Is it common practice to color grade a scanned, color-timed interpositive for digital viewing?

[Moon Adnan]

Hi all,

When a film goes through a traditional photochemical finish, my understanding is that the answer print process establishes the final approved color timing, which is then carried into the interpositive, internegative, and ultimately release prints.

If the color-timed interpositive is then scanned to create a digital master for non-film viewing (streaming, online video, DVD/Blu-ray, etc.), is an additional digital color grade normally necessary?

How closely does the approved photochemical timing represented by the interpositive translate when that IP is scanned and viewed on TVs, computer monitors, phones/tablets, etc.? Is the subsequent digital grade generally just a matter of matching the scan to the approved film appearance, or can substantial additional grading be required?

For practical purposes, I’m wondering whether I would need to digitally grade a short film after already going through answer prints and approving the photochemical timing, if I wanted both film prints for projection and a digital version for online viewing and festival exhibition.

I recall reading on David Mullen’s The Love Witch (2016) thread where he mentioned that they still graded the scanned interpositive, but that there was considerably less work to do because much of the intended look had already been established photochemically.

Could one reasonably scan a properly timed interpositive and use that scan with minimal correction, or is a separate digital grading/mastering pass essentially always necessary to reproduce the intended appearance on digital displays?

I appreciate this community, I have learned a lot from you all! Thank you 🙂 

[Tyler Purcell]

So yea, there is a big difference between how you'd do it today and in the past. 

You'd make a contact print of every roll of original negative. Those contact prints would be used to understand where the negative needs to be adjusted. Sometimes you do this on a light table, but you can also use a color master, which is machine that takes the negative and captures it with a TV camera, so you can kinda get an approximation of what the color will look like. It then spits out the tape for the optical printer. Then you can go and make a new contact print from that and get it in the wheelhouse. This would all be done as the film is being shot for a big movie. 

Once the camera negative is cut, you can translate those color numbers to the cut. Then roll by roll you can do contact prints of the cut negative and then those contact prints are used to fine tune the look using the same methods above. Usually at that point, it's down to balancing between scenes and shots, which is about all you can do anyway. 

Once everything has been approved, the edit is good and such, then they will make a colored IP, which is at this point a protection element more than anything else. Some movies will strike multiple IP's, but suffice to say with limited print runs today, it's only for protection anyway. Any theatrical print would be struck from the original camera negative anyway, especially if you're doing just a few. 

From there you have two choices; DI from the negative or DI from the IP. 

If you're doing the DI from the negative, you have to start all over again from scratch. However, it would be sharper and offer far more dynamic range (with HDR scan) which allows much better manipulation in post. 

If you're doing the DI from the IP, you're stuck to the IP's dynamic range and unless you're shooting large format, there is a degrade close to half the lines of resolution when you go OCN>IP. So even with S35mm, you're kinda losing quite a bit of resolution just making an IP.  The other thing to keep in mind is that if you're shooting spherical S35mm, you will need to do an optical reduction to make prints anyway, unless you're doing an external soundtrack, you wouldn't want to scan the reduced IP anyway. If this is for 16mm, I would not bother. Photochemical process only really works well for 35mm and larger formats. 

In terms of finishing from an IP, I get why a cinematographer would want that as their finished look; because a colorist can't make changes. What's baked into the IP, is forever the look of the movie. However, I do feel that scanning the original camera negative and retaining all that high res and color science goodness, is something that just needs to be done. Scanning the IP as reference, ok fine that's not a big deal. But you will never grade your film perfectly photochemically, so it's better to have a beautiful OCN scan to work with for the digital version and mimic the print look, than it is to start with an IP.  Thats just my opinion, but I've tried it, just never been happy with the results from IP. 

[David Mullen ASC]

It’s very common to transfer from a color-timed IP for home video if the negative was cut. You have to be very careful scanning spliced negative and if it was A-B rolled then you’ll have to do some digital editing and adding transitions. Most of Nolan’s movies on home video are transfers from an IP.

I just did a D.I. of a feature from original 35mm negative this spring and it was an intense 10-day job. If we were using a color-timed IP it could have been done in 5 days if not less. But of course to create a color-timed IP might take three answer print passes so that’s also a time issue.

[David Mullen ASC]

Keep in mind that if the negative was cut with the intent of creating projection prints, one reason the transfer of the color-timed IP goes faster is that the colorist is just trying to create an accurate digital representation of the look of the print, they aren’t adding creative Power Windows or a lot of dynamic adjustments even if they could since that’s not in the print version.

[David Mullen ASC]

There’s always the need for a digital grade even if the IP is color-timed. And there are “trim passes” made for different digital release formats — Rec709 home video, P3 DCP, HDR versions, etc. They all need the eyes of a colorist to make sure they look correct. 

[Brandon Paterno]

On 8/17/2026 at 7:34 PM, Tyler Purcell said:

If this is for 16mm, I would not bother. Photochemical process only really works well for 35mm and larger formats. 

news to me!

[Tyler Purcell]

9 hours ago, Brandon Paterno said:

news to me!

When you drop resolution in half between OCN and IP, it really shows up with 16mm. It's even worse because 16mm projection doesn't use the full width. Having been striking 16mm prints from the original negative for years, it's always been a "meh" result compared to 35mm and larger formats. The benefits are generally outweighed by the trade off's, especially due to A/B roll blind splicing required for 16mm assembly/printing. You don't need to do blind splicing with 35mm and larger. 

[Moon Adnan]

Thanks for your detailed perspectives and responses! @Tyler Purcell @David Mullen ASC

David, do you/people worry about the resolution drop Tyler refers to when scanning the IP instead of OCN for digital release?

[David Mullen ASC]

If you scan a 35mm IP of a 35mm movie at 4K for a 4K release in cinemas and UHD, it should be a good representation of the movie's resolution plus a good representation of a Vision release print. Sure, you'd get a bit more resolution, dynamic range and color information by using the original negative.

You do what makes the most sense for your budget - if your negative is A-B roll cut, especially a checkerboarded 16mm negative, it might make more sense to work from an IP (unless the A-B roll is just for a few lab dissolves, then scan the negative and do an online to recreate the dissolves). If you are doing a D.I. from camera negative rolls with no plan to cut the negative later, then you couldn't make a contact-printed IP anyway of the final cut.

[Tyler Purcell]

53 minutes ago, David Mullen ASC said:

You do what makes the most sense for your budget - if your negative is A-B roll cut, especially a checkerboarded 16mm negative, it might make more sense to work from an IP (unless the A-B roll is just for a few lab dissolves, then scan the negative and do an online to recreate the dissolves). If you are doing a D.I. from camera negative rolls with no plan to cut the negative later, then you couldn't make a contact-printed IP anyway of the final cut.

With Resolve, you can instantly cut and remove the blank spots between the shots in the A/B rolls and get an assembly of the scanned A/B negative in a few hours. I just did a scan from an original camera negative A/B roll and it took me around 3 hours to conform the entire 6 reel A/B roll feature without any issues. The bigger issue we've found is related to opticals like transitions and titles. Some of the films we've restored, have 35mm opticals, so that helps greatly. Others simply you have to steal from the blow up 35mm IP, which generally look very different then the 16mm source. I've been lucky and for some restorations, we've found the actual elements used to make the dissolves, so we've just done them digitally. Tho, I prefer using the original credit elements for sure.