NX1 rumours - for what it's worth...

Started 7 months ago | Discussions thread
ttbek
Contributing MemberPosts: 985Gear list
Like?
Re: NX1 rumours - for what it's worth...Well ttbek I for one............................
In reply to pixelpushing, 7 months ago

pixelpushing wrote:

ttbek wrote:

As many other devices with much greater demands tends to prove out (cameras + 4G cellular + Wifi + Bluetooth + 8 core CPU + hardcore 3D games + tons of apps running in tandem + etc.).

Mind naming one with a longer battery life to battery mAh ratio?

Longer than what? A battery can only contain X capacity amperage vs. Y rate and quantity of drain.

Than the NX300

I had this conversation with someone before on here. All the devices named have much larger battery capacities than the NX300 and when scaled they turn out to be less efficient by a significant factor.

NX300 or NX?

NX300

The NX300 runs Tizen, the NX runs Android. Neither camera has been reviewed to have exceedingly poor battery life. As I've cited here previously, the NX has received high marks. Your mileage may vary, but isn't it possible some folks around here got preproduction models, or needing fw revision?

I'm not saying the Galaxy NX had poor battery life.

The culprit here was a total lack of adaptive customization of UI for this device. Tizen would do the very same thing if it were the fully open 'desktop' configuration, too, I have no doubt. Instead, Samsung crafted a simple menu system (that I don't care for vs. older NX cameras, myself, not intuitive at all IMO) so the OS wasn't tripping over itself and draining the resources.

Having looked at what source code is available, it's not running a stripped down shell, rather it's running a full desktop with a full screen app on Tizen with compositing and all that jazz, so not actually as far from running a 3D game the whole time as one might think.

I do not follow what you're saying here. The NX300 is definitely running a shell. Whether you want to deem it a full screen app or whatever, it's an opaque UX that the user cannot circumvent to run other apps that could draw more resources.

What I mean is that a lot of things are ticking in the background that are not needed, it's a shell, but it's not a very stripped down one.

The Desktop manager is Elementary by the way, which you can install on pretty much whatever Linux distro you have at hand to try out instead of Gnome or KDE or whatever else is being used by default on that distro.

But you cannot do that on an NX300. As has been noted, all the source talk here has netted the NX300 absolutely zero development. The reason I was reading wasn't that nobody cared, but that there were still compiling components missing at some level. I don't recall.

You can compile it, it's missing the source for the camera app, but there are binaries for it in there. There's zero development because nobody cares about the part we can get to, might as well just work from the Tizen repositories instead if you're interested in the non camera app pieces.

Anyway, as I said, NX300 has a locked UX. It's not fiddling with apps or a cell radio or BT, or maintaining any other hardware. Just a camera and WiFi.

Actually, those parts of the source you can get to and fiddle with, but no ones really interested in turning their camera into a router or something, it's the camera app source we want to play with. Someone (not me) that's familiar with USB drivers could probably convince the NX300 to act as the host and connect up all kinds of peripherals, and the drivers are in there because it's a full Linux kernel. Actually, if you put the NX300 in the DLNA mode it runs an open X server and you can display an application running on your computer onto your camera (this I have actually tried, and it works, but functions too slowly for video, I attempted VLC, lol).

With a proper camera UI that stopped all resource draw while in that mode and a smidge of optimization for the NX, I have no doubt Android could have easily handled that system.

To do that in Android is a much larger undertaking than I think you realize. I should also point out that most hardware doesn't draw a whole lot of power on mobile devices, the vast majority of power draw is screens and wifi modules.

I make no claims to be a coder or engineer... But I do know there's a cell radio, BT and a host of always-on apps that keep a (for example) Galaxy S4 up and running. It's not inconsiderable.

That's part of my point.

There are concurrent processes involved that are different for every device -- even from one carrier to the next. And a pixel density difference of say 20%, or even a different model of display, can make a substantial difference in battery economy.

CPU and graphics only really hit heavy on power when they're under heavy load. Consider for a moment why we have 8 core CPUs in our phones...

couldn't that have to do with Android needing more cores to feel fast.

No, because older Android phones with 2x and 4x cores still function fine.

What clock speed and what version of Android?

They probably do ok with recent versions of Android? I do not own an Android device that operates at a speed acceptable to me for camera use. Maybe I just need to get out and use a more recent version. I also consider this poor support from manufacturers because they are not offering a path to upgrade (for my devices) to newer versions of Android. I'll also mention that most of those older Android phones are sporting processors clocked quite a lot faster than the NX300's. The NX300 is running on an a9 cortex processor at 800 MHz with the Neon media processor. This processor can be configured with between 1 and 4 processor cores, I think the NX300 has just one, but please correct me if I'm wrong here.

It has more to do with how you're taxing the system with resolution density of screen, per pixel depth and any bloatware that happens to be on board including the wireless apps running under the hood. Also, phones and tablets need to have robust graphics to run 3D games and handle the camera. Maybe not an APS-C but it still draws battery power.

Anyway, we can make Android run as fast as Tizen by throwing a bunch of hardware at it like the phones do.

Excuse me, but do you happen to have benchmarks of similar or same hardware running Tizen vs. Android? Because I don't think we've established Tizen>Android in terms of speed handling a variety of devices without a shadow of doubt.

I don't, but I know where my bets would be.

And then we can follow it up by throwing a huge battery capacity at it, which the phones are also doing now.

Was thinking like the Oppo N1 going up to 3610 mAh etc.. when compared to pre Android/smart phones.

2600mAh isn't that huge. The Windows Nokia Lumia 1020 has a 2000mAh and most comments about that are 'small battery'.

I didn't cite 2600 mAh. Where did you get this number? I see people saying the Galaxy NX has a 4,360 mAh battery, is this incorrect? We would then expect even larger capacity for the NX1. Another device someone compared to in the other thread was a Galaxy Note 10.1 I think, which does sport a large battery at 8,220 mAh. They were implying that Android was more efficient than Tizen based on these devices.

Also, notice how popular those extra power packs are becoming? This isn't about fundamental properties of Android and Tizen, merely about the state they are in now, and how much work would be required to make them suitable for a high end camera.

And I have yet to see solid data that Tizen is better suited.

Tizen has been targeted from the start (well, depends where we call the start, it has some roots that go back to Meego) at somewhat more critical embedded devices in my opinion.

Examples?

Nevermind, I was thinking of the car industry picking it up, which a number of them are, but on further poking around it looks like they're just using it for "infotainment."

Which is about what I'd get out of my NX300 if I were shooting a lot. Several review of the NX claimed battery life was excellent. If I had a phone (iOS or any other OS) with a 20MP full mirrorless APS-C camera attached I would be totally floored if it lasted a full day of heavy use. Even my Galaxy S4 doesn't do that.

Mine doesn't by the way, on the NX300 I can go through 2 batteries in 4-5 hours.

Talk like this lacks relevance when we aren't doing equal benchmarks of performance. What you do with your camera in 5 hours is probably entirely different than what I do.

I just meant that not everyone is satisfied with NX300 battery life.

This is of course, with turning it off when not in use for a while. But with the Galaxy NX, I couold only get around 4 hours use, and a lot less photos than the NX 300 took, before I got the warning of a low battery.

It seems to me Samsung needs to release some kind of patch or update to limit the draw of resources. Have you run any kind of battery monitor app?

Others have pointed out to me that for them the Galaxy NX lasts longer than the NX300, but once you scale the battery capacity you see that it is much less efficient.

That's irrelevant, though. Either the NX lasts about what you'd expect or need out of it, or it doesn't. Scaling battery capacity to claim 'inefficiency' is bordering into fanboy OS wars, IMHO and put as respectfully to you as I can, because I cannot stress enough the help you bring to this forum and my own personal experience. Cheers and all that.

Not exactly, because 3rd parties can offer batteries of differing capacities (of course limited by energy density, but there is still a reasonable margin there). So it is possible that you could be dealing with a Galaxy NX and an NX300 with similar battery capacities. I'm not concerned at all about how long the Galaxy NX battery lasts, but the energy efficiency is relevant because it gives us some indication (though I would appreciate the ability for us to be able to use a better indicator, this is indeed a very crude indicator with lots of room for error) of how busy the the components are.

My point is, and imagine us sharing a pint and haggling over this, is simply that batteries are relative in performance only to the device. You can't say an entirely different device (like a non 4G camera without EVF, BT, 4G or huge 4.8" high density display) is 'more efficient' because it runs the same with a smaller battery. No?

True, but this is partially to do with what people expect from a given platform. That is, when we hear something is running Android, we may expect it to have 4G and run that, this is of course extra power draw, so we could say that Android isn't being less efficient because it's not intrinsic to Android, or we can say Android is being inefficient because it's doing things that don't need doing in a camera when I'm taking pictures. Like what I say later about the phones.

Anyway, while citing mAh numbers, they were intended as a more general point of comparison to refute claims that Android was more efficient based on long running times of devices packing a lot of hardware because they were also packing much larger batteries. The core reason I would argue that Android is less efficient is because it's all running on top of a Java virtual machine, so at an intrinsic level if Tizen and Android were both fully optimized, then Tizen will run faster and use less power. Of course it's possible that Tizen is coded more poorly than Android and thus loses that advantage, but I don't think they did that badly and suspect that they did much better considering that the Linux foundation is directly involved with this project, that is, I think the Linux foundation has better standards for code that is allowed in than Google does.  I'm kind of on a a tangent here though, because we were concerned with the practical amount of modification required rather than ideal optimization scenarios.

I may be exaggerating how much of a difference this makes in practice due to the relatively poor performance of Java devices that I have had actual experience with. That said, Jellybean was a bit of a game changer, but I still think that Tizen requires less optimization for the same performance than Android for a given camera architecture in practice. You're right that you are also getting less functionality, but I still think it's fair to compare because the extra functionality is not core to camera functioning in the same way that I call my old Samsung dumb phone more efficient than new smart phones, because it's better at being a phone and lasts way way longer. Of course it doesn't do any of thousands of things a smart phone does, but for being a phone, it does a fine job of it.

My only real experience with a post Jellybean device was running Android x86 on my netbook (dual core, 1.6 GHz), which was fairly snappy once I convinced the graphics hardware acceleration to work. The issue I had with that is stability because almost no apps would work properly as they are still compiled for ARM and are only usable on x86 via Intel's byte level interpretor, which is just not cutting it yet.

I've rambled here and probably failed in my logic as well someplace. The message I'm trying to convey is that in two devices that are similarly snappy running Tizen and Android out in the wild, the Android device is typically packing more hardware and a larger battery. This bit is limited of course to my experience of the two. For instance, the Galaxy S4 didn't seem much more responsive than the NX300 to me, but the NX300 is potentially a single core device at 800 MHz, maximally a 4 core device at 800 MHz with a 1130 mAh battery vs. a device with either a 1.9 GHz Quad-Core Processor or a 1.6 GHz Octa-Core Processor with a 2600 mAh battery. From a hardware standpoint the two devices are night and day. Also, this article supports the NX30 using the single core version of the cortex a9, and so it is even more likely the NX300 is also using the single core version as well: http://www.gcps.org.au/2014/01/samsung-nx30/  As far as I can tell the Galaxy NX runs on a 1.6 GHz quad core and still has the cortex a9 in addition to that, dedicating the 1.6 GHz quad only to Android and the cortex a9 only to image processing/focus/other camera stuff.  Maybe I'm missing another processor separate from the cortex one for the UI and other Tizen stuff of the NX30 and NX300 to run on?  I think though that for those models the whole thing is running just on the a9s.  So, despite all that extra power in the Galaxy NX, people on here still find it to be less responsive?  That's a real question by the way, not rhetorical, I haven't used a Galaxy NX and am genuinely asking.  I might be that the Galaxy NX was thrown together quickly with a relatively stock Android, but I can say from looking at the source that the NX300 and NX30 use pretty much stock Tizen, so if one is more responsive than the other near stock, then one of them will take more modification and optimization to match that performance, right?  For me it goes to the tradeoff I mentioned earlier, am I looking for a device that does everything, or for a device that is just a really good camera, and for me, if I'm in the market for something marketed as a "pro" camera, then I'm looking for the later.

I think I'm going to reread this post in the morning and feel like an idiot, I'm blaming that hypothetical pint we were sharing.

 ttbek's gear list:ttbek's gear list
Canon PowerShot SX10 IS Nikon Coolpix L3 Samsung NX300 Samsung NX 30mm F2 Pancake Samsung NX 18-55mm F3.5-5.6 OIS +6 more
Reply   Reply with quote   Complain
Post (hide subjects)Posted by
(unknown member)
(unknown member)
(unknown member)
Taffy MOD
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark post MMy threads
Color scheme? Blue / Yellow