PapaJohn's Newsletter #18 - Movie
Maker 2 and Photo Story 2 - Sept 11, 2004
Exploring: Audio
- Enhancement, Generational Losses, Sync with Video,
Conversion.
There are many posts are about audio and video tracks
being out of sync. I ran into a sync issue the other day
while working on the 'Mustang Sally' video for
the reception page of the new wedding website -
www.Jill-Mark.PapaJohn.org - I
couldn't work it out, and had to finish the clip using the unenhanced audio
track.
That got me thinking about and exploring some of
the nooks and crannies of audio. I thought I'd share what I found with you
in this newsletter.... some things worked and others didn't. We can
learn from both.
The intro section will describe the sync issue that I ran
into on 'Mustang Sally....and the
mini-tutorial will explore some questions I explored as a
result of it, and as I put together the 'Wedding Party Arriving' clip, the
next one for the site. Here's a list of the assorted topics in the
min-tutorial section:
• Audio Wave Patterns
• Wave Pattern Differences - DV-AVI Versus
WMA
• MM1 Narration Files are One Channel
• WMA Generational
Testing
• Plus! Audio Converter
to Get an MP3 File
• Use TMPGEnc to
get a WAV file
• Use iTunes
to Enhance the Audio Track
• Does the Process
of Ripping a WAV File using TMPGEnc, enhancing it with
iTunes, and capturing back in MM2 as a WMA file result in any sync
issue?
Before getting into it more, a few notes about some things going
on...
Notices
• I find myself looking at what I'm currently working
on, and taking one aspect of it as a newsletter topic.... that works well for me
as it helps me explore the topic more fully, get some value-added for my current
project, and hopefully makes an interesting newsletter for
you.
• The Creative Zen is on its way back to Microsoft. It was a loaner
to test and write about.... fun while it lasted!!!! I moved the
associated blog page into the Distributing section of the website. Portable
Media Centers will be a convenient way for many to share their movies so it'll
stay on as a permanent page.
• On Tuesday I received an e-mail from Sliq MediaTechnologies. They
told me about 50 people who downloaded their WMSnoop utility
last weekend referenced the newsletter. They were appreciative,
and open to comments about the utility.... so pass any along.
• My son handed me one of those mini-DVD discs from a
Sony DCR-DVD200 camcorder that he's using, one of those that burn the video
directly onto a disc in the camcorder. He said he'd like to know how to edit the
video files in Movie Maker, and how to get them into the My Videos menu of
the Media Center software. It's a good topic as more and more people purchase
such camcorders and have the same questions. Now I have a disc to check the
options.
....on to the topic of the week
The Audio Issue(s) on 'Mustang
Sally'
Some Background Info
I took the wedding footage with two camcorders.... my
older Sony Hi8 TRV-615 which records in analog with FM quality stereo
sound, and a digital Sony TRV-80 with its digital recording and higher
quality audio.
For 'Mustang Sally' I needed to use the continuous audio track
from the digital tape, and mix the video from both
camcorders.... and the audio/video sync was critical for the entire
length of the video as there were lots of tight shots of the singer (he was
a guest at the wedding, not a member of the band - or he wouldn't have gotten as
much coverage in the video).
I dubbed the analog footage to the digital camcorder so I had both
tapes coming into the computer via firewire as DV-AVI files.
Because the sound track was important for this video, I thought
I'd rip it from the DV-AVI file, enhance it a bit in iTunes and bring it
back into MM2, something I've done before a number of times.
I previewed it in MM2 and captured it as a WAV
narration file in MM1 as it played.... I wanted a WAV file for iTunes.
I listened to most of iTune's 22 default equalizer
setting choices and, when I selected one I liked, I played the file in
iTunes as I captured it back into MM2 as a WMA narration file. I'd
done this before but not with such a sync-critical project.
Audio/Video Sync
During project editing I found myself wrestling with a
significant audio/video sync issue. I'd align it in
one place and find it off in another. I could only keep the alignment going for
a short period on any of the clips..... and some of the clips weren't
short enough to keep it in sync throughout.
When I put the enhanced WMA narration file over the audio of the
DV-AVI file in the project, I saw a problem..... there was a steady
drift out of sync between the two. By the end of the video the new
enhanced WMA sound track was 2-1/2 seconds (about 75 frames)
longer than the audio track of the DV-AVI file.
I ended up dropping the enhanced audio track completely, and
using the un-enhanced track from the digital camcorder tape. But, while
doing it, I decided to return to the topic and study the issue(s)
further to learn what happened, and what can I do about it the next
time.
Visual Patterns in MM2 to Aid
in Syncing
..... wave patterns of the DV-AVI and WMA files were
very hard to match up. I found myself tying to sync it mostly by listening to
both tracks playing together in the project preview.
Where did it go off in Duration? Why? And why are
the Wave Patterns so Different?
I didn't know. But the nature of the video was such that I needed great
syncing throughout... so I tossed aside the new enhanced audio
track and used the original unenhanced sound track to finish the
video.
That sets the stage for this tutorial, exploring some
audio items.
Mini-Tutorial: Exploring Some Audio
Items
Here are the
topics I explored and what I found:
Audio Wave Patterns
The differences in the visual wave patterns with DV-AVI
files, WAV and WMA files are significant enough to make it too
hard to study precise drifting causing of sync issues, and to use the
different types of files in sync-critical situations.
For easy syncing in a project, use a WMV video file with the
audio track ripped to a WMA file.... the duration of the files and the
visuals of the wave patterns are identical. Here's a file with the
patterns about 14 minutes into the project.
WMV and WMA Audio
Tracks
Another easy
way to work with syncing is to use two copies of the same DV-AVI file, one
copy for the video/audio, and another for the Audio/Music
track.
DV-AVI File - On Video and
Audio/Music Tracks
Harder to work with are a DV-AVI file and a copy of the
audio track ripped to a WMA file.
This WMA file was 00:00:08 less than the DV-AVI file
it was ripped from. That's only .08 seconds in 9 minutes... it equates to about
2 frames. I could live with the slightly shorter audio track but the
differences in wave patterns make it hard to sync video clips with the
audio track.
Ripped WMA and the
original DV-AVI
.... when playing them back individually, they sound
the same. The wave patterns just look different.
Wave Pattern Differences - DV-AVI
Versus WMA
The properties of the audio track of the DV-AVI file
I as studying are 1024 kbps bit rate, 2 channels, 32 kHz
sample rate with a sample size of 16.
The two properties that differ in the saved WMA file are
the bit rate of 175 kbps, and the sample rate of 44.1
kHz.
I had picked the highest quality option when saving.... does a
change in the quality setting vary the patterns? I'm looking for
visual wave patterns that will align with the original DV-AVI
file. The 160 kbps file looked the same as the 175 kbps one, and was
the exact same duration.
I worked down the quality settings list - 128 kbps was next....
and was the same.... so was 96, 64 and 32 kbps. I don't want to
go any lower in quality than that, not for wedding videos.
If the patterns didn't change with the bit rate, then maybe
it's because the DV-AVI file has a 32 kHz sample
rate versus the 44.1 of the WMA file..... I checked that by
creating a custom WMA profile that uses the 32 kHz sample rate that
aligns with the DV-AVI file. This one took a lot longer to render than the
others and didn't provide the answer... the differences in visual patterns
were still significant.
Custom Audio Profile at 32
KHz versus DV-AVI file
.... so using a ripped WMA file from a DV-AVI file
wouldn't work for my sync needs.
Wave Pattern Differences - DV-AVI
Versus WAV
WMA files are the usual for MM2, but MM1 captures narrations as
WAV files, which I know iTunes can play and enhance.
I compared a captured narration WAV file from MM1 against
the DV-AVI file. The visual wave patterns of the WAV file looked like the WMA
files saved with MM2.... but not like the patterns of the DV-AVI
file.
MM1 Narration Files are One
Channel
The WAV file properties from the MM1 capture were 352 kbps
bit rate, 1 channel, 22 kHz sample rate, 16 sample size. What
happened to the second channel? Did MM1 capture in mono, expecting voice
narration and not music?
I went back and did some more testing of the MM1 narration
capture and found all of the WAV files to be single channel..... so it must
assume voice narrations.... MM2 is capable of switching to higher level quality
when it detects music playing during a narration capture and I guess MM1
isn't.
I started with 2 channel stereo sound and wanted to maintain it,
so WAV files from MM1 won't get me there.
WMA Generational
Testing
Last year I tested generational losses when saving WMV and
DV-AVI files..... the WMV files lost the last frame of each generation, over and
over until they stabilized with one frame left... it didn't lose that one. I
wondered if saving a project as a WMA file in MM2 has similar losses.
And it does.
This DV-AVI file that the audio was ripped from was
9:05:48. When saving just the audio to a WMA file, each generation lost a bit.
Here's the lengths of each generation's WMA file:
1st generation save to WMA - 9:05:40
2nd generation - 9:05:30
3rd - 9:05:20
4th - 9:05:10
5th - 9:04:100
6th - 9:04:90
7th - 9:04:80
8th - 9:04:70
9th - 9:04:60
10th - 9:04:50
11th - 9:04:40 - for this one I added a still image to
the project so it would save as a WMV and let me compare the wave patterns
to see where the duration loss was occurring.
It was good to see that it wasn't a drifting situation. The loss
was clearly and cleanly at the end of the file, just as the WMV generational
losses were.
The wave patterns were still close enough to see that the
full second had come off the end of the file.
11th Generation - One Second
Less
Such WMA generational losses are not critical to syncing, or
for usual project work. It's more of an interesting technical item.
Plus! Audio Converter to Get an MP3
File
Because MM1 narration files wouldn't work because
they were mono, maybe converting the saved WMA file to an MP3 would
provide a path to the iTunes equalizer.
Plus! Audio
Converter
..... here's my error message. I don't have the
needed MP3 codec on my laptop, which needs to be purchased.
Plus! Audio Converter - Error
Message
If you have the Plus Pack and the codec, it's a path
you can take to get your audio track to other software for
enhancement.
Use TMPGEnc to get a WAV
file
....from the original DV-AVI..... I'll use the settings of PCM
format, 44,100 Khz, 16 bit, stereo, 172 kb/sec.
It took a couple minutes to get a 96 MB WAV file
that was 9:05:48 in duration.... the same as the DV-AVI file.... no
generational shrinkage. Now I have a file to play in iTunes and enhance
with the equalizer settings.
Use iTunes
to Enhance the Audio Track
Once I had a WAV file that iTunes can play, I used it's
equalizer with the presets and tweaked it manually. The 'Wedding Party Arriving'
clip had a hiss that I wanted to minimize.... I started with the
Classical style preset and tweaked the sliders. Moving the 4K slider
down removed the hiss.
iTunes
Equalizer
I'm ready to play it in iTunes and capture it in MM2 as a WMA
narration file, using my stereo mix option. It captured fine and I had
my enhanced audio track.... and because syncing wasn't a concern, the rest
of the project was normal and easy. The video is online with the enhanced
audio.
Does the Process of Ripping a
WAV File using TMPGEnc, enhancing it with iTunes, and capturing
it back into MM2 as a WMA file result in any sync issue?
I couldn't easily tell from looking at the wave patterns... so I
did a test. I added audio 'markers' at the beginning and end
of the file, and at each full minute point.... the marker is a
short sound effect clip from Sound Dogs.
I rendered the test clip as a DV-AVI
file.... extracted the audio as a WAV file using TMPGEnc (File > Output
to File > WAV).... then to iTunes to play it with the equalizer on and
capture back in MM2 as a WMA file. I compared the starting and ending audio
tracks.
The syncing of the markers at the start and at each full
minute point looked exactly like this final marker.... perfect sync
throughout.
Test File With
Markers
I can enhance my audio tracks in iTunes and work with them in
MM2 without the concern of introducing audio sync issues.... but the wave
patterns of the WMA file are different than those of the DV-AVI audio track....
not an issue for this project but something to consider for many.
Closing
I never did find what caused the
2-1/2 second sync issue in 'Mustang Sally', but I'll keep my eyes open for it to
happen again.
The rule of thumb to start with DV-AVI files and stay with that file type
through the editing phases doesn't seem to help when it comes to wanting to rip
off the audio track and enhance it..... maybe one of the readers will tell me
what audio file type aligns with the wave patterns I see in MM2 when
working with DV-AVI files.
It seems that, with the perfect alignment of WMV and WMA files, and the
equalizer of the Windows Media Player, it would be lots easier for me to switch
to using WMP instead of iTunes to enhance my audio tracks. This might be
especially so now that WMP10 has been released.
If I leave the enhancement to the end of the project and have confidence
there won't be any syncing issues associated with the process, I can edit
with DV-AVI files right up to the last minute.... that seems like a good path
to take the next time.
I hope that packing this newsletter with lots of individual items helps
you more than confuses you....
Have a great week...
PapaJohn