DV or DVD capture quality help from the experts

[fopi]

I need a little help from the experts.

I was wondering what is the quality difference between a
- DV quality captured then edited and finally DVD burnt movie
and
- DVD quality captured then edited and finally DVD burnt movie?

In other words, do I gain in quality if I capture in DV format, then edit the movie (mainly cut, rearrange scenes, add music, not too many filters and other fancy stuff) and then export it to a DVD format?
Or is capturing in DVD format good enough since I will be exporting the end movie to DVD quality anyway?

Thanks for help.

[Ken Berry]

The answer is relatively simple -- at least to me! If you have the opportunity to capture and edit in DV format, do so, every time!

You are capturing from DV format, which for that style of video, is about as good a quality as you will every get. (Remember I am talking about DV, not HDV.) Moreover, DV is meant to be edited, so you can do so freely.

The algorithms used with the format, moreover, mean that it is essentially lossless. The purists would say that there is some loss, but in effect you can re-code DV (to DV) many times over and never be able to pick the deterioration on the screen. You finish your editing, then convert to DVD-compatible mpeg-2. If you maintain high quality settings in the mpeg-2, then the final result is really indistinguishable from the original DV on your final DVD (though in reality the higher compression rate used by mpeg-2 inevitably means that there has been some data thrown away, and hence some quality loss).

Now if you capture direct to DVD quality mpeg-2, it depends. First, you can set various capture qualities -- HQ, normal, long play, extended play... Each will have some deterioration which becomes more noticeable the lower down you go. With DV, you capture what you have in exactly the same quality. Period.

If you capture with the highest quality DVD/mpeg-2 settings, then you will get good quality, no doubt. But the problem is that, unlike DV, mpeg-2 is not really meant to be edited much, or if you do so, it is better to use a specialised mpeg editor like Womble. Mpeg-2 moreover is essentially a lossy format. Every time you recode it entails some loss, just like a jpeg. If you start out with high quality settings and maintain those settings, then the quality loss is not going to be really noticeable even after two or three recodes, but after that you might start to notice some artifacts creeping in.

Video Studio minimises this by the process they call SmartRender, which means that if you start with mpeg-2, and recode to mpeg-2 after editing, then only those parts of the video which have actually been edited, are recoded, so only those parts will suffer any loss in quality.

Some people also suffer problems like out of sync audio and video when they re-encode mpeg-2. Admittedly this has occurred less after VS8, but it still happens.

Overall, as I said at the beginning, it is just that much safer to capture and edit in DV format when you can, and encode to mpeg-2 once and once only, when the editing is finished. You might save time by capturing to mpeg-2 (if your computer can manage the load), but whoever said that video editing was meant to be fast???

[2Dogs]

From my own experiences of capturing directly to mpeg-2 and creating a DVD from that footage, I would say that there is hardly any difference in the finished DVD quality when you only make simple cuts, add titles and so on.

The great advantage is a time saving - since the miniDV footage is converted to mpeg-2 in real time as you capture. If you are then able to maintain the same project settings, you will be able to enable Smart Render when creating your DVD or the single mpeg-2 file if you're using the "suggested workflow".

The downside - there are inevitable artefacts at transitions, which you may not notice. They'll generally appear as brief flashes or flickers, and VS has always done a great job of hiding them if you try to step thru the affected video frame by frame in the preview window. Open the file up in Virtual Dub, however, and you will clearly see the offending artefacts - macroblocks, the inevitable result of VS trying to deal with trimmed mpeg footage that doesn't contain complete groups of frames. If you disable Smart Render, you can avoid getting the macroblocks - but then the whole video is re-encoded, with some inevitable loss of quality, and you have lost the advantage of time saving.

So, if the time saving is really important to you, by all means use the direct to mpeg-2 workflow. Otherwise, and especially if you have a recent fast dual core pc, the "classic" DV avi on the timeline workflow gives you the best quality and offers the greastest flexibility in output choices.

[skier-hughes]

The answer is relatively simple -- at least to me! If you have the opportunity to capture and edit in DV format, do so, every time!

same here

[Ken Berry]

Another comment for fopi: you haven't filled in your system details, so we have no idea what sort of computer you are using. If it is a recent Core 2 Duo or something similar, you can probably capture direct to mpeg-2 with no problem (though with the same comments by me and 2Dogs regarding editing it, applying).

But if you have an older computer not overly well endowed with resources, trying to capture direct to mpeg-2 can be a hit and miss affair. It is a process which is very demanding of system resources. The data is coming in from your camera at a rate of 25 Mbps, which is high. The computer has to try to apply a new codec (mpeg-2) to the DV signal, and thus compress it significantly. If your computer doesn't have sufficient resources, then the incoming data starts to build up in what is called a transcode buffer (i.e. DV waiting to be transcoded to mpeg-2). If the computer can't process and convert the data fast enough, then inevitably the transcode buffer will fill up, and the whole capture process will stop, either permanently, or at least until VS can process the data and empty the buffer, allowing capture to start again.

Just so you have the complete picture (in more ways than one!!)

[etech6355]

Great answers guys.
I happen to edit in the DV format because VS & other programs also perform smartrendering of the DV. I'll write back to tape which has to be in the DV format. Along with that I can also output the DV from the computer to my Sony DVD recorder via the firewire port. Not all DVD recorders will do this, I'm fortunate enough to have one that does.

Depending on the project I may capture in Mpeg if it's a simple project with simple cuts and all (as posted above). I'm also forced to work in the mpeg2 editing format with the videos from a dvd recorder or the analog capture card.

As 2Dogs posted VS does a nice job of editing mpeg, it's pretty good for making Titles, cuts, splicing (and mixing dv with mpeg2) etc. My biggest choice for DV is when using transitions, color correction or anything that involves a complete re-encoding of the video (or every frame of the video).
I think that DV has less loss then Mpeg2 when every frame needs to be re-encoded. So it does depends on your project.

That would be a quality comparision, for reliability the DV format has proven itself over & over again to be the most reliable format to edit in.

Thread edited by etech.