HD Video Assist

I'm going to shoot with a F900/3 at 24p and the producer asks me to record the video assist in order to make an offline editing later. I use to work with the Miranda mdc700, but it seems that the output is NTSC.

(I work in PAL). the other choice could be the Miranda DVC 800, which has an firewire output that could be interesting. Then I have some questions (if anybody can help!)

The thing that worries me is the Time code... specially if I had to work in NTSC. anybody knows how can I solve the problem of the TC if I mix NTSC, HD and PAL?

(wow)

If I use the dvc800... how long is the longer firewire cable without loss of signal? (ooops, maybe this question is for the video list) and if I use the dvc800... the audio and the TC is embedded in the firewire? the offline editing will be with an Avid xpress and the online with a XPRI.

Any thoughts with this?

Pol Turrents
DoP Spain, Barcelona

Tell them "No".

They are hiring YOU as the HD expert. YOU know that the best way to make dallies is my doing downconversions after the tape is shot.

Just tell them I bet that they will listen...

Good luck

Dave Satin
Video Engineer

>Tell them No...that the best way to make dallies is my doing >downconversions after the tape is shot.

MUCH AGREED! I haven't been in a position to try and push this (simultaneous recording of downconversions have not been asked for on my sets in my particularly small market so far) but I certainly think it best. There's enough crap hanging off the camera already, particularly if you're syncing multiple cameras/sound recording devices.

A local rental house who owns and could eventually make a profit from renting on-camera downconverters even agrees and would also like to see this go "back" to post.

Jay Farrington

Don't bother with on set down converts....you'll need a video geek to push record on a beta deck and the time you waste will never be made up.

Nick Hoffman 600 DP NYC

>Tell them No...that the best way to make dallies is my doing >downconversions after the tape is shot.

<snip>

>There's enough crap hanging off the camera already, particularly if >you're syncing multiple cameras/sound recording devices.

From my own experience, there is definitely a place for on-set downconversion. I agree that hanging more stuff on the camera is not desirable. However, if there is a central place to which the SDI's from the cameras return, such as a DIT rack or a switcher for the DP, it is quite feasible and relatively economical to record high quality downconversions on the set. A separate rack (in a road case if need be) can easily receive DA'ed SDI from the cameras and sound directly from the mixer. In our case, we used 4 Evertz 9155Q afterburners, 4 channels of Rane audio delay, 4 channels of RANE audio DA, and 4 Panasonic DVC-PRO50 decks to downconvert and record 4 24P cameras on the set. Picture and T/C were derived from the SDI and the sound from the mixer was delayed and recorded in sync from analog inputs.

This system, operated remotely with a BUF controller, provided a number of advantages over offline downconversion. The recorded output was available for review on the set, if necessary, in sync, without resorting to in-camera playback of the HD tapes. The NTSC tapes were immediately available for digitizing by post without waiting for off-line downconversion. The HD masters were not subjected to any unnecessary passes in tape machines, thus lessening the likelihood of machine induced damage. And high quality double system sound, in sync with picture, is immediately available to post without the need for separate digitizing or transfer.

Plus-8 put together our rack and once the initial crocks were worked out, the system turned out to be both workable and practical.

It's also certainly possible to REMOVE something from the camera using a setup like this. There's no reason not to run the video village feeds from this downconversion and not even bother with a Miranda or other onboard downconverter on the camera.

John Gilman - Zoundz Audio

>"I'm going to shoot with a F900/3 at 24p"

24 or 23.98?

>"I use to work with the Miranda mdc700, but it seems that the output is >NTSC (I work in PAL)."

There is a PAL version: MDC-700 PAL

>"the audio and the TC is embedded in the firewire?"

YES embedded and frame accurate. You can also have up to 3 TC burned in i.e. HDCAM TC, "down convert" TC with 3:2 pull down and external (audio) TC.

You don't need to "to push record on a beta deck" as you can slave a DVDeck or DV Drive to the HDCAM via the DVC-800.

Thanks,

Gilbert Besnard
Director, Product Development
Miranda Technologies Inc.

>...The HD masters were not subjected to any unnecessary passes in >tape machines, thus lessening the likelihood of machine induced >damage...[A posting suggesting that avoiding extra tape passes is one >reason to make work dubs on set.]

REPLY / COMMENT :

A recent posting suggested the above as one of the criteria favouring making work tapes on set, as opposed to making them in post.

In the last 20 years, I've watched tape being shuttled all morning, still framed while the crew goes to lunch, and shuttled all afternoon, with no ill effects. From that perspective, it seems unlikely that making an extra tape pass to make work dubs will harm anything.

I would therefore not worry about making an extra tape pass when deciding whether to make work dubs on set, or in post.

Lew Comenetz,

HD Video Engineer.

On set downconversion is typically cumbersome (more wires & attention) and problematic.

Especially in NTSC, as you need to make sure that the 2/3 pulldown is sequenced properly with the timecode (something that many of the less expensive "monitoring" downconverters do not deal with....). Sony and others have a convention as to where the A frame in the pull down hits on 30 frame timecode, and if the downconverter is not referenced to that code, you could have problems conforming the 24P masters to your 30 frame edl.

The other problem however, is you won't know of any problems (or omissions) in the original HD recording if you do not make your editing dailies from a properly sequenced downconversion of the HD tapes. Bad surprise potential in online!

In PAL you don't have to worry about the 2/3 sequence, however, you may need to deal with an offset in code from the latency in the downconverter (delay is also an issue in on set NTSC dailies).

The downconversion path in the studio vtr deals with this latency as the SD sdi output has both audio and timecode embedded, so that delay caused by the downconversion is matched by audio & timecode.

Phil Squyres

In case your recording audio in camera (as with the F900) you'll simply use the afterburner with an HDSDI feed from the camera which includes audio, timecode, and picture all in sync and the afterburner will pass the downconversion keeping everything in sync as well.

Keith Collea
Video et al
LA, CA

>In case your recording audio in camera (as with the F900) you'll simply >use the afterburner with an HDSDI feed from the camera which includes >audio, timecode, and picture all in sync and the afterburner will pass the >downconversion keeping everything in sync as well.

That's certainly true if you have Afterburners with the audio option and IF you are certain that the camera audio is not having any problems of the sort that occur in certain situations. One of the virtues of sending the sound directly from the mixer to the downconvert rack is that there is no possibility of a problem with camera sound causing a problem on the downconverted tapes. Effectively, the downconverts are also double system sound masters. Since the SDI from the cameras is dependent on many more variables, this seems like a more secure approach. There's no possibility of the level controls being accidentally changed, for instance.

Just my $.02

John Gilman - Zoundz Audio

John Gilman writes :

>The HD masters were not subjected to any unnecessary passes in tape >machines

Seems to me that running the HD originals (sic) in continuous passes for offline downconverts wouldn't be much of a risk. And it would give you the additional advantage of catching any glitches in your HD tapes that you wouldn't otherwise see until it's too late. (Granted, real-time downconverts do provide a backup original of sorts...)

Dan Drasin
Producer/DP
Marin County, CA

Phil is correct that their are many issues with downconverters, however the Miranda DVC-800 is an exception as is the Evertz Afterburner in that those two units do everything right.

Phil said "Sony and others have a convention as to where the A frame in the pull down hits on 30 frame timecode, and if the downconverter is not referenced to that code, you could have problems conforming the 24P masters to your 30 frame edl."

The DVC-800 3:2 sequence respects this " A Frame convention " and the unit also adds an "Avid compliant" White Flag in the vertical interval.

Phil said "In PAL you don't have to worry about the 2/3 sequence, however, you may need to deal with an offset in code from the latency in the downconverter (delay is also an issue in on set NTSC dailies)."

With the DVC-800 the "downconverted" TC/Audio/Video are re-timed at the output to be perfectly frame accurate.

Phil said "The downconversion path in the Sony studio vtr deals with this latency as the SD sdi output has both audio and timecode embedded, so that delay caused by the downconversion is matched by audio & timecode."

This is exactly what the DVC-800 does too. I have a .pdf of the Miranda manual if anyone needs it, or you can request one from their web sit.

Jeff Blauvelt
HD Cinema
NY/NE (Westport, CT)
LA/West (Los Angeles, CA)

>I have a .pdf of the Miranda manual if anyone needs it, or "you can >request one from their web sit.

Why don't you offer to Geoff to post in the cinematography.com website? in other hand, I just talked with my producer, and he understood that is not correct to use the video assist to make an offline editing...thank you to all of you that answered my question!

Greetings,

Pol Turrents
DoP Spain, Barcelona

>Seems to me that running the HD originals (sic) in continuous passes >for offline downconverts wouldn't be much of a risk. And it would give >you the additional advantage of catching any glitches in your HD tapes >that you wouldn't otherwise see until it's too late.

You're under the mistaken impression that the people doing the downconverting actually look at the footage while it's being run. Unless the footage is really interesting (and most raw footage from most sources is boring as hell), that rarely happens.

Downconversion is a commodity business with little profit. In most facilities and production/post companies with gear, downconversions (and Avid captures, by the way) are done by the lowest of the low on the experience scale, internswith the attention span of a ferret or people who were in shipping and receiving last week and want to move up. They're given the most basic of instructions and left, usually on the night or graveyard shift, to fend for themselves.

The idea that these beginners are also acting as QC is not realistic.

Bob Kertesz
BlueScreen LLC

Bob Kertesz writes:

>You're under the mistaken impression that the people doing the >downconverting actually look at the footage while it's being run

I doubt its the point that some lowly assistant looks at it, in my opinion the point is if these tapes are subsequently used for editing, the editor will look at it and will certainly spot any problems. If the tapes he/she uses are copies of the original it is a much safer bet than if the editor were to have his/her own "masters" which wouldn't show the problem. This is how editors have always worked. What if they like frames that haven't actually been recorded just because the video assist rolled before the camera (in continuous TC situations, of course). I agree that the less stuff hanging off the camera the better (its taken years to free movie cameras up, why take these ridiculous steps backward?) and the editors should have copies of the original, not their own masters.

Roger Simonsz
DP/Operator
(Trying to move forward) in Paris