CurtPalme.com Home Theater Forum

[kal]

Newer Barco's like the 's' series 808s, 1208s, 1209s, and Cine series (as well as my Barco Cine 8 Onyx clone) have the following service menu adjustments:

- Blue gamma slope (bar scale with about 6-7 ranges)
- Blue breakpoint (0-99)
- Red midlights (0-16)
- Blue midlights (0-16)

Anyone have any official documentation on what exactly what they do? The manual is somewhat useless. I've played with them and can offer the following comments but any additional info is appreciated:

From what I can tell, the blue gamma slope plays with the blue gain: When setting it higher, you get more blue in the higher IRE's. The blue breakpoint sets the 'breaking point' where the increased blue output starts from a range of 0-99. Setting a lower number seems to start the increased blue at higher IRE's only. Setting a higher number seems to bring the point where blue starts to increase down to lower IRE's. It doesn't seem to have much affect under 50IRE however, even with the blur breakpoint set to 99.

I've been playing around adjusting my greyscale recently and have always had a blue hump in the middle that dies down at the higher IRE's and goes too low like this (see the blue line):



If I increase the blue gain to pull the right side of the blue higher, the middle area goes even higher still. Not good. What I want to do is lower the blue hump in the middle and raise the right side of the blue so that a have a flat line.

By using the blue gamma slope and breakpoint, I've managed to only add some blue gain to the higher IRE areas (above 60-70 IRE) to flatten out the line. I'm still playing with perfecting this so no new readings yet or graphs yet. (The ones above were when I used a Minolta CA-100 to do greatscale - I'm now using a Spyder2 sensor with the free HCFR software - quite the bargain! Great greyscale for $60! I've heard that the Spyder3 ($165) is even more sensitive so it may be a better option for someone buying new).


Red and blue midlights seem to add a hump in the middle IREs Default is blue=0 (no hump) and red=7. Haven't tested midlights very thoroughly though since I don't need to use them.

Kal

[kal]

Newer Barco's like the 's' series 808s, 1208s, 1209s, and Cine series (as well as my Barco Cine 8 Onyx clone) have the following service menu adjustments:

- Blue gamma slope (bar scale with about 6-7 ranges)
- Blue breakpoint (0-99)
- Red midlights (0-16)
- Blue midlights (0-16)

Anyone have any official documentation on what exactly what they do? The manual is somewhat useless. I've played with them and can offer the following comments but any additional info is appreciated:

From what I can tell, the blue gamma slope plays with the blue gain: When setting it higher, you get more blue in the higher IRE's. The blue breakpoint sets the 'breaking point' where the increased blue output starts from a range of 0-99. Setting a lower number seems to start the increased blue at higher IRE's only. Setting a higher number seems to bring the point where blue starts to increase down to lower IRE's. It doesn't seem to have much affect under 50IRE however, even with the blur breakpoint set to 99.

I've been playing around adjusting my greyscale recently and have always had a blue hump in the middle that dies down at the higher IRE's and goes too low like this (see the blue line):



If I increase the blue gain to pull the right side of the blue higher, the middle area goes even higher still. Not good. What I want to do is lower the blue hump in the middle and raise the right side of the blue so that a have a flat line.

By using the blue gamma slope and breakpoint, I've managed to only add some blue gain to the higher IRE areas (above 60-70 IRE) to flatten out the line. I'm still playing with perfecting this so no new readings yet or graphs yet. (The ones above were when I used a Minolta CA-100 to do greatscale - I'm now using a Spyder2 sensor with the free HCFR software - quite the bargain! Great greyscale for $60! I've heard that the Spyder3 ($165) is even more sensitive so it may be a better option for someone buying new).


Red and blue midlights seem to add a hump in the middle IREs Default is blue=0 (no hump) and red=7. Haven't tested midlights very thoroughly though since I don't need to use them.

Kal

[danOO00]

Hi Kal

I had an almost identical blue hump when I first started doing grayscale with HCFR and the spyder2 on my NEC PG. A bit of additional electronic blue defocus flattened out the hump so its almost nonexistent. The extra output of a defocused beam will compensate for the loss at the high end.

I dont know squat about the barcos, but try a little bit of blue defocus each run through the grayscale until you get to the point where the hump is as flat as it can be without the blue halos becoming distracting.

Dan

[kal]

[kal]

[danOO00]

Did you try it out with blue focus at 100? I am not sure if the over/under focus matters at all, but I remember someone mentioning it in a thread a while back.

Also my electronic focus on the PG was not able to compensate enough for the blue, so I defocused it some more on the deflection board pot. This location is probably PG specific, but I would guess your set has a similar hard adjustment. If you aren't seeing blue halos, I would guess that you can defocus some more to see a difference.

[WTS]

Hi Kal,

That's one ugly blue line you got there and the red doesn't look that good either. I had to play with the blue and red adjustments too but in the end I got all 3 colours ruler flat for my grayscale. I don't recall exactly what I did as it was a while ago, my blue focus is set to the best focus not defocused. I use the HCFR probe and software.

[kal]

[kal]

[zGman]

Greetings,

I have spent a lot of time fiddling with my Barco's (G808SLC, P16 tubes)
and Colorfacts, and have achieved some decent results. Here are some thoughts.

You are right, the manual has almost no info - a little more can be picked
up in the service manual theory of operation.

The red and blue midlights are very powerful and key to good results.
Blue midlight will end up at zero, red (which operates in reverse) somewhere
around 4 to 6.

You shouldn't need much blue defocus at all, just a couple clicks. And it is true
that one way is better than the other, but its been a year since, and I can't recall.

I found that the contrast setting was totally basic to getting good greyscale,
and ended up with brightness at 42, contrast at 75 for adjustment - 72 for movies, etc.
If you try to run contast too low you will never get rid of the blue hump.

It is easy to chase the bias settings down into the weeds, and that will kill the top end.
In other words, if you find too much of one color at low ire, raise the others to compensate
instead of lowering the first.

It is best to start with the blue Gamma functions at factory defaults, but the midlights as above.
The meters can be very misleading on the lowend, the main focus is to get the G2 pots set correctly,
(never mind the LED's) and then use the lowend patterns on displaymate and colorfacts testpats
to ensure that all the tubes come up at the same time when you go from a 0 IRE to a 5%IRE
full screen (colorfacts testpats)

Once you can get them to come on at the same time, then you can step through the
fields and see if the lowend is roughly grey. Small Bias adjustments can make a huge
difference - and interact with Gain, so don't try to fine tune the bottom end until
you have dialed in in the top and midrange.

I wil be working on Barco's again soon, so maybe we can come up with a more
refined procedure.

Cheers,
Galen

[kal]

Thanks for the input Galen - very good advice all around.

[kal]

Thanks for the input Galen - very good advice all around.

[kal]

[kal]

[zGman]

Hi,
OK, that's a lot of questions, but more fun than actually working....hohoho!

1st, yes - very backwards, bias, gain, cut off, etc - mostly Marquee stuff for a year!
However, my all notes show cutoff for lowend, and Gain for high end for Barco.

Midlights - very much control middle part of response curve, not just 50IRE, but(!)
red works backwards (eg. up is less, down is more) So, I always end up Blue=0,
red about 4 or 6.

Yes - set G2 pots with default settings, and brightness/contrast at 42/75.

I look into the lenses to set G2 pots(thanks Terry), the meter cant see it.
with the lowend block pattern from colorfacts testpats look for the lowest
block (2) to blend with the background and the second two (4) to just stand
out from the background.

The tube will not be completely off in the background, nor will the tubeface
be completely off at 0 IRE. (I consider this "idle", the tubes have some
excitation but no light is being projected. You can test this with a lens cap.)

The correct setting will have all tubes come off idle at the same time from 0 IRE
when you go to the 5 IRE full screen white pattern, and hopefully not too
far from grey - but really its hard to see color at 5% or 10%, 20% is where you
can make a difference with adjustments and have some small chance of getting
a reading - but trust your eyes more. Displaymate has color block patterns,
and you can use the "low red" "low yellow" low green" and "black" blocks to
see if you are tracking right at the bottom.

I know that you may find 65 Contrast setting to be "plenty bright", but that's
not the point. The high end (Gain) settings will determine the actual tube drive level.
But what I have found is that the control circuit for Gamma seems to be calibrated
to a contrast setting of 75 or 80, so if you are only running 65 then your gamma
settings can't track properly. It is possible to set the Gains to get high output at
lower contrast settings, but it won't track. This is not a random observation, but
something I worked on for hours and hours.

I found the only way to get a greyscale within a couple hundred
degrees of 6500 from top to bottom, was to set those defaults of 42/75, set the G2 pots,
and then set the overall Gains with a colorimeter and a 90 IRE input to get 6500 approx.
Then you can check your white crush with the appropriate colorfacts test pattern.

Another thing that's important, is that its also possible to push the high end too far,
you can max the tubes out at 100 IRE, and then response goes nonlinear. In other
words, say I get 7 FtLamberts at 100 IRE, 6.5 FtL at 90 IRE, and 4.8 FtL at 80 IRE.
This is what I would call "crushed white", very hard to tell the difference between
90 and 100 IRE.

If you want to see the Blue gamma controls in action, put up a horizontal 20 step
pattern full screen and try sweeping the breakpoint, sometimes you can see the
blue shift through the pattern. I usually end up with breakpoint at about 1third scale,
and slope about 2thirds scale.

Let me know if this is helpful, I may fire up the G808S LC tonight or tomorrow
and get some numbers for you,

Cheers,
Galen

[kal]

[kal]

[r.bauer]

[perisoft]

[zGman]

Yes, that's true - both PJ's had new tubes, and could be very bright.
Are you saying, that for an older tube the solution might be to reduce
gain for green and red so that the blue can keep up. I could see that if
the blue was reaching an inherent limitation in output, there wouldn't
be any headroom for the blue gamma circuits to work properly.

Kal, your luminance curve looks pretty good to me. Good luck
with the greyscale tonight!

I am trying to make some of the stupid little bent metal (.030")
catches that go in the scan reverse switches, they fall
out when the silly little plastic spring retainer breaks...
Thanks - Barco!

G

[WTS]

Hi Kal,

It's been quite a while since I did the initial gamma setup where I adjusted out the humps using the midlights and the gamma curve adjustments. But I do recall it took me a lot of back and forth and trying different setting to finally get each tube flat lining. No I never did keep a file of the results even for the last time I tweaked the gamma. Normally I just setup the probe and fire up the laptop then set it for continuous sampling then go up and down the scale from 0 to 100 ire and tweak the gains/bias's as I go until they're all flat and overlapping each other from 20 to 80 or so then it's good to go. It doesn't take me that long to just tweak things.

I agree 100%, the factory setting are terrible and a properly setup greyscale makes a huge improvement.

[kal]

[kal]

[r.bauer]

[kal]

[kal]

[zGman]

Hi Kal,

I just do the G2 pot setup while inputting the test pattern, I haven't seen the "compensation"
effect that you spoke of, and haven't had any problem getting good results....unless of course
there really is some difference in these machines beyond the branding change....?

Yes, long arms and a steady hand...hohoho! not fun, especially balancing on a HT recliner!
actually it is helpful to bungee up the cover plate, and sometimes get a bit of help to make
certain the driver stays centered on the pot...

You asked about the colorfacts test patterns...I can send them, but at one time they were
available on the website. There are several very nice features, and the lowend pattern
has much finer graduations than a typical step pattern, which allow you to see more exactly
when the tube is just off "idle" (ok, ok, so I'm a car guy....)

The other thing I like is the 5% increment full screen patterns in R,G,B and white, which can be stepped
through using the arrow keys on a keyboard, which makes it easy to see when the tubes come on,
and easy to measure the relative luminance of each IRE step.

Displaymate is another great set of patterns, I can't recall where I found those, but it is very handy.

One thing that I had meant to mention, is that I would start all these adjustments with the focus, astig,
and convergence as perfect as possible. Light output is very dependant on these, and the blue defocus
should be just an intermediate step, after things are roughed in, and should not be more than a few clicks.

I still have a little work swapping lens bases to get my BG808S LC fired up, and the other BG808S
that's here has decided to act up, no doubt jealous of its neighbor with new tubes.....jeez...!

Cheers,
Galen

[WTS]

Hi Kal,

All I've ever done is adjust the G2 as per the book before I do the gamma. THe G2s rarely drift off more than a hair on my cine8, I doubt yours is any different.