[kadajawi]

I'm on the run, but just saw this video:

You Tube

You Tube

[ThorSanchez]

Suck it up and use either the USB cord or a card reader to upload or use your picks on your Android or iDevices and get on with enjoying the camera.

It can be very difficult to get photos from, say, a K-3ii to a phone. Many Android phones don't support exFAT, which is the file system Pentax SD cards have to use. You have to get a card reader, probably a USB-C adapter cable, purchase a 3rd-party app that enables exFAT support, then you should be able to manually transfer files from the card to the phone... phew.

I tried this a year or two ago and gave up, the exFAT apps for Android at the time were less than optimal, and being a degreed electrical engineer I didn't quite have the technical chops to make it work. So when I go on vacation I'm lucky that my kids like to drag around their laptops which I then steal when we're at a hotel or equivalent to upload some photos to either my Google site or social media for the family and friends back home. Or just to do backups to the cloud.

Is it really too much to ask that the camera make that a bit less of a chore? There are now toasters, refrigerators, my car and lots of other things that seamlessly integrate with the internet (whether there's any real reason for them to do that or not). But my relatively recent camera, whose job is to take photos I need to get off camera to do things with, doesn't. At all. I'm basically using a high-capacity floppy drive with a proprietary file system every time I want to do something with a photo.

And God forbid I forgot to correctly eject the SD card with the computer after I'm done transferring pics, or the camera won't recognize it at all. If I can't get back to the computer it'll need to be formatted. This is in 2019.

[Kunzite]

Android IS written in Java, but it is well optimized.

Ha ha ha ha.... wait, you weren't joking...

2) The Milbeaut is... well, slow. Smartphones use processors that rival notebook processors from few years ago, or even current ones.
[...]
If you were to go for Android, you'd have to use a proper smartphone processor too. IMHO, sticking with stuff like the Milbeaut is holding back cameras at this point.

There's a confusion here: what you want is not to go Android, but to use more powerful, non-Milbeaut processors - which can be done without Android/Java.
The GR III for example is using a Linux-based OS. I suspect future Pentax cameras might do the same.

I also have the feeling you're assuming the most powerful Qualcomm, but most by far Android devices aren't using such a beast.

[kadajawi]

So yeah, WiFi+Image Sync is a terrible implementation of technology, can't help feeling Android + other apps/community development would help solve many of these issues.

I'm pretty sure its inevitable, give it 5yrs and I think you will at least see the top tier camera brands embrace Android as their OS and it being a success, I just can't think this tech advancement (that we have in phones) will be ignored in the dslr market, in fact I bet its in development right now and a race for the Canikony to release a camera being hailed as the first that's 'doing it right'.

To be fair, the Panasonic implementation of WiFi tethering (let alone wired tethering) is, while not great, quite usable. You don't need Android for that, though it can be a decent base to work with. Perhaps.

All we need is to have our cameras get hijacked to mine cryptocurrency or send adds to Facebook (Yeah I know I am not displaying the new "Branding")

No, embedded systems do not require general purpose OS's.

Android systems can be as locked down as you want. Heck, they may even stop you from installing apps, or only let you install approved apps. And I'm really not too worried for hackers to exploit a flaw in a camera that is sold a few hundred thousand times at best. It's not an interesting enough target, except perhaps for academic purposes. State sponsored actors though may be a different thing, true. But it would be relatively trivial to detect malicious behavior through unexpected heating up etc.

You raise many good points.

1) In computing there are many kinds of "fast." Yes, Android on modern smartphone hardware is "fast" in throughput but that does not mean it can guarantee that it will always perform some task within some number of microseconds. Human perceptions of latency are probably 100X slower than that required for the internals of a MILC or DSLR camera. True real-time computing isn't about blazing speed, it's about strict control of the machine's resources to guarantee that things ALWAYS happen within some time frame. Android's version of Linux and Android's version of Java are not suited for true real time apps (see Real-time Android: real possibility, really really hard to do - or just plain impossible? - Embedded.com if you want more gory details). Of course, camera makers could fork Android, replace the Linux kernel and replace the current version Java but that would probably break compatibility with all those Android apps that likely depend things the way they are and create a maintenance nightmare every time a new version of Android comes out and any changes need to be replicated in the camera version of Android.

2) Milbeaut is slow in some ways but is fast in others (it can do 112 operations per clock cycle per core). Overall, you are right that it does need an upgrade. But it's possible that upgrade might be a next generation Milbeaut Image Signal Processor which are being used in mobile phones. Again, it's about the different kinds of "fast" in computers -- sometimes specialized circuits are much faster at specialized tasks than general-purpose CPUs and they are much much lower in power.

3) Why would a OS optimized for touchscreen phones and tablets (Android) have built-in support complex multi-button, multi-dial UIs? There's not reason they would. But the bigger issue is on the app developer side because every camera body is different. Unless the app is super simple (using only the shutter button and one e-dial), it probably wouldn't work on many models. And the reason app developers are willing to develop stuff is they know their app might be useful to the billion or so people on Android. If the total user population is only the 100,000 that use a particular model of camera with a particular layout of buttons, dials, and screens, there's less room for revenues.

1) Fair enough. But what exactly is this time sensitive? Android can constantly capture images and just save the last few when you push the button. Surely that will be enough to get the timing right? And when I look at the performance of my K-5... we've put up with some terrible lag. Any decent Android does better. Or the camera becoming unresponsive when saving (something that affects Panasonic too). I fail to see what is this time critical about a camera, but maybe you can give me an example.

2) Personally I'd drop any antiquated Milbeaut processor. I'd want my camera to offer me computational photography like my phone does. With different aims, of course. I don't need any fancy JPEG features, but if the camera can take many exposures, align and stack them up for a higher dynamic range raw file? Yes please. As an option of course (as framing would be less predictable), but still it should be there. I wonder if there is any advantage to a Milbeaut compared to a Snapdragon 855+, except for power consumption perhaps (however since the processor is faster, it will only use more power for a short while and gets through tasks quicker. Also, there is a ton of specialized circuitry in such a Snapdragon processor. From handling the storage subsystem, to supporting multiple displays (internal and external), the image signal processing, the compression of videos etc., even some higher end features like combining multiple shots are implemented in silicone.

3) Android supports keyboards. they tend to have a ton of buttons. It also supports mice and all sorts of other input devices. I would be very surprised if it couldn't handle a few buttons and dials. After all, it's Linux underneath. Do you really think Linux can't handle those things? Especially since there are already Linux based professional cameras?

To be clear, a camera, espiacially a dslr is a relatively high performance image processor, and to load it down with an android operating system will reduce the overall performance.

Remember a camera is first and foremost a camera not a global entertainment and communications platform. Also consider as others have cited, the lifespan, while I do have a K1 MKII I still use all of my bodies back to my *istD which is now 16 up years old. Many anderioid platforms like apple devices become obsolete and almost unusable because the media criteria for web applications rapidly outpaces the hardware, but a camera still takes pictures.

Suck it up and use either the USB cord or a card reader to upload or use your picks on your Android or iDevices and get on with enjoying the camera.

Apple devices do not use Android. And they do not become obsolete quickly. Yes, phone manufacturers, at least on the Android side, tend to stop supporting older devices far too early. However this is a) not because it is impossible to support the older hardware, and b) it doesn't impact core functionality. I have a phone that was launched with Android 1.5, the final update it received was to Android 2.3. It works as it did back then. Why wouldn't it? Yes, it's too slow and limited in storage to display modern webpages. It lacks the APIs modern apps need. But those things aren't what a camera needs. If a camera manufacturer were to release a phone with Android, and, under the assumption that it works stable, never release an update again, in 20 years it would still fulfill its purpose. It would still take photos as good as on day 1. You just wouldn't be able to play Angry Birds on it. Or, more realistically, the latest Lightroom.

Ha ha ha ha.... wait, you weren't joking...


There's a confusion here: what you want is not to go Android, but to use more powerful, non-Milbeaut processors - which can be done without Android/Java.
The GR III for example is using a Linux-based OS. I suspect future Pentax cameras might do the same.

I also have the feeling you're assuming the most powerful Qualcomm, but most by far Android devices aren't using such a beast.

You are right that I am not necessarily wanting to go for Android, but for stronger hardware and software. Now much development is being done on Android, which might make it (in trimmed down form) an ideal platform. You could also implement easy image editing software and social media for photographers who need to be FAST. Take a photo at an event, or of news, touch it up within a couple of seconds and put it out there. There is value in that. No need to reinvent the wheel, as you might have to when using some other Linux distribution. Much of what you need should be there already, from easy wireless connectivity to image processing or an upgrade system.

Smartphone apps usually are quite flexible, as they have to run on a wide varierty of devices. They can adjust accordingly. If there really is a camera app for your DSLR, it could be programmed for different button and dial layouts. But I'm not sure it is even necessary. I see more value in a) being able to script the camera, say, create an app that does timelapses. b) being able to implement advanced image processing techniques and to improve them. Maybe by inventing a scripting language and a plugin system to process files.

[clackers]

Android is a modified Linux.

FWIW, the Ricoh GR III runs an embedded Linux with touchscreen and so too maybe the K-3 successor.

The WiFi transfer I have on my Sony, Pentax and Samsung cameras is appallingly slow because of hardware, not software.

The chips are traditionally designed for very, very low power consumption.

[Lowell Goudge]

Apple devices do not use Android. And they do not become obsolete quickly. Yes, phone manufacturers, at least on the Android side, tend to stop supporting older devices far too early. However this is a) not because it is impossible to support the older hardware, and b) it doesn't impact core functionality. I have a phone that was launched with Android 1.5, the final update it received was to Android 2.3. It works as it did back then. Why wouldn't it? Yes, it's too slow and limited in storage to display modern webpages. It lacks the APIs modern apps need. But those things aren't what a camera needs. If a camera manufacturer were to release a phone with Android, and, under the assumption that it works stable, never release an update again, in 20 years it would still fulfill its purpose. It would still take photos as good as on day 1. You just wouldn't be able to play Angry Birds on it. Or, more realistically, the latest Lightroom.
.

I never said the two devices used android. But both companies method of out dating apps on old hardware and making most of the social media type apps Especially IS the same.

You want to have the camera work on social media platforms yet fail to recognize these are the apps specifically that get cut off for support if you don’t update hardware.

So fine you can still use the camera but it will not do all the functions you want to use it most for by putting android on the camera in the first place.

As for your other comment about reading the cards, a micro use compatible SD card reader is $8.99 on Amazon. That is a lot cheaper than the cost of getting camera companies to support android and by default also requiring a version supporting IOS

[PDL]

Just remember a few things:

• DPR ripped the Pentax K-1 II over the "accelerator" which could be seen as applying aspects of "computational photography" to images. The addition of more work being done in the camera results in, more heat, the requirement for more on-board memory (heat again and space on the internal boards) and power requirements.
• The statement that "Android can take buffer 10 images before you push the button - reducing lag", implies that you want a Mirrorless camera. Ricoh/Pentax has stated that they are sticking with DSLR's not mirrorless. Doing this is an always on sensor that will build up heat as a byproduct where heat is the enemy and significantly reduce battery life .
• Even the embedded OS in cameras eventually ends up being abandoned. Just look at your latest firmware version and look up its age - in years for most models, at least the ones I use.
• JPEG engines are not upgradeable by firmware updates. A Prime whatever is a Prime whatever forever. Then again, I really don't shoot JPEG.

Yes, UNIX/Linux can be made small and locked down. I even have a Rio MP3 player that uses BSD UNIX. I worked for a medical systems company that used a embedded UNIX (VxWorks) and in their later models of devices used embedded-XP (Yeah - Microsoft). For small embedded single use systems a custom designed OS is far better than a General Purpose OS that has fragmented into the mess that Android is today.

As for Computational Photography, why would I want to let some faceless group of engineers decide for me what is a "good" photograph? Ever since I first picked up my SF-1, I was frustrated by the limits imposed on the hardware to do just because the "engineers know best".

---

Last edited by PDL; 11-08-2019 at 06:17 PM.

[stevebrot]

Android is a modified Linux.

...and may be programmed* using an API accessible using a modified (semi-proprietary) JDK, though APIs for other languages are supported as well. I played around with Android development a couple of years back and decided that it was too big a mess for the trouble involved. Even with capable device emulation, many features (particularly in the area of camera features) are poorly supported through the standard interfaces.


Steve


* most commonly

[angerdan]

If the camera firmware would be open source, this allows a lot of stuff. Like this awesome project:
https://en.wikipedia.org/wiki/Magic_Lantern_(firmware)

[Kunzite]

Android systems can be as locked down as you want. Heck, they may even stop you from installing apps, or only let you install approved apps.

What is the point, then, if not to install poorly written Java applications that do nothing useful but hog resources?

1) Fair enough. But what exactly is this time sensitive? Android can constantly capture images and just save the last few when you push the button.

No, that isn't Android. It's something that can be implemented on any camera, if the hardware allows (and isn't it done in some Olympus?)
And you wouldn't want some things that are Android, like:
- very slow start up times
- high battery consumption
- the camera app taking seconds to start up
- camera freezes while the garbage collector runs
etc.

2) Personally I'd drop any antiquated Milbeaut processor. I'd want my camera to offer me computational photography like my phone does. With different aims, of course. I don't need any fancy JPEG features, but if the camera can take many exposures, align and stack them up for a higher dynamic range raw file? Yes please. As an option of course (as framing would be less predictable), but still it should be there. I wonder if there is any advantage to a Milbeaut compared to a Snapdragon 855+, except for power consumption perhaps (however since the processor is faster, it will only use more power for a short while and gets through tasks quicker. Also, there is a ton of specialized circuitry in such a Snapdragon processor. From handling the storage subsystem, to supporting multiple displays (internal and external), the image signal processing, the compression of videos etc., even some higher end features like combining multiple shots are implemented in silicone.

There are reasons to drop Milbeaut, yes - maybe there is a much more advanced version, however the current one holds back the cameras a lot. The USB controller, the limited amount of RAM usable, and yes, even the performance.
The GR III, with whatever processor it's using, has at least double the write speed of my K-1.

3) Android supports keyboards. they tend to have a ton of buttons. It also supports mice and all sorts of other input devices. I would be very surprised if it couldn't handle a few buttons and dials. After all, it's Linux underneath. Do you really think Linux can't handle those things? Especially since there are already Linux based professional cameras?

Professional cameras? How about the GR III?
This indeed is not an issue.

You are right that I am not necessarily wanting to go for Android, but for stronger hardware and software. Now much development is being done on Android, which might make it (in trimmed down form) an ideal platform. You could also implement easy image editing software and social media for photographers who need to be FAST. Take a photo at an event, or of news, touch it up within a couple of seconds and put it out there. There is value in that. No need to reinvent the wheel, as you might have to when using some other Linux distribution. Much of what you need should be there already, from easy wireless connectivity to image processing or an upgrade system.

Smartphone apps usually are quite flexible, as they have to run on a wide varierty of devices. They can adjust accordingly. If there really is a camera app for your DSLR, it could be programmed for different button and dial layouts. But I'm not sure it is even necessary. I see more value in a) being able to script the camera, say, create an app that does timelapses. b) being able to implement advanced image processing techniques and to improve them. Maybe by inventing a scripting language and a plugin system to process files.

We both want stronger hardware and software. I simply believe that Android would be a mistake; I'm using Android-based devices daily, they do the job but I'm far from impressed.
I also know how fast and frugal software can be.

My reason to want stronger hardware and software - perhaps this is why we disagree - isn't to run various apps; but, first and foremost, to strengthen the basic camera's functions. Improved responsiveness and speed (through faster processors); faster SD write speeds and USB transfers, USB-C (new SD/USB controllers); larger buffer (ability to address more RAM)...
Android comes in direct conflict with some of these requirements. I do not want to pay for the downsides.

I do not care at all about social media sharing from my camera; the smartphone can take care of that (although Image Sync needs improvements). I would not have a constantly connected camera, by the way (that is, a camera whose battery would not last a day because it's busy to keep up with what happens on Facebook).

As for advanced image processing techniques... that's what they started with the Image Accelerator, right?

[kadajawi]

What is the point, then, if not to install poorly written Java applications that do nothing useful but hog resources?


No, that isn't Android. It's something that can be implemented on any camera, if the hardware allows (and isn't it done in some Olympus?)
And you wouldn't want some things that are Android, like:
- very slow start up times
- high battery consumption
- the camera app taking seconds to start up
- camera freezes while the garbage collector runs
etc.


There are reasons to drop Milbeaut, yes - maybe there is a much more advanced version, however the current one holds back the cameras a lot. The USB controller, the limited amount of RAM usable, and yes, even the performance.
The GR III, with whatever processor it's using, has at least double the write speed of my K-1.


Professional cameras? How about the GR III?
This indeed is not an issue.


We both want stronger hardware and software. I simply believe that Android would be a mistake; I'm using Android-based devices daily, they do the job but I'm far from impressed.
I also know how fast and frugal software can be.

My reason to want stronger hardware and software - perhaps this is why we disagree - isn't to run various apps; but, first and foremost, to strengthen the basic camera's functions. Improved responsiveness and speed (through faster processors); faster SD write speeds and USB transfers, USB-C (new SD/USB controllers); larger buffer (ability to address more RAM)...
Android comes in direct conflict with some of these requirements. I do not want to pay for the downsides.

I do not care at all about social media sharing from my camera; the smartphone can take care of that (although Image Sync needs improvements). I would not have a constantly connected camera, by the way (that is, a camera whose battery would not last a day because it's busy to keep up with what happens on Facebook).

As for advanced image processing techniques... that's what they started with the Image Accelerator, right?

Fair points. I think the app side is... not mandatory, thus Android isn't. At least for me. Others may benefit more from having an Instagram/Facebook/whatever is new app. I've read about a photographer who is hired to do events for companies that sponsor the event, and he has to post some photos right away. I can imagine apps that are specialized for certain tasks that we may not be aware of yet. But in most cases, the camera interface itself should be where you are. And background tasks need not run, unless the user explicitly wants it. Same goes to connectivity. In most cases, you'd want it to stay off. You can go online when you actually need to send the photo. (However having a Google Drive, Dropbox, ftp client etc. in the background could be useful for journalists). Having Android would give the device more flexibility to cover any photographers needs. Having to transfer the files to your phone first is several steps too many.

Start up times... good point. It would have to suspend to some fast memory I guess in order to be back up quickly.

Battery consumption: Compared to a DSLR perhaps. But using more modern processors may also benefit battery consumption. I wouldn't be too quick to jump to conclusions there.

I don't think we disagree that much, we just disagree a bit about the downsides of using Android. While added software can be a benefit, as I wrote above (CHDK-style scripting should be enough for MY needs), my main gripe really is with the hardware they put in these cameras. High end smartphones are absolute beasts, with extreme write speeds, super fast UIs despite being very heavy, even the battery life is impressive.

And most of all I want the advances in photography that come from smartphones to also benefit bigger cameras. You're right that they can probably be ported to other platforms as well. I want a line of lenses that is optimized for maximum performance when software is added to the mix, so that they can be more compact. Especially in mirrorless cameras. Panasonic seems to go that route a bit with their MFT offerings, and I like the results. The images are flawless, despite coming from relatively tiny lenses.

To be honest, I am not too concerned with battery life, if the circumstances are right. I've used more than one battery a day on the K-5 sometimes, and I'm used to swapping batteries with my mirrorless camera. As long as I get through a day with a few batteries and can get them charged over night (quick charging would be useful! Via power delivery!) I'll be fine. It's not like the batteries take up a lot of space in my bag. And the battery can't be drained when the camera is off, that's crucial.

---------- Post added 09-11-19 at 14:45 ----------

Just remember a few things:

• DPR ripped the Pentax K-1 II over the "accelerator" which could be seen as applying aspects of "computational photography" to images. The addition of more work being done in the camera results in, more heat, the requirement for more on-board memory (heat again and space on the internal boards) and power requirements.
• The statement that "Android can take buffer 10 images before you push the button - reducing lag", implies that you want a Mirrorless camera. Ricoh/Pentax has stated that they are sticking with DSLR's not mirrorless. Doing this is an always on sensor that will build up heat as a byproduct where heat is the enemy and significantly reduce battery life .
• Even the embedded OS in cameras eventually ends up being abandoned. Just look at your latest firmware version and look up its age - in years for most models, at least the ones I use.
• JPEG engines are not upgradeable by firmware updates. A Prime whatever is a Prime whatever forever. Then again, I really don't shoot JPEG.

Yes, UNIX/Linux can be made small and locked down. I even have a Rio MP3 player that uses BSD UNIX. I worked for a medical systems company that used a embedded UNIX (VxWorks) and in their later models of devices used embedded-XP (Yeah - Microsoft). For small embedded single use systems a custom designed OS is far better than a General Purpose OS that has fragmented into the mess that Android is today.

As for Computational Photography, why would I want to let some faceless group of engineers decide for me what is a "good" photograph? Ever since I first picked up my SF-1, I was frustrated by the limits imposed on the hardware to do just because the "engineers know best".

What DPR says should be irrelevant, what matters is the results. And battery life should be less of an issue as it is made out to be. Carry spares. If you have a charger and can charge in camera, that's already a good start. Especially when there is quick charging.

Fair enough. Personally I think sticking with DSLRs entirely is a mistake. I'd like them to offer both. But that is another discussion for another day.

Embedded OS in cameras tend to get abandoned pretty fast IMHO. At least I didn't see many updates to my K-5, and even fewer that weren't just there to make it compatible to new hardware.

Maybe our ideas of computational photography differ. Yes, there's a lot that artificially pumps up JPEG images (looking at you, Huawei!), which we may not want (but that's for JPEG... and beginners may actually want it). But there's also a lot of technology that WOULD benefit professional photographers. Most of all (to me) the capturing of multiple short exposures (underexposed as they may be), lining them up and then combining them into a raw file with massive dynamic range and low noise floor. Longer exposures run risk of filling up the pixels in brighter situations - if you flush that buffer regularly and add up the data, you can collect far more data. Also, you can do some image stabilization this way/work together with the sensor shift. Basically I want better raw files to work with. Of course they need to stay true to reality.

[Kunzite]

Fair points. I think the app side is... not mandatory, thus Android isn't. At least for me. Others may benefit more from having an Instagram/Facebook/whatever is new app. I've read about a photographer who is hired to do events for companies that sponsor the event, and he has to post some photos right away. I can imagine apps that are specialized for certain tasks that we may not be aware of yet. But in most cases, the camera interface itself should be where you are. And background tasks need not run, unless the user explicitly wants it. Same goes to connectivity. In most cases, you'd want it to stay off. You can go online when you actually need to send the photo. (However having a Google Drive, Dropbox, ftp client etc. in the background could be useful for journalists). Having Android would give the device more flexibility to cover any photographers needs. Having to transfer the files to your phone first is several steps too many.

Start up times... good point. It would have to suspend to some fast memory I guess in order to be back up quickly.

Battery consumption: Compared to a DSLR perhaps. But using more modern processors may also benefit battery consumption. I wouldn't be too quick to jump to conclusions there.

I don't think we disagree that much, we just disagree a bit about the downsides of using Android. While added software can be a benefit, as I wrote above (CHDK-style scripting should be enough for MY needs), my main gripe really is with the hardware they put in these cameras. High end smartphones are absolute beasts, with extreme write speeds, super fast UIs despite being very heavy, even the battery life is impressive.

And most of all I want the advances in photography that come from smartphones to also benefit bigger cameras. You're right that they can probably be ported to other platforms as well. I want a line of lenses that is optimized for maximum performance when software is added to the mix, so that they can be more compact. Especially in mirrorless cameras. Panasonic seems to go that route a bit with their MFT offerings, and I like the results. The images are flawless, despite coming from relatively tiny lenses.

To be honest, I am not too concerned with battery life, if the circumstances are right. I've used more than one battery a day on the K-5 sometimes, and I'm used to swapping batteries with my mirrorless camera. As long as I get through a day with a few batteries and can get them charged over night (quick charging would be useful! Via power delivery!) I'll be fine. It's not like the batteries take up a lot of space in my bag. And the battery can't be drained when the camera is off, that's crucial.

Indeed we do not disagree that much: the problems we want solved are similar. There are a few points though, about allowing user apps for example, and in general transforming the camera into something more similar to a smartphone.

Android user apps... are not something I'd want on my camera. That is, the entire Dalvik VM environment, apps that you can't control when they start and when they stop, and so on. I'm not impressed by claims that high-end smartphones are super fast... no, they're not as fast and efficient as they could've been, and they will lag and stutter (maybe not from the beginning). My cameras' performance remained constant.

High end smartphones seems to have impressive battery life just because you don't remember how much the battery lasted before

I want lenses made right, not botched then "corrected" via software. "Flawless" results, ahem, that's a bit like praising the battery life

[stevebrot]

If the camera firmware would be open source, this allows a lot of stuff. Like this awesome project:
https://en.wikipedia.org/wiki/Magic_Lantern_(firmware)

Of course Magic Lantern is not firmware per se, but an open-source hack on top of the proprietary Canon firmware exposed on some Canon bodies. I think of it as running the camera in a sort of debug mode. As for the claims of safety in the Wikipedia article, Google "magic lantern bricked". Ooops!




Steve

[kadajawi]

Indeed we do not disagree that much: the problems we want solved are similar. There are a few points though, about allowing user apps for example, and in general transforming the camera into something more similar to a smartphone.

Android user apps... are not something I'd want on my camera. That is, the entire Dalvik VM environment, apps that you can't control when they start and when they stop, and so on. I'm not impressed by claims that high-end smartphones are super fast... no, they're not as fast and efficient as they could've been, and they will lag and stutter (maybe not from the beginning). My cameras' performance remained constant.

High end smartphones seems to have impressive battery life just because you don't remember how much the battery lasted before

I want lenses made right, not botched then "corrected" via software. "Flawless" results, ahem, that's a bit like praising the battery life

I think my idea is to make the camera flexible in order to cover all photographers needs, rather than just my needs. That's why Android would be useful. But of course it would be necessary to deactivate it all - for those who don't need them. There has to be a way to block background apps etc. In that case the performance should remain the same.

Of course if you were to run a bare bones OS on smartphone hardware, it would be even faster. But functionality is worth some trade-offs.

Well, yes, phones used to last longer. But they did nothing. It's like saying an abacus lasts longer than a notebook. I suppose that IS true... In any way, if you would leave a smartphone in stand-by all the time and occasionally make a phone call (which is all that those old bricks used to do), then battery life would be great too.

I'm more about results. And if the choice is between an optically perfect lens that's huge and 1,5 kg, or one that is 500g and small but requires software to correct it... I'll take the small one. Now if I were to use the same lens with analog cameras, that'd be different. Look at the Huawei P30 Pro. It has a f1.6 lens covering a 1/1.7" sensor, and it's tiny. They are doing a ton in software to fix it (at least for JPEGs, with raw you're on your own... and it's BAD, but can be fixed). But it's better than having a well corrected lens, which would be much bigger.

[PDL]

What DPR says should be irrelevant, what matters is the results.

Well DPR and its ilk, are the 800 pound gorillas being that they are usually places where people go to get "opinion" and guidance. When using Mirrorless/DSLR's/Rangefinders etc. the overall position is that anything that "interferes" with RAW processing is running against a huge headwind. Since we are talking about Pentax here, we are talking about a minority manufacture who has a real issue getting new customers. If they decide to go out on a limb and allow third party applications to be loaded onto their devices to "add features" not under their control they will die. Ricoh may not seen as being all that responsive to new things, but they are not stupid. So, when a off brand App bricks your camera or produces really bad image data who do you go to for repair? Will Ricoh just flash your camera back to its original settings for $300 bucks a pop and send you on your way?

So, who would be the gate keeper for these "Apps"? Ricoh? Google? You?
How would you propose to keep garbage "Apps" off of the camera. Just how much free memory/storage are you talking about? If you are talking Android, do you really want Google to control what goes on your camera? I for one do not.