Frame Accuracy

[Adam Schmidt]

I have had this problem across the board with almost every deck that I have used. First off, some decks are better than others but it has been hit and miss. Mostly I find that using either LTC or VLTC (during strip and ETT) but not both works much better. Also digging around in the deck’s menus and turing on Duplicate timecode from LTC/VLTC to VLTC/LTC works better than the mac sending both (got your manuel?). In fact, i have seen beta decks read off one frame on an external reader, but be on frame on FCP and the deck display. Try switching between LTC and VLTC in the deck control menu. Also make sure that you have run your BBG properly (That is: deck ref in —> deck ref out —> IO ref —> 75 Ohm Terminator [all over 75ohm bnc]) , leave deck termination off, and let the monitor free run w/o a bbg feed. If your on house black (multiple BBG outs), run only one ref signal output loop in the order described, and don’t independently feed the deck and io.

Adam

[Nick Price]

Adam
have you found that these methods have fixed this for good? As with Bill, my playouts are accurate sometimes, but not always. Recently changing to LTC & VITC in the device settings in FCP has proven more accurate, but i spend far too much time worrying if each program will be accurate during the playout. not good for the heart pressure.

Also on another note, i very often get what i can only describe as render errors, during edit to tape. and sometimes during just normal playback. They consist of a tiny line, that appears to have been displaced from elsewhere, flashing up often only on one field, so when looking at the frame still you cant see it, but when spooling thourgh on a digibeta it appears on the second frame. I go back and rerender that small section and it disapears. But another might appear during a repeated playout!

any thoughts?

nick

[Bill Portune]

Thanks for your input Adam.
This has been a major source of frustration for me as my Avid in the same setup was Rock-Solid with accuracy for 7 years!
I’m going to finally set aside some time for testing, and I’ll try to post my results.

[Duncan Craig]

Well nice to know I’m not on my own.

Been having these problems laying back to SP and Digi since the start.
I had one of the first IO’s in the UK. And there’s always little quirks.

Luckily for me I only layback commercials to tape, so 1’40” is the most time I ever waste.

I’m not running the latest drivers, or OS. The last time I upgraded was about 9 months ago.
I worry to much about creating even more problems.

[Trevor Asquerthian]

As it was explained to me QT has an inbuilt inaccuracy in that it does not start at an exact offset from when it receives the play command, so your video card will put the movie at a frame edge, but the frame edge may be +/- a frame from where you really want it to be.

It was also explained that QT have an API to help rectify this, but most card manufacturers have yet to implement it.

I am interested in any cards that have had this implemented (AVSync I believe it is) because I’d like to test them for a different application.

Of course this may all be hogwash as the boards should be full of FCP editors complaining of ETT being off by a frame – but then maybe they don’t check! I guess the machine control would be right, so if your deck is going into record at 10.00.00.00 you are only going to see a problem if the playback is delayed, not if it is advanced. And then you’d only see it if you checked!

Trevor

[Trevor Asquerthian]

[Bill Portune] “Q
My shows are long-form (55 min) so can’t really stop and check for accuracy.
Does anyone have a tip for more accurate edit-to-tape? “

I can give you a method for checking, before you have laid down 65 minutes…

make a note of the timecode your first cut is at (that is the first cut AFTER the vtr is in record)

take the burn in output of your vtr and stick it in a vhs (or better) machine…

set it recording then do your ETT…

whilst ETT is running roll back the vhs and see what timecode your first cut is at….

repeat as necessary

HTH

If anyone wants to test the accuracy of their ETT then do the above test – only leave the VHS running as you do multiple ETTs (i.e. restart the ETT after the cut you are using as the sync guide) – do report back… 20/20 would indicate a locked down system.

[Bill Portune]

Great tip (should have thought of that myself).
I’ve also found that lengthening my preroll time to 5:00 instead of 3:00 has improved accuracy/consistency.

Thanks again!

[Trevor Asquerthian]

[Bill Portune] “I’ve also found that lengthening my preroll time to 5:00 instead of 3:00 has improved accuracy/consistency.
“

any chance of a consistency score out of 20 attempts?

[Bill Portune]

5 sec preroll 20/20 good attempts

3 sec preroll 17/20 good attempts – attempts 8, 16, 19 the first cut after in-point occurred 1 frame early.

This was with Mastering with no bars, slate, or black prior to first frame of program. On the DVW-2000 I set the TC to int/preset and black the head using FCP. Accuracy on the machine has been flawless with both assemble and insert edits. It’s just the FCP material that’s been 1 frame early at times.