Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

A similar thing I've noticed in the last ~6 years is in regards to smartphones, and in particular their cameras.

Android is pretty standardized these days, and even a fairly cheap Android phone will probably do most smartphone things (make calls, send texts, watch YouTube, etc.) reasonably well, but a corner that seems to be cut in cheaper phones is almost always the camera, and that kind of makes sense.

If you're buying your phone online, you can look at the more-or-less objective specs of the phone like how much RAM it has, or how fast the CPU is, but you can't really tell how the photos will look [1], and as such it's easy to put a cheaper sensor or lens in there, especially in the cheaper phones which (I think) have lower margins.

[1] Even if the listing has sample pictures, you don't know how representative they are of the actual experience; they could just have taken the sample photos with ideal lighting, with a tripod, in a controlled studio, for example.



I realized exactly this after some frustration with a few phones. I found GSMArena that has extensive reviews including sample standardized photos.


GSMArena is such a valuable resource.


And with a history at that. They have reviews dating back to 2004. It's fun to check out what were the biggest concerns throughout the years.


Yes, and those crappy sensors will still be 200 megapixels (which will be used as an advertising point), as most people don’t realize that a fantastic 24 MP camera is better than a crummy 200 MP one.


Those 200 MP sensors are 200 MP in name only. They're so small the individual pixels are smaller than the wavelength of light they're trying to capture. That's why when you use them, the max image size is usually around 50MP because binning is required physically


That is insane (not your comment, the fact that the sensors are that small)!

Here [1] is the first result when I searched, it's the powerfully-named Omnivision "OVB0B" sensor.

It has a resolution of 16384x12288 pixels (=201,326,592 pixels in total) meaning individual pixels are 0.61 µm. Visible light is generally around 0.4 µm to 0.7 µm so the sensor is definitely in that size region with its pixels. Wow. Thanks for pointing this out.

[1]: https://www.ovt.com/products/ovb0b/


Worse; the crappy sensors will be only 8 megapixels or less, but the camera software will inflate that with AI to 200 megapixels.


This has been happening for ages in the trail camera market.


I'm happy that the new metric is sensor size. A bit of an upgrade in metric that mildly correlates to picture quality.


That's good as long as they quote the sensor size in millimetres for the width and height (e.g. 36 mm × 24 mm "full frame").

What is atrocious is the inch format - e.g. 1 inch sensor, 1/2.2" (decimal and fraction!). It somehow refers to the diagonal length of the vacuum tube and not even the active image area. https://en.wikipedia.org/wiki/Image_sensor_format#Table_of_s... , https://en.wikipedia.org/wiki/Video_camera_tube#Size


True, the lens is usually more important than the sensor, remember that the hubble space telescope is only one megapixel.


and the speaker is 300 watts*

*300 watts for 1e-20ns


Yeah, definitely; megapixels are a more or less objective metric, but they're only one factor to image quality, and nowadays not even really the most important one.


Pixel count is objective, but never meant much. Sensor area is much more important, but doesn't make for a sexy number, and it often doesn't make it into the spec sheet.


It is easy for programmer-types to think that pixel counts are objective, simply because familiarity with units, or perhaps because pixels are discrete countable physical locations on a sensor.

An optical system composed of multiple lenses doesn't obviously correspond to a certain resolution, but the confusion is merely the literature taken at hand: instead of obsessively comparing smartphone advertisements, it is physically possible to read books and course notes on physics, optics, ...

There is nothing impossible about publishing spatial frequency responses etc, or legally mandating such information to be present on any advertising materials as soon as the MPixel counts are advertised.


> pixel counts are objective

Pixel counts are objective. But the problem is that while the pixel count sets the upper bound on the image quality, all cameras don't achieve that upper bound - especially due to the Bayer color filter matrix.

To hit this point home, try taking a photo on a smartphone with a 12-megapixel camera. Now take the same photo on a professional camera, say the Canon EOS R5 with an expensive L-series prime lens. The R5 photo has 45 megapixels, so go and downsample that to 12 megapixels to match the phone photo. The pro camera photo will look much better than the phone photo, thus demonstrating that the phone camera is not maximizing the limits of the 12 MP picture format.

(Also, to make the comparison fairer, take a landscape photo of faraway scenery and buildings so that there are no effects of focus blur from having a large lens aperture. Everything in the photos should be in sharp focus.)


None of this negates the fact that the spatial frequency resolution plots for the lenses, or lenses+sensor combined system would adequately describe these facets under discussion. So what if a manufacturer with their 4X monochromatic pixels that really only constitute just X color pixels, but insinuates the 4X numbers to be the full color pixel count? The spatial angular frequency plots of the compound lens(es)+sensor would quickly become the "figure of merit" to check regarding these facets. Of course nothing would prevent them from stating lies and providing false plots, but then any laboratory could start underwriting (measuring and publishing) such resolution specifications.


I’d be curious re your thoughts on something like a iPhone 16 vs my d6 MkII with a 24-70 or 200 EF lens. The telephoto obviously wins at distance but for family portraits I’ve found I get better photos with the phone and I’m not sure why. I chalk it up to iPhone image processing voodoo vs my setting choice and raw editing skills.


Pixel 7 pro phones latest camera was the biggest disappointment for me. I usually have been happy with pixels but it has trouble focusing on infinity, and the image stabilization on the 5x zoom lens is so jumpy it's worse than nothing at all. I thought about trying to get it warrantied or something but was already sick of spending hours at the Verizon store trying to get my trade-in credit.

And to get the full advertised resolution (which you have to enable in the photo toggles) you have to tolerate a delayed shutter button and a "long exposure" for it to do the computational photography or something. So it's useless to expect full resolution with kids or action.


The Pixel 7 Pro was the worst phone I've ever owned. The battery was terrible, the camera was terrible, sometimes it just wouldn't boot up, everything about it was awful. I swore an oath that I wouldn't buy a Pixel phone ever again.

I've heard that the Pixel 8 was better, but I'll never know.


Fellow P 7 Pro owner here: yeah the 8 Pro is better. Still wish I'd got something else, but at least audio through the USB-C port doesn't have a buzzing caused by the clock on the chip.


I just went back to iPhone for the time being. When this phone breaks I might give Android another try but it will definitely not be a Pixel.


In all fairness, I did recently take a 30x zoom video of an animal using the pixel, and it got compliments specifically on the zoom capability of the phone when I shared it to the people who had been there at the time.


The camera (and/or the camera software) was awful. There was no way to get a good photo out of it, even with 3rd party apps. And so many cameras now won't allow 3rd party apps access to the full data either, so you are stuck with the junk-ridden garbage that comes with the phone.


I dumped a decade of android after the pixel 6 and 7 due to the pathetic camera. It couldn’t take an picture of a flat piece of paper without distortion!


The most blatant example of this is Samsung phones artificially replacing photos of the Moon with a high-resolution composite.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: