field order ??

[mrhafez]

Hi
In which case we must selest field order ? When I capture a PAL DV tape which one must be used ? ( frame-based or upper or lower frame ?

Thanks

[etech6355]

DV tape is Lower Field First when capturing via the firewire connection.

[mrhafez]

Thanks

[brocot]

Files recorded with WIN Tv 2000 (Haupage), is that "Upper field first"
(Its analogue i think)

[Black Lab]

Analog is generally Upper Field, digital is generally Lower Field (except for hard disk and dvd types, which are generall Upper Field - confused yet? ).

[Ken Berry]

Not to mention capturing analogue material via a mini DV digital video camera as a passthrough device, Digital 8 camera as passthrough or actually playing analogue 8mm or Hi8 tapes, or external capture devices that have a built in DV chip. All these capture as DV and so are lower field first...

[brocot]

Ok, thanks.

I do mi recordings on WINTV Pci card 350.

DVD quality CBR 5200 Bitrate (upper field)
then i cut out the commercials whit Ulead11
I create the new file (Upper field, CBR 4200 Bitrate)
At last i create the DVD (CBR 4200) ISO file. (4,2 Gb space)
Nero burns it on the DVD.

Best result is on DVD+R.
The DVD-Rw disk are sometime not so good.

Is this Ok?

[Ken Berry]

I note that you downgrade the bitrate from the original 5200 kbps to 4200, which of course will result in a lessening of quality of the final DVD. But I am assuming you do this to fit more video on a DVD (approx. 2 hrs with 4200 instead of approx 90 minutes with 5200...) Otherwise it looks fine -- though of course you need to make sure the final DVD also uses Upper Field First...

[brocot]

Oh, is it better to record on the same Bitrate as the final file.
I thought that it was better to change it, to rebuild the whole file.

ps: In the setting of the WIN Tv record setting, there is e field where i can make the choice between "GOPs 15 ore 6" . Standard it is on "15"
I have not the intention tho change it. But what is the meaning of it ?

[Ken Berry]

No, not quite. It is the other way around -- it is better to make the final file use the same bitrate as the original recorded video, if it is in the same format, of course. Apart from frame size, it is essentially the bitrate which determines the final quality of a video. Basically, the higher the bitrate, the higher the quality the final video will have.

Normally, a bitrate of, say, 8000 kbps will allow you to burn about 1 hour of high quality video onto a single layer DVD using standard high quality LPCM audio. A bitrate of 9000 would allow a marginally higher quality, but a little less time for the video on the same DVD. If using a bitrate, particularly a constant bitrate, higher than 8000, though, you have to take into account that some (many?) stand-alone DVD players have difficulty playing back bitrates higher than 8000 kbps so your DVD could appear to jerk in those places or even stop playing. In theory, though, you can use a combined audio and video bitrate which goes up to 9800 for a standard definition DVD and still be within the international DVD standard.

But some people are less interested in the quality of their final DVD, and just want to fit more video on the disc. So to do that, you lower the bitrate. That will lower the quality, but correspondingly, it will increase the amount of video you can burn to a DVD. Thus, a bitrate of 6000 will allow you to burn about 90 minutes of video to a single layer DVD, and the quality will still be good, though not as good as the DVD using 8000 kbps. And if you lower the bitrate to 4000, you will be able to burn 2 hours onto a DVD, though the quality will only be average. Below that, the quality drops off more rapidly.

With all these figures, you can squeeze another 10 to 20 minutes of video onto a disc if you use a more highly compressed audio format like Dolby or mpeg layer 2 audio. The more compressed the audio, the smaller the audio files on a disc, and correspondingly the more space there will be for video!

As for GOPs, they are Groups of Pictures and are the building blocks on which mpeg-2s are built. An mpeg-2 is a stream of GOPs, and each GOP consists of frames which are encoded in different ways, or carry different types of information. They always start with what is called an I-frame, which is essentially the master image for that GOP. As you know, an mpeg is a compressed format, and the compression is largely achieved through the GOPS. After the full I frame (which contains all the information for that sequence), there are a series of what are called P, B and D frames, each of which takes a different aspect of information from the I frame, but together they make up the image.

If there is not much movement in the video, then the subsequent frames don't have to do much work and merely repeat what is in the I frame. But if there is a lot of panning and zooming or other motion, then they have to do a lot more work each showing new information. This is also related to the difference between variable and constant bitrates.

As you have seen, GOPS can also vary in size, though 15 is very common. Generally, though, a smaller GOP, like your alternative of 6, makes an mpeg more editable because there is more information in each of the P, B and D frames, and if you make an editing cut in the middle of a small GOP, less of the information will be lost. Cut early or in the middle of a larger GOP, and the more information is lost, and the more likelihood that there might be a slight jerk between the preceding GOP and the next new one after the editing. The audio could also get out of sequence very slightly, but over time that would build up and result in the common problem of out of sync audio and video that afflicts so many people who edit mpegs!!

Specialised mpeg editing programs like Womble or Video ReDo, achieve their good results largely because they analyse the GOPs on a video, and better coordinate the cutting etc so that GOPs are cut cleanly at the end of a sequence, just before the next I frame, rather than in the middle. And they can 'fix' GOPs on badly cut videos.

That is all, of course, a very simplistic explanation of a rather complicated subject. My apologies to the experts who will see lots of holes in it!

[brocot]

Thank You,

I'm very satisfied whit the result of the final DVD.
Only the recordings of large files (8gb) results in problems. Using the time line : VS11 friezes.
Only exporting the file whit Neroshow time is the solution. After this operation it is possible to edit in VS11.
I have 2gb ram in the Dual core pentium4 processor.