View Full Version : Capture formats Inquiry

10-18-2003, 06:30 PM
Hello All;
Ever since I got T2 about a year ago, and now T3, I have been capturing all files to uncompressed .RTV which are HUGE. I am wondering if this is necessary when using footage from a mini-dv camcorder ? Since Mini_Dv isn't broadcast quality (Or is it ?), is uncompressed .rtv really necessary ? Would I be doing just about as good if I were to capture as one of the compressed dv formats which would require a whole lot less HD Space ?

In other words, by inputting via. S-Video or Firewire, and then outputting to DVD, am I gaining any quality using .rtv or would I be doing just about as good if I were to use a dv .avi file (Or another compressed format) which would take up a lots less HD space >

Opinions, advice, etc ...


10-18-2003, 06:38 PM
RTV will look better if doing alot of layering and re-rendering, but for single pass capture and save from a mini-dv, it is a bit of an overkill.

Try capturing the same footage using DVType2 and RTV and compare. You'll be suprised.
(capture with s-video this way)


10-18-2003, 06:41 PM

10-19-2003, 01:57 AM
Transfer your file via firewire as native DV, use Type-2 DV, and you will have the BEST possible quality based on your source material.

You CAN'T make something that is DV better than DV by using RTV, that's silly.

Seriously, a bit by bit transfer of a digital file without uncompressing it is HIGHER quality than playing the video through analog connections just to save it as uncompressed.

100% of the background rendering in VT-Edit is done as uncompressed so you lose NO quality no matter how heavily you layer or composite.

VT-Edit simply uncompresses the DV as it plays it, producing the maximum quality.

Using RTV for DV/DVCAM/DVCPRO25 is simply a WASTE of HD space.

10-20-2003, 07:00 PM
For some excellent geek speak on compression formats and TMPGEc, hop over to VTNT and read the thread Tempgenc and VT3, nice comments from Dr Croos of Newtek.

hit messages on the left.

10-20-2003, 07:00 PM
My apologies for haste,
It is Dr Cross of Newtek.

10-20-2003, 08:21 PM

10-21-2003, 12:26 AM
Because all NLEs don't work like the VT[3].

Seriously, I'm sure many on line finishing systems don't support DV directly in real-time like VT[3] does.

Also, ALL other DV based edit systems that I know of, if they need to do a background render do it as DV, thus REDUCING quality. VT[3] doesn't do it this way as I've already explained.

If you're not altering ANY bits, I assure you your quality isn't going to go down.

10-21-2003, 03:30 AM

10-21-2003, 09:12 AM
No - FireWire is completely lossless. You can transfer to your hard drive, to tape, to another hard drive or 2, through a few generations of tape, back to your hard drive, and barring tape dropouts, it'll be bit-for-bit identical.

11-04-2003, 02:19 PM
'You CAN'T make something that is DV better than DV by using RTV, that's silly"

now thats a false statment.

just capture any shot with sky in it in dv format through firewire, and than capture it again through analog as rtv uncompressed.

now try keying the sky...youll notice hugh artifacts in the matte edge in the dv format which means:
you cannot add data throuh analog capture but u certainlly can change the quality(a vague concept but bear with it) of the image.

try working with mainconcept's m-jpeg codec, pretty good and not too expensive,they also have lossleg jpeg compression codec but thats costly.
i presonally avoid firewire capturing , it keeps the color more accurate that analog capture, but introduces a large collection of artifacts (jagged edges,blocks if u try keying and just about any artifact known to man )

good luck.

11-04-2003, 04:44 PM
Sure, if you're looking to soften your video image, going analog is a good way to do it, I don't deny that.

And yeah, I can see how that would improve keying in some situations.

But if you want clean keys, don't record DV to begin with.

11-09-2003, 05:49 PM

You said

"But if you want clean keys, don't record DV to begin with."

If you have a mini-DV format and elect to use the firewire capture, VT3 automatically makes this an AVI. Since I have updated to even the new patch, TED does not like a whole timeline full of AVI's. It just plain runs horrible.

I have utilized the S-video capture to ensure that I get a stable timeline if time is of the essence. Is there any way of getting around this?

11-09-2003, 09:43 PM
You mention "mini-DV format", well, if you've recorded to tape, in "mini-DV format", you've already saved it compressed and sampled in a way making it less than optimal for keying.

If you want to record live from your camera via the Y/C connection, that would prevent this.

The new patch was a rather flawed fix, I'm sure there will be another coming soon, and if you are using DV at all, you should e-mail Dr. Andrew Cross at Newtek for the latest DV patch till a public one comes out.

I have no problems doing all DV with it.

As for getting around the problem you mention, sure, simply select "Program Out" in the capture panel instead of "Main In", and then you can capture via firewire to ANY codec you want.

But until you have the patch from Andrew I mentioned for DV or a new public one, be aware there are A/V sync issues with firewire captured DV. So, IF you don't want to save as DV now anyway, I would recommend you use the Y/C in for now.

11-20-2003, 06:33 AM
You mentioned the DV patch from Dr. Andrew Cross. When I sent an e-mail to support attention to him, they said they only knew of one DV patch. They sent me a zip file named: WindowsMediaLib. Is this the file that you refered to in your reply, since I am not sure what I was expecting.

Also you said that you had no problems with DV in your system. What codec were you using?

11-20-2003, 07:33 AM
Originally posted by AntAT
But, as far as I know, there is a decompression/recompression going from the camera or deck thru the firewire to the computer's hard drive and saved with a dv codec. Then once you are done editing, if going back out thru firewire you have to render the project to a single dv file (decompress then recompress into a new dv clip). I have to side with Eugene's analysis here (though I love working with RTV's).

With reference to your first statement about losses on the way from the camera to the HD via Firewire: to the best of my knowledge, this is incorrect. There is no degradation (recompression).

On the second point, while it's true there may be some miniscule losses associated with recompression when exporting via Firewire after editing, two things to note:

First, it has been explained that any losses are truly trivial, because the real "damage" is done on the initial (in camera) compression, and re-compression generally adds little to this.

In any case, this thread was about going from VT-Edit to DVD, not back out via DV ... hence the second point isn't relevant in the setting Eugen was discussing, even if we were allow that there is some minor re-compression penalty in your scenario.

12-06-2003, 05:51 PM
my 2 cents:

I have been all RTV since I first got the VT2 more than a year ago. We use SDI in/out from Digibetas and BetaSP (A/D conversion). Obviously, very pleased with the quality. I have been experimenting with compression recently based on some of Eugene's suggestions. I had a project that we were re-editing that ran about 25 minutes but had well over 300 elements most of them RTV's. My drives filled up before I got half way done.

I went into the filebin and right-clicked, convert to PicVideo MJPEG every RTV (except the animation ones with alpha), then deleted the RTV's. Worked marvelously. Re-couped about 60 Gigs of drive space and captured the rest of my clips using the same codec. This project is not for broadcast, but for IMAG, so I hit it with level 17. Remarkable quality even at this level. I have used level 19 with some of my broadcast spots and I can't seen any difference from RTV, but I am not trying to chromakey either.

RTV will still be my first choice, but when needed PICVideo MJPEG can't be beat. BTW, it is 4:2:2 as opposed to DV which is 4:1:1. Cost under $100.