Dilemma of 4K Restore From The IP or IN Source

[Frank Chang]

Hi All. I posted this message at the "Film Stocks and Processing" section

yesterday and soon after I realized it is likely in the wrong section.

Sorry for this dupe post.

We recently pulled two reels that needs to be restored. And there are more reels

that are in this similar configuration. Here are the facts.

All the reels are from the mid 80s. The two reels are the same clip, one internegative

and the other interpositive. There are no splices on IP nor IN. The IN is the final edit/release,

and the IP is not the final edit. There are no other IP reel found that is final edit like the IN.

The IP is also a little longer than the IN.

The only problem is that the IP seems to be raw/source reel or perhaps a one light

print or workprint or perhaps a direct copy from the original camera negative? In other words,

the clip on the IP have each scene that are separated by 2 or 4 blank cells. Throughout the reel,

it contains the RGBs info on the cells. Each scene are usually about 10 seconds or less. The IN,

however, is complete and final with all editing done. And the IN comes with paper works from the

processing lab for all the RGB configurations for the IN reel.

Here are the questions:

I just find it strange that there are no IP that is the same final edit as the final IN.

So, the question is if it is worth the cost and time to scan the IP for the 4K restoration, since it is

at least one generation early? or just scan and 4K restore the IN instead?

Any ideas/guess out there for the reel that is marked IP, if it is a raw/source reel or perhaps a one light

print or workprint or perhaps a direct copy from the original camera negative?

Frank

[David Mullen ASC]

Can you get them to scan a frame of each type so you can compare them?

[Frank Chang]

Hi David. Do you mean compare them for image quality?

 

By the way, I did notice this morning that one of the fade in/out is not in the IP but it is in the IN.

[David Mullen ASC]

Yes.

[Frank Chang]

Ok, I will have the other dept do a test scan of each type. Thanks.

[Perry Paolantonio]

We almost always recommend scanning from elements as close to the original as possible. You should get a sharper image that way because you're skipping generational loss.

 

That being said, it's often more work to do a scan from the original neg, because things like optical effects might need to be recreated, color correction will almost certainly be shot by shot, and you're more likely to pick up annoying white dust specs when working off of negative.

 

The original lab notes and timing info may be useful for determining general color correction settings, or for figuring out where there should be fades, for example. but the actual numbers for the printer lights are probably useless as they would have been specific to the lab (and possibly even the printer) that they were run through.

 

The cost/benefit analysis is something that only you can determine. I would do as David suggests and get a comparison of the same shot from both elements and see what looks best. It's going to be more work to recreate things, but it's not that hard, and you may find that you get a much better end result because you're working from something closer to the camera original.

 

-perry

[Frank Chang]

Thanks, Perry.

[Frank Chang]

Hello David. Just a follow up. We did a test scan for one scene (about 10 seconds) and it does appear that both have the same quality

and can not tell a difference between the two.

 

The original vault person retired years ago and the record keeping back in the days are almost next to none.

The things that are worth noting are that these IPs are not the standard IPs I am used to. For one things, past IPs we

have worked on are almost all final (i.e. same as IN). These IPs, however, look like it is direct copy from the OCN.

So not only it is not color-timed (it does at least contains clipboard images for each scene about the color configuration),

it is missing all the optical effects. And it contains 2 to 4 blank cells that separate for each scene. No splice.

Generally I would have called this as a workprint reel. Just pondering why it is called IP?

Edited November 10, 2016 by Frank Chang

[David Mullen ASC]

If the IN's are color-timed but the IP's are not, then the IN's weren't struck from those IP's...

 

People make un-timed IP's from negative for various reasons -- perhaps they were protection copies or someone else needed a dupe copy in a hurry, or perhaps someone wanted to use them in an optical printer (without timing them first), or they were timed to match the negative but not the overall scene if that hadn't been cut together yet. You could be seeing something prepped for optical printing onto a dupe negative.

 

It's called an IP because it's a positive copy of the negative made onto duplication stock.

[Frank Chang]

I see. Thanks David.