Hacker News new | ask | show | jobs
by jakecopp 2399 days ago
If all the advancements in smartphone photography is in software, why aren't DSLR/mirrorless manufactures doing it too?

I don't want my camera to have a touchscreen/social media/wifi but it'd be cool if Adobe Camera Raw/some alternative could do this stuff!

3 comments

IMHO they've basically missed an opportunity here for many years. They'd be in a perfect position to offer those things, combined with a much better/larger sensor, which enables even better images. On smartphones it's a matter of necessity, as the sensor is (fairly) crappy in comparison, but on a DSLR it could still be a benefit. Personally I'd be perfectly happy to get a pre-processed DNG from the camera instead of having to do this afterwards. And then give me the raw files to do it manually as well.

Perhaps they're trying not to cannibalize their lower market segments or think that professionals would never use those things (on which they might be correct). But I can definitely see that computational photography beyond raw->JPEG conversion with a color profile could have its place in a DSLR.

While it would be nice to load some post-processing script onto the camera and get all that done by the camera without having to bother with it yourself, all that processing will cost battery life, as well as response time. Having to charge your DSLR every day is a huge cost in convenience, and making a dedicated imaging machine unresponsive would be terrible.

Additionally, you are always going to move your pictures out of the camera for proper viewing, so it doesn't make a lot of sense to provide the best possible picture already in the camera.

I don't know if the software they bundle with the camera is any good at this computational photography though (I've never checked to see if there are linux versions), but it better be if it indeed helps image quality.

Yeah, it's definitely nice to get ~1000 images per battery charge. But when uploading to the PC infrequently you'll then have to sort photos into groups to process individually (something I already hate with panoramas). Personally it's something I'd rather not do, at the expense of only getting ~200 photos per charge. It's also not necessary to compromise, as those special modes would be, well, special modes. So to preserve battery I could just as well shoot raw as normal.
Idk, I feel like we have far better software on our computers, and mirrorless/dslr's will always beat our phones in terms of raw specs.

This is more of a "Now anyone can capture decent astrophotography with just their phone!" then some revolutionary new thing if that makes sense? Basically they have to push their phone's sensor as far as it'll go and use software to remove the noise, instead of using a better lens/sensor setup (which is space and cost prohibitive for a phone).

You're still welcome to do extensive post-processing on your computer (I don't want my camera doing any processing), and indeed that's what any astro-photographer will do if they wanted.

That said I've captured some amazing night/moon photos with my A6000 that I have never been able to achieve before, and that's without any extra processing.

> why aren't DSLR/mirrorless manufactures doing it too?

Most of them are historical companies and are probably very old school / slow to adapt. Google has probably access to better software engineers than Nikon or Canon which seem to barely be able to develop a working bluetooth/wifi sync.

They think they're still hardware companies in a hardware-first world. The problem is we've started hitting some limits in the hardware which software can improve on dramatically, but they are terrible at software (for the most part).