01-23-2013, 10:40 AM #1
- 7 Posts
- 01-23-2013, 10:45 AM #2
- 01-23-2013, 10:45 AM #3
"PureView" is only meant to point out that it's advanced camera tech. The 920 is real PureView because they focused on camera (in this case, the floating lens). Just because something is getting an oversized camera sensor doesn't mean it's "more PureView" than what we already have.
Now that I've said that, yes, it appears to be true (WPCentral had an article on it as well). I don't doubt that they're working to keep advancing the camera tech in phones, and this would fit right in with their Lumia release cycles thus far.
- 01-23-2013, 11:20 AM #5
It's a marketing term. It means nothing. Something without meaning can't be real or fake. Just as their are no real or fake Retina displays.
- 01-23-2013, 07:17 PM #7
I am really curious about how they plan on making the Phase 1 system work on the WP8 supported SoC .. I see two options.. they figured out how to use the custom DSP/GPU inside the camera module found on the 808, or there is a new SoC out there that can support the processing needed for oversampling 41 million pixels down to 5/8Mpix, which also happens during 1080p video. None of the currently known SoC can support that kind of processing..
The other option is.. don't do oversampling and just give out the raw jpeg, but then the gains in quality would be almost non existent since they are still using 1.4 micron pixels.
Also, the Xenon flash requires a mechanical shooter... and that requires precise synchronization..
I can't wait to see what they've done :) If anyone can do it.. its Nokia.. they pretty much have the whole 41Mpix system (sensor,DSP, optics) ready to go.
- 01-23-2013, 07:56 PM #9
In this case I think you missed one option... just use a 20 MP instead of 41 MP sensor.
If I read the article correctly, it was stated to be 808 like... that might mean "not identical to".
As the resolution says nothing about sensor size, picture quality could end up being virtually identical to the 808, while offering less digital zoom but still working with current SoC's
Just a hunch...
- 01-23-2013, 09:56 PM #10
The 20Mpix option is very interesting, and it might work.. There might be two alternatives with that.
So.. its all about pixel sizes.. lets say they take the 1/1.2'' sensor from the 808 and cut that down to 20Mpix instead of 40Mpix, which (as far as I can tell) would give them physically bigger pixels, and then then can further oversample those bigger pixels to get a better quality jpeg. The 1/1.2'' sensor cut into 41Mpix yields 1.4 micron pixels, which is no different than any of the other current top smartphones for sale, where the 808 gains a huge advantage is the fact that they take 6-7 pixels and combine them into one "bigger" one, while cleaning all the garbage (noise) in the process.
roughly .. this is what happens with the phase 1 system
So lets take that 1/1.2'' sensor and see how big the pixels would be if it was cut into 20Mpix instead of 41.
The 808 shoots at 38Mpix max (7152 x 5368). The size of the sensor is 1/1.2'' (10,67 mm. x 8 mm.).
There are 1000 microns in 1 mm.
10.67 х 1000 = 10 670 microns
10670 / 7152 = 1.49** microns/pixel ... I think the math is off because I am taking 38Mpix instead of 41. The 808 does shoot @ 1.4 microns for sure.
For 21Mpix (which would probably be around 18 for 16:9 and 19 for 4:3)
I will use a picture from this Amazon.com: Sony DSC-RX100 20.2 MP Exmor CMOS Sensor Digital Camera with 3.6x Zoom: SONY: Camera & Photo for reference. It has a sensor cut into 20Mpix, and the images come out at 5472×3648
So same thing again:
Our hypothetical Lumia EOS shoots at 20Mpix max 5472×3648. The size of the sensor is 1/1.2'' (10,67 mm. x 8 mm.).
There are 1000 microns in 1 mm.
10.67 х 1000 = 10 670 microns
10670 / 5472 = ~2.0** microns/pixel
Just to put that in perspective, the Nokia N8 shoots 1.75 microns, and to this day.. 2+ years later, there is not other smartphone (other than the 808) than can beat it in most conditions.
That is the first option.. use the sensor from the 808, cut into 20 instead of 40, and do a bit of oversampling on top.
The 2nd option is to use a smaller 20Mpix sensor, which would yield 1.4 micron pixels, and oversample those.. but see.. its one thing to have 6-7 pixels available for oversampling (808) and another to have 2-3 pixels available.. the effect won't be the same. The advantage here would be a smaller "hump".
Then there is the question of the Xenon flash, which as you can see here (808):
requires big capacitors for it to make any sense.. so the hump remains an issue, until they come up with mega powerful short burst LEDs which would eliminate the need for a xenon.
And finally.. there is something I don't know enough about, but seem to be an issue. The ratio between the pixel sizes and the sensor is directly related to the lens somehow, but I am not sure how. It seems like the bigger the pixels are, the bigger the lens needs to be.. but I am not sure at all on this one. I just know that there are issues with that.. especially for a phone.
Why are bigger pixels important.. they can collect more photons (light), and from there.. less noise, more detail.. here is how it looks for the DSLRs..
I am not saying that pixels sizes are everything in imaging, there is way more than that.. but it is a pretty important part.. especially for compact/cell phone cameras.
- 01-23-2013, 10:31 PM #12
01-24-2013, 07:40 AM #15
- 229 Posts
- 01-24-2013, 07:58 AM #16
However, I doubt Nokia can pull that trick off. Nokia implemented OIS by floating the entire sensor, but doing so with a sensor of the 808's size sounds like an impossibility to me. Of course they could take the more traditional mechanical IS route, but that will be BIG. I doubt they could fit that into a smartphone.
Nokia is great with cameras, but they aren't magicians either. I suspect we will get either OIS or oversampling, but not both... unless they really do want to ship a brick, or move OIS into front mountable lenses...
Lets not set ourselves up for disappointment by expecting too much.
- 01-24-2013, 10:20 AM #21
- 01-24-2013, 03:01 PM #23
- 01-24-2013, 03:33 PM #24
- 01-24-2013, 03:48 PM #25
Getting back to DxOMark, the way they score, they measure various things and average them. Problem is they are basically doing daylight testing only. Okay, so let's say they got 66 for the 920 and 72 for iphone/GN2. Now, if we do the same set of tests handheld in lowlight, we probably get something like 66 for the 920 and 10 for iphone/GN2. And let say you use your camera in low light about half the time so you weight them equally. Now you get (66+66)/2 = 66 for the 920 and (72+10)/2 = 41 for iphone/GN2. You get my drift? Of course, iphone/GN2 users won't be using their cameras all that often in low light...because they can't! lol
- By boneycat1 in forum Windows Phone 7Replies: 0Last Post: 01-14-2012, 11:32 AM
- By btbam91 in forum Upcoming & Rumored DevicesReplies: 2Last Post: 12-22-2011, 10:33 PM
- By johnnieblack in forum Windows Phone 7Replies: 9Last Post: 10-12-2011, 04:41 PM
- By johnnieblack in forum Windows Phone 7Replies: 1Last Post: 10-10-2011, 01:43 PM
- By Winterfang in forum Windows Phone 7Replies: 9Last Post: 07-02-2011, 09:12 AM