Definitely, the color detail looks straight out of a postcard — shall we say, just like back in the PureView days?
The camera issue is actually pretty well known. Our main camera is quite good, but it’s still waiting for some fixes, so I can’t really say that the camera itself is poor quality.
As for the background blur, what aperture does this camera have? We could probably tell from that whether the background blur can actually be improved or not. Or, actually, SFOS has a landscape mode — how about trying a shot with that?
I have hoped this is software thing, which hopefully will be improved. To my knowledge almost full camera stack is open source so I hope so wizards from community would give it some love
A little off-topic, but today I noticed that the sound quality of recorded videos is very bad. I did several videos today of a rock/metal concert - and there are so much sound details missing that it’s no fun watching that.
iPhone 13 Pro which I have for work phone does alter the image via AI rendering (and I’ve seen it do weird things to the photo when a bit out of ordinarie is photographed) thus making it debatable how much of your scene is a real photo of the location.
I just downloaded HedgeCam 2 from Aurora Store and the difference between the amount of detail compared to the native camera app is big. It’s night here in Finland so I can’t take similar car photo, but this is 100% software related problem. I’ll post some test shots tomorrow.
It would be very interesting to have the photo metadata to determine the following: focal length, shutter speed, and aperture.
In fact, with the same lens on a camera (especially an SLR with a depth-of-field tester), you get very different results depending on the aperture and shutter speed you choose. And in the tests, these details should be provided for each photo. I’m not sure it’s possible to “objectively” compare a photo taken with the JA2 to the same photo taken with another smartphone, such as the XA2.
To make a fair comparison, we’d need to be able to force each smartphone to take the photos at the same shutter speed and with the same aperture. As far as I know, this isn’t possible because we’re in “auto” mode, which we can’t really “tweak.”
It’s true that being able to retrieve the raw files would be a real plus
I totally agree, it would be much more scientific and rigorous.
Nevertheless, even as is, the miss of details, the colour treatment and “repainting” thing is pretty obvious.
Especially when compared to a 2018 phone (XA2), would I say.
Also the native Advanced Camera seems to do a better job.
But this app produces 50 megapixels pictures, where the stock cam produces 12,5 megapixel, IIRC.
Is that the only cause of the difference or is there really another treatment involved?
To make a fair comparison, perhaps should you try HedgeCam 2 with the same resolution as the stock app.
That is indeed interesting - no only it seems to have HDR stacking in the second image (or just HDR in-sensor, without stacking) - but there are many more details! Gives one hope
In the first comparison I was worried the “blur” is just weak sensor vs iOS oversharpening (which is also bad).
Maybe a little off-topic, but what’s HedgeCam? I see it’s a fork of OpenCamera which I’m familiar with, but I can’t find what it changes/improves over OpenCamera anywhere.
E: also why does it have in-app purchases? None of the descriptions mention this either.
tl;dr I’d just like to say that while an actual aperture will influence DoF, we can’t be sure there is one, and comparing depth-of-field as computed from the f stop recorded in EXIF may be misleading.
This is because the aperture openings become very small when the focal length is short, causing deteriorated image quality. Probably also because an aperture and related actuators (as we know them from SLRs) might be tricky to build and to fit inside a phone camera.
The EXIF info may report different apertures, but these are usually achieved with grey filters (ND) rather than by changing the diameter of an aperture. (Some digital compact cameras do use a single aperture at (say) f/5.6 combined with ND filter(s) to achieve other “f stops”.)
