The new DCP of the digital restoration of The Tales of Hoffman was the final matinee screening at Cornerhouse in Manchester before the move to HOME. The post-screening discussion was led by Andrew Moor of Manchester Metropolitan University. Andrew wrote a piece on the film for Criterion’s website and also co-edited a book on Powell and Pressburger’s films with Ian Christie. The discussion was dominated by the audience members who were primarily music/ballet/opera fans. Since I know little about any of these art forms I found this illuminating but slightly frustrating and here I want to focus on the film as an Archers production from Michael Powell and Emeric Pressburger.
The Tales of Hoffman is interesting for several reasons. It represents in some ways the fruition of Michael Powell’s long-held desire to make the ultimate ‘composed film’ – to marry music, dance, theatre and film as a single coherent work. But to do this Powell had to work quickly and cheaply at Shepperton in order to comply with the Archers’ contract with Alexander Korda’s London Films. The film was really Powell and Pressburger’s last attempt to deal with Korda and after this production they bought themselves out of the contract and took three years off – a long ‘rest’ for such an active partnership.
Powell commissioned a new English libretto for the opera. Emeric Pressburger had less to do on the script this time – although unlike Powell he had actually ‘experienced’ the opera, playing “second fiddle in the orchestra in a production in Prague”. Powell’s plan was to record an opera performance conducted by Thomas Beecham (the originator of the project) and then to ‘compose’ the film on a silent stage with actors miming to the playback. He thus created one of the earliest forms of ‘music video’. This approach also helped him to use ‘real’ ballet dancers, ‘real’ singers and ‘real’ actors. Only two of the cast, the Americans Robert Rounseville and Ann Ayars, were both singers and actors in the narrative.
The Tales of Hoffman was the only opera written by Jacques Offenbach (who mainly produced operettas) and he died a few months before the completed work was first performed in 1881. The story is based on three tales written by the German Romantic writer E. T. A. Hoffman between 1814 and 1818. The opera uses a fictionalised version of Hoffman himself as the hero of each story with the framing device of the ‘telling’ of the tales in a tavern. For the film the Archers added a ballet sequence at the beginning and the end, placing the tavern sequence as a potential meeting place for Hoffman and the ballerina. There are many descriptions and analyses of the opera and the BFI website features an extensive look at the restoration with images from the film and other materials (which they don’t want to offer for download – the images on this blog were obtained from other sources).
The great coup for the production was to persuade Moira Shearer to dance in two sequences. Made into a star by The Red Shoes, Shearer was sought by many film producers but refused them all, only agreeing to work with Powell. Alongside her the Archers were able to cast many leading figures from the ballet world. Just as important for the production was the creative team of Hein Heckroth and Arthur Lawson in production design and art direction, Reginald Mills as editor and Chris Challis as DoP with Freddie Francis as operator.
I think this screening completed my ‘set’ of Powell and Pressburger films. Although I can’t really appreciate the music or the dances, I can admire the cinematic ‘composition’ that the Archers created and especially the genius of the set design, performances and camerawork/editing. In a sense the film takes us back to Powell’s early experience with Rex Ingrams in Nice in the 1920s and to Pressburger’s early career in Germany. What is most fascinating for me is to see all the links to the Archers’ early Technicolor successes. The final tale is set on a Greek island and the designs reminded me to some extent of The Life and Death of Colonel Blimp (the Western Front battlefield) the prologue also reminds us of the meeting of British and German officers in the bar café at the early part of Blimp. Elsewhere we had overhead shots and a staircase reminiscent of A Matter of Life and Death and the whole film referred constantly to techniques developed for Black Narcissus and The Red Shoes. The casting too includes many of the ballet stars from The Red Shoes (Shearer, Tcherina, Helpmann and Massine) plus the third of Powell’s great loves of the period, Pamela Browne as Niklaus, Hoffman’s companion (a male part usually played by a woman in the opera).
Perhaps the most important outcome of watching The Tales of Hoffman for me was that it sent me back to reading the second part of Michael Powell’s long autobiography Million Dollar Movie. I first read it on publication in 1992 and I had forgotten many of the stories. He gives rare insights into the production process and the battles with Korda. All lovers of P&P’s work must have mixed feelings about The Tales of Hoffman. In one sense it represents the peak of their achievements in ‘composed’ films. Powell himself rates it as a ‘bulls-eye’ for The Archers in their four Korda productions of 1949-50. I think I prefer A Small Back Room (1949). Hoffman does not have the same glorious melodrama feel of The Red Shoes and it did seem to me that the camera felt slightly more constrained in its movements during the ballet scenes. Sadly the last three Archers films though all interesting and entertaining did not raise the spirits in quite the same way as their 1940s’ films. Nevertheless it would be interesting to see digital restorations of Oh Rosalinda! (1955 in ‘Scope), The Battle of the River Plate (1956) and Ill Met By Moonlight (1957) – the last two both in VistaVision.
Here’s the trailer for the Hoffman restoration. Even if you don’t know opera or ballet, it’s a real treat for the eyes:
This vies with Phoenix as my film of the year (i.e. seen in a UK screening). It’s a perfectly formed art object that is both engaging and moving. It has been celebrated around the world and has recently been in UK cinemas after a winning a prize at the London Film Festival a year ago. We’ve had to wait a year but it has been worth it. I’ve watched the film twice now and in between screenings I spent a couple of days researching the work of the director Pawel Pawlikowski for an introductory talk. I enjoyed the research very much because it seemed that as I re-watched clips from the earlier films I’d seen and sampled some of the director’s TV documentaries from the 1990s, I began to see the continuities and the links between ostensibly different projects. Ida has been seen either as Pawlikowski’s ‘comeback’ film or as a revelation for those who have not known about the earlier films. Whether or not the concept of a director with ‘personal vision’ as an auteur still has mileage, there is no denying the continuities between Pawlikowski’s films.
Pawel Pawlikowski left Poland with his mother some time in the 1970s and eventually arrived in 1977 in the UK aged 19. His career as a filmmaker began in the late 1980s with documentaries mainly focusing on quite controversial figures in Eastern Europe. My notes for the film introduction can be downloaded here: Pawlikowski Background Notes.
Ida is a short film (just over 80 mins) and a plot outline would suggest that relatively little happens. A young woman, Anna, who we presume has grown up in a convent is now 18 and in preparation to take her vows when she is told that she has a relative, her aunt Wanda, and that she should visit her before she takes the final decision to commit to Christ. Anna is not sure what to expect but Wanda eventually provides surprising information about Anna’s Jewish family, including Anna’s birth name, ‘Ida’. Reluctantly at first (in Wanda’s case) the pair undertake a road trip to uncover the past. They meet relatively few people and the ‘action’ is limited but there is so much going on in the unspoken exchanges between characters and in the presentation of sound and image that we experience an immensely rich narrative.
There is no better way to introduce the film than to give the director’s opening statement (from the Press Pack):
Ida is a film about identity, family, faith, guilt, socialism and music. I wanted to make a film about history, which wouldnʼt feel like a historical film; a film which is moral, but has no lessons to offer; I wanted to tell a story in which ʻeveryone has their reasonsʼ; a story closer to poetry than plot. Most of all, I wanted to steer clear of the usual rhetoric of the Polish cinema. The Poland in Ida is shown by an ʻoutsiderʼ with no axe to grind, filtered through personal memory and emotion, the sounds and images of childhood . . .
It seems to me that Pawlikowski succeeds in each of these aims. (Though the ‘outsider’ bit is possibly something for Polish audiences to comment on and they haven’t taken to the film as much as audiences in France or even the UK as far as I can see.) Much of the strength of the film does come from the director’s perspective – as an insider who became an outsider and who now returns unafraid to ask questions and ‘re-present’ the past. The power of the film comes from the astounding attention to the detail of the visual and sound images and the performances of the cast, especially the three leads.
The film was shot digitally using an Alexa 4:3 camera and the raw footage was then processed to create a monochrome film with a traditional Academy ratio and then further processed to add the grain effect of the filmstock used in the early 1960s. This process is described by ‘Benjamin B’ in his blog, ‘The Film Book’, which carries two features complementing an initial feature in American Cinematographer. These are a must read for anyone interested in the filmic image. There are several notable features of the images created in the film. As Benjamin B comments in his analysis of a short sequence, Pawlikowski and cinematographer Lukasz Zal, create pure cinema, ‘showing’ not ‘telling’ the story through a combination of acting, camerawork and sound related to a carefully structured narrative outline. The film also offers good examples of the old adage about needing a great deal of artifice to represent an image of ‘reality’. To achieve the ‘effect’ of natural light and simplicity in the depiction of the convent required careful placing of key and fill lamps. The effect works very well.
The two central characters of Wanda and Ida are played by Agata Kulesza, a vastly experienced actor and Agata Trzebuchowska a non-actor and this fits the narrative perfectly as Wanda has to drive the narrative and Ida has to respond to what happens. I was intrigued to discover that Agata Trzebuchowska had seen Pawlikowski’s earlier British films even though she was not involved in the film industry. I’m delighted that films do travel more extensively than might be apparent from the relatively meagre information we get from outside the US/UK film world.
I think Ida is going to be one of those films that “keeps on giving” – offering up new insights into how it can create meanings through camerawork, lighting, design, sound and performance. It has also prompted me to find the few John Coltrane recordings that ought to be played more often. The most striking aspect of the beautiful visual compositions is that they often place the characters in the bottom third of the frame, utilising a low horizon in landscape shots and producing a great deal of ‘sky room’. I haven’t quite decided what this means but it is distinctive and it certainly suggests a dialogue between characters and their environment. It also reminds me of the big skies in Academy frame compositions by directors like Kurosawa Akira and John Ford.
Since I started this posting several weeks ago the European Film Awards in December made Ida the big winner with a total of five awards – two for the film itself, two for Pawel Pawlikowski as writer and director and one for cinematography (shared by Lukasz Zal and Ryszard Lenczewski who both worked on the film). If there is any justice the film would win at the Oscars as well. If you haven’t seen it, the DVD is now out but do try and find it on a big screen.
Global film production, post-production, distribution and exhibition has now reached a point of no return in relation to ‘digitisation’. Writing about the experience of watching films in cinemas has become problematic because the industry is in a state of flux and it is easy for any of us to get confused about what is happening. This posting is an attempt to lay out the current state of digital film and digital cinema as I understand them. Please add any other anecdotes, explanations, suggestions for additions etc. as a comment.
What we finally see on a screen as a moving image sequence depends on at least four separate processes. The first is image capture. It is often difficult to determine the exact format that was used in shooting a film. Today it is possible to shoot on virtually any known format from 16mm film through to an iPhone or a toy camera. It doesn’t really matter because anything can be digitised. Of course, the more image data that is captured (the higher the ‘resolution’), the more options are open in post-production. However, it is still possible to lose the advantage of high quality images if they are not processed correctly.
Editing is now routinely digital since all the source material has been digitised. This was the first part of the process to be converted. During the post-production process it is possible to manipulate images so that they resemble different sorts of film material. The end product of post-production is a ‘digital intermediate‘. This could still be printed back to 35mm film for distribution and projection but it is now most likely to be processed to produce a digital master and a DCP – a digital cinema package for digital cinema exhibition. The same master will also produce a range of digital formats for digital download, digital TV and DVD/Blu-ray, each of which will have different specifications. Films are now edited/post-produced in the knowledge that they must look good on several different formats The distributor creates the DCP and the exhibitor must ‘unlock’ it and decompress and download it for projection using a Theatre Management System (TMS) which places the film in a projector menu alongside ads, trailers and other material and probably commands to mask the screen, open curtains, lower lights etc.
So this is the basic process. Unfortunately it isn’t quite so simple in practice. The Hollywood majors want to remain in control of distribution in the major territories and so the seven studios (reduced to six when MGM became part of Sony) set up the Digital Cinema Initiatives (DCI) in order to create an international standard for digital cinema. You can access the specification and background details here. Any distributor or exhibitor that wants to handle Hollywood product going into cinemas must comply to the DCI standards (set by SMPTE, (Society of Motion Picture and Television Engineers) and confirmed by ISO). DCI compliant technologies are part of ‘D-cinema‘. These standards are concerned with the creation of DCDMs (Digital Cinema Distribution Master) and DCPs and their projection. From the outset, the DCI standard was designed to include both 2K and 4K specifications. This means that either a 2K or a 4K DCP can be sent by a distributor to a cinema. The cinema may have either 2K or 4K projectors and it is possible to ‘extract’ a 2K file to project from a 4K package. Similarly, a 2K package can be ‘up-sampled’ by a 4K projector. Major cinema chains globally are now beginning to invest in 4K projectors – but most films are still being distributed in 2K DCPs.
Norway was the first territory to become completely digital for cinema distribution and other European territories are approaching 100%, although in larger territories with many small single screen independent cinemas the process may take longer. However, D-cinema is not the only digital cinema technology. Lower resolution digital formats have become known as ‘E-cinema‘ and in India E-cinema is in operation via satellite distribution, supplying smaller rural cinemas while in metropolitan multiplexes DCI-compliant prints are projected. The Indian satellite distribution system may produce a lower resolution image but the economics of the system make more sense and it’s possible that this form of E-cinema might be more suitable elsewhere in other parts of Asia and Africa. Since anything ‘sub 2K’ is classed as E-cinema there are already a range of E-cinema sites in Europe and North America, small community cinemas or screening rooms projecting from DVD or Blu-ray. (Sorry Keith, but for some exhibitors, Blu-ray is de facto a theatrical format.) Similarly, most film festivals now accept films on a variety of digital formats including HDCAM SR/HDCAM from Sony and the slightly lower specced DVCPRO HD from Panasonic. Unfortunately some also accept Digibeta or Beta SP. The problem is that digital projectors need careful treatment by knowledgeable projectionists to get the best out of different formats and in a festival context, even the best technicians don’t have time to tweak settings between showings so films that look great at one festival look terrible at another. (This rant from a US website offers an interesting perspective on the problems of preparing a film/digital file to show at festivals in North America – there is a lot of sensible info here and I certainly recognise the problems as seen by festival audiences.)
The unresolved question for cinemas that have still not converted is still who pays for this conversion to digital? The so-called Virtual Print Fee (VPF) is supposed to work by ‘spreading the load’ between the distributor and the exhibitor but it doesn’t work for everyone and especially for small distributors.
New distribution and exhibition practices
Rumours are circulating in the UK about the new distribution practices in a digital environment. I’ve heard stories that distributors are not maintaining the DCPs of films beyond their ‘normal’ release. The hard drives can easily be re-cycled/re-used so once a film has finished its run, the print won’t be kept. I don’t know if this is the policy but in the last couple of weeks I’ve had two education screenings. The first was for a 1990s film, a classic already re-released in the 2000s. It’s just too old to have been released as a DCP so the distributor sent a Blu-ray disc. This was an improvement on the last time I showed the film a few years ago when they sent a DVD. The Blu-ray looked very good on a very big screen. Apart from a few over-dark scenes I wouldn’t have noticed standing at the back of the auditorium. The second film was released as a ‘specialised film’ title earlier this year and I watched it on a DCP. Imagine my surprise when we were sent a slightly battered 35mm print. Fortunately, the cinema still had a working 35mm projector. The audience didn’t seem to mind but somebody asked me if the scratches had been added for authenticity and I don’t think that they were joking! The serious point here is how geared up are the distributors to handle education/festival/archive/repertory bookings? Is Blu-ray going to be what we can expect after an initial release on DCP?
Yesterday I heard about a new multiplex that has opened locally. It is completely digital and I’m told that the manager can virtually run the whole operation from his office, ‘dragging and dropping’ films and ads onto different projectors via the TMS. Pretty soon the films will arrive in the cinema by satellite in the UK and another ‘technical operation’ will be removed.
Overall, I’m happy to see the more consistent quality that we get from DCP, especially in multiplexes. But it comes at a cost in terms of employment and ‘de-skilling’ of projection staff. This was recently demonstrated in the UK by the widely-reported incident in which a projectionist in a multiplex, presumably looking after several screens at once, projected the first few scenes of a gruesome Cert 15 horror film to an audience of young children expecting a family animation. I’ve also been told horror stories about satellite links going down in live broadcasts – these now include Q & As with directors as well as live feeds of opera, ballet etc. My feeling is that satellite is a necessary evil in countries with transport problems, but I’m not totally convinced by the current technologies available in a country as densely populated as the major urban centres in the UK.
Finally, there is the important question about formats for proper film archive storage. Digital is not a good long-term storage medium since the longevity of physical discs and tapes etc. is not yet proven. But just as important, each time the technology improves, archivists will need to maintain a working example of each playback device. Keith is our local expert on archives, so I’ll leave him to write about these issues. He has already pointed us to the website of the FIAF.
There is some confusion over the broadcast formats of the the three Red Riding films, so I’ve taken a screen grab from each film and measured each image in terms of the pixel matrix to calculate the aspect ratio.
I calculated this image to be 1086 x 608 pixels on my computer screen (it’s scaled down here) and that equates to a screen ratio of 1.79:1. I may be one or two pixels out given the way I use the grabbing software, but no more than that, so I’m fairly confident that the Channel 4 image is 1.78:1 , i.e. the standard 16:9 of the modern widescreen TV set.
Using the same procedure on the grabs from 1980 and 1983, these came out as 1086 x 476, equating to a screen ratio of 2.28:1, which is slightly less than the cinema projection standard for CinemaScope of 2.35:1. I find this a bit strange. No doubt Channel 4 alienated a small proportion of viewers by showing the films in ‘Scope (especially given how murky 1980 becomes). But why compromise on 2.28? Why not 2:1 or the full 2.35? Is this in any way related to the use of Super 16 or the Red One digital camera? Or is this just Channel 4 ‘house style’? Of course, it could also be an issue to do with how the TV signal is broadcast or received. Mine came via cable, set to letterbox for my 4:3 TV set.