The Alpha 580 – a three-way view

Once I had a quarterplate hand-and-stand camera, vintage 1920s. Attached to the front standard was a small reflex viewfinder, giving a miniature composition you could use at waist or chest level. On the same standard was a folding wire frame, with a companion eye-sighting window flipping up from the side of the body. This gave a direct view from eye level. But for the most accurate framing and focusing, a groundglass screen at back could be used with the shutter open and a viewing hood folded out.

Those three ways of viewing have never been available in a modern SLR. Until now! The Alpha 580 (for which you can also read 560 throughout this review, give or take the sensor) is the first modern SLR to offer three entirely different viewfinder systems, all with their own unique focus and exposure methods. There have been cameras made by Alpa and Praktina which had optical finders tucked in alongside their pentaprism, and Rollei invented a finder which could switch from eye-level to waist level at the flick of a lever. But the Alpha 580 offers three through-the-lens systems and it’s unlikely any DSLR will do so again.
This is a 10-page article – please use the navigation bar at the bottom to move on to the next page, or click the ‘Continue Reading’ link to view as a single long article (this function is not very reliable though and may produce an ‘undefined’ error)


The three finder methods
To clarify, the 580 has the two principal modes of the A3xx and Axx series predecessors – conventional mirror-prism eye level finder, and the so-called Quick Live View on the tilting rear screen. This live view is not fed from the shooting sensor, instead a small CCTV camera sited in the prism housing presents a webcam quality view of the focusing screen itself.

Ken Rockwell found this labelling on Sony cameras annoying, but then, almost anything annoys Ken… funny thing is they have missed out a big extra label for REAL live view with contrast-detection AF
Like the little reflex viewer on the 1920s camera, this allows easy waist or chest level composition, with limited accuracy. The focus screen itself only shows 95% of the final image, and the Quick LV camera crops this further to show just 90%. Depth of field can not be seen any better than by eye (e.g., badly) and although the 580 has a stop-down preview button this function is disabled in Quick LV. But Quick LV keeps the fast, accurate 15-sensor phase detect AF module in operation allowing 5 frames a second shooting.
Since the QV camera system is not the actual sensor, it not only shows various finder markings as seen by the eye, it also does not reflect colour or exposure settings precisely. It can simulate them using manual exposure or +/- control, and use auto gain to operate in low light. It also takes over exposure metering when used, replacing a standard 40-sector honeycomb TTL metering module with 1200 colour sensitive zones. In the latest models including the 580 it allows Face Detection and Smile Shutter. So it’s not all that bad, but the image is only VGA and can not be enlarged to check focus.

Focus Check LV in action, including showing the effect of +1 over-ride and of tungsten WB in this shot
From the A450 (brief blip on the radar with a fixed rear screen), A500 and A550 onwards Focus Check Live View has been added to Alpha cameras at this level. This is true sensor-fed LV showing 100% of the picture with accurate preview of contrast, exposure and colour. It can preview depth of field perfectly and magnify the image 7.5X or 15X (7X/14X for the 560). Until these latest models appeared, AF was not possible in this mode, it was strictly as labelled – ‘Focus Check’ using manual focus.
With the 580/560, contrast detect autofocus has been enabled along with 1080p HD video. I believe this could be have been implemented in the 500/550 generation, as I fed the Focus Check LV via HDMI cable to a television and the output quality was both excellent in quality and not limited to any shorter duration than typical HD clips (in 2009). Sony chose not to, as they did not yet have the lenses needed for contrast detect autofocus on the mainstream Alpha mount system

19 Comments

  • David,
    Thank you for the review and helping me understand my camera better.
    I do have many questions,, but i will only ask one.
    Which Auto-focus mode do you feel is most accurate when using Minolta Primes such as the 50mm f/1.7?
    The reason i ask is from experience with terrible back-focusing when using a a560 with my Minolta lenses ,, which are tack sharp on my a100.
    After reading your review i see it may have just been a matter of me not understanding the camera and the different focus modes.
    To be honest ,, i dont really care about the Live view, i use the optical viewfinder for everything. Yeah, im that old.
    Anyway, thanks again.

    • I always use single shot, centre AF point. The A580 has not proved immune to FF/BF issues though and I am afraid it tends to go with that type of AF sensor despite continued improvements. I have only recently started using the 50mm f/1.4 lens after getting the A77, it’s the only camera I have owned so far which focuses it accurately.
      David

  • The Sigma is a much better lens. Since you have a choice between OS and SSS there really isn’t an issue. The Sigma is so much better at the long end of the range, optically.
    David

  • Thanks David. Now that I’ve gotten an a55, I’ve been thinking of selling my Tamron 18-250 and getting the Sigma 18-250, for the potential of better IQ and less distortion, as well as the OS for video. Your review comments stopped me for the moment. From your reviews, it sounds like you’d recommend this kind of switch (it’s affordable), so long as the video would be in the 18-100 range or so. Am I correct in that?
    BTW – do you still find that you get better IQ at 200-250mm from the Sigma compared to the Tamron? Are they pretty much equal from 18-150? Thanks again for your advice.
    William

  • I did not use a tripod; at times, I used a binocular support (a monopod with a sort of Y-shape at the top) because I was so cold I could not hold the camera still at all. The stabilisation on the Sigma 18-250mm seems fine for movies up to a middle setting, maybe 100-150mm. At a certain point (always at 250mm) it becomes impossible to pan or move the camera without the distorted effect. I think this is due to the type of stabilisation used in that lens, and some other Sigma lenses might behave differently. What I do know for sure is that the in-body SSS does not produce the same effect.

  • David,
    I was struck by your comment about video and the Sigma OS system: “Don’t think that in-lens OS will be your solution to longer clip lengths before overheating shuts the system down. It doesn’t agree with the movie mode, in contrast to SS which works well.”
    Q: Did you use a tripod when making the video on YouTube? If so, could the “ripple” effect due to the OS stabilizer being on and the lens trying to stabilize something that is already still — the same kind of difficulty as leaving the SSS on when using a tripod?
    Is there any reason to get a Sigma OS lens for a55 or a580 video — or is it always best just to go with SSS and a steady hand?
    Thanks, William

  • In terms of camera handling I would rate Minolta 7D as best camera made by Minolta/Sony. Suppose A700 and A900 also were designed by Minolta, but ‘improved’ by Sony and both are already one step down.

  • The D-lens distance encoder is great for ADI calculations, and in that case focus produces a distance measurement rather than the other way around. But the D encoders are mechanical, and are too coarse to allow the camera to finely control focus motor speed for silky smooth contrast AF.
    I feel sure the encoder in the E-mount lenses is a high resolution encoder (probably optical) for that reason. Since contrast AF doesn’t really care about distance, I’m pretty sure the LA-EA1 and Alpha cams don’t even use the D encoders.
    With much less mechanical backlash (gear slop) than body-driven lenses, the SAM and SSM lenses contrast focus smoother and quieter (and the LA-EA1 with not drive motor can only focus with them anyway). But E-mount lenses will always contrast focus much better than any of them because there is precise feedback of focus group position and speed to the camera.

  • Excellent article! The best ever explanation of the three viewing systems in use on these newest Alphas.
    One thing regarding the suitability of SSM and SAM lenses for contrast detection AF is the lack of a focus group position encoder in the 8-pin A-mount lens system.
    The E-mount lens system is most suitable for contrast detect AF not just because the focus motor is smooth, but because the control system has knowledge of the exact position of the focus group.
    I believe that is the purpose of the extra 2 pins on the E-mount lenses. It is the A and B phase of an electronic position encoder on the focus group.
    Such an encoder allows the control system to move the focus group very precisely in executing the contrast AF scheme. All A-mount lenses lack this encoder (not needed for phase detect AF). A side benefit of the encoder is the camera body can enter manual focus check magnification automatically whenever the focus ring moves the group.

    • The D-lens specification is supposed to report a fairly accurate focused distance and the initialisation of the A-mount lens, on startup, has always included a ‘focus range check and park’ process. We need to wait and see how things develop. Pentax devised a very good way to get single shot CDAF with any lens, including screw drive – a simple process by which the camera does a ‘big’ focus range check, maps the result; homes in on the approximate position for sharpness, and does a second smaller sweep around around this; then it moves to the detected peak contrast position and if necessary does a tiny final shuffle; and locks. I don’t think the story is over yet.
      There are two obvious options open to camera designers. One is very simple – detect focus from the focusing screen in an SLR design. Optical devices like the split image RF or microprism create phase contrast in a form the eye can see, and what the eye can see can also be detected by a sensor. The second option is go down the Contax N body route, and build a fine focus mechanism into the sensor carriage – have a sensor which can move forward of backward over a small but functional range such as 3mm. Used with a manually focused lens this could allow enough video focus adjustment with shorter focal length (below 100mm) lenses to enable video CDAF.

  • Now that you have reviewed extensively both the Alpha 55 and the Alpha 580 which would you choose if you needed only one camera for family and vacation shots, no studio?

  • The 2s timer for MLU would not work for me. So often, I raise the mirror and wait for the right moment, perhaps waiting for the wind to drop or for someone to move into/out of frame.
    Thanks for a couple of really interesting reviews – this one and the A55.

  • With all the latest Sony models – NEX, SLT, 560, 580 – you do not see image noise in typical shooting conditions until ISO 3200, provided you process the raw file sensibly. That does not mean cranking up to maximum NR either. I am just going to add a link to a full size version of the antiquey furniture shop interior which is at 1600. I’m sorry these large file download links are restricted to subscribers, but I need to limit bandwidth and obtain subscriptions alike. Anyone with an existing YUDU or Payloadz subscription, or a magazine subscription paid up, can email me [email protected] and I will manually upgrade their Free registration on the site to Normal or Premium status as appropriate.

  • David:
    That report was extremely interesting (and for me timely). Especially enlightening was your discussion of High ISO low light handling by Sony vis-a-vis Nikon’s strategy. This would explain some of the results I’ve seen on various sites and from photos I’ve received from people working with the A580. The statement you made about ISO 1600 being virtually noise-free caught my attention. My A850 has fairly strong noise from ISO 800 and above (raw, right out of the camera). So I was wondering can you quantify the raw file low light ISO advantage of the A580 over the A850 (or A900) in terms of stops? I promise this will be my last low light question.
    Regards
    John

  • I would figure that the 2s timer would be enough to tame most vibrations, even with FCLV’s double-dip of the mirror. Of course, once you nail the focus, you could leave FCLV, wait a few extra seconds, then trip the shutter. Yes, it’s not as good as having the a900’s multi-fire MLU mode, but I’d say it’d get the job done for 99.9% of the people who relied on 2s timer before.

  • Bingo! It’s only a 2 second respite, but in FCLV mode combined with the 2 second selftimer, the mirror lifts for the 2 seconds and the shutter closes – both optical and screen finders are black. Then the shutter fires and the mirror briefly flips to return to FCLV mode.
    This doesn’t work with the wireless remote release, if set to remote and the 2 sec button used, you just get a rapid-bleep 2 sec delay and normal firing. It does work with the wired cable release.

  • My error – I was thinking about the lack of it in Focus Check LV mode when testing this. I’ve altered the paragraph involved to state this. For me, not having the mirror drop and rise in FCLV mode would be the most important (probably impossible) change. That’s because FCLV is the focus and viewing mode I would use on telescopes, microscopes, macro bellows and similar setups needing zero vibration. Of course I have also been working with the Alpha 900 for two years, and that has a true mirror lockup mode not just a pre-lift.
    But you have given me an idea. I have not tried combining the FCLV mode with 2 second mirror prelift or 10s self timer. I’ll check this out right away and add something to the review.

  • David,
    The a580/560 does have Mirror Lockup in the 2s self timer. It’s right on the Sony specs pages for the cameras.

Leave a Reply