I started a bit of a script for a partial tutorial on the SI2K camera-recorder a while back but became distracted and yes, a little bit lazy.

I think most owners would have nutted out what they need to know by now about the feature set and there is likely nothing much else for them to need to know. .

Serious mainstream production as I observe it, seems to have sidelined the traditional movie camera style self contained camera-recorder and gone for the mini head on a cable to laptop computers or dedicated LAN recorder systems. Maybe my observations are wrong as I live in a cloistered part of the world called Perth, Western Australia, so correct me vigorously. ( but don't dopeslap too hard ).

My presentation was going to be more focussed on in-field how-to's, some gotchas and personal preferences as to the configuration, rather than a fullform tutorial. That is for the manufacturers and vendors to provide.

Before I resume putting any more effort into this thing, any suggestions as to the worth or otherwise of continuing on would be appreciated.


Like a fool, I forgot to copy and paste that much of the rough draft that I have prepared. I had proposed to post the info as short video clips rather than one huge diatribe which will send people to sleep.

ROUGH DRAFT THUS FAR. ANY ADDITIONS OR ALERTS AS TO MY MINSINFOMRING PEOPLE ARE MOST WELCOME.

UNAUTHORISED SI2K VIDEO TUTORIAL. - ROUGH VOICE-ONLY SCRIPT.


You gotta be asking, who is this guy. Certainly does not have the charm and charisma of Philip Bloom or Tom Guilmette or the wealth of experience of Charles Papert and Rusty Geller.

What you see is what you will get. Spoiler warning. I will not be pretty. I do not sell secondhand horses or used cars, so that is something in my favour.

Most of you who view this clip will likely already be users or fans of the SI2K camera system and will know of its two iterations, a self-contained file-based camera-recorder and remote tethered camera configuration using a computer, which has become generically known as the Slumdog rig.

My focus will be on the camera-recorder.


HISTORY BACK.

The SI2K camera shares its genesis with the RED camera system in that both were gestated in part on the dvinfo.net web forum.

Both were born out of frustration with what was perceived as a loose cartel of self-interested videocamera manufacturers grudgingly releasing product improvements well short of what was possible in a mass production environment in order to protect their higher-end product lines.

The SI2K grew out of a hypothesised concept of mounting a 2/3" sensor in an existing 16mm film camera body, gutted of all its mechanicals and containing the electronic hardware to manage the image. The recording system would be based as much as was possible on store bought components, would be file based and recorded to a hard disk drive.

What follows in this presentation are my personal preferences and experiences, not endorsed by the manufacturers and vendors of the SI2K system. Viewers are cautioned that they should conduct their own researches and should not accept any of my observations without challenge.


MY SETUP.

So here is the camera-recorder body, in this case P and S Technik's ex-demonstrator, well used as there is some skin and hair missing here and there.

There is the modified CTF840-SH touch-screen controller on Noga Arm, OLED side-finder, D-sub audio breakout cable, CRU Dataport25 digital magazine.

My own added choices are, Anton Bauer Titan power supply/battery charger, an accessory battery pack modified with a 4pin XLR fitting, and ENG/EFP style quick-release mount, in this instance a Panasonic TM-700.

I have substituted a lighter VGA cable for the original as it is conveniently more flexible and places less mechanical strain on the socket in the monitor. The cabling is bundled with velcro tomato ties, a cheap consumable available at your gardening store.

I covered them with gaffa strips, more for appearance sake than function. Yellow makes cars go faster so should help the camera too, don't you agree??.

I replaced the standard Noga Arm with a shorter model. Inexperienced crews attempt to force the monitor to their viewpoint without loosening the release knob. Their leverage over the longer arms forces the locking friction.

In my experience, it wore prematurely at the the monitor end from this abuse, required increasingly more tighening force on the knob until the circlip at the monitor end tore its groove out and that was that. You can fix them with a thicker piece of welding wire or cut down keyring but it is at best a temporary solution.

I added quick-release plates on top and bottom of the arm. The vendor has hidden these gems under a bushel. They are not readily found and do not seem to be promoted. You can buy the shorter arm with one quick release fitting as a kit from BHPhoto, a dvinfo.net sponsor and Adorama.

You can buy directly from the wholsaler-vendor, Sixteen-Nine Incorporated, who will also sell the quick-release plates separately. The plates are not cheap but they are of good build quality.

They eliminate that one vexing impediment to quick setup and strike of this camera. That is the mounting and dismounting of the monitor.

The camera-recorder is no lightweight. For oldies like me who have dragged a CP16R film camera around, the feeling is familiar. A tripod with a 100mm ball is about as lightweight as I would care to support the camera-recorder upon. I have used a Miller DS10 but it is a big ask.

For audio, I use a Sound Devices MixPre three-into-two mixer to feed the camera with a feedback line to the soundman. Audio levels in the camera-recorder are set to max as default and they need to remain there. For feedback (tape monitor) on the MixPre, I found I had to set the trimmer slightly hotter.

I made a custom snake cable to connect the MixPre to the camera-recorder.

The other SI2K onwer-operator here is Steven Rice. He uses a Zoom H4 recorder in mixer mode, mounted to his camera on a second Noga Arm for agile-portable recording via one of the USB inputs.

I have tried it but found it too tricky to reliably set up under pressure. I find unacceptable latency in playback on the camera-recorder with a Zoom H4 or Zoom H4n in the chain.

If I use the Zoom H4n, I prefer to record double-system and feed separate analogue audio to the camera audio for the sync track.

The manufacturers of the camera-recorder recommend double system sound. For voice work, with care, via a good mic and the MixPre, the camera audio is quite acceptable to me.

From a practical viewpoint, the camera audio does not conform to industry standard. Use a 1K tone to set unity gain through the audio chain and you will end up with very low audio levels in the file.

A workable method is to experiment with both sustained and peak real audio levels on the camera display to match as nearly as possible, the sustained and peak levels of the mixer levels display and use the conventional method of monitoring camera audio back at the mixer.


GOTCHAS.

A power cut during the recording will result in a damaged file.

A hard powerdown with the side power switch before the system has closed the file off will also damage the file.

The preferred method of powering down the camera is to exit DVR, enter the SI folder in the Windows screen and select "Shut Down". It has been put there for a reason. Use it and you automatically reduce the risk of a damaged file to a battery going flat or some fool hanging a toe in a cable and ripping it out of the plug.

DVR2.0 now includes an automatic file repair operation which commences on the next boot-up, so for practical purposes the following applies only to DVR 1.1.681.

Incomplete copying of files. - If a folder-to-folder copy/paste run or copy/paste of higlighted multiple files is commenced, a damaged file in the list will cause the paste operation to terminate at that file and no furthur files will be pasted. All the system will do is stop the operation as if it has completed.

Best practice is to check the properties of both the original folder and copied folder to ensure identical numbers of files and bytes of data have been transferred.

An attempt to play a damaged file in DVR or Windows Media Player in the camera will cause a freeze crash in the camera, also in an editing computer, requiring a hard shutdown and reboot. A damaged file should be noted on a shotlist. If it is not critical it best be deleted to prevent any unexpected hold-ups or glitches later on if the director wants to play back footage.

If it is to be retained as a "best take" or vital footage, then it should be copied to a "damaged files" folder for later repair on a computer and the original deleted. Single damaged files will copy and paste successfully one at a time.

Depending on the version of Cineform software used by the editor, the filename of the damaged clip may need to be changed to an eight-character filename plus extention to comply with MS-DOS command line conventions, eg., "DAMGD001.avi"


DISASTERS.

The camera-recorder is a robust combination. Sensible handling and freight forwarding should not worry it. The Mini head is particularly robust and is known to have survived severe impact.


The camera-recorder body is water resistant but not waterproof. The socket panel on the right side is angled so that water running along its attached cables should fall off an upward bend at its lowest point before reaching the plug. The exception is the XLR style 4xpin power plug which is water resistant as part of its design.

Normal best practice of enclosing the camera and lenses in a weathershield should be followed. Cabling into the rear of the camera should be forced to hang low or be looped so that any water runs will drop off at the lowest point.

In the event of a partial camera flood whilst it is on a floor, the power source should be disconnected immediately without powering off the camera switch.

If there is a mains supply to the camera then obviously safety comes first. The power should be remotely switched off and the location made safe before anyone goes near the camera. If it is only supplied by battery then the battery should be immediately removed.

The camera should be handled gently to avoid any surges of water within which might reach critical components.

If the camera has to be hoisted quickly out of deepening water, then it should be lifted as smoothly as possible, nose slightly upwards no more than about 10degrees and tilted gently to the left not more than about 30 degrees.

Any water inside will move to the rear left where the least amount of vulnerable circuits exist and exit through the fan vent, then more slowly through the gaps in the rear.

Remove the digital magazine, but make sure you keep it level and do not rock it about. Set it level or with the black plastic front very slightly higher for water inside to drip out. Unscrew the two rear screws. The black plastic front and its attached frame should then slide out of its case with the hard drive attached.

Mop off any free moisture. Hard disk drives are water resistant but the circuit board on the drive may be damaged if there has been power present during the water entry to the enclosure.

Assume that the camera is damaged and return it to your vendor for repair. Any "good luck" attempt to operate the camera may add to any existing damage. If you have the allen keys to fit the fasteners, there will be no harm in removing the CPU unit and mopping out all the water you can find on it and inside the camera body before shipping.

I have experienced a camera sensor fogging due to moisture ingress and condensation during heavy rain and a lot of splash from covers flapping in strong winds. It is obviously best not to let this happen. Fungus from an infected lens getting into a damp space between the filters and the sensor would definitely be not good.


The SI2K camera-recorder seems to be fairly heat tolerant. The fans kick in to high flow when a certain temperature is reached. That said, for optimal life expectancy of any electronic components, excessive heating should be avoided.


As with all camera-recorders, rapid temperature changes should be avoided, especially in humid environments. The assembled camera is awkward to bag wrap. A large beach towel might be a good substitute to slow the warming or cooling and to wick off any condensation which might occur.

Camera fan vents should be kept free of obstruction as they are essential for CPU cooling. A little ambush awaits if two recorder units are being used with the camera heads on umbilicals and are being kept close. They should be separated by about 12 inches or so and turned so that warm air from one is not directed onto the other.

If they must be closer, a barrier should be placed between the two. Otherwise one camera may ingest the hot air from its neighbour and overheat.


THE IMS MOUNT ADAPTOR SYSTEM AND IMS TO PL-MOUNT ADAPTOR.

If using a mix of PL-Mount and other lens types with their own IMS adaptor, beware of stacking their IMS adaptor and the PL-Mount. An IMS adaptor will mount to a PL-Mount. This then places the lens well forward of its normal position.

If you discover your non PL-mount lenses will not focus at all, it is likely you will have stacked the IMS adaptors and the IMS to PL-Mount adaptor will be the likely culprit.


Coke-bottling the lenses. If you catch your camera assistant offering the lenses up to the camera one-handed without bracing to the camera with the other, take the recalcitrant out the back and give him a bait.

Good glass needs protecting. On many lenses, the rear element is unprotected and vulnerable to damage if offered up to the camera by the bottle-opening technique.


RANDOM UNSYSTEMATIC NAMING OF FILES BY THE CAMERA.

You will first observe this when attempting to play back your last clip after a shutdown and reboot. Your first thought will be that the camera did not record. When you dig furthur you will find that the order of the clips has somehow been re-arranged and that filenames are moving about or seeming to disappear.

Because the file names are partly comprised of time code, their naming is system clock dependent. The system clock itself is dependent on the memory battery. If the battery goes flat, the clock resets to zero with each boot up.

Filenames on subsequent boot-ups will be generated by a timecount forward from day and time "zero". They will become interposed with older files depending on just where the timecode their name includes happened to be when the shot was done. The files will be there but no longer separated by dates or in the sequential order in which they were shot.

CONT'D BELOW.

ADVENTURES WITH THE DATAPORT (aka. Digital Magazine or D: Drive).

I do NOT recommended the practice of hotswapping the dataports. The drive may still be spinning. Sudden angular changes in its position as you move it around may introduce excessive loads on the spindle bearing which may seize or run out from that point on. The drive may become unreliable.

If a dataport is dropped, no furthur work should be recorded to the enclosed drive. It might be on its last legs and only just capable of yielding the files already recorded. If you take a chance and record more data to it, you may be consuming the very last skerricks of its usable life.

The drive goes into extended continuous duty, when copying out. This extra workload may be the last straw.

If the dataport case appears to be undamaged and you cannot continue without it and are prepared to take a chance, your best insurance may be to relieve any stresses the impact may have introduced.

Unfasten the two rear case screws of the dataport. Slip the frame out of the case. Loosen, but don't remove the side screws which fasten the hard drive to the frame. Check that the assembly is square and not lopsided, then retighten the screws.

This may relieve sideloads imposed on the PCBs inside the enclosure, on the USB board above the drive and upon the connecting socket inside the enclosure.

Check the screws which retain the USB PCB in the frame. You might choose to loosen then retighten them also. However you also risk a cracked corner coming entirely off the board, so they are best left alone until your project is complete.

Check that the rear of the enclosure has not been dented in at the rear corners such that the screw ends press on the little upright transverse PCB inside the rear.


SETTING BLACKS AND DEFECTIVE PIXEL MAPPING

Desirable practice is for the camera to have time to warm up and the internal temperature to stabilise before the first set black and first take of the day. If the camera is to roll immedately after startup, then the blacks should be reset between each take until the camera temperature stabilises, otherwise a fixed pattern artifact may appear in the image.


TRANSPORT AND SHIPPING.

Check all fasteners in the camera body after shipping or a long commute to a location. The front upper side panel allen screws and the countersunk bottom cover retaining screws seem to be the most prone to loosening.


THE IMS ADAPTOR MOUNT SYSTEM AND ITS TRICKS.

The IMS and PL-Mount threaded clamping rings have no latch and retain themselves and the lenses by wedge friction alone.

If the camera is to be operated unattended on a moving vehicle or in a dynamic environment, it is prudent to secure the IMS mount ring and the PL-Mount ring against loosening and releasing the lens to a violent demise.

One method might be by using gaffa tape or a bridge of semi-hardening bathroom sealer and attaching a separate short lanyard to the lens.

Another might be to swap out the small limit screw on the clamp ring for a longer screw and to tighten this screw through to interference on the bottom of the groove it locates in. This will lock the ring. There is risk of overtightening and stripping the thread or breaking the screw off for this is not what the original screw was intended for.


MINI HEAD SUPPORT.

The mini head is often used remote from a recorder on a cable in a dynamic environment. If the head is attached by only one of the available two pairs of threaded holes in the upper and lower surfaces of the case, any remaining threaded hole should be used to attach a safety lanyard as a hedge against the camera becoming detached.

.
POWER.

On the camera specimen I have, it appears that of the two power conection options, the Anton Bauer battery source has priority if both the external XLR 12V input and the Anton Bauer input are supplied with power. The external power connection will not become active util the Anton Bauer battery is removed.

Drawing 12V power from an automotive cigarette lighter style accessory socket is not recommended unless a desperation circumstance occurs. This socket style often has unreliable conductivity and is prone to disconnect or become resistive due to vibration.

A "low battery" warning may be displayed in DVR and the camera may shut down when the fans kick in after boot up. This may lead the operator to start the engine. Vibration may then cause more problems. The automotive charging system may also be brutal and spiky, depending on the age of the vehicle.

If there is an intermittantly open joint on the car battery poles due to poor maintenance, spikes of up to 60VAC may be present on the 12V supply to the camera with damaging results.

So you don't believe me?? Ponder this story.

When starting a stalled vehicle with jumper cables with uninsulated clamps, I was electrocuted and temporarily paralysed when one battery became disconnected and the two alternators happened to go phaselocked AC.

The voltage was AC because the diodes in one alternator had failed and those in the other also failed at that moment. No one had the wit to switch the things off. They just stood there and laughed like fools at my predicament.

If powering the camera via a mains adaptor, it is desirable that no other high draw appliance, like lighting, electric kettles, power drills, heaters or airconditioners share the supply. HMI lights induce noise into the mains power supply waveform which can disrupt computers.


If using a heavy cable such as a snake from the D-Sub connector in back of the camera -recorder body, the cable should be supported away from the lightweight USB cables and plugs immediately below.


MAINTENANCE AND TROUBLESHOOTING.

Memory battery. This is the likely culprit if the order of the filenames alters from their shooting sequence. It is a standard computer memory battery and is accessed from the bottom of the camera.

MORE TO COME.

I am not an SI-2K user, but I have suggestion if I may:

Why not put together a PDF e-book with images and see how it does? If it receives good feedback and gets you more requests for clarification or specific information you can then gauge whether or not you have an audience.

Hope this helps.

Suresh.


Thanks for your response. Like everyone else, I have competing tasks chasing my available time.

This post was a sort of backhanded survey to see if any effort on my part would be justified. Although I did not ask, it was a hint also for anyone else who would want to add their bit as well.

The population of SI2K camera-recorder operators is not huge, certainly not in the league of the RED enthusiasts and fans in the wings.

The number of responses I have received thus far would suggest that most operators are comfortable where they are right now. They are also likely more advanced in working the system than I.

As for an e-book? - Over time, this forum has served well for raising issues and spreading handy hints. My draft notes above are a compendium of bits and pieces I have posted. They are as much as is needed to put my limited knowledge into one place.

I think the documentation of any 'real-world' experience - whether in the form of videos / e-book or blog would be really useful. We all use the camera in different ways. I have just finished a few days of aerials in a KenyanNational Park using a wrap-around strut mount with the si2k mini head cabled back through the cessna door to a P&S recorder strapped to the copilot seat, with the monitor mounted on the wind-screen brace. Before that I was filming tiny fish eggs at c. 20x magnification.
We all learn from each other and I imagine Bob has had more experience with the camera than most of us! I welcome any dissemination of information.

Mark.


Kenya, airplanes and cameras on struts. I sincerely would have liked to have been along on that ride.

The camera C-mount option is very friendly for microscopy imaging with the microscope C-mount adaptor barrels.

I wish I could say I have had more experience. I may have tinkered with it more than most, but others out there have done far more worthwhile work with it.

I've used the c-mount adapter a number of times with Switar RX lenses which, given their age, and design ( originally to be used in front of a beam-splitter) I think hold up really well.
For macro / micro I'm using a self-built 'macro-bench' with the camera held stationary in a horizontal plane and the 'stage' moved by motorized linear slides and joystick-operated micro-controllers - that way I can get rotational and linear movement along x, y, and z axes. I don't much like changing to the c-mount adapter, especially under dusty field conditions. I much prefer the P&S 'professional' Nikon adapter and then using a Nikon to RMS thread adapter. That way I can change relatively quickly between micro-nikkors and RMS mounted Zeiss Luminars.

Mark.


C-Mount. For similar reasons, I ended up making an IMS to C-mount adaptor for an old Kinoptik 5.7mm I have used for interiors of aircraft and cars. I offset the flange a little bit rearwards to allow for the rearwards setback needed because of the thick IR/Antialiasing filter.

With the "Professional" IMS-Nikon mount, take care that another crew member does not force the clamping ring if it does not want to go. Check that the fixed pin which replaced the pop-out lockpin on the genuine Nikon mount is correctly in the groove of the Nikon lens. This pin will press inwards a little furthur if abused.

You'll know soon enough as the lens will not mount up properly, may turn with the ring as you tighten or will be a bit weird in other ways.

Some older Nikon lenses vary in the dimensions of the mount, both in overhanging shoulder depth and diameter. The newer style Nikon F-Mounts which P+S initially styled their "Professional" mount pattern from, have a shallowere should depth. They do not grip the older lenses properly but may jam the aperture ring where it overhangs to the rear.

I understand that shortly after introducing their professional mount, P+S discovered the issue and cut the shoulder depth deeper and narrower to correct it. I think the newer NIkon mounts had a shoulder depth of 0.7mm and the older mounts had a depth of 1.7mm.

If the clamping ring begins to wobble excessively, you will have one broken shouldered pillar screw in there. The other will be soon to follow. If you are lucky it will break before tightening and allow you to remove the lens. If your luck deserts you, the lens will be stuck in the mount. It can be got out but the process is a bit too involved to describe here.

On the mount which had broken, I cut the shoulder on the silver ring deeper and to the narrower diameter of the older lenses.

That was the easy part. Getting the mount apart with the broken screws still in place was the challenge as was getting the broken screw-ends out.

If you adjust the P+S mount to hold these older lenses snugly the ring will leave newer lenses loose. Adjust for the newer lenses and the ring will baulk on the older lenses, which is when the damage from forcing the ring can occur.

P+S have made allowance for brute force but too many times and the little stud screws whch link the outer ring to the inner clamping "sleeve" will shear off and leave the ring to spin free.

There. A bit offtopic but hopefully forewarned is forearmed. I described this issue some time back so I thought I should reprise in case you or others missed it.

The Switar lenses as I understand them, with their flange to focal distance adjusted for the thickness of the splitter prism might be a better match for the SI2K than others.

I have posted thess links previously to airbourne in-cabin footage with the Kinoptik.

http://exposureroom.com/members/DARANGULAFILM/6e540d13ebbe4de5928714d445c8a62f/ unedited footage - better resolution.

http://exposureroom.com/members/DARANGULAFILM/b32d389c3fc7455f8df597bc917c0c0b/ test assembly - lower resolution. The rest of the scene was shot on the ground with white styro to simulate burned out sky under the wings. As a Cessna traveller, you will recognise the background sound layer.

Thanks a lot for that Bob - as you say 'forewarned is forearmed'!

One problem I was finding with the strut mount was not being able to 'set blacks' once airborne. I was using a Zeiss 9.5mm distagon ( mk 2) so couldn't completely close the iris ( also the aperture gearing on that lens is different from its focus gear (0.8) pitch I discovered, so I couldn't vary the aperture in flight using an RT motion wired follow focus with 0.8 pitch gear as I had planned...)
I found occasionally I was getting fixed pattern noise - which I assume was down to the camera temp rising by 20-30 Deg C after the initial black calibration (?).

I either need to find a way of completely capping the lens in flight ( to set blacks ) or perhaps change to my Canon 7- 63 - but that means extra support (as it is a big chunk of glass to hang off the mini while taxying on rough strips) and an extra motor and a redesigned bug shield to cover the larger front element...

Fixed pattern noise. That one is a very hard call. Airflow cooling to the camera body would make it difficult to judge just the right warmup period on the ground before setting the blacks. The only way around that will be to enclose the camera and lens in a cell with a flap on the front. That would cause heaps of problems due to complexity.

One option you might examine is the new generation of electronically controlled welding helmet eyeprotection glass. This stuff goes very dark, as in look-at-the-sun dark, like a normal welding lens but is apparently transparent until a welding flash is detected by a sensor. I do not know if it is actuated by power on or power off. I also don't know just how good the colour is but it might be worth a look.