AVI file questions

Hi,

Please help me settle a debate.

If you have an original AVI file, then convert it to a QT so you can edit in Final Cut and then lay off your sequence to a HD Cam tape, will that improve the quality? In other words, would the playback on an HD Cam tape look better than playback on mini DV?

I'm saying "no" because the information on the original file AVI is all you have to work with -- you can't make something Hi Def that wasn't recorded in Hi Def just by laying it off to a HD tape, right? Wouldn't that be like taking a postage stamp and blowing it up to fit a billboard then expecting it to look better because it's on a bigger medium?

Is there any way to up-convert the AVI to Hi Def in Final Cut?

Also, is there any difference in quality if you bring in the AVI to your timeline, then render it edit rather than converting it to QT first then importing to FC?

Thanks for your help.

Shelley
MacBoo Pro 2015
16 GB Ram
OS X 10.13
Premiere Pro CC

Quote

If you have an original [lots of words left out] will that improve the quality?

No. I cut out the middle of your sentence because you could replace it with anything and the answer would still be no. As you said, what's in the original footage is all that'll ever be there, and it can't be improved.

Quote

Is there any way to up-convert the AVI to Hi Def in Final Cut?

Not in Final Cut, no. Final Cut doesn't deal with AVI-type media reliably. (It may not deal with it at all, I've never tried. I know it doesn't work, I just don't know the details of how it fails to work.)

Quote

Also, is there any difference in quality

That's a difficult question to answer. Too many variables. "AVI" isn't a format, but a whole family of formats, and there are lots of different possible ways to get out of it and into something useful. What exactly are you trying to do?

"What exactly are you trying to do?"

I'm trying to make an AVI file taken off a surveillance camera look cleaner. Someone told me there was a way to up convert an AVI file to HD and I'm said I don't think that is possible. And was pretty sure that laying off an edited sequence using QTs that were made from AVI and then laid off to HDCam tape would not make any difference in the quality. I know I can make enhance the video some and make it look a little cleaner with some contrast adjustments and sharpening filters but it's not going to change the quality.

Shelley
MacBoo Pro 2015
16 GB Ram
OS X 10.13
Premiere Pro CC

Garbage in, garbage out. If it looks bad originally, making it bigger isn't going to make it look any better. In fact, it will make it look WORSE, as all the compression artifacts will be blown up to 10 times their size.

You MIGHT be able to clear this up with other "forensic" video cleaner applications...like stuff the CIA or FBI might employ. It just isn't in FCP.


www.shanerosseditor.com

Listen to THE EDIT BAY Podcast on iTunes
[itunes.apple.com]

I'm going to assume that this isn't something you yourself sought to do, but something your client asked you to do. Cause otherwise, I'm about to stick my foot in my mouth, and if so, I apologize in advance.

There's this sort of popular misconception that video (and photography) can be "enhanced." It's not an extremely recent phenomenon; we saw it depicted with eye-rolling fetishization in "Blade Runner," for cryin' out loud. But wherever it got started, the notion has worked its way into the popular consciousness that you can take a bit of footage or a still, do some computery stuff to it, and suddenly read that license plate clear as day when previously it had been obscured by dust and rain and a honking great rhinoceros that was standing in front of it.

There are two aspects to this popular misconception that I think are worth noting.

First of all, it actually is possible to do some astonishing things with image processing algorithms. Every year or two a paper comes out of SIGGRAPH that seems like pure science fiction, like the one from a few years back that described a process for turning a short segment of low-resolution SD footage into a single extremely high-resolution still. This technology really does exist.

But the other side of that coin is that it requires extensive custom programming by honest-to-God geniuses, it takes weeks to process, and it usually fails catastrophically, taking your source material and giving you back something that looks like a reproduction of your shot drawn in crayon by a hyperactive two-year-old who was fleeing for its life while being chased by the aforementioned honking great rhinoceros.

None of this knowledge, of course, makes it any easier to explain to a client who asks us to "just enhance it" that it doesn't work that way and they've been watching too much television.

what they said...
blowing something up isn't going to make it cleaner.

can you blow something up in FCP?
yes, but it's close to the worst way to blow something up.

in order of goodness are:

hardware conversion 1: running it through a box like a teranex

hardware conversion 2: capturing via a kona card to HD.

software conversion 1: specialist software like Instant HD Pro, or Compressor

software conversion 2: FCP timeline.

however, my experience with footage that isnt great to begin with is that there isn't MUCH in it.
i had source footage from VCDs, and had that blown up via a hardware box.
i did an A/B comparison with the footage blown up in FCP, and while it was better there wasn't much difference.
(and contrary to my list, the specialist software was AWFULL - it highlighted the artifacts)


CLEANING UP:

as you guessed a bit of a tweak in FCP will help.
but there is a noise reducing plugin for FCP i've been using called Neat Video
[www.neatvideo.com]
that i think is pretty good.
you may want to give that a go.


cheers,
nick

I work with surveillance footage on a daily basis and there is usually not much you can do with it, but you can make some improvements.
1. Most clients with these requests (in my case these are attorneys) have a different perspective on what constitutes an improvement. When ever I tell them don't expect much they are usually over joyed with the results even though to me it still isn't very good.
2. Photoshop is a better way to improve quality for this type of video. You can't process hours of footage but if you can find the short portion that you are interested in, save it out as an image sequence and bring the files into photoshop. Make your adjustments like levels, deinterlacing, and color corrections on each file (if they all require the same adjustments, you can set up an action to make it go more quickly). After they are improved turn them all into smart objects and resize the files as smart objects. You will want to maintain the aspect ratio. Save them back out as tiffs and bring back into the timeline at the appropriate frame rate. The video should now look better than it did.
3. We avoid the word enhance and try to get people to use the word clarify.
4. If you are trying to clarify a stationary object (like a license plate that doesn't move but is obscured by noise and darkness you can save out a short image seq and use photoshop to process stacks-this will eliminate noise from video gain and improve the brightness at the same time.
5. You usually cannot improve something you can't see but you can do comparative image analysis-for instance you may not be able to see someone's face or read clothing logos but you can do class comparisons-in our situation if a defendant was arrested in the same clothing worn when the crime was committed you can show class comparisons like long sleeves, triangular shaped markings (sports logos), stripes on clothing, etc.
6. If your avi was made from a VHS tape, make sure you capture in full frame uncompressed, that will give you the best chance of finding something useful. If the AVI came from a DVR system it probably is the best you can hope for. Most DVR systems don't record in a file format that will play with out some kind of recompression-you might be starting out with a file that is already diminished from the original recording.