ARRI's brand new 4K S35 sensor is mere weeks away

[Andrew Reid]

FULL specs, brochure images and 17 stops dynamic range sample shot

https://www.eoshd.com/news/arri-alexa-35-full-specs-brochure-images-and-17-stops-dynamic-range-sample-shot/

[d shay]

Is there a link to view the brochure?

[ND64]

I don't like to say that 17+ DR is a marketing lie, but even state of the art CMOS sensors can't reach that level. 

[IronFilm]

11 hours ago, Andrew Reid said:

Built in ND filter and a smaller RED / ALEXA MINI form factor.

From what I've read (and what we can see too in the pictures), it is bigger than a Mini LF (which is already a bigger body than a Mini!! And the Mini itself is a very big body compared to Z Cam / BMD Pocket / S1H / FS5 / etc etc) but smaller than an XT (err... a disturbing big camera to be comparing it again!). 

I wouldn't be surprised if within 5yrs they release an even smaller body than the one they're releasing next month. (and/or ARRI's own version of the Rialto system, like they did before with ARRI's M)

[IronFilm]

10 hours ago, Django said:

wow.. was not expecting that kind of form factor!

 

I am not surprised, the Mini is their most popular camera ever. 

And the Mini LF is the hot camera for them right now which everyone was wanting.

10 hours ago, Django said:

built-in NDs also rather surprising..

No surprise here, would be more surprising if it didn't. 

Both the Mini and Mini LF has built in NDs. 

(so too does the AMIRA, but that's not so relevant. Oh, and the ALEXA LF "kinda" has internal NDs. They technically have them, but not in the sense you think of them: https://www.bhphotovideo.com/c/product/1389604-REG/arri_k4_0015600_alexa_lf_fsnd_2_1.html )
 

10 hours ago, MrSMW said:

I might get a pair to shoot weddings with. What's the AF like?

(Not really and I'm not interested in proper cine cameras because I'm not a proper filmmaker)

I know you're joking, but there are people shooting weddings with REDs all the time. 

Give it another ten years and we'll see people shooting wedding with the ARRI S35K 4K sensor as well. 

9 hours ago, MrSMW said:

I guess, as the saying goes, if I have to ask, then I can't afford it, but what is the cost going to be for one of these?

50k+?

I'd be surprised. That's an extreme low end guess, I'd say more likely half way in between there and double the price. (probably very easily close to double your guess once you add in accessories / media / batteries)

9 hours ago, Django said:

Not many folks in the world can afford this, aside from A-list actor/directors like Ben Affleck in the photo above.

But usually no one person buys a brand new ARRI. It's either studios, production companies or rental houses.

Likely in a city such as Auckland (largest city in the country, but still a small city by international standards with one and half million people) then all the rental houses of any substantial size will get it eventually. But you could count on your fingers how many DoPs will own one, and probably a similar number of production houses would own it. 

 

9 hours ago, Django said:

Rest assured pretty safe to say nobody here is in the market so its mostly just an industry look at what's happening on the high-end.

Sure, it is likely zero EOSHD regulars will buy one, but probably quite a few of us will end up working with it at some point in time or another. I'm 100% certain I'll end up working with the new ARRI within the year. 

  

5 hours ago, majoraxis said:

Very impressed that it is not 8k or 6k

That's standard operation procedure for ARRI, chase image quality and not resolution. 

  

9 hours ago, Django said:

Again, its interesting to see ARRI took form them the built-in ND's. surprised they didn't go Sony's Vari E-ND but maybe there is a patent on that one.

Very likely the image quality hit of an eND is why they didn't use them. The people at that level can be very picky about that. 

[webrunner5]

3 hours ago, Eric Calabros said:

I don't like to say that 17+ DR is a marketing lie, but even state of the art CMOS sensors can't reach that level. 

Hell Red has been lying about 20 stops for years on their 8K stuff. I can believe the 17 but 20, hmm they might have made a breakthrough. Arri has been top dog on DR for years.

[MurtlandPhoto]

14 hours ago, Eric Calabros said:

I don't like to say that 17+ DR is a marketing lie, but even state of the art CMOS sensors can't reach that level. 

This is statement is true of basically every other camera company except ARRI.

[ND64]

13 minutes ago, MurtlandPhoto said:

This is statement is true of basically every other camera company except ARRI

There is no exception. ARRI didn't reinvent the CMOS sensor, they just achieved lower noise floor with double readout. State of the art CMOS is limited by 15 stop, which 13 is useable. These are so clean that they're only limited by shot noise. In other words, the noise problem is already solved; so to increase DR you need to work on the other side: saturation level. Unfortunately its so hard to increase the saturation capacity that Nikon celebrated when they achieved 1/4 of a stop more capacity and called it ISO 64, which in reality was only ISO 80.

[MurtlandPhoto]

13 minutes ago, Eric Calabros said:

There is no exception. ARRI didn't reinvent the CMOS sensor, they just achieved lower noise floor with double readout. State of the art CMOS is limited by 15 stop, which 13 is useable. These are so clean that they're only limited by shot noise. In other words, the noise problem is already solved; so to increase DR you need to work on the other side: saturation level. Unfortunately its so hard to increase the saturation capacity that Nikon celebrated when they achieved 1/4 of a stop more capacity and called it ISO 64, which in reality was only ISO 80.

Perhaps I should rephrase. I think there's a true, scientific definition of DR and then there's the marketing definition of DR. The latter has always been squishy for most manufacturers with some taking more liberty than others. My understanding is that ARRI is much more conservative and "real" with their marketing definition of DR. I believe they claim the original Alexa sensor as 15 stops. So, if they say this new sensor has 17 stops I take them at their word that it is two stops better. 

[mrtreve]

I reckon this will be priced a fair amount higher than the Mini LF if it's the full version of this camera (i.e. it replaces the Alexa SXT, not just the Mini) and there demand will outstrip supply by quite a lot. Arri can only build so many cameras plus there's the chip shortage to contend with still.

Rental companies will be looking to replace their fleets of Minis (and SXTs) and they have deeper pockets than your typical owner operator, so I'm guessing the strategy would be to price-out most of this market to manage the demand somewhat.

As usual, it'll probably be hard to even rent one when they come out as they'll all go to high end features/tv/commercials.

At least the LUT salesmen have a new colour science to try and match the Pocket 4K to! Boomtime for the LUTboys.

[Django]

15 hours ago, IronFilm said:

I am not surprised, the Mini is their most popular camera ever. 

And the Mini LF is the hot camera for them right now which everyone was wanting.

No surprise here, would be more surprising if it didn't. 

Both the Mini and Mini LF has built in NDs. 

Ok didn't realise this was in the Mini category.

15 hours ago, IronFilm said:

Likely in a city such as Auckland (largest city in the country, but still a small city by international standards with one and half million people) then all the rental houses of any substantial size will get it eventually. But you could count on your fingers how many DoPs will own one, and probably a similar number of production houses would own it. 

Yeah depends where you're at.

Here in Paris, France we've got pretty big film studios & production houses. ARRI also has offices.

I literally stumble on sets in the city daily with ARRI cams & lights.

My friends production house has 3 AMIRAS. they specialise in high-end Multicam shoots for TV & events.

15 hours ago, IronFilm said:

Sure, it is likely zero EOSHD regulars will buy one, but probably quite a few of us will end up working with it at some point in time or another. I'm 100% certain I'll end up working with the new ARRI within the year. 

Sounds good, let us know when that happens!

 

Quote

Very likely the image quality hit of an eND is why they didn't use them. The people at that level can be very picky about that. 

Oh I know that but for some reason I thought eND was a better (less polluting) solution than standard built-in NDs? could be wrong though..

[Caleb Genheimer]

1 hour ago, MurtlandPhoto said:

Perhaps I should rephrase. I think there's a true, scientific definition of DR and then there's the marketing definition of DR. The latter has always been squishy for most manufacturers with some taking more liberty than others. My understanding is that ARRI is much more conservative and "real" with their marketing definition of DR. I believe they claim the original Alexa sensor as 15 stops. So, if they say this new sensor has 17 stops I take them at their word that it is two stops better. 

This is right on. It doesn’t *really* matter what stop number ARRI say the DR of their new sensor is, they’re very consistent and technical with their numbers. They rate it having two stops more than their previous sensor, so however you’d rate the original sensor, this one will have two stops on it. It sounds like a decent amount of that is aimed at cleaner shadows, with a touch of bonus in the highlights, which makes sense. ARRI is the poster child for highlight handling, everyone points at them and says “just do roll-off like that, it’s perfect,” so it is logical they’d largely maintain that behavior. 
 

I also understand the talk of ARRI bumbling the sensor update, but that’s not gonna happen either. ARRI will have put these in the hands of diehard users and ironed out any minor gripes behind the scenes long before release. ARRI users aren’t afraid to say when they don’t like a camera, and ARRI has the very top cinematographers essentially as beta testers. I’m very interested to see where the future of DR in theaters goes after this new camera is out in the wild. Should be a required bit of kit on DolbyVision sets IMO.

This is 100% square one for an eventual refresh of their entire lineup though. Just like LG cut multiple size TVs from one “master glass,” I’m assuming ARRI can make varied sizes of this new sensor architecture for a new LF, and even stitch them for a new ALEXA-65. I also don’t doubt that they’ll take their sweet time. 
 

The potential update I’d be most excited to see is the AMIRA though. It really is such a cool iteration, the preset toggles enable you to work VERY fast, and it shoulders flawlessly with minimal rigging. The only things it’s awkward for are gimbals (due to length,) and anamorphic due to the 16:9 sensor. I hate gimbals, so if they update it with this new sensor without chopping down the height, I’d be VERY tempted.

[ND64]

3 hours ago, Caleb Genheimer said:

It doesn’t *really* matter what stop number ARRI say the DR of their new sensor is, they’re very consistent and technical with their numbers. They rate it having two stops more than their previous sensor, so however you’d rate the original sensor, this one will have two stops on it.

Seems you think reputation trumps physics. 

[webrunner5]

I seriously doubt Arri is out to fool people. But I admit 2 extra stops is a LOT. Usually, lucky to get a half of a stop on a new model.

[mercer]

Arri should have just listed the DR as having a BUNCH. Or BETTER than any other camera. Maybe INSANE highlight rolloff and DANG you can really see into the shadows and the window is perfectly exposed too.

[Caleb Genheimer]

2 hours ago, Eric Calabros said:

Seems you think reputation trumps physics. 

No, I don’t. You have no physics evidence to refute ARRI’s claim. They do not use a typical CMOS sensor configuration, they use dual gain architecture at high bit depth to arrive at a higher dynamic range than common CMOS sensors, and there absolutely is real physics behind how they do it, because their current sensors ALREADY do it. There’s no secrets (although there are patents protecting it,) and I think it is not only reasonable, but obvious, to conclude that in the many years since their original sensor design, there are improvements in sensor technology that now allow them to improve by a couple stops.

[Caleb Genheimer]

2 hours ago, webrunner5 said:

I seriously doubt Arri is out to fool people. But I admit 2 extra stops is a LOT. Usually, lucky to get a half of a stop on a new model.

“Usually” (other mfrs) new models have new sensor designs, and come out every year or two. ARRI has gone a very long time between updating sensors, I’d expect larger improvements in these rare instances when they’re making core tech upgrades. Most of it being in the shadows makes it very likely to be true. ARRI is heavily biased towards highlight retention. 
 

It could also be though of this way: ARRI uses dual gain circuitry, so ~1 stop of improvement in the underlying sensor technology (CMOS) potentially allows them to “double down” on that (via dual gain,) resulting in an extra 2 stops. Heck, we’re 12 years out from their original ALEV sensor, wafer and chip tech has come a LONG way in that decade. All else being equal, sensors now produce cleaner images even if the DR is very similar. I’m actually super curious to see if any prominent folks will feel this new sensor is “too clean,” many seem to love the ALEXAs for their texture, even if it is subtle. 

[webrunner5]

Yeah, but they are hinting at 20 stops. That would be wonderful if it is that but nobody else is even close to that even today. Sony makes all kinds of sensors; I would think if someone could make something with 20 stops it would be them by now. DR is a Huge selling point these days. Cine camera users don't care about AF, its color and DR.

[HockeyFan12]

I have a very limited understanding of sensor design, but highlight dynamic range should be dictated by full well capacity.

Each photosite is like a bucket. It can only fill with so many photons before it overflows. The reason most of these current cameras are 500-800 base ISO is because they need to underexpose and dig into the shadows to distribute DR since the photosites are only so big and can only hold so much light before being saturated.

I suspect when the S1H shoots 640 ISO in V Log, the sensor is reading out the same signal as 100 ISO in another picture style. It's just digitally redistributing the dynamic range into the highlights.

The Alexa has about half as many pixels per square inch as the C200 (4K S35), S1H (4K in S35), Venice (4K in S35), etc. 

Its photosites are about half as big.

(2880X1620 = 4665600 photosites; 4096*2304 = 9437184 photosites)

So they should hold twice as much light (one stop) before overflowing.

Unsurprisingly, the Alexa has about a stop more highlight dynamic range than its competitors at base ISO.

The Alexa35, however, has pixels that are about the same size as the C200, S1H, Venice, etc. 

But it has 1.5 stops MORE highlight dynamic range than the Alexa Classic. So that's 2.5 stops more than the competition.

Maybe the base ISO of the Alexa35 will be 1600 to account for some of that (it underexposes and pushes an additional stop)? And the sensor design has physical improvements that account for the rest? Regardless, this puts the camera on another level from everything else on the market. But if its base ISO is 1600 that might explain it to an extent. Or maybe I misunderstand this and there are other factors like voltage involved? Regardless, this camera has already far surpassed my expectations. 

[TomTheDP]

I don't see the Mini price falling that much. They are going for 30-35k used while this bad boy is gonna be 70k. The OG Alexa mini is the smallest ARRI camera and that makes it still desirable. Just my opinion, though I'd be overjoyed if it dropped to 15k'ish LOL