[nigel_ht]

I was at the Davis factory in Drammen, Norway in April and measured the projectors at that time. The lamps drop a significant amount of the Red output in a short time frame at the beginning of their life cycle. As a result a gamma look-up table properly designed for use at 0 hours will be radically different from one accurately calibrated at 100 hours.


The lamp stabilizes in its output curve at this point and does not change substantially to the end of its life cycle. The light output will decline gradually but will not change the color balance (spectral output)significantly. It is truly unfortunate that most LCD/DLP projector manufacturers do not design the gamma look-up tables for a lamp at the point in its life cycle that it has stabilized at which represents the majority of its useful life.


I was under the impression that different bulbs (from different production runs) exhibited different color characteristics. If this is true then I would expect most gamma tables to be at least somewhat wrong if not adjusted for a particular bulb.


In any case you've indicated that Thumperized units are now non-linear performing projectors. I didn't notice any (obvious) non-linearity on grayscale when running the calibrations with Avia (by eye) but I wasn't looking for any either. Have you measured any units to determine the correct gamma for a Thumperized unit (say one of the more common UP-1100T) and would this new gamma table hold true for any future replacement bulbs?


Nigel


PS Non-linear or not...my unit performs a lot better than before. The price of admission was well worth the cost and Thumper went out of his way to make sure I was a satisfied customer.

 

[nigel_ht]

Hit the wrong button. The original thread is here:
http://www.avsforum.com/ubb/Forum10/HTML/007668.html.

 

[ghibliss]

I have made modified look-up tables for the UP1100 projector. I have also seen a Thumperized UP1100 which requires an entirely different Gamma look-up table due to the revised contrast and brightness settings. Your Davis projector does not have the benefit of a flashable eeprom for the new look-up table required so further calibration of your projector is basically not possible at this time. If you place a Gray scale pattern from Avia on the screen you should see that some of the bars are Gray and others are somewhat Red or Blue. This is what I referred to as a non-linear Gray scale and is controlled by the gamma look-up table.


Thumper does very nice work and I am not trying to discredit any of his work here. I am simply making a point as to what other elements of the picture are being effected from the changes made to your projector. Everyone has different criterior for assessing picture quality and the one that matters most is yours. If you are happy with the results of the modification than thats what counts.


My personal taste for picture quality has always been for accurate color reproduction first and foremost. If you are interested in the software from Davis that will allow you to modify your look-up table for a more accurate Gray scale than please feel free to send me an email and I will be happy to help you.


Regarding lamps from different production runs: yes their are slight variations from one run to the next from the same manufacturer on a given lamp model. These variations are not great when comparisons are made between them and are typically relevant to the amount of Red content that the lamp starts with. What is more important is how the lamp performs after it has stabilized in its maturity cycle which I have found to take place within the first 50-100 hours depending on the lamp.


I have also found that when developing a new Gamma look-up table for a projector with a lamp optimized for a specific number of hours (100 hours)that when replacing it with another lamp with similar time it will look virtually the same. As stated in a previous post once the lamp has stabilized, its output has only a very gradual decline in output (intensity) spectrally however the output is pretty much unchanged.


What I noticed in Norway while at Davis when viewing projectors in the lab was that only one of the available gamma curves in the projector was good for video (D65 with a new lamp). The other gamma selections produced a somewhat brighter image but at the expense of color accuracy (Reds were way off and appeared to look very muddy/Brown). After a very short time the D65 color temp with an aged lamp of 100 plus hours yielded a color shift of almost exactly 2,000K making the display read close to 8500K. This is clearly to Blue and is a result of the lamp losing its Red content from the gas burning off in the lamp itself.


[This message has been edited by ghibliss (edited 07-31-2001).]


[This message has been edited by ghibliss (edited 07-31-2001).]

 

[mflaster]

So is your concern only about color accuracy? I got the impression that the dark end of the gray scale would be compressed. In other words, if you had a dark movie, you could turn up the brightness, and the picture would look terrific. But then a bright scene would be washed out, too bright. If you adjust brightness/contrast for a brighter scene, you lose your shadow details.


It seems that this could be fixed by boosting the gamma curve so that its slope is > 1 initially in the dark sections, and then have the slope < 1 as you get more to the midrange. This would hopefully generate linear grayscale output.


Is that what you were getting at?


Mike




------------------

Join the NYC-area HT club! Send me e-mail!