Scarlet 3k 2/3" DOF

[Jason Murphy]

I'm still a bit amazed by people who find 2/3" or 16mm "unacceptable" because of the DOF. Many great masterpieces of cinema have been made on 16mm/S16, often by independent and avant-garde filmmakers.

But it doesn't stop there.

David Fincher's Zodiac, one of last year's best (and best shot) films, was shot on the VIPER, which is a 2/3" camera; his film The Curious Case of Benjamin Button comes out this Christmas, and was shot on the same camera.

http://www.apple.com/trailers/paramount/thecuriouscaseofbenjaminbutton/

I guarantee you that 99.999% of the people watching either of those movies won't be thinking "this movie would be good if it weren't for the unacceptable depth of field inherent in the originating format."

And those who feel that somehow a shallow depth of field is inherently more cinematic, or necessary for good or even competent narrative filmmaking, might I also note that both Citizen Kane and Rules of the Game, widely hailed as two of the greatest films of all time, were shot almost entirely in deep-focus, with almost everything in the frame in focus.

All that to say, I can't wait to get my hands on the 2/3" Scarlet. A good digital 16mm equivalent camera is a dream come true.

EDIT: Not saying that you're wrong about the S35 3K either, here, brainburst. I don't personally have a need for it, and I don't think it's likely to happen, since it looks like the only Monstro sensor RED is implementing has 6 micron pixels, but your points are definitely valid.

 

[Stuart English]

3K modes

3K modes

But there should be a 3k option in a larger sensor size!

There is - see the specs on the FF35

 

[Craig Ryan]

"So what's not to like or lacking that prevents you shooting documentaries, news inserts, sports, regional commercials and yes of course independent films?"

"S16 is not bad.. good enough for indies."

Did you even read what I wrote?

Higher pixel density on a smaller sensor means:

• Greater noise.
• Less dynamic range (It doesn't matter what processing happens after the fact. A sensor is an analog device with analog limitations.)
• Need for better, more expensive glass. Name one low cost piece of glass that will resolve 3k on a 2/3" sensor. Standard 16mm glass will not cut it. Low cost Video HD glass will not cut it.
• Video type depth of field. Expensive adapters required for 35mm DOF

A larger sensor means:

• Better S/N
• Better Dynamic range
• Ability to use any decent 35mm glass including SLRs
• 35mm DOF without expensive adapters.

Good enough for the past is not good enough for the future!
The Scarlet needs to be unequivocally better than super 16!

Sounds like you're proposing that the $7k s35mm Scarlet should be 3k instead of 5k for the reasons above; but I don't think we can assume a 3k s35mm Scarlet would reduce the price all that much since you're still getting a sensor almost TWICE as large.

Another thing to remember is that you're theoretically not resolving a full 3k with a Bayer pattern sensor, only about 2.2k I recall being the estimate. So in reality, all the glass currently being used on 2/3" 1080p cams should work fine.

 

[brainburst]

No I'm not

No I'm not

Sounds like you're proposing that the $7k s35mm Scarlet should be 3k instead of 5k for the reasons above; but I don't think we can assume a 3k s35mm Scarlet would reduce the price all that much since you're still getting a sensor almost TWICE as large.

The 5k sensor has a larger cell size so the argument does not hold.
The size of the sensor is not as great a factor in price. Pixel density is a greater driving cost.
They could use the same foundry for all the sensors instead of a separate one just for the 2/3" sensor. Same hardware for the camera same everything else but the specific sensor.That would save money.

So in reality, all the glass currently being used on 2/3" 1080p cams should work fine.

You've made my point! That glass is very expensive! Furthermore the cell sizes on those cameras is larger. We'll end up spending enough on glass to almost upgrade to the 5k which can get away with cheaper glass.

 

[Zach Swena]

There is - see the specs on the FF35

Sure the FF35 can take 3k video but why would I pay $12k for that. If I got that model I would take 6k. The point is there are costumers that would benefit from having a 3k S35 available for a lower price.

 

[brainburst]

Yes but...

Yes but...

I'm still a bit amazed by people who find 2/3" or 16mm "unacceptable" because of the DOF.

It's not "Unacceptable" it's just not as good

David Fincher's Zodiac, one of last year's best (and best shot) films, was shot on the VIPER, which is a 2/3" camera; his film The Curious Case of Benjamin Button comes out this Christmas, and was shot on the same camera.

The Viper is a 3 CCD camera with larger cell size.

both Citizen Kane and Rules of the Game, widely hailed as two of the greatest films of all time, were shot almost entirely in deep-focus, with almost everything in the frame in focus.

Deep focus should be a choice not a mandate.

It looks like the only Monstro sensor RED is implementing has 6 micron pixels....

The 5.4 micron size pixels of the 5k camera are good enough. A ~1inch sensor would be fine.

 

[brainburst]

Are you kidding?

Are you kidding?

There is - see the specs on the FF35

Spend almost 5x the amount when it should only cost at most 2x as much for 3k in a larger sized sensor?

 

[Jason Murphy]

It's not "Unacceptable" it's just not as good

The Viper is a 3 CCD camera with larger cell size.

Deep focus should be a choice not a mandate.

The 5.4 micron size pixels of the 5k camera are good enough. A ~1inch sensor would be fine.

Fair points all.

 

[Zoran Milich]

Does anyone on this list know if one can mount fine Leica M glass on the 2/3 Scarlet? With stills I shoot raw both with the 10mp Leica M8 and the 21mp Canon 1Ds Mark II and find the Leica glass does wonders and is just a hair less in quality behind the Canon due to the sensor size.

 

[Jason Badham]

The 5k sensor has a larger cell size so the argument does not hold.
The size of the sensor is not as great a factor in price. Pixel density is a greater driving cost.
They could use the same foundry for all the sensors instead of a separate one just for the 2/3" sensor. Same hardware for the camera same everything else but the specific sensor.That would save money.

The size of the sensor has a great deal to do with the cost, the same as for microprocessors. When Intel and AMD shrink their CPU's (the photomasks used in the "doping" process - when they go from 65nm --> 45nm) the overall costs go down since they can get more chips out of a single wafer. The larger the chip, the more expensive it is since you're paying for the wafer real estate.

If you had a ten inch square wafer (100 sq. inches total real estate) and each inch cost $1 to manufacture, then your wafer costs $100. Now if the chips produced from that wafer were 1 inch square, then each chip would only cost $1. If you shrunk the chips down to 1/2 inch, then the chips would cost 50¢ each. There are other costs like R&D and retooling (and countless other factors no doubt), but that is the general principal.

 

[Craig Ryan]

I'd love a s35mm 3k Scarlet IF the cost actually did lower from the $7k it's currently priced with the 5k res...but I also wouldn't want to lose the 120fps at REDCODE 42 either. There's going to be a trade off somewhere, and I'm sure there is a much more complex reason for the current specs. I seriously doubt a 3k s35mm Scarlet would be priced under $3k though. I hope I'm wrong of course! Afterall, these specs could change at any time...hopefully it is possible for an in between s35mm 3k brain in the $3-5k range. If not, I'm already set with the 2/3" version.

 

[brainburst]

The size of the sensor has a great deal to do with the cost, the same as for microprocessors. When Intel and AMD shrink their CPU's (the photomasks used in the "doping" process - when they go from 65nm --> 45nm) the overall costs go down since they can get more chips out of a single wafer.
The larger the chip, the more expensive it is since you're paying for the wafer real estate.

This is true, but must be weighed against development costs of a totally different form factor for the 2/3 sensor. Also I think almost all of Red's prospective customers would be willing to pay the extra cost plus profit for a better sensor. All they have to do is ask...

 

[REDneck]

This is true, but must be weighed against development costs of a totally different form factor for the 2/3 sensor. Also I think almost all of Red's prospective customers would be willing to pay the extra cost plus profit for a better sensor. All they have to do is ask...

you have a great point, and i had the same thought myself. why 3.2 microns? why not pull the same amount of data from bigger cells? isn't there a premium paid for cramming everything onto a small sensor? aren't there big advantages to a bigger (but no denser and no more data-generating) sensor?

 

[David Rasberry]

you have a great point, and i had the same thought myself. why 3.2 microns? why not pull the same amount of data from bigger cells? isn't there a premium paid for cramming everything onto a small sensor? aren't there big advantages to a bigger (but no denser and no more data-generating) sensor?

I'm not seeing where reducing resolution on S35 would lower cost at all. You might pick up higher frame rate on Scarlet's processor, but neither the cost of fab nor processing would change much. S35 at $7k is a huge step in price performance over currently available technology and the same is true of 2/3" at $2500. Bigger sensor and cheap glass won't necessarily get you better images. Nobody else is making anything larger than 1/2" in Scarlet price range that is truly capable of shooting large screen/broadcast quality HD. This is a huge upgrade in what is available to low budget indies.

Video from a Canon 5d or NikonD90 still camera doesn't look very good at screen sizes bigger than web video on a desk top computer monitor and neither can be considered a very good acquisition device for motion pictures.

Motion picture cameras have always been more expensive than their still frame counterparts. If 35mm DOF is that important then spend the extra bucks on the S35 and stick a $400 still camera lens on it.
I would rather have a 2/3" Scarlet with really good cine glass on it than s35 saddled with the compromise of cheap still lenses.

 

[Eki Halkka]

isn't there a premium paid for cramming everything onto a small sensor? aren't there big advantages to a bigger (but no denser and no more data-generating) sensor?

If you look at what's out there, in both still and video cameras, high pixel density doesn't seem to be expensive.

The small and cheap pocket cameras have very high pixel counts on small sensors. In the other hand, there's no cheap cameras with big sensors, regardless of pixel count.

The smaller the chip the cheaper is a good rule of thumb.

 

[Tarek S. Kandil]

And i'd like to have a Ferrari that costs the same as a Nissan.

We forget quickly 04 and before that time. This is in fact, a ferrari priced like a nissan.

 

[REDneck]

I'm not seeing where reducing resolution on S35 would lower cost at all. You might pick up higher frame rate on Scarlet's processor, but neither the cost of fab nor processing would change much. S35 at $7k is a huge step in price performance over currently available technology and the same is true of 2/3" at $2500. Bigger sensor and cheap glass won't necessarily get you better images. Nobody else is making anything larger than 1/2" in Scarlet price range that is truly capable of shooting large screen/broadcast quality HD. This is a huge upgrade in what is available to low budget indies.

Video from a Canon 5d or NikonD90 still camera doesn't look very good at screen sizes bigger than web video on a desk top computer monitor and neither can be considered a very good acquisition device for motion pictures.

Motion picture cameras have always been more expensive than their still frame counterparts. If 35mm DOF is that important then spend the extra bucks on the S35 and stick a $400 still camera lens on it.
I would rather have a 2/3" Scarlet with really good cine glass on it than s35 saddled with the compromise of cheap still lenses.

i'm more concerned with the low-light performance

 

[brainburst]

You are not understanding at all

You are not understanding at all

I would rather have a 2/3" Scarlet with really good cine glass on it than s35 saddled with the compromise of cheap still lenses.

You will need really good cine glass to achieve worse results than cost effective glass coupled with a larger sensor.

And it doesn't have to be s35 sized!

Everyone here thinks that the 2/3 sensor is going to be like a baby Red1. It can't be. It will be noisier. It will have worse low light performance. It will require better glass. The photosites are 40% smaller. Unless they've developed a totally different kind of sensor (They haven't said this) it will be a totally different and inferior camera.

 

[brainburst]

I'm not seeing where reducing resolution on S35 would lower cost at all. You might pick up higher frame rate on Scarlet's processor, but neither the cost of fab nor processing would change much.

Of course the cost of fabrication would be reduced. Only one maybe 2 similar form factors to support. The 2/3" is totally different from the others.
In terms of cost savings you can pull more 3k sensors from the same wafer than 5k sensors.

 

[combatentropy]

If 2/3" is only 2.5 stops away from S35 DOF, the only look out of reach would be S35 at 2.8 or below, right? Because if I get a fast lens, I can shoot 2/3" F1.4, and it should look like S35 f2.8 or f4.

And if the look I'm trying to go for is the movie look I've grown up on, what f-stops are movies typically shot on?