| Re: ATTN Rely - Your request - <- Awesome RIB, thanks! |
|
| RIB (rib@rack.com) |
2015/04/16 01:53 |
Path: news.nzbot.com!not-for-mail
From: RIB <rib@rack.com>
Newsgroups: alt.binaries.documentaries
Subject: Re: ATTN Rely - Your request - <- Awesome RIB, thanks!
X-Newsreader: NewsLeecher v6.5 Beta 6 (http://www.newsleecher.com)
References: <m0mtia1dqilchlni53o4caqt51he2eh7kp@4ax.com>
Lines: 85
Message-ID: <hUJXw.171782$mZ3.118263@fx10.fr7>
X-Complaints-To: http://www.newsleecher.com/support/
NNTP-Posting-Date: Thu, 16 Apr 2015 07:53:49 UTC
Date: Thu, 16 Apr 2015 07:53:49 GMT
X-Received-Body-CRC: 2558719132
X-Received-Bytes: 3837
Xref: news.nzbot.com alt.binaries.documentaries:2932
In reply to "Knotty Bitz" who wrote the following:
> On Wed, 15 Apr 2015 16:23:39 GMT, RIB <rib@rack.com> wrote:
>
> > In reply to "Knotty Bitz" who wrote the following:
> >
> > > On Wed, 15 Apr 2015 07:56:50 GMT, RIB <rib@rack.com> wrote:
> > >
> > > > You are welcome.
> > > >
> > > > I've managed to get the file size down to about 650 megs for the AVIs
> > > > But
> > > > strangely I can't go much smaller than that for H.264
> > > >
> > > > The files on the DVD9s are way too big for the quality of the clips
> > > > which
> > > > means
> > > > it's very difficult to get a good size/quality ratio. I swear this is
> > > > done
> > > > on
> > > > purpose.
> > > >
> > > > I'll post one of each when I have more time and then you can tell me if
> > > > you
> > > > are
> > > > interested in the rest.
> > > >
> > > > RIB.
> > >
> > > My prediction of 2/3 smaller was based on doing the one episode, which
> > > happened to compress well. I was probably too optimistic. ;-)
> > >
> > > You know, it probably shouldn't be surprising when we consider that
> > > 50% or more consists of old archival footage - raw, unprocessed, noisy
> > > material, and lots of it taken from analog TV. In fact, I've seen
> > > docs get larger not smaller for this reason. I cranked the xvids
> > > through HB at CRF 23 and got an average of 534 MB per ep. If I'd gone
> > > for CRF 24, the average might have slipped under 500. And that was for
> > > resolution of 640x480. You're probably going for the max (PAL)
> > > resolution, closer vto 720x576. Therefore, if I had the original
> > > DVD-9s, I'd go for the better resolution, use a setting of CRF 24, and
> > > accept what resulted with the filesize.
> > >
> > > BTW I misspokie yesterday. I said that the original poster did the
> > > audio at 81Kbps. I must have had a brain fart. When I looked again it
> > > was actually 192K stereo.
> > >
> > > KB
> >
> > ---------------------------------------------------
> >
> > I did do a trial with a CRF of 24 and thought it looked pretty good although
> > the
> > resulting file was about 800 megs. I'll play around with it a bit and post
> > what
> > I think is the best quality over size. I'm sure you'll let me know if it's
> > crap
> > :-)
> >
> > RIB.
>
> FWIW -
> Using Part 8 as a test, I encoded at CRF 24 but with different audio
> settings. Then I demuxed and measured:
>
> H.264 video = 438 MB
> ac3 passthrough (448Kpbs) = 269 MB
> aac stereo (192Kbps0 = 117 MB
>
> So, at least 150 MB of the file's size is attributable to sticking
> with AC3 5.1 channel soundtrack. I question whether 5.1 is really
> needed (someone who did want it would probably rather download the
> DVD-9). If the track is downmixed to Stereo, even 192K may be slightly
> overkill.
>
> Splitting the difference, I'd consider Dolby ProLogic II at 192K.
> http://en.wikipedia.org/wiki/Dolby_Pro_Logic
------------------------------------
I did pretty much the same and came to the same conclusions. The 650 file is
with the sound downmixed to stereo. The content doesn't really benefit from 5.1
RIB.
|
|
|