Mpeg-2 File format

[benny74]

Hi all,
I have sony HD camera. the file type of the camera is mpeg-2.
my question is what is the best way to edit those files with video studio?
should i leave the files format as they are or convert them to some other format like DV AVI or somthing else.
the most importent parameter for me is the quality of the movie.

thanks

[GeorgeW]

Is that a Standard Definition Hard Drive camcorder, or a High Def camcorder?

What is the Make/model of the camcorder?

Regards,
George

[benny74]

camera model is:
Sony DCR-SR40 with 30 GB of HDD.

[Wolfgang van Ween]

I also have a Sony camcorder that records MPEG2, but to DVD. I keep files in that format and keep fancy transitions to a minimum to enable as much smart rendering as possible. Means I loose no quality to the final DVD and no endless waiting for rendering.

I do however decode into nearly uncompressed AVI (HUFYUF codec) when I need to do fancy stuff, like speed changes or smoothing my camera movements, which I use third-party programs for.

The thing I always have trouble with when cutting directly from the MPEGs is the sound synchronization, so I keep scenes short, and sometimes split off sound to hand-synchronize it later.

Try both and see what better fits your requirements.
Have fun!
Wolf

[benny74]

Hi wolf and thanks for your answer.
do you mean that the camcoder is having a DVD disk in it and it record to the disk directly?
so basically what you say is that if i don't need to do a lot of transitions i can use visual studio just to cut the scenes and burn the files to CD.

moreover if i do want to add some transitions am i need to convert the files from mpeg-2 to some other format? to which format and with which tool it can be done.

thanks a lot

[roy wood]

Hi Benny, I'm afraid you're going to get some conflicting answers to your questions. There are those that will say convert your Mpg to DV/ AVI files because of quality loss, Audio out of sync problems when editing in Mpg.

There are others like myself who will say Rubbish because I Capture, Edit extensively (Multi cut video add Transitions, multiple Filters, Overlays and Audio) without any visible difference whatsoever using only Mpeg2 and I have the advantage of my video files being a third of the size and my rendering times being much shorter.

Since we are all using the same Editing programmes and mine is not a particularly powerful machine I can only conclude that the disparity of results has something to do with the set up and compatibillity of mine and others Machines and Ulead software.

So my advice is the same as Wolfs try it and see, you're video is already Mpeg2 make a short video burn it to a rewrite disc and if you're not happy convert to AVI.

[DVDDoug]

I agree with Roy. Your DVD will be MPEG-2, so leave it as MPEG-2 if you don't have any problems. In general, you are not going to gain anything by converting it to DV. In fact, you will (theoretically) loose quality with the 2nd MPEG encode (MPEG is lossy compression. Data is thrown-away during the compression process.)

However, I am one of those who's had occasional "lip-sync" problems (See my signature/tag-line below! ) and occasional crashing caused by edited MPEGs. And in my case, it was the editing that either caused the problems or perhaps brought-out the worst in an already corrupted file. I never had any trouble with DVDs made from unedited MPEGs. My solution was to aviod MPEG whenever possible, and I bought a special-purpose MPEG editor from Womble.

Note that any "real editing" (transitions, text-overlay, color adjustment, etc.) will require the MPEG to be decoded and re-coded. Using Smart Render and and checking Do Not Convert Compliant MPEGs will help to avoid any unecessary "lossy" re-coding. (Although, some people report that Smart Render contributes to sync problems.)

[roy wood]

Hi Doug, you've read my previous reply, now I do all the things without limitation in Mpg that yourself and many others say should result in degredation but doesn't for me.

I have then duplicated the whole process exactly using an AVI capture of the same video sequence and workflow. On Monitor playback and DVD/ TV I can detect no difference in picture quality between the two.

I do make sure that when I edit, Mpg an 'I frame' is the first frame, of the retained sequence and I also use a high Bitrate, the Audio I edit last.

You mention 'Womble' as a specialist Mpg editor in what way does that work, that makes it Supposedly less prone to loss than 'Ulead'.

[Ken Berry]

I don't think Doug was suggesting that Womble makes mpeg-2 less lossy. Mpeg-2 by its very nature IS lossy. No one can change that, but it can at least be minimised.
In that regard, I grant you that if you use relatively high quality settings and re-code only once or twice (some would say even 3 or 4 times), the quality loss is unlikely to be noticeable to the naked eye. In other words, I agree with your conclusion that in the circumstances you describe as your workflow, you will achieve the same level of quality in the final DVD whether you capture and edit in DV then convert to mpeg-2, or whether you capture and edit in mpeg-2 to start with.

Many people do this as standard practice. But there are other users out there who do have the other sorts of problem Doug described. And for them, capturing wherever possible in DV and editing in that format, gets around those problems.

What Womble is good at -- and I think this is what Doug meant -- is that it was developed specifically as an mpeg editor. And it seems to avoid the sort of problems that some people meet when editing mpeg in VS. Most specifically, it never seems to produce out of sync audio and video.

[Wolfgang van Ween]

...I do make sure that when I edit, Mpg an 'I frame' is the first frame, of the retained sequence and I also use a high Bitrate, the Audio I edit last....

Roy, excuse my ignorance, but how do you do that? How do you tell in Video Studio whether a frame is an I-frame?

And a PS to my own remark about out-of-synchs: this happens when I edit 5.1 Dolby; I haven't had out of synch when I worked with converted files, e.g. MP3 or WAV stereo.

[roy wood]

How do you tell in Video Studio whether a frame is an I-frame?

Hi, an 'I frame' in MPEG contains all the Pixel information to re-produce the image taken with your camcorder the next few frames don't contain all this data only the parts that have changed, then once a certain point of change is reached a new 'I frame' is created etc this is how high quality MPEG uses only one third of the file space of AVI. On playing the rendered video the Codec puts the data back.

If you then cut your video half way between two 'I frames' most of the data in the next few frames is missing hence the 'Lossy Tag'. The more times you do this and render the more you loose.

To avoid this find the point at which you wan't to cut the clip then Single frame forward or back looking for the clearest image this is your 'I frame' make sure you cut to keep it within the part you wan't to keep.

[roy wood]

Thanks for your considered reply and conclusions Ken, unfortunately the way MPEG is usually portrayed a lot of new people may well have been put off at least giving it a try. Despite the advantages of massive file space saving and the time saved in the rendering.

[Ken Berry]

Admittedly, the space savings are huge and the rendering times saved can also be large. But for those people whose computers are not up to the job, trying it your way can waste them an awful amount of time and angst -- and often, a lot more than they might otherwise have saved!!!

Yes, they should try it. But if it does not work, they should at least be aware of the alternate method.

[Wolfgang van Ween]

To avoid this find the point at which you wan't to cut the clip then Single frame forward or back looking for the clearest image this is your 'I frame' make sure you cut to keep it within the part you wan't to keep.

Roy, thanks for that advice - I knew about I-frames, but I had no idea one could "see" them. Actually I just single-frame-forwarded through an MPEG clip (one of the skiing samples provided by ULEAD), and to be honest I couldn't spot a single image in over 8 secs that would be clearer.

What are you looking for, where are you looking at?

[GeorgeW]

Thanks for your considered reply and conclusions Ken, unfortunately the way MPEG is usually portrayed a lot of new people may well have been put off at least giving it a try. Despite the advantages of massive file space saving and the time saved in the rendering.

You do get alot of valid pros/cons to either method. Alot depends on the users' workflow, video source, and knowledge on DVD Authoring (if distributing on DVD).

For instance, at the moment 95% of my video source comes from dv camcorders, so I chose to capture/transfer the dv .avi to my computer, edit all in dv .avi, and then final output for DVD. I also archive the final dv .avi to dv tape -- so that's another reason I try to keep it all DV.

But in the case where source starts out as mpeg, it could be best to capture/edit the mpeg instead. I use this approach when my source is mpeg -- like from dvd camcorder, or from a set-top dvd recorder. I don't bother to convert it to .avi first. However, for a while that was the approach taken by some editors of HDV mpeg2 (converting the mpeg2 to an intermediate codec such as Cineform). But as machines get faster, and software matures with dealing in HD mpeg (HDV mpeg2, AVCHD, AVC-Intra, etc...), the ability to edit the native capture stream gets easier and easier...

A possible danger for those who don't understand mpeg compression, and how to calc bitrates to fit on their DVD -- they capture from dv-to-mpeg, do all their edits (or go to share/create dvd-mpeg), and then try to put too much on their dvd -- potentially causing another render of the mpeg files in their DVD Authoring step.

Regards,
George