● Video Inputs: Two Composite, Two S-Video, Two
Component, Four HDMI, One RGBHV, One SDI
● Audio Inputs: Four HDMI, Two Coaxial Digital, Two
Optical Digital, One Analog Pair
● Video Outputs: One Component, One HDMI
● Audio Outputs: One Coaxial Digital, One Optical
● Dimensions: 2.2" H x 17" W x 10.4" D
● Weight: 8 Pounds
● MSRP: $2,995 USA
The VP50 is the immediate
successor to the
VP30 which I reviewed in February of 2006.
A lot has happened since
the VP30 was initially released. Anchor Bay Technologies (ABT) has
now made available the Precision De-interlacing daughterboard - which converts the
VP30 into one of the best de-interlacers that I've seen to date. Born
with the code name "Ducky Lucky" (or DL for short), the ABT102
daughterboard handled SDTV sources only.
The VP50 is the next evolutionary step beyond the VP30. Not only does it
contain the "Ducky Lucky" algorithms (yet another Dale Adams masterpiece
algorithm), it finally supports HDTV de-interlacing.
Just like the VP30 add-on, which supports crazy cadences, and maintains
a proper lock - so does the VP50. The VP50 behaves almost identically to
the VP30, with better de-interlacing of HDTV sources.
The VP50 also adds a few more toys, some of which I found to be
useful and some that require more work. The Gamma Correction feature is
a new addition that gives you the ability to change global gamma
settings (mostly useful for projectors) or individual color gamma values
(mostly useful for . . . uh . . I can't think of where this might be
Essentially, what's missing is the ability to custom control
various data points along the gamma correction graphs instead of having
them conform to the usual correction formulas. This would have allowed
us to use the unit to improve display calibration from good to perfect,
albeit with a reduced color range (i.e., this could cause increased
banding in some cases).
The company who pioneered this idea (Lumagen)
has been improving it throughout the years, and we've already seen
similar features coming from companies like Pixel Magic in their Crystalio II product.
The implementation here is mostly a proof of
concept, which shows that the unit is capable of such features, but that
the missing ingredient is still not there (PC software to customize the
graphs, or preferably on screen menus to control each point like with the Lumagen and Pixel Magic products).
ABT has pioneered the use
of adaptive audio inputs for their products. The VP50 is no
exception, and it contains a single analog audio input (I would have preferred at least one more), two optical
digital and one coaxial digital audio input, as well as
the ability to decode audio (standard formats only, not the HDMI 1.3
formats!) from the four HDMI inputs.
The audio inputs serve as both a
switch and a lipsync delay engine that adapts itself to the source as
necessary. I personally prefer to use a fixed delay and have all my
audio driven through my A/V receiver, but many people prefer a single
source that switches both video and audio for their system. While
"normal folk" would typically use their A/V receiver to do both - the
ABT way is to use their system for the task. The advantages are clear,
but the main disadvantage was recent findings that in some combinations
of sources and receivers, there were a few issues like audio
dropouts, noise bursts, and other problems. ABT has been hard on the heels
of these issues, and they hope that the upcoming firmware releases will mark
them a historical footnote.
With regards to video inputs, the VP50 is no slouch. It contains no less
than four HDMI inputs, more than enough for most people, two analog 5xBNC RGBHV
inputs, two composite, and two S-Video inputs.
Picture quality (PQ) on all of these inputs is exceptionally clear, and I have not seen a
particular problem with those inputs.
The VP50 is capable of transcoding
analog HD video to digital and vice versa (unless HDCP is active).
Of course, most people are going to purchase the VP50 for its HDTV
capabilties, but let's start with the SDTV processing. Using
the DVDO (Spears & Munsil) test disk shows how far the ABT algorithms
have taken us in recent years. I remember not too long ago when using
the video mode was my equivalent of water torture. With fixes
like the Faroudja DCDi algorithm, video was barely edible - but add
something a bit more complex like 2:2 cadence or bad edits and you're in
the twilight zone of video de-interlacing artifacts - jaggies, combing,
line twitter, judder, and practically anything else you can guess.
"Ducky Lucky" algorithms take the ABT102 and push it slightly further -
they further reduce the already tiny number of cases that I watch a
display and mention how bad the image looks. The image is simply
maximized to the best the source can be shown, without adding artifacts to
Next, I subjected the unit to a test round of torture. This is a
test DVD that I recorded, composed of the toughest 2:2 and badly edited
content available to me. It includes sequences with unmatching cadence
subtitles, overlay sequences, noisy sequences, badly edited video
montages, and so forth. The VP50 passed with almost a perfect record - I
found slightly fewer combings than with the VP30. This puts the VP50 as
the forerunner for SD de-interlacing.
Nothing is perfect, however, and the VP50 had a tough time with a Jackie
Chan movie that apparently has something strange that causes
most de-interlacers to think Film mode is enacted. The only way to get
the unit to stop combing was to put it in Game mode. The test sequence
was sent to ABT for review.
The frame rate conversion of the unit, in par with the capabilities of
the VP30, is exceptional, and even when I do 50->60 Hz
conversions (necessary for some displays that are incapable of
supporting native rate at 50 Hz), judder is kept to a minimum.
generous number of Genlocked (source locked - preventing lost or
repeated frames) rates for each source, more than enough. Handling of 24 fps is still
pending, to be be tested with more devices that provide it as an output (Sony BD
players) and receive it as an input (Sony projectors).
If we add HDTV to the mix, we have to talk about 1080i-to-1080p
processing. To test that, I primarily used a Pioneer PDP5000EX plasma
display. First, we set the display to native rate (dot
by dot) and used the VP50's built-in test patterns (ten of which I
designed) to ensure that perfect matching
was enacted. I did not need to change to custom timing, but this would be
the point where a more experienced user could adjust the VP50's output
timing to ensure everything is ok.
Once native rate mapping had been
tested and I calibrated the display, it was noticeable that the VP50 needed
a slightly elevated brightness setting on its analog and digital inputs.
This is strange and has been reported to the company. There should be no
reason why calibrated digital outputs would need anything other than 0
brightness on digital inputs.
I used several HD sources and content, including 1080i60, 1080i50 (2:2
cadence), and 720p (using only 3:2 60 Hz content). Extra details, brought
out by the new processing were easily seen. The picture was exceptionally
clear, and even 720p sources looked better with the VP50.
It should be
noted that the unit senses 3:2 on 720p sources, which allows it to
convert the signal to 48 Hz or 72 Hz without a problem. As the
Pioneer plasma supports
72 Hz, I tried this, but saw no real benefit (projector owners will
probably see a slightly better image with less judder).
My typical HD test sequences, composed of the typical mix of Sci fi
movies (The Fifth Element - Lilu's birth scene, several Star Wars sequences)
produced amazing results. Details were extra ordinarily crisp. However,
even when I fed the plasma interlaced sources, it did pretty good on its
own. I would expect that the reason behind this is that most 1080i
content has fairly good cadence locks. The area where the VP50 "should"
shine is 1080i video mode. However, I had a tough time finding anything
that was in 1080i video mode to test with.
The VP50 uses the familiar ABT user interface, which is
very friendly and easy to get used to. Custom input and out aspect
ratios should cover every conceivable user requirement. My main
annoyance with the custom input aspect ratio was the inability to assign
custom aspect ratios to 2-3 buttons on the original remote. Discrete
timing information posted on ABT's website will let anyone with a Pronto
(or Pronto discrete code compatible) remote to directly access aspect
ratios. I guess I'm old fashioned, but I prefer to learn the remote
control button features directly from a remote when possible.
Output profiles are very useful and can be assigned to be recalled on a
per-input and even per-format basis. This allows you to quickly switch
from your plasma to your projector (analog vs. digital outputs) and
could be a nice prelude to a multi-zone processor (no consumer multi-zone
processor has been announced so far, though!)
The Prep feature is one innovation I found very interesting. As a
processor owner, you are often told to configure your sources to output
interlaced signals rather than de-interlaced. After all, a processor can only improve on
something that hasn't already been "corrupted" by another processor.
However, there are quite a few situations where you can't really conform
to that requirement. Many people already have a terrible DVD player that
they want hooked up and improved upon. Several STBs only support
progressive HDTV output, even when the original source material is SDTV. De-interlacing in these cases
is usually pretty bad, and a processor
can't really do much to help.
Prep was basically designed to overcome this core issue. It senses these
conditions and re-interlaces the original data and detects (using similar
techniques that are used to detect cadence) the two field sources and
then de-interlaces them again, this time doing it properly.
How well does
this work? To be fully honest, the verdict is still out. I need to test
this with a lot more sources to be fully convinced that it works. I'm
still convinced, of course, that processors should get the source
material as closely as possible to the original content (i.e., SDTV
should be sent as interlaced format, 720p should be processed as such
and 1080i should be passed as is). My intial tests show that Prep does
improve upon typical de-interlacers found in most DVD players and does
make a difference when presented with progressive sources. How much of a
difference? Like I said, I am not convinced the difference is night and
day, I would say that it's far closer to optimal than what you would get
without a processor in the middle.
Something that ABT has not improved upon in quite a while is the scaling
algorithm. While it was sufficient for 720p
projectors, which needed 480i (SD) scaled to 720p, representing a 50% increase in the number of scan lines,
1080p displays show that the ABT scaling algorithm needs to be addressed
quickly. Feeding SD content to such a display, particularly one that has
letterbox marks shows clear ringing that is the result of a less than
ideal scaling algorithm. Processors like the Lumagen HDQ ring far less
given the same input and output conditions. Of course, given a less
rigorous scaling ratio (e.g., with 720p displays), the situation is far
less noticeable. However, ABT should start investing in a better scaling
algorithm in the very near future.
A nice addition that users have been asking for in recent years, is an alternate silver
front panel for the unit. I personally believe that there are
two types of processor owners. The first are the practical users that
hide the unit, along with other black anodized equipment, in the back
of the room and set to show absolutely no visible lighting.
The other kind puts everything right in front where everyone can ask why
that person (let's use me as an example) has fifteen different boxes doing
what most people only need four pieces of equipment to do. The most
visibly challenging pieces in my equipment rack are processors, which
don't really have any visible reason for being on the rack. Such
people need their devices to be as flashy and obscurely designed as
possible (see Faroudja, Extron, and Algolith box designs for good
examples of what I mean).
Unfortunately, both the silver and black
front panels don't really deliver the goods. ABT should hire a designer
to get out of the rut - practical boxes are out and good
looking pieces are in. The VP50 has all the potential of being such a
piece, it has nice buttons and back lit LCD in front. It just needs a
better box design to go along with it for those users. I know of at
least one case (a scaler by Key Digital) where users revolted against the typical black anodized
design of a processor and simply redesigned it and started producing
alternate front panel designs.
Last but not least, the VP50 hosts an optional precision SDI input.
With the proliferation of HDMI enabled players, even those that are
capable of passing along interlaced output (e.g., Oppo 970), I see less
of a need to support SDI. It primarily serves the good old CRT projector
community, but those are becoming a rare breed that suffer from a
serious lack of HDCP disabled sources. Although I still own an SDI
player (my good old trusty Panasonic RP56), I think ABT should eliminate
input and employ the valuable back panel space for something a bit more
useful, or just space the HDMI jacks farther apart.
In summary, the DVDO iScan
VP50 is a great
digital video processor. It still needs some work (audio glitches, aspect ratio, some
features like the overscan), but it's probably the best deal for the cash at
this point in time, particularly for anyone needing HD de-interlacing.
- Ofer LaOr -
Terms and Conditions of Use