Archived from groups: rec.video.desktop (
More info?)
"PTravel" <ptravel@ruyitang.com> wrote in message
news:372ia7F56qns4U1@individual.net...
>
>> >
>> > Capturing to DVD-compliant mpeg means CBR, not VBR.
>> >
>>
>> Site a creditable source for that.
>> VBR MPEG2 Video is a
>> very commonly used format, and in no way prohibited in any
>> DVD Standard reference I've read.
>
> You're right and I mispoke. What I meant was a good transcoder does
> _multipass_ VBR -- single pass will not yield as good results. Though, of
> course, it is possible to do VBR on realtime capture to mpeg, it is not
> possible to do multiple pass. I always think of VBR in the context of at
> least 2-passes, since otherwise it's simply a best guess based on the
> current content and whatever may have been buffered.
>
>> There is a reference to
>> CBR Audio as a preferred and expected format.
>>
>>
>> > Editing mpeg beyond simply non-frame-accurate cuts requires
>> > re-rendering/transcoding.
>> >
>>
>> You must know that you are nit-picking, to an illogical extent.
>
> I'm not nitpicking at all. Mpeg cannot do frame-accurate cuts without
> re-rendering. That will seriously, and negatively, impact the quality of
> the re-rendered video.
Totally wrong, with any of the "Smart" MPEG NLE, as I explained
the video data is not changed in any way. There is a change in the
number of frames in the GOP(s) effected and in the related GOP
header data, but no change to the image data. There is only one (or
two if the previous GOP header needs to be changed also) frame
that is rebuilt, with the changed header data but the same image data.
All the other frames in the GOP are bit for bit copies.
>Transitions cannot be applied, color and gamma
> corrections cannot be made, titles cannot be added, mattes cannot be
> included, compositing cannot be done without re-rendering which will have
> a
> serious, and negative, impact on the quality of the re-rendered video.
>
There is a need when applying such effects, using available editing
software, to both render the video (so that you have all the image data
available for each frame) and (after applying such effects) to re-encode
to DVD compliant MPEG. But you are grossly exaggerating any negative
effect. When your output remains close to the same parameters as your
input MPEG, there is seldom any discernible difference between them.
I have used a Non-Destructive Editor, many times to add text effects
and VirtualDubMod filters to remove Logos and for other purposes;
with some care and the use of a multi-pass encoding the results are in
no way inferior to the original MPEG.
If it is the case that you can't take a DVD Compliant MPEG source
and apply such effects without ruining your video; I would suggest
that you keep trying. If you haven't tried it because you "Just know
it won't work", try it some time. Remember Bumble Bees and B-52s
aren't suppose to be able to fly either.
>>
>> The "new" MPEG Editor's re-encoding of the 30th of a sec.
>> frame that includes the effected GOP header; can't be detected
>> by the human eye.
>
> It depends on where the cut takes place. If it's the 18th frame from a
> GOP,
> then 18 frames will need to be rerendered and that will certainly be
> detectable by the human eye.
>
Wrong again, as I explained the only frames that get rebuilt are
those that contain the GOP Header data, and even that one first frame
of the GOP is built using bit for bit the same image (based on the fully
rendered frame). The other 17 frames, in your example are bit for bit
copies of the image data in those frames exactly as they are,
unrendered. That's one reason they only take seconds while your
approach can take overnight, certainly hours in the best case.
>> None of the image data is effected at all, the
>> same 1s & 0s are copied to a new file. So frame accurate
>> MPEG cuts are not a problem.
>
> I don't know how you can say this. If you're saying it's not important to
> be able to cut to a single frame, I disagree. As an example, I frequently
> edit to music and an error of as little as 6 frames one way or the other
> produces a noticeable jolt.
I'm saying that you CAN make Frame accurate cuts with the new
MPEG editing tools. There are certainly times when being able to make
such cuts is critical. That said, there are also times when an "I-Frame"
cut is all that is needed. For the most part removing commercials from
captured TV video, is easily done on that basis. The transition to most
TV commercial breaks is both extensive and of the "Fade to Black" type.
There are often several "I-Frame" cut points that are "in the Black".
> If you're saying the re-rendering that will
> result in a cut to single frame doesn't impact the quality of the video, I
> similarly disagree; re-transcoding mpeg produces noticeable (and rapid)
> degeneration.
>
As I have described the new "Smart" Process where only one or two
frames per cut get, in even the slightest way, rendered then changed and
then re-encoded ( what you insist on calling "re-rendering"); there is NO,
none, zero "degeneration".
As to: going from MPEG and rendering that so that each rendered frame's
image data is used to encode to a low compression format native to an
editor; and then in that editor combine it with other data also in the
editors native format; and finally that editor's output being encoded to
DVD compliant MPEG with close to the same parameters as the original
MPEG; IF DONE CORRECTLY, with the right programs, there need
be no "degeneration". If you try to "improve" or make any significant
changes to encoding parameters, from those of the original MPEG then all
bets are off.
>>
>> The use of DVD Shrink as a part of the Creation of a DVD,
>> is just to allow for the occasional to rare case where the
>> capture/encoding settings you are using and/or the content
>> of the video (with VBR set) produce results that are slightly
>> over sized for your intended purpose.
>
> Then you agree that using DVD Shrink does have a negative impact on the
> quality of the video.
>
It can, slightly, under the "right" conditions, for a few seconds total
of
an hour's worth of video.
>
>> Its use saves the
>> time and effort of re-capturing/encoding at different settings,
>> just to avoid a very minor if any noticeable impact.
>
> Noticeable, yes. Minor? It depends on what your personal tolerance
> happens
> to be. The methodology you outlined will _not_ produce the best quality
> DVD
> that can be created using consumer gear and software. It may be more
> convenient, or faster, or cheaper, but it's not going to produce better
> video. Maybe the quality of the video that results is sufficient for you
> and the OP. It isn't for me.
>
Not having seen Your results, when you have tried the approaches
described, if you have indeed have ever actually given it a try, I have no
way to judge.
Why don't you use the free trial of VideoReDo
www.VideoReDo.com to cut up a MPEG file and compare the resulting
file with the original file; see for yourself.
Let's see your post here after you've made the comparison.
Luck;
Ken