Potentially Awesome Double8 Pin-Reg 2K Workflow

[Paul Korver]

So we all know that some of the charm of Super 8mm is the lack of registration. This problem is not only due to the camera systems (lack of pressure plates and plastic cartridges)... but it's also due to the fact that all current Super 8mm film transfer systems are non-pin registered.

 

Lately I've been thinking it could be cool to be able to create a very clean small-format image with not only superior D8 camera systems but also a super high-rez scan. Our new 4K Scanity film scanner at Cinelicious opens up some new and interesting possibilities for this with double 8 film. Basically when shot with a Double 8 camera the image is already much more stable due to the pressure plates (check out Bolex H8 Rex cameras). And on the Scanning side our Scanity matches an ARRI Scanner in steadiness but with much better dynamic range and it's a lot faster.

 

The work flow would be:

 

- Shoot Double 8

- Process as 16mm (ie. DON'T SPLIT THE FILM!!!)

- We scan the unsplit D8 as 16mm at 4K resolution (15fps!!) pin registered on the Scanity.

- We split the frames digitally during color correction and render out 2K per 8mm frame, pin-registered images.

 

The only problem is that currently the optical perf camera is only calibrated to standard 16mm and 35mm perf pitches. D8 film stock has a half-perf pitch of 16mm (meaning twice the perfs) I'm going to see if we can address this with the manufacturer to see if we can program the Scanity to recognize the smaller D8 perf pitch... but this would be a non-issue if D8 cameras could run normal 16mm film stocks. Does anyone experienced with D8 camera systems anyone know if you can run modern standard 100' 16mm spools in D8 camera systems such as the Bolex H8 Rex Double 8?

 

I have a feeling that Guy at CamerasPro would probably be the person to ask but I thought I'd put it to the Forum.

 

Best,

 

Paul

[Brian Drysdale]

The problem would be that the camera claw pulls down one frame at a time and if you've only got 1/2 the number of perfs (two frames per perf) the claw will try and find a perf or attempt to make one.

[Rafael Rivera]

Inspiring to see how much thought you've given this, and how seriously you take film, even a small gauge one!

[Dom Jaeger]

Hi Paul,

 

it certainly would be very interesting to see a high quality scan of double 8 footage.

 

As Brian pointed out, 16mm film won't work in a d8 camera, but unslit d8 will run through a 16mm camera or projector. The projected 16mm image will show 4 x 8mm frames, with the 2 on the right upside down and running in reverse. So I would imagine it's possible to scan d8 as 16mm, with each 4k scan capturing 4 frames. If you can digitally split that in half, presumably you can split it into quarters?

 

The only issue I can think of is if the camera alignment of claw pulldown to gate aperture is off, and the exposed 8mm frame is not centred between perfs. Then when viewed as a 16mm image the 2 left 8mm frames will be higher or lower than the 2 right frames, and possibly exceed the scanned area. But that can be adjusted fairly simply in most d8 cameras.

[Andries Molenaar]

Don't want to spoil a good idea. But of course the non-split footage can only be newly exposed. How much Double-8 gets exposed these days and how much could possibly end up at your scanner? It is certainly a niche market. It could grow a bit if newly exposed Super-8 would have to cease one day.Double-8 can more easily be manufactured from 16mm film then Super-8.

Edited May 2, 2011 by Andries Molenaar

[Will Montgomery]

I've actually already done this on a Spirit here in Dallas; not sure how the Scanity would handle it but it was just twice as many perfs so the Spirit only used every other one.

 

Just shoot regular 8mm, process at Dwayne's (or wherever) and tell them not to split the film in the notes section. Then transfer it on your Scanity or Spirit with the 16mm gate. You will have a big advantage since you're used to dealing with small format and the grain. My test had very large grain that I never get when Cinelicous transfers my Super 8mm and you mentioned that it took some work to get that process down. Plus what we did was only 2k as their 4k Millennium was down.

 

I'll post some samples later in the week.

 

Yes you can still get regular 8mm film from John Schwind at:

 

http://www.zerelda.com/internationalfilm/internationalfilm.html

 

You're also right in thinking that you'll need a really good reflexed regular 8 camera because focus actually maters :rolleyes: especially at this small of a gauge.

 

If I can dig up the film I can send it to you for a test if you're interested... it's prepped and ready for 16mm transfer.

[Paul Korver]

Thanks for the feedback all.

I guess I'll have to try to pursue getting the Scanity to recognize the 1/2 perf pitch of D8 on 16mm film. Hopefully this won't be too difficult as it should just be inputting new parameters into how the perf camera is interpreting data.

 

In terms of the resulting resolution and how I'm coming up with the term "2K" my calculations are as follows:

 

-In 4K mode the Scanity scans 3420 pixel resolution across the 10.26mm R16mm horizontal negative area.

-The resulting pixel resolution 4.5mm horizontal negative area of the R8 images on D8 film would then be calculated by the simple (3420*4.5)/10.26 = 1500

-What would be amazing about having a pin-registered HDR scan of R8mm at 1500x1100 is that technically speaking, at the native 1.36:1 regular 8 aspect ratio, this scan resolution would fit perfectly in a 2048x1080 2K DCP deliverable.

 

Inspiring to see how much thought you've given this, and how seriously you take film, even a small gauge one!

Thanks Rafael... we love all film formats big and small. Unfortunately most of the R&D has been done for the larger formats since a lot more volume being shot.

 

 

Don't want to spoil a good idea. But of course the non-split footage can only be newly exposed. How much Double-8 gets exposed these days and how much could possibly end up at your scanner? It is certainly a niche market. It could grow a bit if newly exposed Super-8 would have to cease one day.Double-8 can more easily be manufactured from 16mm film then Super-8.

You are correct Andries... we're not looking at this as a "cash cow" or even a business model. It's just a passion project enabled by the fact that we have already invested in the technology to do so and if all it takes is a bit of firmware tweaking to be able to make this huge leap in image quality then why not?

 

I've actually already done this on a Spirit here in Dallas; not sure how the Scanity would handle it but it was just twice as many perfs so the Spirit only used every other one.

 

Just shoot regular 8mm, process at Dwayne's (or wherever) and tell them not to split the film in the notes section. Then transfer it on your Scanity or Spirit with the 16mm gate. You will have a big advantage since you're used to dealing with small format and the grain. My test had very large grain that I never get when Cinelicous transfers my Super 8mm and you mentioned that it took some work to get that process down. Plus what we did was only 2k as their 4k Millennium was down.

 

That's interesting Will... but at HD or 2K across D8/16mm wouldn't result in a picture area useable for anything other than SD. Plus not pin-reg, HDR, yada-yada. In addition to a huge leap in resolution and picture quality doing this on the Scanity is pretty cool due to the speed of ingest. Since we can do 16mm 4K at 15fps... and that would be ingesting 4x frames per 16mm frame which would net us 60fps 2K 8mm scanning speed!!! (of course there would be added time/cost in splitting up the 4K quadrants w/pan&scan, rotation etc but if a client was tech / post savvy enough to do this on their own then it would be added savings for them).

 

In terms of granularity anyone that doesn't like grain should obviously not shoot R8. However I'd be very interested to see Witter Kinotechnik's lower speed reversal stocks... 100D, Velvia 50D, PlusX or FomaPan. And in terms of scanning reversal which is high dynamic range in terms of film-density... the Scanity has no competitor.

 

-Paul

[John Sprung]

... and that would be ingesting 4x frames per 16mm frame

 

OK, so the good news is that you wouldn't have to worry about recognizing the D8 extra perfs -- the machine would just use every other one. It thinks it's doing 16mm. The bad news is that the pin registration would apply to only one of the four images.

 

Once you have it scanned, each scanner frame consists of four images with frame lines, four camera frames. So, you can do even better than pin registration just by recognizing the camera aperture frame lines in the data. That's a computer programming job. The only limitation is that you'd be SOL on dark night exteriors.

 

 

 

-- J.S.

[Will Montgomery]

Plus not pin-reg, HDR, yada-yada

After all these years I never seem to get that one right. So a Spirit is not pin registered? I thought that it was because you can't see the sprocket holes on the 16mm gate but maybe that's just where the gate crops. I guess that's why the different perf pitch didn't effect the Spirit.

 

I seem to remember perfs mattering to the Spirit for something... Is the Millennium pin registered?

[John Sprung]

The Spirit isn't pin registered because it's a continuous motion line scan device. It's sprocket driven. The sensors are one photosite high by 2048 across, one each for RGB. How many lines you clock out per turn of the sprocket is easily changed in the electronics, so it's easy to set up for 4, 3, or 2 perf.

 

 

 

-- J.S.

[Patrick Cooper]

A lot of potential there. Sounds like you can give regular 8mm transfers a new lease of life. Any possibility of trying something similar with double super 8? Obviously, more modification would be required regarding the perfs.

 

In terms of granularity anyone that doesn't like grain should obviously not shoot R8. -Paul

 

I have seen some transfers of Fuji 64D and Kodachrome (exposed on regular 8mm) and in both cases, the footage looked fairly clean and fine grained. The transfers were done with frame by frame Workprinter and Workprinter/ Sniper clone equipment (basically pointing the video csmera at the projector gate and capturing with a computer.)

[Will Montgomery]

So when is this going from theory to production? Do you need a sample to work from? I can shoot this weekend...

 

My regular 8 cameras are a little weak however, I use Canonette Ciné's which I love for portability but are not exactly the sharpest cameras available.

 

Maybe someone with a Bolex could shoot a roll or two for you?

[Paul Korver]

So when is this going from theory to production? Do you need a sample to work from? I can shoot this weekend...

 

My regular 8 cameras are a little weak however, I use Canonette Ciné's which I love for portability but are not exactly the sharpest cameras available.

 

Maybe someone with a Bolex could shoot a roll or two for you?

 

Hey Will,

It may take longer than expected. Reading the half-sized D8mm perf spacing is on the manufacturers to do list but pretty low on the list due to low demand for that format.

 

I'll post again if/when we are able to do it.

 

-Paul

[Will Montgomery]

Results of Regular 8mm Spirit transfer via a 16mm gate...

 

Here's that transfer I was talking about earlier. No grain reduction which is probably needed for this film. Of course you can zoom in and rotate as needed from within the Spirit but I thought it would be fun to post what the 16mm gate sees...

 

[Jean-Louis Seguin]

Results of Regular 8mm Spirit transfer via a 16mm gate...

 

Here's that transfer I was talking about earlier. No grain reduction which is probably needed for this film. Of course you can zoom in and rotate as needed from within the Spirit but I thought it would be fun to post what the 16mm gate sees...

 

 

 

 

Why are sprocket holes moving around so much? Is this the registration one can normally expect from a Spirit?

 

Jean-Louis

[Will Montgomery]

Why are sprocket holes moving around so much? Is this the registration one can normally expect from a Spirit?

Good question. Paul? what do you think?

 

It was a cheap 8mm camera so the framing registration could probably jump around due to the camera but the sprocket holes should be steady I would think.

 

Since I learned on this thread that Spirits are not pin registered maybe in 16mm it does move a little. Or maybe the extra sprocket holes make it less stable?

[John Sprung]

Why are sprocket holes moving around so much? Is this the registration one can normally expect from a Spirit?

 

The holes move around, but the middle image on the left looks quite steady. Perhaps they registered it digitally off of the frame lines?

 

 

 

-- J.S.

[Jean-Louis Seguin]

Yes, I guess the left side looks more steady than the right side, so I assume the left sprockets on the left side are the ones used to register the film.

Perhaps the film was not perfectly perforated.

 

Cheers,

Jean-Louis

[Paul Korver]

Spirits aren't pin registered. That's what's causing the perf drift.