Pixii Camera – true monochrome performance with a Bayer layer?

(tested against OM-1 and Phase One IQ260 Achromatic)

I have to thank the kind people at Pixii for lending me their new camera after I expressed an interest in testing its monochrome imaging abilities. So I’m not writing a general review of this new and innovative camera although I will mention a few observations in passing that may not be apparent from reading other reviews or visiting their website.

As a Canadian, I’m naturally a bit of a Francophile. That’s why I take offense to those who cast aspersions like referring to French engineering as oxymoronic. They clearly know nothing about the Citroen DS, the TGV or the Ariane rockets, amongst just a few of the examples that immediately come to mind.

And like all those highly successful engineering projects, innovation is the key and common trait. The Pixii is also an innovative business model, because they allow owners to continuously upgrade the camera hardware without having to purchase a new model. The original model made its debut in 2020 with an 11 MP APS sized sensor and its current 3rd iteration features a Sony sourced BSI 26 MP sensor, but original owners were able to have their cameras upgraded with this improved sensor by the factory (for presumably an unspecified fee). The complimentary firmware upgrades have also improved performance and usage.

The camera is a manual focusing rangefinder, which is in keeping with the Leica esthetic – right down to the the ventral battery compartment that bears a strong resemblance to the vintage Leica access for film reloading.  I find the lack of a handgrip makes one handed shooting difficult, but I suppose the need for manual focusing necessitates constant two handed operation.  The handgrip does matter in terms of weight perception and handling balance.  The Pixii is very similar in size to my OM-1 and with a similarly attached lens, the OM-1 feels like half the weight when in reality it weighed 30g more!  There was no operating manual in the box (perhaps by design?) but I’m glad that the controls are minimalistic and generally intuitive with settings viewable on the dorsal OLED screen.  There is no back viewscreen to preview images but there is a well integrated phone application that will allow this function.  There is only an electronic shutter which will be susceptible to artifacts when shooting fast moving objects as well as being incompatible with flash use which is perhaps why only a cold shoe is available.  The camera operates in Aperture Priority mode (the A on the big dial) or Manual with adjustable shutter speed.  There is no memory card slot but this unit came with 64GB of built in storage. 

The main deficiency is the placement of the rangefinder focusing window. This position doesn’t allow the use of adaptors and non Leica M39 lenses because the flange of the adaptor has to be wide enough in diameter to support these lenses and thereby blocks this window. Even some wide barreled native Leica M39 lenses will block this window which really questions the original thinking behind this design.   (bottom image is a Canon EOS to Leica M39 adaptor).

Despite the outer flange diameter of the m43 lens mount being very similar to the Leica M39, its unlikely an adaptor exists that will allow unhindered operation of the rangefinder focusing window since its proximity is so close.

My chief interest in the camera is their assertion that the camera is able to produce a true monochrome image even with the Bayer layer in place. Since the transmission characteristics of each RGB filter in the colour filter array is known, it should be possible to calculate the original photon values hitting each pixel well and reproduce the raw data of a monochrome sensor. Monochrome imaging is not some affectation. Monochrome images will be of higher resolution and absent of any visual artifacts especially along edges in colour images caused by inaccuracies generated by Bayer layer interpolation algorithms. But unlike a true monochrome sensor, the images will not benefit from even higher resolution possible without the presence of an anti aliasing filter, almost two stops of increased light sensitivity from an absent colour filter array layer, and the ability to image in the infra red spectrum without the presence of an IR blocking filter.

All images shot at ISO 800 and f/2.8 but Voigtlander 35mm f/1.4 in the Pixii and Leica 15mm f/1.7 for the OM-1. All were shot about the same distance away from the target given the similarity in field of view of both lenses. All images were shot in raw and developed with Capture 1 software for the Pixii and Phase One camera.   The Olympus raw files were developed with OM Workspace.  Despite the pixel oversampling that occurs during the OM-1 50 MP high resolution mode, the images of all three are very similar in quality with no discernable interpolation errors, even along the transition edges.
The Phase One was shot with a Schneider Kreuznach 80mm LS f/2.8 lens also providing similar field of view as the other lenses. Again monochrome performance of the Pixii and the OM-1 are very similar and the only obvious difference in performance is that color tiles #1 and #2 are distinctly different in the Phase One’s true monochrome image.
From top to bottom is Pixii, OM-1, OM-1 50 MP mode, Phase One. The three Bayer layered cameras appear to be noisier than than the Phase One monochrome camera under high magnification – which is to be expected. A monochrome camera should have higher signal to noise ratios is all its pixels and interpolation can introduce not just errors but noise into the image.


Painting500 copy
I wanted to test the cameras with a colour subject that had irregular transitions. The order from top to bottom is OM-1, OM-1 50 MP mode, and Pixii. The standard OM-1 example shows an interpolation error of having a black line painted over the perimeter of the red clothing and finally the oversampling 50 MP mode pays dividends because this error is not present. Since oversampling means the true RGB value of each pixel is known, interpolation is not necessary and such an error does not result. The Pixii somehow managed not to produce this error at all!
From top to bottom, Pixii and Phase One. The monochrome examples are more similar in this example because the brush strokes of the painting tend to disguise the noise that we saw in the earlier example with the colour guide.  Its possible that there is more resolution in the Phase One image since even the fabric nature of the underlying canvas is visible in the painting but not so in the Pixii image.

The Pixii does a good job at delivering a monochrome image.   Their website concedes that while their monochrome image does not fulfill all the characteristics of an image delivered by a monochrome sensor, this is balanced by the ability to image in full color with much better than anticipated Bayer interpolation.  Besides, most people are unfamiliar with what a true monochrome image looks like and will never notice the differences, that being a very high resolution noise free and artifact free image.   Most people are more concerned with how the contrast and shadows are manipulated in a monochrome image.

1 Comment

  1. Hey, thanks for the Pixii – Phase One blurb. I have the Pixii, LOL. And I like the quirky little rascal. Francophile like you, too. Worked for my uncle for two and one half years in France and did grad school out on Sherbrooke in Montreal (McGill, of course). Damned fine school. Damned fine camera. I have the X2D and wonder how great the gulf is between the X2D and the Phase One. Will you test them?



Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s