Sony introduces the a9

[Ilkka Nissilä]

One problem with that wish-list is that it is irrational to have IBIS in a mirrorless camera.

From a complexity and cost point of view, perhaps. But since there is an EVF and no mirror (and in the case of the A9, the mechanical shutter doesn't have to be used), the camera itself is causing no vibration, and so it may be possible to shoot successfully hand-held at slower speeds than with a DSLR. To facilitate this, compensation for hand shake with IBIS is a reasonable approach.

However, personally I prefer fast shutter speeds that stop the movement and wide aperture to isolate the subject (to create images where the subject is sharp and surroundings are blurred), instead of broad depth of field and slow shutter speed, which results in a sharp surround and a blurred subject. So for me VR is mostly needed in medium telephoto lenses where subject-freezing shutter speeds may not be enough to stabilize the viewfinder and eliminate hand shake. I do use VR occasionally with shorter lenses and while it can be useful, it is not a large effect. For static subjects I use a tripod for much greater options of depth of field and to eliminate remaining uncertainty over shake (with non-supertelephoto lenses at least). I think with any image stabilization system whether sensor-moving or implemented in the optics, cannot guarantee sharpness the way a good tripod can, and sometimes introduces a bit of blur in itself (due to noise in the measurement of acceleration and imperfect compensation). This is one of the reasons why I systematically use a tripod for static subjects instead of relying on technology. Further advantages of the tripod approach is more precise and reproducible framing and composition, the possibility of capturing several frames at different exposures or with different focus, without the composition drifting, and the ability to use tilt for enhancing the sharpness of near-to-far scenes precisely (hand held, it would be very difficult and results more random than I'd like).

But I can see the benefits of IBIS for certain types of e.g. environmental portraits, etc. using a shorter lens, and I'm sure I'd come up with uses for the technology if it were implemented in my cameras.  But I'm unlikely to choose a system based on this feature by itself. The optical viewfinder is far too important for me. I might buy a mirrorless camera for the silent shooting feature, once mature and not ridiculously priced. I expect that the lack of moving parts would translate to lower purchase and service costs for the user, but I'm not sure if this is happening yet. I like to shoot concerts and dance, and there have been circumstances where the camera sound has been a distraction to someone sitting near me, which is where a camera like the A9 would be excellent option. But the Sony is very expensive! 5300€ for the camera body and 2000€ for a 135/2.8 lens.

[bjornthun]

IBIS is definitely a rational thing to have in a mirrorless camera. As, Ilkka says, environmental pirtraits. Another application is when the use of a tripod is impractical, or simply not allowed. Sometimes even flash photography is not allowed.

I think that the A9 may find uses where a silent camera is desired, and some of the venues where IBIS is useful could coincide with where a totally silent camera is desired. This could include some sports like like golf or tennis, where you're not allowed to disturb the players, not even with the sound of a shutter. These are of course different sports from what we see in Sony's commercials.

Personally I wouldn't buy a camera without IBIS, since IBIS adds flexibility. Minolta/Sony and Pentax offer DSLRs with IBIS, so the feature is possible with an OVF as well as an EVF.

[MFloyd]

IBIS, the new techno acronym.  I'm not sure, if the generic IBIS, is as efficient as the lens optimized VR.

[David H. Hartman]

However, personally I prefer fast shutter speeds that stop the movement and wide aperture to isolate the subject (to create images where the subject is sharp and surroundings are blurred), instead of broad depth of field and slow shutter speed, which results in a sharp surround and a blurred subject. So for me VR is mostly needed in medium telephoto lenses where subject-freezing shutter speeds may not be enough to stabilize the viewfinder and eliminate hand shake. I do use VR occasionally with shorter lenses and while it can be useful, it is not a large effect. For static subjects I use a tripod for much greater options of depth of field and to eliminate remaining uncertainty over shake (with non-supertelephoto lenses at least). I think with any image stabilization system whether sensor-moving or implemented in the optics, cannot guarantee sharpness the way a good tripod can, and sometimes introduces a bit of blur in itself (due to noise in the measurement of acceleration and imperfect compensation). This is one of the reasons why I systematically use a tripod for static subjects instead of relying on technology. Further advantages of the tripod approach is more precise and reproducible framing and composition, the possibility of capturing several frames at different exposures or with different focus, without the composition drifting, and the ability to use tilt for enhancing the sharpness of near-to-far scenes precisely (hand held, it would be very difficult and results more random than I'd like).

+1, very well said.

Underline = extra agreement or ++1

Dave Hartman

---

"Another application is when the use of a tripod is impractical, or simply not allowed. Sometimes even flash photography is not allowed." --bjornthun

I find situations where I can't set down a tripod and VR really helps so another +1.

---

It's often said that VR can't stop subject motion. A tripod can't either unfortunately.

Perhaps VR is better compared to a monopod than a tripod. Neither VR nor monopod holds the composition as a tripod does. Both give a few extra stops of help with hand and camera shake.

VR should come with a guaranty that it will always deliver one from hand shake. It's not there yet.

---

I'm finding the A9 interesting but I'm still doubting the EVF. I'd like to have a look. I wonder if Best Buy will stock it. Otherwise I'd have to drive about 75 miles to see one.

Dave Hartman

[Ilkka Nissilä]

One possible drawback of IBIS is that since the sensor is able to move, heat management can be more difficult.

Jim Kasson has been studying the dark noise from various cameras and I recall that A7RII has some increase in dark noise in a series of long exposures relative to the A7R.  A7RII has IBIS which wasn't present in the A7R.

http://blog.kasson.com/the-last-word/a7r-self-heating-for-long-exposures/
http://blog.kasson.com/the-last-word/a7rii-self-heating-for-long-exposures/

Now, it might not be an issue for most people who don't do e.g. astrophotography with a series of long exposures, or long exposures at night, but the same phenomena may be affecting the camera during recording of 4K video; although I have no idea of how common this problem is but some Sony users are reporting overheating of the camera in some circumstances. I don't know if the heat is generated in the video processor or the sensor, battery, or all of them. However, I suspect IBIS plays a part in making it more difficult to keep the sensor cool during extended usage since one can not fix a large area heat conduction path from the components if they have to have significant freedom of movement.

Anyway, this is a minor issue but as 4K recording gets more common, heat issues need to be attended to. I haven't had camera heating problems as I don't do much video, but I did experience the SB-900 overheating when I was shooting  formals for a July wedding in bare sunlight. Before, I had never experienced overheating of the flash as I had been using it for some time only indoors, and I didn't use the full flash energy, but outdoors it became necessary to squeeze as much as possible from my three speedlights and the SB-900 dropped from the game very quickly. I since got SB-910 and SB-5000 which contain progressive improvements to the issue; I think the SB-910 has changed hardware to minimize the issue and the SB-5000 finally has active cooling so one can use it with full output without concern. I think Nikon managed this problem quite well although the best efforts are seen in the newer flashes.

I think IBIS is best fit to a mirrorless (or SLT) camera as if you implement IBIS in a DSLR, the viewfinder no longer shows the precise framing of the image and it may also lead to problems in positioning the AF sensor point accurately on the subject unless the AF point position markers are moving according to sensor movement also in the viewfinder. In a mirrorless camera system, the merits of IBIS are more clear than in a DSLR and I feel it makes sense for Nikon and Canon to stick with in-lens VR / IS technology for DSLRs to keep the accuracy of the viewfinder and AF point position. Finally since in-camera stabilization was invented by Minolta and further developed by Olympus and Sony (if I've understood the history correctly), I think there would likely be license fees for Nikon and Canon to pay if they want to implement IBIS in their cameras. When they make mirrorless systems, they may consider it although not invented in house. Canon made a few IS prime lenses (24mm, 28mm, and 35mm) and Nikon the 24-70/2.8 VR and 16-35/4 VR to cover applications where stabilization is needed on a shorter lens. There are also f/1.8 stabilized prime lenses (35/1.8, 45/1.8 and 85/1.8 VC) from Tamron available.

[Akira]

With all the drawbacks of IBIS, I have to admit that it is addictive.  The IBIS in Olympus E-M5 MkII was fantastic.  True that the IBIS won't contribute to reduce the subject movement, its positive effect is undeniable.  It worked admirably even with the adapted Nikkor Ai 300mm/f4.5 and Samyang 7.5mm/f3.5 fisheye so long as you set the focal length correctly.

I didn't find any heat management problem with E-M5 MKII because of the floarting sensor, neither with Panasonic GX8, possibly thanks to their smaller m4/3 sensors.

[Ilkka Nissilä]

With all the drawbacks of IBIS, I have to admit that it is addictive.  The IBIS in Olympus E-M5 MkII was fantastic.  True that the IBIS won't contribute to reduce the subject movement, its positive effect is undeniable.  It worked admirably even with the adapted Nikkor Ai 300mm/f4.5 and Samyang 7.5mm/f3.5 fisheye so long as you set the focal length correctly.

Point taken.

I didn't find any heat management problem with E-M5 MKII because of the floarting sensor, neither with Panasonic GX8, possibly thanks to their smaller m4/3 sensors.

It is likely that there is less heat generated by the smaller sensor. Also I suspect the problems with larger sensors that are occasionally reported mainly affect extended long-exposure shooting and long continuous 4K recording (possibly in hot climate outdoors during the day) and are likely something that most users will not run into in their day-to-day use. Regarding astrophotography long exposure noise management, there seems to be an issue with median filtering

http://www.sonyalpharumors.com/specific-a7sii-astrophotography-fix-request/
http://blog.kasson.com/the-last-word/sony-a7rii-long-exposure-spatial-filtering-with-fw-3-30/

These may be the side effects of Sony trying to mitigate long-exposure noise with algorithmic processing in the firmware, leading to stars being mistaken for hot pixels and erased by the algorithm.

Nikon used to have shadow clipping at high ISO settings in older DSLRs just a few years ago but in the new cameras (e.g. D810) this has been reduced to a minimum and in the D5 I don't recall that it is present. I suspect Nikon used shadow clipping at high ISO to reduce visibility of banding and other types of artifacts in general photography but this meant that averaging long exposure high ISO shots to bring out nebulae then didn't work as well as it should have. In the newer cameras they've been able to solve these problems in a way that shouldn't hinder astrophotography.

I think it would be better if manufacturers offered an option to shoot and store a "true RAW" file with no processing whatsoever but this hasn't been the case in general. I guess they're worried that sensor testers find out about the noise and the product's reputation suffers, or that people use the raw RAW file and run into the problems which could have been mitigated by processing in firmware.

[Akira]

I shot a ten-minute movie with NEX-5R in Tokyo during a very hot summer day (35 degree Celsius, possibly over 40 degree on the asphalt), and the camera became formidably hot.  After the shooting, I had to remove the battery, leave the battery door open and raise the tilting LCD screen to cool the camera down.

Also, I always wonder how the camera can tell the hot pixels from the stars when it does long exposure noise reduction.  For the landscapes of starry nights, there should be no problem if some stars are erased, but for the astronomical photography, it should be a serious problem.  You would have to stack the frames or use a dedicated camera.

[longzoom]

Tony Northrup at Sonyalfa Rumors: A7rII vs A9 - A7rII is sharper, more DR, less noise at high ISO. And way cheaper, I'd add.  So its what I've already said - I do not need 20fps, for such the huge loss of quality, to summarize. Let us see the direct comparison on the field.  LZ

[bjornthun]

I shot a ten-minute movie with NEX-5R in Tokyo during a very hot summer day (35 degree Celsius, possibly over 40 degree on the asphalt), and the camera became formidably hot.  After the shooting, I had to remove the battery, leave the battery door open and raise the tilting LCD screen to cool the camera down.

Also, I always wonder how the camera can tell the hot pixels from the stars when it does long exposure noise reduction.  For the landscapes of starry nights, there should be no problem if some stars are erased, but for the astronomical photography, it should be a serious problem.  You would have to stack the frames or use a dedicated camera.

Actually, that's a problem, hence the infameous "star eater" algorithms.

[Akira]

Tony Northrup at Sonyalfa Rumors: A7rII vs A9 - A7rII is sharper, more DR, less noise at high ISO. And way cheaper, I'd add.  So its what I've already said - I do not need 20fps, for such the huge loss of quality, to summarize. Let us see the direct comparison on the field.  LZ

I didn't know Tony Northrup (of Tony & Chelsea?) has been involved with Sony Rumors.

Actually, that's a problem, hence the infameous "star eater" algorithms.

That problem is not only with Sony, but applies to the algorithm in general?

[bjornthun]

Tony Northrup said more:

1. shutter & mechanical shutter on a9–DR, ISO 25,600 noise, rolling shutter. No e-shutter penalty except sync
2. Sony a7R II vs a9 raw: a7R II is sharper, more dynamic range, lower noise at high ISO. Still want a9 bc it works better & IQ is still great!
3. Checking results, Sony a9’s focusing equals 1DX, D5 & D500 (the best in the world) + higher framerate (10-20 FPS real world) + no blackout 👏
4. Our #SonyA9 review will be up 4/27 @ 11am ET because embargo. #SonyAlpha

In point 2, IQ of the A9 is still great, i.e. no big loss in image quality.

Note also point 3 about AF.

Akira, I don't know enough about noise removal algorithms in digital sensors. I guess the "star eating" may be a problem for scientific astro photography.

[longzoom]

Tony Northrup said more: In point 2, IQ of the A9 is still great, i.e. no big loss in image quality.

           The rest is priority of the second kind, as for me, of course. Direct image quality is not equal to the D750, so far. The next Sony's A9R with its 43MP will be close to what I, personally, is waiting for Nikon's mirrorless. Or A99II, with its pellicle mirror. In this case Nikon doesn't need any adapter for the last AFS-G lenses. Will see soon, I believe. LZ

[David H. Hartman]

For Nikon shootests who haven't warmed up to the electronic view finder ready yourselves for an EVF replacement for the D810 or D750. Nikon traditionally rolls out new technology in a lower model SLR so I think this trend may follow for new technology in a dSLR.

I look for a Nikon D5s and D500s not a mirrorless replacement. As interesting as the A9 is I don't think it's ready for most sports venues.

Dave Hartman who would like to be an Ambassador for Sony so he could be a Double Agent for Nikon.

[simsurace]

IBIS, the new techno acronym.  I'm not sure, if the generic IBIS, is as efficient as the lens optimized VR.

I would be inclined to say that it's probably more efficient, and especially efficient when combined with in-lens stabilization. E.g. On the Olympus cameras with some of the new supertele lenses you can attempt shots at shutter speeds that are completely unthinkable with Nikon's current VR implementation. I don't know whether IBIS makes sense on a DSLR, maybe you need a continuous stream of image data so it would only be applicable to LiveView. So I don't know whether Nikon will ever want/be able to compete in this regard. Moreover, moving the sensor to stabilize the image does not have the same detrimental effects on OOF rendition that the Nikon VR sometimes exhibits.