Pages: 1 [2] 3 4   Go Down

Author Topic: I guess we are going to see more MF CMOS introductions…  (Read 12091 times)

narikin

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1376
Re: I guess we are going to see more MF CMOS introductions…
« Reply #20 on: January 26, 2014, 11:23:23 am »

2. There is ZERO actual demand for chips bigger than the 60/80MP Dalsa chips.

Where on earth do you get these silly assertions from?

There's lots of folks out there who would buy a 100mp back. Me for one.

Just because something isn't needed for your photography, does not mean it isn't needed by someone else's!
Logged

EricWHiss

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2639
    • Rolleiflex USA
Re: I guess we are going to see more MF CMOS introductions…
« Reply #21 on: January 26, 2014, 12:21:57 pm »

I think a 56x56mm sensor for Hy6, tech (and V) could work commercially, maybe with 7-8um pixels (49-64 megapixels). Problem is that Hassy and Phase don't have any reason to make such a back as their cameras don't work with that format.

I like your idea!  :-)   If you include the V, the Hy6, RZ, tech, aerial, reproduction markets its not that small of a group.   Who knows maybe they'd consider it? But my guess is they have less control of the sensor pitch than anything else - so it might come through with higher pixel count which is okay with me.
Logged
Rolleiflex USA

Sareesh Sudhakaran

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 546
    • The Indie Farm
Re: I guess we are going to see more MF CMOS introductions…
« Reply #22 on: January 27, 2014, 06:26:55 am »

Hi,

Sorry, I am not familiar with cinema.

Can you explain the benefits of 70mm, for those of us who are shootings stills?

Best regards
Erik


70mm (just a pseudo-term like Super 35mm) is somewhat equivalent to 6x7 medium format. E.g., IMAX is 70.41 × 52.63 mm. Other than IMAX, "70mm" is also shot on 65mm film (Todd-AO 48.56 x 20.73 mm) - and projected back on 70mm.

There are movies being shot in IMAX and 65mm even today - The Dark Knight Rises, Samsara, The Master, etc. Not small movies. Obviously, there are cinematographers and filmmakers enamored with the "medium format look" and want to translate that into cinema. The lenses used are 'cinevised' versions of Hasselblad, Mamiya and Schneider lenses. What is stopping more productions are the cost and weight of film (and the camera, by consequence).

As you can see, the Todd-AO sensor size is not that big. It is pretty close to what medium format cameras have today. In fact, it is damn close to what Sony has just released.

A thought experiment:

What if Sony put down money to make a digital 70mm film (they produce movies) shot on their own medium format digital camera (they make cameras and computers) and then distributed that globally (they distribute movies globally) to cinemas (they are fighting a battle for digital projection), Blu-ray (they co-invented it) and the Internet (Luckily the only place they have no clue about) - to be seen on Sony 4K televisions? If the demand has been created by someone else, they will quickly push large sensor cameras into the market.

What makes me think they will? They introduced a 35mm full frame VG900 camcorder (but no Cinealta camera). They invented a completely new mount - E, and have also adapted it to heavy cinema cameras (called the FZ mount). The Sony A7R must have been in their sights when they did so. If they have agreed to a few MFDB sensor orders (that surely is in the lower four figures, optimistically), what is in their sights for tomorrow?

For years Leica have marketed the 'superiority' of their cameras. They did this by using the scarcity principle - low volumes at a large cost. What this does is create a huge market of people who lust after their cameras but can't afford them. Sony is the type of company that goes after such a market.

Suddenly Sony comes out with an A7R that rivals the best any of them (Leica, Canon, Nikon) have to offer. They started by selling some of them sensors. In the MFDB world, Phase One and Hasselblad have marketed the 'superiority' of the format (which is mainly the size of the sensor) for years. Sony has now started selling them sensors...

Sony never tries to invent a market. They just wait for it to mature and then 'arrives'. They don't care about any particular segment, they want to dominate the entire market from end to end. Let's see the pattern here:

  • Michael says a million visitors come every month to LuLa. Why would they come, if they are not interested in MFDB? Assuming only 25% of them are real humans, that's about 250,000 per month. That's not a small market.
  • Steve Hendrix and Doug Peterson have consistently maintained that MFDB is selling quite well. Obviously, the combined marketing efforts of the industry as a whole are working. This is why Leica and Pentax showed interest.
  • Why did Nikon feel the need for a 40 MP DSLR, if not for the demands of MFDB fans for cheaper '40 MP'?
  • Every time there's a MFDB discussion, the topic turns to 'I wish it were cheaper'. What the MFDB manufacturers and their marketing have created is a huge market of people that lust for MFDB but can't afford it.
  • Now Sony has shown interest. They are the masters of 'good enough products for cheap'.

The good news, one hopes, is, a 70mm cinema camera. That will definitely pique a lot of filmmakers' interests (in numbers greater than the total size of the MFDB market). These filmmakers won't demand still cameras of the same size, because the digital 70mm camera will shoot 8K (40 MP) frames at 24 fps. And just in time for the 2020 8K broadcast by NHK.

This will also put Sony ahead of Red and Arri, and back at the top of digital cinema pantheon. One camera to rule them all.

Just a hunch. Anybody see any benefits?
Logged
Get the Free Comprehensive Guide to Rigging ANY Camera - one guide to rig them all - DSLRs to the Arri Alexa.

hjulenissen

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2051
Re: I guess we are going to see more MF CMOS introductions…
« Reply #23 on: January 27, 2014, 07:08:22 am »

Not at all! Ideally a sensor should outresolve the lens. If the lens outresolves the sensor you will get artifacts like moiré, false colours and jagged lines. An actual pixels image will be awful, but downscale that image and it will be beautiful.
I think that we are in agreement, but your words could be interpreted in a direction that I disagree with.

Downscaling does not make an image "better". Rendering it at smaller sizes and/or increasing the viewing distance may hide small artifacts, though.

Ideally, we want to maintain the highest possible resolution end-to-end in our scene reproduction activity (all else being equal). The total degradation of the scene by all of the components (among other things) tells us how up-close it is possible to view the end-result.

-h
Logged

ndevlin

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 679
    • Follow me on Twitter
Re: I guess we are going to see more MF CMOS introductions…
« Reply #24 on: January 27, 2014, 08:18:41 am »

Where on earth do you get these silly assertions from?

There's lots of folks out there who would buy a 100mp back. Me for one.

Just because something isn't needed for your photography, does not mean it isn't needed by someone else's!

The number of people who would plunk down $60 or $80K for that (which is what the economies of scale would demand) is countable on fingers and toes, and roughly equals the number of people who would complain unrelentingly about the poor corner-quality of their wonder-back   ;D

But we all gotta dream! (for me it was a digital Mamyia 6 -- would have killed a man with my bare hands for one)

- N.
Logged
Nick Devlin   @onelittlecamera        ww

Doug Peterson

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 4210
    • http://www.doug-peterson.com
Re: I guess we are going to see more MF CMOS introductions…
« Reply #25 on: January 27, 2014, 09:52:20 am »

2. There is ZERO actual demand for chips bigger than the 60/80MP Dalsa chips.

Just as a reference point, I'm very sure that if/when P1 announced a 160MP sensor we'd sell a few in the first hour.

In addition to the "usual" (architecture, interior, landscape) who desire massive prints there are also the niches of aerial, scientific, industrial, cultural heritage, film scanning, and more.

Don't get me wrong - I don't think I'll ever be recommending a 160mp sensor for portrait work or fashion. But just because YOU have zero demand for higher resolution doesn't mean there isn't more than enough demand for such a sensor to justify it's creation.

torger

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3267
Re: I guess we are going to see more MF CMOS introductions…
« Reply #26 on: January 27, 2014, 10:00:49 am »

There's already are huge sensor cameras on the market, for mapping that is. Integraph's DMC II250 uses a 250 megapixel Dalsa sensor of 96x82mm size (plus four 42 megapixel sensors of smaller size with own lenses to make a multispectral capture). I'm assuming it makes Phase One's digital backs seem cheap ;D
Logged

JoeKitchen

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 5022
Re: I guess we are going to see more MF CMOS introductions…
« Reply #27 on: January 27, 2014, 10:04:13 am »

The resolution that we have right now, I feel, is more than enough, as do many other APs I speak to.  What concerns me is how poorly the high resolution backs handle lens cast.  I would like to see a 50 MP full frame CCD that has the same capabilities as the IQ260.  This, I feel, would be the perfect compromise between high res and large enough pixels to handle LCCs.  Many APs I have spoken to also agree with me on this; no one really wants to deal with those uber big lenses Rodenstock keeps on releasing.  (You can not even leave the Roddie 32mm on the camera due to the weight, what is up with that?  And why did they not just design it to mount on a stronger Copal 1, which is only marginally bigger than the 0?)

Unfortunately, Phase One does not seem to really be concerned with tech cameras.  Not counting the dealers and techs I speak to, who have all voiced concerns about the SK wides with the new backs, the hand full of times I have spoken to someone that works for P1, they smooth it over and just say that the software handles that.  I am not sure if this is just marketing hype or that they just do not know? 

I shoot with a P45+ right now and love it.  Maybe I will upgrade to an IQ260 in 2 or 3 years, but after that ... not sure.  If they continue with this pointless MP race, than the IQ260 will probably be the highest res I would ever want to go to. 
« Last Edit: January 27, 2014, 10:30:14 am by JoeKitchen »
Logged
"Photography is one percent inspiration and ninety-nine percent

torger

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3267
Re: I guess we are going to see more MF CMOS introductions…
« Reply #28 on: January 27, 2014, 10:20:46 am »

I guess a not too small pixel size in combination with back illumination process would greatly improve the angular response.



as far as I know only cell phone sensor with extremely small pixels are manufactured with a back-illuminated layout today. Not sure what the reason is, but I'm assuming it's hard to make large sensors back-illuminated.

Angular response is however not the reason sensors got back illuminated, it was to improve sensitivity (ie quantum efficiency I assume) especially with extremely small pixels as found in cell-phone sensors. There's not the same gain for larger pixels as found in 135 and medium format sensors so it's not sure that technology will find its way to larger sensors, at least not for the same reasons.

I'm also one of the users that think that improved angular response is the single most desired feature of future medium format sensors, and if that will cause a halt in the megapixel race, I'm fine with that. With back illumination it might be possible to combine small pixel size with wide angular response though, I don't know. Today it seems hard.
« Last Edit: January 27, 2014, 10:51:34 am by torger »
Logged

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
Re: I guess we are going to see more MF CMOS introductions…
« Reply #29 on: January 27, 2014, 12:57:26 pm »

Hi,

The Sony RX 100II has a back side illuminated sensor, it is 1" size.

Best regards
Erik


I guess a not too small pixel size in combination with back illumination process would greatly improve the angular response.



as far as I know only cell phone sensor with extremely small pixels are manufactured with a back-illuminated layout today. Not sure what the reason is, but I'm assuming it's hard to make large sensors back-illuminated.

Angular response is however not the reason sensors got back illuminated, it was to improve sensitivity (ie quantum efficiency I assume) especially with extremely small pixels as found in cell-phone sensors. There's not the same gain for larger pixels as found in 135 and medium format sensors so it's not sure that technology will find its way to larger sensors, at least not for the same reasons.

I'm also one of the users that think that improved angular response is the single most desired feature of future medium format sensors, and if that will cause a halt in the megapixel race, I'm fine with that. With back illumination it might be possible to combine small pixel size with wide angular response though, I don't know. Today it seems hard.
Logged
Erik Kaffehr
 

EricWHiss

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2639
    • Rolleiflex USA
Re: I guess we are going to see more MF CMOS introductions…
« Reply #30 on: January 27, 2014, 01:09:30 pm »

Two thoughts:

2. There is ZERO actual demand for chips bigger than the 60/80MP Dalsa chips.
 

Nick, not ZERO, as I myself would be interested in one - 56x56.  If they made a sensor to fit 4x5, I would also be interested.   I think the larger formats have a look that is hard to replicate with 35mm.
Logged
Rolleiflex USA

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
Re: I guess we are going to see more MF CMOS introductions…
« Reply #31 on: January 27, 2014, 01:56:53 pm »

Hi,

Thanks for explanation!

If you feel that a large sensor is needed to achieve IMAX quality I can see that there is a significant market for large sensor video cameras. Such cameras could be very expensive and still be a bargain.

From what I have seen from stills, my 24 MP full frame camera is a good match for my Pentax 67 using Velvia, so I feel 24x36 would do, but larger is mostly better.

So you think 8K is around the corner?

Best regards
Erik


70mm (just a pseudo-term like Super 35mm) is somewhat equivalent to 6x7 medium format. E.g., IMAX is 70.41 × 52.63 mm. Other than IMAX, "70mm" is also shot on 65mm film (Todd-AO 48.56 x 20.73 mm) - and projected back on 70mm.

There are movies being shot in IMAX and 65mm even today - The Dark Knight Rises, Samsara, The Master, etc. Not small movies. Obviously, there are cinematographers and filmmakers enamored with the "medium format look" and want to translate that into cinema. The lenses used are 'cinevised' versions of Hasselblad, Mamiya and Schneider lenses. What is stopping more productions are the cost and weight of film (and the camera, by consequence).

As you can see, the Todd-AO sensor size is not that big. It is pretty close to what medium format cameras have today. In fact, it is damn close to what Sony has just released.

A thought experiment:

What if Sony put down money to make a digital 70mm film (they produce movies) shot on their own medium format digital camera (they make cameras and computers) and then distributed that globally (they distribute movies globally) to cinemas (they are fighting a battle for digital projection), Blu-ray (they co-invented it) and the Internet (Luckily the only place they have no clue about) - to be seen on Sony 4K televisions? If the demand has been created by someone else, they will quickly push large sensor cameras into the market.

What makes me think they will? They introduced a 35mm full frame VG900 camcorder (but no Cinealta camera). They invented a completely new mount - E, and have also adapted it to heavy cinema cameras (called the FZ mount). The Sony A7R must have been in their sights when they did so. If they have agreed to a few MFDB sensor orders (that surely is in the lower four figures, optimistically), what is in their sights for tomorrow?

For years Leica have marketed the 'superiority' of their cameras. They did this by using the scarcity principle - low volumes at a large cost. What this does is create a huge market of people who lust after their cameras but can't afford them. Sony is the type of company that goes after such a market.

Suddenly Sony comes out with an A7R that rivals the best any of them (Leica, Canon, Nikon) have to offer. They started by selling some of them sensors. In the MFDB world, Phase One and Hasselblad have marketed the 'superiority' of the format (which is mainly the size of the sensor) for years. Sony has now started selling them sensors...

Sony never tries to invent a market. They just wait for it to mature and then 'arrives'. They don't care about any particular segment, they want to dominate the entire market from end to end. Let's see the pattern here:

  • Michael says a million visitors come every month to LuLa. Why would they come, if they are not interested in MFDB? Assuming only 25% of them are real humans, that's about 250,000 per month. That's not a small market.
  • Steve Hendrix and Doug Peterson have consistently maintained that MFDB is selling quite well. Obviously, the combined marketing efforts of the industry as a whole are working. This is why Leica and Pentax showed interest.
  • Why did Nikon feel the need for a 40 MP DSLR, if not for the demands of MFDB fans for cheaper '40 MP'?
  • Every time there's a MFDB discussion, the topic turns to 'I wish it were cheaper'. What the MFDB manufacturers and their marketing have created is a huge market of people that lust for MFDB but can't afford it.
  • Now Sony has shown interest. They are the masters of 'good enough products for cheap'.

The good news, one hopes, is, a 70mm cinema camera. That will definitely pique a lot of filmmakers' interests (in numbers greater than the total size of the MFDB market). These filmmakers won't demand still cameras of the same size, because the digital 70mm camera will shoot 8K (40 MP) frames at 24 fps. And just in time for the 2020 8K broadcast by NHK.

This will also put Sony ahead of Red and Arri, and back at the top of digital cinema pantheon. One camera to rule them all.

Just a hunch. Anybody see any benefits?
Logged
Erik Kaffehr
 

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
Re: I guess we are going to see more MF CMOS introductions…
« Reply #32 on: January 27, 2014, 02:00:41 pm »

Any MF camera with an exchangable back. Also technical cameras.

Best regards
Erik


But to put on what?
Logged
Erik Kaffehr
 

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
Re: I guess we are going to see more MF CMOS introductions…
« Reply #33 on: January 27, 2014, 02:10:25 pm »

Hi,

The enclosed image illustrates pretty well what I mean. The images were shot at around 3.5 using a 150 mm lens, with three different cameras.

- Left Hasselblad 555ELD, with a Zeiss Sonnar 150/4 (probably f/8) with a P45+ back (6.8 my pixel pitch)
- Center Sony Alpha 99, SAL 70-400/4-5.6 at 150 mm and f/8 (6 my pixel pitch)
- Right Sony Alpha 77, SAL 70-400/4-5.6 at 150 and f/8 (3.9 my pixel pitch)

The three different images obviously cover very different FOV (Field Of View), but the test is intended to demonstrate the potential benefit of small pixels.

The Sony image were downscaled to the same image size the P45+ has. It could be stated that the right image shows what a 130-160 MP medium format camera would be capable of.

Best regards
Erik

I think that we are in agreement, but your words could be interpreted in a direction that I disagree with.

Downscaling does not make an image "better". Rendering it at smaller sizes and/or increasing the viewing distance may hide small artifacts, though.

Ideally, we want to maintain the highest possible resolution end-to-end in our scene reproduction activity (all else being equal). The total degradation of the scene by all of the components (among other things) tells us how up-close it is possible to view the end-result.

-h
« Last Edit: January 27, 2014, 02:12:09 pm by ErikKaffehr »
Logged
Erik Kaffehr
 

Theodoros

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2454
Re: I guess we are going to see more MF CMOS introductions…
« Reply #34 on: January 27, 2014, 02:33:00 pm »

Hi,

The enclosed image illustrates pretty well what I mean. The images were shot at around 3.5 using a 150 mm lens, with three different cameras.

- Left Hasselblad 555ELD, with a Zeiss Sonnar 150/4 (probably f/8) with a P45+ back (6.8 my pixel pitch)
- Center Sony Alpha 99, SAL 70-400/4-5.6 at 150 mm and f/8 (6 my pixel pitch)
- Right Sony Alpha 77, SAL 70-400/4-5.6 at 150 and f/8 (3.9 my pixel pitch)

The three different images obviously cover very different FOV (Field Of View), but the test is intended to demonstrate the potential benefit of small pixels.

The Sony image were downscaled to the same image size the P45+ has. It could be stated that the right image shows what a 130-160 MP medium format camera would be capable of.

Best regards
Erik

There is moire with your P45+ Erik… and some less so with your A99…
« Last Edit: January 27, 2014, 02:35:17 pm by T.Dascalos »
Logged

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
Re: I guess we are going to see more MF CMOS introductions…
« Reply #35 on: January 27, 2014, 02:39:14 pm »

Oh yes,

There is always moiré on the P45+ at medium apertures if there is pixel level fine detail in the picture. The Sony Alpha 99 (6 my) has some, the Alpha 77 (3.9 my) has virtually none.

Moiré and large pixels go hand in hand, stopping down to f/16 eliminates moiré on the P45+ but it also reduces sharpness. The sample below is a good example of this (left f/8, right f/16). Best way to avoid moiré (and other aliasing artefacts) is to reduce pixel size.

Best regards
Erik


There is moire with your P45+ Erik…
« Last Edit: January 27, 2014, 02:45:08 pm by ErikKaffehr »
Logged
Erik Kaffehr
 

Joe Towner

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1365
Re: I guess we are going to see more MF CMOS introductions…
« Reply #36 on: January 27, 2014, 02:55:08 pm »

There will be a market for it, just how big a market is the fun question.  I'm interested in what DT/CI see folks trade-in on a cmos based unit, and how many folks come back to ccd based on whatever workflow.

The higher ISO will be welcomed in a lot of segments - weddings or places where you're not shooting staged shots.  Just another tool for the bag - an extremely expensive bag.
Logged
t: @PNWMF

rainer_v

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1194
    • http://www.tangential.de
Re: I guess we are going to see more MF CMOS introductions…
« Reply #37 on: January 27, 2014, 07:28:58 pm »

The resolution that we have right now, I feel, is more than enough, as do many other APs I speak to.  What concerns me is how poorly the high resolution backs handle lens cast.  I would like to see a 50 MP full frame CCD that has the same capabilities as the IQ260.  This, I feel, would be the perfect compromise between high res and large enough pixels to handle LCCs.  Many APs I have spoken to also agree with me on this; no one really wants to deal with those uber big lenses Rodenstock keeps on releasing.  (You can not even leave the Roddie 32mm on the camera due to the weight, what is up with that?  And why did they not just design it to mount on a stronger Copal 1, which is only marginally bigger than the 0?)

Unfortunately, Phase One does not seem to really be concerned with tech cameras.  Not counting the dealers and techs I speak to, who have all voiced concerns about the SK wides with the new backs, the hand full of times I have spoken to someone that works for P1, they smooth it over and just say that the software handles that.  I am not sure if this is just marketing hype or that they just do not know? 

I shoot with a P45+ right now and love it.  Maybe I will upgrade to an IQ260 in 2 or 3 years, but after that ... not sure.  If they continue with this pointless MP race, than the IQ260 will probably be the highest res I would ever want to go to. 

+1
Logged
rainer viertlböck
architecture photograp

eronald

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 6642
    • My gallery on Instagram
Re: I guess we are going to see more MF CMOS introductions…
« Reply #38 on: January 27, 2014, 11:30:53 pm »

Just use a bad lense and your Moiré will magically disappear :)

Edmund

Oh yes,

There is always moiré on the P45+ at medium apertures if there is pixel level fine detail in the picture. The Sony Alpha 99 (6 my) has some, the Alpha 77 (3.9 my) has virtually none.

Moiré and large pixels go hand in hand, stopping down to f/16 eliminates moiré on the P45+ but it also reduces sharpness. The sample below is a good example of this (left f/8, right f/16). Best way to avoid moiré (and other aliasing artefacts) is to reduce pixel size.

Best regards
Erik


Logged
If you appreciate my blog posts help me by following on https://instagram.com/edmundronald

gerald.d

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 438
Re: I guess we are going to see more MF CMOS introductions…
« Reply #39 on: January 28, 2014, 12:29:09 am »

Any MF camera with an exchangable back. Also technical cameras.

Best regards
Erik


You think Sony would make their own MFDB to put on other manufacturers' cameras?
Logged
Pages: 1 [2] 3 4   Go Up