S16mm - 35mm Blow-up

[Sam Care]

I was wondering how doing a S16mm to 35mm blow up optically versus a Digital Intermediate affects the editing process, if you want to use Final Cut Pro? Is one method generally significantly more expensive than the other? Is one method better if you know your going to have to create 35mm theatrical release prints and later release on DVD as opposed to just making one print for a short or feature to screen at festivals?

 

Thanks. Sam

[Stephen Williams]

I was wondering how doing a S16mm to 35mm blow up optically versus a Digital Intermediate affects the editing process, if you want to use Final Cut Pro? Is one method generally significantly more expensive than the other? Is one method better if you know your going to have to create 35mm theatrical release prints and later release on DVD as opposed to just making one print for a short or feature to screen at festivals?

 

Thanks. Sam

 

Hi,

 

A good DI will cost more, and could be be better. Optical Blow ups and DI's vary in quality, you will need to see some tests.

You can use Final Cut in both cases, for the blow up the neg will need to be cut. When you telecine your rushes I would have the keykode in picture if going the optical route.

 

Stephen

[Dominic Case]

There are so many factors that affect the final quality, it is impossible to say categorically that a DI blowup will be better (or worse) than an optical blowup. In theory though, an ideal DI should be a little better than an ideal optical blowup.

 

If you go via DI, the choice of format (2K vs HD) will make a difference, as will the choice of where it's done: some houses are better at DIs than others, believe it or not!).

 

If you go via optical blowup, there are different routes too: you can go direct to print (often not recommended, as the blow-up procedure does put the negative at a higher risk, with no fall-back if it suffers from wear and tear after a couple of prints); or via a 16mm or a 35mm IP. But different houses will produce different results - the lens used for the blow-up stage, for example, can affect the outcome in terms of sharpness and flare.

 

You will need frame-exact neg matching to A&B rolls if you choose to go for a direct blow-up, or for a 16mm IP. For a DI or a 35mm IP it's better to cut the neg with handles to avoid splices on good frames. There are many more vartiable that your lab should advise you on.

 

Finally, a DI will allow you to do a lot of things in the digital mode (fix-ups; "optical" effects such as speed changes, freezes, etc; and digital colour correction) much more readily than you can in the all-film process: but it WILL cost you more.

[Dan Goulder]

Using a Spirit to do the D.I. transfer, instead of a film scanner, has really closed the gap in terms of cost. You can send the output to HD or data files. "Walk the Line" is a recent example of a movie done that way(Spirit to data). Some will argue that this isn't a true D.I. if it doesn't involve a film recorder. On the other hand, you may find the output from the Spirit even more to your liking, as a subjective matter of taste.

[Chris Burke]

A very good, seasoned and working DP just told me that you could expect to pay about $1,000.00 per minute for a descent Super 16 - 2k - 35mm DI. I heard it said right here on these boards about a year or two ago that a feature should expect to pay 200K to 150K. That was two years ago. So 90K for today's prices sounds about right. DIs are becoming quite common place, even for the indie short. The price difference is getting smaller all the time. Shop around, you may be able to find a good price for a HD-D5 DI. Even a HD-SR DI. It won't add any grain like an optical can.

[David Mullen ASC]

Using a Spirit to do the D.I. transfer, instead of a film scanner, has really closed the gap in terms of cost. You can send the output to HD or data files. "Walk the Line" is a recent example of a movie done that way(Spirit to data). Some will argue that this isn't a true D.I. if it doesn't involve a film recorder. On the other hand, you may find the output from the Spirit even more to your liking, as a subjective matter of taste.

 

They used the 4K Spirit at Modern Film & Video, then downrezzed to 2K.

 

And you don't mean "film recorder" you mean "film scanner", as in versus a datacine/telecine. The Arrilaser is a film recorder (a device for transferring digital TO film.)

[Dan Goulder]

And you don't mean "film recorder" you mean "film scanner"

Yeah, ya know, one of them film thingies.

[Stephen Williams]

A very good, seasoned and working DP just told me that you could expect to pay about $1,000.00 per minute for a descent Super 16 - 2k - 35mm DI. I heard it said right here on these boards about a year or two ago that a feature should expect to pay 200K to 150K. That was two years ago. So 90K for today's prices sounds about right. DIs are becoming quite common place, even for the indie short. The price difference is getting smaller all the time. Shop around, you may be able to find a good price for a HD-D5 DI. Even a HD-SR DI. It won't add any grain like an optical can.

 

Hi,

 

The cheaper route Via HDCAM-SR is not 2K its HD!

 

A high end DI is still very expensive, you pay for what you get in some respects.

 

Stephen

[Dirk DeJonghe]

For shorts films with no distribution a direct blow-up may be the most cost-effective solution. You get very high picture quality, lowest cost, and the option to make an IP/DN later using the same grading data. No fancy grading effects are possible.

 

For feature films, very few direct blow-ups are made because the distributor won't pay the price. The production company has to provide either a duplicate negative via IP blow up or a DI negative.

 

Both scanning and recording prices have gone down considerably in the last couple months, a 2K DI with full digital grading is now only somewhat more expensive than a IP/DN blow up if you take the video mastering into account.

[Max Jacoby]

They used the 4K Spirit at Modern Film & Video, then downrezzed to 2K.

I will add to this that there are several different Spirits out there, the 4K and 2K version being the most recent ones. But there is also the older Spirit, which is far more common and which is not 2K resolution but 1920 x 1445. On top of that it subsamples the color information by half, down to 960 pixels. Also it is not pin-registered, so there is still a quality difference between this Spirit and a dedicated film scanner.

[Keith Mottram]

Hi,

 

The cheaper route Via HDCAM-SR is not 2K its HD!

 

A high end DI is still very expensive, you pay for what you get in some respects.

 

Stephen

 

As Max said older Spirit 2K models (which as Thomson sight says were used for Amelie and O Brother Where Art Thou) scanned 1920 accross, modern ones 2048. if we take 2048 as true 2K then were talking about a resolution loss of less than 7% which will not be noticable, so to say SR is not 2K is a bit perdantic. On a 1.85:1 ratio a 2048 scan will need 1107 vert as apposed to a 1920 scan needing 1037. The colour space is more important, but dual link 10bit SR will be indistinguishable in my opinion to data- once blown up no one will notice that 7%. If you had a post produciton FCP workflow enabling the level of data flow for SR then an SR reconform could be a desent master and not prohibatably expensive. You would also have everything required for all digital deliverables- from HDCAM/ D5 for festivals to NTSC/ PAL sd versions. To get to this stage you just need to hire an SR deck for laying back your TK to tape and as many hours as your project needs in the suite. This will be nowhere near $90000. If there is just match grading and simple colour correction (as opposed to multi layered and tracked correction) then this could be a very cheap and simple process. Of course transfer via Arri laser is another cost entirely. I have no idea how much an optical would cost, but I cant see how it would be a better option.

 

Keith

[Stephen Williams]

The colour space is more important, but dual link 10bit SR will be indistinguishable in my opinion to data- once blown up no one will notice that 7%.

Keith

 

Keith,

 

Have you seen a side by side test? I ask as the HDCAM SR DI's I have seen have been very average.

 

Stephen

[Keith Mottram]

Keith,

 

Have you seen a side by side test? I ask as the HDCAM SR DI's I have seen have been very average.

 

Stephen

 

I have seen a 1920x1080 data to 35mm and an SR to 35mm and could not tell the difference, though obviously it is hard with no split screen. Having worked with Spirit 2K 2048 DPX scans and 1920 '2k' re-scans (of the same take) I can say that there is no noticable difference (and in this case the 1920 SR was 4:2:2) once the DPX was placed in the 'uncompressed' HD timeline (and therefor moved to the same codec/ colour space). I looked for major differences, but could not find them - maybe I should have tried even harder, but after a certain point any miniscule difference would be lost upon transfer to film or downconversion anyway.

 

This is obviously not an exact test, but with the information I have I think that my previous post is correct. I have found SR 'compression' to be outstanding and I feel that an average DI is to do with the rest of the post chain rather than wether you master to data or SR. I would be interested in anyone else who could explain why this should be otherwise (apart from the 'unnoticable-just below 7%' and assuming you are working in 10bit colour).

 

Keith