View Full Version : S.I. Interesting things ahead.


Bob Hart
March 7th, 2013, 08:53 AM
Ari Presler has some interesting things relating to SI2K ahead. No spoilers yet but when I can open my big mouth here to spill the beans, I shall.

Sareesh Sudhakaran
March 11th, 2013, 04:46 AM
Can I guess (or hope)? SI-4K with a Super35 mm sensor shooting RAW for $9,999?

Brian Drysdale
March 11th, 2013, 06:41 AM
A Super 16 sensor would be more unique with perhaps 2.7k, to give a good sensitivity/resolution balance. There are a lot of high quality Super 16 lenses out there and you don't lose the compactness.

Bob Hart
March 15th, 2013, 11:31 AM
Sareesh.


I don't think you will be seeing a SI4K Given that everybody else and their dog and cat are already doing one, it would not be much of a business model for a smaller scale builder.

The Kinefinity cam which shoots a S35 frame and downsamples to 2K and codes to DNG and cineform have let that horse go already at an asian price point. Their sensor is 4K nd rthey say they intend to progress the camera to 4K recording. At the higher end industrial level, P+S have gone it alone and have it covered with the X35 which was to go from 1920 to 2K during its product life.


Bryan.

Out of courtesy to Ari at SI, I should hold my tongue for the time being. Suffice to say, he is not sitting on his hands.

Ari Presler
March 15th, 2013, 12:26 PM
There are way too many camera products announced before they are ready for prime time. I will tell you we are "looking" at a next generation SI-2K product targeted for the optical format segment that Brian is referencing....Super 16 and compact.

Would it be appropriate to call the camera an SI-2K, if its resolution is higher than 2K?

Anyone interested in having a live streaming channel directly out of their camera?

Phil Stone
March 16th, 2013, 03:31 AM
I loved using the SI2K over the last two years, we have just replaced it with a couple of cameras that are sadly twice the size. There is now a huge gap in the market for something that can actually replace the size of this camera & have the same latitude as some of the newer cameras on the market. If they come out with something new at the same size as the mini head with 13+ stops I would buy it direct for sure. SDI out makes a lot of sense also.

Bob Hart
March 16th, 2013, 08:50 AM
Compared to some other owner-operators of the system, my personal preference has been to stick with the P+S recorder with camera docked. Except for a glitch which shook down to a locally replaced HDD, this combination has been robust.

My past wishlist has included a 35mm sensor 2K retrofit kit or exhange deal for the S16 Mini head insides and upgrade of motherboard and backpane kit to i7 with two Lemo LAN ports to enable on-body side-by-side 3D or beam rig on the front via a dummy Mini body insert attachment.

With the trend for Lego style modularity over fully enclosed systems, that is now unlikely to happen. There would be need to be too much expensive alteration of the existing large P+S body system.

High volume asia-based builders are closing in on the 35mm digital cinema space as evidenced by RED Corp's recent litigation. There is a squeeze also coming from the SinaCam style products which are small sensor 1920x1080 and the most recent adventures by GoPro.

Compact and agile 2D and 3D digital cinema seems to be the remaining niche for the SI2K. Within that scope, improvements to or replacement of the Mini head may be commercially viable.

I wonder what sort of intererst there is among SI2K owner-operators as well as newcomers who might be drawn to the system.

Sareesh Sudhakaran
March 17th, 2013, 05:07 AM
There are way too many camera products announced before they are ready for prime time. I will tell you we are "looking" at a next generation SI-2K product targeted for the optical format segment that Brian is referencing....Super 16 and compact.

Would it be appropriate to call the camera an SI-2K, if its resolution is higher than 2K?

SI-3K surely sounds cool. There's no one else with a 3K camera on the market, so it definitely stands out and still adds a 'version' element as well!


Anyone interested in having a live streaming channel directly out of their camera?

Ethernet? The mini was ultra-cool in this regard, but I'm not sure anymore. A lot of processing (encoding, conforming, 'correcting', etc.) needs to take place prior to a live stream - too many devices, formats and business models.

If it was a simple live telecast, there's HD-SDI, with an 'established' infrastructure and workflow.

Possibly naive question: Are you competing against the BMCC? Small form factor, 2.7K resolution, 3G-SDI out, and recording RAW? For some reason, when the BMCC first appeared, most people compared it to DSLRs; but it reminded me of the SI-2K mini.

Bob Hart
March 17th, 2013, 09:39 AM
As well as BMCC, there are Ikonoskop Digital A-Cam, Digital Bolex, Kinor, Kinefinity aspiriing to the same digital S16 sportsground for SI to go up against.

I assume Ari's questions are based on what is doable now or not too far down the track.

I would also not write off ethernet just yet.

Would the existing SI2K owner-operator cohort want third generation improvements integrated within the existing P+S and PC notebook platforms, something new and self-contained, or something which is new and more capable but has some legacy compatablity with the existing platforms?

Furthur feedback from more SI2K owner-operators may be helpful.

Brian Drysdale
March 17th, 2013, 11:05 AM
There was a lot of interest in the 2/3" Scarlet, but I'd assume the price point of a next generation SI camera would be in professional price bracket, rather than the BMCC range or even the fixed zoom 2/3" Scarlet.

Regarding the RED patent case, the divide seems to be compressed RAW larger than 2K, although with Cineform RAW being around at the same time, if not slightly before REDCode, perhaps there could be wiggle room. However, 4k seems to be the current magic number for RED.

Bob Hart
March 18th, 2013, 09:56 AM
QUOTE:
--------------------------------------------------------------------------------
There was a lot of interest in the 2/3" Scarlet, but I'd assume the price point of a next generation SI camera would be in professional price bracket, rather than the BMCC range or even the fixed zoom 2/3" Scarlet.

Regarding the RED patent case, the divide seems to be compressed RAW larger than 2K, although with Cineform RAW being around at the same time, if not slightly before REDCode, perhaps there could be wiggle room. However, 4k seems to be the current magic number for RED.
---------------------------------------------------------------------------------

My interpretation of Ari's comment is that SI will not make promises it cannot deliver on.

Whether SI can deliver at a competitive price point with the other players one can only guess. I think their selling point can draw on the demonstrated reliability and robustness of the Mini camera head and mount system itself. Anecdotally at least, I have not heard of one ever laying down. For robustness and reliablity, existing SI2K owner-operators might pay a small premium to remain in a familiar environment versus migrating to a new system and abandoning their SI2K investment entirely.

My ideal would be for two variants of a new S16 "Mini" head, one which embraces all the improvements the other S16 Digital Cinema aspirants are spruiking but not necessarily yet delivering.

This variant would be fully functional and complementary to the latest capability of computers. In keeping with the furthur miniaturisation trended by competitors, the new Mini head would need to be similar to them in its physical footprint, especially given its niche in POV and 3D applications.

The other variant would be legacy and provide sharpness and dynamic range improvements to the extent that the existing P+S recorder body and older laptop installations can cope with. Desirably this variant would be built into a Mini-style body as a direct plug-in replacment for the existing Mini head.

Alternatively - owner-operator retrofit kits for the existing Mini bodies might be released to contain end-user costs, - existing Mini heads might be sent in for infit of the new camera core on an exchange basis - an adaptor dock to allow the new camera body form to dock into the existing Mini dock in the P+S body.

I have camera body Serial 1019. Assuming that bodies have started at Serial 0001, then potentially many of those owner-operators might be interested in a new compatable Mini head. A more valid indicator of potential uptake by existing SI2K owner-operators might be the degree to which the new software DVR2/SI3D has been adopted by non-3D users.

Catering for a new market and still providing legacy compatability could be a messy task and may not be a viable business plan. We shall have to wait and see.

Bob Hart
March 18th, 2013, 10:10 AM
The RED case might be as much about retaining the claim they pegged out as the underdog everybody "roots" for and which they are still mining. For how long they can do so is moot for RED is no longer a seller of a few hundred units any more. Other patent holders including Ari Presler and David Newman are acknowledged in the RED patent.

It might be suggested then that RED by pursuing its intellectual property rights may be also collaterally protecting the intellectual property rights of others they have drawn from by licencing deals.

Apparently if intellectual property is not affirmatively defended, a case is effectively constructed by the holder for the infringers. This is apparently a driver of the action as well.

Ari Presler
March 18th, 2013, 11:49 AM
The SI-2K platform has always been about image quality, workflow and small form-factor. We would want to solve problems the current crop of products have not yet been able to do or solve them significantly better. It has to be application and use driven.

One application/use that is frequently done with SI-2K is POV and body worn applications with nominal or no compression.

The most common request for improved functionality has been increased sensitivity/dynamic range, to better match the current pool of cinema quality cameras. This would be our first and foremost objective.

The drawback of going higher resolution within a fixed optical format, is it requires the use of smaller pixels to fit more of them in the same square mm area. The question remains...more pixels or more sensitivity?

Also, keep optical format within 2/3"/Digital-16 (11-12mm diag) for use with B4 and small C-Mount or IMS primes or go for larger Super 16.

Finally, if target is for use with smaller C. Should we also go for a form factor even small than the current Mini. (Note: These form factors immediately differentiate it from the other cameras on the market.)

Chris Hurd
March 18th, 2013, 11:57 AM
It's good to hear from you, Ari -- see you soon at NAB (thanks for starting this thread, Bob)!

Brian Drysdale
March 18th, 2013, 12:04 PM
Also, keep optical format within 2/3"/Digital-16 (11-12mm diag) for use with B4 and small C-Mount or IMS primes or go for larger Super 16.


I'd tend to go for the Super 16 sized sensor. You can use various optical devices to allow B4 mount lenses to be used. The extra area over the smaller 2/3" allows room to manoeuvre with larger pixels.

Edit: Perhaps a 2.5k sensor might be the best balance.

Sareesh Sudhakaran
March 19th, 2013, 06:30 AM
... The question remains...more pixels or more sensitivity?

Also, keep optical format within 2/3"/Digital-16 (11-12mm diag) for use with B4 and small C-Mount or IMS primes or go for larger Super 16.

Finally, if target is for use with smaller C. Should we also go for a form factor even small than the current Mini. (Note: These form factors immediately differentiate it from the other cameras on the market.)

I'll take sensitivity over resolution any day. 15 stops of DR in 1080p at ISO 800 - who has that even today?

I can see how many who own and use 16mm and 2/3" lenses will find a small sensor interesting, but where I'm at, these lenses are not readily available. It's one thing to hear about a decent lens that someone uses in another continent. But it's extremely frustrating to not have access to the same.

Why not develop your own mount (like the Sony FZ/E) and then supply adapters for any kind of lens? E.g., a sensor like the one used on the BMCC has extreme resolution demands that not all lenses can meet.

Rohan Dadswell
March 19th, 2013, 04:25 PM
The S.I. already has an interchangeable lens mount system - PL, Nikon, Canon FD, B4 & C mount (and probably others) so you can fit just about any lens to the system.
A micro 4/3 mount would be great so you could use the Voigtländer f0.95 range.
Super 16 size chip would work fine.

Bob Hart
March 20th, 2013, 01:41 AM
With the original Mini head, for side-by-side POV 3D, the IMS-Mount system limits the narrowness of the interaxial distance to about 72mm if my feeble memory serves. If you take the IMS ring off and use the disk-shaped C-mount which fastens by four screws, I think you can get it down to 66mm.

I think the shortest flange to focal of the IMS adaptors available is 17.5mm or thereabouts for the C-Mount. Micro 4/3rds is 19.5mm. A Micro 4/3rds adaptor might be doable in IMS but the lens receiver as with the C-Mount adaptor would be dished which may make the lens aperture control awkward to work with if it is in the traditional rear position of stills lenses.

There is as little as 2mm available workspace behind the Micro 4/3rds flange for bayonet lugs and lens barrel structure sticking through, depending on the actual diameter of the 4/3rds mount which I do not know.

Closer than 17mm to the sensor, the camera body becomes coned in towards the sensor panel, so the rearwards intrusion of some wide-angle Micro 4/3rds lenses might impinge on the structure of the camera.

The diameter of the lens rear body would need to be less than 50mm to allow internal clearance within the shoulder chamber of the IMS-Mount less a wall structure of 1mm thickness in the IMS-Micro 4/3rds adaptor.

There is a SI camera variant called the Nano. It is a rehoused version in a small case with no sharp corners by Radiant Images created specially for bodycam POV work. - Think of the film "End Of Watch".

It is fitted with only a C-mount. I understand one can have tighter 3D side-by-side with physical lens size being the practical limit. Someone may like to correct me here.

Given that POV and 3D as compact agile and bodycam have become SI's niche, my guess is the new tech will appear in that form.

The larger boxey Mini head with IMS-Mount on rails tethered to a laptop or homespun recorder seems muchly used for the same work as the full P+S body combination. It is up against RED's Scarlett and all the other S16 wannabes, so may not be as viable a commercial development to kick off from as the Nano style.

There were three versions of the Mini head, one without LEDs, one with three LEDs across the rear top, one with three LEDs down the rear right.

My imagining is that unless the LEDs were fed by wires, each version had a unique PCB with LEDs hardmounted on it in the LEDs versions. I would expect that for my wishlist Mini head budget retrofit kit, the LEDs might be eliminated or hanging on long tails, with a prewired 12pin Lemo panel socket hanging on a tail.

An entire front case panel for the camera might be redesiged to accept a different fixing pattern for screwholes etc., for a different PCB layout and to preserve precision between the sensor and the focal plane. The average kitchen table mechanic fitting it all together is not always going to get it right, so an assembly including front case panel would be the safest bet for a kit.

So the kit might be comprised of a replacement front panel with threaded fixing holes for moving the IMS-Mount across from the old Mini, already fitted with sensor PCBs and revised OPLF/IR panel.

The remaining rear case may be the least of any added expense to be saved by an end-user retrofit kit although it is a complex piece of machining, not cheap to remake.

Shimming the IMS-Mount to correct collimation would be the end-user's responsibililty.

There are good reasons for not issuing a user's retrofit kit. SI's control over quality assurance would end at the Fedex gate and end-users might badmouth the product if they can't get their end of the deal right due to hamfisted incompetence.

My guess then, if there is to be a Mini head retrofit suited for the P+S body, it will be a fully built-up piece as an entire replacement. That would be the only means by which SI could control the quality assurance up to the point the end-user's DNA is smeared all over the product and the power goes on.


As for S16 versus 2/3" my personal preference is for S16 as it enables me to extract a little more apparent sharpness from my lenses across a slightly wider sensor. Except for the Voigtlander f0.95 some of the Micro 4/3rds lenses may not be as sharp as the S16mm T1.3 lenses which are sweetest at about T3 or thereabouts.


Rohan.

Did you ever recover your stolen gear in the end.

Lior Molcho
March 28th, 2013, 12:42 PM
Hi Ari!

As a long time supporter and customer, and having hands on experience with the new F-series from sony, and other vendors, I can safely say that 5 years ago (!) SI-2k was so far ahead of the game that only now other manufacturers are trying to accomplish their workflow, not to mention the ACES workflow from the ASC and the Academy which are still working on full integration of RAW/metadata workflow.

I still shoot on the SI every week, but it's getting harder to "sell" it to customers, not because of RED hype but because 4K (as silly as it is) is going to be the de-facto standard or at least 3K.
clients like buzzwords - that's the fact of life I see, so my personal recommendations would be:

1. Dont touch the DVR software! it's beyond perfect, it's cutting edge and amazing!

2. 3k resolution - its a must have for clients, regardless of what us techies know is best.
eventually were all selling gear to the producers of a show

3. Offer a full package camera and sensor for around the same price point as sony (15-20K)
pro DP's dont care about gear, about rigs etc. they just want a camera that shows a picture they can light.

4.Native ISO at 800. it's almost an industry standard.

5.Get on board the ASC ACES workflow and integrate ACES inside the workflow, get the cinematographers support, they still talk about the SI-2k there :)

6.Shooting in prorez would seal the deal - most times working with the Alexa, people just use it's prorez,
again, its just the way people LIKE to work, not necessarily the smartest or what I would do but is the working reality I see in LA

7.And last but not least, demo the hell out of the camera - CANON finances so many short movies to demo new equipment, Panasonic now does that, they pay high-end, ASC members to shoot short films to demo their cameras(not young filmmakers) and it helps spread the word in "the industry"

Love my SI-2k! got it tatood on my bum :)

Lior