After several viewing sessions I am glad to report that I am thrilled with the VW95. Indeed it is a very
nice upgrade to my RS20 which I’ve had for 3 years, and I spent several hours with A/B testing to draw
many conclusions.
Note that I am just a hobbyist and these are my laymen opinions. You should not make any decisions to
buy or sell a projector based on this. Everyone has different tastes so YMMV. There is no substitute for
seeing a projector you are interested in first-hand and evaluating it based on what is most important to
you, before making such decisions.
First some background on how I did the A/B testing…
For the A/B testing I used a Lumagen Radiance XS to output the same signal to both pjs over HDMI.
Memory A on the Radiance was set to display video on the RS20 and no video on the VW95. Memory B
was set vice-versa. Each memory also contained the appropriate calibration grayscale for a dE of less
than 2 from 5-100%, a gamma of 2.22, and a gamut that measured nearly perfectly for Rec 709.
The Radiance is an excellent product and provides a great deal of calibration flexibility and features. I’ll
have more to say about the Radiance in a separate write up, but for now suffice it to say that I find it an
essential tool for the HT enthusiast. The addition of the Mini line was a great idea and provides a nice
price point for a highly functional unit.
The Radiance made the A/B testing rather easy. I could just select memory A on the Radiance and see
the RS20, or memory B and see the VW95. Due to HDMI handshaking with the projector, there was
about a 3 – 4 second delay when switching between the two projectors.
For all but the darkest scenes the little bit of light that comes out of either pj when they are not getting a
signal did not impact viewing of the other (I even tested this). For more critical black level viewing,
however, I also used cardboard to completely block any light from the non-active pj.
The VW95 was ceiling mounted and running in low lamp mode with Auto Iris 1. The RS20 was table mounted
about two feet closer to the screen. It was in low lamp mode with the iris fully opened, which is the mode
I always run in. I recently replaced the bulb in the RS20, so the lamp is still quite bright. In fact the
measured lumens were approximately 10% higher on the RS20 compared to the VW95.
I watched all sorts of HD material – video and film from cable TV, and clips from several Blu-rays. Over
the course of two nights and many hours of viewing, I watched and paused well over 100 different
scenes of various types (dark, bright, video, film, etc) and flipped back and forth to make comparisons.
Overall impressions from A/B comparison
The first thing I should say is that, without the benefit of an A/B comparison, the RS20 still looks amazing
to me. I still go Wow when I see it. However, that changes when viewed in the context of the A/B
comparison, as the VW95 easily beats the RS20 in most categories.
The best way to describe my overall impressions between the two is that the RS20 looks somewhat
“murky” and a bit dull in comparison. I was amazed to see the additional contrast, pop and vibrancy
that the VW95 has – especially considering that not only were both pjs were running a 2.22 gamma with
a finely tuned Rec 709 gamut, but also because the RS20 was actually 10% brighter.
I wondered how this could be. Both pjs were well calibrated and yet the one that’s putting out 10% less
lumens is displaying a cleaner, vibrant, seemingly brighter and more “contrasty” image. I suppose this is
due to the higher ANSI CR or ON/OFF of the VW95. Then again, can on/off CR really come into play in the
context of a single frame? I wouldn’t think so since the pj is still dependent on the native CR in that case,
but whatever the reason, the VW95 had a more dynamic and vibrant picture.
I will now address some characteristics that people often ask about…Black Level
This area was very close and was hard to judge from still frames, mainly because of the time it took for
the HDMI switching. And during such handshaking the picture was getting bright, then dimmer, then a
little flash of color etc, so I would lose my frame of reference by the time the new image was displayed.
That said, when I watch the VW95 in motion and watch dark scenes, I find it to have more satisfying
black levels then the RS20. I do not know if this is because the black levels are darker, or because the
image has more vibrancy and pop to it (which can then make the blacks look darker relative to the
brighter whites), but at any rate the black levels are very satisfying to me. Also on a total fade to black it
seems the VW95 is darker, based on seeing the light on the screen and the hand-puppet test (and yes
you can still see hand puppets with the VW95).
Convergence
Given the many reports of excellent convergence from other owners, I was surprised to see that my
convergence wasn’t so great OOTB. I needed to move red down 3 and over 2, and blue over 2 using the
panel shift function. This was not the zone function, but rather the full panel shift option. Unlike the
RS20 where a single click of panel shift makes a large change, the changes are much finer with the VW95
adjustment.
After making the convergence adjustments on the VW95 using a single pixel grid, the convergence was
excellent. The only area that is a bit off was about 20% of one side of the screen had green off by a half-
pixel, perhaps more. I am sure I could use the zone adjustment to correct this. However I cannot see any
fringing or other such issues caused by it, so I am leaving it alone. I figure why engage additional
processing and possible artifacts when it’s not a problem to begin with. When I run out of things to
tinker with I may experiment with the zone adjustment, but for now I really see no need.
The VW95 makes it very easy to turn on and off the panel shift function. I used this feature to A/B many
scenes and still frames to compare the VW95 against itself both with and without the shift function
engaged. I could not detect any negative effects from using the panel shift function like I did.
Focus and Sharpness
I found it a bit difficult to get the focus just right. I prefer a manual lens focus so I can rock it back and
forth and find the sweet spot. Unlike the RS20, the VW95 is rather challenging to see individual pixels
and the grid even when right up at the screen. I used a magnifier card up at the screen to zoom in on the
pixels and set the focus at the point where they were most defined. I then checked edge to edge, up and
down, and the focus was very even. I exchanged screenshots of my focus with other VW95 owners and
we agreed our focus and pixels looked about the same up close at the screen.
The picture is sharp, but once again after years of yearning for a sharper picture I still feel like it can be
sharper. It is a bit sharper and more focused than the RS20, but my expectations were higher after
hearing for years how much better the Sony lens is and how much sharper it is. Like was the case with
the RS20, I still find for my personal taste that I need to use the sharpness control to sharpen things up
to taste, even though technically it is artificial sharpening. For now I settled on a sharpness setting of
22. The picture looks nicely sharp and focused, but not a night and day difference compared to the
RS20. My Samsung 63” plasma (PN63C8000) looks far sharper, but that is probably not a fair
comparison.
Clarity and Fine Details
This is one of the areas that the VW95 is a far stand out above the RS20 and other pjs I have owned. The
picture is very clear and crisp. The VW95 does an amazing job with fine details and textures. The picture
is also “silky smooth”.
Reflections off of water and other objects and other such shiny surfaces are real stand-outs. One such
example of this was a close up of jewelry with diamonds shown in a HD commercial. On the VW95 the
diamonds were sparkling and glistening much more than on the RS20. Likewise there is a scene in
Baraka where they are inside a structure that is made up of all glass and panning through a hallway and
showing the ceiling. The reflections and sparkles coming off of all the glass and crystals made for a jaw
dropping moment on the VW95 and does not glisten nearly as much on the RS20.
Scenes with fine details and particles look rather striking as well on the VW95. For example snow, dust,
smoke, sand, leaves on trees out in the distance, etc look more real and defined than I am used to
seeing. All these properties combine to deliver a very life-like picture, particularly when the HD quality
is very high.
Shadow Detail
The shadow detail I see on the VW95 is outstanding. What is crushed in the blacks on my RS20 is shown
in great detail on the VW95. A fellow RS20 owner mentioned that shadow detail is a function of gamma,
which I agree with. He says he can see level 17 and 18 on test patterns. On the RS20 in my light
controlled room in total darkness I definitely cannot see 17 and never have been able to, and 18 is even
a bit questionable. And my gamma at 5% is at 2.2 and not crushed and my black level (brightness) is set
correctly. Raising the brightness even a single click not only still does not resolve level 17, but also raises
the entire black floor.
There were many scenes where the shadow detail was much better on the VW95 than the RS20. One
such example is in Monsters, Inc. Toward the end of the movie Sully is trying to escape with the girl. He
opens a door and looks inside and sees a long and narrow tunnel. On the RS20 it is mostly black with
some highlights and details. On the VW95 the tunnel is much more lit up and much finer details and
highlights can be seen.
Motion and Artifacts
As would be expected, the motion handling on the VW95 is very good and far better than the RS20. I set
the MotionFlow to Low, and in some cases to High (do not notice much of a difference between these
two settings). Watching sports on the VW95 is a lot more enjoyable. Gone are the zebra-stripe red
fringing from referee jerseys and scrolling text as can be seen on the RS20. Although I will note that
while scrolling text looks great on the VW95, it still doesn’t look 100% perfect (but close enough).
Hockey is very enjoyable to watch on the VW95. The players indeed look like they are gliding freely. This
is especially noticeable after play stops and they just coast around a bit. It gives a good sense to the
feeling of “being there”, like I am looking over the glass from the lower level. Football looks great too,
with hard thrown spirals keeping form and shape nicely. The super slow-motion replays are a joy to
watch too.
On such example of the VW95’s superior motion handling jumped out at me when I wasn’t looking for it.
There is a scene in Baraka where they are showing several huge waterfalls. As I was A/B-ing this scene I
noticed that the water was keeping its full form, shape and definition as it rolled over the falls and was
falling. You could still make out its particles and structure while it fell just fine. On the RS20 it still looked
like water of course, but many of these fine details were lost in the motion.
I have watched this scene many times on my RS20 and of course it never dawned on me that some of
these details were getting lost. Without the benefit of an A/B test I would not have appreciated the
improvement that the VW95 added to scenes like this. It seems that whether the little things like this
are directly noticed or not, they all contribute to the VW95 providing that “being there” feel more-so
than other projectors I have had.
It should be noted, however, that my VW95 is not artifact-free. There are some occasions where a
moire-like effect pattern can be seen as the camera pans or zooms over certain types of patterns. In
some cases it is very subtle and quick, and in other cases it is more pronounced and a distraction.
One thing that may provide a hint is that I have only noticed this while watching video on broadcast HD.
I have not noticed it at all on movies (film) on broadcast TV or on Blu-ray movies or the XBOX/PS3
gaming.
I have done some very basic and preliminary troubleshooting and haven’t been able to pinpoint the
source of this yet. So far I tried turning off the panel shift function, turning off Motion Flow, turning off
Sharpness (min) and yet the artifact is still reproducible. When I have a chance I will try bypassing my VP
to eliminate that as well. The good part is that I have clips on my DVR where this occurs so I can easily
test different settings and experiment. If anyone has some ideas on what types of settings or changes to
try in order to pinpoint the source of this, please let me know!
Uniformity
Overall the uniformity of the VW95 is very good. There is a very slight pinkish tone that can be seen in
some full field gray test patterns. It might be my imagination, but it seems to have improved since I first
got the unit, and I have not seen this in any real world material (although at first I thought I may have). It
could also be a newer calibration setting I am using (maybe I was driving red too hard before).
The brightness uniformity is also very good, but not perfect. About 20% of one side of the screen
vertically seems a tad bit darker when viewing a full field 100% pattern. But again nothing that is
noticeable with real-world content or that jumps out at me.
My VW95 may have very slight brighter corners. It is hard to tell. Your eyes can play tricks on you in the
dark, especially as dark as things get with the iris fully clamps down on a 0% pattern. Most of the time it
seems there are no bright corners. Although sometimes I think I may see just a hint of them. Even
though I am very sensitive to bright corners, I am happy to report that even if they are there it is very
slight and does not affect the PQ in any way and is not really noticeable (if I’m not quite sure I am seeing
them when gawking for them, I am not going to worry about it for sure!).
The uniformity is one area I give the RS20 an advantage. My RS20 is nearly perfect in this regard, as far
as having no color shift, being evenly lit and not even the slightest hint of bright corners. That said, I do
not see any real world benefit from this advantage, and the VW95 still seems to have a darker fade to
black as mentioned earlier.
Calibration
As has been mentioned by others, the VW95 OOTB settings are very good. dE was off by about 3 on
mine. However I am perplexed at why the default gamma is 8. This gamma measures very poorly,
basically falling off a cliff in the upper end. In one of AVS member SOWK’s calibration reports you can
see how this gamma measures by looking at his pre-calibration numbers. I set gamma to OFF and
calibrated the unit using the Radiance 21pt grayscale.
I did not use it or measure it, but SOWK reported that using the RCP causes some linear tracking issues
at other saturation levels, and that this did not occur with RCP off. This sounds like the same or similar
issue than what RS20 users have faced. I used the Radiance’s gamut controls to dial in Rec 709, starting
with the Wide 3 color space which I found necessary to get enough room to move the points exactly
where I needed to. It may be possible to do this with the Normal or Wide 1 setting. I didn’t stick with it
long enough to find out (it seemed like I would not have enough room so I changed direction before fully
experimenting), and another AVS member has reported good results doing so.
I did some A/B testing on the VW95 between my Rec 709 calibration in the Radiance compared to the
default gamut in Normal mode on the VW95. It is not very easy to switch back and forth instantly so it is
hard to compare exactly. But it seems that the default Normal mode is very good. It is very close to my
calibrated results if perhaps a tad bit under saturated compared to mine. If you do not have a Radiance
or the means to calibrate one, it is probably best to just run in Normal mode with RCP off then to worry
about it or mess with RCP.
Also note that the Wide 1 color space is very tastefully done for those who like a more saturated gamut.
In this mode green, yellow and red are basically pushed out, but not obnoxiously so. For some animated
movies I could see running in Wide 1 just to add more punch and fun to the image.
I did not use Sony’s ImageDirector software, but if you do not have a Radiance you can supposedly use it
to dial in the RGB gamma for the 5-95% range. This would be similar I suppose to the built in custom
gamma controls in the RS20, which can be used with great patience and some trickery (well
documented elsewhere). I understand that the later generation of JVCs made the custom gamma easier
to work with, but I have not used it.
Much has already been said about the CMS and gamut on the RS20. For those that follow my posts, I
spent nearly 100 bulb hours years ago trying to come up with a compromise of settings that tracked
decently across various saturation levels and still provided a gamut close to Rec 709. In comparison to
the Rec 709 provided by the OOTB settings or calibrated settings on the VW95, the CMS settings I was
using on the RS20 left the picture rather under saturated.
In order to provide an apples-to-apples comparison of the VW95 and RS20, I used the Radiance to dial in
Rec 709 on the RS20 so that it was nearly identical to that on the VW95. Therefore in my assessment I
did not notice any benefit in color on the VW95 versus the RS20. However, if one does not have a
Radiance or other VP that can deliver good CMS results, then the VW95 may offer an advantage in color
reproduction as well.
As far as measurements go, as I reported a week or so ago I measured 120,000:1 on/off on the VW95
with auto iris on in low lamp mode. This compares to about 30,000:1 or 50,000:1 that I get with the
RS20 (depending on its lamp and iris setting) without an auto iris. In my configuration the VW95 is not as
bright as the RS20. The VW95 with about 80 hours on the bulb is putting out 325 lumens from near the
back of the throw range. This output seems somewhat lower than what other VW95 owners have
reported.
Dimensionality (2D viewing)
On the VW95 once in a while certain scenes come on that are just draw dropping in terms of depth and
dimensionality in 2D viewing. The RS20 is no slouch in this area either, but the effect is more
pronounced on the VW95.
On the couple of occasions that friends have seen the unit, different people have remarked with “look at
that!” and “wow!”, and asked me if they were watching 3D. Clearly they have not experienced true 3D
at home to think that, but the point remains – the fine details, shadow detail, high contrast and other
characteristics on the VW95 contribute to great depth and dimensionality.
3D
I find the 3D on the VW95 rather excellent and enjoyable. It is not perfect and there is ghosting at
times, but it easily exceeds my expectations. Note that my impressions are based on the MonsterVision
/ Optoma 3D glasses. I rarely use the Sony glasses as they are reserved for the last to arrive.
But I will
say this about them – they are not as bright, not as clear and introduce a little color shift and “dulling”
compared to the MonsterVision / Optoma glasses. Also the battery seems weak. It’s flashing after just a
few hours of usage. I need to look more closely at it. Maybe they are not turning off or the batteries are
not great (both pairs do the same). Yes it is a noticeable and appreciable difference between these
glasses, for certain.
The first thing I should mention is that, although I am still experimenting, it seems that a warm up period
of 30 – 45 minutes is necessary for the 3D to work very well with little or no ghosting – at least with SbS,
TnB or 720p60 frame packed. In these modes the 3D is pretty unwatchable to me from a cold start
(rather blurry, lots of ghosting, and bothersome to the eyes). I believe this is similar to what others have
reported on different manufacturer’s pjs as well. The 1080p24 frame packed from Blu-ray may not be
affected nearly as much by a cold start – I have to go back and test this though.
In 1080p24 frame packed the glasses run at 96hz (48hz per eye). And in all other modes they run at
120hz (60hz per eye). I believe the idea is that with 24sf material they can take advantage of the native
multiples of 24 which I suppose is better for motion without the need to alter the frame count. However
I am very sensitive to flicker so in this mode I see some flicker particularly on bright scenes and it can be
somewhat distracting at times. Note that many people are not sensitive to flicker and will not likely
notice it. In the other 3D modes at 120hz I do not see any flicker.
I like watching the 1080p24 frame packed 3D with Motion Enhancer set to High. Note that the Motion
Enhancer control is only available in this 3D mode, unfortunately. I did not initially realize the extent of
the downside of this, but after spending a night trying some very fast scrolling PS3 3D games like Sonic
Generations, I certainly do.
More specifically, very fast non-stop scrolling 3D games do not look choppy, but it is harder on the eyes
and a bit blurry compared to playing on my Samsung plasma in 3D. I believe the plasma has its own motion
handling that does runs in 3D mode, which could account for the difference. I am reaching out to
Sony engineering to see if they will consider adding support for Motion Flow with SbS, TnB and 720p60
frame packed 3D.
It should also be noted that Dark Frame Insertion (DFI) is not available at all in 3D, which I don’t think
actually matters because I have yet to find a situation where DFI offered an improvement anyway.
Overall I think the 3D is very well done. It is plenty bright for me, even with darker movies. 3D gaming
through the PS3 looks great as well. In high lamp mode I am getting about 525 lumens. This of course
does not account for the light loss through the glasses, which is estimated by some at around 50% (in
general, not specific to the VW95). On my 106” diagonal Firehawk 1.3 gain screen that would work out
to about 10ftL through the glasses. It may not sound like a lot, but it is plenty bright and punchy for me,
and exceeds my expectations for brightness.
Summary
The RS20 is an exceptional projector and puts up a good fight considering it is 3 generations old at this
point. Without the benefit of the A/B testing I would have been harder pressed to realize the
differences. Although the VW95 is clearly better to me and worth the upgrade, keep in mind that it is
not a day and night difference so expectations should be kept in check (I originally went into this
upgrade thinking some of the differences, like sharpness, would be greater than they are).
Overall I am very happy with the VW95, and the direct comparison to the RS20 helped me appreciate
the VW95 even more. The characteristics that make the VW95 really stand out are its great looking
contrast and the way white really pops, satisfying blacks, its exceptional display of fine details, shadow
detail, and motion handling.
Screenshots
Attached are some screenshots I took of the VW95 using my iPhone 4S at a resolution of 3264x2448.
Three are in this post, then 3 follow in the next post. In order to fit
them here I loaded them in Photoshop, cropped them to remove most of the area outside where the
screen was, and resized them to have a width of 1280 (the max allowed here). The height will vary based
on how I cropped. Some resolution and detail was lost in the process (although the original pictures did
not pick up the fine details to begin with). No other changes were made.
Standard rules apply here regarding screenshots – they are just for fun and to give you a rough idea of
how things look on my screen. For the most part the colors are similar to what is in the pictures. The
pictures do not come close to showing off the real beauty that is on screen though. For example there is
a ton of black crush in the photos, many fine details (in the blacks and otherwise) are lost, they are a bit
out of focus (I did not use a tripod) compared to the real thing, and some dithering may have occurred
from the camera.
Any questions? Just let me know!
