames said:

You have literally no idea how the hardware works. Have you touched an actual camera in your life? 

lol, why are you so mad?

I have a fairly good understanding of how the hardware works. You on the other hand don't seem to understand it since you start talking about nonsense like lowering the resolution to "magically get better images". That's not how pixel binning works.

 

59 minutes ago, Imbadatnames said:

Not even the original photos fml

What are you on about? Why do you say they aren't the original photos?

I took the original photos, cropped out the same section and then put them in a single photo.

Are you afraid to do the comparison because you have talked so much shit about Samsung? It would be pretty embarrassing if you picked a Samsung picture as the best looking one after having talked so much shit about them, right? Especially now that you can't blame it on how the iPhone is "knee capped" by turning off as much software post processing as possible for all cameras.

Shouldn't it be easy to pick out the S21 Ultra picture from those three since Samsung's phones apparently struggle so much because of the 108MP camera?

 

I can post the full pictures after you have decided which one you like the most, if you want.

[LAwLz]

21 minutes ago, TOMPPIX said:

https://drive.google.com/file/d/15o4CsM0-pt3iAnOxuyvPAE65kfERbanA/view?usp=drivesdk

Some photos i took with 50 ISO 1/8 Shutter on both pictures

Mi10t pro 108mp camera

Feel free to delete if this is against some rules.

Only 10MB? That's surprisingly small for a 108MP picture.

I wonder if Xiaomi applies higher JPEG compression to keep file sizes reasonable, or if the subject you captured just lends itself really well to compression (not a whole lot of color).

 

The S21 Ultra sits at around 30MB for a 108MP image.

[TOMPPIX]

Some interesting pictures comparing the 108mp camera to a Lumix GX80

https://www.dpreview.com/forums/thread/4486119

It's gonna be interesting to see what the 200mp sensor can do.

[TOMPPIX]

1 hour ago, LAwLz said:

Only 10MB? That's surprisingly small for a 108MP picture.

I wonder if Xiaomi applies higher JPEG compression to keep file sizes reasonable, or if the subject you captured just lends itself really well to compression (not a whole lot of color).

 

The S21 Ultra sits at around 30MB for a 108MP image.

Yes I think it compresses the photos a bit too much, would be nice to get a third party app with 108mp support. Found one.

https://www.celsoazevedo.com/files/android/open-camera/dev-wyroczen/f/dl0/

[Imbadatnames]

5 hours ago, LAwLz said:

What are you on about? Why do you say they aren't the original photos?

I took the original photos, cropped out the same section and then put them in a single photo.

Are you afraid to do the comparison because you have talked so much shit about Samsung? It would be pretty embarrassing if you picked a Samsung picture as the best looking one after having talked so much shit about them, right? Especially now that you can't blame it on how the iPhone is "knee capped" by turning off as much software post processing as possible for all cameras.

Shouldn't it be easy to pick out the S21 Ultra picture from those three since Samsung's phones apparently struggle so much because of the 108MP camera?

 

I can post the full pictures after you have decided which one you like the most, if you want.

I’ll address this as it shows your greatest faux pas. You don’t do that when you’re asking to compare photos. For a start they’re not full size, they’re just a pixelated mess when you zoom in so you actually can’t compare them. Personally the easiest way to see which is a Samsung is to get a neutral photo from an actual camera and see which is furthest away colour wise and that’s the Samsung, apple are generally more true to life in my experience so that would be closer. 
 

They’re camera phone photos they’ll all be bad, there’s a reason why I have and use a Mirrorless system when I want a good photo rather than my phone which is mainly just for sending photos of work to people and taking photos of documents for reference. 
 

 

[LAwLz]

14 minutes ago, Imbadatnames said:

I’ll address this as it shows your greatest faux pas. You don’t do that when you’re asking to compare photos. For a start they’re not full size, they’re just a pixelated mess when you zoom in so you actually can’t compare them. Personally the easiest way to see which is a Samsung is to get a neutral photo from an actual camera and see which is furthest away colour wise and that’s the Samsung, apple are generally more true to life in my experience so that would be closer. 
 

They’re camera phone photos they’ll all be bad, there’s a reason why I have and use a Mirrorless system when I want a good photo rather than my phone which is mainly just for sending photos of work to people and taking photos of documents for reference.

Good thing I got this picture then, which was taken at the same time as the others, of the same subject. It's taken with a Fujifilm mirrorless camera with an APS-C sensor.

Feel free to use that as a reference to determine which of the three other pictures are the best.

 

 

[Dracarris]

16 hours ago, Gamer Schnitzel said:

MP literally does nothing for photo quality at this point which is why it is actually turned off by default when you use the camera. The default lense is probably going to be 16 MP and you have to press an extra button to use the high MP lens.

As you said, MP does not do anything for picture quality if lense and sensor size are the real bottleneck. It's like streaming an 8K video with a typical 1080p bitrate. The actual information is simply missing, no matter into how many pixels you subdivide it.

 

There's a very good reason professional cameras with vastly larger sensor sizes have nowhere near the MPs discussed here. If this would be actually beneficial to picture quality they would've long adopted it.

 

This is again the big numbers=equal better=braggin rights thing. And it's getting absolutely ridiculous at this point.

[Inkz]

7 hours ago, DildorTheDecent said:

>125W charging

 

What a time to be alive.

Honestly I am scared for the battery's long term health if it regularly charges at that speed. 

[Imbadatnames]

11 hours ago, LAwLz said:

Good thing I got this picture then, which was taken at the same time as the others, of the same subject. It's taken with a Fujifilm mirrorless camera with an APS-C sensor.

Feel free to use that as a reference to determine which of the three other pictures are the best.

 

 

Are you linking the other photos or not? Also that image is only 1000x480 can I have the original image please 

[mr moose]

13 minutes ago, Imbadatnames said:

Are you linking the other photos or not? Also that image is only 1000x480 can I have the original image please 

 

You have nothing to fear except choosing the Samsung image as the better one and then finding out the photos are legit and your claims are bogus.  

[BetteBalterZen]

But ey, who the hell cares what sort of black magic the "camera" does? If one photo is better than the other, then it is better. 

[Imbadatnames]

31 minutes ago, mr moose said:

 

You have nothing to fear except choosing the Samsung image as the better one and then finding out the photos are legit and your claims are bogus.  

There’s no point comparing images that are 33x smaller than the original 

[mr moose]

1 minute ago, Imbadatnames said:

There’s no point comparing images that are 33x smaller than the original 

 

No of course not.  You can't see them.