Firstly it isn't uncompressed!
If you truly want uncompressed a 16bit tiff or dpx sequence would be the norm, around 2TB/hour depending on res. Audio would have to be as separate file.
You haven't mentioned how this footage was shot, if it was 24fps I assume you're getting it with 2:3 pulldown as 60i?
Is this old S8 or a movie you shot on a modern camera with modern stock?

The BMD codec is pretty much the same as the SonyYUV codec but I'd have to defer to Glenn as to just how Vegas will treat this in 32 bit. Certainly it should open OK in V8 or V7.
To get RT playback you'll need a RAID 0 disk array.

Bob.

Thanks John & Bob for the quick replies.

The S8 if from footage shot back in the late 70's on an old Canon Super 8mm camera (w/ sound)
The outfit I am using for transfer claims it is completely uncomppressed.
See here http://www.mymovietransfer.com/Serial_Digital_Interface.html

Using Rank Cintel Turbo II Telecines.
I wonder if I stick to their regular DV format if It will be easier to work with and if I notice any final quality loss.

Ted

Before you do anything, have you seen this film?
It could be worth getting the very best transfer possible done or not, really depends on how it was shot, how it was processed and how it has been stored. If it smells of vinegar you might have a real problem.

Aside from that. As you're working in NTSC then the 4:2:2 part will give you more to work with than NTSC's DV 4:1:1, moreso when going to DVD's 4:2:0, having 10 bits compared to 8 bit to work with gives you more also if you're planning on grading the footage and you have a pipeline that can use it. On the other hand if the film is in really bad shape nothing is going to bring it back from the dead.

We did a job a few years back transferring regular 8mm as the client only had a S8 gate for their telecine. We transferred the R8 on a pretty worn out Elmo telecine to DV and dubbed that to digital betacam. Our work was intercut with the S8 from their telecine. This was quite well shot footage taken by a pro with a good camera. I sure couldn't see the difference when it made it to air. One caveat though, we were working in PAL, some things get a lot different in NTSC.

The company you plan on using are charging afforable prices by the look of it. If you can, get the footage on HDD and Digital Betacam tape(s) Two reasons. a) DB tape is extremely robust, it'll outlast your HDD. b) If you really have a problem handling what's on the HDD any dub house can dub it to DV for you.

When they say "uncompressed" they probably mean relative to what comes out of the Cintel i.e. no further compression.

And take on board what John is saying, getting back to the original frames can help, having to wrangle interlace problems only adds to the task.

Bob.

Ted:

I tried those guys and the Video Conversion Experts and the Video Conversion Experts guys transfer came out a little better. Of course it's more expensive.

It seems you are trying to make a best effort on this and I believe the raw capture will make the biggest difference more than anything. The Video Conversion Experts supposedly use a hardware frame by frame capture method using adapted equipment the studios transferred all the old films to digital with. Something like 20 CPUs are in those things.

Bottom line though was it still couldn't revive some of the films I sent in. There is also a lot of inherent inconsistency shooting 50 foot reels of film as they are shot on different days, in different light, etc.

The guys here are right on that you don't want to introduce any interlace, jitter or flicker problems. You are bound to have plenty of that anyhow.

Well, I looked all over their site, and I didn't get my main question answered. It is more important than your question about compressed vs. uncompressed. To whit: Have they added pulldown? They don't answer that question. You need to ask them. It is VERY important.

The good news is that they are using a Rank Cintel. This is the ultimate film transfer equipment, as good as it gets. It transfers the film one frame at a time and, if you specify that you want to receive your video without pulldown, it will let you edit the film in native 24p.

You absolutely, positively want to do this, and it is far more important than whether or not they compress the video.

Now as for compression, since you are starting with Super8, compression, per se, will probably not degrade it much, since there isn't a lot of detail to begin with. However, that last statement depends entirely on the what sort of compression is used. Since you are going to take delivery on a hard drive, you are not limited to just the two choices of uncompressed vs. DV compression. As pointed out by others already, DV compression has yucky color space issues. The compression itself (the JPEG part of it) really isn't an issue, but the colorspace is. You should find out what other formats (if any) they can deliver. If it were me, I would want the result delivered in 24p using something like Cineform (which can do SD) or HuffYUV (lossless, but compressed) or something else that will not only reduce the disk space required, but as already pointed out, will let you play back and edit with full 24p framerate, without having to have a RAID, something that most people outside of the high-end HD (not HDV) industry should ever need.

You guys are reminding me that I have thirty rolls of 1930s 16mm film that I was supposed to transfer last week and I'd better get to it ...

I really appreciate all your input on this.
Jeff, are you saying that the Video Conversion experts have a progressive output?
Listening to John & Bob, sounds like that would be the best way to go, Want to stay away from interlaced.
I will check out the Video Conv. experts.

Ted

Two items to be consider:

Professionally shot Sound super 8 is usually 24 fps. However, amateur (home movie stuff) is much more likely to be 18 fps. For example, I use a Super 8 sound digest of "Showboat" to set up my transfer machine. It is 24fps, produced and distributed by MGM. However, the amateur sound film that I'm transferring right now is nominal 18 fps. I've seen very few amateur sound S8. but they've all been 18fps.

Also, since Rank Centel is frame accurate, how are they resynching the sound to the frames?

yes, what I have is amateur S8 at 18fps with a sound strip along the side of the frames. I also wonder how they sync that sound to the video if they transfer 1 frame at a time, which I assume isn't real time ?

You'll have to ask them how they sync the sound. It should be very easy for them to do, because I am sure they know exactly what the offset between the mag sound strip and the film should be for that film technology. Knowing that offset, they can line up the sound start with the film. It would be another good question to ask, but I would guess that given the equipment they are using, this should be the least of your problems.

Remember, you want to receive the transfer in a file format that is progressive (since film is, by definition, progressive). Edit it as progressive and make sure that Vegas shows the media as progressive (change this if it gets it wrong). Your project properties are progressive. Then, if you encode to DVD, use the pulldown option and encode as progressive. You should get stunning results.

This does raise an interesting question though, one that I've never worked out. If the original film was shot at 18fps how do you pull that down to any video frame rate?

Bob.

This does raise an interesting question though, one that I've never worked out. If the original film was shot at 18fps how do you pull that down to any video frame rate?

Good question. I should have reacted to the earlier post about amateur sound being 18 fps instead of 24. I haven't dealt with Super8 sound, so I haven't had to deal with this.

The answer, of course, is that you cannot use the pulldown flag for DVD. The DVD spec, unfortunately, only specifies one type of pulldown, which repeats fields in order to achieve the correct playback speed when the original material is 24p. Whenever possible (i.e., whenever dealing with 24 fps film) you want to encode at 24p and use DVD the pulldown flag (so the DVD player adds the pulldown during playback) because the encode is infinitely cleaner (because you are encoding only 24 "things" per second instead of 30 frames, actually 60 fields -- since you only have so many bits per second, which is your encode rate, a LOT more bits are available to encode each frame). It is not a subtle difference at all.

However, for 15, 16, or 18 fps (standard silent speeds), you are stuck with adding your own pulldown. You have to do this before you encode, and you then encode at 29.97 (again I am taling about NTSC here). You can do this with Vegas by specifying a 29.97 progressive project, and then disabling resampling for each piece of media imported (this assumes the media header has been set to 18 fps progressive and that no pulldown has been previously added).

However, I have never liked this result, although it probably would suit most people. Instead, I frameserve out of Vegas into an AVISynth script I wrote that applies exactly the pulldown I want. I spent a lot of time looking into this and trying different things. Unlike 24 fps where the 3:2 pattern is the only sensible alternative, when you get to these slower speeds, you have more options.

For 24 fps, I use this script:

# Pulldown for 24 fps
separatefields()
SelectEvery(8, 0,1, 2,3, 2,5, 4,7, 6,7)
weave()
AssumeFPS(29.97, true)

This is the traditional 3:2 pattern. You take field 0 and field 1 from the film and combine those for the first frame of video. You do the same with the next two fields of film (yes, film doesn't have fields but the video of that film -- even if progressive -- does have fields, but they just don't have any temporal displacement from each other, just spatial.). Then the fun happens. You repeat field 2 and combine it with field 5. This may seem like too big a jump, but it isn't. Normally 4 would go with 5. So, at this point we've repeated one field. The next frame of video is made up with field 4 and 7. And then finally we get back to fields that "belong" together in the final pairing of the sequence by combining field 6 and 7. The key is that you always have to combine an even and an odd field, and the odd field has to be second. This is because even and odd fields occupy specific horizontal places (weave your fingers together like you are praying and note how, if you are right-handed, your right thumb is on top). If you put the fields in a the reverse order, you get that familiar "studder" that happens when you reverse the fields.


For 18 fps, you can actually just repeat ever third frame and you get what you want. This looks like this:

# Pulldown for 18 fps using repeats normal-repeat-normal-weave-normal
separatefields()
selectevery(6, 0,1, 0,1, 2,3, 2,5, 4,5)
weave()
AssumeFPS(29.97, true)

While this is really clean, it tends to accentuate problems when the camera pans. Therefore, I like (and use) this pulldown:

# Pulldown for 18 fps using all weaves normal-weave-normal-weave-normal
# (I like this the best)
separatefields()
selectEvery(6, 0,1, 0,3, 2,3, 2,5, 4,5)
weave()
AssumeFPS(29.97, true)

And for 16 fps, I use this:

#Pulldown for 16 fps using all weaves
separatefields()
SelectEvery(16, 0,1, 0,3, 2,3, 2,5, 4,5, 4,7, 6,7, 6,9, 8,9, 8,11, 10,11, 10,13, 12,13, 12,15, 14,15)
weave()
AssumeFPS(29.97, true)

I have not had to deliver 16 or 18 fps for PAL, but if anyone needs that, I could develop a pulldown pattern for that, if anyone is interested. You obviously cannot do the usual trick of treating 24 fps and 25 fps as the same, and just re-timing the audio, which is my understanding of how film is usually treated for PAL.

Ted:

The VCE guys can offer 720P for progressive.

However, I don't see 18 FPS being served to you as progressive. You will be served in 29.97 by their equipment. Speaking of equipment, I looked at the Cintel site and they are certainly the film scanner leader. Their high end equipment is $750,000, before adding their accessories at $50,000 a pop.

Here are typical m2p file details for my 18FPS projects scanned:

File Details

Here is a 12 second sample of what I consider good 1960s 18 FPS video:

18 FPS Video Sample 40 MB file

Thats about all the color and contrast you will get without PP when you get it.

Jeff:
When I look at the VCE site, the cost goes up quickly when they put on their "hollywood restoration". To me it looks they are doing color correccting for this. Could I get the same results by going with the less expensive transfer and add the CC within Vegas.
In the clip you are showing, which VCE process did you use?

Let me share my limited experience with transferring amateur super 8 sound. The amature super 8 sound is not very sophiscated and the frame rate is far from exact. The frame advance is slaved off the audio track drive in both the camera and the projector. If you choose a specific frame rate then you will have to stretch or shrink the audio track to match. That is okay for small changes. I worked the other way around. I transferred ~4000 feet of amature super *mm sound using Roger Evans' WorkPrinter Pro. Before I transferred the film I spliced 1 frame with an audio blip at the beginning and end of each roll. I transferred the movie on the WorkPrinter and the audio on a sound projector. I imported the audio and video into Vegas and stretched the video (creating pulldown) until the visual audio blip frames matched the audio blips (resample off). Then I moved the audio track back the offset between gate and audio head (super 8 amature is 12 frames if I remember correctly) and it was in perfect sync.

Ed B

Very interesting. Could you elaborate a little more about how you acquired the audio? How did you get it from the projector into gthe computer?

How did you get it from the projector into gthe computer?

Some sound projectors have a line-out which you can use directly. Others, like the Kodak EktaSound, have an aux speaker out where you need to use an attenuating cable or a resistor divider to lower the signal amplitude. Plug the cable into your audio card and record with Vegas. You will most likely will want to do some noise reduction in Vegas to remove the camera noise, hum...Remember this not today's electronics. :)

Ed B

ECB - good tip for synching the sound back to the video.

For real-time transfers (using a modified Sound S8 projector), I've found that a fairly simple method is to just adjust the voice pitch via Sound Forge. While the dialog is speeded up, it usually sounds natural unless the original camera was running slower than nominal. If that does not work, then it becomes necessary to stretch the video and audio in Vegas with "maintain pitch" turned off.

For you experts - How does Vegas stretch video? Does it interpolate between frames, or does it duplicate frames to create the length needed?

RalphM

How does Vegas stretch video? Does it interpolate between frames, or does it duplicate frames to create the length needed?

If the video event Properties Smart Resample is checked the frames are intrepolated. If Disable Resample is checked the frames are only duplicated.

Ed B