[HN Gopher] Sony announces a9 III: first full-frame global shutt...
___________________________________________________________________
Sony announces a9 III: first full-frame global shutter camera
Author : ksec
Score : 228 points
Date : 2023-11-09 04:50 UTC (18 hours ago)
(HTM) web link (www.dpreview.com)
(TXT) w3m dump (www.dpreview.com)
| geerlingguy wrote:
| I love seeing global shutter on this camera. It's been amazing
| getting shots with cameras with global shutter for me.
|
| Unlike film grain, the weird artifacts for sports/action, camera
| moves, rotary motion... it's all bad looking and disappears
| completely with global shutter.
| buffington wrote:
| Not just global shutter, but shutter speeds up to 1/80000,
| flash sync the same (which is simply bonkers, but possible due
| to global shutter), and pre-recording?
|
| I love my A7IV, but the A9III makes it look like a toy.
| semi-extrinsic wrote:
| FWIW, you could do shutter speeds like that on the Canon
| PowerShot series of cameras all the way since ~2005 due to
| the unofficial CHDK - Canon Hackers Development Kit.
| sho wrote:
| While I love ridiculous specs as much as anyone, let's
| remember that 1/80,000th of a second is.. nuts and as far as
| I can find there's only one flash unit that can even do that,
| the profoto pro-11, which is a $20k proposition before you
| even buy the damn camera. And you will get very, very far
| from its maximum output for that kind of duration.
|
| So it's very cool, yes, but unless you have a ton of money to
| spend and a very very good reason to need such extreme
| speed.. you're never gonna use it.
|
| I am certainly looking at my trusty but aging a7sii with an
| increasingly contemptuous eye, though...
| numpad0 wrote:
| Flash doesn't have to be faster. Max flash sync speed means
| up to this speed the shutter is guaranteed to be fully open
| when exported flash signal triggers the flash. Below this
| speed is totally fine, above this could result in partially
| lit images.
|
| e.g. A camera's max sync speed is 1/60s, and you connected
| a flash unit over sync cord, and set the camera to 1/250s,
| ignoring the lightning bolt sign on the dial. You take a
| picture, and alas, top half/left half is pitch black and
| only bottom half/right half is shown correctly, because you
| exceeded the max speed and mechanical circuit to trigger
| the flash did not close until image was already partially
| taken.
|
| Or, e.g. A camera's max sync is 1/80ksec., with a random
| flash unit on top. You set it to that 1/80ksec, push down
| the shutter, the camera triggers flash, flash starts
| discharging energy onto Xenon gas tube. As the light
| intensity goes up, the image processor on the camera pulls
| down internal !global_shutter_trigger signal to lock image
| data onto per-pixel RAM on the sensor. 1/80ksec. later, the
| signal is released and intensity is recorded for retrieval.
|
| These faster sync speeds are useful for situations when a)
| you want to use flash && b) the room/environment is too
| bright for a flash. Now you can keep ISO value to not way
| low, aperture kept open to F2.0, etc. without overexposing
| the image(can't do F0.3 at 1/80k; aperture is limited to up
| to F1.8 with some body recognized lenses attached,
| according to product pages)
| nayuki wrote:
| FYI you cannot use the abbreviation "1/80ksec". It may be
| interpreted as (1/80)ksec or 1/(80ksec). But you actually
| intended to mean (1/(80k))sec.
|
| The correct ways to write it include 1/80000 s, 1/80 ms,
| or 12.5 ms.
| sudosysgen wrote:
| Isn't it 12.5 microseconds?
| nayuki wrote:
| Yes. Edited now.
| mnw21cam wrote:
| The only reason you need a flash duration that short is if
| you really want to freeze something that is moving very
| quickly. You don't need the shutter duration to be that
| short as well, unless you also want to exclude light that
| hasn't been produced by your flash.
| fassssst wrote:
| It's for when you want to shoot at huge apertures in full
| sunlight or directly into strong lights without having to
| grab an ND filter.
| geerlingguy wrote:
| For me, I've often wanted to shoot with flash at 1/500th or
| so outdoors, but most mirrorless cameras only go up to like
| 1/160th. Back in the DSLR days, there were a few camera
| bodies that would go up to 1/500th. Its insane to see flash
| sync beyond 1/1000.
| formerly_proven wrote:
| Note that a full power strobe will burn about 1-1.5 ms,
| so below 1/1000 you start to loose the ability to balance
| flash versus ambient via shutter speed.
| vouaobrasil wrote:
| Weird artifacts and rotary motion are for all PRACTICAL
| purposes gone with fast, stacked shutters as in the Nikon Z9.
| The global shutter's main benefit is flash sync. Grain has
| nothing to do with global shutter....
| Espressosaurus wrote:
| Even with a fast readout, PWMing LEDs can still present
| issues for cameras like the Z9. The global shutter removes
| that concern entirely with no need for fine-tuning--still
| need some logic to identify the peak brightness but don't
| need to fine tune frame rates by fractions of a Hz to reduce
| banding.
|
| A global shutter has been the holy grail for awhile for
| cameras like this in wide production. Yes, you can get a
| machine learning camera with 20 megapixels and a global
| shutter, but you'll need to strap a whole PC to it to do the
| data capture. And it'll be manual focus and manual aperture
| to boot.
| zuminator wrote:
| Far as I can tell, OP isn't saying that grain has anything to
| do with global shutter, they're saying that rolling shutter
| produces artifacts, which unlike other photographic artifacts
| such as the often pleasant ones produced by film grain, are
| not at all aesthetically pleasing.
| geerlingguy wrote:
| This is what I was getting at, yes.
| jillesvangurp wrote:
| > Global shutter is a method of ending an exposure where all the
| image data is captured simultaneously. This is distinct from most
| existing shutter systems (electronic and mechanical) that start
| and end progressively: working their way across the sensor.
|
| Buried at the bottom of the article. This is interesting for
| especially video and sports photography. Some cameras with slower
| sensors suffer from the rolling shutter effect where fast moving
| subjects get distorted because the subject or the camera is
| moving. Think e.g. lampposts looking diagonal when shot from a
| fast moving car. Capturing all the pixels in one go solves that.
|
| Cropping the sensor is a common way to counter rolling shutter
| with current cameras. Basically, that means you are not using all
| of the pixels on the sensor and capture a smaller area. Aside
| from the quality, this also means that you don't use everything
| the lens captured; it's cropped away. So, it looks like you are
| using a larger focal length than you actually are and you can't
| capture at the full width of the lens.
|
| Lots of cameras use cropping to support higher resolutions and
| frame rates because they can't read out the sensor fast enough.
| This new sensor can use the full resolution and doesn't have to
| crop.
|
| And like with photography, you use shutter speed and aperture to
| control what the video looks like. 1/30th of a second is actually
| a relatively long exposure for a photo. And that's also
| potentially a lot of rolling shutter to deal with. An instant and
| fast readout means you have more wiggle room to play with this
| creatively or shoot in very bright or low light.
| fullspectrumdev wrote:
| Thanks for the explainer - I was wondering what the
| significance was, and something I do is mess about with cameras
| for fun.
| cubefox wrote:
| The article also mentions the downside of global shutter: lower
| dynamic range.
| Espressosaurus wrote:
| Not just the global shutter, but 120 FPS with full AF and AE is
| huge too. The A9III is going to be the best sports camera out
| there for a few years I think. Of course, you're going to pay
| through the nose if you want access to that tech.
|
| I can't wait for them to gain back the dynamic range and for it
| to roll down to less expensive cameras in the future.
| clan wrote:
| Agreed. Worth mentioning as well: flash sync speed. As you are
| not dealing with a shutter or progressive capture this will be
| a thing of the past.
| dheera wrote:
| > Cropping the sensor is a common way to counter rolling
| shutter with current camera
|
| Not in my experience. For full-frame 35mm digital cameras, the
| most common way to counter electronic rolling shutter effects
| is, ironically, a very fast _mechanical_ shutter.
|
| In digital cameras with mechanical shutters, which include all
| Canon/Nikon/Sony/etc. consumer DSLRs and mirrorless cameras,
| the electronic shutter programmed to stay open for much longer
| than the mechanical shutter, which allows there to be a time
| instant when all pixels are simultaneously electronically open.
| The mechnical shutter then opens and closes during that time.
|
| The rolling shutter on a Sony a7 is about 26ms. Mechanical
| shutters can beat that easily, and can be as fast as <0.2ms.
|
| (Electronic shutters can be faster on a per-line basis. But
| they will still take 26ms to roll through the image for a
| rolling shutter sensor. So if your electronic shutter is set to
| 0.01ms you'll still be seeing 0.01ms exposures at each line but
| it will take 26ms to roll through the image.)
| Espressosaurus wrote:
| Cropping is common in video. For stills, you're correct.
| mnw21cam wrote:
| I don't think you're right about mechanical shutters being
| faster. The type of mechanical shutter on most cameras like
| this is a focal plane shutter[0]. It has two curtains - an
| upper one and a lower one. When the shutter is closed, the
| two overlap. For a long exposure (typically anything longer
| than 1/100s although some cameras are faster) the lower
| curtain will fall all the way, then the upper curtain will
| fall to close it off again. But at faster shutter speeds, the
| upper curtain will start falling before the lower curtain has
| reached the bottom, and so you can get exactly the same
| rolling shutter effect with a mechanical shutter as with an
| electronic shutter. For very fast shutter speeds, the
| exposure is a narrow slit between the two curtains that
| travels down the image at a much slower speed than the
| exposure time, and this is why most cameras can't synchronise
| with a flash at speeds faster than 1/100s or so. So therefore
| the type of mechanical shutter in most DSLRs and mirrorless
| cameras effectively have a rolling shutter transition time of
| 4-10ms, which is much more than the 0.2ms you claim.
|
| Video-specific cameras are where this might be improved,
| because you can have a shutter which is a rotating disc, and
| this can allow faster opening/closing times, because it
| doesn't need to speed up from a standstill for each frame.
|
| There exist central shutters which are mechanical shutters
| that act as a global shutter, but these are unusual because
| they need to be built into each lens you attach to the
| camera. These are typically limited to a maximum speed of
| 1/500s.
|
| [0] https://en.wikipedia.org/wiki/Focal-plane_shutter
| vouaobrasil wrote:
| Most higher end cameras like the Panasonic G6, Nikon Z8/Z9,
| etc. hardly have any rolling shutter. You'll certainly NEVER
| see rolling shutter in any PRACTICAL application for video
| unless you pan like a true madman, either in photo or in video.
| (I am a professional photographer and videographer....)
|
| In fact for sports and wildlife, you'll NEVER see rolling
| shutter in pretty much any existing camera, either in photo or
| video. For photography it's been like this for decades and for
| videography recently.
|
| Where rolling shutter does appear is in the use of the
| electronic shutter only, but the Nikon Z9 already solved that,
| same with the Sony a1.
|
| The main benefit is flash sync and a very SLIGHT improvement in
| some types of distortion when shooting at very high shutter
| speeds. I'd argue that the Nikon Z9 is a much better camera if
| not wanting flash sync due to its lower base ISO.
|
| So to reiterate, the main benefit is not the reduction of
| rolling shutter, especially since any camera in this price
| point already has minimal rolling shutter.
|
| Global shutters are COOL, and the reduction of rolling shutter
| is COOL, but not the main point, though of course most future
| cameras will probably switch to global shutters at some point
| if the base ISO can be lowered.
| andrei_says_ wrote:
| Global shutter allows for much higher flash sync speeds.
|
| I shoot dance, often in close proximity, and rolling shutter
| on canon r6 was a reason enough to upgrade to r6ii.
| onetimeuse92304 wrote:
| > I'd argue that the Nikon Z9 is a much better camera if not
| wanting flash sync due to its lower base ISO.
|
| Yeah. I think a lot of people focus too much on those "high
| end" specs and too little about small, practical, quality of
| life things.
|
| I remember how I instantly liked Fuji X100 for its built in
| ND filter. Yes, you can screw a filter on by yourself. Then
| you suddenly find yourself on a sunny beach and you wish you
| had your photo bag with you.
|
| Not saying global shutter isn't nice feature, but that 250
| base ISO can be a bit of a hassle depending on what kind of
| photography you do just like not everybody is going to
| benefit from the new shutter.
| ISL wrote:
| The faster available shutter speeds will compensate for the
| faster base ISO except where the superior dynamic range of
| a lower ISO is needed or when a longer shutter speed is
| simultaneously desired.
| vouaobrasil wrote:
| Well to be fair if the base ISO is 250, even the typical
| 1/8000 will be enough for most purposes except maybe
| f/1.2 in sunlight....
| onetimeuse92304 wrote:
| True that, unless you also want slower shutter speed for
| some reason. But that's pretty rare.
| PaulHoule wrote:
| Never could get why people liked ND filters until I got a
| 50mm f/0.95 lens for which exposures are scary short close
| to sunset...
| bearmode wrote:
| They're pretty essential for any situation where you
| can't just stop down the exposure in the camera. If you
| want a long exposure shot of some crashing waves on a
| bright day for instance, you might not be able to just
| increase the shutter speed or narrow the aperture without
| ruining the intended photo.
| onetimeuse92304 wrote:
| I personally like to make dreamy photos wide open with
| relatively long shutter to accentuate movement. Some of
| it in full sunlight. Yes, ND filter is essential then.
| nayuki wrote:
| It's not essential. You can take hundreds of short-
| exposure photos and blend them in software.
| lvl102 wrote:
| It's not a "very SLIGHT improvement" as you write. It's VERY
| SIGNIFICANT. Even the Nikon Z9 has roughly 4ms rolling
| shutter speed. Light travels faster than you imply here. A
| lot faster.
| sixothree wrote:
| At 120 fps that's half the frame time and probably the
| entire shutter time.
| davidgay wrote:
| The Z9 has no (mechanical) shutter.
| sixothree wrote:
| I mean is 1/250 "shutter speed" which seems more normal
| for lighting than 1/120. But I guess for video it doesn't
| matter and could be 1/120.
| yakubin wrote:
| For video you typically set shutter speed to half a
| frame. The most common frame rate is 24FPS, so you get
| 1/50 shutter speed. It's lousy for a photo, but good for
| video , since you want motion blur, so that the video
| isn't janky. You control light through ND filters and
| LED/tungsten lights, not through shutter speed, since
| shutter speed affects motion blur and you want to make
| that choice independently of exposure.
|
| 120FPS is good for sports or wildlife, but those are
| special. For normal stuff, 24FPS is good. And even with
| wildlife, you're probably not recording at 120FPS all the
| time, only during quick action. Otherwise you're just
| wasting space on your memory card with identical frames.
| Battery as well.
| AnimalMuppet wrote:
| It's not the speed of light that matters. It's the speed of
| objects in the field of view (or, more specifically, the
| speed of their image on the sensor compared to the speed of
| the shutter across the sensor).
| zero_iq wrote:
| Also worth noting that, unlike physical objects, images
| are not bound by the speed of light. Patterns of light
| and shadow can move across a sensor at unrestricted
| speeds.
| neogodless wrote:
| I'm confused what this means. Are patterns of light and
| shadow not also light, and bound by the speed of light
| (on the upper end)? How can patterns consisting of light
| (or the absence of it) move faster than light?
| yakubin wrote:
| https://physics.stackexchange.com/a/48329
|
| In other words, speed of a projection of light from 3d
| space to 2d space may be higher than the original speed
| in 3d. (Because one dimension gets squished to 0, so
| movement in this dimension is perceived to be instant.)
|
| It's like a diagonal of a cube 1x1x1 has length sqrt(3),
| but if you apply orthogonal projection onto R^2, its
| image will be a diagonal of a square and it will have
| length sqrt(2). Shorter distance -> shorter time to
| travel.
| jmholla wrote:
| > It's like a diagonal of a cube 1x1x1 has length
| sqrt(3), but if you apply orthogonal projection onto R^2,
| its image will be a diagonal of a square and it will have
| length sqrt(2). Shorter distance -> shorter time to
| travel.
|
| This example doesn't make sense to me. In that analogy,
| wouldn't anything on that diagonal appear to move more
| slowly in 2D than the same thing moving along the
| diagonal of a face? The cube diagonal would make it move
| farther than it does in 2D space.
|
| I remember seeing a simulator in my optics class that
| combined multiple wavelengths of light. The interference
| pattern moved faster than the speed of light, but that
| was fine because information wasn't moving faster. That
| was just the result of adding them together.
| yakubin wrote:
| But when you move the laser emitter in your hand you're
| controlling the speed in that 2d space, not in 3d. You
| don't ever affect the position of photons in the Z
| dimension. So you're not constrained by speed in 3d which
| would later be slowed down after being projected. So you
| move your laser emitter along the diagonal of a face with
| velocity v. And the _perceived_ light which would get
| projected onto a plane needs to match the position of the
| emitter on the face. Which creates the illusion that
| light travelled along the 3d longer diagonal faster than
| at v (in order to match the projection which describes
| how you /camera sensor see the light). But in reality the
| light never travelled along this longer diagonal. It's
| only an illusion. And it is this illusion that we're
| measuring the speed of. Photons on this diagonal arrived
| straight from the emitter, i.e. each of them appeared in
| only one point of the diagonal throughout its entire
| history. In other words, the photon at the beginning of
| the perceived movement is a different photon than at the
| end. They travelled along different paths. And when some
| photons were at the diagonal, some others were on their
| way there.
| hotstickyballs wrote:
| Shine a laser into space and the image of your laser can
| be much faster than the speed of light. Nothing actually
| moved faster than the speed of light though.
| neogodless wrote:
| What do you mean by "image faster than light"?
|
| How is an image not light?
|
| Or do you mean a captured image may show items from
| different points in time?
|
| But that's only relevant after the photo has been
| created, not during the window of time that a sensor is
| capturing light.
| Zak wrote:
| Stand a meter away from a wall and wave a laser pointer
| such that the spot travels back and forth between two
| points a meter apart in one second. Move two meters away,
| but keep your movement exactly the same; the spot now
| moves two meters in one second.
|
| Move two light-seconds away and do the same movement. The
| spot now moves two light-seconds in one second: twice the
| speed of light. Of course it takes two seconds from when
| you turn the laser on to when an observer at the wall
| would see it, and four seconds before you see the spot on
| the wall, but the spot itself moves faster than light.
| neogodless wrote:
| Ah, so for the sake of capturing conceptual / perceived
| "objects", the global shutter, at least, can do a better
| job at what would be perceived during a short period of
| time that the shutter opens and captures each pixel.
|
| A rolling shutter might capture points along the way but
| leave gaps in comparison. In the laser pointer example,
| you'd probably want a longer exposure, but the global
| shutter would still give you uniform capture better
| matching what your eyes / brain perceived.
| jdietrich wrote:
| _> You'll certainly NEVER see rolling shutter in any
| PRACTICAL application for video unless you pan like a true
| madman_
|
| Fast lateral movement doesn't just happen on a whip pan. Any
| scene involving fast vehicles, high speed machinery or strong
| vibration is likely to have some level of rolling shutter
| artifacts. Many perfectly normal tracking shots will have
| some artifacts, particularly if you're shooting handheld.
| Modern fast-readout sensors are very good, but they aren't
| perfect; to a great extent, we've just got used to seeing
| mild rolling shutter artifacts and don't immediately
| recognise them as an image quality defect.
|
| There are still lots of good reasons to use rolling readout
| sensors, but there are also many real advantages to global
| sensors.
| veidr wrote:
| > lots of good reasons to use rolling readout sensors
|
| Like what? Just curious, but the only reason I know of is
| that they're cheaper, or you already have them.
| sudosysgen wrote:
| They also tend to have better dynamic range and noise
| performance.
| ptsneves wrote:
| Global shutter is essential for serious drone photography or
| photography from platforms with undamped vibrations like
| machinery. It means you will have discontinuities in features
| of the picture. You will need to re-stich the photo or post
| process it. For video it is just bad. An example of such pixel
| perfect uses was we were able to know the size of a marking on
| the ground with centimetre accuracy by counting the pixels.
|
| Modern gimbals are amazing and do away with much of the issue
| but if you are needing pixel accurate features global shutter
| is the way to go.
| ptsneves wrote:
| An example of global shutter issues
| https://youtu.be/qqsDMJ7iyM0?t=86
| account42 wrote:
| You mean an example of _rolling_ shutter issues.
| ptsneves wrote:
| Yes thanks for the correct. I cannot edit the typo
| anymore :(
| smogcutter wrote:
| > 1/30th of a second is actually a relatively long exposure for
| a photo.
|
| As an aside, in video & motion picture cameras the frame rate
| isn't the exposure length. You need to account for the time the
| shutter is closed. So the real exposure is about 1/60, still
| slow but much more manageable.
| vanderZwan wrote:
| > _Some cameras with slower sensors suffer from the rolling
| shutter effect where fast moving subjects get distorted because
| the subject or the camera is moving. Think e.g. lampposts
| looking diagonal when shot from a fast moving car._
|
| These days this is the least of your worries with a camera with
| rolling shutter. I got a Nikon Z-fc, knowing from the reviews
| that it would have a rolling shutter, because I don't mind
| these types of distortions.
|
| What the reviews didn't mention is that many LED based lights
| actually alternate red, green and blue LEDs in quick
| succession. And you see that in photos as horizontal streaks of
| rapid changes in white balance. This also happens with
| projectors that are LED based, and CFTs if their drivers aren't
| fast enough.
| plus wrote:
| Are you sure that's an artifact of the light source emitting
| colors sequentially, and not an artifact of the camera sensor
| detecting colors sequentially?
| vanderZwan wrote:
| It only happens with _some_ light sources. If it was the
| camera it would happen with all of them.
| m463 wrote:
| I hope global shutter makes its way to other brands so we can get
| away from the banding and jelly cam that we put up with when
| using digital cameras.
| vouaobrasil wrote:
| That stuff is already mostly gone with fast stacked shutters as
| in the Nikon Z9. In fact, even cheaper cameras like the
| Panasonic G6 don't have noticeable banding.
| vr46 wrote:
| Very cool technology, for those who don't want, need or can't
| make creative use of the visual artefacts that come with other
| cameras.
| threeseed wrote:
| Nokia has had something similar for a while now where the sensor
| readout was fast enough to prevent the roller shutter effect.
| Either way this was never a huge issue for photographers since
| you could always fall back to using mechanical shutters.
|
| The more interesting part of this camera is whether this is a
| trend towards photography being the process of extracting a frame
| from a video rather than taking a series of shots.
| blt wrote:
| How does this compare to the capabilities of DSLR with a
| mechanical shutter?
| threeseed wrote:
| So the global shutter will have no benefit today for most
| users.
|
| In the future though it means we can do away with the
| mechanical shutter which in theory means smaller, cheaper, more
| reliable cameras.
|
| However the capturing photos when you hold the focus button
| down feature will I think be genuinely useful for pretty much
| everyone. Especially in family situations with kids, pets etc.
| whyenot wrote:
| > So the global shutter will have no benefit today for most
| users.
|
| It will have a benefit for anyone who is also recording video
| using their DSLR, likely the majority of users. It completely
| eliminates rolling shutter. Now, is it a benefit that is
| worth paying $6,000 for? That is a good question! For most
| users, probably not.
| Almondsetat wrote:
| > Now, is it a benefit that is worth paying $6,000 for?
|
| This tech will certainly reach all levels of the consumer
| market. Plus being able to shoot with flash at any speed
| will indeed benefit tons of amateurs who will be able to do
| away with the stupid artificial limitations present for the
| last 150 years
| threeseed wrote:
| > who is also recording video using their DSLR, likely the
| majority of users
|
| I doubt that the majority of users are using their DSLR to
| film video.
| jiggawatts wrote:
| > In the future though it means we can do away with the
| mechanical shutter
|
| The future is a year ago: The Nikon Z9 and Z8 cameras have no
| mechanical shutter, and their electronic shutter speed is so
| fast that there is no practical issues capturing even fast
| motion.
| SulphurCrested wrote:
| More capable mechanical focal plane shutters with a flash sync
| speed of 1/200s or faster do not show objectionable rolling
| shutter effects for most practical stills photography,
| including wildlife and most sports. Better mirrorless cameras
| use the same types of shutters.
|
| Older film cameras may have had flash sync speeds of 1/60s, and
| rolling shutter was sometimes a problem with them.
|
| Mirrorless cameras also have a fully electronic shutter, for
| silent shooting and shooting video. (Some DSLRs could also
| shoot video this way.) In all but a handful of very expensive
| cameras, it takes a long time to scan the sensor and rolling
| shutter is obtrusive - in many cameras it is as slow as 1/15s.
| In that handful of expensive cameras it is around 1/200s, like
| a fast mechanical shutter. Nikon's expensive cameras don't even
| have a mechanical shutter any more.
|
| That last bit of speed, from 1/200s to the 0s of a global
| shutter, is detectable in certain examples Sony showed off in
| their presentation. Most of these examples relate to high-end
| sports photography.
|
| Flash sync at any speed means flashes can be built more cheaply
| and cycle faster. Currently, to shoot faster than the flash
| sync speed the flash must emit a long burst instead of a single
| pulse. For this reason, some cameras used a leaf shutter, which
| can sync at any speed, but they were typically expensive to
| make and sometimes less reliable.
|
| There are related problems to do with the refresh rate of LEDs
| in the field of view, which are solved by the global shutter
| and somewhat clumsily worked around in cameras without one by
| trying to sync the shutter with the refresh rate. Again, high-
| end sports photographers are the ones who usually encounter
| this problem.
| xyzzy_plugh wrote:
| > will also add C2PA authentication metadata. C2PA is a
| combination of the efforts of the Content Authentication
| Initiative and the separate Project Origin initiative.
|
| First I have heard of this. My initial thought is "gross" but I
| wonder if it may actually be beneficial for the average consumer,
| prosumer even.
|
| The spec is absolutely vomit inducing, though -- more ASN.1!
| Arcanum-XIII wrote:
| Working for a broadcast company with a good chunk of news
| content, authenticating content is one of the major we face. It
| has always existed, but it's getting worse and worse. That's
| the goal of this system, not for consumer.
| Regnore wrote:
| I think ideally we'd get to a point where end users could
| easily see who signed an image so if someone was claiming an
| image was from cnn it could be validated. I imagine the end
| goal would be to put a warning on images that aren't from a
| signed source or maybe not displayed at all by default -
| similar to the uptake of https
| akoboldfrying wrote:
| Agree, but I also think that signing at all should be
| optional. That is: Give the camera user the option of
| strongly identifying themselves (with the added trust that
| brings), or remaining anonymous (as is the case now).
| sudosysgen wrote:
| I mean even if the setting isn't there it's easy to strip
| the metadata.
| blopker wrote:
| This is the first time I've heard of C2PA authentication[0]. It
| looks like they are digitally signing images from the cameras now
| to create a chain of authenticity. There's even TruePic[1] that
| is an authentication service that will show people if an image
| came from one of these cameras, overlaid on the image (or video)
| with some Javascript. Interesting way to fight AI images and deep
| fakes. They are also letting people register their deep fakes[2],
| but I'm not sure why someone would do that.
|
| 0: https://c2pa.org/ 1: https://truepic.com/ 2:
| https://truepic.com/revel/
| Almondsetat wrote:
| The new Leica camera already does it, give it a look for a
| working example already in the wild
| choppaface wrote:
| C2PA is more about tracking origin and edit lineage for e.g.
| press photographers. It does not determine if an image is a
| deepfake or not. Create a deepfake then take a picture of it
| with a C2PA camera, now you've got a C2PA-stamped deepfake.
|
| But supposed you take a C2PA photo and then edit its contents
| with e.g. generative in-fill. The updated C2PA stamp from that
| action will get noted and can be audited later on.
| chpatrick wrote:
| > Create a deepfake then take a picture of it with a C2PA
| camera, now you've got a C2PA-stamped deepfake.
|
| Okay but then it's a deepfake from the BBC or some other
| known source, which would be bad for its reputation if people
| found out.
| abdullahkhalids wrote:
| Both newspapers and social media routinely reconvert the
| original camera image into some smaller size or more suitable
| format. How is this system going to work in practice?
| Regnore wrote:
| The newspapers and/or social media would sign their scaled
| pictures - presumably they could also provide the original
| for external validation as well
| beenBoutIT wrote:
| Could it be used to track down journalists or prove that a
| certain journalist took a specific photograph?
|
| Sounds like the photographic equivalent of guns that put
| unique stamps on their fired shells - theoretically making it
| easier to ID the shooter.
| EMIRELADERO wrote:
| Considering this might be targeted by nation-state adversaries
| for propaganda uses, I don't see the effectiveness, at least on
| that threat model. AFAIK even if the private keys are buried in
| some trusted computing chip, there's very little a focused ion
| beam can't do. And that's in an extreme security scenario, I
| doubt those cameras even reach that.
| huytersd wrote:
| What do you mean? An ion beam can read the private keys from
| the onboard chip?
| EMIRELADERO wrote:
| It's essentially a million-dollar silicon modding tool. It
| can, in 99% of cases, read down to the individual
| transistor level.
| huytersd wrote:
| Then why can't the three letter agencies in the US unlock
| an iPhone? I don't think it's that easy. In that case all
| encryption would be useless if you had physical access to
| the machine.
| zimpenfish wrote:
| > Then why can't the three letter agencies in the US
| unlock an iPhone?
|
| They almost certainly can but I believe the point of the
| FBI making a big song and dance about unlocking that
| phone (which they did unlock by themselves, btw) was
| about trying to force Apple into allowing TLA backdoors
| via the court of public opinion.
| jsmith99 wrote:
| It only works for things where the key is stored on the
| chip or phones where the key is stored in a TPM or
| equivalent and relies on a PIN to release it, as opposed
| to typical use of encryption where the full key is
| entered to unlock. An attacker with physical access can
| probe the chip but it's risky - one slip and it's gone
| for ever. This technique is most useful when all chips of
| this type have the same key so you only need to
| successfully crack one, any one. Ross Anderson's book
| Security Engineering uses the example of Sky Pay TV cards
| https://www.cl.cam.ac.uk/~rja14/Papers/SEv2-c16.pdf
| huytersd wrote:
| With all the encryption I've used these past decades the
| only time I have entered a full encryption key was for
| cryptocurrency. Aren't 99% of use cases the case where
| the key is stored on the client device? It seems like the
| main deterrent then just comes down to the failure rate
| of the probes.
|
| On a tangent, a bad actor could then release HDCP keys
| for TVs from a big brand like Samsung and effectively
| invalidate all content protection for all the TVs they
| have already sold (afaik those can't be remotely
| updated). If those keys are then revoked, there would be
| millions of bricked Samsung TVs.
| brokensegue wrote:
| Pretty sure they use a kdf to generate the key from your
| input so you don't need to enter the " full key"
| TaylorAlexander wrote:
| > Then why can't the three letter agencies in the US
| unlock an iPhone?
|
| Who says they can't?
| jsmith99 wrote:
| Assuming they use different embedded keys per camera (which
| is unfortunately unlikely) instead of the same key on every
| chip, this can be defeated by demanding to see the intact
| camera before accepting the signature. Ion beam assisted
| probing is a destructive process.
| imhoguy wrote:
| I get a flashback of the yellow printed dots. Smells a privacy
| threat to me. I wonder if that fingerprinting can be disabled.
| nonrandomstring wrote:
| Useful for source authentication in news to combat fakes. I am
| sure if you don't _want_ your identity associated with a
| picture there will be plenty methods to avoid this. Indeed some
| journalists will also need anonymity - but perhaps signed
| metadata on the GPS location /timestamp, so any camera for
| serious photo-journalist reporting will have to have those
| features.
|
| Perhaps less interestingly, single frame exposure will defeat
| the accidental recording of sound into pictures recently
| reported to be a side effect of rolling shutters.
| varispeed wrote:
| I think we are going to see more of this. I doubt this is
| about "deep fakes", as in reality the press doesn't care
| about it as long as it generates clicks and revenue.
|
| It's more to do with surveillance and being able to more
| easily track someone trying to play investigative journalist,
| whistleblower etc. and show them their place.
|
| For instance, if someone takes a photo of a big corporation
| exec giving a brown envelope to a politician and then another
| journalist show them to interested parties as a courtesy
| before publication, they could find out who took the picture
| and get them killed.
|
| This is very dangerous and stupid what Sony is playing with.
| zamubafoo wrote:
| Seemed fairly dystopian before you get to the examples. It also
| seems fairly trivial to bypass even if it's through crude brute
| force methods (ie. screenshots/copying framebuffers).
|
| Is this essentially just enforcement through building a moat
| around editing applications?
| alphanullmeric wrote:
| When's it getting put in a phone?
| kart23 wrote:
| it's never made sense to me that there's still not a single
| compact full frame camera. there are plenty of 35mm film cameras
| that you can literally slip in a pocket and still have pretty
| good lenses, what's stopping a compact full frame camera these
| days?
|
| https://casualphotophile.com/2019/05/17/ricoh-gr1v-film-came...
| Darthy wrote:
| I've got the smallest digital full frame (the Sony RX1).
| Compared with your linked analog Ricoh GR1v, the Sony is bigger
| because of its much more capable lens. Putting the Ricoh lens
| on the Sony would have limited the quality a lot. It's a
| question of "do the elements fit together?".
| coldtea wrote:
| There are a few Leicas (Q3), a Sony (RX) and some others.
|
| On APS-C (Fuji) or compact-design-but-with-changable-lens there
| are a lot more options.
| dboreham wrote:
| One issue is that you need a bunch of space behind the focal
| plane in a digital camera (the thickness of the sensor + its
| mounting system + PCB + cooling gap) while film cameras can
| mount the film (~zero thickness) right against the back door of
| the body. Another is that people expect EVF now, which requires
| more envelope volume than a compact 35mm OVF. Generally people
| paying $$ for a camera expect it to have removable lenses which
| adds again to the physical envelope.
|
| There have been some Sony camera with fixed 35mm focal length
| lens and FF sensor. Like this:
| https://www.bhphotovideo.com/c/product/1316682-REG/sony_cybe...
| selimco wrote:
| There's the sigma fp which is the smallest I believe.
| vouaobrasil wrote:
| Sony a7C. Very compact...
| geerlingguy wrote:
| It's still a bit of a thick back; compared to the thinner
| rangefinder film bodies it's a little thicker still.
|
| What I don't like most about all my Sony cameras is the hand
| grip. It feels made for a robot, and I miss the nice organic
| shape of Canon and Nikon grips.
| jsmith99 wrote:
| There are pocketable FF camera with prime lenses and pocketable
| 1" sensor cameras with zooms. Those old 35mm film pocketable
| zooms had terrible apertures but a RX100M3 with a 1.8-2.8
| aperture has a comparable effective aperture to some FF
| mirrorless with kit lens.
| _ph_ wrote:
| Compared to what modern 35mm cameras can deliver, these cameras
| had very bad resolution. Yes, you could build such a camera
| today with a 35mm sensor, but the results wouldn't really
| justify the effort. If you want good compact cameras, get one
| of the smaller mFT cameras like the E-M10 or OM-5 and mount a
| pancake lens on it.
| kart23 wrote:
| "bad resolution"??? 35mm film can deliver 87 MP, more than
| any ff digital camera i know.
|
| https://www.kenrockwell.com/tech/film-
| resolution.htm#:~:text....
| sudosysgen wrote:
| No, it can't. That's at a very low contrast rate, meaning
| that any detail at the 87MP level is going to be extremely
| muted and barely noticeable even if it was already high
| contrast. So basically you'd only ever get that resolution
| if you were taking picture of a black/white striped test
| chart, nothing else.
|
| A better benchmark is MTF50, at which you can get about 50
| line pairs at ISO 100-50 which corresponds to about 10
| megapixels. If we want to be nice we might go as low as
| MTF20 where you'll get about 18 megapixels of low contrast.
| I'm looking at Velvia 50, the same film as Ken (https://www
| .ishootfujifilm.com/uploads/VELVIA%2050%20Data%20...)
|
| Ken also makes the point that the Bayer sensor reduces the
| resolution - and it does - but he forgets that film has
| different MTFs for different wavelengths, sometimes even as
| bad as half for red compared to green.
| justinator wrote:
| A camera like this will make it a lot easier (and probably
| cheaper!) to verify who wins competitions, such as road bike
| races where a 200km race can be won by millimeters.
|
| And oh hey! dpreview.com lives!
| justinator wrote:
| Here's an example of how the rolling shutter effects bike race
| photo finishes,
|
| https://inrng.com/2012/04/photo-finish-camera/
| justinator wrote:
| And here's what's really going on with photos like this,
|
| https://en.wikipedia.org/wiki/Photo_finish#Strip_photography
| Toutouxc wrote:
| I don't think so. You've posted a link to Wikipedia in an
| adjacent thread that clearly says that photo finish cameras do
| NOT use traditional 2D sensors, but a "1D" single-strip sensor
| with very high refresh rate. They intentionally use an effect
| similar to rolling shutter where instead of the currently
| exposed row moving up and down the sensor, there is just one
| row and the photographed subjects move against the camera.
|
| Yes, you could capture a similar photo by aiming a global-
| shutter camera at the finish line and firing the shutter
| exactly when the first contender triggers a laser beam or
| something, but that would be much more complicated and
| expensive than just capturing a single slice of space and
| letting time flow along the perpendicular axis.
| justinator wrote:
| _> would be much more complicated and expensive than just
| capturing a single slice of space and letting time flow along
| the perpendicular axis_
|
| I'm not quite sure. You're replacing a custom rig and
| technicians, with an off the shelf part. Sounds like it could
| be easier and cheaper (my point).
| alpaca128 wrote:
| _> Sony says doesn 't compromise on ISO performance or dynamic
| range. [..] Sony isn't quoting a DR figure for the camera._
|
| Who are they fooling?
| Espressosaurus wrote:
| No one that bothers to look at the minimum ISO spec of 250.
| jacquesm wrote:
| I remember a whole thread about dpreview no longer posting new
| articles and yet this one seems to be very recent.
|
| https://news.ycombinator.com/item?id=35520495
|
| Sony is rapidly becoming a major factor in the professional world
| as well. A family member works for a pro camera shop and he's
| absolutely lyrical about the Sony line. My biggest gripe is that
| it is hard to interface to the camera for things like shutter
| control, what looks like a USB cable really isn't (there are a
| ton of other wires running there that you can just barely see
| that control these kind of functions). But otherwise it's a very
| nice piece of gear.
| cromulent wrote:
| Gear Patrol bought them, so DPReview lives on.
|
| https://www.dpreview.com/site-news/8298318614/dpreview-com-l...
| jacquesm wrote:
| Ah cool, I totally missed that. Thank you!
| cesaref wrote:
| Looking at this history of this problem, it first appeared with
| the adoption of focal plane shutters, which appeared widely with
| Leica 35mm cameras, and is a technology used by most SLRs. I know
| of earlier examples, the Speed Graphic large format camera uses a
| focal plane shutter, not sure when this was first used.
|
| In the medium format world, lots of camera systems used leaf
| shutters, and these do not suffer from rolling shutter problems.
| The exposure is controlled by an aperture within the lens itself
| which opens the pupil to expose the film/sensor. The problem with
| leaf shutters is the minimum exposure tends to be limited to
| 1/500th of a second, although I believe there are some 1/1000th
| of a second leaf shutters in some lenses. Even some medium format
| SLRs have leaf shutters, for example, the Hasselblad 500 series
| cameras, and the Mamiya RZ67. Leaf shutters allow flash sync at
| all shutter speeds, which is why it was a popular technology for
| studio work.
|
| In 35mm land there are some leaf shutter cameras, they tend to be
| fixed focal length designs. The olympus 35RD has one for example.
|
| Anyhow, interesting tech, it's not relevant for the sort of
| photography I do, but i'm pretty sure the videographers will
| appreciate it!
| wkjagt wrote:
| The Leica Q has a leaf shutter that goes up to (down to?)
| 1/2000s.
| jeswin wrote:
| A handgun bullet travels 1200 ft per second. At 1/80000 of a
| second, the bullet travels 0.18 inches.
|
| It won't be very good, but you can actually shoot a bullet with
| this. With enough light of course.
| asimpletune wrote:
| This might be a strange thought, but the idea of the global
| shutter got me thinking about the promise of a global shutter and
| the strange implications it could lead to regarding special
| relativity. I'm not an expert but, as I understand, one aspect of
| special relativity is that there is no shared now. Like, the idea
| of simultaneity is different than how we experience it, because
| two observers can literally witness a different sequence of
| events and there are no contradictions. (I hope I got that right)
|
| In a traditional shutter, each row of pixels are read one at a
| time, so there is an inherit arrow of time. However, with a
| global shutter, each row of pixels is "exposed" simultaneously,
| theoretically. So what I wonder is what would be the implications
| of this with respect to special relativity.
|
| For example, if a global shutter camera were used to adjudicate a
| close race, would it be possible for one row of sensors to
| witness a contradictory result with another row? Sorry if I'm not
| explaining this right but I'd be curious to hear what a more
| knowledgable person's thoughts on the matter are.
| Espressosaurus wrote:
| There are no implications regarding special relativity. The
| propagation speeds of the triggering signal are well
| understood. If for some insane reason it's not considered fast
| enough to simply ignore delays from one side of the chip to the
| other (wherever the sync signal comes in), it's straightforward
| to put in delay lines to ensure everything triggers at the same
| time to an accuracy that depends on the amount of silicon you
| want to lay down for those delay lines relative to the
| capacitors.
|
| Most likely a couple nanoseconds one way or another don't
| matter (even at timescales of 1/80,000 s) and those ever-so-
| slight differences are ignored.
| asimpletune wrote:
| I meant in a theoretical sense, like as a thought experiment.
| If the shutter were an idealized, global shutter, where each
| row of pixels are exposed instantaneously, what would be the
| special relativity implications, if any?
| Espressosaurus wrote:
| Unless the sensor is half over the event horizon of a black
| hole I'm failing to understand how my answer doesn't short-
| circuit the thought experiment to ground.
| wayfinder wrote:
| If two global shutter cameras take a scene at the same time,
| they would have the same image. If they don't, then they
| wouldn't.
|
| Global shutter cameras are also not new. CCD sensors are global
| shutter and we used to use CCD sensors for cameras decades ago
| until we switched over to CMOS for most applications due to
| their larger size and better dynamic range. I have a global
| shutter HVX200 from 2005 in my closet.
| asimpletune wrote:
| So the whole point of what I was saying is that special
| relativity proves that whether two spatially separated events
| occur at the same time is not absolute and it depends on the
| observer's reference frame. My question has nothing to do
| with engineering and wasn't intended practically.
|
| The point was to imagine light from some event, reaching two
| parts of the sensor at the same time, being exposed at the
| same time (ie the global shutter), and yet asking if those
| two parts of the sensor could show conflicting versions of
| that same event, since they are nonetheless separated by
| space.
| nayuki wrote:
| There is no problem. The sensor is a rigid assembly where all
| parts are motionless with respect to each other. Simultaneity
| is a well-defined concept for the sensor.
| DrNosferatu wrote:
| Curious to see how it films a spinning airplane propeller.
|
| Also, can the global shutter be turned off when you don't need
| it, so you can get higher dynamic range?
| amelius wrote:
| How is this accomplished? Is there a memory cell close to every
| physical pixel? Did they sandwich two dies together?
| pestatije wrote:
| yes, its a 2 layer stacked-Cmos sensor
| driggs wrote:
| Global shutter might be the most impressive feature, but also
| notable is that the a9 includes C2PA authentication metadata.
| This is a cryptographic chain of provenance on digital media,
| starting at the hardware camera itself.
|
| In addition to provenance itself, C2PA allows for asserting that
| an image should not be used for AI training / ML mining.
|
| https://c2pa.org/specifications/specifications/1.3/explainer...
| nayuki wrote:
| What stops someone from photographing a printed picture of a
| deepfake?
| ngcc_hk wrote:
| For many it is irelevant. No doubt sony and video professional in
| sony cam would enjoy another upgrade opportunity. But for normal
| people, it is not the camera but also the lens etc. one has to
| think about. Not to mention still have F lens to migrate. (Or
| canon ... and Manu of these can be used by my Z9.)
| lvl102 wrote:
| It's about time Sony threw in a PCIe5/6 slot on these cameras.
| Ability to simply insert an internal 2-4TB drive would be very
| welcome. These CFE cards are ridiculously expensive for what they
| do.
| cfn wrote:
| The latest Hasselblad already ships with an internal 1Tb SSD so
| there's hope.
| e12e wrote:
| > [Sony] also says it will be introducing firmware updates to the
| a7S III and the a1, adding features such as breathing
| compensation. The a7S III will also gain DCI 4K capture. (...)
|
| > All three cameras will also add C2PA authentication metadata.
| C2PA is a combination of the efforts of the Content
| Authentication Initiative and the separate Project Origin
| initiative.
|
| Wait, they're adding image attestation in software to existing
| models?
| Avamander wrote:
| Funnily enough I can imagine this awful DRM being one of the
| few things that would help us avoid deepfakes and similar in
| the future.
| M2Ys4U wrote:
| It's not DRM, it's optional provenance metadata.
|
| So, think of it as cryptographically _signing_ the image
| rather than _encrypting_ the image.
| jauntywundrkind wrote:
| Everyone wishes there could be new photography or videography
| features and Sony could with effort to reengineer some of these
| processing pipelines to do more (ex: Focus Breathing). but
| adding some workflow capabilities is off the main & intensive
| path of the device & easier to tack on.
| e12e wrote:
| As for price:
|
| > The a9 III will be available from Spring 2024 at a recommended
| price of $5999.
|
| > The 300mm F2.8 GM OSS has two XD linear motors to allow it to
| focus fast enough to work with the a9 III.
|
| > The 300mm F2.8 will also cost $5999 and will also be available
| in Spring 2024.
| tlhunter wrote:
| The two products announced are mostly catered for professional
| photographers at the Olympics. The camera body and the lens are
| both very expensive, over $10,000 for the two combined.
|
| That said, the global shutter is interesting, and will hopefully
| find its way into consumer grade products in coming years.
| Personally I'm excited about this "obsoleting" existing used
| cameras so that I can finally afford them.
| touisteur wrote:
| Oh does this open software lenses and optical beamforming? The
| progress of acquisition tech is just staggering. ADCs in the last
| 10 years have just gone to incredible sampling rates,
| compacity... The future looks fun for the linear algebra
| signal/image processing people to go back to the drawing board:
| now that you have teraflops and these sensors available, let's
| start again.
___________________________________________________________________
(page generated 2023-11-09 23:02 UTC)