I've been looking forward to see reviews since I finished implementing the RAW DNG profiling and calibration for +Cyanogen on the OnePlus One a couple months ago

+Tek Syndicate just published one in video and they're pretty enthusiasts about the results 🙂

A tip about converter software: +Albert Manduca, who shows his experience with RAW editing uses +Adobe Lightroom which is great as it's one of the software that implement all the DNG capabilities I used during profiling.
* Color calibration
* Noise profiling
* Lens vignetting correction and sensor color uniformity

DNG files are the RAW sensor data plus metadata that describe shooting conditions and how to transform what the sensor sees into a corrected image representing colors as they are.
Compared to proprietary RAW formats, DNG is self-describing.

Smartphones' captures require more correction than other cameras due to their physical constrains.
Some converters support DNG but not all its features.
Typically, noise, vignette or color uniformity won't be corrected as expected, color conversion incomplete, exposure compensation not applied so be sure to use a fully fledged RAW editor!

By the way, did anyone made a comparison with a Lumia DNG?

#supercurioBlog #camera #DNG #cyanogen #oneplus

Source post on Google+

Large change in how disk encryption is done in Android 5.0 Lollipop preview

In previous versions of Android, encrypting your phone aka encrypting the /data partition meant that you had to keep using a PIN or password, that would serve both as:
– decryption key to enter during the device boot (that is a component to decrypting the disk decryption key? need confirmation on that)
– lockscreen PIN or password

As you can see in the attached captures, the behavior is entirely different in Android 5.0 Lollipop (here on on Nexus 5).
The first capture show what you see after encrypting your phone after entering a PIN for the screen lock.
Just like before, during boot, the PIN will be asked during boot and to unlock the device.

But now, as you can see in the second capture, you can switch back your screen security to Swipe, or even None.
Then, no password will be asked during boot or in the lock screen.
Yet your data partition is still encrypted.

Question: Where is the disk decryption key stored, and how is it protected?

I look forward to learn that exactly, hopefully a security researcher will take a look at how this new implementation functions, and if it actually provide any security benefit after you revert Screen security to a PIN to Swipe or None.

A question to you, as all my 5.0 devices are already encrypted: can you also now trigger disk encryption without enabling a Screen security first?

Yes, you can now indeed encrypt a tablet with Screen lock being set to Swipe or None.

#supercurioBlog #encryption


In Album Android 5.0 Preview encryption

Source post on Google+

Shoot; after adding support for latest HCFR versions I realized that they're unable to display the high precision measurement correctly

As you can see from simulated measurements shown here, both Luminance and Gamma graphs are crapped up on current HCFR 3.1.6, compared to the old version (from April 2012).

On the CIE 1931 gamut and saturations graphs, the saturation targets are actually nice on the new version, but the errors when visualizing the curves are a deal breaker.
Well, a little bit of time lost it seems, but I'm still targeting a public release soon.

#supercurioBlog #color #calibration


In Album Old vs New HCFR visualisation

Source post on Google+

But that's not all – this is actually the first phone with an AMOLED screen that is as faithful in rendering hues properly – color and grayscale errors are minimal

The only downside to the Note 4's panel is its gamma value of 1.97, which is below the reference value of 2.2 – the iPhone 6 Plus is close to perfect, at 2.18. In practice, this means that the Note 4 delivers a punchier, more contrasty image than it should, though the effect is not so overdone as to be annoying or distracting.

Average gamma: 1.97
Incompatible with:
– grayscale errors are minimal.
– this means that the Note 4 delivers a punchier, more contrasty image than it should

Conclusion: +PhoneArena authors need some more training on display analysis.

Assuming the average gamma is indeed 1.97 and it is not measurement error and/or inadequate measurement methodology.
A gamma that's too low means the response curve is too high, too bright.
A gamma of 1.97 will give a fairly washed out appearance and the appearance of lack of visual contrast and punch to the images.

Gamma at 1.97 is a pretty large deviation: grayscale error can't be minimal.
It also has the exact opposite effect than what +PhoneArena describes: "punchier, more contrasty image than it should"

It's nice to have some data, but some appear to be invalid and conclusions are contradicting the data.
Not quite there yet +PhoneArena

#supercurioBlog #display #color #measurements #critic

Samsung Galaxy Note 4 vs Apple iPhone 6 Plus
Nowhere else does the rivalry between Apple and Samsung cut as close to the bone as with the iPhone 6 Plus and Galaxy Note 4. Encroaching onto true Samsung territory, it’ll be on the 6 Plus to prove itself better than the Note 4, and that will be no small feat given the years of experience that sprung the latter on the scene.

Source post on Google+

Capture photo's spectrum instead of only 3 channels somewhat matching an inconsistent RGB color model?


Scientific term for that is: multispecrral imaging http://en.wikipedia.org/wiki/Multispectral_image

#supercurioBlog #camera

Originally shared by +PetaPixel

New Imaging System Promises 12-Times More Color Sensitivity than a Traditional Sensor

Most traditional image sensors work their magic by picking up Red, Green and Blue, but a new system promises to use Foveon-like technology to detect a whopping 36 color channels.

That's 12 times the color sensitivity!

New Imaging System Promises 12-Times More Color Sensitivity than a Traditional Sensor
What’s the use of an image sensor that’s 12 times more color sensitive than the human eye? We’re not entirely sure, but thanks to a team of researchers at

Source post on Google+