When Sony acquired Konica-Minolta in early 2006 they also acquired the Minolta DYNAX 7D DSLR
Konica Minolta Dynax 7D Sony a900 and TTL integrated phase detect AF (PDAF) sensor technology, developed from AF technology originally introduced 20 years earlier in the 1985 DYNAX 7000AF SLR.

Most DSLR's nowadays use area detection PDAF and the Sony alpha 900 is no exception.
SONY alpha 900 TTL_CCD_PHASE DETECT AF_diagram


Phase detect AF with f/2.8 sensors will operate between -1EV to +19EV @ ISO100 @ 20°C (Sony alpha 900 manual states 0EV to 18EV @ ISO100 @ 20°C). It struck me that prime focus images of the filtered Sun in white light and the Moon ought to be bright enough for the camera AF to function. The 2/27 day old crescent Moon is as bright as EV3, depending on the zenith distance, within the operating limit. (PDAF will not work with a Fabry-Perot H-alpha etalon filter because the monochromatic beam is polarised and produces interference within the AF sensors).

Focussing the camera manually is awkward and time consuming. The question was how to connect the camera, with an AF lens fitted, to the telescope in such a way that it could operate unimpeded. There remained the nagging question of AF hunting at low light levels. Would the PDAF be able to respond consistently, once the focus had been adjusted manually?

How to incorporate the actuator part of the AF system into a camera adapter? I looked into the feasibility of dismantling a standard lens and removing the optical elements, leaving a barrel assy & AF motor. It ought to work, but the cost of a secondhand Minolta Dynax or Sony standard lens is 350. In any case there was the possibility that the AF unit might not move the camera the correct distance because the lens barrel would be sending focal length info at odds with the OG prime focal length, and the unit could well fail owing to the mass of the DSLR body.

Recently I thought of a way of economically and easily being able to make use of AF. I completed a special afocal adapter in APRIL 2010 designed to enable me to take images of either the Sun in white light, or the Moon, and use the Sony alpha 900's phase detect auto-focus facility.

lathe operations:

1) Chuck up (3 jaw) & face off both ends & centre - check centricity with DTI (if out by >5thou use 4 jaw)

2) Drill down 3.50" x 3/8"

3) Open up 3/4"

4) M/c I.D's to 2" (2.030) & 1.7 (1.750)

5) M/c O.D. 2.540 to match PVC tube I.D. (drive fit) & face off @ 1.50 from datum

6) M/c shoulder 2.700 x 0.25

7) M/c 2.000 O.D. to dead

8) M/c 2.030 I.D. & polish - rounded tool

9) M/c 1.750 I.D. & chase

10) Part off 2.000 O.D. @ 1.25 from datum.

Basically it is a straightforward afocal standard 50mm f/1.4 lens + 30mm x 80 eyepiece arrangement. The problem is that if you simply screw the eyepiece to the camera lens and fit it into the rackmount, the lens SSM AF motor will be overloaded. So what I did was design an adapter that would enable the eyepiece to float on the end of the lens, with sufficient room for the AF lens axial movement (about 1/4-inch). The adapter has a 2-inch nosepiece that fits into the rackmount, and the camera, with standard lens and 30mm eyepiece (Wollensak), screws into it. (the 30mm Wollensak eyepiece has a M49 male thread under the eyecup; all that is required is a stepper ring M55 to M49 ).

The afocal amplification is 50mm/30mm = x1.667 . With a x1.6 Antares Barlow the net amplification is x2.667. The clear aperture of the 50mm f/1.4 lens is 35.7mm & the rear element 30.5mm. The field stop of the Wollensak eyepiece is 46.5mm, and the eye lens diameter 38mm. I obtain an unvignetted 30.5mm field, and a partially vignetted full frame image circle 43.3mm. (See vignetting @ x1.667 (FINE JPEG 10.3Mb) flat field frame)
. The Ramsden Disc (exit pupil) lies on the camera lens iris (even with the iris stopped down to f/22 there is full frame field coverage). This is important because otherwise the field is vignetted, and the image plane not flat & therefore distorted. (The eye relief is 22mm and matches the distance between the eye lens and the camera lens iris). The re-imaged exit pupil must lie at infinity, or at a very long distance. The field of view of the eyepiece is &asymp80°, and exceeds the 47° field coverage of the standard lens. This is another very important consideration. If the field of view of the eyepiece is similar or less than the camera lens, the field will also be vignetted.
Er = eye relief    Ep = exit pupil = eyepiece focal length / OG focal ratio

At x1.667 the Solar image size is 14.8mm (under-sampled). At x2.667 it is 23.8mm (slightly over-sampled). The Fx frame is 24mm x 36mm, so either image will fit on the sensor.

I tried it out on Saturday afternoon (April 17th 2010), comparing the best of my manually focussed (MF) images, focussed using a 3.3x Magnear viewfinder (net viewfinder magnification 2.4x), to the AF images. The PDAF was consistently more accurate, and never failed to find true focus. All I had to do was half depress the shutter button, & the AF locked onto the image plane, initially set visually. The image of the Sun @ x1.667 was dim because I had the OG stopped down to 70mm by a Thousand Oaks ND5 filter. The image was fainter than the Full Moon. The Auto exposure was not used because the lens has to be held wide open (f/1.4) so I used aperture priority to set the aperture, and then shutter priority to set the shutter speed (1/30s @ ISO100). The AF set to <LOCAL> was sensitive and rapid enough to follow the seeing (AIII) induced focus shift. AF was set to slow drive speed to allow for the additional mass of the eyepiece being moved by the SSM, and prevent hunting.

I next I tried it with my Herschel Wedge, and an ND3 filter, at full aperture.

The 6082 dural billet 2".75OD x 4" long, cost 28. The stepping ring from SRB-GRITURN 7.40. The sleeve, pipe flange, and spacing tube were fabricated from 68mm black PVC fall pipe offcut obtained from a local DIY merchant. The dural nosepiece was machined dead to the fall pipe ID, the pipe ID scored, and the nosepiece super-glued in place. The nosepiece ID was roughed out on a vertical miller and turned on a Myford 7-inch lathe. It was very obliging of Mr. Paul Schofield, the Ways and Means officer at Blackpool & District Astronomical Society to make his workshop availlable.

Having successfully used my Surplus Shed 30mm EHQ Wollensak for afocal photography, using my Somy 50mm f/1.4 autofocus lens, in my recently invented autofocus adaptor sleeve, I was intrigued by the possibilities of using a Baader Hyperion eyepiece, which reputedly has a better field correction.

At Leeds AstroMeet 12th November 2011, my girlfriend generously bought me one of the latest in the range, the 24mm. It is smaller than the 21mm, does not have the removable Smyth converter, and has the widest field stop of any 1.25-inch push fit 24mm eyepiece. The Baader description says the field stop is 27mm, but it measured up at 28.4mm, equating to a 69°.4 afov with almost zero angular magnification distortion.

The dealer from whom the eyepiece was purchased, was the newly formed northern branch of GreenWitch. I asked the dealer if he had any copies of the Hyperion accessory catalogue, but he hadn't, so I downloaded it from Alpine Astro's website. That's when the fun and games started.

I wanted the adaptors for both afocal and projection photography. The part numbers on Baader's chart did not match Alpine Astro's who had decided to use their own. In the end I got my girlfriend to drive me to Rother Valley Optics the Saturday before Xmas, after 5 weeks of frustration trying without any success to get the afocal adaptor, and the extension tubes off either David Hinds, the UK Baader import agent, or either GreenWitch, or several other UK dealers.

Rother Valley Optics were able to supply all the projection adaptors, but the M43 to M55 stepper ring needed to couple the eyepiece to my camera lens in my autofocus adaptor sleeve had to be put on back order. In the New Year I tried several camera shops including Denton Photo Optics who offer a full range of stepper rings, but the ones they supplied were in plastic instead of black anodized duralamin alloy. In the end, I had to get two stepper rings, M43-M46 & M46-M55, from Ian Broomhead of SRB-GRITURN, Luton.

The 24mm Hyperion is a good eyepiece. What lets it down is the difficulty in obtaining the adaptors needed for afocal photography, which is ironic, because this is precisely what the eyepiece was intended for! Over the years I have acquired lots of stepper rings at camera fairs, for a song. Typically however no combination would couple the eyepiece to the camera filter thread.

One wonders why the eyepiece has an M43 rather than a T-thread (M42) standard thread. Had this been the case I could have obtained a single stepper ring from a high street camera shop.

There is also no information I could find about the eyepiece's lens arrangement. The other Hyperions have 8 elements in 5 groups, with a two lens removable Smyth converter. Since the 24mm does not have a Smyth converter, one is left wondering if the lens arrangement is 6 elements in 4 groups, or, since the Baader website informs it is an entirely new design, another lens arrangement entirely.

This eyepiece gives x2.08 amplification so the image is correctly sampled close to the Nyquist-Shannon criterion. However repeating the field-of-view equations shows the field is heavily vignetted.

Er = eye relief    Ep = exit pupil = eyepiece focal length / OG focal ratio

Bearing in mind the rear element of the Sony 50mm f/1.4 lens is 30.5mm, the fully illuminated field diameter of 27mm, the beam entering the camera lens is not wide enough to evenly illuminate the sensor. However the Solar image will be 18.5mm diameter, and unvignetted.


If you own a high end DSLR with PDAF (Canon EOS D series - TTL_CT_SIR; Nikon D series CAM-3500FX / CAM2000FX / CAM-1000FX; Olympus E3 series; Pentax K series - SAFOX VIII; Sony alpha series - all alpha DSLR models use TTL PDAF) this is a far simpler way of focusing a whole disc Solar or Lunar image. There is no need to keep checking the LCD screen at maximum magnification, and tweaking the focus; nor is there any need to connect your camera to a laptop to obtain accurate focus. PDAF is rapid, right up to the time the mirror is raised. And it operates for every frame without you needing to intervene. It is important you use a fast standard lens with a supersonic wave motor, f/1.4 or faster for an Fx format sensor, f/2 or faster for an Dx format sensor. Set the AF area to local, and if the DSLR has Live View, ensure it is disabled (only the backup contrast AF operates in Live View mode - with the exception of the Sony a350/390).

This afocal projection AF setup will not work with contrast AF. Contrast AF is too sluggish and unresponsive to follow rapid seeing fluctuations, and it is nowhere near sensitive enough to detect contrast boundaries of filtered white light solar images, or crater detail at the Lunar terminator. If you wish to follow my example you must use a DSLR with PDAF.

Minolta's PDAF technology was developed from the Leitz Correphot system based in part on the work of Norman L. Stauffer, <see US Patent 4185191 22JAN1980>. Honeywell licensed the TCL2 system to Leitz in 1976 who demonstrated a fully functional auto-focus SLR at Photkina.

Minolta in turn licensed Leitz's Correphot system and by 1983 had developed a CCD slit array AF system which they integrated into their Dynax 7000AF in 1985. Unfortunately for Minolta as it turned out, Stauffer had by this time extended Honeywell patented AF technologies to include systems that broadly overlapped Minolta's system <see US Patent 4384210 17MAY1983>.

Minolta's PDAF infringed Honeywell's Visitronic TCL AF system, patented in the mid '70's by its engineering manager Norman L. Stauffer. Contrast AF, also invented by Stauffer, was used in the 1977 Konica C35AF. Stauffer filed numerous patents between 1973 & 1983, covering most aspects of passive AF. The patents were all assigned to Honeywell who successfully sued Minolta in 1991 following a 4 year patent litigation which cost Minolta $127.6m in compensation; Asahi, Canon, Konica, Nikon & Olympus were obliged to pay royalties totalling $303.1m. Honeywell had been importing Asahi Pentax SLR cameras into the USA through their Heiland division in Minneapolis since 1959, having acquired Asahi Pentax exclusive distribution rights over Sears, Roebuck & Co. who had marketed the Asahiflex under their Tower brand.
Honeywell's Heiland division retailed Asahi Pentax SLR's throughout the 1960's
including the 1964 SV clip on meter
or 1964 Spotmatic TTL meter
and 1970's ES AE TTL metering

Honeywell had originally intended entering the stills camera market in the early 1970's. Much of the initial work which Stauffer and his engineering team undertook was based on earlier work in the 1960's incorporated in the first auto-focus slide projector in 1964. This technology was licensed to Kodak who used it in their Ektagraphic AF slide projector in 1969.

The level of investment required in manufacturing compact rangefinder and SLR cameras proved too much for Honeywell, who instead decided to license their patented technology to Leitz. Leitz only developed working prototypes. Asahi produced the first AF SLR, the PENTAX ME F in 1981, but it was not a fully integrated AF system, and proved a commercial failure. It was Konica who incorporated Honeywell's TCL AF miniaturised module into the C35AF (1977/78), and subsequently Minolta, who took Leitz's Correphot system and devised a fully integrated AF placing the lens motor unit within the SLR body. By 1985, Honeywell had abandoned any idea of using their patent rights in cameras of their own manufacture. They effectively sat on their intellectual property rights, and exploited their position to litigate against Japanese camera companies who were making all the headway in realising working SLR AF systems during the 1980's.


The development of SLR AF technology in the late 1970's and 1980's is interesting. Leitz Camera AG was the first camera company to develop an auto-focus prototype with the Correphot, demonstrated at Photokina 1976. It was based around the Leicaflex SL2.
A secondary mirror deviated part of the rays to a pattern recognition system. Correphot stems from "Maximal correlation" between two elementary light rays passing through opposite edges of the lens from the same point of the object. The control unit was given the identification CK1. Production of the SL2 ceased in 1976. At the time Leitz was in financial difficulties and opted to produce the R3 in collaboration with Minolta.

The AF control unit was then fitted into an R4 MOT prototype, ident CK3 in 1978, and R4 MOT CM2 in 1980. The Leica R4 was developed in co-operation with Minolta. Leitz AG never put any of these AF prototypes into production. Indeed Leica R series 35mm SRL cameras have never featured AF. The rumoured R10 DSLR anticipated to feature phase detect AF was abandoned mid 2009. Ironic, when one reflects Leitz Camera AG began exploring ways of focussing a camera automatically back in the mid 60's. Leitz Wetzlar Gmbh & post 1996 Leica Camera AG was run by an extremely conservative board of directors. It was considered that Leica SLR users preferred to focus the image themselves, and therefore auto-focus was superfluous. This argument is sustainable for an SLR but not DSLR whose sensor is far flatter than 135 format roll film and places excessive demands on visual focussing in critical conditions.

Nikon purchased rights off Leica which led to their F3AF in 1983.

Following Leitz's Correphot CK3 system demonstrated at Photokina in 1978, came the Pentax K2MD AF zoom,
using Honeywell's Visitronic system. It was impractical, it required a 4AA battery pack, and weighed 1.1kg, it was also slow, had low discrimination, and could not focus on a low contrast object.

At Photokina 1980 Ricoh unveiled a 50mm standard lens equipped with a Visitronic module,
and Canon demonstrated an A1 body with a 35-70mm f/4 Zoom with an electronic rangefinder and focus drive motor
It was bulky and impractical, but was subsequently used by Zeiss in its 1995 Contax AX.

Photokina 1981 saw the Pentax ME-F for the first time, with a powered 35-70mm f/2.8 zoom lens. It too proved sluggish and unresponsive except in bright light. This camera is generally recognised as being the first auto-focus SLR, although this is only true in the sense that it remained in production until 1985. It most certainly was not thee first AF SLR. <see US Patent 4307947 29DEC1981>

Photokina 1982 saw a Canon AL-1 fitted with integrated Quick Focus. It was not a powered lens focus system, the focal plane was defined in the viewfinder, and contrast was evaluated and focus indicated when it reached a maximum. In real life shooting situations it quickly showed its limits however. <see US Patent 4557580 10DEC1985>
CANON AL-1 1982

Also at Photokina 1982 was the Olympus OM30 with a 35-70mm f/4 zoom lens fitted with a phase detect AF drive unit,
and an interesting Contax prototype based on a 1980 137-MD. The AF drive was fitted within the mirror box, and movement to the lens transmitted by a fork located in the bayonet. Curiously Zeiss abandoned what proved a fruitful system. It was however picked up by Minolta, Nikon and Pentax.

At Photokina 1983 it was Nikon's turn to reveal its modified F3 with a bulky TTL AE DX-1 pentaprism. The phase detect AF unit was fitted into the bottom of the mirror box, and the drive fitted into the lens. The CCD sensor was provided by Honeywell.

Then came the Minolta Dynax 7000 AF bombshell unveiled at Photokina 1985. A fully integrated phase detect AF with both sensor and drive unit within the mirror box. It was a huge success, and the engineering architecture was copied by Nikon and Pentax.
Nikon's response at Photokina 1986 was the F501, which used the same Honeywell AF module but this time with the motor in the camera, and a modified F bayonet.
NIKON F501 1986
In 1987 Canon introduced its own MOS AF sensor in its EOS650 QD,
CANON EOS650 QD 1987
& Pentax the SFX which was marketed as the SF1 with the new K-AF mount.
PENTAX SF-1 1988
& in 1989 Nikon their F4.
NIKON F4s 1989
NIKON F4s 1989

Bear in mind that Nikon, Ricoh / Pentax & Olympus all used Honeywell Visitronic AF units. These were sold by Honeywell under license to these companies. Honeywell were only too pleased to make some profit from their TCL2 AF system, which had languished following well over a decade's investment in time and manpower. What Honeywell did after Nikon, Minolta, and Pentax had begun to crack miniaturisation, caught Minolta completely off guard. Minolta had designed and developed their own AF system, used it in their Dynax 7000 AF SLR and sold it in the US badged as the Maxxum. Honeywell waited until sales had taken off before issuing a patent infringement suit. It was a cynical ploy intended to recompense Honeywell for their inability to enter the SLR market a decade earlier. After all why bother going to all the time and trouble of building your own consumer SLR AF cameras when you can exploit the US legal system to milk the innovators.

The suit filed against Minolta by Honeywell has gone down in history as one of the most egregious abuses of patent nonuse. In the case of Honeywell vs Minolta, the compensation price indicated by the jury for damages was $95,490,000 and the judge determined, with added interest for a year, damages amounting to $127,500,000 for 178,000 sets of $535 products. The costs were calculated from court material presented by Honeywell. Objects for compensation included cameras, lenses, covers, and many additional goods not directly related to Honeywell's patented AF technologies. That meant the licensing price for the product was $71 or 13.4% of shipment price. The ratio is generally determined as 3%. During the litigation Honeywell campaigned for a ban on the import of all Minolta Maxxum cameras.

Honeywell could have elected to sue Minolta at any time following their first foray into the AF market, in 1983. But they didn't. Instead they waited until sales of autofocus SLR's and compacts had begun to take off. Minolta decided to fight the suit on the basis that their AF technology was of their own design and development. They rightly argued in court that their technology was fundamentally different to Honeywell's telemetry system. Honeywell's lawyers in turn responded that Stauffer had patented the basis for Minolta's phase detect AF system prior to 1983 (the patent was filed in 1981 but not granted until 1983, by which time Minolta had already designed their own system). Minolta stood their ground, until Honeywell sort an injunction banning the import of Maxxum SLR cameras in 1991. This was yet another cynical ploy by Honeywell, intended to discredit the Minolta brand image. It worked, Minolta decided it was in the interests of the company image to settle rather than stall Honeywell in protracted legal proceedings.

What the other Japanese camera companies, who had been using Honeywell's TCL2 Visitronic modules under license, were slow to realise, was that once Minolta caved into Honeywell's opportunistic litigation, Honeywell would turn on them. That's why they all had to pay additional royalties, until Honeywell's patent rights expired in 1995. It ended costing them dear. What they should have done is file a combined counter suit, and take Honeywell to court in Japan. The outcome would have been very different. Honeywell could get away with muddying Minolta's image in the US, but they would not have been able to do so against the combined might of most of the major Japanese camera manufacturers.

PhotoHistory 1992 (Heisei 4) <> Chronological history of photography in Japan in 1992.
In February, the federal district court in Newark, New Jersey orders Minolta Camera Co. to pay Honeywell Co., Ltd. 96 million dollars in damages for patent infringement concerning autofocus technology. Honeywell also demands patent royalties and a ban on Minolta AF camera sales in the U.S. Minolta agrees to pay Honeywell 127.5 million dollars even while still claiming that they did not infringe on patent rights. They decide to settle to avoid a prolonged court battle and damage to the corporate image.
Honeywell subsequently announces that it will seek settlements from other camera manufacturers such as Nikon, Canon, and Olympus. Honeywell had took the case to court in 1987.
In August, Honeywell announces that a settlement has been reached with Nikon and Canon over its AF technology patent infringement. Nikon is said to have agreed to pay Honeywell 45 million dollars by 1995 when the Honeywell patent expires. Canon accepts a similar settlement. Olympus and Konica are reported to settle out of court. (Asahi Optical, Ricoh, and other firms are supposedly pursuing a similar solution.) Honeywell says that it will receive 124.1 million dollars from a number of firms including Matsushita Electric and Kyocera.

Kyocera took an entirely different approach following the experimental modification of the 137MD in 1982/3.
In 1993 Kyocera brought out the Contax RX which featured a built-in digital focusing aid, "Digital Focus Indicator". Zeiss had eschewed motorised CCD AF technology, opting instead for a means of visually determining best focus by measuring the circle of confusion and presenting the information in the viewfinder, together with the depth of focus in the form of of a pair of Focus Indicator Scale modes. The technology was based on TTL phase difference.
This was followed in 1996 by the Contax AX, a seriously weird Zeiss-Kyocera approach to the problem; move the pentaprism and mirror box towards the film plane and all the associated film transport internals. It had the advantage it worked with any Zeiss Contax C/Y AE mount lens, but despite the speed of the coreless drive motors it could never be as responsive as a dedicated AF lens. The pentaprism movement also seriously effected eye clearance at the viewfinder, and hence field-of-view accessibility.
The TTL PDAF and motor were contained in the bottom of the camera body. This made the body far more bulky than the RX, but still slightly smaller than the Canon EOS 1 or the Nikon F2. Needless to say it didn't catch on! The reason Zeiss took this approach was because they were reluctant to alter the helicoid focusing device in their camera lenses. It was a zero backlash, very lightweight plastic thread which could not take the torque imposed by an AF drive unit & maintain the critical lens spacings. The consequence was this oddball technological blind alley. Ironically of course this technology lent itself to astrophotography with a telescope.

Zeiss eventually gave Kyocera the go ahead to make an AF lens SLR, which resulted in the 2001 N1.
The new body had a different mount, the N mount, incompatible with the previous Contax-Yashica or Zeiss Contax lenses which used the
C/Y AE 48mm trefoil bayonet and had a flange to film plane distance 45.5mm.
The only adapter Contax supplied was the NAM-1 which enabled Zeiss T* 645AF lenses to be used with the N1 body.
If you wanted to trade up to the AF Contax N1 you needed to invest in a new outfit. Pro's opted for Canon or Nikon outfits instead which by this time were all AF compatible and had been for about a decade.

The German SLR camera industry represented by Leica Camera AG in the R series SLR & through Zeiss in the Contax, produced in collaboration with Kyocera, failed to recognise the significance of phase detect AF technology during the 1980's / 90's and discovered too late its crucial importance in DSLR's. Leicaflex and Contax DSLR's could not maintain the pace of development set by the Japanese competition as their sales revenue rapidly diminished during the first decade of the C21st and have consequently fallen by the wayside.
You could of course move up to medium format, the 2008 Leica S2 (Kodak CCD 45 x 30mm 6 micron 7500 x 5000) will set you back a mere 20k.
Why, when you can buy the 2010 Pentax 645D (Kodak CCD 44 x 33mm 6 micron 7264 x 5440, AF SAFOX IX) for 7k?


Fortunately all the patented auto-focussing technologies assigned to Honeywell expired by 1995 and Honeywell has not been able to renew or extend them. The irony is Sony, who are probably one of the world's greatest innovators of electronic technologies, took little part in the development of SLR camera AF, which fell into their lap following the merger of Konica and Minolta in 2004, and their subsequent acquisition of the camera manufacturing and design rights in 2005/6. It was both Konica and Minolta who did most to take ideas and turn them into a workable reality, which is why Sony's PDAF is so much more effective than most of the competition. And if it hadn't been for Honeywell's corporate greed, Minolta would have been in a far better position to weather the Japanese recession during the 1990's.

And what of Honeywell's contribution to the SLR camera industry? There was none really. All Honeywell did was distribute the Pentax Spotmatic, imported under licence from Asahi and distributed exclusively in the USA as either Heiland Pentax or Honeywell Pentax, e.g. Spotmatic F
& Sp500 (as Sp but 1/500s top shutter speed & no self timer), and the Spotmatic IIa & SL (same as SP - no TTL metering)
& ES
& ESII which used Honeywell's Strobonar electronic flash (same as ES but with self timer).
A fitting testimony to a globalised US fortune 100 conglomerate.
Ref: Auto-Focus Patents registered to Norman L. Stauffer & assigned to Honeywell:


The move now is to DSLR Live View and HD Video. Personally I find these features utterly pointless in an optical viewfinder DSLR. After all a DSLR is a digital version of an SLR. The SLR is a stills camera. It is designed to "Capture the Moment." If I wanted to shoot video I'd buy a video camera. The argument I read time and time again trotted out like some Buddist Mantra is, "Photojournalists, particularly the Paparazzi and Sports photojournalists, want to include video clips in their coverage." Photography magazine editors are especially guilty of this, unquestioningly regarding Live View to be desirable.

Fine, but why then use a stills DSLR that shoots HD Video so badly (CMOS staggers) and is singularly ergonomically ill suited to the task? Why not use an HD Video camera and take the stills off the video? Most of their stills are intended for magazines, newspapers, and HD TV. You do not need an Fx format hi-res DSLR for that type of work. And why do top end DSLR's with Live View not have articulated LCD panels? If you're going to shoot video above the heads of the milling Paparazzi, wouldn't an articulated LCD panel be desirable? So why are such features only to be found on entry and enthusiast level DSLR's?

Live View has little to do with the specialist demands of photojournalists. Camera manufacturers are not stupid. They know what their target customer's desire, and they're trying to tempt point and shoot compact digital camera owners to upgrade, so they add features common in current point & shoot cameras. But there is a snag. Phase detect AF will not work with the LCD display set to Live View mode. An ancillary contrast AF system is used instead (with the exception of two Sony entry level DSLR models).

Consider for instance the Canon 450D. This is what the Canon 450D DSLR user has to do to make use of Live View (the same applies to the 500D; 550D; 50D; 5D MkII & 7D). From the custom functions menu enable Live View, press button on the camera back and the mirror flips up to allow light to pass directly through to the image sensor. The Live View is then displayed on the LCD panel. To auto-focus, press <*> button on camera back. If contrast AF mode is selected, it will attempt to auto-focus, if Quick View / Normal AF focus is selected the Live View LCD display will momentarily black out. The latter option is the more rapid of the two and the more accurate. Having achieved the desired focus, you then capture the image.

To be fair this problem is peculiar to Canon (& some Nikon) DSLR models. Canon have been working on the problem. But the problem I have outlined is just one of the reasons why I dislike Live View and especially Live View AF. Basically its a problem that need never have arisen but for the specious demands of photojournalists, and the inability of point & shootist's to comprehend what an SLR camera has that their point & shoot camera doesn't.

Sony were first to work around the Live View auto-focus problem and incorporate their ingenious solution in the 2008 Sony a350, & the 2010 a390. Sony redesigned the Penta-Prism replacing it with a Penta-Mirror, the forward segment being a tilt mirror with an auxilliary sensor. When the reflex mirror is down, the beam from the lens is normally fed through to the eyepiece. When Live View is enabled the Penta-Mirror tilt mirror segment deflects the beam onto the auxilliary PDAF sensor above the eye lens and the secondary imaging sensor. Live View is thus PDAF enabled. The advantage of this system is that the viewfinder image is bright. Sony have stated they have no intention of incorporating Quick AF Live View in whatever DSLR supersedes their a850/a900 models. (As of July 2010 there is rumoured to be a replacement but there is nothing definite).

Canon, as per usual, went down the overly complicated route. On Christmas Day 2008 Canon filed a patent application with USPTO # US2008/0316353 A1 describing an articulated reflex mirror incorporating an auxiliary articulated PDAF feed mirror. I find it difficult to see how the proposed solution could be used with an Fx format sensor. The swept space of both mirrors is about 30% greater than a reflex mirror with a fixed PDAF feed. The patent hasn't been issued yet, but I envisage Canon using this technology only in their APS-C format DSLR's. If it were to incorporated into their APS-H & Fx format DSLR's they would become almost as bulky as a medium format DSLR.


The patent application describes two different solutions.

Embodiment 1:- In Live View mode the reflex mirror is tilted back so as to direct the beam onto the PDAF sensors at the bottom of the mirror box. When using the Penta-Prism optical viewfinder the reflex half-mirror is tilted forward to the usual 45° position and an auxiliary sub-mirror directs part of the beam onto the PDAF sensors. When the image is captured the reflex half-mirror flips up and is articulated backwards whilst the auxiliary sub-mirror is articulated downwards to clear the light path.

Embodiment 2:- The mechanism is similar to that described in Embodiment 1 except the optical path splitting system is configured differently. It comprises a half-mirror, a sub-mirror and the driving mechanism. Whereas in Embodiment 1 the optical path to the optical viewfinder sensor is changed by either directing the beam to the PDAF sensor in the bottom of the mirror box, or onto the sensor, in Embodiment 2 it is kept directed to the PDAF sensor until the last possible moment, when the viewfinder blacks out, and the sub-mirror folds up under the half-mirror. In Live View mode the half mirror splits the beam between the image sensor and the PDAF sensor. The disadvantage of either variant of this system is the viewfinder image is dim because the reflex half-mirror acts as a beam splitter.

Canon's US Patent application followed an earlier patent filed in August 2004 and granted August 2008. <see US Patent 7,414,664> Clearly Canon, well aware of the problem, attempted to devise an original solution, but as of mid-2010 have not implemented it.

Nikon have published no patent application or made known how they intend addressing the problem. If they do not wish to pay royalties to either Sony or Canon they will have their work cut out. Since Nikon procure their sensors from Sony, my gut feeling is they will use Sony's system under license.

Of course this difficulty only occurs because DSLR's follow the traditional 35mm SLR design. If you do away with the SLR optical viewfinder (invented to get around the rangefinder viewfinder parallax problem & viewfinder framing with different lenses), and either use an electronic viewfinder (EVF), or a Live View LCD panel, then you do not need any of these complications because there is no need for an articulated reflex mirror. At present digital viewfinders do not have adequate resolution or refresh rate to satisfy prosumer DSLR users. But that is only a matter of technological development. It is not unrealistic to predict the DSLR in its present form to be a hybrid technology which will ultimately be superseded by cameras akin to the mirrorless touchscreen Panasonic Lumix G2 or the radically different Sony NEX.

Chris Lord

This page was created by SimpleText2Html 1.0.2 on 12-JUL-2010

Top of Page
Return to Home Page