[Mike Most — account bouncing, bad address]

There is no such thing as 720p HDCam. HDCam is only 1080, i or p. The only Sony recorder that supports 720 is HDCam SR, a completely different format, and no current HDCam camera model supports 720.

I think asking for such a thing, given that the delivery is for the Internet, is tremendous overkill. If you have the ability to do so, consider recording standard definition either uncompressed into a computer, or compressed into a low-loss recorder like a Wafian. Proper lighting and distance of the talent from the green screen will be much more of a factor than whether you’re recording in SD or HD for your application, especially if compression is not an issue.

[Mike Most — account bouncing, bad address]

There is no current HD videotape recorder that is uncompressed. The only uncompressed HD videotape recorder ever made was the D6, which never sold well (none that I know of in the US) and is, to my knowledge, no longer manufactured. All current HD tape formats – including HDCam, HDCam SR, and D5 – are compressed. The “least compressed,” and also the one blessed with the best generational performance as well as the most superior error correction, is HDCam SR. It is also the only format that is accepted as a universal delivery master by all current US television networks, regardless of format (it supports 1080i. 1080p/24, and 720p/60). It has to a great degree supplanted D5 as the preferred mastering format as well.

[Mike Most — account bouncing, bad address]

Your first problem is that you don’t want to “pull down” the sound. If anything, you want to pull it up. You shot video, not film. In film, the running speed is slowed down during telecine. With video, it isn’t changed at all. The only change it’s going through is what you’re doing to it by using a 24p project, which is running it faster than it was shot – i.e., you’re speeding it up. The easiest solution is to go to a software version that has a 23.98 project type. In this case, you want to bring in the sound as is. It should sync just fine, because you’re not changing anything from the “real time” speed it was shot at. The key to staying in sync from any double system elements is that they both need to be running at the same speed. In a 23.98 project, everything would be running at “real time,” exactly as it was recorded. Nothing has been slowed down and nothing has been sped up.

You’re not the first person to a) believe that there is such a thing as a 24p standard definition video format (there isn’t), and b) believe that because something is recorded at 23.98, it should be treated like a film transfer, even if it isn’t one. The only time “pulldown” becomes an issue with sound sync is when the picture is shot on film at 24fps, because telecine slows the running speed of the picture to 23.98 during the transfer. Therefore, the sound must be slowed down the same amount in order to stay in sync. With video, there is no telecine, no “pulldown,” and thus no speed change.

[Mike Most — account bouncing, bad address]

>Except for Douglas Trumbull.
>He’s the guy who did the effects for 2001: A Space Odyssey, and who got ILM up & running for Star Wars.

Although Doug is responsible for untold contributions to the visual effects industry, and is considered one of its great visionaries, he had nothing to do with ILM or Star Wars. In fact, he turned George down when it was offered. The people most reponsible for the early ILM work were John Dykstra and Richard Edlund, with Dykstra being the more “hands on techy” of the two. You’re probably confusing Trumbull with Dykstra – something many people who aren’t directly involved in the effects industry do, because along with Dennis Muren, they’re some of the most accomplished people the industry has.

Doug was, of course, responsible for “2001.”

[Mike Most — account bouncing, bad address]

Actually, it’s very common, but with one major exception: a new HD submaster is created that consists of only printed takes, and the DVCam reflects that submaster, not the original camera tape. All daily duplication can be done from this HD submaster, and it also is the source for final assembly. So the only difference in the workflow being described here is that in this version, there is no HD submaster, and thus the DVCam time codes don’t relate to anything other than the DVCam – a critical flaw in the workflow, and one that has no easy workaround other than changing all the time codes once the material is digitized, if you’re cutting on Final Cut. If you’re on an Avid, this is much less of an issue because you can enter the original HDCam time code into the TC24 column and directly generate an EDL for assembly purposes by selecting this column as the source time code in EDL Manager. Since most high end television is cut on Avid systems, this is probably the reason the Canadian post house offers to do it this way. I guess in the end this is, unfortunately, another lesson in just how lacking Final Cut is in terms of media and database management. By the way, the system of creating an HD submaster and a DVCam downconversion is used for just about every single camera HD originated television program currently on the air, with the exception of a few that use a system originated by Sony that involves Lockit boxes and digital sound recorders to “auto sync” in post.

The various suggestions here of using DVCPro HD are valid only if the digitizing is done by the post house directly from the camera tapes, with the sound sync done by the editor – not a bad workflow, by the way. Going to DVCPRO HD tape is not a valid option because there is no such thing as a 720/24p DVCProHD video format. If you send 24 frame material to the machine, it will record as 720/60p, but there will be no “frame flags” to identify the cadence – these are only present when you are shooting with a Panasonic camera. With no way to recognize the cadence, there will be no way to remove it, thus going to DVCProHD tape is not an option in this type of situation.

[Mike Most — account bouncing, bad address]

No. The native aspect ratio of 3 perf S35 is approx. 1.77:1, so if you scanned using a 2048 horizontal pixel count, the height of the frame would be 1157. If you were doing a 2.39:1 scan, the numbers would be exactly the same as 4 perf, assuming the horizontal pixel count was held at 2048 – as would the actual image area on the negative.

[Mike Most — account bouncing, bad address]

The numbers you gave would indicate a 2.39:1 aspect ratio, which is sensible for anamorphic material. However, anamorphic material would not normally be unsqueezed during the scan. The frame would be scanned “full frame” at 2048×1556 and the unsqueeze process done in the recorder when the film recording is made. The only other way you would get what you’re saying is if the picture was shot “flat,” usually in Super 35, for anamorphic release – although in this case, a full frame scan is usually done (again, at 2048×1556) to allow for repositioning during the DI session.

It is, however, possible that in order to save the amount of storage needed, that a flat scan was done specifically for the intended aspect ratio, in which case what you were told would be correct.

[Mike Most — account bouncing, bad address]

None of that should make any difference. If they really need this to be “hard 60” they can do that during the online. Personally, I’d be curious to know why they feel they need it.

If they’re really onlining in 720p, my guess is that this means 720p/60 (because, once again, that’s the only existing video format, there is no such thing as a 720/24p video format), and if that’s the case, you might want to talk to them prior to offlining, and find out if they want a 60 frame EDL or a 24 frame EDL. Asking questions here might be interesting, but only if you ask them first.

[Mike Most — account bouncing, bad address]

>>Not possible on Nitris, it does not support 720p.

That’s not quite correct. It doesn’t support 24 frame variations of 720p, primarily because there is no 720p/24 video format. It does support 720p/60 (59.94), because that is a standard video format.

>>no it is not, the timebase is not slower nor odes it have less frames, it is a counting system >>that allows for realtime downconverts whereas 24.0 needs to be delayed to playback on SD >>equipment.

Also not quite correct. It DOES run slower, but it does not have fewer frames. It simply makes a second a bit slower than a real time second. The “counting system” you’re thinking of would be drop frame time code, which makes up for that real time discrepancy by skipping some numbers in the counting sequence (frames 0 and 1 every minute except for multiples of 10 minutes). On some tape formats – HDCam SR, for instance – “hard 24” formats can be played back as “video speed” formats directly, with no delay. In fact, on a properly equipped SR deck, you can even play back 23.98 as 25, or vice versa, with no delay.

[Mike Most — account bouncing, bad address]

You haven’t explained a few things, in particular, what your online format/deliverable format is. Assuming that it’s 1080/24p, which is the most common HD deliverable format, you really would be much better off cutting in a 24 frame time line. You also haven’t mentioned what kind of online system you’re assembling on, which also impacts what you deliver. In most cases, a “proper” facility is likely to cross convert your 720 material directly to 1080p on playout using the deck as the cross-converter, so if your offline is in a 24 frame environment, this becomes a bit more straightforward, as all of the timecodes will directly match.

Of course, my first question is why you went the “hard 24” route in the first place, when life could have been much simpler if you went with 23.98 as your base video frame rate – since all the Final Cut presets are predicated on that, as is almost all post sound work. There is absolutely no difference in the number of frames, just in how fast they play. Having a one to one correspondence with film frames is not affected.