Can someone help "PTravel" make a frame accurate cut with VideoReDo, he says he can't.

"PTravel" makes the claim that you can't make frame accurate
cuts in MPEG. (See the thread: "What product do you think is
best for AVI to DVD?") So I finally made this proposal:

"PTravel"'s response:

I've experimented with VideoReDo when doing video extraction from my Tivo
(which produces non-DVD compliant mpeg2). VideoReDo did not let me do
frame-accurate edits though, of course, it didn't matter when all I wanted
to do was chop out commercials.

Luck;
Ken

 

I don't need help with VideoReDo because I have no use for it. As I've
said, I edit video projects, I don't trim commercials out of TV programs
that I've captured using USB toys.

VideoReDo, as well as the other tools you've identifed, are probably fine
for hobbyists who want to compile a DVD of Survivor episodes. They're not
appropriate for anyone doing any serious editing and authoring.

If there's a way to make VideoReDo do frame-accurate edits without rendering
or re-transcoding, that's very nice. It's also of no particular interest to
anyone who wants to create video projects from AVI (you might, btw, look at
the subject line of the thread which was about going from AVI to MPEG).

 

Don't forget. There are a lot of people out there who actually do think that
MPEG-2 is somehow a higher quality video than DV. So, the toy maket abounds.

 

Your camcorder's DV Quality is of little interest to most. It is only
an intermediate stage on the way to MPEG-2 in a DVD. But you
are right that MPEG-2 is valued by many; for instance; anyone who
makes a DVD, all the major movie studios, the broadcast networks,
cable and satellite systems.

Still if you want to spend many hours turning your "Higher Quality"
DV into DVD compliant MPEG-2 that could be slightly "Better" in
some way, have at it. Me I'll settle for MPEG-2 video captures that
are ready to author into DVDs what, when played, provide video
that looks exactly the same the source.

Luck;
Ken

 

Sorry, but that makes absolutely no sense. The format is of interest to
anyone who wants to edit video. Mpeg is a delivery format, and not one that
is amenable to editing.

Well, yes, but that's like saying computers are of no interest to anyone
producing a DVD because they are only an intermediate vehicle on the way to
a television set.

None of the major movie studioes, none of the broadcast networks, and no
other content producers edit projects in mpeg. Cable and satellite are
delivery systems -- their only concern is bandwidth.

And that's the whole point. It's not "slightly better." It's dramatically
better. Retranscoded mpeg is not broadcast quality. DV-codec-encoded AVI
is above broadcast quality. Some people are happy with VHS quality video --
perhaps you're one of those people.

But cannot be edited like an AVI. Your VideoReDo is not an NLE. It doesn't
do titles, transitions, allow color or gamma corrections, permit
compositing, do insert edits, etc. It's useful for one thing and only one
thing: simple cuts edits of mpeg video.

Utter and complete nonsense. The best commercially-produced DVD will not
look exactly the same as the source video. A good transcode of an AVI will
come close, but the differences are easily detectable. A real-time mpeg
capture, retranscoded and compressed by DVD Shrink will not even come
remotely close.

 

You can keep changing the question with all your strawman arguments,
but you still won't and obviously have never in the past TRIED to produce
good results using any process than the one you espouse. You say you
have tried VideoReDo, and it won't make frame accurate cuts. No one
else that has ever tried it would come here and make that claim.

You say there are tremendous problems with the quality of MPEG that
hasn't spent time as AVI. When called on that you say the issue is that
MPEG can't be edited. When directed to the new MPEG frame accurate
editors, you say they can't work or change it to be about adding various
effects. You even had to add DVDShrink to the real-time capture process,
when you know it was mentioned only as a means to allow a slightly over
sized capture to remain usable, not as a routine part of the process.

Your assertion that Direct to DVD Compliant MPEG Capture, cannot
produce results that are indistinguishable from that video source (not the
output of the movie camera on the movie lot, what the guy at home has
as a source.); flies in the face of, not only my experience but many others.

If done properly it is the easiest and fastest way to produce a true copy
of the viewing experience provided by the source. ( Cable, Satellite, VCR,
TV tuner)

It is certainly the case that if you are working with material that
requires the type of editing that would actually require a re-encoding if
it were already DVD compliant MPEG, like your camcorder DV, then
it would be only common sense to edit it in the format it already is in.
This does not make your assertion that "MPEG can't be edited" valid.
It certainly doesn't mean that your characterization of the quality of
MPEG-2 that was made in real-time and without the "benefit" of
having spent time as AVI, is any more valid.

Luck;
Ken

 

And you're the only person here who would claim that VideoReDo is a NLE.

Ken, who cares if it can produce frame-accurate edits. The bottom line is
this: if you're editing video, use video editing software. VideoReDo is not
substitute for FCP. Premiere, Liquid Edition, or even Studio or Microsoft
MovieMaker.

If you're editing commercials out of TV shows that you've recorded, use
VideoReDo. NLEs are overkill and, clearly, you're unconcerned with quality
anyway.

Wrong, wrong, wrong, wrong. Talk about "strawman arguments." That's not
what I said and you know it. Last time:

Trying to edit mpeg for anything other than your Survivor DVD compilations
is pointless because:

1. Transitions, corrections, compositing, etc. require re-rendering which
will SERIOUSLY impact the quality of the video in mpeg.

2. REAL editing programs either don't handle mpeg, or don't handle it well.
REAL editing programs don't do frame-accurate mpeg editing, except, perhaps,
a specialized few. I wouldn't know because REAL editors don't work in mpeg.

No. Again you're wrong about what I said.

Mpeg can be edited -- just not well.

What new MPEG frame accurate editors? VideoReDo? Oh, please.

Wrong again.

You know what? This is really pointless. You have a bug up your butt for
some reason and are more interested in making this about personalities than
about video.

Hey, you're the one who recommended DVD Shrink which, of course, requires
retranscoding and degradation of the video.

Then you and many others have very, very, very little experience.
Commercial-quality mpeg is the result of multiple-pass transcoding. The
best realtime hardware transcoders (which are used by the satellite and
cable providers) don't produce output comparable to a commercial DVD.

See, this is what I'm talking about. You're speaking from the perspective
of someone copying television shows and removing commercials. THAT is not
video editing. I don't doubt that you're satisfied with the quality of
video that you get from programs like VideoReDo and DVD Shrink.

Spend a little time over at rec.video.production and www.dvinfo.net where
the professionals gather and run your theories by them.

I never said that MPEG can't be edited. I said it can't be edited easily or
well.

Your mischaracterization of what I said may not mean it, but the reality of
video capture certainly does.

 

2. REAL editing programs either don't handle mpeg, or don't handle it
well.

mpeg.

And that's a fact. Even using the Main Concept plugin for PP, mpeg editing
is clunky at best.

For simple slicing and dicing of material already encoded into mpeg format,
V-Redo, Womble, etc are fine. And they can, in fact, do frame accurate
edits. But there's no way anyone can say that mpeg'd material is the
preferred format for video editing. Unfortunately, people using the Hauppage
line of products have no choice because _all_ of their captures are hardware
mpeg encoded. Since I use an ATI card, I prefer to capture to Huffyuv avi's,
edit in Premier and encode with Procoder. And I'll match my two-pass VBR
encodes against a five year old hardware encoding algorithm anyday of the
week.

The Hauppage/Video-Redo/Vegas crowd in here can be downright arrogant
sometimes.

Easiest and fastest usually do not add up to quality.

 

And No One ever said anything different, Except "PTravel" saying
that you can't do frame accurate edits.

Just who is being arrogant here? No one has said the way you
capture, edit, encode, author, or burn DVDs produces bad results.
No one on this side of the issue is saying you can't edit AVI.
We aren't saying the results that you see every day are impossible,
or the result of some lower standard of "Quality", as your side
often does.

I myself have often recommended your old, slow, method of
capture and DVD production to those who are working with
unedited material.

With the process that I and many others use, I save up some
of the video entertainment that arrives at my house, into DVD
format. When I playback that archived video it looks and
sounds in no way inferior or different from the way it looked
when it first arrived. Your side keeps saying this is a lie, that
it can't be, that there must be some "Degradation" that I am
unwilling to acknowledge. Isn't That arrogant?

It is your side that is saying, that your way is the Only REAL
Way. You are in effect saying that since we are not trying to
make movies from scratch, our faster and easier methods
don't count and must be inferior to your methods, just
because they aren't applicable to movie making. How
arrogant is that?

If "PTravel" had said you can't do a crossfade transition in
MPEG, I would have let it go. Those with the objective of
saving up professionally edited material, never need to do
such editing. But his false claims that frame accurate editing
of MPEG would require the introduction of horrendous
degradation, because (first the whole video then even if a
few frames) must be "re-transcoded", do impact on our
efforts.

All in all it appears to me that you old guard, the only way
to get good results is my way crowd, are the ones with
your noes in the air.

We will never agree, but the issue is here in this thread
for all to see, and at least one false claim hasn't gone
unanswered.

Luck;
Ken

 

And you corrected me by pointing out that there are some products around
that could. Of course, those products aren't NLEs, and none of the
consumer/prosumer NLEs can, but I let that go.

Your problem is your insistence that kludging together re-transcoded mpeg
results in video quality indistinguishable from edited AVI transcoded by
decent software to mpeg.

No. You just said that producing DVDs in this fashion results in negligible
improvement in quality which is, of course, completely and totally wrong.

And no one has said you can't edit mpeg. What you cannot do is edit mpeg,
particularly in the way you've described, and get high quality results.

As I've said repeatedly, you must have a very different standard for
"quality" video. What are you watching it on, anyway? Honestly, I simply
don't believe you when you say you don't see any difference between you're
twice-transcoded single-pass encoded on the fly mpeg and a properly produced
DVD.

So I was right -- you make Survivor DVD compilations. What you don't seem
to understand is that whatever it is that arrives at your house is either
OTA broadcast, cable or satellite. None of these delivery vehicles provide
video of even DVD quality, and certainly not approaching that of DV-25.

Then you've probably got cable and a not particularly good television,
because applying DVD Shrink to mpeg results in a demonstrable reduction in
quality.

It would be if it were wrong. However, what's arrogant is your insistence
that kludged-together hobby software is a perfectly good substitute for a
good NLE and transcoder. YOU know that it's not. The problem is that
newbies who ask for information may believe you.

Nope. It's the only way to get high quality video results. Your method
won't even come close. Does it work? Sure. Does it work as well as a good
NLE and transcoder working with DV-code AVI? Absolutely not.

No. We are saying that, if you want to make DVD compilations of OTA
broadcasts or cable, and don't care whether you get the maximum quality of
your video, than your technique is fine. It is not a substitute, however,
for doing it right.

I did say that. You just ignored it.

And that's just crap. You said VideoReDo does frame-accurate cuts and I
said fine. You, however, persisted in simply ignoring that:

1. Real-time single-pass mpeg encoding doesn't begin to approach the
quality of a good software-based 2-pass transcoder.

2. VideoReDo isn't an NLE.

3. MPEG video can't be easily color or gamma corrected, used with
transitions, composited, or any of the the other things that constitute
"editing".

4. DVD Shrink re-compresses, i.e. re-transcoded mpeg, with a resultant loss
of quality.

The only way to get good results editing video is the way that I've
described -- your way won't work. Your way is good for one thing: Survivor
compilations. If you had said THAT, I wouldn't have bothered following up
to the rest of your nonsense.

 

PTRAVEL, Ken is kid, I think. Pretty much of a novice. I think you're
wasting your time.

Ken, Any compressed image is NOT ever going to be better than its
source.

I highly DOUBT that video-re-do does frame accurate MPEG cuts. More
probably just cuts at I Frames. Its a toy anyway.. who cares. ?

 

You express the attitude, reverence for the truth, and that amazing
ability to evaluate things you have obviously never tried, that marks
the rest of your crowd. What I actually said is there for all to read.

Luck;
Ken

P.S. I'm one yoar younger that Grandma:

One evening, a grandson was talking to his grandmother about current events.
The grandson asked his grandmother what she thought about the shootings at
schools, the computer age, and just things in general.



The Grandma replied, "Well, let me think a minute, I was born before
television, penicillin, polio shots, frozen foods, Xerox, contact lenses,
Frisbees and the pill.



There were no credit cards, laser beams or ballpoint pens. Man had not
invented pantyhose, air conditioners, dishwashers, clothes dryers, and the
clothes were hung out to dry in the fresh air and man had yet to walk on the
moon.



Your Grandfather and I got married first and then lived together. Every
family had a father and a mother. Until I was 25, I called every man older
than I, "Sir"- - and after I turned 25, I still called policemen and every
man with a title, "Sir".



We were before gay-rights, computer dating, dual careers, day-care centers,
and group therapy. The Ten Commandments, good judgment, and common sense
governed our lives



We were taught to know the difference between right and wrong and to stand
up and take responsibility for our actions.



Serving your country was a privilege; living in this country was a bigger
privilege.



We thought fast food was what people ate during Lent. Having a meaningful
relationship meant getting along with your cousins.



Draft dodgers were people who closed their front doors when the evening
breeze started. Time-sharing meant time the family spent together in the
evenings and weekends - not purchasing condominiums.



We never heard of FM radios, tape decks, CDs, electric typewr iters, yogurt,
or guys wearing earrings. We listened to the Big Bands, Jack Benny, and the
President's speeches on our radios. And I don't ever remember any kid
blowing his brains out listening to Tommy Dorsey.



If you saw anything with 'Made in Japan' on it, it was junk.



The term 'making out' referred to how you did on your school exam. Pizza
Hut, McDonald's, and instant coffee were unheard of. We had 5&10-cent stores
where you could actually buy things for 5 and 10 cents. Ice-c! ream cones,
phone calls, rides on a streetcar, and a Pepsi were all a nickel.
And if you didn't want to splurge, you could spend your nickel on enough
stamps to mail one letter and two postcards. You could buy a new Chevy Coupe
for $600, but who could afford one? Too bad because, gas was 11 cents a
gallon.



In my day, "grass" was mowed, "coke" was a cold drink, "pot" was something
your mother cooked in, and "rock music" was your grandmother's lullaby.



"Aids" were helpers in the Principal's office, "chip" meant a piece of wood,
"hardware" was found in a hardw are store and software" wasn't even a word.



And we were the last generation to actually believe that a lady needed a
husband to have a baby. No wonder people call us "old and confused" and say
there is a generation gap.!



And how old do you think grandma is???



Read on to see -- pretty scary if you think about it and pretty sad at the
same time. This is something to think about. How time has changed...





Grandma is 58
(born 1946)

 

Given the GOP structure of VBR or CBR MPEG-2, whether the video is captured
to or encoded to that format, frame accurate editing is not - and I repeat
NOT possible. The program and the manufacturer that claims this frame
accurate editing is in short stretching the truth, and anyone who believes
it is easily fooled.

MPEG-2 when encoded, whether using VBR or CBR to a GOP with the IBBP frame
structure has full frames only about 1/3 of the time. Every frame in between
I frames is made up of information from subsequent and previous frames in
the stream. These "so-called" frame accurate MPEG-2 editing packages may
have the ability to take the information from the BBP frames to create a
pseudo "full frame" but that is still NOT frame accurate. The only way these
programs could come near to frame accurate editing would be to negate the
BBP frames near the cut in favor of the nearest I frame.

If this is not the case then the whole structure of the GOP would be trashed
by the cut. The GOP cannot have a structure of IIBPIPBIPPI, which is what
you get if true frame accurate editing is allowed. It's purely simple. The
MPEG GOP standard is 15 and is divisable by 3. The explanation below (which
is information from the Motion Pictures Engineers Group proper) shows that
it is impossible to do frame accurate editing on MPEG files with the IBBP
frame structure. ONLY I frame MPEG can be editing in frame accurate mode.
The GOP (Group of Pictures) Pattern is the arrangement of frames in an MPEG
video stream. The GOP consists of a variable number of I, B, and P frames.
There is not a standard way of formatting the groups, and different
manufacturers may choose quite varied GOP arrangement. One may use I, B, and
P frames in specific groups arranged as IBBP, while another may use an
arrangement which incorporates on I and P frames, and others I-frames only.

The GOP interval and size will determine the pattern of the frame types in
the MPEG stream. It is common practice to keep the number of pictures in the
GOP at a number that is evenly divisible by the interval. So, if the
interval is 3 then the GOP has to be divisible by 3. MPEG has a standard GOP
of 15 with the interval of 3.

Using an IBBP arrangement will yield a low data rate video with the best
quality. When you use I-frames only the quality is much higher as is the
data rate also higher. When you want to keep bandwidth to a minimum and
maintain quality it is good to use the IBBP arrangement as the GOP.

 

----From here to the next break is this poster demonstrating that -he-
doesn't
know how it can be done, therefore he can't accept that anyone else has done
it. I am sure there would be problems if the Frame Accurate MPEG Editors
actually worked the way this poster seems to think that they
do. ------------

This may come as a shock, but the standard only sets the upper limit
for GOP size (and current practice has extended that dramatically, check out
MPEG4 and there is an extended DVD (xDVD) that plays in most DVD
players, See the TMPGEnc 3.0 XPress help files). All the cuts I've seen
with VideoReDo result in two smaller GOP, each with a normal structure.
By the way the PAL standard has the 15 frame per GOP limit with NTSC
it's 18. There is no restriction, in the standards or in practice, to any
smaller
GOP all the way down to the I Frame only variety.

Let Me speculate for a minute, you have never used VideoReDo or any
other MPEG editor that "claims" to do frame accurate cuts; have you?


---- All the below is mostly quite accurate, if irrelevent. --------------

----------------------------

If it were true that : "The program and the manufacturer that claims
this frame accurate editing is in short stretching the truth, and anyone who
believes it is easily fooled." Then the fact that the MPEG that I cut that
way, starts or ends on the exact frame that I want, must be a really good
illusion. If we are back to "it's not a REAL frame accurate cut" because
it wasn't done your way, I give up.

Luck;
Ken

 

Ken, take a deep breath and look at that statement in a mirror.
Your worldview appears to be remarkably limited to what you
see on your OTA/cable/sattelite. Go down to a TV station and
see what real video looks like before it has the life compressed
out of it. It will blow your mind and completely change your
attitude.

In a way you are correct that MPEG compression and even editing
of such is a "new generation" and is rejected summarily by many
of us. But what you don't seem to realize is what a remarkable
step DOWN in quality and functionality MPEG encoding and
"editing" is. Perhaps it is excusable if you have only MPEG
compressed video to work with.

TV shows go through a long production and distribution chain
which includes MPEG compression if you are watching cable
or sattelite. We attempt to store and edit video at the highest
quality available and hold off MPEG compression until the
last possible moment in order to preserve enough quality that
the video will be good enough for you to consider it worth
archiving and "editing" when it finally gets to your house.

 

TV shows go through a long production and distribution chain

When I was making the transition from C-Band analog satellite TV reception
to the mpeg'd format of Dish Network, I was surprised that I generally could
not tell the difference in "quality" on my 32" Sony XBR. Sure, you could
tell on certain situations where there's fast motion and low light levels,
but for casual viewing they were both virtually identical, to me... on my
system.

Using the method I've previously described, I burned a DVD of material I had
captured and played it back on my system. It was easily watchable. I then
went over to my brother's house and viewed it on his 55" monster. It was
almost unwatchable. Now I had previously done this with my C-band system and
the shows were virtually indistinguishable from his normal OTA picture. So I
had to conclude that even though _I_ couldn't easily discern the apparent
"quality" differences between the two transmission systems on _my_ system
didn't mean it wouldn't be apparent on another system.

So what's the point of this long-winded post? "Quality" is determined not
only by the "quality" of the original signal but in the manner in which it
is displayed as well.

 

Well the fact still remains that, "life compressed out of it" or
not, that is what most of us have for source video. It is the
"Quality" of video that we non-professionals have access to,
at home. It is to this that we must make our comparisons.

I am also curious how capturing the analog signals of such
inferior sources in the less compressed AVI DV format will
allow you to, in the end, produce a DVD that is "better"
than this source. Just how much "better" than this source
can come from using your techniques? I know that using a
multi-pass encoding technique you can do a better job
compressing the source into a smaller MPEG, but the
source (the analog signal that you encoded as AVI)
contained only so much image data.

If my process gets enough of the analog input's image
data properly stored, so that on playback, it is virtually
identical to how it looks when the analog signal is
played directly; How much more of the analog signal's
image can your process retain?

I am very aware of the effects of compression on video. I
would remind you though, that to make a DVD you must
compress the video. With the Real-time capture to DVD
compliant MPEG approach there is only the one encoding,
one compression. With your approach isn't it that the
analog signal is first encoded as AVI, image altering editing
applied, and then transcoded to DVD compliant MPEG?

As to the issue of the "functionality" of MPEG Editing;
if it provides all that is needed, in this case Frame accurate
cuts, then it -is- functional. Keeping in mind the nature
of the sources coming into the home and that most of
us are capturing, Professionally edited material with
commercials and other unwanted parts; the Only editing
required is to cut out some material.

Yes and I'm sure we all appreciate the Professional
Videographer's efforts, but that's part of the point, we are
working with their Output, not their Source.

If I can convert the analog sources that I have coming
into my home, into a DVD that, when played back looks
just the same as it looked when it arrived, then I've done
all that I want. Then the procedure I use will have done
all that I can ask of it. That it does this in Real-time, is a
much appreciated bonus.

If I were making a movie, I would certainly be using a
process similar to yours. But as long as I am just saving
up entertainment, that has been professionally produced,
and edited, the faster,easier and just as good as it looked
when I got it approach is, on balance, at least as good as
yours.

Luck;
Ken

 

Agreed, and interesting, but you used the same capture
procedures for both trials, right? It wasn't one was properly
captured directly to MPEG and the other properly captured
to AVI first then encoded to MPEG.

One wasn't MPEG that had been Cut when it was AVI and
the other cut with a frame accurate MPEG editor?

I have a "SharpVision" Projector that lets me view any of
my video on a 100" screen, in a darkened room. Under
that screen I have a 27" S-Video capable monitor/TV. I
can route the same S-Video signal to both. When I do so
with any analog (S-Video) source the image is a little softer
on the projection screen than it is on the monitor and the
blacks could be a little blacker. When I playback one of
my "homemade" DVDs I see the same effects looking from
the monitor to the projection screen. I haven't seen any new
effects, as you describe, looking at the 100" screen.

Luck;
Ken

 

Agreed, and interesting, but you used the same capture

Yes, the only variable was one source was C-Band analog and the other was
mpeg'd.

As I said before I don't capture directly to mpeg, I capture to Huffyuv avi
and then encode with Procoder with 2 pass VBR. I don't believe that a 5
year old encoding algorythm hard coded into a chip is going to give the same
"quality" as a dual pass VBR encode with a current, top quality s/w encoder.
How the footage was edited has absolutely no bearing on how it was
displayed.

The _source_ of what you are viewing is the same, that's why you see the
same "effects" on both displays. Now if you're trying to say they both look
exactly the same to you and you can't tell the difference in pixelization
between a 27" monitor and a 100" projection, then I have to question your
ability to perceive video.

 

But they weren't even the same source, how's that for a
variable? The only point I could see to that paragraph was
that Dish Network's output can't stand up to being enlarged,
but your C-Band can. In both cases you are taking the
analog signal from the receiver/box and producing the video
using only your technique. It only provides information on
some effect that you notice using your technique. There is no
comparison of the techniques under discussion.

But how it was edited is a factor in contention, in this thread.

To make a Valid comparison the source Must be the same.
What I did and am now saying is that whatever differences there
are between the monitor and the projection, they don't change or
increase when I play one of my DVDs. The same differences
appear when using the analog S-Video signal from my DirecTiVo
DVR as appear when using the S-Video output of my DVD
player as it plays one of my DVDs. No new or more grievous
artifacts appear.

Given that the "Quality" of the DirecTiVo S-Video output, I'm
not surprised that it can be Blown up without problems, this is
also some testament to the "Quality" of the Phillips SAA7114H
A/D and the Broadcom BCM7040 "KFir-II" Encoder Chips.
Combining your and my experience might suggest that the S-Video
output of the DirecTiVo DVR is closer to the analog output of your
C-Band Receiver than it is to that of a Dish Network Box. All
this is interesting, but has little to do with the theme of this thread.

Luck;
Ken