Rumor: Sony A7s will have 4K video
- Thread starter vanvideo
- Start date
Nirv
Carbonite Member
That my friends is what's commonly known as a deal breaker.
Just a basic 24p point.
Don't put a long lens on and move your camera quickly - it will look bad.
In fact, don't move your camera quickly with any lens - it will look bad.
These tests are like the guys who point cameras at junk in bad light then blow stills up to 300% and point out the faults.
Why bother with "tests" that represent things that no competent film maker would ever do...because they're competent?
Don't put a long lens on and move your camera quickly - it will look bad.
In fact, don't move your camera quickly with any lens - it will look bad.
These tests are like the guys who point cameras at junk in bad light then blow stills up to 300% and point out the faults.
Why bother with "tests" that represent things that no competent film maker would ever do...because they're competent?
Absolutely a deal breaker for all the directors looking for that awesome zoom and wiggle shot.
Nirv
Carbonite Member
Yeah, or on a shoulder rig running alongside your talent or maybe in a car on a uneven road or on a steadicam orbiting the talent.
the camera isn't fit for purpose as a video camera. I'm sure if you shoot on a tripod by candle light it'll come into its own though.
joe 1008
Veteran
On a 5D a 50mm lens is the absolute limit for handheld, even with a rig and the intention to hold it steady. The A7s doesn't seem to look better.
Samuel H
Major Contributor
ok, so, just trying to see if this is completely off the charts...
I stopped the youtube clip in two frames, one panning left and one panning right, with same focal length (both before the zoom-in), showing the ghosting from two consecutive frames of the A7s. I took those screenshots, corrected the perspective distortion, and measured intra-frame skew vs inter-frame displacement, in pixels.
The results are to be taken with a grain of salt the size of an SUV, but here they are:
* if the A7s was set at 24p, the rolling shutter is around 31ms (average of four numbers that go from 26ms to 35ms)
* if the A7s was set at 30p, the rolling shutter is around 25 ms (average of four numbers that go from 21ms to 28ms)
This would put it at the bottom of my RS table, but not completely off the chart. It's a deal-breaker for action shots, but not for regular stuff. I'm not bothered by the 29,4ms rolling shutter of my NEX-5N, though of course the 15ms of the GH3 or the FS100 would be a lot better.
Make that a grain of salt suitable for amateur mountaneering...
edit: this is the kind of image that I'm using for this measurement, intra-frame skew is around 75% of inter-frame displacement.
I stopped the youtube clip in two frames, one panning left and one panning right, with same focal length (both before the zoom-in), showing the ghosting from two consecutive frames of the A7s. I took those screenshots, corrected the perspective distortion, and measured intra-frame skew vs inter-frame displacement, in pixels.
The results are to be taken with a grain of salt the size of an SUV, but here they are:
* if the A7s was set at 24p, the rolling shutter is around 31ms (average of four numbers that go from 26ms to 35ms)
* if the A7s was set at 30p, the rolling shutter is around 25 ms (average of four numbers that go from 21ms to 28ms)
This would put it at the bottom of my RS table, but not completely off the chart. It's a deal-breaker for action shots, but not for regular stuff. I'm not bothered by the 29,4ms rolling shutter of my NEX-5N, though of course the 15ms of the GH3 or the FS100 would be a lot better.
Make that a grain of salt suitable for amateur mountaneering...
edit: this is the kind of image that I'm using for this measurement, intra-frame skew is around 75% of inter-frame displacement.
FrameFarmMedia
Veteran
The GH4 is supposed to read out the sensor 50% faster than GH3 or so I have read.
Samuel H
Major Contributor
if true, that's nearly global shutter (GH3 is at the top of my chart, in real-video-camera territory)
but it sounds too good to be true, my guess is that this measure is in a per-pixel or per-line basis, and RS will be the same
G
Guest
Guest
Yes, the rolling shutter on the A7s is fairly sad. Although, for most things it is the lesser of evils. I am glad a 'baby alexa' is now accessible to some consumers. I hope for more internal recording soon.
I hate the separation of pro image and consumer image. I can understand on the refinement side of things, but there has been far too large a gap of 'pro' versus 'consumer' in the digital imaging world.
If I want, I could go make wet plates and have incredible archival quality photos of friends and family for under $1000 dollars for camera and setup. But all these digital cameras under $1000 are rather sad, until more recently with sony/nikon.
The large gap in image quality has created an issue for professionals, since people just assume it is the 'camera'. I am glad more recently when I let people use my cameras and they realize a 'snap shot' is a 'snap shot' and generally looks similar across all camera platforms. That perception may never go away, but at the least, I hope there will be less of a separation of pro and consumer image quality. Cinema will always be more refined, and rightfully so, but it is on the hobby/artist side that much of the nuance is appreciated, and it is the 'dumb' consumer that needs the latitude since they don't understand lighting and exposure. So, it make more sense for consumers to have better specs than pros.
I hate the separation of pro image and consumer image. I can understand on the refinement side of things, but there has been far too large a gap of 'pro' versus 'consumer' in the digital imaging world.
If I want, I could go make wet plates and have incredible archival quality photos of friends and family for under $1000 dollars for camera and setup. But all these digital cameras under $1000 are rather sad, until more recently with sony/nikon.
The large gap in image quality has created an issue for professionals, since people just assume it is the 'camera'. I am glad more recently when I let people use my cameras and they realize a 'snap shot' is a 'snap shot' and generally looks similar across all camera platforms. That perception may never go away, but at the least, I hope there will be less of a separation of pro and consumer image quality. Cinema will always be more refined, and rightfully so, but it is on the hobby/artist side that much of the nuance is appreciated, and it is the 'dumb' consumer that needs the latitude since they don't understand lighting and exposure. So, it make more sense for consumers to have better specs than pros.
joe1946
Veteran
Just like the $12,000 Canon EOS 1D C rolling shutter test with much slower camera movement :
sammysammy
Well-known member
wonder how good the 60p 1080p on it?
roxics
Veteran
Agreed. In my opinion the only difference between pro and consumer should be ergonomics.
Right about SUV size salt lol.You do understand that its a video of a video played on a monitor and the monitor itself will have ghosting.
G
Guest
Guest
big +1 (as well as the standardization that only pros truly need. ie mounting systems, I/O portals and connections, certain codecs, lens mounts, etc, etc, etc)
Samuel H
Major Contributor
yes, it has ghosting, and that's exactly what I used to calculate this (check the image I posted)
Ian-T
Veteran
It does look impressive but I'm not really sure. The very first part of the clip is almost pitch black. Yet by the time we get to the end of the clip we see the guy handling the fish and gear and walking back and forth as if there's plenty of light around him. Makes me pause and wonder if it was really that dark (in the start). Maybe it's just me...
G
Guest
Guest
true. the fire is not clipping, meaning, the ambient light is good.
Beat me to it. It's practically dead at 1600 2.8@1/50th
There isn't any light there. The new PD170?
Or maybe there's just a lot of usable latitude in the 8-bit SLog2??? More than we're used to seeing from any other DSLR shooting video available?