Re: Editing HV20 HDV Footage


"Smarty" <nobody@xxxxxxxxxx> wrote in message news:NHFuj.9637$eg3.1096@xxxxxxxxxxx
"David Ruether" <druether@xxxxxxxxxxxx> wrote in message news:47bafef2$0$17341$4c368faf@xxxxxxxxxxxxxxxxx

I just did two tests. In the first, I placed two HDV clips on the PE-4
timeline with a 1 second crossfade between and hit the "Enter" key
to cause render. Result: it took 5 seconds and then played the two
clips + crossfade.

I was not sure, but this seemed to be the test you were describing...
Since it was so fast (too fast to accurately measure the time...;-), I then
did the two tests below, identical ones for each program, using footage
captured by the program used for checking.

I then put two clips in Ulead 11+ and did what
was necessary to stretch a 45 second crossfade over the first clip
and follow it with another 45 second clip. Ulead took 3:05 to output
a file of the 1:30 timeline (on another attempt, it took far less - and
then I realized it was using renders it had previously made, so I
switched in fresh clips). I did the same with PE-4, which took a
"killer-long" 3:25 to output similar files from the 1:30 timeline.
Whew! S-o-o-o-o much longer!!! 8^)

This is very surprising to me. Why should Ulead take only 88 seconds on my older, slower single core machine versus more than
twice that amount of time on your dual core machine at 185 seconds (3 mins 5 seconds)? And then deliver over twice the speed on
Elements, cutting my time by more than half.....?? This really is puzzling.....

Please tell me again what you did (almost certainly I used a different
test, but it should not matter since 45 seconds of the timeline would
need to be rendered with each for the same type of 45 second transition,
and 45 seconds of each timeline should have been simply copied by the
programs), and my comparison would be fair.

I think I can live with this,
given the greater "security" with PE-4, and the, for me, the much
superior output. Settings for both were defaults for HDV NTSC
(though the normal HDV frame rates/formats are named confusingly
differently in the two programs). My computer is a home-brew
Intel dual-core 2.4GHz (not the fancier version of this CPU, though)
with 2 gigs of RAM (not the fastest version), and "plain-vanilla" old
hard drives, run as singles and not in RAID.

Not sure what you mean by greater "security" with PE4?

So far, no oddities and unexpected problems with the footage or with
editing HDV in PE-4, unlike Ulead 11+.

And I assume the "much superior output" refers to the "wubble" / "hop" problem you are reporting.

Yes, plus my need to recompress everything in Ulead to avoid
these problems. The resultant footage from Ulead from doing this
varies from somewhat softened to terrible (quite soft, and even
looking doubled in places). Why should I bother with this
program when PE-4 works better for me, appears to be more
dependable, and is about equally cheap? It may yet be that some
quirk happened in Ulead that may not happen again, but I don't
want to take the chance, and I'm really (mostly...;-) liking PE-4
now. PE-4 does have a couple of operational quirks that drive
me up the wall - but at least they don't affect the output quality.

I recommend (and should have made clear) things like a slow
pan/zoom with much detail across the frame (stones on beach,
trees, weeds, etc.) and also a person walking slowly. These will
easily show what I've found, if it is there at all for you (it would be
nice if it were not...;-). For me, the above is how I often shoot.
If I used nothing but cuts instead of slow crossfades and mask
based transitions, the busyness of more "violent" transitions may
well conceal the problem (as would busy or still-action footage).

When I go back and do another round of tests with the real version of Elemenets, I will start with different clips with less
motion per your latest suggestion above.

My complaints with Ulead would probably never be seen in most
music videos, staticly framed shots, or "run-and-gun" news coverage...;-)

I missed that one - but it does a cute trick when you start a file
save - the screen dims...;-) I like the large preview window (you
can lift the top of the timeline area to just close in on the 16:9 area,
giving more space for the timelines). The (mostly) real time previews
are nice, too, and they are sharper than Ulead's. I also like the lack
of need for proxy files, which take a long time to render at startup in
Ulead and which can be easily lost. There are also more "possibilities"
than in Ulead, but you need to look around for them. Some controls
hide with grey on the darker grey background, and turn white only
when you use them. I did find keyframing harder in PE-4 than Ulead,
and adding a transition to a keyframed clip removes the effect
of the keyframing at that end of the clip. The overall effects that can
be applied to clips are nice, and give MUCH control over "zillions"
of image characteristerics, all controllable with sliders or automatically, as in Photoshop (but with the same good/bad averages
for success with the auto adjustments...;-).

I too will give Elements some high marks for a very attractive interface. I have yet to use it enough to comment intelligently on
how its' feature set compares to Ulead's.

We used to run "races" of computers with 10 seconds of Mini-DV
with a + cc10 red shift to see how long each took (the "fast" Macs
with FCP took F O R E V E R, and my antique P-III 700MHz did it
"lickety-split"...;-).

I have a love/hate relationship with FCP and the Mac. I dumped my latest 8 core Xeon MacPro after a short ownership, having also
owned a prior 4 core Quad Xeon MacPro, based primarily on speed or lack thereof. I am presently eyeing the new Penryn-based MacPro
8 core which (finally) does the memory I/O properly and benchmarks a lot faster than my 1 year old 8 core did.

Ah, hardware...! I try to avoid the computer treadmill, generally
upgrading only every few years, and only when useful (and usually
by dropping inexpensive parts into my old box). I've sold most of
my camcorders, though the HV20 is FUN! I'm still an unrecovered
still camera nut, though I have yet to buy a DSLR body to fit my
shelves full of lenses. The audiomania is quiet for now, though....;-)
I like to "run cheap" in all my manias, if possible...;-)

Richard Crowleys comments regarding recompression I believe to be absolutely correct, so it may be something as simple as the
trial version with its' Adobe logo adding a big time penalty.

I doubt this, since as I recall, the logo is added only to rendered footage, which would need to be rendered anyway...

I disagree. It should be entirely possible for nearly 90 of the 92 seconds in my test movie to be produced without any
recompression whatsoever if an Adobe logo were not being added for the trial version HDV output file, since the renderer only
needs to form new frames and then new GOPs immediately before, during, and immediately after the crossfade. The rest of the video
should be untouched. Since Adobe chose to watermark their clips in the trial version, it is my opinion, and Richard's, that
Elements thus is forced to make new, watermarked frames for the entire video length of 92 seconds, rather than just the 2 second
crossfade interval, thereby adding a lot of extra rendering time. Why do you assume that the majority of the video content would
"have to be rendered anyway"?

Smarty

No - the reverse. I assumed the "A" appeared only on changed footage,
not all footage, but I barely tried out the trial version before the "real"
one arrived, so I could easily have been wrong about that. If I was
right, then the trial version would not recompress the whole timeline
segment, but if I was not right, then it would and that would explain why
rendering was so much slower than Ulead. But it seems unlikely that
Adobe would offer a trial that would be that difficult to use (they would
just want to mess up your transitions and effects, but give you normal
rendering speed or they would never sell the "real" copy after such a bad
experience!). I'm too lazy to fire up my friend's laptop with the CS3 trial
on it to see who is right, though...;-)
--DR
--DR


.

Relevant Pages

  • Re: Editing HV20 HDV Footage
    ... The entire timeline of 92 seconds is being ... rendered in both Ulead and Elements, and there is not other content, merely two 45 second clips separated by a single crossfade. ... render to an HDV output file and let's see what time you get. ...
    (rec.video.production)
  • Re: Rendering the timeline!!!!
    ... > following filters are used for the rendering of MPEG-1 files. ... is the average filesize or time duration of the clips that you are using? ... timeline really include an AVI Decompressor AFTER the MPEG Decoder? ...
    (microsoft.public.win32.programmer.directx.video)
  • Re: Editing HV20 HDV Footage
    ... have made a ton of HD DVDs with Ulead using HDV content and just have NOT ... My chief complaint from the start is that Adobe Premiere just ... sorts of HDV rendering time trials are published and updated all the time, ... Thanks for the report, though I think it may be flawed (though I think ...
    (rec.video.production)
  • Re: Premiere Pro problem
    ... marked off on the timeline for you. ... (In the monitor window in premier, ... You select the clips you want from project sources, ... so in about 3 clicks or so (after selection of clips) you ...
    (rec.video.desktop)
  • Re: combining clips
    ... and sell books. ... B and C on the timeline in that order: ... Movie Maker 2 - www.papajohn.org ... so now I have several clips that I wish to combine. ...
    (microsoft.public.windowsxp.moviemaker)