Whilst watching my short film I hope you realise how much effort, dedication and soul went into producing it; only taking 4 hard months of pre-production, 4 days of hell in production and 2 months of vigorous post-production. One can see the level of thought and passion through the images on screen. The journey was a long but one that had taught me far more than I had ever imagined. Creating such an ambitious project was part of my plan, I knew it was going to be difficult, but I wanted to prove myself as a media producer that anything could be possible as a student. Majority of people were shooting films in a house, I was shooting it in a nightclub. I wanted that extra bit of showreel material that would sell myself to employers, this was the guy with the biggest student project and he was the one who pulled it off.
Working together with a very talented writer I was able to make my ideas heard, the writer would do careful research into each part of the story so that the film would be realistic and legitimate. Everything in the film had meaning; nothing was ever random or just put there for being for the sake of it. Everything was governed by an interesting ideology; this gave depth to all of the characters as well as the events in the script. It was the importance of understanding the writer’s vision that allowed me to transfer his script to visuals.
I learned that money is the one of the biggest parts of filmmaking; I could only rent the best camera with a hefty amount of cash, I could only use the best sound equipment by paying a big rent fee. And if I offered to pay less or not at all, then people would just drop-out – like the first DoP and multiple sound recordists. Actors dropping out plagued me less, but I should have issued legally binding contracts to actors taking part in the production to minimise this – I only learned this when Levis dropped out a day before the shoot, which almost ruined it. Mentioning expenses meant nothing, you always had to have that extra bit of cash to offer as payment for people’s time. It was really important to do so. People who are good at what they do and charge nothing is impossible to find in today’s society, on top of expenses you have to pay people, and even a small amount will suffice. Even if you do show people your showreel and state that you are fantastic, there is no guarantee that someone will watch it and fully commit themselves to you. Contracts and payments are the way forward.
And then there were the extras (an integral part of the film that would actually make it believable as film) a part that I completely overlooked, and to no surprise, failed at. But this made me learn that organising extras was another science that had to be respected and followed. I had to remember that they were not getting anything from it in terms of showreel so I had to offer them something other than payment and arrange and confirm them individually. I was lucky enough to meet a person on the first day that would prove to be an invaluable contact in getting people to act as nightclub extras thanks to his son’s company, Destiny Live Performance. It made me learn that making good contacts as a media producer was crucial because the more people you know the more you can organise, the more links you make and the more professional you get as a filmmaker. The more you can achieve. This took me from zero experience of organising extras to knowing exactly what to do if I needed them in the future.
Directing on the set was insane. At points I just wanted to drop dead on the floor and shut my eyes to the world – but I couldn’t, I had to prevail with an iron fist. Last time I worked on a 15min short film was back in the 2nd year of college (I went from 2 years of HND to final year top-up at Coventry University) and we didn’t even arrange any other actors apart from our college group. It was child’s play compared to this film. You’d probably imagine me being completely mad going from something as petty as this to something as grand as ‘End of Nights’ but I needed to fill in that missing gap of familiarity with proper film sets, working with proper actors and a professional crew (which I also ALL produced with my fellow producer). Just 4 days of production gave me first-hand experience in working on a big set – it is not for the weak, it is not easy and it is certainly one hell of a task. As the director you have to be the man in charge – you are the one calling all the shots – everyone is under your command. You have to make very important creative decisions in order to make the film succeed, especially when you want it to have no similarities to the work of students. You have to tell everyone what to do and how to do it the way you want it. Now that I can easily sit down and look back at it, I realise that I loved every minute of it. I constantly push myself to the limit because I just want to get better; I want to make great films. I believe I have what it takes.
Project management as a whole was satisfactory. It was rough at parts and only got serious a week before the actual shooting, but can you fault them if I managed to complete the film to the highest of standards as well as keep it affordable? There is room for improvement and I know what needs to be improved and taken care of the future. Legal documents are of utmost importance, sure we got the release forms after the project, but the forms in pre-production were just as crucial. This project has made me realise the importance of every little stage in the process.
In post-production I have never in my life worked with a Red or 4k footage and this project gave me that chance to. I enhanced my technical skills and devised an incredibly high-quality workflow to handle such epic footage. As an editor you gain experience with every project you work on – the bigger and more ambitious projects grant you massive rewards upon completion – you look back and think, wow, I did that. You devise clever little ways of bypassing silly errors and mistakes that were made in production and you make the film flow and give it depth and meaning, this is the part where the script transforms into fantastic visuals. The power of grading gave me insight into how the colours can emit a dominant meaning and combined with talented sound design and an interesting and fascinating story, a legendary film is born. A film I created.
On a finishing note I will quote my friend once again: “It is better to have great ideas but have to work within realistic limitations than weak ideas that you have to try and polish into something better”. I hope you enjoyed my film.
Producer, Director & Editor of “End of Nights”
My last bit of concern was with exporting to a delivery codec (H264 being the most popular one, in a widely international MP4 container) of the highest possible quality. After multiple test exports from Premiere I noticed serve cases of banding and other artefacts, even when set at an incredibly high bitrate. I did some research and found people complaining about the same thing too, issue being Premiere’s poor implementation of MainConcept AVC (Premiere’s H264 encoder).
The best H264 encoder in the world is the open source x264. But how would I export directly from Premiere using x264? I could go from Premiere to Intermediate to x264 encode but I would rather skip the middle part because it’s pointless, wastes time and resources and ‘could’ result in quality loss. It takes about 8 hours to export 15mins of 4k footage into 1080p (has to downsize, crop the shot, crop to aspect ratio, sharpen, any other effects, etc). The answer was either X264PRO or Sorenson Squeeze. The former attached a watermark under trial usage so the latter proved far more superior – you configure the preset in Squeeze itself and then it appears in the Premiere export window, which very neat and user-friendly. X264 offers far more advanced finely tuned profiles (film, grain, etc, based on content), presets for encoding (ultrafast to placebo, the slower the encode the better the quality) and AVC profiles. To obtain the best possible export (and bearing in mind that it takes 8 hours already to process 4k) I learned that:
- Selecting the placebo (slowest) preset will add on an extra 4-5 hours. This allows the encoder to utilise every single tool it has to compress and retain every single amount of detail. This makes the redcode raw footage to be almost indistinguishable with this x264 export.
- Tuned for a ‘film’ look to retain grain and colour with the best possible compression & quality AVC profile (high) and a level of 5.1 to support it all the advanced settings the placebo preset supports.
- Making each frame a keyframe ensures each frame is of the highest possible quality – no need for reference frames.
- A high bitrate of 35000Kbps is merely a bonus to ensure quality is upheld throughout; based on all these settings the encoder will handle the rest. A bitrate of 12000Kbps is suitable for delivery (1.2GB).
- I am cropping the resolution to 1920×816 in this program; every important setting should be set in this program because I will not have a chance to do it on Premiere.
- A nice and subtle sharpen filter to make everything ‘pop’
This gives me a total file size of only 3.7GB, which is decent. At this stage I do not need to compress it further because this is suitable for the web as well as film festivals. I am a quality freak and I have discovered the secret recipe of insane and efficient quality. I have visually tested this x264 export against DNxHD/Cineform HD/ProRes 4444 and it is to the eye, identical. Don’t get me wrong though, intermediate codecs are used to preserve the data the eye doesn’t see for post work. x264 actually holds its own at an incredibly low file size in comparison. This is the reason why the encoder has won many awards in recent years, it is simply the best.
Creating the poster
I was tasked to create the poster thanks to my knowledge in Photoshop. I wanted a really unique design and avoid mainstream poster clichés. The writer really wanted to include the male and female symbols in the poster so I came up with a clever idea to incorporate that. The male symbol is split with Ted and Levis to imply that they are fighting over a woman; Grace who is within the crossing female symbol which implies the winner will take her home. All of this amongst a nightclub inspired background. “Assert Your Dominance” the tagline reads, the whole premise of the film. The one who dominates, is the one who wins.
Post-production on this project was fantastic. I wanted to push myself to the limits by working with a Red for the first time and that I did. It was a whole new experience; I had never in my life used a 4K camera, let alone a Red. The professional post-production program from Red was incredible to use and gave me some really important experience that I can put onto my CV. I was also able to bring life to this film and increase my editing skills a tenfold. I spent many sleepless nights fine-tuning every single edit, crop and adjustment to absolute perfection. Each 15 min film that I do gives me much more knowledge and experience than I can imagine. I love every minute of it and I just want to make more films.
Many people mistake the difference between grading and correction. Correction is when one adjusts the footage to how it looked like through your eyes on the day of shooting and grading is where one adjusts the colours to the desired film look and style. It is generally better practice to colour correct all of your shots first to match. I believe that the best grading is the one that you do not notice; it is the one that gives the feel of the film that pulls the audience in. The test graders on ShootingPeople were so bad I knew I needed to this into my own hands and handle this like a pro.
I had a few possible options for grading and these included: Adobe Speedgrade, Magic Bullet Looks/Colorista and RedcineX Pro. Whilst the first two tools are very powerful on their own I found that the only program that comes to giving Red footage justice is RedcineX Pro. This goes back to the idea that red cameras record compressed Bayer data. This so-called compressed raw data allows white balance, gamma, sharpening, RGB, FLUX control, ISO to be adjusted in post-production. The sheer power of this program and format surprised me, so I decided to put it to the test and colour correct the footage first, and then grade it.
It was surprisingly easy to use (thanks to the advice of the DoP), all that one needed to do to colour correct is adjust the Reds, Greens, and Blues as well as the FLUT Control option. Once all the footage was corrected it could be graded. For grading, other options could be adjusted, for example Kelvin, Tint, Brightness and the Lift/Gamma/Gain setting. The beauty of this program is that the workflow was hassle free. You adjust the look of the image in RedcineX and it is instantly saved to a .RMD file next to the .R3D footage file (non-destructive workflow). Then back on Premiere, one updates the shot to match the just-saved look. Downside is that you have to update each edit to match the new look, however I came across a small utility called ‘Reset RED RMD’ thanks to Fallen Empire Digital, which batch reloads all the RMD files in Premiere saving you a lot of time.
Another really cool feature of RedcineX is the various colour sciences you have available to choose, for example you have REDgamma1-3 for gamma space and REDcolor1-3 for the colour space; REDgamma2 and REDcolor2 were used for the first scene to make it as flat as possible and then colours adjusted and for the rest REDgamma3 and REDcolor3 was used for that glossy, deep/rich colour look then adjusted to look beautifully vile. The founder of RED, Jim Jannard, said this about it:
“At 1st blush, you may just see subtle differences in REDcolor3 but it is a complete remapping of the color space and REDgamma3 makes a much better image on your monitor.. REDcolor3 is probably the bigger update but REDgamma3 is a welcome improvement.”
At first Premiere did not read these colour spaces I had to manually update it to support them, which was a tricky task but I was hit by a significant difference when I did. 5D footage was graded using the Magic Bullet Looks, a plugin I used from the first year. I passed on from using its presets and now have the ability to create my own; it’s fairly powerful but does not come close to RedcineX.
Enough with the technicalities. For the look of the film I wanted a something to match the likes of Se7en, Fight Club, Irreversible, Harry Brown, Company of Men (also our main film influences). I wanted a depressing, gritty, disturbing and dark feel to it. An almost dystopian atmosphere. The councillor scene is dark and dull, the outsides have a green cast to them for the stylistic purpose of making it look vile, cold and dangerous, and the club is all red to demonstrate the themes of madness, insanity, sex and those of a red-light district. The toilet is incredibly green again to show sickness and depravity and the rest is just dark and dull. It’s a cruel and hostile environment with no mercy for the weak.
The film is not meant to be pretty to look at. It is meant to be disgusting and horrible yet maintaining a level of art. The world the characters live in is not pretty. It’s harsh and depraved.
I honestly enjoyed grading; the software was so powerful it gave me great pleasure and experience working in it. One can really see the difference in colours from the rough cut (no colour adjustments whatsoever) to the final graded cut. It’s phenomenal, really. Makes me think I have potential to be a good colourist after all and that I should never have doubted myself! But it’s a massive learning curve and that’s what I’m here for, to learn.
A draft grade of the film: https://vimeo.com/65898419
Premiere’s powerful title system allowed me to create the rolling credits with ease. It was incredibly user-friendly; I was able to create some fantastic looking credits and titles from the first time of using the tool, meaning I can be even more efficient next time.
Believing that I was not up the task of colour grading the film, I set out to find an experienced one from ShootingPeople. I agreed with my fellow producer that we would send in the trailer as a test to see if the applicant was up to the task. Now I needed to find out how to sent it to another person for post-production grading.
The whole point of an intermediate codec is to use it to compress the footage with no visual loss with a file size that is easily distributable for other post-processing, in this case, grading. The best codec for this is ProRes4444 but it is not available to export to it on a Windows machine which I found very upsetting. After some heavy research I found the alternatives were DNxHD and Cineform HD, both capable of outputting to 10bit 4:2:2 and 4:4:4 chroma subsampling. Typically gradists require the highest quality so 444/422 was essential. After extensive testing of both of these codecs I found out their negatives;
- You are stuck with its resolution presets of 1080p/720p
- Cannot output in a custom resolution or anything higher than 1080p
- When imported back into Premiere one experiences a massive gamma shift
- Set of codecs have to be installed on Windows/Mac
- Cineform HD
- I established that the codec was incompatible with the latest versions of Mac OS so I could not use this codec at all to pass onto Mac users
- For some odd reason did not output to 444
- Codec has to be installed
In conclusion it made me furious to find out that Windows users have to go into great lengths to output to a lossless codec when Mac users could simply output to ProRes4444 with a custom resolution – a format that even Windows machines can read and edit by having Quicktime installed. For this reason alone I think Mac editing is superior.
Two people applied to grade the trailer from ShootingPeople so I sent the trailer to them. The results were startling… they were simply bad and almost comparable to the work of first year students. I was let down but happy to know that I did not have to export the film just yet for grading. I would rather keep exports as minimal as possible to keep quality at 100% and do the job myself.
One of the test grades from ShootingPeople: www.youtube.com/watch?v=UUGqmHc5Loc
A poor effort and not what we were going for. It does not like filmlike and there is no style to it. I know it is a trailer but we were able to decide best on that draft.
Post-production on End of Nights began on the 20th March 2013 as soon as I received all of the Redcode Raw files from the DoP – total file size was 251GB. My initial thoughts were on the Red workflow on my Windows machine but I was told by the DoP by Adobe Premiere (my editing program of choice) handled the red files natively with no conversion needed, so editing was going to be as easy as working with DSLR footage. Placing the first file onto the timeline was exciting, and after clicking play the footage played smoothly at ¼ the resolution (since REDCODE is a wavelet codec, the files contain several lower resolution versions of the video, which means the 4k file supplies 1080p footage also). The footage looked stunning. As if I was looking into a portal into the past and seeing the events unfold as I did on the day, the quality was simply unparalleled. Whilst there are a number of different 4k resolutions (http://en.wikipedia.org/wiki/4K_resolution) the one I decided to work at (sequence setting too), as well as having most of the footage shot at, was “4K Ultra high definition television” at 3840 × 2160. This allows me to scale the footage down to 1080p effortlessly (mathematically too). Support for 4k projection and playback is still obviously poor domestically (only cinemas can afford these projectors), so 1080p was the best and only choice.
I love editing, it brings all the separate footage that you have to life. I believe that the editor has the biggest power of the telling story through his edits; he can make it fast paced or slow and drag on the shots for as long as he likes. This process was going to last around 2 months, all the way up to hand in – I was sure to give myself enough time to make the standard as high as possible.
Editing was relatively straight forward using traditional techniques of continuity editing (majority of the film) with instances of montage editing (nightclub dance scene), jump cuts, with L-cuts and J-cuts for dialogue. Me and fellow editor worked very closely on the first rough cut – we edited scene by scene in separate sequences on Premiere to keep everything as organised as possible. After the assemble edit I took over the project personally from the rough cut leading all the way up to the final edit; I was a perfectionist and wanted the edits to flow as authentically, cinematically and professionally as possible – every detail was important. This had to be my best work.
Whilst it may seem to be a controversial choice, but we edited the video first with absolutely no audio guidance. This helps a great deal because you want to see the video flow first with no distraction or the temptation of constraining to the audio. Once we saw the video flow, only then I could start to sync up the audio. This is perhaps an editing preference but you learn so much over the course of 3 years media production you find the more comfortable and professional ways of working. Of course there are little *incredibly annoying* errors in continuity that you have to work around (heads turning differently, hands in different positions, etc). Nevertheless we managed to avoid them every single time whilst keeping up that incredibly high standard of editing; but unfortunately there was just one we could not avoid – and it had to be in the very first scene of the film. Half way through the councillor takes the phone and holds it with 1 hand, and in the next shot she’s holding it with 2 hands. There was nothing we could do about it. HOWEVER, trials of the rough edit established that out of 30 people no-one noticed this error. I tried to lessen the flaw by cropping the shot a little more, it’s still there, but thankfully less visible.
I made sure a rough edit was available to show to the crew members and cast on the final day of shooting (scene 9) so that I could get some feedback. I managed to get two very important pieces of feedback.
1) The conversation from around 1:38-2:44 needed to cut to something else, otherwise it just looked like one massive long take, I decided to cut to their backs. At first I did at times when they were not looking but it looked out of place, so I carefully edited it based on continuity, e.g. when they looked at each other, as well as trying to mask the very shaky camerawork (the shake was intentional but I wanted less of it).
2) In the very original cut (earlier than the rough cut) When Ted gets hit by a can and rushes off to the bar it seems as if he gets to the bar unusually fast. I was given advice to cut back to a few seconds of dancing at that stage and then cut to him getting to the bar, I followed this and it did naturally look better.
Once these issues were rectified and my fellow editor was happy with it I was able to export a rough cut out to Vimeo. The first issue however took far longer to sort because I want perfection at the stage so it was only up till the final edit I was finally happy with it.
I was very secretive with this film and refused to show it to anyone other than the team who made it – this is because I can trust these people professionally. NO, they are not just my ‘friends’ they are people who want what’s best for the film. They are a dedicated team who want to see this film succeed. Why would I get strangers or even worse, the students/people of Coventry to give ME feedback? They probably want to see me fail so it is more than likely their advice will be poor or of a low standard. Furthermore I want most people to see the completed film from the first viewing to get hit by the full power of the film’s message. My editing is my strong point so I know what I am doing; when spending so much time with the film you start to see what exactly needs fixing, work and tweaking. You know what you can and cannot do. I want my own style to it and no-one else to interfere.
When editing a film the second most important thing other than continuity is the flow & pace of the film itself. Each edit has to be paced correctly. It is very hard to describe, but you cannot draw out certain shots for too long or short, but with others you can, you have to make a creative decision each time for the exact length of a shot before cutting to the next. The art of film editing. Of course there are various edits and techniques, but overall it is not something that can be fully taught, but something you learn from experience.
I received enough feedback from the assemble edit and first rough edit so it was more than enough to lead me onto locking the final edit. It was now my task to fine tune every single edit as well as crop the footage, make it adhere to the 2:35:1 cinemascope aspect ratio and eventually colour correct and grade it.
https://vimeo.com/64468259 – Rough cut
2:35:1 is the most traditional cinematic aspect ratio (and my favourite) so I was very determined in releasing the film at this format so I had to find out how!
After a bit of old fashioned research I found a very good tutorial on how to do it here: http://whoismatt.com/cinemascopetutorial/ the only problem was that my footage was 3840 and the math was henceforth slightly different. I calculated that 1920×816 is the resolution that I would need to export to and that 132 pixels would need to be cut from top and bottom of 1080p footage, this meant that 264 (simply doubling that worked) needed to be cut from top and bottom of 3840px footage – I made my own black bars overlay on Photoshop and imported them into premiere for guidance.
The key is to edit at this aspect ratio because you need to reposition every single shot to how you want the audience to see it, with the black bars working as an overlay to aid my repositioning I was easily able to adjust the motion setting on Premiere to make the shot position relative to the aspect ratio. Voila, beautiful 2:35:1 and professional too.
Without bars – (top right) Motion position adjusted to make the shot look right once black bars applied
Throughout the rough cut feedback people were expressing that more close-ups were required and the long shots should have been even close, perhaps to the point of medium shots. I was a little annoyed that we sometimes did not get that range of close-ups so I wondered what I could do to fix that problem. This was remedied with the ‘crop’ effect on Premiere. Some very simple and subtle cropping is very powerful because I am effectively moving the audience closer into the action and pivot points of various scenes. With 4k footage you get the power and ability to crop with minimal quality loss, the image is so damn big in the first place so have the ability to crop everything to 1080p at the very minimum. This occurred in some parts of the film, from the councillor scene bringing the medium shot closer and tidying up some bits of the gang – some is not even noticeable.
Initially we worked up a rough cut of the dance scene, fast montage style edits whilst trying to tell the story at the same time. Everyone is dancing, everyone is having a great time, yet Ted is still reluctant to dance because he is a coward. This is the one of the most exciting parts of the film and I wanted the edit to stand out, be special. After listening through hundreds of creative commons (a slight better quality in music than completely free music) tracks on Soundcloud.com we eventually decided on a specific track for the dance scene Karel Craft – Three Elements by Bake The Break [NETLABEL]. It was incredibly fast-paced and had a great beat that I could sync the video to the audio. I wanted the viewer to mentally dance to the music or at least head nod to the tune. They need to realise how pathetic Ted is, great song, it’s his opportunity to GET the girl, but he doesn’t. A shame. A variety of close-ups and medium shots were used and edited in very fast succession to the beat in the song to give the impression of an insane club atmosphere. Final result is something that I am incredibly proud of.
Scene 9 was shot on a Canon 5D. I’m not sure but one may start to see some loss in visuals. The 5D footage looks really poor in comparison with everything else; it feels like it has no depth whatsoever, it was however, suitable for this one scene when the Red could not be afforded. To put it bluntly, red footage looks like film footage, 5D footage looks like DSLR/slightly-improved-camcorder footage. You can make 5D footage look more film-like but it will never come close to the Red.
Since 50% of the film is sound I decided to pass the audio editing to a top sound designer, who was also the sound mixer on set (he used a 744T to mix). After I had synced all of the audio tracks to the video and locked the edit I exported the OMF file to him so he could use his professional editing software. He was probably the most ‘genuine’ professional person on set and when I was asking him to be careful with a certain track in the edit he replied, “Don’t worry, I’m a surgeon” that phrase alone echoes his incredibly high level of understanding. He sent me the first draft shortly and I was able to give him feedback for improvement, same for the second draft. He was very easy to get along with and knew his area well. All the way until I was perfectly happy with his sound design.