Any ideas on this telecine unit?

[Pawel Saladziak]

Hi. I just won the auction:

 

http://cgi.ebay.pl/ws/eBayISAPI.dll?ViewIt...em=250252671106

 

Has anyone ever seen such a thing?

 

Seller claims it is a transfer unit, my all hope is that it is capable of frame by frame mode, so I could hook it up with DSLR to make home 4k-5k telecine and then modify one of my cameras to ultra16.

 

So, once again, any ideas how it works? I paid this today and since it will come from germany there are still a few days to wait for having it in my hands.

 

 

regards

Pawel

[alexandros petin]

Hello

 

I am wondering if your dslr is going to handle 4000 frames of a 100 roll

Isnt it going to crush ??

i think these have a maximum number of shutter use.

[Pawel Saladziak]

for modern DSLR max is about 10000, but as far as i know here in poland when you buy a camera in bigger store you get warranty for the whole camera for 2 years :)

 

 

p.

[Nate Downes]

for modern DSLR max is about 10000, but as far as i know here in poland when you buy a camera in bigger store you get warranty for the whole camera for 2 years :)

 

 

p.

Warranties usually do not cover normal wear and tear, and shutter fatigue qualifies as such. I used to sell cameras, and I had one kid spend $600 on a nikon D50 for just this, and his warranty was voided due to the way he used the camera.

[Jaron Berman]

Slightly different, but on my 1DmkII, the shutter is guaranteed 200,000 cycles. If you get more (which I did), lucky you. Sent in to Canon Professional Services, $228 to replace the shutter, clean everything and ship it back. Sucks to have one go down, but now I essentially have a new camera. $225 every 200,000 cycles...it's your world, you do the math and see if it's worth it.

Edited June 6, 2008 by Jaron Berman

[David Auner aac]

for modern DSLR max is about 10000

 

Our local Nikon Pro rep told me about a guy who put 416000 shots on a D200 in little over a year. Only then it went down. And two filmmakers who used two D200 for a film at 6fps or so. Both cameras took over 250 thousand without any problem. So on the better DSLRs (40D, D300 and up) at least 100 - 150 thousand should be the norm.

 

Cheers, Dave

[Paul Bruening]

After I burn up the shutter on my Kodak DCS DSLRPro C I'm switching over to Canon Rebel XSi (the key to the Canons is the 65mm macro lens). They're only $800.00 a pop and should go down in price after a while. They're pretty fast on the sub-processing and data moves and store 4K RAW and JPEG proxies together straight to hard drive. At the price I can throw them out when the shutter blows and still have a scan cost of $0.005 per scan. If I keep two on hand and rotate them out with Canon service my scan costs could get down to about $0.0015 per scan. I'll go ahead and do the math: 2 hr feature (173,000 frames) scanned with handles in 4K RAW for about $250.00.

 

That's why I built my rig. I don't care how many times people, here, imply that I'm stupid for going this way. It's the cheapest DI on the friggin' planet.

[K Borowski]

After I burn up the shutter on my Kodak DSC DSLRPro C I'm switching over to Canon Rebel XSi (the key to the Canons is the 65mm macro lens). They're only $800.00 a pop and should go down in price after a while. They're pretty fast on the sub-processing and data moves and store 4K RAW and JPEG proxies together straight to hard drive. At the price I can throw them out when the shutter blows and still have a scan cost of $0.005 per scan. If I keep two on hand and rotate them out with Canon service my scan costs could get down to about $0.0015 per scan. I'll go ahead and do the math: 2 hr feature (173,000 frames) scanned with handles in 4K RAW for about $250.00.

 

That's why I built my rig. I don't care how many times people, here, imply that I'm stupid for going this way. It's the cheapest DI on the friggin' planet.

 

I'm not implying you're stupid. I'm not implying anything; I'm bluntly saying you're not being realistic with the time component or the cost components, and you still have technical problems and workflow problems.

 

You're going to have to scan more than 2 hours of footage for a two hour feature. How else can you actually select the best takes? Realistically, you'd need to scan at least 20 hours of footage to get a 2 hour feature digitized, probably 40 would be more like it. 5:1 is pretty spartan.

 

How are you going to align frames? There's going to be some variance. Howabout the time involved? You can't set up your machine and walk away, unless you want to risk trashing the entire OCN. You've got to sit there and baby-sit your scanner through the entire run.

 

I'd recommend you try to get a Kodak HR 500 or other still scanner that can handle long rolls of film (they're optimally designed for at least 100 feet at a time) and then try to adjust the tolerances so that they're tighter and will give you good registration, and then see about having someone write software to enable you to scan 3- or 4 perfs at a time instead of the prerequisite 8 that a scanner of this type is set up to do.

[Paul Bruening]

This rig telecines and scans. I shoot the footage once through with a 24P XL2 in SD straight to hard drive (It does the pull-down in-camera, either 3:2 or 2:3:3:2). The NC's gate has been cut to shine the Kodak edgecode through and captures it in the SD frame. I use the SD for workprint in Premiere to make up the scanner's shot list on the edgecode data with a head marked frame on each roll. When I'm ready to conform the scans, I load up the rolls and assign the scans using the shot lists from the workprint. Bruce "G'Day" McNaughton's software guy, Paul Moorcroft, provided the GUI software to control Bruce's single frame/24fps motor on the NC in conjunction with the DSLR. I can, literally, scan exactly only those frames that match the workprint's, named in whatever file system I choose. The GUI does this automatically per roll, so, I can sort-of walk away from it. I throw the EDL and files over into Adobe AE and conform the scans to the WP. Then I do all the other cool shiznet that you get in DI including compositing, Maya 3D CGI, and color timing. I can keep it all 4K since the cost of SATAIIs is getting down to under $200.00 per TB. Then my dual core, 8 MoBo render farm/work stations does the crunching.

 

My biggest problem is that this system is interdependant. I can't take Aaton or Arri code and plug it into my pipeline. Shot lists have to be made up by hand in those cases. As long as I'm going with my own productions, it's okay because I use the integrated system from start to finish. I never conceived of it as a customer service. It's only after my heart attack ate all my money that I wished it was a more versatile system.

[alexandros petin]

hello again. i hope you didn't misunderstood my question.i wasn't implying anything.i was wondering myself about that after my FM2 crushed after some years of moderate use and wanted to know the dslr 's threshold.

 

:-)

[K Borowski]

hello again. i hope you didn't misunderstood my question.i wasn't implying anything.i was wondering myself about that after my FM2 crushed after some years of moderate use and wanted to know the dslr 's threshold.

 

:-)

 

A good rule of thumb, Petin, is that you have 50% failure at around 100,000 cycles with a focal-plane shutter. That means that your camera's shutter will fail, on average, after 100,000 shots. DSLRs may have other parts that wear out before 100,000 shots, but I'd still plan on your camera's useful shutter life being 100,000. You can replace shutters, at least on SLR models that are still made, a big if with digital's planned obsolescence, for about $150-200.

 

Paul, it may still be worthwhile to get a professional telecine, at least partially when you consider the amount of time all of this takes.

[Paul Bruening]

Hey Pawel,

 

Has the unit arrived yet?

[Pawel Saladziak]

No, it has not. It comes from abroad so probably it will take few more days. I'm really excited about this :)

 

And by the way, Paul, passed thorugh your mails about your scanning rig, what do you think about using a Plustek 7200 series scanner as a scanning device? In case I would not be able to use the unit I'm waiting for I was thinking of making a custom one stem ultra16mm movement that would replace the film adapter that goes together with this scanner. Some more software for batch control and voila :) what do you think about it?

[Paul Bruening]

Hey Pawel,

 

You're going down a road many of us have gone down. Few have followed through on these projects because there are a lot of challenges to overcome. And, no matter how you go, there's some capital investment involved. For most, they'd rather just pay for the scans than pay for the uncertainty of DIY DI. I believe their hesitations are completely legitimate. Especially a factor is how the scan market is turning. If you catch a boutique post house the size of mine, you might get them in a desperate time or in a price war and snag decent quality scans (2K or less and unprocessed) off of them for as low as $0.05 per scan. Sure, they won't be Arri-scan quality. But, they may be good enough for a dubious, indie project.

 

Film scanners and flatbed scanners seem like an obvious choice for conversion. When I was making decisions about them 3 or 4 years ago, the big problem with them was speed. They're slow. Really slow. Part of why they can get the resolutions that they do is because of the time they can take getting the maximum interpolation from the sensor. Many folks here figured that slow capture time might not be a big factor for Indie producers since they can wait things out in favor of not spending much on scans.

 

I assumed that scan speeds would improve as the years have ticked by but they haven't improved that significantly. Film scanners are still slow. Flatbeds are slow as well. So, first, you have to get the illusive numbers on each scanner's scan speeds at various resolutions/pixel depth and decide whether that unit is a useful approach or not. You pretty much have to find someone who owns that particular scanner and get some specific numbers on the ACTUAL scan speeds. The salesmen don't know and will likely give you optimistic numbers. The tech staff of manufacturers know to avoid answering those kinds of questions about their products. I've tried through the phones and got really horrible information through all channels. It's a game they all know how to play.

 

That leaves you with DSLRs. Things actually have gotten better with DSLRs over the last three years. Prices have gone down, resolutions have gone up, and processing speeds have improved as well. There are problems with that approach. Don't get me wrong. It's a hassle trying to cobble together technologies not professionally designed by trained engineers to match with each other.

 

So, before you proceed with buying and ripping into any film scanner, I'd recommend you find someone who uses that scanner and get some REAL numbers on its scan speed. If you can live with it, then proceed to the next questions. As a reference to compare with: A DSLR scan system can snatch frames in 4K+ RAW and JPEG proxies and store them to hard drive within anywhere from 12 seconds to 5 seconds each. Multiply that with 173,000 frames to scan in a 2 hr movie. Twice the scan speed can cut a scan project down from 5 months scanning to 2.5 or 2 months to 1 or 20 days to 10. Those seconds per scan add up to a hell of a lot of time committed to baby-sitting equipment.

 

Does that help? I have a bad habit of running down un-asked-for directions. Ask again if I missed something you wanted to know.

[Pawel Saladziak]

Hey Paul, do you know that Pawel is Polish translation for Paul? :)

 

Okay, thanks for your opinion. The telecine unit arrived, unfortunatelly no use with that, pity for my waste of money :(

 

I decided to try your way with the DIGI-SLR-style setup since after some rough calculations speed of scanner based solution is much to low.

 

And I got this, taaaaa-daaaa:

http://cgi.ebay.com/ws/eBayISAPI.dll?ViewI...em=320291042012

 

It is in perfect condition and is able to run 1 frame per second (it has a film cooling system :)

 

So now I will try to couple this with trigger of some low end digital camera for tests, that will put pictures directly into my computer.

How do you think - do I have to change the light source for a weaker/different bulb? Or I can use the original one?

 

best regards

Pawel

[Paul Bruening]

Hey Pawel,

 

Look around for a camera that can macro down to your 16mm frame size. That's a tough challenge. Few can (Canon's MP-E 65mm lens can but it ain't cheap). You can shoot a screen instead. But, I can't tell you with certainty what that will look like or what's the best surface for that. You'll do better with a camera that can tether to the computer. Canon and companies that big supply control software for some of their cameras. That will help. Registration of the film image will be an issue. The projector will handle a percent of that but software my take up the rest. Adobe AE has a frame registration feature. It's really designed to stabilize shaky-cam but I've heard it can smooth out registration problems too. I haven't used it. I can't account for how well it works. Controlling the projector along with the camera will be an issue. I had Bruce's software guy, Paul Moorecroft, write a utility for me using Microsoft's .NET system. If your projector runs at 1 FPS and the camera can shoot at and store files at 1FPS or faster this may not be too difficult. You only need 2K sized RAW files or less for 16mm. More than that won't be that useful and will slow down your system. Also, you'll probably want to do 2 passes on your film. Videotape the first run for a workprint. Then second pass scan only the frames you need for conforming to your finished work print. You probably don't want to scan all of your footage since that would grind up a lot of shutters (2 hours of footage = 173,000 frames scanned. 10:1 shooting ratio = 1,730,000 scans. A DSLR's duty cycle is about 100,000 to 250,000 cycles). If you go a telecine/scan approach, you'll need a way to sync your video workprint to your footage for two reasons: You need to know what frames to scan and you need to link your dialog to your picture. I had the edge of my aperture and pressure plates ground out to reveal Kodak's edge numbers. I shoot that in my video frame so I know exactly what frames to scan. I don't have to shoot in 24P, absolute sync with my telecine/video camera since I have the edge numbers, but I prefer to. I can't account for how your video camera can deal with the difference between PAL and film speeds. It may smear the film image. Someone, here, may know how well that works out. You could simply project some film and shoot the screen with a video camera to test. I'm still working out how to control my projector speed accurately (23.976 fps). Slating the head and tale of your takes will help with audio sync on your workprint. My Frankenmitchell has an automatic, head and tale, bloop light and tone for this (a la Bruce).

 

The projector's lamp will be a problem. Projector bulbs don't last very long. 30 hours is an average. This unit may have a longer lasting bulb since it is an analysis projector. I don't know it so you'll have to get the number off the lamp and check with a manufacturer on it's life span. I'm using a mini-candellabra (E 11) bulb. It's a halogen type (use halogens or ordinary tungstens for their even color distribution) that comes in 250, 500 and 1000 Watt units. They're pretty long lasting in the 800 to 2000 hours range. Those are American numbers. I don't know if they are available in Europe. The bulb is longish and makes a linear light pattern if lensed. I'm shooting mine through an old style slide projector and bouncing it into my projector using mirrors. I'm doing it this way because of heat issues. Your projector might take the Mini-can socket and bulb with an easy modification in it's existing lamp house. The shorter light path may allow you to use the 250 W without frying your film. Mini-can sockets and bulbs are a lot cheaper than projector bulbs as well.

 

There's probably more that I've overlooked but this should give you something to chew on. I do wish you the best of luck with the project. If you can figure out a way to do it cheaper and easier, I'm all ears.