[Andrew Rendell]

Sorry about the tone, just finished a fast turnaround (cut a 47min doc in 11 days). If I wasn’t tired I’d have said basically the same thing but in a more gentle and encouraging manner 😉

BTW, this one:
https://www.channel4.com/programmes/terror-at-sea-the-sinking-of-the-concordia

It’s on 4OD for the next 4 weeks if anyone’s interested. I think it stands up as pretty good even without considering the speed at which we cut it! (But then I’m biassed, ok.)

[Andrew Rendell]

You can’t ask a book a question, but you can read the words!

I think that if you don’t know that in an interlaced scan records a field 1, with it’s lines numbered 1, 2, 3, 4, 5, etc, AND THEN records a field 2, with that’s lines numbered 563, 564, 565, 566, etc, (that’s for HD, the field 2 line numbers for PAL and NTSC are different ones), you really ought to learn some technical information ASAP.

[TBH, I’ve always thought that describing what everyone knew as field 1 and field 2 as “odd & even”, “top & bottom” or anything else were simply examples of deliberate obfuscation designed to add a layer of unnecessary jargon that only confuses people.]

Start with this
https://www.itu.int/dms_pubrec/itu-r/rec/bt/R-REC-BT.709-5-200204-I!!PDF-E.pdf

and every time you see something that you don’t understand, ask someone or get on the internet and find out.

[BTW you can have a picture that’s taken all at the same time in an interlaced signal, i.e., both fields actually recorded by the camera at the same moment rather than a 50th of a second apart, but that’s called Segmented rather than Interlaced to avoid confusion.]

[Andrew Rendell]

If you mean the click at the start & end of the clip as it passes the “play head”, that’s because frame based editing doesn’t put the cut at a “zero” crossing point, so at the cut there’s a high likelihood that there’ll be a voltage jump, which translates aurally into a click. It won’t happen every time, YMMV but it probably will something like half or three-quarters of the time.

The way to deal with it is either tweak the cut point in an audio suite (where you have sample accurate cutting) or add a fast dissolve to the cut (either keyframed or simply add a cross fade effect with a duration of 1 frame).

[Andrew Rendell]

Not sure what you mean by “frame by frame”. I’d choose a free video track to put the text on, then edit in a series of text clips (mark the in and out points on the sequence and overwrite). What you need is a transcript of the film as a text file, then you make up the first one, choosing the font, size, etc, then you copy/paste each bit of text into the same format for each clip.

I’ve done that loads of times for short sections of interview (e.g., foreign language translation) but I wouldn’t volunteer to do a long film like that if I didn’t have to! On the positive side, you’ll become pretty good at spotting the start and ends of words when audio scrubbing if you do it that way (and that’s a useful thing to be able to do).

Good luck.

[Andrew Rendell]

I haven’t looked into bundling a player, but what I do routinely is to sent a Quicktime in h.264 for the video and AAC for the audio. That plays back fine for most people, and works well for the majority of my clients (delivering files that will be re-encoded for the web or sending viewing versions of broadcast programmes for approval or compliance).

For corporate clients I’ll tend to give them a Quicktime (codecs as above) AND a Windows Media file as well (letterboxed in a 4:3 frame). Takes a little longer to do the extra encode, but saves having people phone up saying it doesn’t work on an old PC that they don’t have administrator permissions to add a program to and they have to show it to someone now.

BTW if you do a 2-pass encode the quality is a lot better than a single pass one for the same data rate.

[Andrew Rendell]

I’m in the UK – PAL world. Reference level here is ALWAYS -18dbfs, nothing else is ever used in broadcast.

[Andrew Rendell]

If we are talking UK, then Ofcom are specifying quality assessment for DVB-T as defined by the ITU for digital broadcasting (e.g., the 5 point assessment scale and various technical things). So to do fast turnaround you really need to have scopes/PPM/full quality monitoring in the edit. Is FCPX sorted for those things yet?

[Andrew Rendell]

I’m not sure, but something does come to mind; dithering.

Dithering is the deliberate addition of a very tightly defined (and very low, basically 1 bit p-p) amount of a specific type of noise to the signal. That greatly improves quantisation errors at the expense of raising the noise floor, which has the subjective effect of reducing what you hear as noise (which sounds counter intuitive, but it’s a well established technique and I’ve heard some tests that have been quite impressive).

I couldn’t be sure, but that’s my guess. TBH -79dB is pretty good and it’s much more likely that if you have noise problems they’ll be from background sound, mismatched equipment or electronic interference than the recording.

[Andrew Rendell]

I generally use one that looks like this
https://www.planetmov.com/Stock_Footage_Elements/Vt_Clock.html

(or when I’m at a facility I’ll use whatever they have on the system, which is usually a variation on the above)

[Andrew Rendell]

I stand corrected