PDA

View Full Version : Is Batch Capture part of an outdated workflow?


billmi
06-07-2007, 01:56 PM
OK, don't take my head off here - I'm not saying it *is* outdated, just pondering how much I've liked the idea of batch capture in a work flow because it is what I am used to, vs. it really being an advantage anymore.

In light of this thread:
http://www.newtek.com/forums/showthread.php?t=69349
and of previous discussion about it not being included in Speed Edit (yet?) it got me thinking and asking myself questions as I sit here during an Autochop capture. I was really surprised to not see batch capture in SE 1.0, especially since it's in most NLEs at the same price point. But now that I think more about it, is it really as important as I've been thinking it was?

In what situations would I want to batch capture anymore, given faster alternatives?

1 – Capturing initial footage, that was shot in separate shots, to individual clips .

Arguably that is the number one reason I've used batch capture in the past – and Autochop takes care of it. For a long-form webcast show that I do I'll often be shooting paintball games that I don't want split up into separate shots when captured, but want the games separated, and the interviews split. I resisted Autochope her, until I realized it was faster to capture with Autochop, then grab all the clips that make up one game and throw them into a subproject, dealing with them from then out as a single clip on the timeline. This newer approach turns out to be substantially time saving vs. shuttling through the tape to find my in and out points for batch capture, and then batch capturing.

2 – Capturing footage that was shot in a single continuous shot, but which needs to be broken into shots by scene for editing.

Previously I would have thought this a great argument for batch capture, but the more that I think about it, that's the slow way to do it.

Capturing the tape as a single file, then scrubbing through it in SpeedEdit, and chopping it with the C Key, will create instances that refer to the clip. They can then be dragged en-masse to a folder in the filebin, and edited with just as if they had been created as separate clips by batch capture.

There are two differences between this approach and batch capture. First, physically on the drive there is only one actual clip that is being referenced, but since I'm doing the editing in Speed Edit, that doesn't matter – my instances handle in the Filebin timeline and story board exactly the same as if they were referring to separate files that were created by batch capture.

The second difference is speed. Instead of spending my time shuttling the tape back and forth like I did in the 1980s to find my in and out points, I'm scrubbing through the video off the hard drive to find those points, like with a non-linear editor. It's hands down faster, for ultimately the same result.

3 – Recapturing footage later, to be able to re-edit a project – First I so rarely am going to come back to a project that I hardly ever worry about this, but there is occasionally something I know I will have to re-edit, and archiving the already captured footage to random access media just makes so much more sense. For me, a typical project has at most a couple hours of raw footage. With Raw DV footage, that's 4 Data DVDs to back it up. For bigger projects, the cost in my time to set up and do a recapture, making sure the filenames are all correct so they will be red by the project, etc. is going to be greater than the cost of an external hard-drive big enough to simply back up the files in one swoop.

4 – Getting a tape from a client, and having to edit clips based on their EDL – If I could import a capture list and autocapture from that list, then yes, batch capture makes sense here, but if I'm going to have to generate that batch cap list myself to edit to their list, it's going to be faster to capture at once, and break up instances, as in number 2. Even if I'm keying in the in and out keys with the 10-key pad rather than shuttling to them, it's still faster to cut after than batch capture, because the capture itself happens in a single pass.

5 – Limited drive space – This might be an argument on my notebook where I've only got 20 Gigs free for video – it's advantageous to capture just what I am going to use.

SBowie
06-07-2007, 02:06 PM
Numbers 3 to 5 are sufficient justification in my view.

UnCommonGrafx
06-07-2007, 02:10 PM
I think we can convince ourselves of anything.

That said, I believe that bc is still important for 3,4 and 5. For 1 and 2, that's about the workflow one would create for one's own sake.

As a freelancer, I generally want and use it for old edits and edls given me. I never let anyone give me a handwritten edl, so re-keying isn't something that is a concern. And I've sat with the edl import with VT in order to be able to import from an excel sheet for EXACTLY this reason.
I'm just happy that I have Jeff White's offering, DMC(?-not at the box to recall the name correctly), to capture DV material as the control in VT hasn't been reliable for me.

It was a thing of beauty to watch a lost edit come back to life, ca. VT3 or so...

Ivan
06-07-2007, 02:17 PM
as the control in VT hasn't been reliable for me

and here I thought it was just me... Assuming that Andrew isn't just blowing smoke about the VT5 upgrade being VT6/7ish this should be fixed in VT5 right?

Ivan

Bobt
06-07-2007, 02:23 PM
You forgot the most important one.
Producer comes in with timecoded shot list. Says take
tapes A,B,C capture em and we will work with them on the timeline.

Then make me a timecode burned version of the edit (normally at the end)
and all the source footage.
He thats the tape tells you he would like shot X replaced by shot Y over the phone. What are you say BUT I don thave it on disk anymore? DOH!.

Time code is still a large way we communicate remotely in video. Its a common basis. So no timecode/batch capture err umm. Now what?
Capture an hour of footage to disk and pretend you know which guy in the rred shirt he is talking about??

Any compression job relies on batch capture. Initial edit sessions do as well.
Least thats what happens here in Chicago. I have producer as a friend did some Industrial spots for him. Did some other work based on that job. Time code was in all of em. Some was simple DV capture. Rest was BetaSp (lots of that footage out there. )
Anyway that was a year ago.. Now I just sit and code and wonder..
Bob

Bobt
06-07-2007, 02:25 PM
OOPs you listed it number 4.. Read to fast. 4 and 3 is a big thing.
then 5

JackJ
06-07-2007, 02:52 PM
One question about the 'capture it all and then use the instances' approach. How much additional overhead is involved in working with one huge clip instead of several small ones? As far as SE and the ram is concerned, is it just a matter of handling the data-rate of the source video required, or does finding the section to play back out of the larger file somehow add significantly to the load the computer has to deal with?

Ivan
06-07-2007, 03:12 PM
Well, for starters, how long does it take your one long clip to conform? As for overhead, I would guess it's only drive space, except for conforming SE would ignore what is not showing. However that isn't the only concern. What do you do when you are in the middle of 5 big projects and you need to recapture the video for a past project? If you have one big clip do you have room? Suppose you have 3 or 4 clips on each of six one hour tapes do you want to recapture 6 hours or just the clips? Will it grab from the same starting point or do you have to tweak it after you bring it in?

Ivan

Jim_C
06-07-2007, 03:45 PM
How much additional overhead is involved in working with one huge clip instead of several small ones?

Working with SE on one of my systems, which I admit is underpowered, one 2 hour mpeg2 file on the timeline brings it to it's knees when editing. Everything is click, wait.
If I capture a couple of hours in 15-30 minute segments. I can fly right thru the editing.
DV clips are not so bad, but mpeg2 gets boggy.

So yes, it is am under powered machine, but it definitely demonstrates that it takes more to work with a large hunk instead of smaller ones.

And I use BC for items 3 and 5 mostly.

lcress
06-07-2007, 10:16 PM
Just my $.02...

I really don't have unlimited disk space. And I don't want to rangle a huge file and try to figure out where that great shot of the kid with the dog was.

I prefer to dump all my DV footage to vhs (and or DVD) with a timecode window. Review my shots (repeatedly), pick the best ones, make a preliminary EDL and digitize. By the time I am at the digitize stage, I have already edited the project in my head and have a direction to start editing at the machine. Plus, I can review the vhs without a computer or beating up my more expensive DV deck or camera.

Also with my EDL notes in hand I can easily find a pickup shot on the tape because I know where it is by TC.

It is not uncommon for me to have 6 to 10 hours of footage to go through. Having all that on the system is just too much wasted space. Also, I work on 2 to 3 projects simultaneously - the amount of disk space to hold all that is prohibitive.

I need Batch Capture. SE needs BC. However, there must be something inherently problematic with writing the code for BC because that was always a weak area in SpeedRazor. Even so, I could not have accomplished my editing projects without it.

Once I have a definitive EDL I can go back to it again 6 months later a recapture just the sections I need. This saves time and disk space.

Working with smaller chunks also has the advantage of being able to name them as you go, which increases organization in the filebin. With one huge file - where are the exterior shots, the close-ups, the POV's, etc? The system likes smaller clips and chokes under large ones.

Batch Capture is a professional element missing out of SE.
However, we are still at Ver 1.0. If we see it by 1.5 I'll be happy. A bug free BC may be a long way off, but I would rather have a workable version for the time being rather than be without it completely.

Verlon
06-07-2007, 11:41 PM
Well (as one who brought it up in another thread), what about Lightwave? Suppose I shoot a wedding with two cameras. I am mapping some video clips onto 3D objects (see the 3D Arsenal WEAP for examples). Now, instead of a 5MB clip to map, I have a 39GB DV clip AND I have to look long and hard for the scenes I want.

If I should want just a few of these, right click and save as is fine. The more I need, the more troublesome this technique becomes.

Further, suppose you have some rather bad footage (we aren't all Michael Bay). Now you have a couple of GB dedicated to the out of focus part where the drunken brides maid wants to discuss politics. Personally, I'd rather have that space back....

I guess it really depends on what you want to do with the footage and how you like to work.

The Batch capture style where you shuttle the tape around seems outdated to me, but autochopping into separate clips seems like a convenient feature.

UnCommonGrafx
06-08-2007, 12:15 AM
hehe, That's where LWConnect comes to the rescue!!

(OF course, this then leads to a bug in LW where it can't separate a clip onto separate objects: last object loaded is the setting the clip takes.)

That's a little ot, though. But LWC would be the ticket here... Even with 3DA scenes.

toasterhombre
06-11-2007, 08:22 PM
SE does do autochop with scene detect.

Well (as one who brought it up in another thread), what about Lightwave? Suppose I shoot a wedding with two cameras. I am mapping some video clips onto 3D objects (see the 3D Arsenal WEAP for examples). Now, instead of a 5MB clip to map, I have a 39GB DV clip AND I have to look long and hard for the scenes I want.

If I should want just a few of these, right click and save as is fine. The more I need, the more troublesome this technique becomes.

Further, suppose you have some rather bad footage (we aren't all Michael Bay). Now you have a couple of GB dedicated to the out of focus part where the drunken brides maid wants to discuss politics. Personally, I'd rather have that space back....

I guess it really depends on what you want to do with the footage and how you like to work.

The Batch capture style where you shuttle the tape around seems outdated to me, but autochopping into separate clips seems like a convenient feature.

Verlon
06-12-2007, 09:20 AM
But you have to keep the whole thing on your hard drive unless you render out the specific clips, right?

Now if I want to grab 7-8 clips (from 1 long single shot), no big deal.
27-28 clips, it starts looking like something I need to automate
107-108 clips and its something that NEEDS to be automated.