DV captures

[stymie]

I'm capturing miniDV to my PC with WinDV and Scenalyzer. Any recommendations on whether to use the automatic scene detection which saves the clips based on time stamp or capture the video to a single file? Also, I see an option in Scenalyzer to turn off scene detection. Is there an option to turn off scene detection in WinDV?

[DB83]

In WinDV's settings set Discontinuity Threshold to zero to turn off automatic scene detection.

[stymie]

Thanks. I started capturing with the scene detection turned on. But, I'm thinking for ease of workflow, editing, converting, archival, etc, that scene detection should be turned off. Any opinions?

[dellsam34]

I agree, Breaking the footage a part is much easier than stitching pieces of it together. So capture the entire tape and choose your scenes later.

[Alwyn]

For ease of editing, I'd go for scene detection. You'll end up with a lot of DV files but they will be dated so they will be easy to manage, and you won't have to manually scene-split them.

That said, Scenealyser may be able to split them up afterward if you capture a single file; I haven't used it for many years apart from a short burst recently when I found it didn't play nice with Windows 10. So for archive purposes, if you capture as one file, you may not have the luxury of using Scenealyser to split them if needed. There may be other programs around that can do that though.

It would appear be easier to archive as one file, but then it would have to be "Tape1" etc. This becomes a nuisance if you have one event over two tapes, whereas if you split the files on capture, you are not then constrained by the tape.

Archiving split files isn't actually that bad an idea, because you can easily set up folders for the dates of the DV files, using as many folders as you have dates (or events). The concept of "Tape 1", "Tape 2" is not needed.

My recommendation? Split/Scene Detect them on capture and archive into folders based on date and event.

[Cornucopia]

If you do this on a regular basis, like the pros have done, you'll soon find that keeping it as a single entity is much easier to manage. And you can always (and usually SHOULD) rename a bland generic name like Tape1 to 2020-09-21_MyFavVid_Tape1of2, etc.

Scott

[Alwyn]

Scott, I'm not a pro but I do have 3tb of DV and HDV files, all stored over 20+ years, in named, dated folders. I couldn't actually think of anything worse than having "single entities". With perhaps 4 or 5 events on each tape (and sometimes spanning tapes), trying to find specific events would be a nightmare.

[Cornucopia]

Guess you don't use an NLE with a database that supports named tape & clip & timecode & tag referencing, esp. one that can reference multiple projects. I also have multiple tb of Dv data on multiple raid drives, also from over 20 years. With the right database, proper search keyphrases can get me my source data clips in seconds, or if it is truly deep-archived, I can restore in minutes/hours.

Btw, this is also a best practice, including when referencing AVCHD and other structured-file based materials (using equivalent of ISOs).


Scott

[dellsam34]

It depends on how the tapes were shot, Skilled people shoot scenes in seconds and every time a change of angle or position. An average Joe would turn on the camcorder and hit record until either the battery dies, the tape run out or need a bathroom break. If your tapes are the latter then sure you can choose scene selection, If the tapes have a large number of scenes then you want to capture the whole thing in one file.