NAB 2014 Sony A7S ‘NORMAL PAN SPEED” Rolling Shutter Test
Here are a couple of non scientific tests made with the A7s. The video on top has been taken by Andrew Wonder and shows you the A7s Jello effect. As all video makers know this is an effect that every CMOS camera has and the A/s will make no exception. It’s also obvious that no one will ever shoot a video with the A7s shaking the camera like that 🙂
The A7s crazy ISO capability is pushing a couple of sites/people to post some unscientific ISO test. Photographybay posted the ISO comparison image you see here on top but admits for himself:
I missed focus on a couple of the shots and this is a terrible lighting scenario, but the A7S shows a lot of promise for low light and high ISO work. This is also probably a preproduction camera, so the results should not be taken as what the final production model is capable of delivering.
OneLoneDork posted a short completely ISO impression test on Youtube.
A7s Preorder or pure notification store links:
USA: Sony US, Amazon and BHphoto.
Europe: Sony DE, UK, ES, FR, IT, NL, SE, FI, PL.
Asia: Digitalrev
0 Responses
“Every”? No CMOS has a global shutter?
Some CMOS sensors with global shutters exist. In fact, at least one of them is made by Sony (http://ww2.ptgrey.com/exmorgs). They appear to be mostly confined to industrial applications, however.
There are CMOS sensors with global shutter, but as far as I know, they are all specialist video chips. Sony has one, the IMX174. Whether this will appear in a general purpose still camera is another issue, as it requires shifting all the stored charges to a temporary location before the A to D conversion process can be performed (which takes time). That would take up silicon space, and might have issues over heat generation too larger sensors.
https://www.youtube.com/watch?v=tkh5eWECgbU
Yeah, I’m thinking RED’s sensors, only half the size but the pixel pitch is similar.
If you got a sensor half the size we’d lose half of our reasons for buying the a7S.
I’m just saying that it exists…
well, there are also people who want an APS-C camera with low MP count.
RED’s sensors are APS-H sized (not true S35mm which would be ~half the size). Also they don’t have a global shutter. Last I heard (from a RED rep) they were working on a electro-mechanical shutter (not on the chip) and even that would reduce sensitivity by ~2/3 stop. As far as I know there are no 36X24mm or even ASC-H CMOS sensors with global shutters.
I thought they were s35…Then what have I mistaken for RED? Surely there was some s35 sensor with global shutter.
Possibly. The Sony F55 I know has a global shutter. As does one of the Black Magic camera. S35 is also much smaller than full-frame 35mm. And all global shutter CMOS sensors seem to take a hit in sensitivity. Personally I would give up a stop of ISO to have a global shutter in the A7S because that would probably mean higher flash sync speeds on the stills side.
It’s still currently a rare feature. Sony f55 has a global shutter, it’s about $25k. Some of the black magic cameras have them.
Try to find me one that does what the a7S does. You won’t, because (1) they are rare and (2) global shutters reduce sensitivity.
I’m just saying that the admin’s exaggerating.
I don’t believe so. The only cameras in use that have global shutter have different circuitry than CMOS- Sony F55, Blackmagic Production- both of which take an ISO sensitivity as well as dynamic range hit.
CMOS sensors with very fast rolling shutter (so that it’s barely noticeable): Alexa, RED, F5
And Mechanical Shutters- which only exist in the F65.
Honestly it’s almost impossible to tell rolling shutter on an Alexa or RED so best solution without compromising ISO and dynamic range would just be a powered-up CMOS sensor (which a pro-sumer dslr body like the a7s probably can’t handle)
What a terrible ISO test. From these images, I would get the impression that the A7s has noticable chroma noise from ISO 200 onwards … :-/
You can’t read much into such an uncontrolled test on a pre-production camera, although if people are expecting some form of miraculous leap in high ISO performance, then I think they need to readjust their thinking. In reality, it might be a modest advance on the Nikon D4S, but that is likely to be it. CMOS sensors are getting close to achievable limits on QE, read noise and the like.
I suspect the super-high ISO is aimed more at video, where the very nature of a moving image means noise is less perceptible and opens up more options for temporal noise reduction.
did you realized that the lenses where behind a glass with a lot of reflexions??? that can introduce a lot of false CA and noise, and was like a 800% not 100% as he mention
Yet Sony still won’t announce the price, wonder what they’re waiting for …
At 2.5fps and with ~300 shots per battery, it is a 1200 Euros camera. 🙂
Hmm 300 pics per shot and 2.5fps make it a 1200 euro camera? Why? I saw it in person and it has the same if not better lowlight ability as the Nikon D4/D4s. I saw no moire or aliasing. Decent codec with a bit rate that should meet BBC standards. 4k clean output I have to admit I am disappointed it doesn’t record internally but not a big deal with the atomos shogun 4k recorder it’s pretty small and a nice monitor. 300 shots per battery is not great but batteries are cheap and these are very small so not a big deal. But I started in the days of film when you had 12,24 or 36 shots. This camera doesn’t need more then 2.5 shot. It a low light photo/movie monster
It will only be a low light photo monster IF it can focus at the same speed it can in daylight. It is fine that the focus can go down to -4 EV but not if it does it at snail pace.
For a lot of things it does not need to be faster than 2.5 frames a second but for some things it does need to be. To only be able to do 2.5 frames suggest the camera is very under powered.
Will be interesting to see how long the batteries last for video work.
Would have been nice if they had given the option of 4K recording in a grip that could have been attached.
Expecting a camera to focus in low light as quickly as it does in bright daylight is unrealistic. I missed a shot yesterday in low light because the canon 1dx I was using failed to grab focus.
And from experience if you are renting one and don’t calibrate the AF before hand. More often than not they are out…
Are you referring to micro adjustments per lens? It’s not a rental but we do beat it up like one. I’m not sure how micro adjustments would affect speed or low light focusing.
No sorry just an observation of the fact that although the 1DX has a great autofocus system. It can all come to nothing if the micro adjustment is no calibrated often…
The A7 cameras are not lightening fast in daylight. The A7s is therefore not in a position to be any slower than this in low light or it is just going to look silly compared to a D4s or D4 which is much quicker in the same lighting situation. It is OK to be a little slower but not to have the same speed as a A7 or A7r in low light which is snail pace. You cannot make a camera that has a great sensor in low light and a AF system that goes down to -4 EV and then mess it up with slow AF. Well you can make it but lets hope Sony have not done that with the A7s.
Comparing it to a d4s is unfair. It’s like comparing a Formula1 car to a compact sedan. It doesn’t have to be as fast as comparatively giant (and much more expensive) d4s or a 1dx geared specifically for fast action. If it’s faster than the a7 I say it’s a step in the right direction
I have shot with film cameras offering more than 2.5fps. Even they were (not all) not single press/shot deal. The Nikon D4/D4s offer many times this possibility. These have to be factored in while the camera gets a realistic price.
Agree, why should it cost more than A7? because?………..
New sensor production + r&d (a7 uses Sony’s most prolific sensor from a99, d600, et al.), 4k output bringing a comparatively low cost 4k solution for video production (under $5000 total being low cost vs other options) and the insane low light / low noise ability.
Okay can someone please do a brain check…How hard is it to read a spec sheet before spouting drivel??….The A7s shoots up to 5 FPS…Don’t just believe this numpty….5 FPS….exactly the same as the A7…
As for battery life…It has very little to do with the number of frames…Last week I shot a job in high frame rate mode (yes 5 fps) and happily got almost 2000 frames per battery..battery life has more to do with how long the camera is continuously on, than number of frames..If doesn’t take must power to release a shutter…
Here is the “drivel” (from: http://store.sony.com/a7s-full-frame-mirrorless-camera-zid27-ILCE7S/B/cat-27-catid-All-Alpha-a7-Cameras?_t=pfm%3Dcategory)
“Continuous Shooting Speed : Speed Priority Continuous shooting: Max. 5.0fps,
Continuous shooting: Max. 2.5fps”
I hope you can understand what those numbers mean and also what “Max” stands for. 🙂
Did you not just read your own post??…Continuous Shooting Speed : Speed Priority Continuous shooting: Max. 5.0fps…Which part of 5.0fps is difficult to understand?? Are you suggesting that 5.0fps is not usable?? I use it all the time…..
Yea i shoot over a 1k of photos before my battery runs out. Just depends how long your camera has been on and other factors.
Battery does run short but i have 4 so its not really an issue for me. Infact i did a timelapse of about 2000 shots and still had 70 percent battery left.
Ah! The “Lartigue effect”!
Im waiting for the A7 vs A7r vs A7s before i make my decision on what camera i want. My main priority is sharpness and High Iso standards
If these samples were shot with the 55 1.8 as the image on top suggests, they might be good evidence for in-raw correction: There is green CA at f8 on a 12mp sensor (ISO 100) which is rather odd considering such things were not reported on the much higher resolution production models of the A7 and A7r.
My guess is that they are actually geometrically correcting lateral CAs (which is imo a totally fine thing to do).
Bulls%%t. The crops are from the center where LCA is non-existant. And SEL55F18 has very little LCA due to symmetric design. Correcting LCA in-raw defeats the very purpose of raw. If Sony indeed does that, the proper way to prove that is taking shot with masked lens contacts so camera would not know which lens it uses.
Didn’t notice that the crops are from the center. You’re right. However, I think that wavelike patterns like these: http://www.leica-boss.com/2013/12/the-problem-with-sonys-zeiss-fe-sonnar-t-fe-35mm-f2-8-za-sel35f28z/
still speak for some kind of in-raw correction if you like it or not.
Obviously the best thing for us would be if the manufacturers disclosed all the details about their lenses so third party people can implement their own models but that’s never going to happen.
Sony A7s is literally distorting reality with so much jello.