[Heboil]

I have an Infocus IN76 projector (almost a year now). My bulb is dimming and I am curious when one needs a replacement. I use the Avia test disc to set brightness and contrast etc, and the brightness is set correctly still but things look darker overall. Should my brightness settings be set correctly to a test disc and you still get a dim (ish) looking picture?


I know people say replace when you want to, but I just looking for guidelines. With my Stewart screen with 1.25 gain and my pretty bright picture, I think I was projecting an image too high in fL according to the masses (almost TV brightness). As my bulb has dimmed, I am getting probably closer to theater brightness of an image. If it just takes getting used to, that is fine. However, if I am below theater-like fL I'd like to know.


Thanks.

 

[CT_Wiebe]

You need a video "Illuminance" meter. The least expensive one is the AEMC CA813 (around $160) - recommended for our use. More professional versions cost $900 and up (like the reference Minolta LS-100 - over $3000).


The "dimming" you are experiencing is normal and (as you pointed out) is much more obvious when you push the brightness to the max when the PJ is new. With your screen, you should have been using the "low lamp" mode to start with (that would have given you a very good picture (still way brighter than in a movie theater). Your IN76 is one of the brighter PJs to start with.

 

[Michael TLV]

Greetings


General rule of thumb is after the bulb reaches 60% of predicted life span.


If you have a bulb that is rated for 2000 hours ... anything past 1200 hours can start to get punishing to watch. Sure you can try to push it to 2000 ... but you will be moaning and complaining for 800 more hours of your viewing time watching some awful looking dim images.


The more the bulb dims ... the worse your contrast ratio gets. You might have started out at a rated 5000:1 ... but it could be 600:1 by 1200 hours ... and it gets even worse after than ... 500 ... 400 ... 300 ... 100 ...


higher end companies that focus on servicing the clients ... like Runco ... recommend a bulb be swapped out at half life. 1000 hrs on a 2000 hr bulb.


Regards

 

[Gary Lightfoot]



Quote:
Originally Posted by Michael TLV /forum/post/0


The more the bulb dims ... the worse your contrast ratio gets. You might have started out at a rated 5000:1 ... but it could be 600:1 by 1200 hours ... and it gets even worse after than ... 500 ... 400 ... 300 ... 100 ...

Why is that? Surely the black and white levels dim equally and the CR remains the same?


Gary

 

[Michael TLV]

Greetings


Guess the answer is that they don't. The drop off seems to be faster on the bright end and slower on the dark end.


Explains why the image looks flat ... no pop.


Of course we all know in real life that no properly calibrated projector does anything like 5000:1 ...


It's closer to 500-600:1 when properly set up. dropping from 600:1 down to 300 then 200 then 100 is a big deal. Images really start to look like those first generation Sony LCD RP sets ... with contrast in the 60-80:1 range.


Regards

 

[Gary Lightfoot]

I still don't understand how that can happen, since the lamp is just outputting less light, and how the pj produces the image has no effect on that - the lamp doesn't dim differently for a dark image compared to a bright image since the lamp output stays the same (ignoring lamp AI or DI that some manufacturers implement).


In effect what you're saying is that if you have a new lamp and go from bright mode to eco mode, you should lose on/off contrast, but you don't, so I don't see how having a lamp dim naturally will lose CR either.


A dimmer image can have less pop merely because it is dimmer and how the eye can perceive colour saturation etc at lower luminance levels. It's like looking at the same image/lumens on a high gain and low gain screen - the high gain will have more pop, but no more or less contrast. I've measured the on/off CR of my Optoma, and that has never reduced even though only 400+ hours have elapsed, but the lumens have measured lower.


Gary

 

[Michael TLV]

Greetings


Never been a fan of on off contrast ratios ... given the procedure seems to vary between manufacturers.


Ansi type is always better for closer to a real representation of actual material. I'll pop in my 1600 hour bulb back into the projector and tell you what things read ...


Regards

 

[Gary Lightfoot]

I use the same method most here seem to use - calibrated black and white levels (pref at D65), and the meter close enough to get a good reading at black without overloading the meter for white, and using 0 IRE and 100 IRE fields from a video test disk (you're then measuring the actual values for video black and video white). On/off will tell you the black level for a given white level, so is an important measure, especially if you want to know if the pj can produce the full contrast range of the source material.


ANSI is white and black squares and although it's good as a reference between different projectors ability to show mixed content, it still doesn't give a true representation of video material IMHO. I.e, a DLP with 500:1+ ANSI and 2700:1 on/off can still have over 1500:1 in a particular scene depending on the content - we rarely see an image of 100% white and 0% black in real images, most of the time it's ranges in between. There's currently a project on the forum developing test images to try and measure more accurately that type of content.


I'll be interested in the results of the old lamp.


Gary

 

[Michael TLV]

Greetings


Some numbers ... wonderful numbers.


Some background info first ... and bear in mind that YMMV as usual.


BenQ 8700+ dlp projector.


1620 hours on old bulb ...

135 hours on new bulb ...


Real time calibrated CR's are so much lower than those fake ones given by the specs.


The 8700+ had a CR rating of 2500:1 at the time. And it actually falls into my Calibrated CR rule of thumb ... that it is almost always about 6 to 15% of what the rated number is.


Measurements taken by Minolta CS200 color meter at 12 feet.


Ansi 16 box method.


Old Bulb ... 36:1 contrast ratio

New Bulb ... 53:1 contrast ratio


full on/off method (100% full field and 0% full field)


Old Bulb ... 226:1 CR

New Bulb ... 309:1 CR


At 6 ft.


Old ... 326:1 CR

New ... 616:1 CR


You are right that the blacks also go down in light output ... but not as much as the bright end. Reason for this ... (don't know right now)


309 versus 226 may not look like much to people numbers wise but visually ... it is significant enough to bring "life" back into an image sorely missing it.


As expected ... these are ISF calibrated readings ... as I am ISF ...


I'm sure if I put the meter closer to the screen ... the light output readings might go up ... but that won't change the result.


Regards

 

[Gary Lightfoot]

Those numbers look very low for both on/off and ANSI contrast readings. Is there any ambient light on in the room? Do you have the light meter on a tripod so that it doesn't move between readings?


I get much higher numbers and not dissimilar to the likes of Greg Rogers in the case of the Optoma H78/79. It's not unusual to get around 40% less than advertised numbers when calibrated. It would be interesting to see what you measure for an HD1


I assume you recalibrated both greyscale and B&W levels for each lamp, but I still don't understand why the contrast should change since it's just a light source that doesn't change - it's just 'on', and the imaging devices are the things that are changing - mirrors are either on or off in the case of DLP.


If what you say is true, then just by adding a brighter lamp should increase the on/off numbers, but that's not the case, since most high CR pjs tend to be dimmer, not brighter.


If you put the meter closer to the screen, you'll start getting errors with the black level readings as it approaches the meters limits of sensitivity (if taking on/off measurements that is).


Gary

 

[Michael TLV]

Greetings


You don't need to know why sometimes ... although it would be nice. Suffice it to say that this is a behavior that is expected as the bulb ages and at some point ... it really does get punishing to watch. One might count the pennies they are saving, but what's the point if the experience is ruined.


Kinda like eating all the food in the fridge eventhough it is now past expiry dates and bordering on rotting.

Mind you, if you don't want to believe that this type of behavior exists, that is okay too.


Regards

 

[Gary Lightfoot]

I like to know - I'm just curious.

I'm not disputing that image brightness reduction can reduce image quality, in which case a higher gain screen, smaller screen from the outset or a new lamp will increase the FL (and if used with an ND filter may help keep FL closer to cinema levels for longer), but the numbers you report and the idea that a lower wattage lamp has less CR than a higher wattage lamp is something I've seen or heard of before.


Were your on/off measures at 12ft or closer? Do you use a colorimiter or dedicated light meter for those readings? Just curious why your numbers are so much lower than what I've seen measured elsewhere (and personally). Did you recalibrate each lamp to ensure they were both at D65 and the white and black levels were set correctly? It can make a difference to the outcome.


Do you have any links or sources that state that lamp aging goes hand in hand with contrast reduction?


Gary

 

[Michael TLV]

Greetings


Every bit of information has to start somewhere and I haven't seen anyone else write anything about it. So this is the first.


Mind you, most of the calibrators I talk to just say ... Well Duh! ... when it comes to this point. Just obvious to many.


Of course everything is calibrated ... I wouldn't have it any other way.

Whether it is 6500 or 7500 or 9500 doesn't really matter though as long as it is the same for everything. Contrast ratios don't care about color temps.


I suggest you go do your own tests if you don't like the data. Or better yet ... pay someone to do it.


I've already said everything was done at 12 feet. I'm not going to go skew my results. I don't do stuff like that.


Regards

 

[garyfritz]



Quote:
Originally Posted by CT_Wiebe /forum/post/0


You need a video "Illuminance" meter.

Isn't the Y component of an xyY measurement (e.g. with Spyder2/HCFR) a measure of illuminance? If someone already has a Spyder or other sensor, isn't that a "good enough" measure of ftL or cd/m^2 for a casual amateur calibrator? Especially if you don't need a guaranteed-accurate-to-3-decimal-places measurement, but are looking at relative brightness?

 

[Michael TLV]

Greetings


But how much light output is enough? 10 ft-l ... 15 ft-l? more? less? And how far back is he taking those readings from?


The rule of thumb for CRT front projectors had been 12-15 ft-l from what I recall. It was supposed to be similar to the theater. But can't be sure how far back the meter is to take those readings and what is the test pattern that they are using?


In the TV world ... we have been taught to use a 100% ire windowbox as our guide to light output ... but different test discs have different sizes of window boxes so who is correct?


It's hard to go by specific light output numbers if you find the image to be too dim for your liking or too flat as a result.


The little test I did comparing the old and the new bulbs gave me light output readings of 6 ft-L versus 2 ft-L ... 135 hours versus 1650 hours. Only 1/3 as bright by 1650 hours.


Regards

 

[Gary Lightfoot]



Quote:
Originally Posted by Michael TLV /forum/post/0


Greetings


Every bit of information has to start somewhere and I haven't seen anyone else write anything about it. So this is the first.


Mind you, most of the calibrators I talk to just say ... Well Duh! ... when it comes to this point. Just obvious to many.

Obvious that contrast reduces as the lamp dims with age you mean?

Quote:
Originally Posted by Michael TLV /forum/post/0



Of course everything is calibrated ... I wouldn't have it any other way.

Whether it is 6500 or 7500 or 9500 doesn't really matter though as long as it is the same for everything. Contrast ratios don't care about color temps.

Indeed, if both lamps are calibrated to the same CT then the measurements will be comparable, other wise it will make a difference - if you have a higher colour temp you will have more green/blue than red for D65, so you have more lumens for white, and hence a higher on/off contrast ratio. You can test this easily and I've done this myself - at D65 my Optoma gives around 2250:1 on/off contrast. With a dE of around 9 at 100 IRE (all lower IREs less than 3dE) I can get around 2650:1 on/off, so colour temp does indeed make a diffeence to contrast figures, and that's why there is a difference between manufacturers figures and calibrated figures for both lumens and contrast. If you make sure that both lamps have been calibrated to the same colour temp, then the readings should be comparable.


I think the lack of pop you mention is more down to lower foot lamberts rather than contrast reduction. At lower levels of luminance colour saturation appears reduced for example.

Quote:
Originally Posted by Michael TLV /forum/post/0


I suggest you go do your own tests if you don't like the data. Or better yet ... pay someone to do it.

I have done, and that's why I questioned why there is a drop off with contrast when a lamp dims since I haven't measured this myself - I've always been able to get pretty much identical results when measuring the same pj when new and at around 400 hours.. I don't need to pay for this.

Quote:
Originally Posted by Michael TLV /forum/post/0


I've already said everything was done at 12 feet. I'm not going to go skew my results. I don't do stuff like that.


Regards

Taking on/off readings closer to the pj gives you more accurate readings for black (for the reason I gave earlier) and thus more accurate on/off contrast readings. Taking them further away can give inaccurate black level readings and skew the results - I discovered that when taking my very first contrast readings when I measured an NEC HT1000 and got 3100:1 (out of the box settings) when measuring at the screen (around 11 feet from the pj). Taking the readings from around 2ft gave much lower numbers that were more consistent with other results I'd seen. After I'd acquired Colorfacts I was able to get 2007:[email protected] using an FL-D filter.


Measuring CR in that fashion (close to the pj) allows me to get similar results to those obtained by the likes of Greg Rogers of Widescreen Review, so I'm happy that my method is consistent and accurate. I can also get accurate ANSI results for both pj and room using other tried and tested methods so I know how much impact my room is having on the simultaneous contrast of the image.


If a dimming lamp does have a big effect on contrast reduction like you say, I'd like to know why. It doesn't appear logical for the reasons I mentioned in earlier posts, but if it is happening, I'm curious as to why. Surely some of those who say 'duh' are equally curious?


Could it be perhaps that when a lamp gets so dim, the black output limitation of the pj optics/light path will still allow a similar amount of scattered light out, but the white output exceeds the scatter limitation so is more measurable as it drops? I wonder if a graph for this has ever been plotted.


Gary