How to set a CRT to 12ft lamberts?

Status
Not open for further replies.

Peter Parker

Distinguished Member
Joined
Apr 6, 2001
Messages
14,739
Reaction score
5,002
Points
2,878
Location
UK.
I had posted a question regarding 12ft lamberts on the digital forum on avs, and it occured to me that I wan't sure how you guys would go about doing this with a CRT.

At the cinema they should (but don't always) set the projector up so that without film in the gate, you will get 16ft lamberts of reflectance from the screen. With film in the gate, it should be around 12ft lamberts - I assume for a full white field.

There are various ways of doing this with a digital pj and I think this is imperative IMHO or the image can look more like video than film.

It's less critical for a CRT, but how would you set a CRT for this level of reflectance? Can you drive the tube contrasts to get a certain brightness level and how does that correlate to setting the white levels with a test disk?

Gary.
 
Thank you for bringing a new word into my vocab

the "lambert" , never come across that in the past but I understand your question and will try and answer.

"It's less critical for a CRT, but how would you set a CRT for this level of reflectance? Can you drive the tube contrasts to get a certain brightness level and how does that correlate to setting the white levels with a test disk?"

drive tubes to get a certain brightness level? - yes with the brighness control BUT hi bright will give an over white image. Much more complex is to use the G2 settings which drive each tube but next answer I think is more what you were looking for.

White levels on a test disk? - you can set the colour temp you require so I guess run the test disk and then calibrate the k temp to suit using a colur balance meter. most crt projectors I know have built in colour balance settings, some have user option to define your own.

level of reflectance? - Usually this can be achieved by the screen gain as CRT's are not know to be bright!

Hope that is of some help to you. Basic I know.
 
Thanks for the input MrH - much appreciated - and basic is good right now. :)

ft lamberts are just a measurement of reflectance from the screen, and you can measure it with a light meter if you wanted to know what you were actually getting. Crteaman mentioned that he got a pretty bright image on a 14ft wide screen which made me wonder about this adjustment.

Test disks like Avia or Digital Video Essentials have some test patterns on them so that you can set the white and black levels correctly - set them wrong and you can lose shadow detail or bright detail like clouds in the sky.

There is also a full white field so that you can project a full white screen and that's how you can measure and determine the maximum ft lamberts and use it for on/off contrast (with a full black screen).

Normaly I'd use the contrast control to set the white level using a test disk, and that would then be a fixed setting otherwise the image could be compromised for white detail - I think the bright white you mentioned would do the same thing by over-driving the contrast so I can see why you wouldn't want to use it.

Using the individual tube G2 controls - are they individual brightness and contrast settings (gains and cuts) for each colour/tube? I guess you would adjust the colour balance with those if the colour temp was off (using a colorimiter to get it correct).

Screen gain makes sense if you don't want to over-drive the contrast control (it's the same for digital). What's the max gain a CRT can get away with before it can cause hotspotting or other image problems? Since you don't have a lamp, I assume you can go much higher than 1.3 gain?

Thanks

Gary.
 
What we should be seeking to emulate with our pictures is reality not film,when a movie is filmed its the limitions of film that prevent accurate portrayal of reality
Remember a cloudy day is 100-300 ft lamberts!!! :smashin:

.(Most movie projectors unless regularly serviced struggle to produce 12 ft lamberts..)


What is thought to be a current ideal by ISF for the home is 20 ft lamberts-though SMPTE spec right upto 22 ft lamberts.

Now,can you achieve this with crt?

Yes..A G90 firing on a resurfaced 11 foot wide torus with Stewarts latest 2.8 gain screen, calibrated to D65,contrast 69 -achieved 25 ft lamberts +

(*With a normal 1.3 gain flatscreen the G90 can exceed 14 ft lamberts and retain good linearity as long as limited to 72 ins wide )

This is where the Sony G series crt`s are good..but the wider beam spot creating more brightness is a tradeoff of ultimate resolution in comparison to something like the Cine 9 with a tighter beam spot.(Which also is no slouch in the brightness stakes!!)

It is hard though to describe the average ft lamberts produced by the average crt in the average room..there are so many different factors to tak into account..screensize,lighting etc..etc....

Suffice it to say that the average 8" crt would on a 72ins screen, pan out at about 7-8 ft lamberts if calibrated to D65 for instance..after 3,000-4,000hrs of use? about 6 ft lamberts.

Now,in a good light controlled enviroment on say a 72 ins screen -8 ft lamberts is actually fine,whats oft forgotten is that its not JUST about brightness..in some respects our eyes are tricked into thinking the image is brighter than it is with crt because of the awesome dynamic range-which is unmatched by ANY other technology at present.

Contrary to what some think a double stack of crt `s are not vastly brighter..but the blacks are vastly increased-tricking us into thinking its brighter than it really is! :thumbsup:
 
Projectiondream
How can a wider beam spot produce more brightness? Surely it will do the opposite!
The way you are setting up your double stacked projectors must be wrong. the blacks should be about 3 lumens higher (not quite vast) and your whites should increase by 50%.
 
Hi Roland,

Interesting point - I was wondering about stacked CRTs the other day (Art Sonneborn over on AVS has two G90s) and wondered how the black and white levels are set in relationship to each other - I assume you stack for more brightness and/or a bigger screen, so the white level should increase noticably (I would guess at double but you say 50%), but you want the black level to remain the same, so is the lumen increase from black to white non linear? Ie, less increase in black to preserve full black out and more increase for white to allow a larger screen or brighter image. Is that possible or do they just double the lumens equally from black to white, or is just by 50% as you mention above?

I assume this also means that if the white and black levels are set correctly, and you increase the white level by 50%, the CR should increase as well:

One CRT producing 300 lumens at the screen for full field white (for arguments sake), and measures say 8400:1 calibrated CR (using Willam Phelps G90 figures) with black set for good visible shadow detail, so you will have 0.036 lumens for black.

Stacking to increase lumens by 50% gives 450 lumens full white and if you can keep full off at 0.036 the CR will now be 12600:1. If the black level increases by the same amount (50%) then the CR stays the same, but I assume you'd just set the black level on both CRTs down (using a test disk?) so that should keep the black level the same as it was with one CRT.

Why wouldn't you be doubling the lumens by stacking? Is there a reason for that such as reducing tube wear perhaps (just a guess).

TIA

Gary.
 
I'd have thought that if both projectors are working at same output capability then stacking would raise black level and white level equally. If you use the dual projector capability to drive them both less then I can see why you wouldn't get twice the light output....

The size of the beam spot would largely seem irrelevant to me with regards to ultimate light output from a CRT. The smaller the beamspot the more lines you need to draw to energise the same amount of phosphor as you would with a larger beamspot. Either way the light output is related to how much phosphor you energise and for how long not how big the spot is that's drawn across the face of the tube.

Gordon
 
If you set the contrasts individualy so that when combined you'd have double the peak white level, but set the black level of them combined you should in theory be able to achive the same black level as the single pj shouldn't you? That way you double white level but keep the black level as before and double the CR.

Can it work like that and is the increase in CR visible?

Gary
 
I'm not sure Gordon - I'm assuming that if you were not crushing at 0.036 lumens for one projector, you'd not be crushing at 0.036 lumens for two projectors. Of course I don't know if 0.018 lumens from each would actually be displaying any detail so doubling up to make 0.036 would still result in crushed blacks so you make a good point. William's CR was from a G90 that had been set for good shadow detail rather than overall CR so it would probably be fair to say that a lesser black level would indeed be crushing black.

I hadn't considered the gamma curve and that might look kinda strange in the low ires though if crush was occuring. Or are there other gamma aspects to consider as well?

Fascinating subject. Have you and Roland done many stacks and other than lining them up, is it just a case of setting them up as two seperate projectors rather than as a single projector with twice the controls so to speak?

Gary.
 
Hi Gary,

I promised I would interject, there is a lot of confusion and incorrect information here, I have done stacked units and will have stacked nines soon. ;)

Speedbird( and a few of the local crt guy's) got to see the stacked 8" units just prior to the first event and you may want to ask his opinion, I think it runs along the lines of "wow" and "jesus" etc.

The dark/black scenes using just one unit were all very well but when we popped the lens caps of the second unit....wow! there was so much more info in the blacks you never knew were there, and as for peak whites...

Hope 'Speedbird' is looking foward to the nines :thumbsup:

Its getting late now and things have been beyond busy so I will try and add to this over the weekend, adding my take on all of this, I will try and address your original question regarding FL and crt's.

james
 
Hi James,

Glad you could make it! (I know you've been busy..). This thread has become quite interesting (for me anyway :) ), so if it wanders from the original question I don't have a problem with that.

Gary.
 
Roland @ B4 said:
Projectiondream
How can a wider beam spot produce more brightness? Surely it will do the opposite!
The way you are setting up your double stacked projectors must be wrong. the blacks should be about 3 lumens higher (not quite vast) and your whites should increase by 50%.

Must be wrong? blimey...well we can all be wrong Roland :)
But..
Let me clarify here,Sony were notorious for defocussing the blue to gain more light output..all the G series levels seemed too high..very bright..but I thought at the expense of ultimate focus..now how much the beam spot comes into it..I would think another "ruse" to increase brightness-just my opinion.

I think the wow factor comes from the greater blacks and the perception of "more detailed" detail :)
The gain in brightness was not as much of a noticable factor to ME-but then the human eyes is receptive to contrast and arguably notices more of a "dynamic" in increased difference between the absolute black and the absolute white.


But 20-25 ft lamberts would be great-and can be achieved..Why then constrain oneself to aiming for 12?
I think that was my main point!!! :hiya:
 
Just thinking...

It's possible that when removing the lens cap from the second pj, you raised the black level (more lumens) so made more shadow detail visible.

I've seen it said that many CRT owners will crush black detail to achieve the full blackout effect at the expense of good shadow detail as it's a preference. William Phelps calibrates for good shadow detail which is why his G90 CR is around 8400:1 - he raises the black level a little so that detail is visible. To achieve full blackout and good shadow detail at the same time, you may need to use an HTPC with modified gamma curve so the lumenance above 0ire ramps up at a differnt rate compared to the normal gamma curve of 2.5 (I asume that's the aim of CRT gamma?). I'll see if I can get a pic of Verns gamma to see how he achieves it. I wonder if his gamma has a hump in the very low ire's rather than a sweeping curve.

Gordon - when using DVE to set black level on a CRT, do you set it for below black and how does that effect shadow detail? If you set it for 0ire/D16 do you gain more detail and less blackout so a reduced CR?

Gary.
 
Gary Lightfoot said:
I've seen it said that many CRT owners will crush black detail to achieve the full blackout effect at the expense of good shadow detail as it's a preference. William Phelps calibrates for good shadow detail which is why his G90 CR is around 8400:1 - he raises the black level a little so that detail is visible.

Personally I like to see as much shadow detail as possible and find the full black out effect less important, so my black level is set a little higher than some.

I don't have an answer to your original question, but you're welcome to bring your light meter along and take some measurements.
 
Hi Henry,

I think I would tend to do the same as well so that you're not missing anything.

I think there's a good example of this from the opening of Star Wars where Darth Vader enters the rebel ship - the background is bright white, and the head/torso close up of Vader is of course very dark. If the brightness control (black level) is set too low, blacks will be crushed and you'll not see Vaders mask or chest plate detail correctly. This is where a test disk such as Avia or DVE will allow you to set the correct levels so that you are seeing what is intended. The THX Optimode is mainly designed for the disk it's on, as it is often set to show below black detail, but it will get the display into the ballpark so to speak for other movies.

I'll bring my light meter along and see what kind of light output you have. I don't think achieving 12ft lamberts is critical with a CRT as I've found that a decent level of contrast can give the impression of more brightness than is actually present (in a light controlled environment). I've seen a very watchable image from a digital with only 4ft lamberts and was very surprised at how 'dim' that measured. It makes me wonder how long the eyes have to become accostomed to a dark room before the eyes can see dark images properly. If I walked into a blacked out room from normal daylight 4ft lamberts would probably seem a lot dimmer than if I'd gone from somewhere darker or had longer to become accostomed to the reduced light output. That pj was achieving just 0.0018ft lamberts of black which compares favourably with the G90 at 0.0014ft lamberts if it was producing 12ft lamberts at white (if it was 10ft lamberts then black would have been 0.0011). Of course the CR was lower on the digital (as was the peak white) so the shadow detail wasn't there like it would have been on the G90.

I've heard it said that cinemas deliberately make you sit around in the dark so that you become accustomed to the reduced light. It also gives them a good excuse to show a few adverts. :)

Gary.
 
There are many contributing factors that need to be considered and understood before arriving at the CRT’s ‘useable’ peak FL (light output), and note as with resolution they all have a perfect sweet spot.

Vince is indeed correct on the subject of using defocus to gain more light but this only works on blue, never forget how lazy the human eye is to blue. Anyone who has observed /calibrated or just plain played with the blue focus (mainly electrical focus) on a CRT will have noted the point of peak focus is actually the dimmest.

Units such as the Sony G series actually use defocus on blue, after you have spent ages focusing it then applies X amount of points of defocus (unless of course you know that number to trick its circuit).

Almost every part within the PJ will have an effect on its peak light output, power supplies and focus design are at the forefront of the contributing and determining factors.

For example, I often work with two particular crt units both of which use the Identical tube type the 9” panasonic tube yet one will allow you to hold focus all the way to 100 (max) on contrast/light output whilst the other starts going soft rapidly as you push it over the 60’s.

Whilst pure FL/ light output are important there are two other very contributing factors to consider, the first is contrast ratio, “riddle me this” “which projector would look the best, out of the following?”

Number (1) whose light output is 20 fl but can’t properly resolve the bottom 30 IRE of a grey scale or Number (2) whose light output is just 10 fl but can achieve near true blacks?

So which is it, which has the greater contrast ratio? We must notice here that the challenge of maintaining a high contrast ratio occurs in the amount of the min value and not, as might otherwise be supposed, the Max.

The crt theoretically can be made to produce the very same level of luminance as a slide Pj, so the Max does not need to decrease.
It is the increase to the Min which debases the contrast, even the most minute increase in the min will have a dramatic effect on the overall ‘ratio’

It is my firm belief that a projector/display device can suffer from having too much light output, as some of you will know unwanted light from the unit itself, is very bad indeed, just as unwanted light from other sources can be disastrous.

Without ‘dark’ there can be no contrast and without contrast there can be no imagery.

So what is this ‘mystical’ second contributing factor you may just be wondering, well I will attempt to explain all in my next post below, a kind of part two (and maybe more)

james
 
James: Good post but I think your example needs better explanation. You can crush 30 ire down to black and have EXACTLY the same black level as on your other projector....After all you are just making sure that black detail(or actual light output ) starts to come out of the projector above 30ire.....so it's on off would actually be greater as it has the same black level but more light output. It would lack dark detail of course.

i presume what you mean is that the poor greyscale of your "30ire" unit means that you need to actually ramp up brightness to let the detail be seen....resulting in the raised black and the reduced contrast.

Of for the day now so look forward to your next post with some more of your excellent thoughts on how you've got on with stacks.

All the best,

Gordon
 
Great stuff James.

Obviously No2 will have the higher CR as the black level is smaller and therefore is divisible by a greater magnitude. It's simple maths - 10 ft lamberts divided by 0.001 FL will give 10,000:1 CR, but 20 divided by 0.01 FL only gives 2000:1. It doesn't take a great deal of reduction in black level to boost CR, but it does take a relatively larger amount of white to achieve the same increase.

Greater light output only has merit if the black level can be maintained or there is ambient light in the room (see my other post). For example, a digital on a screen that is too small will need a filter to dim the image down to cinema levels - relatively easy to do, or you can juggle the screen gain or size. Too much light can be a bad thing in that it can bring out artefacts from the source so setting the black level is imperative.

The second magical element wouldn't be ambient would it?

Gary.
 
Good point Gordon.

Having a good black level with no detail is not ideal by any means and although the on/off CR will be exceptional you'll be losing a lot of shadow detail and missing the action.

I've seen a good example of this and will try to find the image that shows this. I have to cook a BBQ now so won't be back for a while.. :)

Gary.
 
Before I reveal the magic word , I would just like to say this is not a fixed panel trashing crusade, but simply a few passages for those such as Gary who have expressed a real interest into the inner workings and functionality of this mystical beast known as the crt.

So giving the fixed panel a nod it deserves is its ability to track the ‘steps’ (bottom IRE not withstanding) of a grey scale pattern with a superior uniformity over ‘most’ of the 6500 range, butthis is where many patterns alone can be mis-leading.

In real life, and on Film, grey is not a stepped function but a continuous one, there are innumerable shades of grey which fall in between these discrete steps, and innumerable steps of both colour and hue.

A digital display whilst having many other strengths simply can’t track these constant minute changes in luminance and colour as faithfully and continuously as an analogue device.

So then, time to reveal the word at question… ‘Gamma’

So for those forward looking people who have already decided that the crt has had it’s day they may have forgotten ‘Gamma’ so despite of the fixed panels impressive resolution specs, huge advantages in convenience, setup and tremendous strides made in fill-factor colour saturation,black level and contrast ratio they still do not produce the look of film like a 8” or 9” crt.

And if film is the very thing which floats your boat this is the very thing which is perhaps most important of all, ah-ha so you want me to back this up I hear you ask to
scientifically quantify and explain my rationale….very well here goes.

First, consider how film reacts to exposure from light, its gamma curve(wish I could draw a pic at this point).
At the bottom of the curve is the “D-Min” area, where the films base level plus fogging which occurs when film is exposed to the most minute glimmer of light which prevents it registering any blacker blacks.
Next comes the “Toe” this is the area in which blacks, though still in a non linear fashion (yes, you got it, slightly crushed), start to be reproduced as shades of grey with visible shadow detail.

After this comes, as with many graphs the long straight line section in which changes of light are registered in a linear and continuous fashion as shades of grey (also hue and colour), before the film stocks gamma curve reaches its shoulder (which is analgous to the toe,but at the bright end of the spectrum) and the “D-Max” the maximum density (i.e, the brightest white) the film is capable of.

So although some videophiles may just think of greyscales as these discrete and logarithmic steps as produced by such test dvd’s such as Avia and video essentials from the darkest grey just above black the “D-Min” to the lightest grey just below peak white “D-Max”.

A projector must be capable of a continuous and infinite number of states of illumination and dynamic range, to use this old analogy, a fixed panel is is a light switch on/off whilst a crt is the very same thing but with the addition of a dimmer.
Part (3) to follow soon-ish

james
 
Yes quite correct Gordon,set thus it would indeed lose a little in the blacks, Henry for example sets his black level way lower than i do, i simply use a certain movie (or two) with a very dark scene and turn up the brightness until I can be sure I'm not missing information.

By the way the bottom 30 Ire quote was simply an example not a measured state, just for those who were wondering.

Now for my next technical miracle......cutting the lawn and washing the cars etc tis Sunday after all :D

more to follow;

james
 
Status
Not open for further replies.

The latest video from AVForums

Is 4K Blu-ray Worth It?
Subscribe to our YouTube channel
Back
Top Bottom