Dan Baxter

Here's the story with FF: https://britishcinematographer.co.uk/filmfabriek/ The scanner you mean is the Müller HM73, the first model made by the late Daan Müller. FF were set up to commercialise the project/make it a commercial reality, but according to the webpage he unfortunately passed away about a year into that partnership.

Good stuff guys. Such a waste to have used "film-chains" given the costs involved to shoot, process, and print the film in the first place, and they're obviously not capable of proper color correction for broadcast. I guess that would be less important for TV shows compared with movies where dark scenes need to be boosted up for a lit-room, and obviously running those Rank Cintel Mk3 telecines would have a cost a fortune compared to them.

Yes correct, very different. The entire process is described here: https://www.adapttvhistory.org.uk/16mm/ This is how a TV telecine works: That's a negative on the telecine, but when they were in full use the normal thing would be that the TV studio would have a 16mm TV print for the show that is printed specially for telecine transfer (low-contrast and without subtitles or things they want to add electronically to the broadcast tape). The entire color correction is done by the telecine technician as you see above, and I believe they were actually called "colorists". They transfer the film to broadcast tape (NTSC or PAL), and the broadcast tape can be edited separately after transfer. For example you could remove a scene to make it shorter to fit a 30 minute time-slot with all the commercials you're packing into it. Another issue with TV shows is that by 1987 in the US they preferred to edit their shows on tape and not on film creating the problem that their film-shot TV shows then had to be standards converted from NTSC to PAL resulting in aberrant quality. To address this specific issue the DEFT (Digital Electronic Film Transfer) device was invented. You can read the patent the Snell & Wilcox operator's manual and also see the Doom9 thread and 1990 New Scientist article. Official launch was either 1989 or 1990 and they cost about $500,000. There are DVDs, VHS, and LaserDiscs from that time that used DEFTs to convert NTSC to PAL and it leaves behind a very distinctive artefact pattern. Unlike what's described in the patent but is described in New Scientist, the PAL tape deck was modified to run at 47.952Hz instead of 50Hz, about 4% slower and in-sync with the NTSC deck, achieving the 4% speedup/slowdown when going between formats. It could cope with orphaned fields and was highly lauded when launched. The Prisoner (1967) was shot on 35mm, it would have gone through the reduction printing to 16mm for TV and then those TV prints would have been stored with the broadcasters and temporarily transferred to broadcast tape via telecine for broadcasting. Clearly by the late 80's the cost of broadcast tapes was lower than 16mm telecine, so by 2000 they would have transferred stuff once and then keep the broadcast master tape handy for the re-broadcasting, and only rely on the 16mm copy if necessary. What's really worth saying is that the telecine process itself lasted from the 1960's when it was first invented all the way through to the 2000's - so that was a very long time and reflective of the limitations of videotape and standards conversion between NTSC and PAL. The major difference between that process and "digital scanning" is that there's no computer involved - you're transferring directly onto electronic tape, not a hard drive. And of course, you'll scan the original cut camera negative if available - so for the Prisoner that's 35mm film and not the 16mm low-contrast prints made for telecine.

Ask questions, do your own research, find someone who's already doing what you want to do and befriend them if you can. Identify the skills you already have and build on that.

If you mean this: That is not a NASA rocket, it's a US military intercontinental ballistic missile (ICBM) launch printed on US Military 10-perf 70mm film. That scanner is purpose-built for formats that nothing else really scans (mainly 15-perf IMAX). It would have been printed off a 10-perf 65mm negative.

They all flash, like this: https://moviestuff.info/moviestuff-_mark-ii/ (That shows a FilmFabriek Pictor). Extremely short exposure, and running the film too fast will blur it of course. The Arriscan has a rolling-shutter camera, I think it's now the only scanner with one, everything is global-shutter. So it has to hold the film very steady to do its thing, and it flashes multiple times per frame to get its scan done. The light has a glass integrating sphere.

Hi Florian, I don't work in the industry - but in terms of the film machines I'd say you should make a list of them. Film processors are getting thrown out by labs that are closing, you can still get rebuilt processors through MMT (and ultrasonic cleaners). I'm not sure who else is still doing it. Film printers, you'll need to speak to a lab about how printing works. Technologically the film cleaners have moved off Trike (the manufacture of it was banned under the Montreal Protocol), most of them use either Perc or 3M/its replacement, but the best cleaning solvent now is Isopar-G. The new Cinetech Hydra cleans with Isopar-G. Rewashing is how you get embedded mold out of the film, and rejuvenation machines of the past turned out not to be very archival so you need a lab that knows their stuff. For projection you can still get new portable 35mm projectors manufactured through Cinemec the cleaners for projectors are the Kelmars (they have 35mm and 70mm ones) or the Film-O-Clean for small-format. You can get the media pads for them through Film-Tech and also Film Guard in the US, there's other products available as well (Isopar-G for example - it's faster drying but Film-Guard is designed to be slow drying). Projectors shifted towards platters, and before the platter system was invented this giant contraption was conceived of: Now you can use either platters or big huge reels (6,000ft and even larger). You can read more information about the Loop-Matic in the youtube video description. The telecines and scanners have changed enormously since the 60's. This is a demonstration of how a telecine used to work: More information here. The Rank Cintel "Flying Spot" telecines/scanner used a CRT light source. The next generation of tech moved to Xenon lamps, and then after that to LED. The LED lights in a current professionally manufactured scanner will flash once for each exposure - that's true whether it's R/G/B like an Arriscan, Director, DFT or Bayer like a ScanStation or Blackmagic Cintel. They have different diffusion methods to get the light even, some have glass integrating spheres and some have integrating cubes (light directed up through diffusion material. Prasad now sells the QUADRIGA INSPECTIONscan, and it's a unique device in the market in that it's designed to do film inspection work that is normally laborious for staff. The newest model Cintel from Blackmagic has an inspection option (to make the base damage visible) - but it can't scan edge-to-edge. The InspectionScan makes a damage report including shrinkage finding new damage since the last time it was inspected etc. Now, the scanners. Blackmagic and FilmFabriek make retail product scanners. That's not the same as the other companies, the other companies are making specialised professional-use equipment. So while the imaging tech etc may be similar now, their capabilities are vastly different. That's why some people purchase a cheaper retail product device and are disappointed when it doesn't work the same way as a much more expensive device that carries a monthly support contract to pay. So when you hear people say "Blackmagic need to change the sensor" - it's true it has a noisy 4K CMOSIS chip camera in it, but it's the only machine-vision camera they make and the only one that will work with it - their products are not designed to be modular. Modular means you can change something in the future, eg put that new brighter light for the latest G3HDR+ model into an older one - that can't be done, and neither can the camera. So even if they change the camera in the next model you need to buy the entire scanner - ~USD $35,000 all up, you will not be able to buy the camera itself and install it as some people expect. FilmFabriek is different and you can read their story here. They were always designed as "modular" but retail - so no ongoing support contract to pay. The Home Movie to VHS transfer market from the 1980's onwards used Elmo Telecines, and those were replaced starting in 2005 with Tobin Cinema Systems devices and you read that history here. Urbanski Film can service them now in the US for enthusiasts who like old tech and want to have them running, I'm not sure who services Elmos and Tobins in Europe. MovieStuff entered the scene as well, but those devices made by Roger Evans were different. Roger was a part of the DIY scene as was the late Daan Müller who invented the precursor to what is now the FilmFabriek HDS+, and Roger's goal was to go to frame-by-frame digital 8mm sacrificing quality to achieve that goal. For some reason some of the home movie scanning people purchased them, probably because they didn't realise that an Elmo or TCS TVT-8 is what they wanted (they go straight to NSTC or PAL no computer needed). These days there's a company called Film-Digital that sells those projector-mechanism telecines. When you hear people say that the Blackmagic Cintel just needs a new camera and will be like a ScanStation, there's a lot they're not saying. So here's a photo for reference: That shows the complete Euro Format kit for the LaserGraphics ScanStation. The gates/skid plates for 9.5/17/24mm and the tri-format rollers for the Euro Format. You take all the normal 35/16/8 tri-format rollers off the front of the machine and then put those on, and you put in the appropriate film gate. That one film transport module (the motor) supports 35/24/17.5/9.5/8mm - six different film gauges all at once. Blackmagic Cintel 1 and 2 (the sproket transport ones) can only do 35mm and 16mm, and the capstan model ones can do 8mm but at very modest resolution. Hopefully the photo gets across a thousand words to explain just how different they are (the retail product devices and the specialised professional-use equipment). Blackmagic's Cintels are still professional devices however, some facilities have them sitting right next to the top pro tech gear. For what they can do they're a complete bargain, that's why I think it's unrealistic to think Blackmagic would want to change their business model with it - doing that now would leave behind many of their customers who rely on them. Libraries, Archives, etc. At the very top end you have the Arriscan XT which can even do wetgate scanning, the DFT devices, and LaserGraphics sells the Director. Mid-range you have Kinetta, DCS Xena, the ScanStation and its derivatives. For inspection work you have the Cube-Tech QUADRIGA INSPECTIONscan, it is very expensive, Filmic Technologies sells a device with similar goals in mind for 16mm only - and that's through the regular sales agents (Urbanski Film in the US, Gencom in NZ, etc). I fear that if people compare the price of a Filmic EZ16 and an QUADRIGA INSPECTIONscan they will make the same mistake that some people have made when comparing the price of a Blackmagic against something else. If you want to see how they work you need to go somewhere that actually has one and see it for yourself, or get the manufacturer to show it to you. That way you won't have incorrect expectations/assumptions. So in summary, I'd say the changes in telecines and digital scanners if we're talking specifically from 2000 through to 2024 would be that: 1. it's no longer just for television use and for Hollywood special effects and early digital restorations, 2. the next market was post-production more widely and then the Archive market, 3. the devices have become more specialised and more capable in what they can do, 4. the cost of the devices has come down substantially, 5. there's many different companies making them. 35mm specifically that's Blackmagic, Digital Cinema Systems, LaserGraphics, Arri, Digital Film Technology, Kinetta, and Cube-Tech. The Arris have a sproket-drive transport modules for 35mm and 16mm. They also have a capstan-drive/sproketless one, but it will slow the scanner down a bit to use it. For normal use you'd use the sproket transport. But remember, the Arris, LaserGraphics and DFTs out there in the real world are very different to each other - those companies have improved and changed their designs both in terms of hardware and software compared to their earliest models. Yes, the film transport modules have changed significantly since 2000. Older scanners had what are called "projector loops" in them and that was so that the film would be held steady in the gate while in continuous motion that's also required for sound readers. Intermittent-motion scanners can't really do audio except by image extraction which obviously won't work for magnetic sound. Note that the Director which is intermittent motion and full R/G/B sequential can't do magnetic sound, but the ScanStation can. So that's one advantage of continuous-motion. I don't know what the normal industry terms are, I'm outside of it - but projector loop, film transport module (see photo), sproket drive, capstan drive, direct drive, intermittent motion, and continuousness motion would be the terms I would use. As mentioned above that one transport module does 6 different film gauges all at once - that's a significant difference between it and a professional device from 2000. A Rank Cintel in 2000 could do 35mm, 16mm, and perhaps 8mm but most of them wouldn't have been capable of 8mm. No Euro-Formats etc. You had to feed it low-contrast film only or you'd crush the detail - so negatives and what are called "TV prints" or low-contrast prints for telecines. Now you can put normal contrast and even really contrasty prints through the modern scanners and they have the ability to get excellent shadow detail. That's important for archives, libraries, home movies, and even professional restorations where you might want to scan a reference print for color grading the camera negative, or where you need to use a print because the negatives are gone or are incomplete. My best, Daniel

Your scanner runs on Windows 10 so you could always use its computer. It might be possible to write a Python script for use in Fusion on MacOS, but I'd hardly say that would be easier than using Avisynth+ for which there's a very good front-end, and plenty of peer support. For any other alternative ideas the best place to ask would probably be on the Video Help forums. The issue you have is that there's really only so much you can do with ffmpeg, Resolve etc. What you're wanting to do is simple in terms of writing it out as a flowchart or pseudocode, but a bit complicated for most video editing software to handle. On the other hand pseudocode can be directly translated into a workable script. See example pseudocode below. Load clip clipA = crop frame A from clip clipB = crop frame B from clip clipC = crop frame C from clip clipD = crop frame D from clip Interleave frames from clips A, B, C, D. Output result

It should be relatively simple to do it in a scripting language (ie Avisynth+).

Says who that image stabilisation belongs in post-processing? There was a trend away from the "telecine process" where the transfer goes straight to broadcast tape/finishing format. Most of the pin-registered machines don't even show the perfs in the scan/transfer either as they used the full width of the sensor for the picture. Image stabilisation in-scan is part of the optical pin registration, LG calculate how much the frame has moved between frames to correct for the next capture. The better the optical pin registration is the more stable it should be in-scan, requiring you do less in post. The number of different audio formats (whether hardware or software based) also gives them an advantage over competitors. Software audio extraction on a LG is better quality than AEO Light as you may have discovered if you've done some tests. They're not a film restoration software company, they're a scanner manufacturer. Arri and DFT also don't write Kodacolor decoding software. Doing it the way you're thinking would bake it in to the scan, meaning you can't go back to the start and start-over if you discover a better process in the future for the decoding.

It's not really their job to offer software that does things in the post-processing step as the scanner manufacturer. A $20,000 software licence to make this? Not sure who would buy it. It's good you identified the person working on it, that's where that work belongs. In the hands of someone capable taking it on as additional work for the sake of preservation.

Oh agreed. This is actually a 2015 re-enactment of the telecine process for television: With plenty more information about it on the website here: https://www.adapttvhistory.org.uk/16mm/telecine/ Well that really is all they could use back in the 1980's-1990's for the "home movie" market. Telecine prints for television were printed specially low-contrast so that those Rank-Cintel telecine machines like the one above could do a broadcast-quality job. The Elmo systems only had to be basic VHS quality. Anyway, the market moves on. Maybe someone will see this thread who has "scanned" their home movies with Got Memories and will send you some for a re-scan. 😛

Single-chip CCD I believe. That's what everything low-end used. Ha, not surprising! There's way more dynamic range and colour density in GOOD film than people realise. What you can do (if your workflow allows) is deliver both Prores + h264/HEVC. Well he's been transferring home movies exactly the same way for 22 years, so what was standard 22 years ago (2002) he seems to think is still (or should still be) the acceptable standard today. That's why I mentioned Rank Cintel as doing that stuff "professionally" back in 2002 would have been impossible, and prohibitively expensive even if you did manage to find somewhere that could actually scan 8mm on a real scanner. The TVTs were just there to fill a gap in the market until the professional tech improved and came down in price to allow the "professional" market to properly expand into the Archive/Collection/Home Movie work. Got Memories knows that their quality isn't great by today's standards. There's a video where that Phil guy admits it directly: Once you go above $5,000 which is at the top-end of what some of those Elmos and TVT-HD units cost (as well as older Moviestuff units) there's very few people willing to spend the money. Especially while they're still getting work from customers paying $400+ per order at the moment. $400 is a lot of money for individuals, and for what they're spending they should be getting real professional work not work that was "acceptable" for their market back in 2002. He's not doing any prep or cleaning prior to loading the film, he could at least attach Film-O-Cleans in-line but he hasn't. In fact we probably should have mentioned this way back with the first video and that horrible contraption! My guess is that he just would not want to deal the long threading paths and different process of professional scanners and having to actually build up onto the 400ft reels that he seems to detest dealing with, and that he just wants to get to retirement with the lest amount of expenditure on his ageing equipment as is necessary. Everything he's saying about 400ft reels in that other video, take note that he's putting them back onto plastic reels that look to be in good shape - is he also doing the same when customers send in rusted metal reels and putting the film back onto those? Probably, but he doesn't upload a video showing that. If you have access to a professional scanner you can make a DIY device that can come very close or even match the quality of the professional device, with the advancements in the tech it's relatively simple compared to complicated old devices. However when people build things or they have some of that lower-end tech most of the time they will never really see what they're missing if they don't know what a professional transfer would look like or if they don't have a good idea about how film should look. As Joerg Polzfusz mentioned here, a lot of the assumptions that had been made 30+ years ago about "getting all the dynamic range out of the film" turned out to be incorrect. Here is an example: "In order to provide as much creative freedom in the digital world as is possible in optical printing, it is necessary to digitize the full range of the original negative film. This allows the digital data to be 'printed' up or down without compromise in post production. It also allows color grading decisions to be made in context with elements of the final composite. The full dynamic range of a motion picture negative film can be captured in a 2.0 density range." Really the full dynamic range could be captured with 10-bit log by 1992? Note the information came directly from a professional transfer device manufacturer closely affiliated with Kodak, so a lot of that really old information came from "official sources" and from within the industry. If you had a time machine in 1992 and you could travel to 2024 and do a top-spec scan and then go back to 1992 and compare it, then you'd know that those old assumptions were incorrect - but they couldn't see it for themselves in the 90's because they didn't have the tech we have today.

Well basically yes, but I think you're missing the point of the Elmos/Tobins: they are supposed to be plugged into a VCR or a DVD recorder directly so you can record straight to NTSC (or PAL) with no computer involved. That's what they were always supposed to do, and they do it in real-time unlike a comparable "frame-by-frame" Moviestuff doing 1fps or whatever the speeds were back then. What Got Memories is doing with them is not how they were designed to be run - that's the whole point. They're putting them to a digital file first on a computer that is hard interlaced to NTSC. There's a German company that sells converted projectors using DLSRs for capture and they are cheap at about half the cost of a TVT-8 all run on free/open-source software. They bought-out the rights for those projectors it seems, unlike other companies that converted projectors in the past. That goes directly to a frame-by-frame digital file with no interlacing in real time, but of course they're still converted projectors using DSLRs instead of machine-vision camera so they'll have their own limitations. But for a broke hobbyist that just wants something basic they can use themselves they're OK for the price. Interesting he says "yes I'm in it to make money" in that video. Typical art of the grift. Publicly shaming a competitor to make his very low-end company look better (according to Reddit his videotape transfers are no better than his film transfers). Real companies don't need to shame their competitors. You can see he has it plugged into a DVD recorder (as well as a Macbook) using composite video. He says "this is more than suffice for the average consumer". In the 1980's-2000's that may have been correct, but who is he to decide what is adequate for the "average consumer" now? Clearly this Phil guy knows that people are dissatisfied with low-quality work as someone sent him that film hoping for something better than what the last company did. He puts out a LOT of misinformation. "Reels are spliced together ..." So what? For an actual professional company with a LaserGraphics or a Filmfabriek HDS it's not viable to thread tiny 50ft reels individually onto them as they take far more leader than 8mm/Super8 usually have so of course they have to be built up to 400ft or larger reels. They're not self-threading projector motors like a TVT or Elmo. This idea that they can't be broken down afterwards is his own invention. I discussed this last year with a friend of mine who actually does this work, here's what he said about one job: "I recently did a Super8 job with about 2.5hs of footage on 50ft reels. The customer wanted them off the metal reels so I combined them onto new 400ft archival reels and I didn't have to wind them back onto the 50ft ones, which was nice. I labelled them so the customer knows where everything is." If they customer wants them broken back down then they get broken back down. Look how rough on film his machines are and his handling of his customer's film. He is also saying that his competitor's website is a scam because it has a fake 25% persistent discount, well doesn't his website say the same? 50% off - $20 per reel "after discount" with a $400 minimum order. Oh wait, this page says $15/reel "after discount" and as little as .12/ft. He blames the customers for not finding a better company. 16:30 in while he's complaining about the leader tape (which he should have removed BEFORE putting it through the Elmo) he snaps half a foot off the customer's film leader and presumably throws it away. So I wouldn't say he's being particularly careful with the film he's transferring. Towards the end he contradicts what he said at the start and claims "I don't need the business, I don't need the money". If that's the case why doesn't he just send the customers to a real professional? He also says his purpose is to "raise the standards" (of his competitors) - if that's the case why's he still transferring film the same way that he started 22 years ago? He is also telling people that the sound heads cost a fortune and are super rare, yet the German company I mentioned earlier sells their kits equipped to do sound and they're as cheap as anything. As for everything he says about reviews and feedback, I'll tell you a secret: he deletes any YT comments critical of his work. Go on, try it. 😉

No, this company uses a projector to transfer 8mm. They project it onto the wall and film it with a camcorder or something. 2005 - 19 years ago is when those TVT-8's and super-8 model transfer systems were made available. How much did scanning home movies cost on a Rank-Cintel with an 8mm gate in 2005 (was that even possible)? LaserGraphics didn't make an 8mm scanner until 2013 and Filmfabriek didn't exist either until 2011. If you read Clive Tobin's old website you clearly see it says "Replaces Elmo Transvideo". That AVP guy is the average type of person using a Retroscan to sell scanning services to "hobbyists shooting film" and to home movie clients. He thinks it's good enough, and as long as the service provider think that the service they're delivering is good enough in quality they don't care about matching professional work. Most of them do not understand how much better the scan could look because they've never actually put their test film through a professional scanning system to check how much in the way of colour and detail they're missing. Some of them probably only really do occasional film transfer jobs as well. Most small mom-and-pop places including small film labs are too scared to pay $35,000 for a scanner let alone $60,000. There are people that say publicly that Filmfabriek charge too much! As long as you have places like AMB Media LLC ("Kodak Digitizing Box"/"LegacyBox"/"Southtree") and that Black Lab Imaging place I linked to doing even worse quality work you'll still have the people doing transfers on their 20-year-old TCS TVTs or their 10+ year old Retroscans etc. On the topic of cleaners, they're certainly not going to spend $50,000 for a Hydra if they won't spend it on a scanner!