I have encountered (or I should say discovered) a rather interesting
problem with my Honestech TVR 2.0 and K-World USB 2.0 capture device
setup. However, I believe that the problem is actually due to the
Honestech software rather than the capture device. I cannot truly
determine this, however, because the K-World capture device really
isn't compatible with any other software. Perhaps someone can identify
the actual problem based upon what is going on behind the scenes...

So, whenever I record with this setup combo, I end up with a
decent-quality MPEG2 recording. The video appears normal, and the
sound appears normal on my no-frills $15 multimedia stereo speakers
attached to my PC (running XP).

However, after making the (proper) conversion from MPEG2 file to DVD, I
test the results by playing it on my Philips DVP-642 DVD player and
Sony Trinitron television. When I do this, I finally encounter the
flaw in the audio recording process.

Let me go into detail about the flaw. The flaw presents itself as
garbled, robotic-like sounds and other defects which sound like
compression artifacts that occur during audio output. The defects are
not always present in the audio, but it is present in audio that meets
a certain specification that I cannot properly analyze. In other
words, it is not constantly present, but it is commonplace, and shows
up in many of the audio samples encountered in my captures.

It took me a while to figure this out, but after fooling around with
the sound presets and other sound options on my DVD player with no
success, I started fooling around with the television's sound options.

I come to find out that the problem only presents itself when the
television is in SRS mode, which is a simulated surround-sound audio
mode. When the television is set to output audio in standard, plain
stereo mode, the problem isn't noticable in the least. I know that the
SRS mode doesn't present a problem with most of the audio out there
because I had used this setting for years without changing it. I had
never encountered this problem before. From my experience, normal
MPEG2 and AC3 audio does not ever cause this problem with SRS.

So, the audio from Honetech's TVR 2.0 might sound normal using
no-frills stereo, but after audio processing, or at least the audio
processing used in Sony's SRS feature, a distinct flaw reveals itself.

As a side note, I noticed this garbled, robotic sound problem only once
elsewhere, and that is on the Speed channel. The problem was present
to a lesser degree during certain vocal interview segments on their
talk shows that were obviously recorded digitally that I could hear
while my Sony Trinitron television was in SRS mode.

Can anyone explain this problem with to any significant degree? Can
this problem be removed from the recorded audio using some other form
of audio processing?

This situation has obviously been encountered by more than one
developer of digital audio solutions. And please don't criticize
Sony's SRS audio processing. Like I said before, most audio works just
fine in this audio processing mode; so much so that I really do
consider audio that doesn't work properly with it to be flawed. And
justly so.

So, any thoughts? Thanks.

--

I should also add that this problem also presents itself in Honestech
VHS to DVD 2.0, and I doubt that the problem has been fixed in the new
VHS to DVD 3.0 version.

--

I was just recently viewing some old archives, and encountered it yet
again in some releases from the now-defunct SMR Group. The SMR Group
released their own SMR Codec and a few video tools that are still
archived on the Web.

One of SMR's past DivX releases of the annual children's classic,
"Rudolph, the Red Nosed Reindeer" exhibited this very same problem in
Sony's SRS audio processing mode. This piece may first have been
encoded using the SMR Codec, and then transcoded to DivX, but I cannot
determine if this is the case. I suspect that many of their other
releases may share the same characteristic.

Yes, digitally recorded audio should be robust enough to handle any
type of audio processing not intended to produce negative effects.
Unfortunately, this is not the case with Honestech's current line of
products. It is a real shame, considering that they are the only
developer who offers software that supports capture devices without
built-in encoding hardware on modest systems (~1.0 GHz). Their
software is otherwise FAST! and efficient.

--

On 2007-01-29 00:52:26 -0500, Kevin O'Mara <unknown (AT) unknown (DOT) com> said:

And by mentioning the SMR Codec, I was implying that that their codec
may have been at fault in the video creation process chain. It may be
the culprit behind the faulty audio in their releases. This was not
considered a professional grade codec, even in its heyday. I believe
that its homebrew / hacked status might have had an influence over the
quality of its encodes. They, perhaps, were not as knowledgeable as
someone like DSPguru of BeSweet fame, for instance.

--