Try using a velocity envelope to deal with the speed issue. I've had success with this.

Also, if any of the film has sound that is distorted by the differences in speed, the Vegas utility to change audio speed works very well. I just did this for a client whose 16mm film had badly distorted audio due to the difference between the fps at which the film was shot and the fps at which it was transfered to digital tape.

Jack

If you want to post a short (5-10 second) clip of exactly what you're dealing with, I'd be glad to take a look at it, and see if I can come up with further recommendations.

I do a lot of film transfer (so does "farss" -- he is actually FAR more experienced than I -- so he's a great resource as well). Here's my YouTube page, which you will see has dozens of film transfers, from 16mm, 8mm, and Super8, mostly from the 1920s-1930s:

John Meyer YouTube

Of course, thanks to the YouTube encoding (hi-res didn't exist when I uploaded these) they all look rather fuzzy.

Also, when changing film speed, you should do a test render of 20-30 seconds and experiment with the "resample" setting. You will find this if you right-click on the media (in the media pool, or on each individual event). I generally prefer to disable resample for all film media. This forces Vegas to do pulldown the way it would normally be done in broadcast, i.e., by repeating fields without creating any new fields. By contrast, if you use smart resampling or force resampling, when you change speed to get to 29.97 for playback, Vegas will create new fields by blending adjacent fields. This does provide slightly smoother playback (do a test encode during a camera pan, for instance), but I think it hurts resolution too much, and it lessens the film "feel." However, you really have to decide for yourself.

[edit] I forgot to mention that when you do gamma correction (i.e. brighten your dark scenes) do NOT use the brightness controls in Vegas. Use the Gamma setting in the Levels fX, while looking at the histogram in the videoscopes. Make sure after increasing gamma to then adjust the black base level, otherwise you'll lose contrast. Also, for the best overall control of gamma, use color curves and create an "S" curve by adding and adjusting 2-3 points at the heel and toe of the curve.

Thanks, I'll see where I can post a small clip in the original format so you can see it.

I disabled the resampling altogether because not only it takes more time to render, also because there's enough blended frames already created by the camera to compensate for the difference between 18 and 24 fps, so by resampling I end up with even more, and that's what I'm trying to avoid, the blended frames. In a panning where there's poles or other distinctive objects, they appear double.

I would like to have that script you mentioned with the instructions on how to use it. Does it work fast, or is it too slow to do the pulldown removal? I have hours of footage, so it might be something for me to consider.

When I referred to a "script," I was talking about a script for a separate program called AVISynth. There are various workflows for working with this in Vegas, and which one to choose depends on the format of your original footage and many other factors. However, the simplest way to use it is to feed your footage into the script and then open the resulting video "proxy" in VirtualDub and encode an intermediate file using either a DV codec (for SD resolution) or Cineform (for high def). Then, bring that into Vegas and go from there.

Here is a script, which includes all sorts of commented-out variations, which takes a 29.97 fps NTSC video that was created using 18 fps film footage, and retrieves that original 18 fps progressive footage. I have used this when I have finished 29.97 footage created from film, and this 29.97 footage already contains pulldown that I added. In order to avoid going back to the original footage (and have to do all the color and gamma corrections, etc.), I run this script and get back to my 18 fps original. Since this is somewhat similar to what you are faced with, it is a good starting point for the IVTC needed on your 24 fps footage (the cycle and cycleR settings will have to be changed).

# Script to IVTC an 8mm movie.

# Note: must serve into VD and then direct stream serve into MC

# For this script, serve out of Vegas using YUYV



#-----------------------------

#Telecide and Decimate are part of the Decomb plugin



loadPlugin("c:\Program Files\AviSynth 2.5\plugins\TIVTC.dll")



AVISource("D:\Documents\Dnload\UNPACK\William\Young Life0002.avi")





#-----------------------------

#Inverse Telecine Using TIVTC (TFM) - for 18 fps John Meyer pulldown

AssumeBFF()

tfm(display=false)

tdecimate(display=false,mode=0,cycleR=2,cycle=5)



AssumeFPS(18)



#ConvertToRGB32(interlaced=false)          #Needed to get correct color when serving into MC.

#Levels(16, 1, 235, 0, 255, coring=false)

I've never worked with AVISynth, so I'm lost on what to do here. I know VDub though, but I don't have a clue how AVISynth interacts with it, or with Vegas. I don't own Cineform, so my only choice as a intermediate codec would be Huffyuv, but I would have to buy an external hard drive to transfer stuff and free up a lot of space in my video hard drive.

Isn't there a way to bypass the intermediate steps using some kind of plugin for Vegas that will do the same job? Like for example, setting up a 1920x1080 24p timeline in Vegas, and drop the footage and have some plugin process it? That would be more time efficient for me.

Unfortunately, the plugin support for Vegas is poor, so there is no plugin that will do this in Vegas.

If you own VirtualDub, all you have to do is install AVISynth, edit my script in Notepad so that the "source" statement points to one of your AVI files. Save that, with th extension "AVS". Then, open that AVS file in VirtualDub. VirtualDub will tread the AVS file just as if it were a video file, except that the video will be processed by AVISynth prior to delivering it to VirtualDub. Thus, it will "feel" just like editing any other video file, although if you use CPU-intensive filters in AVISynth, the playback can be slowed down. Fortunately, because of the wonderful architecture of AVISynth (it is a GREAT program), it is VERY fast. Also, the inverse telecine plugin is close to real time, so you should notice any real slowdown.

Actually, if you want, you can download VFApi (another free utility). You can create an AVI "signpost" file from your AVISynth AVS file, and then open that signpost in Vegas and edit from there. This would avoid having to create any intermediate file or use a codec like Cineform or HufYUV. However, I haven't tried it with high-res material. Also, now that I think of it, I haven't tried the AVISynth scripts with m2t files, so you might need another plugin to read those (although I do know that such a plugin exists, and the latest version of AVISynth may have now incorporated that facility natively).

I haven't done a film transfer in years. The somewhat worn out Elmo telecines I used got sold off to another business as they weren't mine. Not in the least sad to see them go, the cameras in them were pretty much stuffed and Elmo stopped supporting them decades ago.
Roger's Workprinters are pretty good if you buy the top of the line units to do a frame by frame film scan. Using a telecine process that adds pulldown leaves you with a mess, preserving the exact frames as recorded on the film gives you a decent shot at restoration and preserves as much as possible for future efforts.
The best effort I've seen at extracting all that can be had from 8mm is a guy in Italy I think. He built a film scanner out of an old 8mm projector. He uses a 720 line industrial camera and a rather expensive matching macro lens to shoot straight off the film.
Anyone that does come to me with film these days I send off to a local post house that can transfer 8, S8 and 16mm to Digital Betacam. It'd probably be better to transfer to a better codec such as DPX or DNG but that gets expensive. It's remarkable how good 8mm film can be if it's shot on good cameras using low speed stock and well processed. If it's old home movies very likely it's going to not be worth that kind of money but you don't know really until it's transferred.
S8 was a pretty common acquisition format for television in the early days, it looked surprisingly good. In fact 8mm (Pro8) is still used in feature movies.

I know none of the above is helping Sebaz. I'm still pretty angry over what one client did with a S16 movie he produced. Lack of care has robbed everyone who donated time and money to the production of any chance of getting a sizable return on their efforts. You never know what something might be worth one day.

Bob.

Bob,

You should check out what Roger (Evans) has done with his latest Workprinters. I had a ninety minute chat with him a month ago, and his newer technology is absolutely fabulous. And unlike the rest of us, he has made a real business out of this and is making VERY good money. Unlike those Elmo units with their old cameras, his units let you use your own camera, although in his top-end devices, he is supplying everything, and is doing the capture in HD, from a cool LED light source, using even more amazing synchronization tricks than in his original Workprinter.

As for Sebaz, if he posts a few seconds of his capture, I'll see what IVTC tricks I can do to get him close to a Cintel/Workprinter capture. Obviously there will be blended frames, but the two steps will be to first get rid of the 24p to 29.97 interlaced pulldown. That's easy. The second step will be to take that 24p native file and figure out how to intelligently reduce it down to the 18p of the original film without dropping any frames. I think this can be done, and it is of course something I spent a year working on when I developed my own film transfer system where I shoot video directly from a shutterless 16mm projector going at full-tilt 24fps, and then through software eliminate all the duplicate fields and end up with a 24p video file (which I then re-time for 12,15,16, or 18 fps).

John,
not spoken with Roger in a while. I did excahneg a few emails with him a couple of years back and was going to give him a big slab of money. Then commercial reality crept in. Nobody wants to pay enough for me to even recover my investment let alone turn a profit and that's working from home with no overheads.

Got to get back to building my second tape oven. 100s of reels of 2" quad tape are waiting.

Bob.

Tape oven?

If I'd known you'd have asked I'd have baked you a tape:

http://www.audio-restoration.com/baking.php

Bob.

Bob,

Very interesting. I just restored some reel-to-reel audio tape manufactured somewhere between 1949 and 1951. It was very rigid and difficult to get back onto the spool, but it didn't shed much and certainly wasn't sticky. I have not dealt much with tape from the 1970s, since I don't handle professional videotape, although I have a few clients that have what I think are the one inch Sony video reel-to-reel tapes that were occasionally found in the home in the late 1960s through the 1970s and also in small corporate setups. I was hoping to get my hands on one, but it fell through. Probably just as well, since I didn't know about this problem.

Bob, I have used moving dry air,( ie. nitrogen) and low vacuum pressure to remove moisture from tapes exposed to extreme conditions. One may want to check out the heating element circuit to ensure that no freq's and magnetic fields can effect the temporary floating tape magnetic particles.