|
Posted by Frank on 10/05/96 11:38
On Fri, 27 Jan 2006 16:30:16 GMT, in 'rec.video.production',
in article <Re: HDV capture on under-powered PC>,
"Nappy" <noemail@all.com> wrote:
>
>"Smarty" <nobody@nobody.com> wrote in message
>news:pLWdnZaFneCCxUTenZ2dnUVZ_v-dnZ2d@adelphia.com...
>> Chris,
>>
>> I'm really not so sure the disk is indeed the bottleneck.
>>
>> Since normal DV capture has exactly the same data rate as HDV, and is
>> captured correctly by Paul's system, I am much more suspicious of the CPU
>> and / or perhaps the bus speed of his PC.
>>
>> The codec has a lot more work to do in encoding the incoming HDV stream,
>and
>> this is where the CPU and bus are probably the issue.
>
>
>It is a 25Mb data stream.
Yes, 25 Mbps in the case of 1080i HDV. 720p HDV is 19.7 Mbps, however.
>The codec doesn't do much on the data's way into
>the hard disk. It certainly does NO encoding. Decoding the stream for
>display is where the extra clock cycles go..
This depends upon the specific capture software being used.
Some programs write the raw HDV MPEG-2 Transport Stream to disk as an
..m2t file. In this case, since the input data is unchanged, a codec is
not involved in the process.
Others change the HDV MPEG-2 Transport Stream into an MPEG-2 Program
Stream and write the data to disk as an .mpg file. No codec is
involved here either since the datastream is unchanged.
Others decompress the (long-GOP) HDV MPEG-2 Transport Stream into an
I-frame-only MPEG-2 Program Stream and write an .mpg file to disk.
Here, obviously, a codec is involved, although all that it's doing is
a normal MPEG-2 decoding operation. No quality loss is involved.
And finally, other capture programs transcode the incoming MPEG-2
Transport Stream to an .avi file using a proprietary visually lossless
codec (the degree of visual losslessness is, of course, in the eye of
the beholder). In this case, and as in all audio or video transcoding
operations, two codecs are involved, one to decode the incoming MPEG-2
data and another to perform the encoding function. Some small degree
of quality loss will result from this transcoding operation, but many
people find it to be unnoticeable and therefore quite acceptable for
their purposes, especially given that the resultant intraframe-encoded
file is much easier to edit than the original intraframe-encoded
long-GOP MPEG-2 HDV data.
It should be noted that most programs that change the incoming HDV
MPEG-2 Transport Stream into another format before saving the data to
disk also usually decompress the incoming lossy compressed HDV MPEG-1
Layer II audio stream and save it in an uncompressed LPCM (linear
pulse code modulation) form. Again, this is done to facilitate
editing.
--
Frank, Independent Consultant, New York, NY
[Please remove 'nojunkmail.' from address to reply via e-mail.]
Read Frank's thoughts on HDV at http://www.humanvalues.net/hdv/
Navigation:
[Reply to this message]
|