Film restoration in the digital domain: A chat with James White

Maria Falconetti in Carl Theodor Dreyer’s The Passion of Joan of Arc (1928). This is how the frame appears in the cur­rent Eureka!/Masters of Cinema Blu-ray disc edi­tion of the film. Below, the image before restoration. 

I first heard from James White, the British film res­tor­a­tion maes­tro, a few years back, around the time the below-mentioned ITV Blu-ray of Powell and Pressburger’s Black Narcissus came out in Britain. White and I recon­nec­ted around last year’s Christmas hol­i­days, just as two very dis­par­ate pro­jects in which he had a res­tor­a­tion hand were com­ing on the scene: the above-depicted ver­sion of the Dreyer mas­ter­piece, and Arrow Video’s iter­a­tion of Lucio Fulci’s notori­ous Zombi, known in Britain as Zombie Flesh Eaters. We figured the time was right to catch up on some of the issues con­cern­ing film res­tor­a­tion in the brave and still rel­at­ively new but increas­ingly per­vas­ive world of digit­al, and thus an e‑mail exchange began.  It took some time, but I think the res­ults are inform­at­ive, and con­tain news that is both heart­en­ing and dis­tress­ing. If you can, you should read in con­junc­tion with Nick Wrigley’s essay “Crimes Against The Grain” in the December 2012 issue of Sight & Sound. Wrigley was one the founders of the Masters of Cinema imprint and worked as a res­tor­a­tion super­visor with White on the Dreyer pro­ject. The inter­view with White follows: 

1) Film restoration
seems to have pretty defin­it­ively shif­ted from a pho­to­chem­ic­al pro­cess to a
digit­al and elec­tron­ic one. When was the tip­ping point for this, and what does
it mean in terms of the eco­nom­ics of film res­tor­a­tion? Not to men­tion the distribution
of restored motion pictures?

It’s cer­tainly true that
over the past dec­ade or so, film res­tor­a­tion has moved from a largely
pho­to­chem­ic­al pro­cess to one rely­ing almost com­pletely on digit­al technology.
When I began work­ing in this field in New York back in the mid-1990s, digital
res­tor­a­tion was still in its infancy and most of the major pro­jects such as
those being over­seen by Photoplay (Napoleon) or Harris & Katz (Spartacus, Vertigo) were being pro­duced almost solely through
tra­di­tion­al pho­to­chem­ic­al pro­cesses. These days, how­ever, if someone embarks on
a new res­tor­a­tion, they will almost always begin by scan­ning the best existing
film mater­i­als and work­ing in digit­al through­out the entire process. 

My time at the BFI
reflects this shift in approach. When I star­ted work­ing there in 2002, film
res­tor­a­tion was still very much the pre­serve of the archive lab, with digital
tech­no­logy hav­ing little to no involve­ment. Restorations from that era
rep­res­en­ted the best that could have been achieved at the time, but in recent
years the BFI has seen fit to revis­it many of their key titles using digital
tools, hav­ing seen the advant­ages these bring. A title like Blackmail for instance, restored last year by the BFI
as part of their silent Hitchcock pro­ject, now looks amaz­ingly improved in its
digit­al incarn­a­tion in com­par­is­on to the pri­or res­tor­a­tion com­pleted in the
1990s. 

I don’t think there’s
been a single tip­ping point that made this change hap­pen, oth­er than
res­tor­a­tion simply mir­ror­ing the film industry at large, and its shift to using
digit­al tools for pretty much everything under the sun. The simplest reason
though, is that the tools just got bet­ter. Most import­antly, the software
became sym­path­et­ic to the needs of archive films and offered solu­tions to
prob­lems that tra­di­tion­al res­tor­a­tion hadn’t ever been able to deal with
suf­fi­ciently. Issues that had always been dif­fi­cult to impossible to fix -
things like image sta­bil­ity, dens­ity issues, deep emul­sion scratches, warping,
regis­tra­tion issues, deteri­or­a­tion brought on by heat or mois­ture, and damaged
or even miss­ing frames – these could now be improved upon in a way that just
could­n’t be done by tra­di­tion­al means. 

Having access to these
new tools has­n’t made the work cheap­er, though. If any­thing, restor­ing a film
has become more expens­ive simply because audi­ence expect­a­tions are so much
high­er now. Fortunately there are so many more out­lets to dis­trib­ute these
titles, so the eco­nom­ics are there to sup­port the work. Mind you, there’s
def­in­itely a lim­it to what most dis­trib­ut­ors can afford. A top-tier restoration
like The Red Shoes or Lawrence
of Arabia can demonstrate
beau­ti­fully what digit­al res­tor­a­tion is cap­able of – these films now truly take
your breath away – but the vast major­ity of film titles don’t enjoy anything
near to the res­tor­a­tion budgets these pro­jects had. So while restor­ing the
major­ity of film titles to a 2K/4K level might be bey­ond the means of some
dis­trib­ut­ors, restor­ing a film to a very high stand­ard in HD is often within
their means. I should men­tion that going this route does­n’t sup­port a film’s
archiv­al basis – restrict­ing your end res­ult to HD does­n’t cre­ate what most
arch­iv­ists would agree is the basis for film pre­ser­va­tion – but it does provide
dis­trib­ut­ors with a format suit­able for the major­ity of cur­rent release
plat­forms (DCP, Blu-ray/DVD, etc). 

In any case, what’s
encour­aging is the high amount of res­tor­a­tion work cur­rently going on
inter­na­tion­ally. Rights-holders of film lib­rar­ies have star­ted to see the value
in restor­ing their titles to make them access­ible for HD broad­cast and online
plat­forms. DVD/Blu-ray sales have helped cre­ate new audi­ences for archive
titles, and the improved present­a­tion qual­ity has made view­ers more aware
and appre­ci­at­ive of the prac­tice of film res­tor­a­tion on the whole. Back when I
star­ted, Criterion was the only com­pany around that devoted time, atten­tion and
a decent budget to giv­ing a film the best present­a­tion pos­sible on video, but
now in addi­tion to well-established labels like Eureka/Masters of Cinema, the
BFI, Milestone and Kino/Lorber, you’ve got all these new labels like Olive in
the US, Second Run in the UK and Edition Filmmuseum in Germany resurrecting
lost cinema or rarely-seen clas­sics.  It’s the work of com­pan­ies like these
that is really help­ing to bring new interest in film res­tor­a­tion. I mean, it’s
great when Casablanca or Singin’
In The Rain gets treated to
anoth­er state-of-the-art res­tor­a­tion, but there now seems to be an actively
grow­ing interest in dis­cov­er­ing new films and film col­lec­tions from all corners
of the globe. A pro­ject like Milestone using Kickstarter to fund a new
res­tor­a­tion of Shirley Clarke’s Portrait of Jason is a great example of the pub­lic tak­ing an active
role in film res­tor­a­tion. Likewise for Distribpix’s recent res­tor­a­tion of The
Opening of Misty Beethoven – that
film now looks bet­ter than any­one prob­ably ever expec­ted it to! 

2) The tools for film
res­tor­a­tion have evolved to the point that the lay view­er is under the
impres­sion that it can per­form mir­acles, which to some extent seems objectively
true. So let’s look at it from the oth­er end of the tele­scope, as it were, and
talk about the things that the tools CAN’T do.

Well, as I said there’s
a great amount of things that digit­al tools can do to repair the issues that
have always blighted older film titles that we couldn’t do just a few years
ago. That said, it’s import­ant to bear in mind that in the most fundamental
sense, no amount of res­tor­a­tion, digit­al or oth­er­wise, can significantly
improve the qual­ity of the image of the film ele­ment that it’s sourced from.
The issues I’ve men­tioned – dirt, scratches, sta­bil­ity, flick­er, miss­ing or
dam­aged frames, etc – these can all be improved sig­ni­fic­antly through the
care­ful applic­a­tion of digit­al tools. But the basic details with­in that
ori­gin­al image, mean­ing the film grain, the level of detail, etc – these can
only be improved with­in the para­met­ers of what the ori­gin­al ele­ment would
allow.

It’s like up-rezing an
image from SD to HD. An increase in detail doesn’t hap­pen simply because you’ve
added more pixels. Of course one can give the illu­sion of increased detail or
sharp­ness through digit­al enhance­ment, but then it’s no longer the film you’re
work­ing with, it’s some­thing else. I work to cre­ate the best rep­res­ent­a­tion of
an old film pos­sible, but that means keep­ing the res­ults with­in the historic
bounds of what that film would be able to achieve at the time – be it its film
grain, the sat­ur­a­tion of its col­ors, the level of its con­trast, etc. A film from
1930 shouldn’t look like a film from 1950, and it most def­in­itely shouldn’t
resemble a film from 2013. Whatever tools you’re using should be done in
ser­vice of what the film looked like at the time of release, not some
ill-advised notion of what it could look like now. That, for me, is probably
the most import­ant thing.

Unrealistic expect­a­tions
are a com­mon source of frus­tra­tion for film restor­ers, as often the best
exist­ing film ele­ments simply won’t allow for a “mira­cu­lous” result.
 The pro­ject is only as good as the mater­i­al it’s sourced from, and not
all films have been sub­jec­ted to the same treat­ment over the years. Often with
older titles the ori­gin­al neg­at­ive has been lost, or is far too dam­aged to work
from. In this case you might be rely­ing on the next best thing, like a
second-generation inter­pos­it­ive or fine grain ele­ment. But there may be
prob­lems with these mater­i­als as well. Often res­tor­a­tions have to make use of
more than one ele­ment, each sub­jec­ted to vastly dif­fer­ent his­tor­ies, and then it
can prove a real chal­lenge to make the over­all res­ults con­sist­ent. To put it
simply, every film is dif­fer­ent, and every res­tor­a­tion has its own set of
challenges.

One inter­est­ing
devel­op­ment of late has been the dis­cus­sion of the dif­fer­ences between what
we’re now able to see digit­ally in com­par­is­on to what was vis­ible on a film
print. The Wizard of Oz scenario
is well-known – in Warner Brothers’ new Technicolor res­tor­a­tion you could
sud­denly see wires hold­ing up the Scarecrow, where­as in all pre­vi­ous releases
of the film, they were invis­ible. Clearly back in 1938 Victor Fleming and his
Director of Photography under­stood enough about the pho­to­chem­ic­al pro­cess that
they could expect the wires to be invis­ible by the time the­at­ric­al prints were
cre­ated in the print­ing chain, but in work­ing digit­ally from the first
gen­er­a­tion ele­ments, the wires are now there, clear as day. Likewise, I’ve
heard it recently remarked that with Vistavision films such as Vertigo we’re now see­ing a level of detail on Blu-ray that was
nev­er inten­ded for the cinema, as the­at­ric­al prints would have been made from
sep­ar­ate matrices reduced for print­ing. So how does one approach situations
like these? It’s an inter­est­ing dilemma.

3) What are your
favorite/favored tools, or the ones you find your­self apply­ing more often? What
have you had a hand in devel­op­ing? 

I can’t claim to have
had a per­son­al hand in the devel­op­ment of any spe­cif­ic soft­ware tool, but I’m
for­tu­nate that I work with such an amaz­ing team of tech­ni­cians at Deluxe Labs
in London, who are always work­ing to find new ways of meet­ing the numerous
chal­lenges this kind of work presents. We’ve worked togeth­er on so many
pro­jects over the years for the BFI, Eureka and most recently Arrow (Zombie
Flesh Eaters) we’ve developed a
great rap­port togeth­er, which is cru­cial when you’re work­ing on archive
titles. 

With pic­ture restoration
there’s a vari­ety of tools at hand that bene­fit some films more than others.
I’m always cau­tious about any so-called “auto­mated” pro­cesses, as
it’s so easy for these tools to have an unwanted effect on film grain and
detail even if applied care­fully. So while it takes far longer to do, nothing
works bet­ter for me than old-fashioned frame-by-frame cleanup, performed
manu­ally. It can be incred­ibly com­plic­ated, time-consuming work, but there’s
just no sub­sti­tute for treat­ing each indi­vidu­al frame with the utmost care.
Full dis­clos­ure, though – I don’t do any of the hand-on stuff myself anymore.
As a res­tor­a­tion super­visor, I over­see all the steps of a pro­ject from the
ini­tial research and selec­tion of film mater­i­als to the review­ing of scans to
the full grad­ing of the film to all stages of pic­ture and sound restoration.
Whenever pos­sible we’ll work with the dir­ect­or or dir­ect­or of pho­to­graphy (if
alive and avail­able) and we always work to deliv­er the highest quality
rep­res­ent­a­tion of a film’s ori­gin­al release with­in the para­met­ers the budget
and time frame allow for.   

4) The two projects
that were the spring­boards for this cor­res­pond­ence were pic­tures that seem to
have almost as wide an aes­thet­ic gulf as a tem­por­al one: Carl Theodor Dreyer’s
1928 mas­ter­work The Passion Of Joan Of Arc, which you super­vised the Eureka!/Masters Of
Cinema Blu-ray ver­sion of, and the afore­men­tioned Zombie Flesh Eaters, the notori­ous 1979 Lucio Fulci hor­ror cult
fave. In terms of mater­i­als, each one must have presen­ted unique challenges.
But there’s also a ques­tion of prop­erly serving a giv­en film’s aes­thet­ic. I
under­stand I’m con­struct­ing a bit of a big tent ques­tion here but I’m
inter­ested in your thoughts on all of these con­sid­er­a­tions. 

One could hardly choose
two more dif­fer­ent films than those two, could they? Though I still think
they’d make a great double bill! Hey, they both fea­ture heavy doses of pain and
suf­fer­ing, both base the reas­ons for their viol­ence in reli­gion of one kind or
anoth­er, and both really know the power of a well-executed close-up. Although
it’s true that only one of the films fea­tures a punch-up between a zom­bie and a
shark…

Anyway, the fundamental
cri­ter­ia for me is always to present the film as close to how it originally
appeared in the cinema. That means restor­ing the film to the highest quality
pos­sible but not employ­ing any means to “update” or
“improve” the image or the sound in some mis­guided effort to
refash­ion it to fit in bet­ter with modern-day expect­a­tions. I want the results
of some­thing I restore to appear as a film arti­fact, not a digit­al one, which
is why I’ve been using the term “pre­serving a film’s photochemical
integ­rity” lately. 

The Passion of Joan
of Arc pro­ject grew out of
Eureka’s interest in releas­ing the film on Blu-ray and the
Danish Film Institute’s desire to see Carl Dreyer’s ori­gin­al Danish version
made avail­able to the pub­lic. As far as I’m aware, Gaumont are still planning
to restore it them­selves, but the ver­sion they even­tu­ally pro­duce will almost
surely be the French ver­sion we’ve all been famil­i­ar with for years. So as this
stood as the only likely chance to have Dreyer’s debut ver­sion seen to
prop­erly, we wanted to make sure we presen­ted Joan in the best and most accur­ate representation
possible. 

We had the good fortune
to be work­ing from the Danish Film Institute’s pre­ser­va­tion mater­i­als, struck
dir­ectly from a first gen­er­a­tion 35mm print dis­covered in Oslo. This is as
lit­er­ally good as it gets with this film, a film we’re amaz­ingly lucky still
exists in any form giv­en the prob­lems it faced from the very begin­ning. For
those unfa­mil­i­ar with the saga this film was put through, the story bears
repeat­ing, so I’ll crib a bit from a piece writ­ten about the res­tor­a­tion for
Moviemail last year:

“The trouble
began just six months fol­low­ing the film’s Copenhagen première in 1928 when the
ori­gin­al neg­at­ive was des­troyed in a fire and the two ori­gin­al prints created
from this ele­ment were sub­sequently lost. A new second neg­at­ive, incorporating
extra foot­age not fea­tured in the ori­gin­al ver­sion, was sub­sequently cut
togeth­er by Dreyer but this mater­i­al swiftly fell vic­tim to mis­for­tune, and was
pre­sumed to have per­ished in a lab fire as well. Although fires like this were
fairly com­mon­place at a time when highly flam­mable nitrate film stocks were
used, the fact that the film had been des­troyed on two sep­ar­ate occasions
seemed to imply that The Passion of Joan of Arc was destined not to sur­vive in
any form.

Over the years that
fol­lowed, sev­er­al incom­plete prints and mater­i­al believed to rep­res­ent the
second lost neg­at­ive re-surfaced; as a res­ult numer­ous ver­sions of the film
have been edited togeth­er and screened for inter­na­tion­al audi­ences, but
Dreyer’s ori­gin­al ver­sion remained lost, pre­sum­ably forever. 

Then in 1981, one of
the ori­gin­al prints struck from Dreyer’s first neg­at­ive was miraculously
dis­covered at a Norwegian psy­chi­at­ric hos­pit­al. The Danish Film Institute (DFI)
imme­di­ately acquired this mater­i­al, which came to be known as the “Oslo
print” and cre­ated a new pre­ser­va­tion neg­at­ive, guar­an­tee­ing a secure
future for the film and keep­ing the ele­ments in optim­al con­di­tions for years to
come. Finally, after over half a cen­tury, The Passion of Joan of Arc could
finally be seen exactly as Dreyer had ori­gin­ally intended.”

Before-and-after res­tor­a­tion images from Joan.

So with all that history
facing you, the respons­ib­il­ity you bear to remain faith­ful to the film and the
mater­i­als at hand should be always first and fore­most on your mind. Given that
we had a spe­cif­ic budget and fairly tight sched­ule to work by, my approach was
to do the best job pos­sible but to be wary of not “over-restoring”
any­thing. The entire film required sta­bil­iz­a­tion due to the shrink­age and
sprock­et wear the print had endured through­out its 80-plus year lifespan. Every
join between shots required manu­ally re-setting as there were bumps on
lit­er­ally every cut – a fairly daunt­ing task for a film as cre­at­ively and
heav­ily edited as Joan. Density
incon­sist­en­cies mani­fes­ted in the form of heavy flick­er­ing had to be
sig­ni­fic­antly reduced. There were numer­ous instances of dam­aged and missing
frames, often requir­ing new frames to be digit­ally inter­pol­ated, as well as a
steady stream of heavy and light scratches through­out. Not all of these issues
could be remedied com­pletely, and some prob­lems could only be marginally
improved. If you watch the restored ver­sion of Joan, the film still exhib­its quite a few of these
issues, although it’s now a frac­tion of what we star­ted with. The important
thing is that none of the work we did com­prom­ised the ori­gin­al photochemical
look and feel of the film, and the ori­gin­al tex­ture of Joan remains. Fortunately most of the feedback
we’ve received on the Blu-ray has been very pos­it­ive, so I think on the whole
people appre­ci­ate this approach, even if the res­ults aren’t 100% pristine,
which would be a very tall order with any film from this era. 

Zombie Flesh Eaters was quite a bit dif­fer­ent, in that it’s a film
that’s been widely avail­able for years, most recently in a fairly decent
Blu-ray present­a­tion from Blue Underground. But Arrow felt strongly that the
film deserved bet­ter, so they decided to fund a new res­tor­a­tion of the film on
the basis that we would be gran­ted access to the ori­gin­al Techniscope
neg­at­ives. A little dig­ging revealed that these ele­ments had­n’t been made
avail­able for trans­fer since the film’s ori­gin­al print run (con­trary to a good
deal of mis­in­form­a­tion that’s been cir­cu­lat­ing on the web), so we had a good opportunity
to cor­rect for some home-video crimes of the past and restore the film to its
ori­gin­al release presentation. 

Techniscope was
a fairly pop­u­lar format in Europe through­out the 1960s and 70s – it was
essen­tially a low-budget ver­sion of Cinemascope, using two per­for­a­tions instead
of the stand­ard four. After the ini­tial neg­at­ive is pro­cessed and edited,
sec­ond­ary ele­ments (Interpositive, Internegative, etc) are pro­duced by blowing
up the image to fill the 4‑perf frame. Having worked on anoth­er Techniscope
pro­ject a couple years ago for the BFI (La Vallée, Barbet
Schroeder, 1972) I knew that using the neg­at­ives provided the abil­ity to
cap­ture a good deal more image area in the film frame than any of the other
print­ing ele­ments in the film chain, as the blow-up pro­cess would have forced
the crop­ping of all sides of the image in order to fit with­in the 4‑perf
frame. So work­ing from 2K scans of these ori­gin­al neg­at­ives, I was able to
bring the col­or palette, the grain rep­lic­a­tion and all the details present on
the neg­at­ive back to its prop­er place, as well as reveal a good deal more
pic­ture area on all sides of the frame for most of the film. None of the work
was excep­tion­ally chal­len­ging, although the foot­age from the famed shark-zombie
sequence was blighted by a lab fault that proved impossible to do any­thing but
digit­ally min­im­ize – for­tu­nately this issue has always been a part of the film,
as it was prin­ted into every ver­sion any­one’s ever seen! 

The reviews we’ve
received for Zombie have been
over­whelm­ingly pos­it­ive – con­sid­er­ing how it was rep­res­en­ted on VHS here in
Britain for so many years, I think people were amazed to see the film looking
as good as it did! That said, It should­n’t really sur­prise any­one famil­i­ar with
Fulci’s work – he’s a dir­ect­or with an amaz­ing eye, the cam­er­a­work by Sergio
Salvati is fre­quently stun­ning, and the com­bined use of loc­a­tions, col­our, and
music by Fabio Frizzi make Zombie
one of the best films of its kind. That said, it does bring up something
inter­est­ing in terms of your ques­tion about prop­erly serving a given
film’s aes­thet­ic. Might some view­ers actu­ally prefer the rough VHS-era
rep­res­ent­a­tion they grew up with, com­plete with faded col­our, hor­rible sound,
video snow and tape dam­age? It cer­tainly makes for a very different
view­ing exper­i­ence, but one I sup­pose that should be treated as just as valid,
as it was viewed by so many people, par­tic­u­larly in this coun­try, in exactly
this way. 

So in a nut­shell, my criterion
is to rep­res­ent the film in as close an approx­im­a­tion to how it originally
appeared on screens as pos­sible. More than being just a tech­nic­al preference
though, I view it as a true respons­ib­il­ity that I wish every­one would take very
ser­i­ously. Which is why I’m dis­ap­poin­ted when Gaumont Pathe chooses to apply a
headache-inducing level of de-graining to Les Enfants du Paradis. Or when Hammer chooses to replace the original
optic­al effects in The Devil Rides Out with new, “modern-day” digit­al ver­sions. Or when so many
recent res­tor­a­tions of clas­sic films com­ing out of Italy are digitally
sharpened to the point of resem­bling a Playstation game. Or when Pathe choose
to put Le Samorai through
so much image pro­cessing that each frame of Alain Delon in his rain­coat and
fedora end up resem­bling an oil paint­ing more than any­thing cap­tured on film.
These are extreme examples, but there are plenty of oth­ers. Given the expense
of restor­ing a film, prac­tices like these should be act­ively dis­cour­aged, as
the res­ults will likely be the only means of see­ing that par­tic­u­lar film title
for years to come. Considering the fact that the use of cel­lu­loid is quickly
van­ish­ing as we speak, and that very soon simply being able to pro­ject 35mm
will be a thing of the past, it’s essen­tial that the tan­gible, tex­tur­al look of
cel­lu­loid is pre­served prop­erly and accurately. 

 5) Back in the
late 1990s, I was at an event in New York City at which the great
cine­ma­to­graph­er Jack Cardiff spoke. In the audi­ence was Robert A. Harris, who
asked him if the lav­ender tint on Black Narcissus was deliberate.
Cardiff’s answer was some­thing like “What lav­ender tint?” and hence a
great error in film preservation/presentation star­ted on its way to being
cor­rec­ted. I believe we first cor­res­pon­ded after the ITV res­tor­a­tion of Narcissus that you were involved in, a pro­ject that
occurred even after the remov­al of the lav­ender. My ques­tion is, how important
is the input of someone involved in the ori­gin­al work, and in the event that no
sur­viv­ors are around to sup­ply input, what are the research roads you and your
team find yourselves most fre­quently traveling?

Well first I should make
it clear that I was­n’t part of the res­tor­a­tion team respons­ible for Black
Narcissus, although I was there in
the peri­phery when it was hap­pen­ing. That res­tor­a­tion was pro­duced by the
archive team at ITV, with Thelma Schoonmaker and Jack Cardiff supervising
through­out the pro­cess. I was for­tu­nate enough to be invited to some
work-in-progress screen­ings in which dif­fer­ent film stocks were being tested to
rep­lic­ate the ori­gin­al 3‑strip Technicolor look of the film. Thinking back,
this served as a pretty good illus­tra­tion at the time of the lim­it­a­tions of
mod­ern film stocks and their abil­ity to rep­lic­ate the look of older formats
accur­ately, and the help that digit­al, if used with care, could provide in this
regard.  I was lucky enough to meet with Jack Cardiff off and on dur­ing my
years at the BFI, and true gent that he was, he was always happy to answer the
numer­ous ques­tions I was con­stantly spring­ing on him (and I cer­tainly wasn’t
the only one). I’ve also had the great pleas­ure of work­ing with Thelma
Schoonmaker, most recently on BFI’s Blu-ray res­tor­a­tion of what Michael Powell
con­sidered his first “true” film, The Edge of the World – an amaz­ing film, if you haven’t seen it.

Anyway, there’s no real
sub­sti­tute for hav­ing the ori­gin­al tal­ent in the suite with you while you’re
work­ing on these films. Unfortunately when work­ing on older films this just
isn’t pos­sible a lot of the time, so it’s imper­at­ive that one has a real
under­stand­ing of a film’s his­tory and the way it’s sup­posed to look and sound.
One of the alarm­ing aspects of the expanse of digit­al tech­no­logy is that more
and more people enter­ing this industry will have had vir­tu­ally no hands-on film
exper­i­ence. How is one sup­posed to accur­ately restore a film if one doesn’t
under­stand the pho­to­chem­ic­al prop­er­ties of the medi­um? Small decisions can have
an enorm­ous impact on a way of film is rep­res­en­ted, and those decisions need to
be informed ones. 

I’ve already mentioned
the import­ance of main­tain­ing the ori­gin­al film grain struc­ture, but grading
the image cor­rectly is just as import­ant. From Eastmancolor to 2‑strip and
3‑strip Technicolor way back to Dufaycolor or the exper­i­ments of Friese-Greene,
there’s been so many ways to inter­pret col­or through­out the his­tory of cinema
that one needs to be very care­ful in how those col­ors are rep­res­en­ted in any
new res­tor­a­tion.  The tempta­tion to make the col­ors pop, to over-saturate,
to make them much deep­er and heav­ier than they would have looked on an original
film print, is one that needs to be kept in check. Working from negatives,
espe­cially those that haven’t been sub­jec­ted to much fad­ing over time, gives
one a great deal of lat­it­ude when grad­ing, and as a con­sequence the films can
be as graded as warm or cool as one likes, so one needs to approach the grading
with care. It’s the same story with black and white titles. There’s a tendency
by some people to push the con­trast on their black-and-white titles a bit too
heav­ily, in an effort to make the image appear bolder and sharp­er, thereby
giv­ing it a look pos­sibly more dra­mat­ic to the mod­ern eye. But going this route
is often incor­rect, as most film print stocks from earli­er years would­n’t have
been able to sup­port this look. As a con­sequence the tones in the grayscale
often get lost and the sub­tleties of the ori­gin­al light and shade are simply
“graded out” in favor of heav­ier con­trast. Fortunately, the adoption
of HD and Blu-ray have brought improve­ments to this area, as the higher
res­ol­u­tion allows for prop­er rep­lic­a­tion of these subtleties. 

My thoughts about audio
are pretty much the same, mean­ing I’m really not in favor of present­ing any
film in an audio format not from its time. So 5.1 or 7.1 mixes for film
ori­gin­ally released in mono I real­ize it’s done as an effort to make young
audi­ences more likely to accept these older films presen­ted in this way, but
it’s his­tor­ic­ally irre­spons­ible and about as con­vin­cing to the ear as
“elec­tron­ic­ally re-channeled for ste­reo” was in the 1960s. 

I’ll give you an example
of the range if choices you can be presen­ted with. A few years ago, I was
work­ing on Terence Davies’ The Long Day Closes (1992). The grad­ing oper­at­or and I had a day to get
things ready before Terence Davies and his DP Michael Coulter arrived, so we
spent part of that after­noon apply­ing an over­all grade to the first couple
reels. Right away we could see that the col­ors held in the neg­at­ive were so
bold and rich, that if we notched up the sat­ur­a­tion a bit it looked just like a
Douglas Sirk film! So for fun we graded the film as if it were, with bright and
lur­id primar­ies on show through­out. Now as any­one who’s famil­i­ar with the films
of Davies knows, he’s extremely spe­cif­ic when it comes to col­or and has often
used it in a cre­at­ive way to evoke spe­cif­ic feel­ings for the past. So as soon
as he and Coulter arrive, the first thing they tell us to do is drain all the
col­or from the image and to start again. As with Distant Voices, Still
Lives (1988), Davies and
Coulter had applied a bleach-bypass to The Long Day Closes giv­ing it a care­fully faded, almost sepia feel
through­out. As the mater­i­al we were work­ing from had­n’t been sub­jec­ted to this
lab pro­cess (this is tra­di­tion­ally done later on in the print­ing ) we could
have very eas­ily graded the film in any num­ber of ways, includ­ing our attempt
at 1950s melo­drama. Of course we would­n’t have done this as we were well
famil­i­ar with the film and it’s print­ing his­tory, but one could eas­ily imagine
a scen­ario in which the per­son restor­ing the film has­n’t done his
home­work, or does­n’t have the film­makers on hand to show him the way, and you’d
wind up with a ver­sion of the film that would be com­pletely wrong, both
his­tor­ic­ally and aesthetically. 

That said, there are times
when one can be too defer­en­tial when work­ing with ori­gin­al tal­ent. I’m talking
about those rare occa­sions when a dir­ect­or has decided this is his opportunity
to “fix” an aspect of the film he was nev­er sat­is­fied with, or
“update” it in some way to bring into line with his cur­rent thinking.
The res­ults can often be inter­est­ing, but pro­du­cing some updated ver­sion should
nev­er be done at the expense of present­ing the film as it was originally
released. I don’t have a prob­lem with Lucas or Friedkin or Coppola re-working
their film every so often provided they allow the ori­gin­al films to survive
intact. It’s when this does­n’t hap­pen that I feel there’s been an abuse of
power, as some­body’s per­son­al agenda has cre­ated a delib­er­ate dis­tor­tion of a
film’s his­tory. I’ve nev­er sub­scribed to the idea that the dir­ect­or or any one
per­son “owns” a film – a film belongs to his­tory, and it’s our
respons­ib­il­ity to ensure that that his­tory is accur­ately represented.

UPDATE: James White writes in, via email: “To answer Mr Kroll’s ques­tion in your com­ments sec­tion, how these res­tor­a­tions are archived really comes down to the level at which they’re restored. The cur­rent industry spec for archiv­ing file-based film work is to store them on LTO5 archive tapes, which, as per good archive prac­tice, should be cloned and the cop­ies stored sep­ar­ately. If a res­tor­a­tion is being pro­duced to the level of 2K/4K res­ult, then the res­ol­u­tion would be high enough to jus­ti­fy pro­du­cing a new Digital Internegative (DI), which was until recently the source for cre­at­ing new res­tor­a­tion prints. Truth be told, there’s still no better-proven long-term stor­age medi­um bet­ter than film, but sadly this seems to be hap­pen­ing less and less – a clear con­sequence of the dimin­ish­ing need for film prints and the drive to go file-based for as many things as pos­sible. That said, one should take heart that restor­ing a film anew would nev­er res­ult in the ori­gin­al film ele­ments being dis­carded. On the con­trary, the avail­ab­il­ity of a new res­tor­a­tion, at whatever level its been pro­duced, would likely mean that the film ele­ments would then be left alone and safely pre­served for years to come. ”