LipTracker™
Frequently Asked Questions
What makes LipTracker™ different
from all other types of lip sync analyzers?
How are the LipTracker™ audio offset measurements
used to correct the lip sync error?
Does LipTracker™ affect the program source material?
How long does it take to measure the lip sync error?
Does LipTracker™ work in any language?
Is a special test signal required?
Does LipTracker™ work with any kind of programming?
How is LipTracker™ better than an operator watching the program
for lip sync errors and making adjustments?
How does LipTracker™ maintain optimum performance even though face
sizes may change during a program?
What happens with LipTracker™ when the face is moving around on
the screen?
Does an operator need to interpret the LipTracker™ results?
Does LipTracker™ interface to video and audio delays from other
manufacturers for automatic correction?
How can the SD version of LipTracker™ analyze HD program material?
Can LipTracker™ analyze the lip sync of an MPEG encoded stream?
Does LipTracker™ accept a Dolby encoded audio input?
|
What makes LipTracker™ different from
all other types of lip sync analyzers? |
| |
| LipTracker™ is the only product on the market
that can analyze and measure A/V synchronization from program material
at any point in the production or distribution channel. Products that
use any form of signal marking technology have two significant limitations.
First, they require known "good" A/V timing at every program
source. Second they require a watermark or other mark embedder at
every source. Watermarking technology then attempts to preserve the
A/V timing downstream using the embedded mark in the video. LipTracker™
has neither of these limitations because it measures A/V offsets directly
from the program material. |
|
|
|
How are the LipTracker™ audio offset measurements used to correct
the lip sync error? |
| |
| The same way it is done now - by manually adjusting
an existing audio delay in the facility. The difficult task is not
adjusting the audio delay but determining the magnitude of the error.
LipTracker™ improves operator productivity by supplementing
subjective and time consuming analysis of lip sync with rapid objective
results measured in real time from the program material. LipTracker™’s
rapid and accurate results allow the operator to quickly decide when
and how to correct for lip sync errors. In addition, all results are
logged for future verification or analysis. The bottom line - LipTracker™
saves time, improves overall performance, increases operator efficiency
and most important, ensures consistent accurate lip sync correction. |
|
|
| Does
LipTracker™ affect the program source material? |
| |
| No. LipTracker™ is completely non-invasive. No
code or watermarks of any kind are added to the program material and
no pre-processing of any kind is required. Any program from live broadcasts
to those that were recorded years ago can be tested in service without
any special encoding. |
|
|
| How
long does it take to measure the lip sync error? |
| |
| LipTracker™ looks for specific sounds and mouth
shapes in the program material so the time to make a measurement is
content dependent. The first result is often generated as soon as
4 seconds after a face is detected and then is updated every 2 seconds. |
|
|
| Does
LipTracker™ work in any language? |
| |
| Yes. The unique method of comparing sounds and mouth
shapes allows measurements to be made with all common languages. |
|
|
| Is
a special test signal required? |
| |
| No. LipTracker™ is designed for on-air use with
live program material. Of course it can also be used for offline testing
with the same type of material. |
|
|
| Does
LipTracker™ work with any kind of programming? |
| |
| LipTracker™ is most effective with material commonly
used in news programming, sporting events, talk shows, infomercials
and other areas where the elimination of lip sync errors is most critical.
LipTracker™ is also a beneficial quality assurance tool for
other types of programming. |
|
|
|
How is LipTracker™ better than an operator
watching the program for lip sync errors and making adjustments? |
| |
LipTracker™ significantly improves operator efficiency.
LipTracker™ eliminates the subjective and time consuming analysis
by a human operator. LipTracker™’s numeric and graphic
displays allow the operator to quickly decide if and how any given
lip sync error should be corrected. LipTracker™ also creates
a log of lip sync errors providing a “proof of performance”
report in the event that questions arise as to where an error originated.
Tests have shown that most people are not consciously aware of lip
sync errors if the audio offset is between 20 milliseconds early and
90 milliseconds late. However, these “small” lip sync
errors cannot be ignored. Research has shown that viewers can become
annoyed to the point of changing channels due to subconscious irritation
caused by these “small” lip sync errors. Even for larger
errors, people generally have difficulty deciding how much error there
is, especially in the pressure of an on-air environment. LipTracker™
on the other hand generates scientifically accurate and objective
results and has infinite patience to analyze the program material. |
|
|
| How
does LipTracker™ maintain optimum performance even though face
sizes may change during a program? |
| |
| LipTracker™ will lock to faces that are approximately
¼ of the picture height (from the top of the forehead to the
chin) or larger. As you might expect, when the face size gets smaller,
lip sync errors become less obvious and therefore less of an issue.
In the absence of a suitable face to lock to, LipTracker™ continuously
searches the input video until it finds one. |
| |
|
|
|
What happens with LipTracker™ when the face is moving around
on the screen? |
| |
| LipTracker™ will stay locked onto a face with
normal head motion and camera pans and zooms. If the face turns to
profile view; looks down at a script etc. LipTracker™ continuously
monitors the video and immediately reacquires the face and resumes
analyzing when the lips return to view. |
|
|
| Does
an operator need to interpret the LipTracker™ results? |
| |
| Yes - in some situations. Consider this example - a
typical news segment where archive video of a politician giving a
speech is broadcast with an off screen news anchor’s commentary.
In a case like this, the mouth shapes in the video are not related
to the sounds in the audio. A human operator will interpret the LipTracker™
results appropriately in situations like this. |
|
|
| Does
LipTracker™ interface to video and audio delays from other manufacturers
for automatic correction? |
| |
| Unfortunately, there is no industry standard protocol
for controlling such devices. Preliminary work on an interface standard
has been started by a SMPTE working group. We are currently researching
this issue to determine the best approach to use. |
|
|
| How
can the SD version of LipTracker™ analyze HD program material? |
| |
| Until the HD version is available, an external HD downconverter
can be used. The downconverter will have a known video delay (check
the user manual for your downconverter) and this delay value can be
used as the LipTracker™ Measurement Offset parameter to ensure
that LipTracker™ operates in the center of its range. |
|
|
| Can
LipTracker™ analyze the lip sync of an MPEG encoded stream? |
| |
| Not while the stream is in MPEG form. The stream must
be decoded to SDI video (standard definition now – stay tuned
for availability in HD) and AES-3id audio to be analyzed. If the decoder
has a known audio/video offset or latency (check the user manual for
your decoder) this value can be used as the LipTracker Measurement
Offset parameter to ensure that LipTracker™ operates in the
center of its range. |
|
|
|
Does LipTracker™ accept a Dolby encoded audio input? |
| |
| No. An external Dolby decoder must be used. If the decoder
has a known audio processing delay or latency (check the user manual
for your decoder) this value can be used as the LipTracker™
Measurement Offset parameter to ensure that LipTracker™ operates
in the center of its range. |
|
|
