Aapo Lettinen

not that I know of. He has used the cards with F5 and F55 but is shooting practically always RAW now so those sxs cards are not needed anymore.

my friend has 4 of the 128GB sxs pro cards for sale here in Finland. I will PM his email to you

As I have seen it, the biggest problem with Kinefinity cameras has been and still is that you have to purchase one to really be able to test it in real production environment (unless one of your friends has already purchased one and you are able to borrow that camera for your tests). So it is this stupid "which came first, chicken or the egg" paradox where one can't purchase a camera which one has not tested personally first in a real production but one cannot test the said camera without purchasing it first. Too risky and expensive for most people to gamble on this. That is how I see their market position: no one dares to purchase their cameras because there is too little user experiences with them and there cannot be more real production user experiences because almost no one dares to purchase them for real high budget productions. The only way to solve this would be to send a bunch of their cameras to all major rental houses, even for free, so that they would be globally available for real productions and local testing. Only then they could get larger market share and sell more of their cameras. Even giving dozens of camera bodies away for free would benefit them more than the current marketing strategy. Just get the camera to real big budget productions. Give them away for free if you have to. No one dares to purchase them now because no one wants to be the first gambler and pay a high price for basic testing which really should be free and easily available. Even RED got this back then and got Peter Jackson and friends to use and market their products. Ridley Scott etc. They were easily available in rental houses too. Kinefinity needs to do the same to become successful.

There was another thread with great regular8 camera suggestions. One of the possibilities would be the Leicina cameras. I am just waiting one of them to arrive to investigate how it could be made crystal sync. Simon Wyss can service them mechanically. So they could be an option for you as well

I think this type of camera would work best as a owner-op system for someone shooting lots of stylistic music videos or "tv series opening credits sequence" type of stuff. But it would not hurt to ask if a rental house would be interested :) One of the things I noticed is that most of the K3:s out there seem to have M42 mount instead of the bayonet. So it is a bit challenging to get a camera which is both the latest model and in good condition AND has the bayonet mount so that the better quality primes could be used. Otherwise one would be limited to the kit zoom unless shooting extreme telephoto where one could use stills lenses as well. The cheap lenses are the whole point of this K3 modification so the bayonet mount is pretty mandatory. Probably I could convert a M42 mount camera to bayonet by myself though, the Krasnogorsk bayonet is very simple construction and I could probably make one with metal lathe and hacksaw if needed :D It would be possible to switch some parts between cameras so it would be possible to scavenge the bayonet mount fronts and good working camera bodies separately. This would be relatively expensive though so would need couple of the cameras to be built to justify the time and expense. If someone would order a single camera I would probably purchase 2 or 3 camera kits from eBay and use the best working one as a base and the others for spare parts. Some of the machining needed for the modification is a bit risky so it is best to have spare parts available if something goes wrong. And then it would be possible to get the gears and bearings which have the least amount of wear etc.

Drafting just for fun. I had some spare time so I checked the K2 just one last time. Took it completely apart and took some measurements and put it back togethere again. Weighed pros and cons of different designs. - It is possible to make these cameras Crystal Sync as said in the previous posts. It is not even that difficult to do. But the real question to ask is WHY. Why one does want to make it crystal sync? what does one want to shoot with it which would require very stable speed? The camera is relatively noisy and it does not have that good of bearings if one wants to shoot tens of thousands of feets with it at a time so I would not want to shoot a feature film with it. So why make one running with Crystal Sync electric motor? And would larger magazines really benefit anyone even if they would look good on paper? What other restrictions the cameras have which would make them less ideal for such a conversion? ------------------------- The cameras are pretty small and lightweight for not being sound cameras and for being originally meant for low budget fast shots. They also have very affordable and relatively good lenses. So I would say that one would want to keep the camera body just as small and lightweight than it originally was and one would definitely want to keep the original lens mount. - these goals negate the use of larger magazines. One would want the camera to run 100ft daylight spools. It is not meant for A-camera use on feature films, it is meant to be lightweight and small camera which is fast to setup and use. - the cameras would work best for pickup / B-roll type of shots and for music videos. - to make the cameras more usable for the music videos which would be their main use, one would need to get a orientable viewfinder. Not want to get it, one would NEED to get the orientable finder. Fixed viewfinder just does not cut it. But making a orientable optical finder takes lots of work and it is less ideal for that use. So I would suggest replacing the original optical viewfinder with a internal 100% HD video tap. So it does not have optical finder at all but has a good video tap which enables focusing. The camera could then be used with gimbals and steadicam easily if needed and low/high angles would be no problem. Makes it also faster to use. - mandatory speeds would be 24.00fps, 25.00fps, 48.00fps and 50.00fps. Others would be possible but not necessarily needed. - Most of the original parts could be used. If wanting to make the camera S16 one would need to manufacture a new ground glass for it. These cameras are not ideal for S16 conversion but for this type of use the quality should be OK. ------------ So I drafted the possible modification this way: - Uses the original camera body and most of the internal parts - original film capacity (100ft daylight spools) - original lens mount to be able to use the relatively good and very affordable original lenses (PL mount would look good on paper but no one would really use PL lenses with this camera in the end. At most they would end up using the original KMZ lenses with K-bayonet to PL adapter. So no need to bother with changing the mount for no reason) - Only basic sync speeds and high speeds. No low speeds necessary though they would be possible if one would need them... - 100% HD video tap. No optical finder at all - Meant for pickups, B-roll and music videos. No one shoots features with this noisy cameras so feature use does not need to be considered (though it would still be fun if someone would end up shooting a feature with one of these) - S16 conversion probably mandatory with replacement of the groundglass and the modification of the gate, sprocket wheel and guide rollers - Might be possible to include all the control electronics inside the camera body (external box would be cheaper but no one generally wants the additional boxes in this type of use). Possibly the motor itself needs to be external, there is very little room for it inside the camera body. Mechanical parts need to be manufactured to fit the electric motor. - Complete overhaul of the camera body would be necessary as well to make it well working again. --------- However this is pretty much theoretical drafting because I calculated quickly that making this type of conversion could cost something between 1200 and 1500 USD which is probably too much for any user. This would be a totally unique camera system for music video stuff so I can see lots of uses for it but the price tag is probably just too much. However, if someone would be interested in ordering one, I could do this kind of modification for the Krasnogorsk 3 camera. It is just time consuming and really that expensive and some of the money would need to be paid up front to enable manufacturing the custom groundglass and HD tap and other parts. Let me know if you are interested of this type of super-expensive but also super handy S16 camera. At least the lenses are cheap ?

No one has been interested in this Eclair modification so I will postpone it about 6 months or so. Thus being able to concentrate on the projects which have had more interest. Today I checked the Eclair Cameflex Type24 motor again and it is fully possible to build the crystal sync electronics inside the motor in the place of the old rheostat. The planned specs are the following: - the motor can function with the original 6 to 8v power source. Maybe the cable connector needs to be changed. I would like the motor to use the original start-stop switch so I would want to keep the connector the same if in any way possible. - the speed sensor attached inside the gearbox. Some custom mechanical parts need to be manufactured. - 4 crystal speeds built-in. The top part of the motor would be a little longer than with the original Rheostat top. - it may be necessary to run a thin external cable from the gearbox to the motor top. Inner cables are very challenging to fit to the motors and would make them much more difficult to modify. Partially this is due to the old 70 years old motors having rusted screws and stuff which makes it very time consuming to disassemble them completely. So it will probably be a thin 4-pole wire running from the gearbox to the top part of the motor. - estimated price of the 4 crystal speeds update to the Type24 motor: about 600 usd + shipping + possible customs and tax. As I said, this project has not gained any interest so far. so I will postpone the project at least 6 months and will concentrate on the other projects in the meantime.

Just a quick update of my working process in case someone is interested. This is basically what I am doing hours and hours and hours at a time. https://www.youtube.com/watch?v=G0fGxUU3NB4 Writing new software or making small changes to the old one. Then uploading the altered program to the Attiny85 microcontroller, transferring the controller to the testing area on my kitchen table and installing it to the breadboard. Then running the motor with different speeds and friction settings and measuring how the speed lock behaves. Then going back to the computer with the results and repeating the process. I wrote a completely new phase locking code just couple of days ago based on all my previous designs. Works pretty OK now. There is also the 240Hz reference oscillator with dividers in the video.

or tape. It would be a good idea to ask the HPC guys about the machining job. They could make a nice quality custom bracket as well

yes, it seems it might be possible to make a bracket which mounts there. It would be ideal if there would be another bracket which would attach the bottom of the camera to another point on the ACL so that those combined with the machined lens tube would total 3 mounting points which should make the system stable enough to reduce the vibrations to minimum. If a double threaded lens adapter tube would not be possible, then it could be possible to just machine a tube which mounts all the way over the lens so that it rests against the C-mount -->MFT adapter of the BMPCC. One could leave a small slot on the side of the tube if the iris of the lens needs to be adjusted regularly. That would be even more secure system than attaching the lens to the ACL by the filter threads. And it would be a little bit easier to adjust by adding shims.

it should be pretty easy to machine a suitable adapter tube to hold the lens in place if you have access to metal lathe. if it is just a correct sized tube with screws or similar to hold it in place it would be easier to do than a threaded one. The bmpc could be supported by making a bracket which mounts to the top of the bmpc cage with two or three screws and the other end attaches to the handle of the ACL by some secure way (screws would probably be the best option)

Tuning a crystal oscillator for a customer's project. The binary counter has built-in inverter-trigger section so it just needs the crystal, two resistors, a capacitor and a variable capacitor to get the crystal working with it. The circuit needs to be tuned with a oscilloscope to get the capacitance and bias right. I tested cheap Chinese crystals and a better IQD one. They could be used with the same bias but the capacitance requirements were different. The IQD one was more accurate though the cheap Chinese one was surprisingly good. There was signifiant differences in the sine waveforms (not shown in these images) so I will probably use the IQD ones in the final devices and the dirt cheap Chinese ones can be used for tests and prototypes. Good to know though that even the cheapest eBay crystals are pretty OK for most uses if you need to use them. The system needs another divider to get 240Hz and 480Hz. One can get 150Hz, 300Hz and 600Hz directly from this single divider. I have other means for generating reference frequencies as well. This type of crystal oscillator setup is just another tool in the box which can be used for the purposes where this type of system is the most beneficial choice. Frequency needs, available connections, size of the system, user interface etc. matter a lot when choosing the best technique for frequency generation.

it would of course be possible to add multiple speeds to this type of all-integrated compact motor but it will make it a little bit more expensive and it will make the upper part of the motor a little longer because then I would need to fit one additional circuit board and the selector switches there which takes quite a lot of space. Probably around 3cm longer motor with 3 or 4 selectable crystal speeds when the single speed version could probably made without altering the motor dimensions at all. It might be possible to add a external signal input for the single-speed motor so that the external sync signal would just override the internal sync and no switches would be needed on the motor itself. then an external box could be used for additional speeds. Let me know what you think. Would a single speed and very compact system be better, or would one definitely need the additional speeds and they should be integrated even if it would make the motor more expensive?

I drafted a simple crystal sync system for the Type24 motor. The system is installed in place of the original rheostat control and would not add much bulk to the overall size of the motor. It uses the original start-stop switch of the motor. Single speed 25fps or 24fps would be the easiest to do and it would use the original battery voltage of about 6 to 8 volts. No variable speed, just a single stable sync speed. Expected accuracy of the speed stabilization would be about 0.01fps. Would anyone be interested in this type of modification? I think it would be much more useful than the external box system because it is more handy and does not add much weight or size to the motor, at most just makes the motor body a centimetre or two longer. I have one spare motor which I could modify and sell if someone is interested (I think 550usd for the modified motor + shipping). Estimated cost of the modification: about 350usd + shipping + possible customs costs if you send your own motor here for the modification. Please let me know asap if you are interested in this type of crystal sync possibility for the Eclair Cameflex. I have limited access to my metal lathe due to the corona crisis and will need to direct my limited development funding to the most promising projects, so any project which does not seem to be time critical will be delayed at least couple of months, even a year. So I will need to know how many would want to order this type of modification so that I can direct resources to the designing work.

The images look like the standard Krasnogork bayonet mount. So should fit the bayonet mount K3 easily. As of the SP16 lenses vs. Krasnogorsk mount, the SP16 mount has larger flanges and the lenses may need modification to fit the Krasnogorsk cameras even when the FFD is approximately the same. So you CAN make them work with Krasnogorsk cameras but don't expect them to work right away without any modifications. For example my 10mm T3.1 needed grinding half of the bayonet flanges away to fit the lens to the K2. Took about an hour with Dremel so it is not just "put the lens on the camera and go shooting" even when you can make them work relatively easily

Nice to hear! I will test various approaches with the Type24 motor when I got time for it. One of the ideas was to try to integrate the control system in the existing space inside the motor. There is very little space for it but I think it could be done. We'll see later :) Originally I was going to collect couple of motors off eBay and adapt them all, then if the modified motors would be good I could have sold the readily modified ones so that people would have had the possibility to purchase a ready to use motor instead of sending their own one in for modification which could take a long time. There was two or three nice Cameflex motors on eBay for affordable price. But someone purchased them all and have not seen more of them since so I will just modify one of my own motors and forget stocking ready-to-use motors for other users ?

Simon, I tried to send you email about the Leicina project but did not get any answer, is the email listed on your user profile working correctly or is the address written without the " - "?

the problem with the Bolex turret is that it is only attached with one screw and it is hold at correct position with spring pressure. So adding a heavy lens will lift it off of its base and creates a gap between the turret and the camera body which of course messes up the FFD momentarily even if not breaking anything. Using the screw-on turret locking thingy one could use lenses like the Vario-Switar or similar size and weight. With heavy and long lenses it is mandatory to use additional support. Personally I use additional support with all lenses which are long barrel or heavier than about 1 or 1.5kg . When talking about stills lenses, most of the heavy and long ones are extreme telephoto or long zooms so they would need the additional support anyway to get a reasonable image out of them. It is pretty easy to modify a basic dslr-style 15mm rod support to work with the Bolex so additional support should be no problem (just use riser rod blocks to get the rods to the correct height)

Testing my refined phase lock + motor control code with the actual Kinor camera with dummy film inside. By quickly tuning the torque response I was able to get the average speed pretty close to the target. It is off by about .03 fps or so and I should be able to tune it to .01 fps accuracy with a little more work. It is possible to get it close even with this simpler phase lock system but the accuracy varies depending on the frequency so I may need to make a response curve to compensate this. Even the current system would be usable for sync sound if one just compensates very long takes in edit by a frame or two if needed. The advantage of this simpler system is that it can be made very very small. I am pretty sure I could fit this system to a matchbox or a little larger space if needed. But I will continue with the more accurate phase lock systems as well. Here is how I tested the system today with the camera. The cheap-o mini oscilloscope is for getting quick reference measurements of the motor speed and I am using the pc oscilloscope for monitoring the average speed and phase lock.

I used Attiny85 on these tests because they can be had in small 8-pin DIP packages which are easy to change between breadboards. There is disadvantages when integrating the whole control logic to one microcontroller (accuracy is one of the disadvantages if the running speed of the longer code is not fast enough) but it was needed on this stage of the developing to test the hundreds of small changes I made to the code. The code Friedemann posted previously was for dividing the crystal frequency in microcontroller to get accurate reference frequencies. It does not do anything else and that code cannot be used on Attiny85 without lots of modifications so I just wrote my own based on other examples found from the web. Making those reference frequencies is easy, the main challenge of these systems is in the actual phase locking fuction and motor control code. When I got my own reference frequency code working with the Attiny85 it was easy to add selectable frequencies to it without compromising anything. My current version has 11 selectable speeds from the same Attiny85. If wanting to do the actual motor controller out of a microcontroller for these camera applications you need to write your own software and that is very time consuming. There is just bazillion different variables you need to take into account and test separately (how large of a phase error triggers how large of a correction without causing oscillation of the motor, etc.). The actual physical tests will take enormous amount of time, a lot of it is trial and error. That is why it took this much time for me to get promising results out of the system. ---- I still think it was a good choice for you to choose the brushless motor route, they are just much easier to control with standard code. A real crystal sync motor can be more accurate but it is maybe 10x or 20x more work to get it working correctly and one needs more tools for it. Like I mentioned previously one needs to have at least the dual channel oscilloscope to be able to work with the real crystal sync prototypes. Otherwise you have no way to know it it's working correctly or not. You can't detect phase locking accuracy and phase errors with simple frequency meters or single channel measuring devices.

I got the free running motor working in pretty OK phase locked crystal sync. Check the thread on my previous post for details.

I got the phase locking working with my prototype. It still needs fine tuning and I will change the phase locking algorithm to get better accuracy and to get it working correctly under 16fps (160Hz). The prototype motor has 10-slot encoder so the frequency is 10 times the fps. So 16fps = 160Hz, 25fps = 250Hz and so on. Look at the lower frequency rating on the "Average", that is the median motor running rate. The upper frequency is the crystal reference frequency coming from the oscillator. The whole motor control code is currently running from a single Attiny85 microcontroller. If wanting three digit accuracy I may need to use different system but as you can see this system is already pretty good. I will test this code version next with the motor attached to the actual Kinor16 camera and then I will get to adjust it with actual film load. The motor already reacts to changing dummy loads correctly so I am sure the current version will already work pretty OK with the camera but we'll see. I will probably post a video when starting to work with the full camera + motor setup :)

The pressure plate should push the film against the edges of the gate to get the frame edges flat against the gate. It also needs to support the center of the frame area to keep it at the correct distance so that the center can be focused at the same plane as the frame edges. so the center support is very crucial for image sharpness. But there is a catch. The film does not stay flat on the gate and it may even be beneficial to let it develop a small controlled curve so that it is easier to make the lens's focus curvature flat (it is easier to let the film plane to be a little bit concave than to make it perfectly flat and correct the focus curve in the optics). So the center area of the pressure plate may be machined to slightly lower plane than the areas which push against the frame edges. This is to let the film to develop the tiny amount of concave curve instead of staying perfectly flat. Don't know which cameras do and don't do this correction on the pressure plate areas but it has some benefits so I assume the newer Arri cameras would use it. If the plate surfaces have the correct smooth finish without defects there should be no problem them touching the "picture area" which is the backing of the film and not the actual image

chips can be a problem but scratches rarely show up unless shooting wide angles stopped down. If you have a mount adapter to attach the lens to a still camera it will be very easy to test if the marks will show and if they do, which are the settings to avoid if not wanting them to show in the final image. I have one 10mm /1.8 made for standard16 and it has one single chip mark close to the center of the front lens. Sadly it shows up badly in the image with all settings and cannot be masked with black etc. or repaired so I may need to get a new lens. It looks like there would be a water droplet on the center of front lens all the time and it shows in all lighting conditions. So it is possible that the marks limits the lens's use but it needs to be tested with a camera to know

just a little follow up with the K2 electric/crystal motor modifications stuff. I had time to check it again and it does not seem to make much sense to do a crystal modification for this type of camera. It is possible of course but the main thing is that the camera body and mechanics are pretty restrictive and the huge amount of gears make lots of noise which would need to be blimped. the viewfinder is not very good for this type of stuff either. And one would want a video assist too of course which would complicate it even more. Larger mags would also be a huge benefit. It would still need to be affordable but that would not be possible because there is so much work to modify it to meet all the expectations. There is one good possibility though if someone has any interest in it. One could take the film movement+mirror+gate assembly and the ground glass assembly and part of the viewfinder optics and the main sprocket drive and some other parts out of the K2 or K3 camera and make a completely new camera body where these would be installed. So one would just scrap the outer shell of the Krasnogorsk camera and build a new camera using some of the inner parts from the K-camera. This way one would get room for the motor and the video tap, one could make a orientable viewfinder for it and use separate larger mags with it and one could also install a more practical lens mount to it. And one could use a belt drive between some of the components to get rid of the large number of gears. And one could make it much more silent to begin with. I don't know if this makes any sense in the end but it would definitely be possible because almost all the precision parts could be taken from the Krasnogorsk camera and only things like viewfinder tubes and the camera housing etc. would need to be manufactured which does not require special tools or that much expertise (just lots and lots of time). This option would only make sense as a DIY pastime project for camera enthusiasts but could be pretty useful and affordable if one does not count his/her own working hours. The final camera should be usable in indie films if one makes it silent enough. If someone would like to try this type of "frankencamera" Krasnogorsk project we can draft something out here and make a project out of it :) I should be able to provide an OK priced simple crystal sync system for this type of camera later this year if someone would be interested. As I said one would need to build one's own camera body for the project to make financially any sense but this type of camera could work as a "semi-opensource" project or something :)