iPhone X(S)(R)(MAX)

OldCorpse

Rockin the Boat
Computational image acquisition is marching apace. Both stills and video. I hate to say it, but unless dedicated cameras catch up with this new techology, they'll start falling behind. Thoughts?
 
I don't think it will fall behind because it lacks those kinds of algorithms/AI/Computational etc. Why? Because in photography and video for professional use there is a bigger emphasis on "pure/clean" images rather than half-assed compromises, and the work flow usually requires some handiwork anyways like removing blemishes, dirt, etc... That said some of the pro-cameras have built in HDR-stacking, pixel shift, but I think most people still do those things manually if the plan is to get "the best picture".

There are several cameras that have had stuff like pixel shift, HDR from RED's by mixing the 1/48th shutter picture with a quick 1/1000th etc... None of them have been game changers, but mostly been of the sorts: "I can only get reliable and good results if these very specific variables are OK" (usually scenes with limited movement, i.e. still life).

Phones on the other hand are much more for pictures that needs to be uploaded right away, i.e. no time or space to edit and fine tune.

Might be different for newspapers that need good pictures fast, but then the pro-cameras have plenty of DR around 12-14 stops, I don't know what normal phones have, but I would guess 10-11-ish without AIs, so they could benefit from a boost on 2-3 stops...

Also, MARKETING, there is no longer much sense in saying this phone has 20mp when others have 40mp, etc. So what is then the next selling point, low light and HDR since HDR is the new 3D. Everybody is getting HDR TVs, then everybody wants something to create HDR with...
 
Last edited:
Yeah, that's been the general consensus since cellphone cameras started getting serious. BUT. At some point, the tech becomes so advanced that the only thing left to differentiate is the talent of the operator. In the same way that despite the m4/3 chips being much smaller than FF, they are at this point sophisticated enough technically that the advantage of it being FF has really diminished on purely technical grounds for the human eye (i.e. not pointless pixel peeping). And that's the issue - at some point, the tech becomes so sophisticated that you really DON'T need a big dedicated camera to shoot most commercial stuff, whether stills or even video. We're not there yet, but note how folks like Soderbergh already shot features on relatively primitive phones like the iPhone 7. What happens when the cellphone cameras become much, *much* better? There comes a point where there is a cross-over effect. So far at least, the old guard camera makers are at a disadvantage vs. tech giants like Apple when it comes to sophisticated data processing. There really is a danger that one day your Alexa becomes irrelevant - what do the camera manufacturers do then? Canon and Sony will survive because they have other businesses, but what of Nikon, Olympus and the more pure play guys? Btw. speaking of stills, I thought it was a pretty neat trick in the latest iPhones where you can take a photo AFTER it was taken and easily adjust depth of field afterwards with just a simple slider... baby steps, but Arriflex, Nikon, Panasonic etc. should be worried, IMHO. Maybe not today, but perhaps not too long from now. Who knows, I guess we'll see! Exciting times, at least for the end user!
 
It will definitely take over if dedicated cameras don't get it together.

It's getting rough in multiple spaces. The EVA1 isn't 100% than the GH5S, and the line given is they are separate departments. All props to the crew and I really like my EVA1 and my GH5/GH5s but the affect of CPU and software in undeniable. Watching the iPhone XS presentation today, I thought the same thing.

Those who say it will never take it have no imagination for how quickly things are progressing. Half baked today, perhaps. In 5 years, not so much.

However, if higher end cameras can start incorporating these software smarts and high end CPUs as options they will always be able to stay ahead... enough to really matter for many people though is another question entirely.
 
At this point the way something is lit plays a huge role in the separation of high and low end. Cameras already look fantastic in some scenarios where lights aren't added by a production but the natural look is still only appropriate part of the time. Painting with light and creating whatever look is called for is still a huge physical limitation that computers (at this point) can't compete with. Physical as in the actual instruments themselves, as well as crew members to set up the lights and accomplish other tasks on set.

I think once any or several lighting situations can be simulated in a believable way is the real day people are in trouble.
 
I don't think they'll surpass, but they're definitely closing the gap, and very fast. At a recent meetup a friend of mine was using a $200 Huawei phone and the pictures she sent us, with computational depth blur, looked stunning. The reasons for regular people to get a big-sensor camera are getting thinner and thinner every year.

But craft will keep pros alive. Because yes, lighting and staging and framing and camera motion all make a difference.
 
Let's keep in mind what the subject is - it's camera capabilities and how do cellphones stack up against pro equipment. It is not about how amateur OPERATORS stack up against pro operators. Yes of course the pros will light better and have better visual ideas and better lighting equipment etc. - but that has nothing to do with the equipment. We are discussing specifically the camera capabilities - it's not like pro lighting can only be utilized with pro cameras. You can use pro operators and pro lighting and so on with BOTH cellphone cams and traditional pro cams - after all, Soderberg used pro folks with the iPhone 7.

All I'm saying is that the day may come where when folks are discussing image acquisition, the DP may not have a strong preference for an Alexa or Sony dedicated video cam, and the cellphone camera might even have an advantage due to size. The original question was: are pro camera OEMs meeting the challenge of the cellphone, or are they asleep at the wheel - if the latter then they'll get bypassed the same way steam engine OEMs were bypassed by better gasoline engines. My feeling is that the pro cam OEMs are not quite appreciating how fast tech is moving and that they are not actually investing in software and data processing capabilities to meet that challenge, or perhaps simply don't have the expertise. Today data processing is a gigantic part of image acquisition, and the cam OEMs are still rolling along traditional lines as if we're still shooting film. They need to re-thing not just data processing but the whole stack from form factor to multilenses - there is a feedback loop between all these factors. Instead, the cams still look like they looked back in the 60's. The image acquisition tech has seriously moved focus and it seems the cam OEMs are not waking up fully to the challenge. YMMV.
 
Let's keep in mind what the subject is - it's camera capabilities and how do cellphones stack up against pro equipment. It is not about how amateur OPERATORS stack up against pro operators. Yes of course the pros will light better and have better visual ideas and better lighting equipment etc. - but that has nothing to do with the equipment. We are discussing specifically the camera capabilities - it's not like pro lighting can only be utilized with pro cameras. You can use pro operators and pro lighting and so on with BOTH cellphone cams and traditional pro cams - after all, Soderberg used pro folks with the iPhone 7.

What I find ironic, is that when we have lit sit-down set-ups and someone comes in and takes a pic(selfie, etc)of the subject, talent, etc. in the light, the cell phone pics are crappy(usually blown out). But when they get out of the light and "just let the phone do it's thing", they are 10x better. I think all the algorithms, etc. are designed for "natural" and "unlit" and when they're in situations that are lit for manually controlled cameras they don't do well, at least not in their default "auto" modes.
 
The thing about digital processing is that you can adapt the process for anything. Today, yes, the algorithms are designed for specific situations, i.e. what most consumers would use the phones for. That doesn't mean those can't be changed at a drop of a hat. It's exactly as when folks developed digital profiles to mimic film - you can create profiles that can mimic specific film stock at specific ISOs. But then the further we move away from actual celluloid, and the more people are used to digital, the compulsion to mimic 35mm celluloid film look will pass into history and a new aesthetic will develop based on the capabilities of the tool. Remember back in the day, you had to have a specific lighting scheme because film stock was so slow - then as faster stock was developed lighting schemes changed. So too with today's lighting. Today's lighting is just as dependent upon the capabilities of the tools, and when the capabilities change, so do the aesthetics. After all, nobody accused newer cameras with faster film stock of not mimicing the look of movies from the 40's and lighting for slower stock. The aesthetic evolves together with the capabilities of our image acquisition tools. The intense real time data processing is revolutionising the capabilities and hence the aesthetic - no use crying about losing the lighting schemes of yesterday - tomorrow is a new day with new aesthetic, that's not a pity, that's natural evolution... it happend in the past and will continue to happen in the future. The only question here is who gets to play the game - are today's camera OEMs going to evolve or are they going to go the way of the Mitchells when technology passes them by. My point is that I see a worrying lack of effort on the part of traditional camera makers in face of the cellphone challenge. Opinions of course will differ. I guess we'll just have to see.
 
I think cell phone cameras are still far off from being practical for cinema. Neither image quality or form factor makes sense. Thats why DSLR's and such still aren't used in big budget productions, they just aren't practical even if the image is fine.

SLR's still aren't up to par with Arri or Red, though its getting close (finally).

As a professional I'd doubt I'd ever use a cell phone for photography. For travel yes, but not for other jobs. An example of my point is how I actually prefer my GH5 body to the XT2 I recently got. I miss my bigger GH5, it was simply more comfortable to hold. Smaller isn't always better. It would be super annoying shooting something on an Iphone with no manual controls.

I do agree eventually the camera market is going to become more niche.

What I find ironic, is that when we have lit sit-down set-ups and someone comes in and takes a pic(selfie, etc)of the subject, talent, etc. in the light, the cell phone pics are crappy(usually blown out). But when they get out of the light and "just let the phone do it's thing", they are 10x better. I think all the algorithms, etc. are designed for "natural" and "unlit" and when they're in situations that are lit for manually controlled cameras they don't do well, at least not in their default "auto" modes.

If they used manual settings they could get a better image. Depends on the phone though. My phone isn't that new, but I was trying to take a picture of the sun set, and it just looked bad. Was kind of pretty but looked totally unrealistic to what I was seeing due to lack of dynamic range in the shadows and highlights. I switched to my XT20, kind of expecting a similar result. It was reassuring though, the XT20 took an image that looked just like what I was seeing.

The one good thing about the competing phone tech is that it will force the "camera" companies to be more competitive as well with their products, which is good for us consumers.
 
Last edited:
Better HDR, resolution, and those sorts of IQ improvements aren't worth much for professional use since they will have limited quality assurance, and nobody likes dealing with weird artifacts in pictures. The thing is that all companies more or less are already down this road in terms of AI auto-focus, they all have the programmers and the tech, maybe not on the same level as Apple's RnD, but they aren't sitting idle by. The only thing I think "we" the professionals are looking forward to is good depth analysis which allows for selecting layers based on depth (it would be a game changer for roto-scoping, green screens would be history for a lot of shots).

I'm sure as hell not against the new innovations, but I don't see the computational power being enough in the near future. Moore's law isn't the same as it was 15-50 years ago.

For ENG's it might also be a source of "integrity" to not let AI push pixels around, but that depends on the level the AI is altering the picture.
 
Better HDR, resolution, and those sorts of IQ improvements aren't worth much for professional use since they will have limited quality assurance, and nobody likes dealing with weird artifacts in pictures. The thing is that all companies more or less are already down this road in terms of AI auto-focus, they all have the programmers and the tech, maybe not on the same level as Apple's RnD, but they aren't sitting idle by. The only thing I think "we" the professionals are looking forward to is good depth analysis which allows for selecting layers based on depth (it would be a game changer for roto-scoping, green screens would be history for a lot of shots).

I'm sure as hell not against the new innovations, but I don't see the computational power being enough in the near future. Moore's law isn't the same as it was 15-50 years ago.

For ENG's it might also be a source of "integrity" to not let AI push pixels around, but that depends on the level the AI is altering the picture.

Just look at the rules for Reuters for stills. Only camera originated JPEGS(cropping and adjusting levels to be correct are allowed).
...Pictures must reflect reality.

https://petapixel.com/2015/11/18/reuters-issues-a-worldwide-ban-on-raw-photos/
 
Just look at the rules for Reuters for stills. Only camera originated JPEGS(cropping and adjusting levels to be correct are allowed).

https://petapixel.com/2015/11/18/reuters-issues-a-worldwide-ban-on-raw-photos/

Yeah I know they are extra hard, but they re-sell their images and need that policy for various reasons, newspapers can decide for themselves what level of "purity" they want, and I think some of them might not be as rigid as others. But still, that's my point that a lot of the professional users can't use "unrealiable" algorithms...
 
At this point things might be headed in the other direction, where news outlets basically fabricate everything and each segment begins with the text - Based on a true story.

As for the cameras; I just picked up an iPhone 8 Plus the other day. I can now shoot 24, 30 or 60fps 4K with optical image stabilization. This is one year old technology and not even the best on the market back then. There is no reason for me to take a DSLR with me when I'm at family or friend functions or at some fair or event that I'm not being paid to shoot. Once these things have global shutter, even better. Even now they blow away every other dedicated video camera I used to shoot casual stuff with. Aside from maybe zoom capability. Heck if I had the choice between this iPhone and the XL1 back in 2003 when I shot my first indie feature, I would have gone for the iPhone. We are spoiled with image quality these days. But that's the consumer world. For the pro world, we still need good ergonomics and mission critical tools. Despite everything I just said I'm not showing up to a paying job with my new iPhone as my A camera. I think that will remain the case for a long time.
 
Overall the Xs feels like the most underwhelming Apple upgrade for a smartphone ever. To be fair the X is not even 12 months old (it launched in November) and was cutting edge (and as I recall they stretched hard to even get it out on the 10th anniversary) so it should not be a huge surprise that not much has been enhanced in 10 months or so.

Still, I have upgraded at every annual increment for years and I think I’ll pass. Normally the processor bump is a 30 second mention in the keynote - this year they spent 20 minutes trying to bolster the CPU benefits and it felt weak. Even the exclusive adjustable DOF is a quasi feature that could be included for iPhone X users. iPhone X can already toggle the blur on off, it’s just an adjustable blur matte but they dressed it up as if it’s an A12 exclusive. Whatever.

Dual SIM, ip68, and 512gb are all interesting but can’t justify the price and the phone has been fast enough for years that I really don’t care about a bump. Larger photo sites are really compelling to me though, but, that’s it.

In 2019 I’d love to see improved battery, reception, reduced weight, and the rumored ultrasonic Touch ID under the the screen surface to bolster Face ID. I actually really love Face ID but I’ve also never entered my passcode manually as much as I have in the last year and it would be wonderful to have both options - you could select either or for most people, or both and for high security govt and corporate.
 
Overall the Xs feels like the most underwhelming Apple upgrade for a smartphone ever. To be fair the X is not even 12 months old (it launched in November) and was cutting edge (and as I recall they stretched hard to even get it out on the 10th anniversary) so it should not be a huge surprise that not much has been enhanced in 10 months or so.

Still, I have upgraded at every annual increment for years and I think I’ll pass. Normally the processor bump is a 30 second mention in the keynote - this year they spent 20 minutes trying to bolster the CPU benefits and it felt weak. Even the exclusive adjustable DOF is a quasi feature that could be included for iPhone X users. iPhone X can already toggle the blur on off, it’s just an adjustable blur matte but they dressed it up as if it’s an A12 exclusive. Whatever.

Dual SIM, ip68, and 512gb are all interesting but can’t justify the price and the phone has been fast enough for years that I really don’t care about a bump. Larger photo sites are really compelling to me though, but, that’s it.

In 2019 I’d love to see improved battery, reception, reduced weight, and the rumored ultrasonic Touch ID under the the screen surface to bolster Face ID. I actually really love Face ID but I’ve also never entered my passcode manually as much as I have in the last year and it would be wonderful to have both options - you could select either or for most people, or both and for high security govt and corporate.

If the articles floating around out there are to be believed, it's more than just a processor "bump", and the Xs(and Xs Max) is the fastest/most powerful cell phone on the market, now. And I'm not saying that the X isn't fast. It's the best cell phone I've ever had.

To me, the phones are already light weight enough and thin enough(actually too thin, IMO. You can't hold them securely without a case/cover). I would however like to see an increase in battery life, even though it's rare when I've used more than 30%-40% of my battery on my X in a normal day and I'd like 1TB of on-board storage, which we SHOULD easily have had, by now. Touch ID under the screen would be great for times when you're holding the phone in a manner that doesn't work with with Face ID.
 
Every year they’re they’re the fastest... same story. And for 2-3 years they’ve been more than fast enough. I stopped feeling like a faster CPU was needed somewhere around the iPhone 6. Unless you are gaming or doing some crazy task on your phone, it’s a “bump” for anything non theoretical.
 
At this point things might be headed in the other direction, where news outlets basically fabricate everything and each segment begins with the text - Based on a true story.

As for the cameras; I just picked up an iPhone 8 Plus the other day. I can now shoot 24, 30 or 60fps 4K with optical image stabilization. This is one year old technology and not even the best on the market back then. There is no reason for me to take a DSLR with me when I'm at family or friend functions or at some fair or event that I'm not being paid to shoot. Once these things have global shutter, even better. Even now they blow away every other dedicated video camera I used to shoot casual stuff with. Aside from maybe zoom capability. Heck if I had the choice between this iPhone and the XL1 back in 2003 when I shot my first indie feature, I would have gone for the iPhone. We are spoiled with image quality these days. But that's the consumer world. For the pro world, we still need good ergonomics and mission critical tools. Despite everything I just said I'm not showing up to a paying job with my new iPhone as my A camera. I think that will remain the case for a long time.

It is the lack of aperture control on the camera that really hurts them for recording motion. Yes they can look very nice but video exposure should never be adjusted via shutter speed and ISO only. I don’t exactly want to be forced to use 1/240th for my family videos when shooting outside in bright sunlight. Perfectly fine for stills but the cameras are not really well thought out from a video perspective.
 
Organic sensors are rumored to be able to sort of dim their light gathering and act as a form of electronic VND. Few years off but when that’s common on mobile - and maybe we’ll even get a variable aperture mobile lens at some point - things may get a little crazy. Especially with things like gyro based hypersmooth (GoPro) and IBiS tech continually maturing. Next 5 years could be a little crazy...
 
Back
Top