Page 2
We count so little that Projector Point categorise their line-up as:
Home Cinema
Office
Education
Ultra Portable
Large Venue
They do have a sub-category called Camera Clubs on their website but they
immediately lose brownie points by suggesting that most camera clubs should
go for DLP (on the grounds of cost). This is despite the practical
experience that the majority of camera club enthusiasts have perfectly good
eyes and are as critical as any professional photographer when it comes to
viewing digital images. When it comes to colour fidelity, LCD and LCoS win
out over DLP by massive and measurable margins.
In order to refine the choices we sorted and assessed the projectors
according to the main performance criteria.
Resolution
Of the 380 projectors looked at, two-thirds were XGA with the remainder
being WXGA or a combination of the new HDMI television standards if 720p or
1080p (mainly for the home cinema types).
For photographers the only resolution of interest is XGA at 1024x768. This
ratio of 4/3 matches many cameras. WXGA is 1366x768 (a ratio of 1.8:1) and
does not match cameras at all, it is the wide-screen standard ratio of 16:9.
This is a total waste for photographers as the space either side of the main
subject area is redundant, not like a monitor where you can use it to store
palettes. The price premium for wasting space by going to this format
averages at 37%.
Restricting the choice to 4:3 format projectors reduces the candidate
projectors to 41. If you then eliminate all DLP projectors on principle, the
number drops again by about 50% – we had no way of knowing exactly with the
database we were using.
Matching the Graphics Card and Projector
This is a topic that does not get a lot of discussion. If your editor's
experience is anything to go by it is vital. My own laptop is an SXGA+
format (1400x1050 pixels). Using any resolution either side of this value
severely compromises the quality of the output to the laptop monitor.
Indeed, the loss of sharpness as you move to the lower resolution is exactly
counteracted by the change of icon size due to the resolution change. This
brings problems when it comes to projection. Most projectors are happiest at
1024x768 pixels resolution whereas my laptop graphics card likes 1400x1050,
a 'no win' situation. A number of the federations that govern the enthusiast
competitions have adopted SXGA+ as an additional, acceptable standard. How
this will impact upon images projected at a downscaled 1024 SGA format is an
unknown quantity, if your editor's experience is anything to go by it will
be poor quality. The move is to a certain extent governed by the popularity
of the Canon projectors available at SGA+ (in say the XEED SX60). It is
noteworthy that although there is a 1024x768 pre-set in the Photoshop
File>New dialogue box there is not one for SGA+ – perhaps the situation will
change as Adobe cotton on to the moves in the clubs.
The on-screen quality can be further compromised by use of keystone
correction. Of interest here is the directive from the Photographic Alliance
of Great Britain that keystone correction should not be employed when
projecting images in competitions under their jurisdiction. To some extent
it depends on just how the correction is made. With only 1024 pixels to play
with, any interpolation to change the shape can be detrimental. Far better
are the optical corrections made via the glassware. Lifting the projector up
to centre screen level* brings in a number of issues in regard to visibility
for an audience as the throw of most projectors is short (ie they need to be
close to the screen).
The wide-screen laptops add further complication resulting sometimes in
partial screen coverage to either projector or laptop screen or, even worse,
everything on screen being squashed into the format. The moral of the story
is to think hard about the various formats that you employ and decide which
should be the overriding factor, then purchase
accordingly.
*This is not centre-screen level, most projectors have built-in optics such
that placing the projector about a fifth of the way up the screen renders a
geometrically rectangular image which covers the screen.
The Screen
You should avoid the more sophisticated 'beaded' screens – they can create
very distracting moiré fringing. The plain matt white screen is much
preferred for digital projection and a painted wall will suffice for many
applications.
In Practice
It is difficult, indeed almost impossible to check out a projector before
you purchase one and the staff at most non-photographic outlets are likely
to display a disappointing ignorance of the requirements for our
applications. We can only report on what we have found with the Epson 1810
as this is a projector we have experimented with. Here is what we
found:
The greatest discussion that you hear when digital images are being
projected in whether or not this or that highlight has been blown out.
A key measurement when viewing a projector should be the point at which
differentiation no longer occurs between highlight detail. To do this we use
the 'highlight' detector of our standard colour audit test chart. Next comes
the shadow detail retention as measured by the 'shadow detector' in our
audit chart. Both measurements are influenced by the projected image size
and we blow up the portion of the target when viewing it. The
differentiation at full-screen-zoom sizes is far less, and the ability to
differentiate from the back of a hall is far less than if you are sitting in
the front row. In terms of colour, the skin tones are usually critical and
influenced by the selected colour temperature.
We started by projecting in a fully darkened domestic room with the
projector at factory defaults, which included zeroed settings for
Brightness, Contrast and Sharpness. The chosen colour temperature was 6500°K
and the sRGB setting for colour space. The shadows were differentiated down
to 15 RGB points, a value that went up to 25 RGB points (ie worse) with
subdued room lighting (of about 80lux – for reference this is the level most
people might use when eating dinner or viewing their television; office
lighting is usually about 200 lux). The highlights were differentiated up to
246 RGB points, a reading that was unaffected by switching on the room
lights. Changing the projector from sRGB mode to Photo mode did not affect
any of these readings.
Next we adjusted the projector contrast and a setting of -19 brought the
highlight differentiation up to 252 RGB points (where we would like it to
be) while retaining the 15 point RGB shadow differentiation. At this stage
we also adjusted the sharpness setting, deciding on a value of -2, although
it was not a very critical change.
Using the settings above we profiled the projector using the GretagMacbeth
Eye One and Beamer using Eye One Match. We choose a gamma of 2.2, a colour
temperature aim point of 6500°K and then profiled the set-up. This retained
our shadow and highlight values of 252 and 15 RGB points but we considered
that the skin tones were a little too saturated.
We then 'reset' the projector and profiled again. Without further adjustment
this gave shadow detail down to 10R GB points and highlight differentiation
up to 250 RGB points. Although the skin tones were attractive, they were a
little 'hot' for our taste and putting the colour temperature up from 6500°K
to 7000°K reduced the saturation (as it forced everything a little more
blue, ie cooler). We then dropped the projector contrast setting to -10
units which brought the highlight separation to 252 RGB points and produced
shadow separation at 5 RGB points. This is slightly better shadow separation
than we achieve with a profiled print, which rarely separates shadows at
better than 15 to 20 RGB points without intervention.
The gamut volumes were measured on the profiles at 993,808 for the first
settings and 1,002,423 for the last set. These values are very close to
those we measured on the projector when we first tested it about 12 months
ago, and in line with typical LCD values. They are much greater than DLP
values.
Our experiments showed that it is worthwhile carefully examining the
projected image and making adjustments to the settings of the projector even
after it has been profiled. While this is counter-intuitive after you have
paid out money for a projection calibration device/ software, it is what we
found!
http://www.pagb-photography-uk.co.uk/misc-pdfs/standards_final.pdf
