They say it is on small reels so I'd asume its film?

Thanks

Then you need a film transfer system. You can either buy one, around $1500 to $2000 or more (not counting camera) , or you can farm it out to the many film transfer companies available.

Dave T2

What I did was project it onto a white wall, use your camera's fireware(oassthru) to connect directly to the computer and use Vegas or SClive to capture the images directly onto the pc.

You will need a 8mm projector, a camera with firewire and a pc with fireware.

I hope this helps.

Telecine

You can also search under my name for "Workprinter" which is one way to go. You can also go over to MovieStuff and look at not only the Workprinter (which is a slow, but accurate way to do transfers) but the Cinemate, which transfers at full speed. I have also invented a new method of transfer, but it isn't ready yet to be fully described.

An often overlooked advantage of Moviestuff's equipment is that they employ a widened film gate for the 8mm/Super8 transfer machines.

You can capture everything out to the sprocket holes. The old projectors had a gate that cropped perhaps 15 % of the frame size, so you may see things on the edges that were never seen before.

As johnmeyer says, the Workprinters are sharp, frame-devices. Since all my transfers are attended, I need the real time transfer of the Cinemate (which still gives a very good transfer).

More and more of my customers are having me transfer to an external HDD, which obviates the need for a camcorder or deck to read in the files from tape.

One very important piece of advice - if you decide to do the "shoot it on the wall technique", be sure to lubricate the film and be sure the projector is cleaned and inspected to avoid scratching the film.

RalphM

Say Ralph, do you know enough about projectors to know what, if any, adjustment can be made to make the film more stable as it goes through the gate? I have invented this device that shoots the video directly from a shutterless projector, via an aerial lens. I set the shutter speed on the video camera really high, which gives me two clean fields (not frames, but fields) of each frame of film. I then get two duplicate fields of each frame of film which I combine into a single video frame. I also get fields that show the film in motion between frames (because the shutter is removed), or which are a third duplicate of the same frame of film. I remove these garbage or excess fields using a variant of IVTC software.

Here's the question. In examining the two fields taken from one frame of film, while the film is stationary, the two fields do not line up horizontally. It looks like, after the claw has pulled down the film, there is some residual horizontal movement. (Yes, horizontal, side-to-side, not vertical). Perhaps I need to adjust the claw, or perhaps there is an adjustment in the gate, or perhaps this is just the nature of the beast and most people don't notice because they don't have freeze-frame captures.

The projector is an Eiki SSL-O 16mm sound projector.

I am actually scanning 8mm film to my computer converting them into avi files and than they are edited in vegas as a normal avi file. It takes time, but very simple and effective as you have frame by frame. What you need is a good negative scanner ( I have the Epson Perfection 4990 photo scanner, my regular scanner) and SmartSoftware Italia, http://smartsw.netfirms.com, the only software converting scanned frames into avi files.The program 8mm2avi has been developed as a tool to acquire and convert regular 8mm and Super 8mm to AVI format. http://8mm2avi.netfirms.com. I am scanning 38 frames at a time and have no problems with the sofware "8mm2avi" Use your own imagination for a setup, if you need any information about my set up ,here is my e-mail: jhirschi@ibl.bm Download the 8mm2avi try it, the final version is $29.00 to get a registration code.The http://8mm2avi.netfirms.com explains the whole procedure. Tell me what you think about.

John

How long does it take to convert a 3 minute 8mm reel?

3min = 180 sec. X 16 frames =2880 frames. I scan 38 frames at a time, so 2880 divided by 38 = 76. 76 X 45 sec. one scan equals 3420 sec. or 57 min. It also depends the speed of the scanner.

While that is mighty slow, I thought it would be much worse. I assume the results are spectacular, although I wonder what resolution you get.

Let's see, 8mm film is 4.8 x 3.5mm which comes to 0.19 x 0.14 inches. I assume you must be scanning at 2400 dpi, although I don't know how you can get the scanner to scan 38 frames at that resolution in only 45 seconds. I guess you've got a better scanner than my old HP 7400C.

Anyway, at 2400 dpi, you'll get 0.19x2400 by 0.14x2400 or 456 x 336.

Hmmm. Is my math wrong? That's rather low compared to what I get with my Workprinter, which lets me use either my DV camera at 720x480, or my HDV camera at even higher resolution.

However, I have posted here before about the "resolution" of film, and if you start with the assumption that 35mm still film (36x24mm) has a resolution of 12 megpixels, and then ratio that down to the 4.8x3.5 mm 8mm film, you end up with a "resolution" of 8mm film of about 0.23 MPixels. Since DV is 720x480 = 0.35 MPixels, it is sufficient for 8mm, and is just about right for Super8. Your approach yields 456 x 336 = .15 MPixels, which is not quite enough, but probably produces excellent results.

Somehow I thought your approach would produce more resolution, so my math must be faulty.

johnmeyer,

I'm familiar with "gate float" but I always assumed it would be vertical in nature rather than horizontal. I use a converted EIKI NT-1 for 16mm captures and the film gate seems to have a well defined slot for the film to pass through.

8mm machines seem to have much less control over the horizontal movement of the film.

How much horizontal movement do you estimate you are seeing?

Don't forget, it is the gate float that is partially responsible fo the soft focus we have come to associate with "the film look". If you get the image too sharp, someone will think it's video, LOL...

Will be out of town for a few days, so don't take lack of further feedback as lack of interest.....

RalphM

Ralph,

Thanks for the answer. I am looking at the service manual at this moment and will spend twenty minutes tinkering. If I can't find a hardware solution, I'll fix it by using motion estimation software to line up the the fields and then mask the edges. Amazing what you can do with the software that's out there for AVISynth.

RalphM You said something about lubricating the 8mm flim? What do you use to lub it? New to this sort of thing and doing some 8mm from the 30's to mid 60's. Also for spliting films what is the best glue for pasting leads? Love this form even if I only understand about 50% of what everyone's talking about. Pushing 60 and the learning curve is dropping fast LOL.

Splicing film: you don't need glue, you need film cement. Or splicing tape. And a film splicer. Film cement probably isn't widely available but a Google search for "cine film splicing cement" will turn up a number of suppliers. Splicing by tape is far more common (but you still need a film splicer) and easier for the rookie to do successfully.

Lubrication doesn't involve oil. Usually done by labs and film libraries. Kodaks will be able to supply solution as well as solution for cleaning. Lubricating the projector gate must be done most sparingly and it is enough to wipe the gate with "nose grease"; wipe the side of your nose with your finger and rub your finger on the gate.

Getting steady images involves friction. Film is pulled down by the claw and let go. It is essential that the film stop there and not slide along further. Generally the claw mechanism provides acceleration and deceleration to the film, so it is stationary when released. However if the gate pressure is too light or the film too slippery then you get jittery images (watch the frame line). Film is held horizontally by a fixed guide on one side and a sprung guide on the other. Make sure the gate and claw are free of emulsion "corns".

If the machine is clean and free of hard accumulations then there is nothing to scratch the film. Never pull the film tight on the spool. This will cause the film to slide over itself and any dirt on the film will cut short grooves. Film must be spooled firmly (so no sagging on the spool).

Kind of interesting that this has come up again as I'm currently scanning 35mm B&W slides. Haven't a clue what sort of film was used but boy have I had some problems with it. Whatever is in the emulsion, parts of it are opaque to IR which sends the ICE system nuts. Turned off ICE and all is well, just have to put up with the grot!

But I digress. I think using optical stabilisation is definately the way to go. You've got two sources of wander in the film, one in the projector and the other in the camera. I'd have thought in the case of 16mm the cameras would be fairly well built but some of the 8mm cameras were pretty cheap and nasty affairs. One problem I found with some of the S8 cameras was the auto exposure system would hunt introducing a very slow flicker, what a pain.

Other thing I noticed from my days of transferring 8mm was a lot of the processing was anything but ideal. At PAL SD res the grain was very noticeable. Of course back them film emulsion wasn't what it is today, that's why 70mm was used, advances in film emulsion killed it off. I only mention this as I think it'd be a mistake to estimate the resolution of 8mm film stocks from 30 years ago based on the figures for todays 35mm camera neg.

As a comparison I've scanned some 3 year old 35mm slides, a full body shot of a 'model' can be blown up to the point where you could count the pores on her skin and still no sign of grain. The 30 year slides even with grain reduction is pretty horrid to very good. The old cameras do take a beautiful image but it's sure pretty low in resolution.

I'd love to have the time to try out the new emulsions in S8 or Pro8, I believe Pro8 blown up to 35mm holds up OK.

Probably the best fim scanner would be a modified slide scanner, there's a few older model Nikons going fairly cheaply on eBay from time to time, they've got a clear front to back film path so probably all it'd need is a modified film strip holder and some stepper motors to automate the whole process, figure on about 20 second per frame though.

Bob.

Thanks serena I have picked up a splicer and it works great. i have search for film glue with no results. Now if I pop in cement I'll get some were. The slicing tape sound like a good way to go also. Again thanks for the in depth answer.

Why not use YESVideo (www.yesvideo.com)?
I signed up because I didn't want to buy the projector and all that stuff, spend time taping the film into video, etc...I'm busy enough doing other things that nmake money...not costing me money.
YESVideo has a quick turnaround time and the costs are reasonable.

Make sure the film glue is Eastman 910 or Dupont 901. These glues actually melt the film together. Scrape the emulsion off prior to gluing and it should last forever and also provides a much smoother transition when passing through the film gate. If you don't think these glues were strong enough they used them to glue fibreglass parts on the wings of aircraft. How do I know? I worked with a boatload of ex Boeing aircraft people and listened to their quality control stories for hours.
JJK

You said something about lubricating the 8mm film? What do you use to lub it?

I'm not Ralph, who probably knows far more. Here's a link that may provide good answers: Cleaning and Lubricating Film.

From what I can tell, lubricants are really for hiding scratches in damaged film that is run hundreds of time, like in a movie theater. It is not clear to me whether it is generally a good idea for home movies that have probably been shown less than a dozen times and will only be run through the projector one more time, for transfer to video. However, you definitely want to clean the film.

Bob,

Interesting stuff. I just scanned over 30,000 slides and negatives for one single project. I can tell you for absolute certain that virtually no B&W emulsion can be cleaned with ICE. This is also somewhat true of Kodachrome. However, if you use Vuescan, it can sometimes remove dirt from Kodachrome, without introducing too many artifacts, although you have to look carefully on each scan to make sure you don't start losing details (all the windows in a distant skyscraper suddenly disappeared in one scan -- although the dirt was gone as well). Fortunately, B&W is very easy to re-touch, at least with Photoimpact, which is the photo editor I use. I actually re-touched almost every one of those 30,000 slides. Took many months.



You've got two sources of wander in the film, one in the projector and the other in the camera.

Thanks for all the information on sources of film wander. While I was perfecting this invention, I kept running the same old reel of film from my dad's estate. It is actually a commercial featurette that HIS dad purchased on a 'round the world trip that took him through India. Thus, it was shot in India, probably pre-war, on goodness-knows-what equipment. It has plenty of movement, frame to frame, which is definitely in the camera. However, the thing I am seeing has to be on my end, because it is movement from field to field on the video (which I can see by separating the fields in AVISynth and then examining them in VirtualDub). It happens on some frames and then not on others. I suspect it may have to do with film shrinkage (this film is from the 1930s). This shrinkage is uneven, and probably causes the film to slide as it comes to rest because it is not making uniform contact with the gate edges. At least that's my theory for now.

I removed the lens and have run the film by hand using the manual advance wheel, and can see nothing unusual happening with the pulldown claw or gate. Of course, I'm sure some things only happen at speed. Everything is meticulously clean.


You mention flicker, and the flicker on this old stuff is horrible. I'm working on building some filters (using things available for Virtualdub), but so far the results have been too subtle to be of much use). This is flicker in the film itself, not flicker from the film to video transfer (which doesn't exist because of the way my film to video transfer works -- each frame is captured exactly, and there is no shutter in the projector).

As for my 35mm "resolution" calculations, I used a figure of 12 MPixels, but I've seen figures that are double, and even close to triple, that number, so I agree completely with your statements. Film "resolution" depends on a lot of things other than just the size of the film, including not only the emulsion, but the camera, the lens, the lighting, the processing, etc.

My modified projector is ready to go. I made a test DVD last night and the results look quite good, although I'm still getting a little bit of a hot spot on one side which is either due to my diffuser arrangement (I replaced the 250W bulb with a 25W bulb of the same design, and then put some frosted glass in front -- I'm shooting directly back at the projector, through the aerial lens that I'm "borrowing" from my 8mm/Super8 Workprinter).

It sure is nice to be able to get individual frame captures, but at 24 fps. Can't beat that speed/quality trade-off. Thus, with this invention, I can "scan" 24 frames each second, and in HD resolution, if I choose.

I really can't imagine scanning 30000 slides. That would take me until 2050. After number 500 I am sound asleep.

JJK

I really can't imagine scanning 30000 slides. That would take me until 2050. After number 500 I am sound asleep.

One of the toughest things I've ever had to do.