First time shooting 16mm film

[Andy Jarosz]

So I'd been living my life in a shell, I kept telling myself. I'd shot everything digitally, and though I attended NAB for my first time this year, I still had not dabbled in the world of celluloid. A few weeks ago, I changed that. The scans came in yesterday.

http://vimeo.com/5634134

http://img107.imageshack.us/img107/6461/16mm2t.th.jpg (http://img107.imageshack.us/img107/6461/16mm2t.jpg/)
http://img268.imageshack.us/img268/3408/16mm1.th.jpg (http://img268.imageshack.us/img268/3408/16mm1.jpg)
http://img33.imageshack.us/img33/7278/16mm3e.th.jpg (http://img33.imageshack.us/img33/7278/16mm3e.jpg)

-Bolex h16
-Fuji 250D stock

I was blown away. Even with my mediocre entry level equipment and experience, fumbling through it with no help, the images were still far prettier then I expected. This requires more intense experimentation :).

[David (dudi) Namir]

Why don't you get this:
http://cgi.ebay.com/Arri-Arriflex-SR2-PL-mount-Color-Video-Canon-Zoom_W0QQitemZ250465188790QQcmdZViewItemQQptZFilm_ Cameras?hash=item3a50e37bb6&_trksid=p3286.c0.m14&_trkparms=65%3A10|66%3A2|39%3A1|293%3A1|294%3A50

[Craig W. Bickerstaff]

What was the cost like, film, processing and scanning.

[Andy Jarosz]

Cost was $30 for the film, and around $100 for the processing and scanning. I went for the cheapest place I could find.

Lithops...this was more of a "fulfill a dream" experiment then anything else. I'll buy the Arri when I'm good and ready :).

[D'Arcy Foley-Dawson]

it's very nice footage. not to be a pain, but i found the music quite annoying, until i turned it off. but good on you for shooting film.

the scan looks amazing for the price. where did you get it done?

[Dan Hudgins]

I did a quick CC on your film scans.

The image with the person I think looks better, more like a film print, the JPG only use part of the Histogram, and the images needed masking to bring up the color.

There is a problem with the scaner's light house maybe, as you see in the grass image the hue is uneven, that was also a issue in finding a gray point to pick on the ARRI light since uneven hue in the scanner light source cannot be fixed with simple CC, its best to adjust the light source.

Uneven light is always a problem in scanners and optical printers, but can be delt with if the equipment is adjusted right. One can make ones own scanner with my DANCINES.EXE (tm) program, and CC with my "freeish" DANCAD87.EXE (tm) if one has a PC with FAT32 disks running W95/W98/W-ME, or FreeDOS using the FAT32 big disk option.

Bs16b.jpg is a scan I made of some Super16 film another REDUSER sent me to try scanning, its ECN, without grain reduction.

1ss_1.jpg was run through my Sharp/Soft filter to cut the grain a little, it works better with 35mm images that are in sharp focus so it can tell the image from the grain better, but it is may be a small improvement over 1us_a.jpg, you can decide. Its hard to judge the JPG images, working with TIF you can see the effects of the filters better, and there would not be histogram gaps from the JPG source images lack of data.

I put a pilot tone generator on my Bolex16 so I can shoot sync sound with it, here is how it works, the 8 frame shaft turns 3 times per second at 24fps, so if you put a 20 tooth steel gear on it then put a small tape recorder head with a magnet on it near the gear you get 60Hz out of the tape head (more like 58 to 62Hz, but thats the point of pilot tone is syncs to the film images). The tape head signal then goes into the left channel of your tape/mini disk/wav recorder and the mic into the right channel. Then you wind up and shoot with the tape recording. To resolve the sound to 16mm mag film you feed the 60Hz pilot tone into a 200watt amp and then through a xformer to get 110vAC to run the sync motor in the mag film recorder, the audio goes to the mag film line or mic input. Then you playback the "resolved" magfilm into your computer using 110AV 60Hz line voltage on the sync motor and you get 24fps AUDIO (48000samples/second ends up 2000 samples per film frame). If you have frame scans, then you find the clap board slate and sync that to the clap in the WAV file you made from the mag film. Sounds like a lot, but most of it is automatic, and you can do many shots at once by making a copy of the stereo recordings of the takes you will be using so you can resolve them all as one group without stopping the mag film. Most labs had such equipment on hand for everyday use. You can also resolve to 35mm mage film, works the same. If you have a crystal motor for the Bolex 16 you do not need to resolve if you record to VHS-HI-FI or MiniDISK etc. since any video camera or WAV recorder records crystal sync.

You can also get a 60Hz SYNC motor for the Bolex 16, we have one, but you need to shoot within reach of a AC outlet, or use a crystal sync 12v to 110v inverter.

The attached images are for viewing on a monitor about gamma 2.22 to 2.4, you may need to rightclick and select open in new tab/window to see the 1024x576 size images on your browser(?). You should step back 3 to 6 feet when viewing the images since 16mm was not ment to be blown up to IMAX size screens...

[Dan Hudgins]

Here is a screen shot of my CC's Levels#1 screen showing the adjustments to the clip points and gammas to adjust for the histograms being smaller than the brightness limits and to bring the sholder of the jacket to gray alignment.

Since I did not know the exact color of the jacket, I made it more or less gray, but if you shoot a few frames (enough for the camera to come up to speed) of a 18% gray card, it makes getting the CC centered since you have something you know was true gray to adjust the "probe" area to.

You can see the "probe" border on your frames sholder, its the XOR rectangle.

You can put the probe anywhere in the frame to sample the color values to see if they are adjusted right, ie mostly white, black, and gray. You can mount 90% white and 2% black as well as 18% gray on a card and shoot that, that way you have three points like in the Kodak LAD image. One can see what the 35mm film LAD looks like at my site.

You may need to right click "open in new tab" to see the full size image 1024x768...

Since the scanner is only using part of the histogram it would be better to get 48bpp uncompressed TIF frames rather than JPG to avoid histogram gaps in the processed images for your output.

I did the grading a little yellowish since for a closeup of a face you want enough yellow dye in the print to compensate for the Xenon lamp in the projector being cold K value, it also helps keep the cheeks from going magenta or pinkish-blue. I kept the 16:9 aspect ratio, but the scans do look quite right H vs. V maybe, were the pixels un-square or something? What were the scale ratios used for the scanner sensor area to the 1920x1080 JPGs? In my control screens the ratio is 4:3 unless you have the letterbox settings used.