Jump to content

Robert Houllahan

Site Sponsor
  • Posts

  • Joined

  • Last visited

Everything posted by Robert Houllahan

  1. Just a followup note. You can get to good but not accurate and with non repeatable control over image dimensional and color by shooting off a screen. The better the screen (lcd etc.) and the better it is calibrated the better the result. making a 24.00fps file and having a camera with a crystal 24.00 fps motor will be best. 35mm will be much more expensive than 16mm either DIY or professionally. A Arrilaser or Imagica recorder will be able to properly and repeatedly record the full density of the film with actual RGB recording per image pixel. A LCD (etc) has co-sighted colors with a very limited grey to grey contrast ratio so there are allot of very fine issues with really recording digital to film that are allot of work to address. The last 20% or so of fine work imaging digital to film is a very steep curve to climb and gets complicated and expensive to do, stuff for features and accurate archival preservation. YMMV.
  2. As someone who has two Arrilaer recorders and has built three generations of 16mm recorders I have a few thoughts. 1. It is relatively easy to film off a computer monitor, I would recommend a sync sound camera not a K3 or Bolex as the frame rate will be off for the whole recording. 2. Setup the monitor and take a spot reading of a grey card to find your exposure. 3. Try to use the most rectilinear lens available, this will keep the image the least distorted. 4. Carefully setup the camera and screen to be parallel to each other. My 3rd generation 16mm film recorder uses a Mitchell pin registered camera and a Eizo Radiforce panel and runs at about 4fps with the camera shutter time and interval between DPX frames being controllable. I also have the panel profiled. You should not try to shoot video that is pushed into the shadows and if that material is then use the log tools in Resolve to elevate the shadows about 100-120ire on the scopes to retain hat detail.
  3. I think it depends on the level of service but in general a full service contract for a machine is about 10% of the purchase price annually. I am happy with the duo of LaserGraphics machines we have at Cinelab now that the "personal" has 2-flash HDR. The Archivist's newer Sony sensor 5.4K camera is very low noise and makes excellent scans.
  4. Digital sensors are linear devices (even when encoded into log) so a ISO change on a digital sensor is an overall dB gain increase. Film negative has a characteristic curve and a photo-chemical push will accentuate that logarithmic curve with more sensitivity in the middle than the top or bottom of the curve. In my experience 5219 pushed +1 is in the same sort of working range as an Alexa in the 1200 maybe to1600 ISO range, with the Alexa having some more shadow detail and 5219 more hilite overexposure tolerance and the mid being similar. YMMV
  5. I would scan HS film at the framerate you expect to deliver at, i.e. 24fps 25fps as then it will have the intended slow motion. You could get a ProRes and DPX scan if you expect to do any speed changes in post.
  6. That client was the US Military Industrial Complex and that was fractions of pennies on the hundred dollar bills they happily burned through. Certainly not any more but it was only within this decade that Hycam cameras were mostly replaced for this kind of work by digital substitutes. The US Air Force had a few very substantial labs of their own. I bet Kodak would do a special run for the overall cost of a master roll now and I was exaggerating a bit about the 25M feet intro cost.
  7. A hair or other small object in the line scan gate on the Spirit does make a streak through the whole frame so it can very well be that is what this is, I have seen similar looking streaks on my Spirit(s) especially because how it dances so far across the frame and out of frame left. I would just send the film back for a rescan.
  8. Was this scanned on a Spirit? It could be a hair in the line scan gate. A movie clip would be more helpful.
  9. I would think that if you offered to buy 25 million feet of film from Kodak they would figure out how to use multiple layers of Vision 3 emulsion without the color dyes and couplers for a theoretical new monochrome stock. Probably at a $1.00/ft or more price.
  10. I have a VV job doing tests right now. I am working on being able to run 9.4K x 7K VV scans for it. We also ran a VV short a few years ago that was really nice.
  11. I have two Visual Inst. Hycam-II cameras (built in 2002) and we used to do allot of work for Pratt&Whitney on FAA mandated tests for jet engines and other high speed camera work for military engines, including for the F-35. The largest test I worked on for them was a engine for the A380 and there was 27 Hycam-II 16mm cameras shooting SO-19 (7219 on Estar) at FR between 6000PPS and 11,000 PPS along about a dozen digital high speed cameras. The test was lit with 500 1K Lowel Omni quartz lights. The newer Hycam-II cameras have new motor control electronics which allow for the camera to get to full speed within 50-75ft of a 400ft roll. P&W would "rent" the lab for a day and we would process the negative and then do a transfer to video, but then cut the negative down to the sections they wanted to really see and make a timed 16mm print. I remember holding about 850ft of negative and thinking it cost about $35 Million to produce it. Picture clarity and detail was really excellent and far exceeded the digital cameras even very recently. The digital cameras would also have issues failing due to the very high vibration environment of a full running 80,000lb thrust engine blowing up. It was only in the last five years that they switched entirely to digital for this work.
  12. The Cintel machine is but 16mm is definitely capable of allot more detail than just 2K when scanned on a machine that is capable of higher res scans.
  13. I am sure he had something to do with it. This was also converted to computer control by Rennie Johnson, I sort of wish it was in it's original config but the PC control works fine too.
  14. Noiiice! i have Harry Walton’s Aerial optical printer at Cinelab that did work on RoboCop amongst others. i recently got a request for a possible Techniscope to 4-Perf anamorphic optical blowup. I am kind of psyched to try out the process. Harry walked me through the basic setup to do it.
  15. Yes 2242 can be used to make a IP from an IN or OCN it is a printer light setup to meet the LAD AIMs Also it is the stock we got so it has to be made to work unless someone has a Tardis.
  16. I cannot see that being the case either I am sure FotoKem is not charging $0.25/ft for 4K 16mm scans on the Scannity. Neither is Pro8mm on their Scan Station.
  17. In all deference to Nikki I am sure she is nice and thorough (we just ran some B&W for her) buying a machine and immediately undercutting everyone else with the same machine's prices is IMO not the best way to start a business. I am sure this can work for a while if you have a scanner in your house and it is just you and a partner running it as a personal enthusiast venture. Start adding cleaners and film processing in a residential neighborhood etc. etc. etc. The Scan Station's ubiquity has largely made film scanning a commodity as DaVinci Resolve has for color grading and there are shops who use Resolve and charge bargain prices and shops that charge top dollar. And there are shops with Nucoda or Baselight etc. Racing to the bottom of the price scale is not a great thing and there are many many pro and other clients who won't work with the "cheapest" shop because they feel like something is wrong when the price is too low.
  18. longer exposures will actually show more of the inherent fixed pattern noise present in the shadow area of the sensor. Allot of the reason for "HDR" multi flash is to work to overcome the sensor noise flaws by pushing a "hot" second exposure to get the densest part of the negative into a better response portion of the sensors dynamic range and by doing so keeping the FPN out of the visible scan. This can get more DR from the shot film especially if it is overexposed but mostly the HDR process cleans up noise from the sensor.
  19. " It's not clear this is the case. It is. I got one. It is a even newer gen Sony camera than the 6.5k one and it is excellent, extremely low noise and higher resolution than the 5k cmosis. The JAI Spark with the CMV20000 is 5120 x 3840. The New Sony in the Archivist is 5.4K x 3.8k And if you can find anyone who has been more of a pain in Stefan and LG's butt about sensor noise on the Scan Station with the 5K camera than me you should buy them a donut and a coffee.
  20. Well to what purpose? on a monochrome sensor scanner? You could control the camera exposure time if you cannot control the lamp but why bother when precision PWM LED lamp control is a trivial thing to do these days?
  21. I honestly don't care if Orwo made a deal with ZUUL and had the formulas for the Agfa 320 smuggled out of Valhalla by Tinkerbell. The only relevant thing is a new color film stock and there are some rolls being shot already so it appears to be real and not vapor. After that the only question is if they can make it consistently and at a fair price and stay in business.
  22. Yeah I have one of those VieWorks too but Rennie found it to be complex to integrate into the Xena especially for the speed. I think the Arri (and probably the director?) do a single 1 pixel X-Y shift per microscan.
  • Create New...