Mark Spano
Forum Replies Created
-
Yep – happens every day on every station. Most of the time stations have a must-air for a commercial, and since specs are thrown around loosely or not at all, you air what you are given. Not many broadcast stations can afford to properly up/down/cross-convert ads to match, so it just goes. That pillar/letter-boxing is best described colloquially as “litterboxed”, which I think might have to be attributed to Ross Daly since that’s where I first heard it.
-
What is your signal path to monitoring? Assuming that it goes Decklink SDI to DigiBeta, DigiBeta SDI to LCD monitor, then you may be seeing the result of your signal being interlaced on the way to tape, then deinterlaced by the LCD monitor. DigiBeta is always interlaced. To get video onto DigiBeta tape, it has to go down as 720×576 / 25i. So the Decklink is doing you a favor by taking your progressive signal and splitting the frames into fields. That gets recorded to tape – on playback, the only true way to see the exact quality of this is to view on a CRT, as all LCD and plasma screens are progressive displays and automatically do a deinterlace process before displaying.
-
[Ken Barnes] “It’s strange but whenever I drop a H264 file straight from the camera onto a new FCP 7 sequence timeline and it says ‘match sequence to clip?’ and I click ‘OK’, the timeline immediately goes to HDTV 1080i (interlaced!) Weird. Do you have any idea why that would be?”
The nomenclature in that dialog is wrong and has been since Final Cut gained HD capability. They’re using “HDTV 1080i” to mean 1920×1080 resolution. It’s just a misnomer. The thing you should look at is Field Dominance. That is the thing that determines whether it’s interlaced (Upper/Lower) or progressive (None).
[Ken Barnes] “Finally, what should the final sequence/timeline be set at for clips converted to ProRes422 (non-interlaced)? Do you have to type in a Custom Frame Size? There doesn’t seem to be a 1080p option on the pull down menu in sequence settings.”
Yes – set as above. “HDTV 1080i” for frame size, “Square” pixel aspect, and “None” for field dominance.
-
Good thing the festival came to their senses…
Now about this other stuff:
[Ken Barnes] “Should I have chosen something else?”
Well, yes. But it’s more than that. If the stuff was shot at 30.0, then this is what I would have done. Batch process in Streamclip to keep frame rate the same and resolution the same, uncheck interlacing (since 5D footage is always progressive), make ProRes. Then, I would have (with Cinema Tools) conformed all of the source footage to 29.97.
The reason why I would conform is that if you bring 30.0 footage into a 29.97 timeline in FCP, there is a constant retiming that will blend a frame or two here or there, and drop a frame here or there, to get it all to work at 29.97. Conforming eliminates this from happening, as it just tells the file to play at 29.97 speed instead of 30.0, and FCP reads it in as if it were 29.97 all along. Audio gets sample-rate converted, but since it’s 5D camera audio, I usually wind up only using that for reference and throwing it away in favor of second-system sound.
[Ken Barnes] “Have I affected the quality of my film by doing this? Would it be worth it to re-import everything de-interlaced (or more properly non-interlaced)?”
This may or may not be noticeable in your cut, probably depending on the length of footage between edits. It’ll look like a slight smear when frames get blended, and I assume you already know what a dropped frame looks like. If the frame rate thing wasn’t an issue, and all you were dealing with was having ‘interlaced’ your previously progressive footage, then there’s only one place where the quality would need judging, and that is: are there any titles/effects used in the timeline? The footage itself does not necessarily suffer when interlaced, because basically you’re taking a whole frame and splitting it up into two fields – no information is lost. However when you do fades, transitions, other effects, those are all now created/rendered in an interlaced timeline, so the motion will be finer (60 field motion over the 30 frame motion of your source footage). That stuff may just look like it’s slightly out of place compared to watching sections where it’s just source footage. These are subjective judgement calls – you have to look at it (on a video output to an actual video monitor) to make a proper assessment.
In the future, knowing how the footage is made in camera goes a long way to figuring out how to handle it in edit. The ultimate determining factor though is you. If you’re OK with the way it looks, then any attempts to ‘fix’ it would have to be weighed against how much better it could be. If the difference in your eyes is negligible, then it’s not really worth trying to go back.
-
[Ken Barnes] “A film festival says they will screen in 4:3 but will accept blu ray”
This film festival is crazy. I don’t know anyone who would ever ask for a 4×3 letterbox Blu-Ray. It makes no sense, considering that you’d be better off giving them a DVD if that’s really the case.
If this is true…
Well first of all, what is your source footage? You say it was shot with 5D – why are you editing in an interlaced sequence? 5D footage should all be progressive. Can you tell what the 5D stuff was originally? It might only be progressive – worse, it might be a different frame rate altogether.
These are some crucial questions. If you understand why I’m asking them, and you’ve already considered that for other reasons, then read on, and we’ll get into your compression/encoding questions.
[Ken Barnes] “I’ve read I should de-interlace at some stage, correct?”
Only if your output requires progressive. The answer to my first question will influence this answer obviously, but the rule is that if you shoot interlaced, keep it interlaced. If you shoot progressive, keep it progressive. Only break these rules if you are required (TV broadcast delivery likely would require interlaced delivery, web delivery should be progressive, etc.).
[Ken Barnes] “Should I convert my FCP sequence before exporting? (including changing the Pixel Aspect and Field Dominance) Or do that elsewhere? Or doesn’t matter?”
The rule for FCP is, whatever happens in FCP stays in FCP. In other words, don’t change resolution/frame rate/aspect/etc. in FCP unless you absolutely have to. FCP is the worst product in the suite to handle resizing and retiming. Compressor here is your friend.
[Ken Barnes] “What’s the Cow skinny on using some of the filters in Compressor (if that’s the weapon of choice)? I’ve read on KenStone.com that it’s important in many cases to fiddle with black levels and sharpening, etc. but that was published a few years back. Somehow I can’t see applying a blanket filter to a whole film when it might only be useful for a few shots or scenes.”
You’re absolutely right. Don’t mess with any of those filters. The only ones that are good for anything really are the ones that overlay (time code burn, watermark). Anything that changes picture quality or audio quality is much better done in an application dedicated to that (Color, Soundtrack Pro, etc.).
[Ken Barnes] “What’s the best way to do it for max quality and what settings are required”
OK. You want to make an SD 4×3 letterbox and burn to Blu-Ray. Hmm. OK. Well maybe you don’t want to do that, but they want you to. OK.
Export Quicktime Movie of your sequence using Current Settings from FCP. Bring that into Compressor. Now you’re going to do the resizing in one step, and then the encoding/burn in a separate step. I’ll give you a couple of presets – put these in your home/Library/Application Support/Compressor folder and you’ll see them. You can also see how I’ve set it up so that you understand what it’s doing. Start with the ProRes Intermediate preset. This will do the resizing and flip fields for SD letterbox. Once that file comes out, you can check it in QuickTime Player and see that it made a proper SD letterbox ProRes file, fields intact. Then you’ll take that file into Compressor and do the encoding step, with the H.264 for Blu-Ray SD and Stereo AC3 presets. You can add a Job Action to create a Blu-Ray image as well, or do the authoring somewhere else.
This all sounds more confusing than it is, only because they’re asking for something way out of the ordinary. I would definitely find someone to call to ask how the projection will happen – there is a strong chance they (being a busy theater in Manhattan) know how to project full-res 16×9 from a Blu-Ray…
-
I find it funny that they give you a complete contradiction in the delivery requirements by saying “Deliver in High Definition” and then listing specs that are standard definition. I would try to check with someone to see if that’s really what they mean.
If they do actually want DVCPro50, then make sure about the field order. Something tells me it might be upper field first. Everything else you’ve got there sounds good.
-
I work in commercials. Commercial advertisers pay for exact amounts of time. That time is also dictated by the networks they advertise with. If a commercial is to be delivered as a thirty second commercial, it has to be 30.0 flat, otherwise it will at best be cut short and at worst be rejected. I would be laughed out of a job if I didn’t know how to do this. Do a frame count, if you want to be clearer about it. 30 seconds, at 30 frames per second, is 900 frames. You can see this easily by setting those in and out points on your sequence (1:00:00:00 in, 1:00:29:29 out), switch to frame numbers, and note the duration. That’s the last frame in a thirty second commercial.
There is only one right answer on this – it’s not subjective. Anyone who tells you differently is misinformed.
-
Usually converters like this will convert TC as well, but on VITC only. Set your record deck to look at VITC and make sure player deck is not blanking VITC lines and the converter should do the job. I’m pretty sure the Teranex Mini does this, so I don’t see why yours wouldn’t either.
-
You have to take into account that setting an out point in FCP means “include this frame”. So if I’m setting in and out in a timeline for a clip that I want to be exactly 30 seconds, and I’m in 29.97fps, starting at 1:00:00:00, then the out point should be set to 1:00:29:29. If you included the frame at 1:00:30:00, then the spot would be 30 seconds and 1 frame long.
-
What are you exporting to? If you’re trying to make a Quicktime from your sequence with 6 tracks of audio, you should check out Shane’s tutorial as it explains it all:
https://library.creativecow.net/ross_shane/multi-audio/1