PCA100 USL colorimeter software

[Everson Soufer]

Hello everyone

First, thank you for accepting my entry on the Forums page. I will try to contribute as best I can to the challenges we face in Brazil with digital cinema.

I have an old USL colorimeter, model PCA-100, but I can't find the software for Windows XP.

I have the software for Windows 7, but it doesn't work well.

Is there any way to install a version of it on Windows 10?

Thank you all!

[Caleb Williams]

Have you tried installing the Windows 7 version using compatability mode on Windows 10? I've moderate success running old applications meant for Windows xp and 7 that way

[Ed Gordon]

I don't have the device you are using but I have run into problems with old devices not working on a new OS. I was able to install the software, but the device was not recognized. I ran device manager (devmgmt.msc) and windows 10 did not support it.. I found a device driver made for windows vista that worked on windows 10.

What version of the software do you have? Version 1.3 was release in 2008 and version 1.5 was released in 2016. These were developed by Ultra Stereo Labs which has been acquired by QSC. From QSC I see the following:

FAQ | How do I get the drivers for the USL PCA-100?

Create a case through your QSC account and include the serial number with your request. We will build the driver file for your specific unit and send it to you with a secure file share.

Link: https://www.qsys.com/resources/news/...-labs-inc-usl/

Windows 10 USB Driver Signing and certification requirements for device drivers installed on the computer require selecting a mode that will bypass the signing requirement.

The USB driver used by the PCA-100, JSD-60, and JSD-100 are native to the operating system and use a co-installer. With Windows 10, the co-installer will not function unless the driver signature enforcement is disabled. The easiest way to resolve this is to follow the instructions provided by the Win10 USB Driver Install Guide.

[Gordon McLeod]

I use virtual box and run it in VirtualBox with xp .. the driver is by Ocean optics and they don't support it it is called omnispam. the calibration files are specific to the meter

[Everson Soufer]

hello everybody

@williams, I've already tried using it in compatibility mode but it didn't work.
@Ed,Gordon, My version is 1.3, but I'll try this possibility you mentioned.

@Gordon,McLeond I say an XP virtual machine, but I hadn't tried this Ocean Optics Driver issue, I'll try this possibility too.

I'll try these suggested methods and get back to you as soon as possible with answers.
Thank you all for your support.

[Gordon McLeod]

hello everybody

@williams, I've already tried using it in compatibility mode but it didn't work.
@Ed,Gordon, My version is 1.3, but I'll try this possibility you mentioned.

@Gordon,McLeond I say an XP virtual machine, but I hadn't tried this Ocean Optics Driver issue, I'll try this possibility too.

I'll try these suggested methods and get back to you as soon as possible with answers.
Thank you all for your support.

The driver from ocean optics is called omnispam. It installs itself with the PCA software

[Harold Hallikainen]

The PCA/PSA/LSS were interesting products at USL. There was a woman who sat in a dark room all day every day calibrating these instruments. I did a prototype of a new PSA, but it never went into production. We were evaluating spectrometers for an updated PCA, but that also never went into production.

[Steve Guttag]

It's a shame because the PSA has proven to be incredibly useful, even in digital cinema. The ability to have 45 light readings (continuously updating) is incredibly useful when adjusting a mirror and showing the effects of gain screen (and why they decrease light, not increase it).

I recently had a conversation with someone that should know better telling me to use a 2.2 gain screen as the recommended brand didn't have a hot spot problem. I inquired as to how they got around the whole physics of how light works and could they provide (essentially) a document showing what the light uniformity is and does it conform to SMPTE ST431-1? For cinemas, the sides of the screen are to be, at least 75% of center (ideally, it is 80-90% of center) and that I've NEVER measured a screen with a gain over 1.3 that could meet it.

I'm going to use Harkness' catalog just as an example...I'm not picking on them as it could be any screen manufacturer but here is a chart showing their gain screen surfaces and the fall off:

image.png

With their 2.2 gain screen, how far off center can you be to have 85% of the light from center? From their chart...about 14-degrees. Maybe, if vertical-video becomes a theatrical thing, that could work. Naturally, with a long-narrow theatre, you can get away with it more because of the angles of incidence are very shallow and will reflect equally (and oppositely) shallow.

Dropping to a 1.4-gain screen you get out to about 28-degrees.

And this is what a 1.3 gain screen can do (Stewart and trust me, others don't do as well) as represented by the PSA-200:

image.png

If I'm balancing machines for a changeover performance, the PSA is again, invaluable as you can adjust the projectors in the booth with the meter down in the theatre at the proper position.

On a flat screen, the highest gain you can use, and meet SMPTE specifications is 1.3. Anyone using 1.4, 1.5, 1.8, 1.9, 2.2 and beyond are, excessively, hot-spotting. And, furthermore, if you are setting your digital projectors to 14fL (48cd/m²) are putting out a darker picture than people using a matte-white screen also setting the center lumens to the same level. All a gain screen does is concentrate the light, not add to it.

If you have a "typical" silver 3D screen in the 2.2-2.4 gain range, this is a typical PSA type reading:

image.png
Over half of the screen is below the SMPTE minimums. That is why silver screens appear dark when the center is set "to spec." The PSA shows it pretty clearly.

Now, you CAN curve a high gain screen...even a silver one.

Here is a silver 2.3ish gain screen, curved:

image.png

If you look across the center row, it is about as uniform as a matte-white screen would be. However, the top and bottom show the gain aspects as there is notable falloff.

Here is the actual screen:

image.png

There isn't any free in this world. Curving the screen does bring about geometric distortions to the picture (barring custom lenses or even electronic warping, that would lose resolution).

I quite liked the PCA-100 too though it struggled with Dolby 3D calibration on red. The VCC-102 was a time-saver as well. Convergence with lasers is proving difficult as what you see from the booth, even with decent binoculars isn't what you see closer to the screen.

[Harold Hallikainen]

The PSA replacement project was right when I started working at USL in about 2007. The new product involved a monochrome camera chip. I wrote code to largely duplicate the PSA. It would put different grid patterns up on the display, and you could adjust the height and width of the grid. It would sample 9 pixels in each of the grid rectangles. These would be run through scaling factors to get luminance. There were a large number of scaling factors, one for every 16 pixels or so. The differing scaling factors compensated for varying sensitivity of the pixles of the camera chip and errors introduced by the lens. Data could be logged to an SD card as csv for offline analysis. I think the project got sidetracked by the JSD-100.

In ISDCF screen luminance tests a few years ago, I remember seeing another instrument that measured luminance at various points over the screen. It was interesting in that it also gave you a "contour map" of the luminance so the hot spot was a circle with concentric circles around it. I have not yet found in my notes who made it.

Some ISDCF notes on luminance measurement include:

https://files.isdcf.com/papers/aa%20Old%20Versions/ScreenBrightnessMeasurementv1_0.pdf

https://files.isdcf.com/MeetingNotes...s-at-ISDCF.pdf - This makes reference to a CR120, which I think is the instrument mentioned above. However, https://www.colorimetryresearch.com/ does not list that model.

[Marco Giustini]

I recently had a conversation with someone that should know better telling me to use a 2.2 gain screen as the recommended brand didn't have a hot spot problem. I inquired as to how they got around the whole physics of how light works

I also had a similar meeting with someone who "should know better" during a 3.2gain screen installation which was being installed to bring "more light".

100% what Steve said, I also like to re-post these pictures when the "high gain" screen subject comes up: these is a rollable matte white screen (TOP) in front of a Silver screen (BOTTOM). It was a test screening room so no sound concerns here. The pictures speak for themselves.

silver1.jpg



silver2.jpg

[Tj Hopland]

How stable were the PSA and LSS products for holding calibration? How about the PCA?

[Steve Guttag]

Rather stable, has been my experience. I would shoot them against other meters.

The LSS meter was never meant to be a precision meter but a comparative one. It is a sub-$1000 meter (including luminance, chrominance and SPL) versus meters that sell in excess of $10K for just the color part. However, as for tracking a projector over a period of time, it was perfectly suited. The PCA stability was solid (particularly after the first generation). As I mentioned, it wasn't the best for Dolby 3D calibration on the red color but it did okay and was my go-to for the bulk of my conventional cinemas during the big Cinema conversion.

I like the Photo Research PR-655 the best of them all. It is incredibly stable and its results are the most repeatable of the meters I've dealt with. For laser projectors, it needs a modification to barely meet specs but still, I trust its readings the most.

[Tj Hopland]

What's supposed to be the issue with lasers vs xenon when it comes to meters? When I have had chances to do side by side comparisons colors have seemed to match but the light reading was off. Have not had that many opportunities to make those comparisons or see what the differences may be with different projectors with the different types of lasers. Perhaps there is a combo of meter and laser type that is/was more of a problem? Do the 'laser approved' meters work equally well on all laser types? How about on xenon?

[Steve Guttag]

The problem is spectral bandwidth of laser light being so much smaller than just a prism separation of white light. Think of it (and exaggerated) of trying to resolve a 1080p image on a 480 monitor. The older/lesser meters have a special bandwidth around 7-8 nm. The "laser approved" ones are under 5nm. So, a meter like the CR-250 just skates in at 4nm. The more expensive CR-300 is 2nm, which is why it would be preferred for screening rooms and other more-critical environments. My personal favorite, the PR-655, in standard form has an 8nm sensor but it can be modified (at the factory) for a 5nm one that just barely meets laser spec. The Qualif (HSG) Specro has a 1.8-2 nm bandwidth (depending on the brochure you read...but sufficient for laser). The Spectro has a xenon/laser selector for better readings from each light source.

[Harold Hallikainen]

Color measurement is interesting. I wrote a paper about it for the LSS-100P. The LSS-100P used an RGB sensor for color sensing, and a photopic sensor for luminance measurement. The photopic sensor has a filter on it that converts the "frequency response" of the silicon sensor to that of a photopic sensor by attenuating wavelengths as necessary to end up with a photopic curve. The RGB sensor values are run through a color correlation matrix to convert them to XYZ values. The matrix is developed during system calibration.

The LSS-200 uses an XYZ sensor for everything (almost). This sensor has filters on it to make each of the silicon sensors inside the chip have an X, Y, or Z spectral response. The sensor values are again run through a color correlation matrix resulting in XYZ values in cd/m^2 for each. They Y value is the luminance. X,Y,and Z are shuffled around to yield xyY where xy is the x,y color value, and Y is the liminance.

The spectrometer based instruments measure the light every 5 or so nm, then multiply the sensor values by coefficients and add the results to yield the XYZ values, similar to the LSS-200. Through this process, the XYZ curves can be replicated very accurately (just adjust the coefficients in calibration). The color measurement accuracy of the LSS-200 relies upon how good the optical XYZ filters are instead of being able to adjust the response as desired (like in the PCA).

On spectrometer instruments, the sensor output value is integrated over the bandwidth of the sensor (like 5 nm). If the bandwidth of the light is only 2.5 nm, the sensor value is still multipled by the 5nm, so it could show up twice what it should.

The LSS-200 avoids this issue, and was found to be quite accurate with laser projectors (I tested it at Real D in Boulder CO). But, we are, again, relying on the optical accuracy of the filters on the sensors.

As mentioned before, color measurement is interesting!

Harold