Peter Finzels Sony VPL-VW10HT FAQ
I wrote this FAQ to describe my first impressions of this long expected projector. Of cause it can't be a real "FAQ", because just the first production samples of the projector were around when the text was written. So I selected the questions to describe in detail what I think are important features. The counter of the10HT showed 4 hours, when I got the set. Before taking measurements, the 10HT was always running for a minimum of half an hour. The projector was attached to a Pioneer 606 DVD Player with the choice of the new "Reference DVD" (625/50 "PAL"), the "Videos Essentials" and the "AVIA" (both 525/60 "NTSC") plus a lot of other DVDs via FBAS, S-video, YUV (more exact YPrPb) and RGB connection. Furthermore a PAL pattern generator, a Minolta precision lux meter, a (also Minolta) color analyzer and various video scalers and a PC were used. The big picture was throw on a 3,17 m (about 124 inches) wide 16:9-format Stewart Screen with StudioTek 130 fabric. This FAQ now represents the end of the test with 34 operating hours on the timer of projector. In the meantime the 10HT is away somewhere else in Germany :-(.
Some words to the history of this FAQ: It was originally written in German language and published at my homepage. Shawn Eyer contacted me, that he had run the text through the mill of an automatic translator at altavista.com so native english readers can share the information contained. Of cause the result was best descibed as "broken english" so I decided to work through the text and add a few corrections. So if you read something weird, it's for sure not from me, blame it on altavista :-) Here in Germany I publish a book about Projektors and I am a freelance writer for "AudioVision" magazine. If you are interested, you may have a look at my website at (only German language). Of couse I run my own home theater, which was the place where this test of the 10HT was made.
04.February 2000, Peter Finzel
What impression you get on the first glance?
As always with a Sony product: a very good impression. The
control panel was taken from Sonys professional devices and is
illuminable by pressing a key. The mains connection at the rear
side is elegantly set back. The front a bit stange on the first
glance, due to the "exhaust pipe" (thats what I call
it). The whole projector is relatively large, and has a
footprintof a medium sized pizzabox. Compared with others,
presently available data projectors it is something near double
size and weight. The reasons: it uses large panels (see
"panel") and an noise-absorbing ventilation (see
"as quiet...").
How well are the basic adjustments correct?
In the case of PAL material with a black offset of 0 IRE
(Pioneer calls this "Cinema" mode) it tracks the PLUGE
test pattern very well. Brightness and contrast remain on 50
respectively 80, the factory setting, you don't see this too
often. I took the sharpness back from the basic adjustment of 50
to 35, because I had to discover double outlines. If you go up to
7,5 IRE black level (what is used in US), the brightness should
be taken down to 35, in order to establish again optimal
conditions. The remaining specification refers to the 0-IRE
adjustment, because that is mainly used in Germany.
Is the pixel structure visible?
Sony used this time (in contrast to the predecessor W400) to
a strictly right-angled arrangement that of pixels (called
"Stripe" structure), while with the W400 every second
line hat a offset of a half pixel ("delta"). However
the structure remained completely invisible. According to my
observations the pixels of the panel are hardly visible, if you
are further away as the double display height (2H). But due to
poor resolution of PAL and NTSC you should sit further anyway,
best 3 to 4H.
How much light does the 10HT actually
deliver?
ANSI lumens are not easy to measure, because many devices
must be measured, in order to get a safe average value. Of cause
only 10HT was here for the test, therefore I can only check its
"mean lumens", measured according to the ANSI method
with 9 test points in the all white picture. With the optimized
adjustments for video, " cinema-black" mode and 6500K
color temperature activated (as you will see, this is a good
adjustment for video) I could get a reading of 396 lumens . If
that sounds not much in the comparison with the factory
specification of 1000 ANSI lumens, wait until the next topic, the
projector will definitly achieve this value. But while delivering
400 lumens, the projector meets the correct color temperature
very accurately, additionally the "Cinema black mode"
reduces the performance of the lamp, whereby the light
performance decreases around approximately 35%. A measurement
with "cinema-black" in the "OFF"-position
reads 590 lumens.
Is the VPL-VW10HT able to deliver 1000 ANSI
lumens according to its specifications?
That happens not very often in the projector business: the
Sony actually delivers more than the promised "kilo
lumen", where a lot of other projectors fail to meet their
specifications. However this is only achieved without
"cinema black" and without regards to the correct color
temperature. I got a reading of maximum lightoutput of 1250
lumens (mean value!) in the 16:9 mode. Of cause all adjustments
where on absolute maximum, so it is not a recommendet setting.
Another measurement under conditions similar to ANSI resulted in
840 lumens. But consider: during this measurement picture
parameters were optimized for - let's say "computer
presentation" with high contrast (only the gradation of the
grey scale has to be fairly OK), "cinema-black " off,
and color temperature dodn't matter. If you go for real optimized
video adjustment, the light output remains with "cinema
black" at 400 lumens and 600 lumens with lamp under full
power.
What about the color temperature?
Short version: as well as never before with LCD. Long
version: The VW10 meets the desired "colour of white ",
which is for video pictures at 6500Kelvin better than all other
LCD (and DLP -) projectors I've seen before. The "low"
adjustment got the best reading on the Color Analyzer. With 100
IRE (white) I could measure color coordinates of x=0,314 and
y=3,42 (which corresponds to 5500K, 6500K is x=0,313 and
y=0,329). The slightly too high y-value reveals a bit too much
green, but it is really close. Still better: the coordinates (aka
the color temperature) remains constant over the gray scale.
Sounds very theoretically, but can be seen in real life. Some
displays get for instance bluish for bright scenes and turn to
red or green tones in in darken scenes. But not the 10HT, if only
all beamers were like that. Just a small warning: once you've
spoted this error on - maybe - your present display, you can
hardly ignore it any more. So better don't look to close :-).
However good gray scele has its price: if you move the contrast
past the 80'sl (70 is actually still better), the color
temperature is out of control. As a matter of fact, lamp does not
supply enough red, so that blue and green take over he picture.
That is true regardless of "cinema-black" on or off.
What about contrast and luminace fall off to
the sides?
The light is distributed very equally throughout the picture
area - the darkest of the 9 testpoints is 96% of the brightest
spot. That is extremly good, especially compared to CRT-Units
being somewhere between 50-60%. Very few other projectors can
draw even. The contrast can't follow, as always with LCD. For
contrast there are basically two measuring methods: On/off and
ANSI. The latter is measured with a test pattern with alternating
bright and dark rectangles and stresses not only the panels, but
also light scatterings in the optics. The VW10 achieves approx.
87:1, good for a LCD device. On/off measurement is not so exactly
defined: some use a white picture with controls on minimum and a
black picture with controls on minimum. If you do it the right
way, you will get approximately 98:1.
Now, is the black level comparable with a
CRT projector?
THE question most of you want know about. Sorry, the answer
is a definite "NO". If you use the optimized video
adjustment plus "cinema black" activated, I could still
measure approximately 5.8 lumens while projecting a black
picture. This is with is an amazingly high value! Just compare
yourself in the table provided in my homepage: a Toshiba MT-1 has
only 3.3 lumens, others LCDs are even lower (but most of them are
worse on maximum light!). However, for a Sony it is typical: the
VPL-S900 had a similarly high low light level with 5,6 lumens.
What are the consequences: well illuminated, for high key scenes
the VW10 does a extremly good job. If it gets darker, the VW10
just can't follow. It's definitely not, that you won't watch it
any more, but who is accustomed to CRTs, will not accept ist.
Despite this more grayish "black" the 10HT falls never
short of picture definition in dark areas, which helps somewhat
to ignore this shortcoming (if you can at all...).
Is the VPL-VW10HT actually as quietly as it
was announced?
Switches it on the first time, and be shocked: the fan starts
up like a hair dryer! However only quite briefly, then the rpm is
reduced dramatically to a really bearable level. I am not quite
sure whether the CPU-fan of my Notebook (on which I write this
FAQ) or the 10HT is noisier - and my notebook has never disturbed
me. But it is rather crucial whether "cinema-black"
mode is enabled: without the fan switches a gear up. So
"cinema black" is also "cinema-quiet"! I
measured he following values: 37dB(A) in the "cinema
black" mode and 39dB(A) without. His closest competitor is
the Toshiba MT-1 with 39dB(A). For your information: all those
readings were found not in a acustic dead room, but in my home
cinema. So it is not a "official" value, but it can be
compared with other readings, I got at the same place, same
position, with the same measuring instrument. The 37dBA is at
present the best result of all LCD and DLP projectors. Only a CRT
(the BarcoVision 708MM) kept up with 37,5dB. Applause for Sony!
Inputs of the VW10HT
The input variety of this projector is a dream also for
seen-it-all professionals: Video and S-video are standard, but
two further ports equipped with 5 RCA-connectors are the stuff,
high end video dreams are made of. They are called simply
"A" and "B" and can be configured for RGB or
YPrPb. These ports work with everything. Really everything,
except three phase current :-) Whether a component output of a
DVD Players, a HDTV Settop box or RGB from a PC: no problem, if
the suitable adapter is within reach. Especially good for us
Europeans: even RGB from a SCART (ever seen one in US?) socket
works immediately. Holy straw bag (german idiom for something you
would say "jeez"), there the Sony engineers really
thought of everything. It's getting almost boring for the tester,
if you cannot find just one hair in the soup. For different
adjustments with different sources the VW10 offers six "
video memories ", which hold basic adjustments such as
contrast and brightness, plus display format and color
temperature. By pressing on a numerical key of the remote each
memory can be activated immediately. Newer than the operating
instruction booklet are four "User" color temperatures,
which can be adjusted only by the dealer in service mode.
Are the YPrPb-Inputs better than S-Video?
Yes! The picture improvements are subtle and maybe you will
not notice them in the first five minutes, but the VW10 puts the
higher color resolution of the component signal into a noticeably
better picture. That may sound subjective, but with test patterns
you can prove it. However you will need a colored multiburst,
which is not to be found on the usual test DVDs like VE or AVIA
unfortunately. I found one on the "Sonopress demo DVD"
(not for sals, a collectors item...) and it turns out clearly:
Color details are represented with components more clearly, with
S-Video smear fine color details. Note: the advantage concerns
only color. Sharpness in the sense of brightness details will not
increase, because the luminance with YPrPb and S-Video are
identically transmitted.
What does it mean that the VW10 can also
process " progressive pictures " in the YPrPb format?
Under normal conditions YPrPb transport video signals with
interlacing, as you know from video or s-video sources. Component
is better, because the signals are not mixed. But YPrPb can also
do better: because it can be used to transmit also signals
without interlacing hence progressive. Those may come from a DVD
Player with " progressive" output, from a Videoscaler
(aka linedoubler) or from a HDTV decoder (not in Europe, as we do
not have HDTV - wake up!!). The VW10 detects the formats 480p
(NTSC after linedoubling, oder DTV), 575p (PAL after
linedoubling), also 720p and 1080i (US HDTV). Because of known
reasons, the latters can hardly be tested here. I managed to test
480p and 575p by attaching a Faroudja DVP-2200, which happens to
produce just this signals at its output. Result: all worked
immediately. The picture profits from the film mode of the
Faroudja, but the combination is of cause unreasonable by the
high price of the DVP. But a DVD Player with progressive
components output - as long as its
"De-Interlacing"-circuits operate as they should do -
could be quite an advantage in repect to the integrated doubler
of the 10HT. <30.3.2000> A friend of mine tried the
combination of Toshiba SD-5109 and 10HT abd confirmed me that it
works perfectly. But keep in mind - the pro-scan output of the
5109 is only active while playing 525/60("NTSC") DVDs!
How easy can DVD Players with Scart-RGB be
connected?
Amazingly well! US-Readers may not be interested by the
importance of this question (as there are no Players with Scart),
but a lot of european players do have RGB-Outputs hidden in the
Scart-sockets. "A" and "B"-inputs accept the
so-called " 15k RGB" signal both with 50 and 60Hz
picture frequency (or: PAL or NTSC) All what is needed: an
adapter from Scart from the DVD Player to four RCA plugs. Three
of it carry the RGB signals (from Scart 15,11,7), the fourth
cable strangely needs the composite color picture signal of the
Scart pin 19, which is connected to the socket marked with
Sync/HD on the 10HT. Then set the respective input in the menu to
"Computer" or " DTV-GRB", both adjustments
work. There are options for adjustment the horizontal size, image
position in addition, all display formats can be used (only in15K
mode, not with higher scan frequencies!). Sony built into the
10HT a thing called "sync stripper", what extracts the
sync impulses from the FBAS mixture. This is, how a true home
cinema projector should be, because now the highest signal
quality can be transported to the VW10 by each reasonable DVD
Player. Advantage of RGB: the whole color bandwidth is used.
Since the VW10 processes s-video signals very well, the
difference to RGB is not too large, however with special test
patterns you will see it (as with YPrPb). Fine color details and
color edges are better defined, so it is well worth the effort.
How to link to a PC
Maybe you will call me crazy, but I tried the connectivity to
a PC with the help of video equipment - to be more more exact I
used (also Sony) DSC1024 digital scan converter. That permits
fast changing resolutions, frequencies and syncs without the
gracious help from Windows - so no need for rebooting, system
crashing and other nice things. Anyhow: If you connect VW10HT to
RGB signals with scan frequencies above 15KHz, its behavior
changes suddenly. As soon as you set the input to
"computer", the "display format" disappears
from the menu. That means: no more 16:9 format (see next topic)!
Contrast, brightness and color temperature remain adjustable, a
new feature called "RGB Enhancement" is added. That is
a kind of sharpness for RGB signals - with inevitably the double
outlines arising. In the menu there are now input adjustments as
phase, the horizontal size and the image position, but all work
only inbetween the 4:3-limits. The 10HT displayed all resolutions
the scan converter could produce. Up to SVGA you can select the
native resolution or the internal scan converter option, that
fills the 4:3-part of the Panel (that equals 1024x768, XGA). The
adjustment "video memory" works also here, as dose the
trapezoid adjustment. All resolutions up to 1280 times 1024
(these down-converted) were detected immediately, both with
Composite Sync (connection with 4 cables) and "separate Sync
" (5 cables).
Can you zoom the PC-Picture to 16:9 format?
Not easily. Of cause, current PC-Software is limited to 4:3.
But when you use PC as DVD Player the full width should be
usable. This is not very good news for all those with the
intention of using a HTPC! I found only one work-around: use the
adjustment "DTV-GBR " instead of "computer",
and use a good videocard like Matrox G400, Elsa Erazor III
(perhaps others) that are able to meets exactly the resolution of
a defined DTV mode, thus 480/60p, 575/50p, 720p, 1080i. If you
can simulate a suitable 16:9 desktop for Windows, your finished
the job. With the "linedoubler" option activated on the
DSC1024, I could simulate 480p and 575p, and with "
DTV-GBR" the format selection. came back. Then use
"full" select and voila - the panel is flooded! I also
used a Elsa Erazor to simulaate 720x576 (that is
"progressive" PAL), and it nearly (as always with PCS)
worked. Maybe some hours finetuning (which I didn't have) and it
would be perfect...
Is the trapezoid correction really usefull?
Very usefull! A common problem of many LCD projectors is that
they must be set up on the height given by the manufacturer
(usually lower or upper screen edge). A higher or lower position
leads to distortions in trapezoid shape. Solutions for this
problem can take place optically ("Lens Shift ") or by
electronic means. The VW10HT uses the latter but a problem with
this type of correction occurs, if the pixel count even is just
sufficient, to show the resolution of the signal. If the
trapezoid correction is activated, the resolution is reduced,
because some pixels remain inevitably unused in the boundary
regions. With other projectors, which have just enough pixel for
the signal it hails artifacts. Not with the VW10: since it has
many more pixels, than are necessary for the representation of
NTSC or PAL (720x576) , the picture must be scaled up anyway, so
there is enough work eraea left for the trapezoid correction.
Nevertheless: Quality enthusiasts will rather set up the
projector in the correct height (screen edge) and switch off the
electronic correction. The balance area of the trapezoid
correction is extremely large
Is the optics how flexible?
Sony set the red pencil here. Both zoom shot and focus are
only manually operated. But honestly: a projector with so
flexible format selection a manual zoom is all you need. The
options for placing the procetor in the room by a biveb picture
width are severly limited by the poor zoom range of (measured)
1:1,23 is very reduced. On the other hand, the optics are made
for short projection distance. Without intervention by the
digital trapezoid correction that the lens center should be on
the height of the screen edge. The short distance is in most
cases very practical, because then you sit behind the projector,
whereby the warm air flow of the cooling fan is directed away
from you. Last news: there are other lenses for change,
especially one for long throw available. But wait for the price:
more than 3500 Euros/Dollars!
How long does the lamp last?
That is one of the few questions, which one can hardly answer
in a test (I hade the 10HT just 6 days at my disposal). During
the last measurement (after a total of 34 hours of operation, 4
at beginning of the test), I found a decrease of about 8%, what
sound quite a lot to me. Of cause it is very hard to measure,
because small variations in mains voltage cause also some
variations of light output. So I took a average of the first to
readings, when the unit was brand new, and executed two more
before I shipped the 10HT back to Sony. The resulting difference
was about 7% in 30 hours. As it is not clear that that the light
decrease is linear, it might not be true, but if it is, the light
output would be at 50% after approx. 200 hours! I appreciate
every comment anybody can give me about this topic. Sony itsself
indicates the average life span in the manual with
"cinema-black" as 2000 hours, for maximum lamp
performance as 1000 hours. A back-up lamp is not very cheap with
1144 D-Marks (that is about 570 Euros or US-Dollars right now).
Anyhow: "cinema black" again proves as very helpful,
because it is a " long-live-the-lamp" mode!
Does the VW10 have a linedoubler?
Of cause! If an LCD projector has a panel with more than 240
of pixel in vertical direction, there has to be a linedoubler.
The function of the linedoubler is to get rid of
"interlace", or in other words, the interlaced video
signals are transfered into a "progressively" scanned
signal. So the question is not, whether an LCD projector has a
doubler, but how well it works. There is still more: because more
than 576 vertical (repectively 480 for NTSC) pixels are
available, an additional Scaler is neccessary. This scaler takes
the results of the linedoubler and boosts it to the full
resolution of the panel. Additionally such a Scaler is very
useful for the realisation of the various display format (16:9
anamorphic, 4:3, zoom, see below). The 10HT has both a doubler
and a Scaler. The doubler has 2 operating modes (DRC-MF
"x4" for moved pictures or " sequential " for
static picture contents).
How good is the internal doubler?
Unfortunately is the handling of interlaced signals is not
the strongest point of the 10HT. Sony used its "DRC"
(Digitally Reality Creation) technique that is also built into
US-HDTV rear projection sets upconverting standard NTSC signals.
Compared to the best at present available linedoubler (Faroudja),
the Sony de-interlacer shows indeterminable weaknesses. Moving
diagonal lines are always accompanied by digital stair steps
("jaggies"). If slower pans over fine detailed
structures result in something probably best described as a type
of "pixelation noise". This error is definitly due to
the integrated linedoubler of the VW10, because if you attach an
external doubler with film mode, artefacts like "pixelation
noise" and "Jaggies" are gone. According to this
observation, the VW10HT may gain a lot of picture quality with
the use of progressive output DVD-Players, because the internal
doubler of the Sonys is not used.
Display formats of the VW10: the Scaler
Sony uses the very flexible scaler in the VW10 to cover all
important formats for video presentation. Full: anamorphic
format, the width is horizontal zoomed to the full 16:9 extend. Full
through: somewhat unfortunately selected name, that means
that the anamorphic format keeps vertically the original number
of lines scales and is zoomed horizontally in such a way that the
format is correct. However with NTSC input there is a small
error, because it is scaled to 1,9:1 instead of 1,78:1. Since the
panel is not filled with this mode, that error will hardly
disturb somebody. Normal: 4:3-format, the extra pixels to
the left and on the right of the 4:3 part is not used. Result:
black strips on the left and on the right. Normal through:
4:3 format with the original number of scan lines ("Window
boxed"). Zoom: Increases a 4:3 picture horizontal and
vertically, so that the panel is filled. On the top and bottom,
the surplus picture parts are lost Ideal for
"Letterbox" movies. Subtitle: vertically
differently zoomed, so that the bottom of the picture is not lost
as much and subtitles can still be seen. Widezoom: strangely zoom
in both directions, so more of a 4:3 picture can be seen, but
distortions occur. Not 4:3, not 16:9, so better keep away from
it.
How does the VW10 cut off in the direct
comparison to a much more expensive CRT projector?
I had the pleasure to compare a VW10HT (14000 D-Marks worth)
directly with a combination of a Seleco SVD800HD CRT projektor
(27,000 Marks) plus a Faroudja DVP2200 video processor (19,000
D-Marks). Both devices were attached in parallel to the same DVD
Player by S-Video connections (the Pioneer 606 is good for
comparison tests, as it offers two S-video outputs!). Both
projectors threw their pictures on a 3.17 meters wide Stewart
16:9 screen with StudioTek 130 fabric. Alternating the pictures
from both projectors showed already the most substantial
difference: muting the picture of the Seleco shuts off its light
output completely. Not like the Sony: when you presses "Pic
Mute", the screen is still slightly bluish. For that reason
I used plan B: put on the lense cap. That is brute force, but it
works. Apart from the picture, the Seleco is louder than the Sony
with "cinema black" adjustment on. Now, "Video
Essentials " chapter 17, chart 13 (Indian) shows clearly the
better vertical resolution of the Faroudja Scalers in vertical
direction. The resolution brooms are completely separated to the
narrow end, while with the Sony the lines tough between the 300
and 400 lines. In horizontal direction the Sony is marginally
better, because the brooms on the left and on the right are
separated somewhat further. On the other hand the brightness
award goes to Sony. The white of the Selecos tended to the
greenish side. The Sony bright white is very accurate, but darker
gray tones have a slightly greenish impact. Now step away to
chapter 19 ("montage of images"): Particularly the
initially dark scenes exhibit the too high black level of the
Sonys. With many LCD projectors subtle nuances are lost in to
high black level, that ist however never the case with the VW10.
I could not find a single detail in the Seleco/Faroudja picture,
which was not to be seen with the Sony. But the problem is
located somewhere else: particularly the scene in the backyard
the CRT-projector created an immense impression of depth, the
picture of the VW10 just being - flatter. On the other side,
bright objects seemed a bit sharper than on the CRT projector,
for example the boat on the river a few pictures further. After
all the test pictures, the real life comes in form of the
relatively dark starting sequence of " Lethal Weapon IV
". Here the high black level disturbs more severly than with
all the test patterns. If Riggs and Murtaugh go into cover behind
the car, the picture misses all the "punch". The CRTs
work much more precise and homogeneous. Further on to chapter 11
(beach walk), the pixel noise of the Sony is really annoying, a
weakness of the integrated linedoubler. That is the domain of the
Faroudja, he rules there sovereignly. Of cause there are coding
artifacts (the approx. 6 megabit/sec. are too low to cover all
the details in this scene), but pixelation noise is not present
with the Seleco/Faroudja duo. Scene change: "Six days, seven
nights " (code 2), chapter 3. A bright scene, no black
problem for the Sony, which is almost an idea sharper than the
Seleco. In horizontal and diagonal edges however clear doubling
artefacts are showing up. More of it in chapter 4 (e.g. beach
chairs) and chapter 8 (jagged wings). A genuine black problem
occurs in chapter 16 (island lunch), there is incomparably more
room depth in the CRT combination. Not easy to draw a summary.
The advantages of the almost 50,000 Marks combination are of
subjective nature, but undisputed they are there. It falls
shorter on total light output, but the better room depth and the
convincing rendering of the (unfortunately frequently occurring)
dark scenes are a strong reason for genuine movie lovers (like
me) to save their money more this expensive combination. Take a
look on the other side: the Sony costs less than a third of the
CRT and delivers image quality not so far away from that of the
much mor expensive combination. Who can manage to live with a
somewhat brighter black and a few artifacts from the
de-interlacing circuit, makes a good business with the Sony. Do
not forget: the doubling artifacts could be eliminated by a DVD
Player with really good "progressive output" in the
near future. But who wants to have the maximim film impact, still
needs an expensive CRT-beamer and a very good (aka expensive)
video processor. Is that not within your financial range, then
the Sony is the best alternative so far. For simple CRTs with
integrated doubler the VW10 is a hard competitor. The question
is, whether one prefers the higher resolution of the Sony
(already the 800 has difficulties to follow the resolution of the
VW10) to the better black of the CRT. In addition also still the
handling problem comes: an LCD device never needs an adjustment
of the convergence.
Copyright 2000 by Peter Finzel. Reproduction only with written permission.