The new Sony A7sII can now be preordered at Sony A7sII at Amazon, BHphoto, FocusCamera, Adorama and in Europe at WexUK. In Asia soon at Digitalrev.
–
Bruno from Lesnumeriques posted those slides from the official Sony A7sII presentation. There are a couple of interesting insights:
- The Large sensor video camera market is increasing (P.S.: larger than 2/3 inch sensor)
- In Europe most A7s buyers were professionals
- A7sII has 2 stops ISO advantage over the new A7rII
Still Planet5D has some doubts about the new A7sII: “The New Sony A7s II is Great, But Is It Awesome Enough To Get You To Upgrade?”.
45 Responses
Curious to see how AF performs, that could make the RX1 sucessor faster without the need of OSPDAF that could be more expensive tech.
I do not see the 2 stops improvement for stills compared to the A7RII. They might be there on pixel level but not resolution adjusted.
My opinion, too.
Considering that the top ISO for the A7r II is 102.400 and for the A7s II is 409.600, that means two stops advantage, right?
That simply means, you can set the A7S II two ISO steps higher. It does not mean that you really get an advantage.
Maybe I’m missing something, but there’s nothing on the slides that says “advantage”. It only states that the 7s II offers two stops over the 7r II.
When the slide talks about sensitivity, it implies this advantage.
Where do I find the switch to select 4.5 stops additional image stabilisation?
I’m not saying you are wrong or anything.
But I guess you also are clever enough to know what they mean stating those extra two stops; marketing!
(even though they are not lying, since the A7s II DOES offer two extra stops! Whether that translates into an actual advantage or not, it’s for the users to find out 🙂 )
No, the per pixel noise of the A7rII is 2 stops higher than the A7s at ISO 102400, that’s what they are referring to.
If they put blue sticker on mediocre lenses, they can lie about ISO too xD
That’s like throwing a Ford badge on an Mercedes. Wait- they both suck. I take that statement back.
Yup, resolution adjusted the it’s the same! But that’s not good marketing, they have to make out that the A7sII is king of sensitivity (when it isn’t) to sell them!
King of low light AF sensitivity 😉
They picked for wording for non-gearheads.
Kind of shocked the number of professionals using the camera isn’t alot higher than the non-professionals… Given that the A7s required a pricey add-on recorder to get the most of the video capabilities and the A7s high price… Didn’t expect many non-professionals would want it.
What do you mean? The main benefit of the A7S was super low light performance, SLOG2, XAVC-S, and beautiful full sensor readout 1080p video. Only the 4K required an addon.
Ahh the land of marketing, where per-pixel noise is apparently a relevant measure…
Well, apparently they’re not saying anything about it at all. Take a look at that picture once again. They are talking about some vague measure of “steps”, not “stops”
Even if you think about stops, the numbers are wrong A7sII is 2 stops above A7rII, but 4.5 stops above original A7s??? How is that even possible?
So first, lets figure out what they meant by “steps”, cause I for one, have no slightest idea.
I can make a guess, where A7sII is 2 stops ahead of A7rII, because both have IBIS and A7sII has 2 stops higher sensitivity than A7rII, while it is only 0.5 stops better than original A7s, but due to “4 stops stabilization” somehow they added the sensitivity and stabilization together and made it out to be 4.5 stops? Maybe that’s why they’re talking about steps, and not stops? LOL
The stupidity of marketing is surprising at times….
It’s quite clear what they’re saying.
It’s not unheard of to use the word “steps” to refer to stops. And if you look at the DXO SNR data then the the A7s is indeed 2 stops ahead of the A7RII in per pixel noise at ISO 102400. That doesn’t change the fact that per pixel noise is the wrong measure of course.
You’ve misread the slide – the A7sII is 4.5 stops ahead of the A7s in terms of image stabilisation, not sensitivity. This is because the A7sII has 4.5 stops of IS and the A7s didn’t have IS (thus zero stops)!
What measure do you want then?
Perceived noise at normalized resolution? That’s non-sense as we don’t downsample our 42mp images to 12 in practice…
If you normalize one thing than you need to do this with everything. If the A7R II isn’t 2 stops worse than it’s neither a 42mp camera…..
I’ve never heard stops referred to as steps. Is there a source?
sure:
“Each increment of 1 in exposure value corresponds to a change of one “step” (or, more commonly, one “stop”) in exposure”
from https://en.wikipedia.org/wiki/Exposure_value
The stupidity of people is surprising at times…
Idiocy of responders is surprising at times…
It’s okay to be flustered; just don’t make a fool of yourself next time.
Thankfully Matt Grum saved me time typing.
Stabilization has nothing to do with low light sensitivity, unless you are shooting handheld…
Steps might be due to translation. Seen this plenty before.
Matters when you don’t full pixel read-out in video from a7RII
A Way To Get Paid $97/hour And More…….After earning an average of 19952 Dollars monthly,I’m finally getting 97 Dollars an hour,just working 4-5 hours daily online.….. Weekly paycheck… Bonus opportunities…earn upto $16k to $19k /a month……….Only a few hour required to understand and start working…..look over here …
.uuj……
➤➤➤ http://GoogleTopCareersJoinLatestOnlineVacancies/get/morethan$97/h…
✥✥✥✥✥✥✥✥✥✥✥✥✥✥✥✥✥✥✥✥✥✥✥✥✥✥✥✥✥✥✥✥✥✥✥✥✥✥✥✥✥✥✥✥✥✥✥✥.
What RX1 successor….
I think if there was gonna be one, it would have happened a long time ago!
Maybe they didn’t have the right tech for it.
The worlds smallest
The worlds greatest
The worlds first
Stop it already- that isn’t special anymore. How about, the cosmological universes first ever…now you have my attention. If martians haven’t shot it then you have my money.
It would be nice if clicking on the slides would enlarge the slides and NOT re-direct the visitor to the amazon account.
0.78x, world highest… Why don’t they make a zoom eyepiece 0.6x-1x. 1x would be something to be proud of xD
Doesn’t need to be worth the extra $$$$ to upgrade from a7s… there are first time buyers too. Also how often are the improvements in the higher model worth the price difference? This is not even a valid question.
With A7RM2 you have choice to make 42mp pictures when you want to and downsample to 12mp (or any mp count) when you want to. A7SM2 does not have such choice. Also, downsampled 12 mp from A7RM2 is still sharper than 12 mp from A7SM2.
I thought i was reading one of my own posts 😀
SNR at normalised resolution would be an improvement!
“we don’t downsample our 42mp images to 12 in practice”
No, most people downsample their images to around 2mp in practice, to post online. Very few print. But if you do print you’ll find the fine noise from the A7RII to be more agreeable than the coarser noise from the A7s, even if the fine noise is two stops greater in magnitude, hence claiming the A7s has a noise advantage is just not true, whether you normalise or not.
With a lower QE the sensor needs to be amplified more for a certain light level. This isn’t just introducing more noise but also lowers the DR, color depth and does bad things to color fidelity.
And at certain thresholds different kind of noise shows up (ex. bias pattern noise or chroma noise). All this can’t be improved with downsampling.
In reality though one camera might just be brighter instead of having less amplification as amp levels are quantified. ISO is sadly not as equalized as it should be…
After all they call it “sensitivity” and not SNR. QE won’t change through downsampling so they dont lie 😉
The downsampling story works in simplified theory but not in practice as
1. things like DR and color depth don’t improve
2. noise increases exponentially or in thresholds (not linear)
3. Not the entire sensor area is used (even with gapeless design)
Not a good thing at all if the market is primarily professionals. This means that this is a specialty niche product, and sales volume will be limited.
To be fair, for this particular model, that may be their target.
For most manufacturers, hobbyists are their real target. That’s where the sales volume is.
so nobody should make product for professional anymore? you see the blue pie chart part that’s labeled non professional? yeah they are hobbyists
Um. Hence my comment.
Looks like Sony has a pretty strange idea of what a large format sensor is…this isn’t even medium format…
You’re wrong on many points here. DR is not affected by quantum efficiency. If you put an ND filter on a sensor to simulate a lower QE then DR wouldn’t change, the difference between the brightest and darkest details that could be captured in once exposure would not change.
Downsampling and measuring SNR isn’t the perfect measure as it doesn’t capture how visually disruptive the noise is, but it gives a more accurate picture (for stills anyway) than quoting per pixel SNR.
And if they’re referring to sensitivity instead of SNR, then the same point holds, the sensitivity of individual pixels is a misleading measurement, it’s sensitivity integrated across the whole sensor surface that matters.
Finally both DR and colour depth DO improve when downsampling.
Yes, that’s about the only case where it’s relevant.