Phase One IQ260 Achromatic Digital Back – a medium format heavy weight.

I first heard of this Danish company called Phase One nearly a decade ago.  Back then, they made very high end medium format digital backs that could fit Mamiya, Hasselblad or Contax bodies.    They still do and now they also make their own bodies but I suspect most of their revenue comes from their raw conversion Capture One software.    They also make institutional equipment to digitize books and cameras used in presumably Western spy planes.

Phase One piqued my interest because I was performing a monochrome conversion to an Olympus 43rds E-500 DSLR by replacing the OEM sensor, a Kodak KAF-8300C, with a KAF-8300M  (https://jimchungblog.com/2017/01/02/the-joys-of-monochrome-photography/).  The M designation implies a sensor lacking the Bayer layer and hence cannot record color data, only grayscale or black and white.   Phase One also made a monochrome digital back, the IQ260 Achromatic.  So did Leica with its Leica Monochrom model.   And that’s it.

I was happily surprised to find the CCD sensor is fabricated by Teledyne Dalsa which are located in Waterloo just an hour west of Toronto, this being the Canadian equivalent of Silicon Valley.   Some of you might even remember the most successful mobile phone before the iPhone was something called the Blackberry, which came from the same region.   The IQ260 Achromatic boasts 60 megapixels, each 6 micron in size with a sensor dimension of 54 mm x 40 mm.  This compares very closely to the dimensions of a 6×4.5 120 film negative.  To better visualize, the crop factor as compared to a “full frame” sensor is 2.5x, and as compared to a m43 sensor is 3.2x.  This was a big deal and a big sensor when it was introduced in 2013, the same time the E-M1.1 made its debut.  So was the price in 2013, $45k USD.  That’s just the digital back, no body and no lenses.  I’m not nearly as frugal as my immigrant parents but I did absorb their lessons and I waited nearly a decade before getting one at a greatly depreciated price.

I wanted to see how it compared to my custom modified 8 MP E-500 Monochrome and if Olympus’ 50 MP and 80 MP high resolution mode might offer equivalent resolution performance even with the Bayer layer.

BlackSwan
Black Swan, IQ260 Achromatic back on Phase One 645DF body, Mamiya 300mm f/4.5 lens. 1/320s, ISO 200.  Even though these cameras are meant for studio work, they can be used in the field for wildlife.  But at over 3 kilograms, they are heavy with a shutter sound like a gunshot.   The large sensor means wonderful bokeh but a 300mm lens that feels like a 100mm on a m43 body.  You have to get in close to the subject to fill the frame.   All images with a monochrome sensor have to be taken with an IR and UV blocking filter since the sensor without the Bayer layer is highly sensitive to a wide spectrum.  I always use the B+W 486 filter because B+W use Schott glass, the same as Zeiss.  The quality of filters is very important to ensuring image quality.
beaver2
Beaver.  IQ260 Achromatic back on Phase One 645DF body, Mamiya 300mm f/4.5 lens, 1/125s ISO 1600. Shot at dusk I got unbelievably close to this beaver before it noticed me and dove to the bottom of the stream with a thundering slap of its tail.   Another deterrent to wildlife usage are the limited number of telephoto lenses available which typically utilize screw driven helicoid focusing mechanism that is slow and prone to backlash issues which prevent small shifts in focusing position once focus is attained.  These lenses hark back to the days of film and most are only available used.  The 645DF body only has three central focus points that cannot be moved.  For birding, this system is not.

To test the resolution performance of the IQ260 Achromatic vs the OM-1 and E-500 Monochrome, I needed a common lens that all three could shoot through.  This is indeed a difficult conundrum because lens adaptors are hard to come by, especially for 43rds bodies which are a rarity.  I then realized the perfect solution, I would use my Astro-Physics Traveler 610mm f/6 refractor as the telephoto lens as it has much more focus travel than a conventional camera lens and I have the adaptors necessary to secure the bodies to the scope.  I mounted the ISO 12233 chart on the garage door and shot from the end of the driveway.  For the telescope, I used an IR/UV blocking filter from Cavision, a Vancouver based company.  This is because a 105 mm B+W filter costs nearly $500 USD.

MTFcurves
MTF Curves generated from the resolution tests, MTF10 in red and MTF30 in green.   The AP Traveler is a serious astronomical instrument with a draw tube diameter of 2.7″ to allow unvignetted usage of 6×7 film cameras from back in the 1990s when it was manufactured (most refractors are limited by a 2″ draw tube).  It features an all spherical three element objective with a central ED element all interfaced with oil. Each scope is hand manufactured and the lens elements hand figured in a custom optical tube that requires no ongoing collimation and is still the world’s most compact 105mm refractor. And each of the 700 units were made in the USA. Only the large Phase One sensor is able to reveal the off axis aberrations of the scope (mainly coma and field curvature) and astrophotography with this camera would require the use of a field flattening optical element. Otherwise for visual observation it is only important that a telescope be optimized for central on axis sharpness.
OLYMPUS DIGITAL CAMERA
Taking a central slice of the ISO 12233 chart showing resolution markings of up to 4000 lines per picture height. All images have been reduced in size equal to the 8 MP image produced by the E-500 Monochrome so they can be easily compared.  All images shot at the base ISO setting of the body which was ISO 200 except for the E-500 which was ISO 100.
Graph
A comparison of resolution performance as measured by HYRes 3.1 software as discussed in earlier blog entries. The takeaways are that 8MP clearly cannot out resolve 60MP even though the pixel density of the two sensors are similar. You do see an improvement with 20 MP and with the 50 MP high resolution image. The 80 MP image is generated by completely shutting down the sensor stabilization and although the telescope is mounted on a robust Manfroto #028 tripod, there was a breeze. Even though I used a remote shutter release it is clear that working outside, subject to atmospheric conditions and wind, the high resolution mechanism was not designed to be used at such a demanding high focal length and sharpness was compromised.
CNTowercomparo
CN Tower.  Mounted on tripod.  (left) IQ260 Achromatic back on Phase ONE 645DF body and Schneider Kreuznach 80mm LS f/2.8 lens. 1.1s, ISO800 (right) Leica 25mm f/1.4 43rds lens on OM-1 body 1/8s, ISO 640, this 80 MP image was downsized to match the 60 MP of the IQ260.  To make the comparison more equitable, the IQ260 was shot through Tiffen RGB filters mounted on the front of the Schneider Kreuznach lens and recombined to create a color image.  Color image sharpness can be affected by chromatic aberration of the lens and color fringing artefacts created by demosaicing.  Both images are very equivalent but the IQ260 still shows a very slight resolution advantage as seen in through the details in the observation windows of the CN Tower and has the more shallow depth of field. 
CityRGBcrop
Toronto Downtown at sunrise. IQ260 Achromatic back on Phase ONE 645DF body and Schneider Kreuznach 80mm LS f/2.8 lens. 1/8s, ISO200 with Tiffen RGB filters.  In this and the previous image, there is color ghosting caused by reflections in the Tiffen filters not designed for this purpose. 
citycomparo
100% crop of images comparing IQ260 (upper) to 80MP image of Leica 25mm f/1.4 on OM-1 body (lower). The IQ260 wins ….. but we are also comparing two very different lenses.

So the OM-1 80 MP high resolution image has the potential to equal the performance of a premier decade old camera, but in a much more affordable and portable package.   I feel with further advances in raw development and computation photography, Olympus can continue to close the gap with larger sensors in ways not previously forseen.

I’m still waiting for the next full moon to complete this article with a comparison between true monochrome cameras, the IQ260 Achromatic and the E-500 Monochrome.  This will also be a good seque to Part 2 of this article involving astrophotography with IQ260 Achromatic.

ColourSuperMoon2016
The Super Moon on November 14, 2016 taken with the E-500 Monochrome DSLR showing real lunar surface colors (with AP Traveler, 50 images stacked per RGB channel).
PhaseONEMoon
Phase One RGB Moon June 10, 2022 (with AP Traveler, 20 images stacked per RGB channel).  So, what does this demonstrate?  Not much.  The quality of any astronomic image is dependent on the stability of the atmosphere above the viewer.  On June 10, 2022 the Moon was barely at 35° of elevation from the horizon.  On November 14, 2016 it was above 55°.  The success at resolving color on this earlier image is due mainly to the thinner atmosphere at this higher elevation and superior atmospheric conditions of that night and not on the camera body at all.   All that can be concluded is that the small 8MP achromatic sensor compares very favourably to the medium format 60MP Phase One in terms of detail resolution which is to be expected since both have approximately the same pixel sizes (5.4 μm vs 6 μm).  The difference is that the Moon occupies only about 3% of the Phase One’s FOV whereas it occupies more than 50% of the m43’s FOV.  

 

StawberryMoonFinal copy
Strawberry Moon at dawn on June 15, 2022. Here the Moon is about 5° above the horizon and although one can stretch the image to induce some coloration, its artefactual and not actual. There is simply too much ground haze (pollution) and poor seeing to be able to separate the very faint color data.  The other factor is the unknown spectral profiles of the Tiffen RGB filters I used.  These were not primarily designed for RGB imaging and the separation of the RGB colors depend on their peak transmission wavelength and shape of the transmission curve.   These may not be distinct enough from each other and hence fail to isolate the RGB signals clearly, especially when the color signal is weak.

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