Pages: 1 [2]   Go Down

Author Topic: 2016 predictions  (Read 9071 times)

landscapephoto

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 623
Re: 2016 predictions
« Reply #20 on: December 26, 2015, 02:30:52 am »

Can't help to notice that the (exclusive to RED)  Weapon Dragon 8K sensor (42mm width) is missing...

Indeed. That sensor is 40.96mm x 21.6mm.

Quote
Also... the Alexa 65 sensor is slightly bigger that the Alexa 765 image area at 54.12mm width (and is also missing)...


The Alexa 65 sensor is the outer blue rectangle. The Alexa 765 is a film camera.

Logged

landscapephoto

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 623
Re: 2016 predictions
« Reply #21 on: December 26, 2015, 02:39:50 am »

But it is still pretty clear that the 50mp chip used by P1/Hassy/Pentax wasn't designed for video, right?

Not only that, but it is also clear that sensors designed for digital cinema are generally smaller than 24x36. The few exceptions: Arri Alexa 65 (rent only, probably less than 10 were built), Vision Research's Phantom 65 (discontinued) and Red Weapon Dragon have sensors totally unsuited for existing MF cameras, since they all use a very wide aspect ratio and comparatively few pixels.
Logged

razrblck

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 482
  • Chill
    • Instagram
Re: 2016 predictions
« Reply #22 on: December 26, 2015, 03:46:45 am »

Yeah, motion video sensors at the high end are usually built for the aspect ratio desired. It would be a waste of silicon area (and, in turn, money) to have 4:3 sensors when all you film is in 2.40:1 or 16:9.

Those sensors are optimized for speed (sustained 24 frames at 6k or 8k for the latest models), proper heat dissipation (to reduce noise after hours of operation) and can be equipped with global shutters on top of other technicalities that clearly separate them from stills sensors that can also do some video.

The 50MP Sony sensor used in MF cameras has nothing to do with cinema sensors, in fact it can sustain a maximum of 3.3fps (all pixels). You'll have to crop and bin the pixels, like Pentax does on the 645Z, to get 24fps which renders most of the sensor area completely useless and limits it to 1080p resolution anyway.

Hasselblad could move to what Nikon, Arri and many others did for years, designing their own sensors and having third parties manufacture them. They should already have enough talent in house to accomplish this.
Logged
Instagram (updated often)

Theodoros

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2454
Re: 2016 predictions
« Reply #23 on: December 26, 2015, 05:31:34 am »


  ....in fact it can sustain a maximum of 3.3fps (all pixels).....


Interesting... do you have a reference for that?
Logged

razrblck

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 482
  • Chill
    • Instagram
Re: 2016 predictions
« Reply #24 on: December 26, 2015, 05:42:16 am »

Interesting... do you have a reference for that?

http://www.sony.net/Products/SC-HP/IS/sensor2/products/

It's the IMX161. Unfortunately the full datasheet is not publicly available, one probably has to sign an NDA to get that.
Logged
Instagram (updated often)

Theodoros

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2454
Re: 2016 predictions
« Reply #25 on: December 26, 2015, 07:54:45 am »

http://www.sony.net/Products/SC-HP/IS/sensor2/products/

It's the IMX161. Unfortunately the full datasheet is not publicly available, one probably has to sign an NDA to get that.
Thanks for that... I wasn't aware of that table, correct me if I'm wrong but if the sensor is of 3.3fps ability at full resolution, won't than mean that the frame rate will be of 36fps if 4 pixels are binned for 4k video?
Logged

Ken R

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 849
Re: 2016 predictions
« Reply #26 on: December 26, 2015, 07:55:27 am »

2016 Predictions

#1 Some people will continue not to be happy with what gear is available now (either not enough DR, Sharpness, some technical issue and/or too much $$) and continue chasing the holy grail of gear with just the right combination of cutting edge performance with no issues at a reasonable price (which is great for companies that sell gear!)

#2 Others will continue making amazing images with what they got...

ok, on a more serious note:

I do not expect any new medium format sensor being offered in 2016. A new 5D mk4 should be out along with a new Fuji X body (or 2!). Sony, Fuji and Leica should continue to expand their lens line for mirrorless and Canon will introduce one more L vII lens. That should be it for 2016 regarding major higher end camera gear releases.
« Last Edit: December 26, 2015, 08:36:36 am by Ken R »
Logged

razrblck

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 482
  • Chill
    • Instagram
Re: 2016 predictions
« Reply #27 on: December 26, 2015, 08:26:34 am »

Thanks for that... I wasn't aware of that table, correct me if I'm wrong but if the sensor is of 3.3fps ability at full resolution, won't than mean that the frame rate will be of 36fps if 4 pixels are binned for 4k video?

Maybe. We only know it has an 8 channel LVDS connection, but other than that we don't know what kind of bottleneck it might have during readout. Don't forget that some of the Dalsa CCD sensors in digital backs also support between 4 and 1 fps, but we all know how bad they are with actual motion video at any resolution.

If you want smooth images with very little hints of rolling shutter, you need a much faster readout than what your video would be. What Pentax managed could very well be the best that can be done with it.
Logged
Instagram (updated often)

Theodoros

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2454
Re: 2016 predictions
« Reply #28 on: December 26, 2015, 09:16:10 am »

Maybe. We only know it has an 8 channel LVDS connection, but other than that we don't know what kind of bottleneck it might have during readout. Don't forget that some of the Dalsa CCD sensors in digital backs also support between 4 and 1 fps, but we all know how bad they are with actual motion video at any resolution.

If you want smooth images with very little hints of rolling shutter, you need a much faster readout than what your video would be. What Pentax managed could very well be the best that can be done with it.

I doubt that... IMO, Pentax (or the rest of MF companies that are using the sensor) is simply inexperienced to other than commercial video and thus are not (yet) capable to take full advantage of the sensor's capability... IMO, Sony did make the sensor aiming it to large format video (as to stay on part with cinema's recent developments)... the reason they don't use it yet on a camera of their own, has (IMO) to do with the mount they use (the E-mount) which is universal to all their products, commercial or professional... OTOH, this doesn't mean that there is no research happening in the labs...

I simply don't believe that Sony would ever bother with a sensor as to make only a few of them annually (which may mean that it's not any profitable... if not the opposite)... Lets not forget that Cmos sensors are cheaper than CCD to make, but they are considerably more expensive to develop and put in production... Other than that, having a sensor that has "just" the pixel size as to deliver 2,4 & 8k video in its long side but 1.5, 3 & 6K video if cropped down to 32.8mm (which would fit ideally in the E-mount) width and the shorter side of it is used as base ...can't be accidental with Sony! ...remember it's Sony we are talking about.... not Dalsa!
Logged

razrblck

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 482
  • Chill
    • Instagram
Re: 2016 predictions
« Reply #29 on: December 26, 2015, 10:18:21 am »

I doubt that... IMO, Pentax (or the rest of MF companies that are using the sensor) is simply inexperienced to other than commercial video and thus are not (yet) capable to take full advantage of the sensor's capability... IMO, Sony did make the sensor aiming it to large format video (as to stay on part with cinema's recent developments)... the reason they don't use it yet on a camera of their own, has (IMO) to do with the mount they use (the E-mount) which is universal to all their products, commercial or professional... OTOH, this doesn't mean that there is no research happening in the labs...

Sony professional video cameras support industry standard PL mount, which covers movie 35mm (more like still APS formats). They do use E-mount on some of the prosumer models.

I simply don't believe that Sony would ever bother with a sensor as to make only a few of them annually (which may mean that it's not any profitable... if not the opposite)... Lets not forget that Cmos sensors are cheaper than CCD to make, but they are considerably more expensive to develop and put in production... Other than that, having a sensor that has "just" the pixel size as to deliver 2,4 & 8k video in its long side but 1.5, 3 & 6K video if cropped down to 32.8mm (which would fit ideally in the E-mount) width and the shorter side of it is used as base ...can't be accidental with Sony! ...remember it's Sony we are talking about.... not Dalsa!

Sony makes a lot of sensors, and many of them get very little use. I'd say this one has been particularly successful for them, considering the low yelds they must have. The bigger the chip, the less you can fit on a silicon wafer and the more catastrophic small issues in manufacturing can become. A tiny mistake can lead to one or more sensors having to be discarded, and when you can fit only 20 or 30 of them instead of the hundreds or thousands of smaller formats it does get expensive really fast. R&D on CMOS sensors is no different than CCD, and Sony has that covered thanks to decades of experience on the matter. Each new design has been most probably already paid in full by previous models, and even if they don't get to sell it's still experience they can put into newer stuff.

Yes, the sensor happens to have the requisite resolution for up to 8K video, but does it have the right on chip electronics to support such a high data throughput at even 24 frames per second? We don't know. It would also need a beefy processor as well as some seriously fast storage to be usable.

Sorry, but the E-mount is a tight fit even for a 36x24mm sensor. It will never be able to fit a 43x32mm one without losing all four corners.

If Sony really wanted it to be a video sensor, they would've made it in a different aspect ratio right from the start. It makes no sense to manufacture all the extra pixels if you don't use them, plus if you make the sensor height shorter you can fit more of those on a silicon wafer reducing costs.

If you are wondering, then, why the Alexa XT and VistaVision are 4:3 (or close by), it's because the former supports anamorphic lenses (and having a bigger sensor helps in not losing resolution) and the latter is a native anamorphic format.

Anamorphic video bigger than 35mm has always been more of a rarity, and after IMAX there was no need for it anymore. Keep in mind that an anamorphic 65mm lens would cover about a full frame 36x24mm sensor, so again no need for anything bigger.

All the signs point away from a video centric sensor. It's more of a general purpose sensor that can take stunning pictures and even manages to be fast enough for video (under certain conditions and limitations).
Logged
Instagram (updated often)

eronald

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 6642
    • My gallery on Instagram
Re: 2016 predictions
« Reply #30 on: December 26, 2015, 01:42:50 pm »

I too think that Sony designed this sensor partly because they needed an 8K capture device for internal use.

The first sample images from this sensor, circulated in the dev community, are more than 3 years old (!).

Edmund


Sony professional video cameras support industry standard PL mount, which covers movie 35mm (more like still APS formats). They do use E-mount on some of the prosumer models.

Sony makes a lot of sensors, and many of them get very little use. I'd say this one has been particularly successful for them, considering the low yelds they must have. The bigger the chip, the less you can fit on a silicon wafer and the more catastrophic small issues in manufacturing can become. A tiny mistake can lead to one or more sensors having to be discarded, and when you can fit only 20 or 30 of them instead of the hundreds or thousands of smaller formats it does get expensive really fast. R&D on CMOS sensors is no different than CCD, and Sony has that covered thanks to decades of experience on the matter. Each new design has been most probably already paid in full by previous models, and even if they don't get to sell it's still experience they can put into newer stuff.

Yes, the sensor happens to have the requisite resolution for up to 8K video, but does it have the right on chip electronics to support such a high data throughput at even 24 frames per second? We don't know. It would also need a beefy processor as well as some seriously fast storage to be usable.

Sorry, but the E-mount is a tight fit even for a 36x24mm sensor. It will never be able to fit a 43x32mm one without losing all four corners.

If Sony really wanted it to be a video sensor, they would've made it in a different aspect ratio right from the start. It makes no sense to manufacture all the extra pixels if you don't use them, plus if you make the sensor height shorter you can fit more of those on a silicon wafer reducing costs.

If you are wondering, then, why the Alexa XT and VistaVision are 4:3 (or close by), it's because the former supports anamorphic lenses (and having a bigger sensor helps in not losing resolution) and the latter is a native anamorphic format.

Anamorphic video bigger than 35mm has always been more of a rarity, and after IMAX there was no need for it anymore. Keep in mind that an anamorphic 65mm lens would cover about a full frame 36x24mm sensor, so again no need for anything bigger.

All the signs point away from a video centric sensor. It's more of a general purpose sensor that can take stunning pictures and even manages to be fast enough for video (under certain conditions and limitations).
Logged
If you appreciate my blog posts help me by following on https://instagram.com/edmundronald

BernardLanguillier

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 13983
    • http://www.flickr.com/photos/bernardlanguillier/sets/
Re: 2016 predictions
« Reply #31 on: December 26, 2015, 05:18:57 pm »

Another prediction for 2016 after reading through this thread!  ;D

Some people will continue to think in 2016 that discussing equipment and taking (great) photographs are 2 mutually exclusive activities. Just like working out and reading books! ;)

Cheers,
Bernard

eronald

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 6642
    • My gallery on Instagram
Re: 2016 predictions
« Reply #32 on: December 26, 2015, 09:29:54 pm »

Another prediction for 2016 after reading through this thread!  ;D

Some people will continue to think in 2016 ;)

Cheers,
Bernard

That's ok - the US is a free country, they're allowed to think.
Logged
If you appreciate my blog posts help me by following on https://instagram.com/edmundronald
Pages: 1 [2]   Go Up