Preroll as DCP vs ALT? Pick your poison.

[Ryan Gallagher]

More a question for my fellow independent cinemas. How does everyone package and handle their preroll content, especially if it tends to change per screening?

We are in the habit of DCPing everything right now, but I could see the benefits of switching. Our 50th anniversary summer series introduced a LOT of per screening extra pre-roll, often film specific, it became as much or more work in DCP-o-matic every day to prep everything per-show per-format etc as our bigger festivals are.

Mostly i'm talking about static slides... we would probably still DCP our in-house video bumpers unless I get 5.1 audio from the ALT source too.

Pros vs Cons?

DCPs leave a bit to be desired in slide transition options, and no support for animations like in powerpoint etc.

My main hesitations for throwing ALT content on screen before any video bumpers is the dance required to get content to fit in film masking, which DCP-o-matic can solve for us. That and I haven't tried sending anything other than ALT178 sources from a laptop, ideally there would be a way to utilize the flat, scope, and full custom 2K rasters on the ALT input.

Missing out on DCI colors is not really a concern, cause most of our slides come to us in sRGB anyway.

But the ability to actually loop a slide deck from ALT would be a big plus.

We commonly lean on OBS for quick and dirty content switching in a non-film context. But even OBS would have fallen short for our pre-roll this season, because this year they wanted different hang-times on specific slides (which DCPomatic can accommodate). It would probably be powerpoint or keynote to drive a pre-roll loop if such requirements persisted on an ALT source. Or just ask them for a power point to begin with and let them get as creative as they want with transitions and animations.

I do like the reliability of everything being in the cinema server, and the ability to utilize our macros for automations within those pre-roll contexts.

There is a way around the lack of animation/transitions of course... and that is to author the slide deck in powerpoint or the like, and capture it's playout as video, but that seems like adding more work when the slide content changes, not less.

[Ryan Gallagher]

In a perfect world one would just pre-build all the slide decks, even if they are film specific. But that was not a reality because the slides pile was a work in progress all season, and there was a special batch of audience submitted "memory" text slides, that they rotated out every week or so as they curated new submissions.

That and you are never 100% sure of the film DCP aspect ratio even when the DCP is fully labeled correctly, untill you see it after you have keys.

[Marcel Birgelen]

OBS has become pretty powerful over the years for video-only workflows and I've used it for live productions. For live productions, we usually use QLab, but that seems to be overly zealous here.

In your case, I'd say I'd start scripting and maybe start looking at the API of my TMS... I'd probably ffmpeg the slide stuff from still images, put that through DCP-o-matic and automatically put out some DCPs per show.

Also, I think pre-shows should also simply stick to "flat", to avoid aspect-ratio headaches.

[Ryan Gallagher]

OBS has become pretty powerful over the years for video-only workflows and I've used it for live productions. For live productions, we usually use QLab, but that seems to be overly zealous here.

In your case, I'd say I'd start scripting and maybe start looking at the API of my TMS... I'd probably ffmpeg the slide stuff from still images, put that through DCP-o-matic and automatically put out some DCPs per show.

Also, I think pre-shows should also simply stick to "flat", to avoid aspect-ratio headaches.

I didn't even think about QLab, we do use it, but there aren't enough macs for the booth to have one just yet.

The ffmpeg step really isn't needed just for slides, since DCPomatic lets you define the per content timing. No reason to make videos first unless you were after fade ins/outs per slide perhaps. Videos for each slide (or the batch) would definitely take DCPomatic longer to process than stills. But when things are more well defined scripting might be a way to go if you are meaning script the whole process from slides all the way to DCP. Our house slides are delivered via a MS share-point resource, but I think there is a way to sync that to a local folder that would be friendlier to scripting, or perhaps a command line way to fetch them from share-point each time.

Is there a way to "push" DCPs to Doremi DCP2K4 via ftp? Or only pull them from an extrenal FTP (we've done that version). A fileserver with FTP capabilities would also enable some things happening more automagically.

Right now they mostly send us 178 slides (which I often stretch to flat during DCP creation), occasionally bumpers will be some other aspect. We use DCPomatic to stick them in the feature aspect (or all aspects if it is content common to the season), but could just as easily use the scaling on the Christie if it was easier to streamline that way. Our ILS lacks a zoom motor, hence why I think we've tended to favor doing the aspect matching work in DCPomatic. That and we only have a top masking motor, sides are manual. So all pre-show is usually custom built to fit in the feature masking.

[Leo Enticknap]

Also, I think pre-shows should also simply stick to "flat", to avoid aspect-ratio headaches.

This seems to be a European vs. American difference. When I worked in cinemas in the UK, all ads and trailers were flat (except for the very occasional one that came spliced onto the start of a scope feature, usually on a print that had been imported from the US and then ultrasonically cleaned, rewashed and polished, which tended to happen if the British censors didn't make any cuts), period. If the feature was scope, a lens change happened at the end of the last trailer. Here, the system seems to be that the entire show is either or, hence both flat and scope versions of trailers being made, even if the scope version is flat pillarboxed, or the flat version scope letterboxed.

Is there a way to "push" DCPs to Doremi DCP2K4 via ftp? Or only pull them from an extrenal FTP (we've done that version)

Yes in both directions. To pull, simply set up your FTP server in the "content feed manager" of the DCP-2K4's web UI, after which it will see it as an ingest source, just like USB or CRU. To push, login to the server using the "ingest" username (forum rules prevent me from posting the password, but Dolby tech support will be happy to give it to you), and upload DCP files into the folder you land in. When the necessary files to play a CPL are all there, the server will process them, and show the CPL as available for use in Cinelister (after refreshing).

[Ryan Gallagher]

Yes in both directions. To pull, simply set up your FTP server in the "content feed manager" of the DCP-2K4's web UI, after which it will see it as an ingest source, just like USB or CRU. To push, login to the server using the "ingest" username (forum rules prevent me from posting the password, but Dolby tech support will be happy to give it to you), and upload DCP files into the folder you land in. When the necessary files to play a CPL are all there, the server will process them, and show the CPL as available for use in Cinelister (after refreshing).

Not being a first run cinema we thankfully have no trailer headaches. Keep a couple handy for CCAP/Hi/Vi testing.

thanks for the ingest tip. Do you mean login with an ftp client? Or ssh and use something like scp or rsync to transfer in files to that folder? Will definitely have to try that.

[Harry Smolin]

We get all of our pre-show content via Screenvision and those come as scope DCPs. If we're running a flat movie we switch after that to flat masking from the studio trailers onward.

[Ryan Gallagher]

We get all of our pre-show content via Screenvision and those come as scope DCPs. If we're running a flat movie we switch after that to flat masking from the studio trailers onward.

Yeah that is why I kinda directed this topic at the independent/classic/non-first-run folks. I know screenvision subscribers get DCPs etc. I'm talking about in-house generated content. Automated masking would be nice... i'd go all flat pre-roll if I had that.

[Carsten Kurz]

We use Powerpoint for our preshow advertising loop. It's a pain in the ass as far as visual quality goes, but we have grown familiar with it, and many people are able to create content with it. I guess Keynote would work just as well, but we use a Windows machine.

[Ryan Gallagher]

We use Powerpoint for our preshow advertising loop. It's a pain in the ass as far as visual quality goes, but we have grown familiar with it, and many people are able to create content with it. I guess Keynote would work just as well, but we use a Windows machine.

Cool, that is the alternative setup I've been considering. When you have video bumpers as part of your pre-roll do you rely on embedded video in powerpoint too? Or do you stick to slides and transition to DCP for anything with video/audio?

[Leo Enticknap]

thanks for the ingest tip. Do you mean login with an ftp client?

Yes - plain FTP.

[Marcel Birgelen]

I know that we built something for a narrowcasting solution years and years ago that was able to capture PowerPoint presentations including transitions and save them as a video, in order to automate the PowerPoint-to-video conversion. That one was built in C, some virtual display driver and a combination of VBA to step through the PowerPoint... We did about the same with Flash animations (you remember those?). Stuff may have become simpler nowadays, Flash is dead and PowerPoint seems to have a PowerShell interface.

So, I found a five-year-old thread on StackExchange where they essentially are trying to do the same with PowerShell.

This seems to be a European vs. American difference. When I worked in cinemas in the UK, all ads and trailers were flat (except for the very occasional one that came spliced onto the start of a scope feature, usually on a print that had been imported from the US and then ultrasonically cleaned, rewashed and polished, which tended to happen if the British censors didn't make any cuts), period. If the feature was scope, a lens change happened at the end of the last trailer. Here, the system seems to be that the entire show is either or, hence both flat and scope versions of trailers being made, even if the scope version is flat pillarboxed, or the flat version scope letterboxed.

Your mileage may vary a bit, I know one place that actually keeps zooming around between flat and scope trailers, often with the transitions still on screen... very professional...

But yes, most places that do it "proper" tend to run the pre-show stuff in flat, then switch to scope for the feature part of the presentation. In Germany, where curtains are still quite a thing, it's not uncommon to see a curtain call between pre-show and feature.

[Ryan Gallagher]

I know that we built something for a narrowcasting solution years and years ago that was able to capture PowerPoint presentations including transitions and save them as a video, in order to automate the PowerPoint-to-video conversion. That one was built in C, some virtual display driver and a combination of VBA to step through the PowerPoint... We did about the same with Flash animations (you remember those?). Stuff may have become simpler nowadays, Flash is dead and PowerPoint seems to have a PowerShell interface.

So, I found a five-year-old thread on StackExchange where they essentially are trying to do the same with PowerShell.

[snip]

But yes, most places that do it "proper" tend to run the pre-show stuff in flat, then switch to scope for the feature part of the presentation. In Germany, where curtains are still quite a thing, it's not uncommon to see a curtain call between pre-show and feature.

Interesting, looks like they got it working in 2018. Getting a video file from pptx is easy if you are willing to manually do it in real time via screen capture, or rely on pptx's built in export to video features.

The automation side is the "challenging" part of course. The two approaches would be what they were trying, or some desktop level scripting/macros to do it via more traditional software clicks, but that kinda relies on a dedicated logged-in instance to manage all that. Powershell approach is more appealing, although having not used Powershell much, maybe it requires an desktop user session too.

If "signage" was all I was after (which would be 80% of it), pretty easy to drive that stuff with one of those purpose built signage mini machines.. and just have it be the ALT input when in film mode. Would cover slides if I could get it to do nightly sync's to a target folder on the network. I think some of them can even play powerpoint files directly with transitions and animations. That idea starts to fall apart when a show demands additional sources... such as a blu-ray and a stage laptop for a pre-film talk. Nothing a video switcher couldn't solve though.

A curtain call for a masking change would be a sweet gig! Even if they were willing to bring our fly-person in, our side masking chains contact the ground, so you have to walk out there and manually set a plumb line between formats. Motors wouldn't even help, unless we shortened our side masking. The dragging is a hidden feature when you need to fudge away from plumb on occasion, can 'stick' about a foot off plumb either direction. But usually only needed if we did not have enough time to adjust a screen file in festival mode, or poorly filed plates.

[Jesse Crooks]

We use Google Slides embedded in a webpage. Our slides are 16x9 regardless of the feature.

[Ryan Gallagher]

We use Google Slides embedded in a webpage. Our slides are 16x9 regardless of the feature.

Generally ours static slides are provided as 16x9 png or jpg (because they are authored for the digital signage machines that drive the lobby monitors as well).

But I have managed to train marketing with regards to onscreen special backgrounds such as the series logo, Q&A backgrounds, or intermission cards to match the feature aspect or just provide a 2048x1080 1.90 aspect slide with all the important content within a safe area that works for flat, scope, and F133. At least those moments are full screen no matter what masking and preset we are in. For those special slides they are even honoring a 5% title safe in addition, because of our keystone. And for the same reason I convinced them to avoid boxier border decorations.