Sign in with
Sign up | Sign in
Your question

Newbie Question - Do I have to encode twice?

Last response: in Graphics & Displays
Share
April 1, 2004 10:48:32 AM

Archived from groups: rec.video.desktop (More info?)

Very new to this DVD burning stuff... I dump my digital video camera
to AVI files about 4gig for 1 hour. Then in ULead VideoStudio I
convert them down to MPEG-1 which is fine quality for what I need. I
would like to fit a few hours on one DVD. The conversion takes a
while (maybe 2 hours) on my AMD Athlon 1600+. Now I have a file
200mb.

All is fine up to this point. Now I want to get these MPEG-1 files on
a DVD. When I run them through a DVD creator like Roxio or ULead, it
takes about 10 hours of 'encoding'. Can I just COPY them to a blank
DVD? Would they play? Is there anyway to copy them and create a
menu? Something like that blue default menu with bars of track info.
I REALLY don't need/want any fancy menus.

More about : newbie question encode

Anonymous
April 1, 2004 6:58:23 PM

Archived from groups: rec.video.desktop (More info?)

Eric wrote:

> Very new to this DVD burning stuff... I dump my digital video camera
> to AVI files about 4gig for 1 hour. Then in ULead VideoStudio I
> convert them down to MPEG-1 which is fine quality for what I need.

Why are you converting to MPEG1?


> All is fine up to this point. Now I want to get these MPEG-1 files on
> a DVD. When I run them through a DVD creator like Roxio or ULead, it
> takes about 10 hours of 'encoding'.

If the files are DVD compliant, then they shouldn't need to be
re-rendered. While MPEG1 is technically DVD compliant, you should
probably be using MPEG2. But the important thing is resolution. It
should be Full, Half, or Quarter D1 resolution. If not, it must be
re-rendered.


-WD
April 2, 2004 3:33:11 PM

Archived from groups: rec.video.desktop (More info?)

Like I said... New to this stuff...

I converted them from AVI to MPEG-1 thinking that they would be a lot
smaller and I can fit that many more clips on one DVD. For example, 6
hours of TV shows. The same way a stand-alone DVD recorder can record
6 hours or so at it's worst quality. Is this not the correct way to
do it?

Do I want to just use some program to convert my AVI to VOB and then
just copy the VOB's to the video_ts directory on a blank disc? Can I
do that?

Eric
Related resources
Anonymous
April 3, 2004 2:34:32 AM

Archived from groups: rec.video.desktop (More info?)

Eric wrote:

> I converted them from AVI to MPEG-1 thinking that they would be a lot
> smaller and I can fit that many more clips on one DVD. For example, 6
> hours of TV shows. The same way a stand-alone DVD recorder can record
> 6 hours or so at it's worst quality. Is this not the correct way to
> do it?

If you want 6 hours on a single DVD, then you're going to want to use
1/4 D1 resolution and somewhere around a 1500 kbit average bitrate. I
don't think it'll look too nice, though. Going from AVI to MPEG1 to
MPEG2 is obviously one step too many. Have a look at www.dvdrhelp.com

> Do I want to just use some program to convert my AVI to VOB and then
> just copy the VOB's to the video_ts directory on a blank disc? Can I
> do that?


No. But authoring programs like DVD-Lab will take your DVD-ready MPEG2
files and just remultiplex them into VOB files. (Without needing to
re-encode).


-WD
Anonymous
April 3, 2004 1:31:01 PM

Archived from groups: rec.video.desktop (More info?)

On Fri, 02 Apr 2004 22:34:32 GMT, Will Dormann
<wdormann@yahoo.com.invalid> wrote:

>If you want 6 hours on a single DVD, then you're going to want to use
>1/4 D1 resolution and somewhere around a 1500 kbit average bitrate. I
>don't think it'll look too nice, though.

That's how I encoded myself, so far, when trying to pack as much as
possible within a DVD. But now I'm doing it (1/4 D1) with CQ 65%. It
may mean a higher bitrate than 1500 when needed -but then, it's
needed. CQ, whenever possible, gives you the best benefits from VBR,
while using 1-pass encoding (faster). There's a drawback, indeed: you
cannot calculate the final size precisely. But better leave it to
chance wether you pack 4 hours or 7 on disc.
Anonymous
April 3, 2004 7:56:24 PM

Archived from groups: rec.video.desktop (More info?)

"Bariloche" <bariloche@bariloche.com> wrote in message
news:8aps609asqoiuniutj28489ptigc2tn597@4ax.com...
> On Fri, 02 Apr 2004 22:34:32 GMT, Will Dormann
> <wdormann@yahoo.com.invalid> wrote:
>
> >If you want 6 hours on a single DVD, then you're going to want to use
> >1/4 D1 resolution and somewhere around a 1500 kbit average bitrate. I
> >don't think it'll look too nice, though.
>
> That's how I encoded myself, so far, when trying to pack as much as
> possible within a DVD. But now I'm doing it (1/4 D1) with CQ 65%. It
> may mean a higher bitrate than 1500 when needed -but then, it's
> needed. CQ, whenever possible, gives you the best benefits from VBR,
> while using 1-pass encoding (faster). There's a drawback, indeed: you
> cannot calculate the final size precisely. But better leave it to
> chance wether you pack 4 hours or 7 on disc.

Using the CQ mode in TMPGenc is by far the best mpeg encoding, but you are
right that the final file size is difficult to control. If you can access
this group- news:alt.binaries.vcd.d which is a discussion group, no files
posted, then they have a method using AVISynth to creat a script that will
encode 5 seconds of every minute of the entire video file using the CQ
setting in TMPGenc. This way you have a pretty good sample of the full
video and can quickly adjust the CQ setting to get the desired final file
size. This site http://www.svcd.cc has the info, but it costs $20/year and
is worth it. If you post a question to the above group, you will usually
get help and a link to the info on using the CQ mode.

Mike T
Anonymous
April 5, 2004 12:42:06 PM

Archived from groups: rec.video.desktop (More info?)

On Sat, 03 Apr 2004 15:56:24 GMT, "Mike T" <mtallent@u$-i$p.net>
wrote:

> If you can access
>this group- news:alt.binaries.vcd.d

Unfortunately, my ISP doesn't follow it, neither can it be found on
Google.

>they have a method using AVISynth to creat a script that will
>encode 5 seconds of every minute of the entire video file using the CQ
>setting in TMPGenc.

It should be easy to write a program that would read the AVI header,
to find out the framecount, then writes down a script

V=AviSource("TheClip.avi")

V1=V.Trim(0, 150)
V2=V.Trim(1800, 1950)
V3=V.Trim(3600, 3750)
....

Return V1+V2+V3...

Is this what they are doing?

I'm thinking of programming a tool that shall do that, and a few other
things as well. Let's see if I can find some time in the near future.
Anonymous
April 5, 2004 8:06:37 PM

Archived from groups: rec.video.desktop (More info?)

On Mon, 05 Apr 2004 08:42:06 +0200, Bariloche <bariloche@bariloche.com>
wrote:

> V=AviSource("TheClip.avi")
>
> V1=V.Trim(0, 150)
> V2=V.Trim(1800, 1950)
> V3=V.Trim(3600, 3750)
>...
>
> Return V1+V2+V3...
>
>Is this what they are doing?

I saw an avisynth script like that for CCE's 1-pass VBR mode and it used
the SelectRangeEvery command.

For example:
SelectRangeEvery(250,25) returns the first 25 frames (1 PAL second) out
of every 250 frames (10 seconds). You could encode 10% of your file that
way and then multiply the filesize by ten to get a rough idea of how big
the completely enoded file would get (in CQ mode).
Anonymous
April 5, 2004 8:27:27 PM

Archived from groups: rec.video.desktop (More info?)

>SelectRangeEvery(250,25)

I just remembered that for best results you should use your MPEG-2 GOP
size as the second number, so if you use a GOP size of 12 the command
would be SelectRangeEvery(120,12) to encode 10% of your avi file or
SelectRangeEvery(1200,12) to encode only 1% of your avi file, etc.
Anonymous
April 5, 2004 9:23:47 PM

Archived from groups: rec.video.desktop (More info?)

"Bariloche" <bariloche@bariloche.com> wrote in message
news:ctv170p6qm8f72qatc4nli5ph8psioh6o7@4ax.com...
> On Sat, 03 Apr 2004 15:56:24 GMT, "Mike T" <mtallent@u$-i$p.net>
> wrote:
>
> > If you can access
> >this group- news:alt.binaries.vcd.d
>
> Unfortunately, my ISP doesn't follow it, neither can it be found on
> Google.

Well, then it is time you need a better usenet provider. Below is one that
has free service, I believed it has limited downloads, but for text groups
this should not be a problem. http://www.teranews.com/

>
> >they have a method using AVISynth to creat a script that will
> >encode 5 seconds of every minute of the entire video file using the CQ
> >setting in TMPGenc.
>
> It should be easy to write a program that would read the AVI header,
> to find out the framecount, then writes down a script
>
> V=AviSource("TheClip.avi")
>
> V1=V.Trim(0, 150)
> V2=V.Trim(1800, 1950)
> V3=V.Trim(3600, 3750)
> ...
>
> Return V1+V2+V3...
>
> Is this what they are doing?
>
> I'm thinking of programming a tool that shall do that, and a few other
> things as well. Let's see if I can find some time in the near future.

I don't remember the details, just that the script allows for encoding 5
seconds of each minute of the video to get a good representation of the
action scenes to find the CQ setting that would give the final file size
desired.

Mike T
Anonymous
April 5, 2004 11:41:55 PM

Archived from groups: rec.video.desktop (More info?)

"Mike T" <mtallent@u$-i$p.net> wrote in message
news:D Cgcc.12392752$Id.2070862@news.easynews.com...
>
> "Bariloche" <bariloche@bariloche.com> wrote in message
> news:ctv170p6qm8f72qatc4nli5ph8psioh6o7@4ax.com...
> > On Sat, 03 Apr 2004 15:56:24 GMT, "Mike T" <mtallent@u$-i$p.net>
> > wrote:
> >
> > > If you can access
> > >this group- news:alt.binaries.vcd.d
> >
> > Unfortunately, my ISP doesn't follow it, neither can it be found on
> > Google.
>
> Well, then it is time you need a better usenet provider. Below is one
that
> has free service, I believed it has limited downloads, but for text groups
> this should not be a problem. http://www.teranews.com/
>
> >
> > >they have a method using AVISynth to creat a script that will
> > >encode 5 seconds of every minute of the entire video file using the CQ
> > >setting in TMPGenc.
> >
> > It should be easy to write a program that would read the AVI header,
> > to find out the framecount, then writes down a script
> >
> > V=AviSource("TheClip.avi")
> >
> > V1=V.Trim(0, 150)
> > V2=V.Trim(1800, 1950)
> > V3=V.Trim(3600, 3750)
> > ...
> >
> > Return V1+V2+V3...
> >
> > Is this what they are doing?
> >
> > I'm thinking of programming a tool that shall do that, and a few other
> > things as well. Let's see if I can find some time in the near future.
>
> I don't remember the details, just that the script allows for encoding 5
> seconds of each minute of the video to get a good representation of the
> action scenes to find the CQ setting that would give the final file size
> desired.
>
> Mike T
>
>

I also prefer a RT "CQ" VBR approach, but then I have a HW
encoder and capture software that works best in that fashion.

On the rare occasion that my captured DVD compliant MPEG2
Half D1 file is too large to fit the DVD or CD I want to put it on;
I just Author it then run it through DVD Shrink. Since with my
normal settings the files usually fit, when they don't it is by a small
amount and they suffer little in DVD Shrink.

Luck;
Ken
Anonymous
April 6, 2004 12:50:44 PM

Archived from groups: rec.video.desktop (More info?)

On Mon, 05 Apr 2004 16:06:37 GMT, erratic <erratic@reply-to.address>
wrote:

>I saw an avisynth script like that for CCE's 1-pass VBR mode and it used
>the SelectRangeEvery command.

Yes, SelectRangeEvery would be easier. But, why using 1-pass VBR mode?
I thought the idea was encoding in CQ, then extrapolating the results
to forecast the final file size.
Anonymous
April 6, 2004 1:30:14 PM

Archived from groups: rec.video.desktop (More info?)

On Mon, 5 Apr 2004 19:41:55 -0500, "Ken Maltby"
<kmaltby@sbcglobal.net> wrote:

> I also prefer a RT "CQ" VBR approach, but then I have a HW
>encoder and capture software that works best in that fashion.

It allows you to capture CQ instead of CBR? How fine -even if I would
not, generally speaking, use it for VHS tapes that need further
processing (cropping, filtering) before the final mpeg encoding.
Anonymous
April 6, 2004 6:03:20 PM

Archived from groups: rec.video.desktop (More info?)

On Tue, 06 Apr 2004 08:50:44 +0200, Bariloche <bariloche@bariloche.com>
wrote:

>Yes, SelectRangeEvery would be easier. But, why using 1-pass VBR mode?
>I thought the idea was encoding in CQ, then extrapolating the results
>to forecast the final file size.

One-pass VBR in CCE is similar to CQ in TMPGEnc. In TMPGEnc you select
a quality value, in CCE a quantizer value, but the result is an
unpredictable filesize.
!