AVS Forum banner
Status
Not open for further replies.

A better mousetrap (colorimeter)

524 views 20 replies 12 participants last post by  Mark Hunter 
#1 ·
Kevin:


This sounds very cool. Do you have a projected date?


Chuck
 
#2 ·
Hi Chuck,


We expect to release this just before the Panamorph ships http://www.avsforum.com/ubb/biggrin.gif .


I've decided to take an ID Software approach on the delivery date this time (ID Software created Doom, Duke Nukem, Quake, etc.)..."The delivery date will be when everything is done".


I know that doesn't help your planning at all, and I apologize for such a snide answer, but there is a bit more involved than software development this time. There are some hardware modifications, vendor negotiations and business agreements to put in place in addition to the software.


The good news is that it's been pretty smooth sailing so far. We will be posting regular updates as this progresses (both here and on the brand-new Dilard.Com web site . The first update: early experiments have been very positive!


By the way, here is picture of what Kevin was describing above:

http://www.dilard.com/dilard/images/spyder.gif


[This message has been edited by milori (edited 04-30-2001).]
 
#3 ·
Very good news. Any chance of a Dilard-based auto black level, color temperature, gamma and shading wizard to go along with this colorimeter?


Will the colorimeter need to be pointed at the screen or into the projection lens? I am concerned about light reflecting back from the room onto the screen biasing the color temperature readings.
 
#4 ·
Quote:
Any chance of a Dilard-based auto black level, color temperature, gamma and shading wizard to go along with this colorimeter
That's exactly the plan! I don't think that we can do much with integrating the colorimeter into the shading wizard, but the others are all in the cross hairs!


I've done all of my experiments with the colorimeter pointed directly back at the projector.
 
#5 ·
light reflecting from the room is definitely a problem because

this device is much more sensitive. Over the weekend i replaced

my screen with black felt from the local cloth store, and the

numbers really came out nice.


Paint all the walls and ceilings flat black, and make sure there

are no windows in the room, and the calibrations come out amazing.


By the way, because of the way this device interfaces with the

computer, 9 of them could be used to do shading. Expensive

(relatively) but possible. 9 tripods or some mounting gizmo.

Yikes.


[This message has been edited by kevin gilmore (edited 04-30-2001).]
 
#6 ·
Speaking of mounting, there will be a tripod mount for the colorimeter as well:

http://www.dilard.com/dilard/images/probe_mounted.jpg


The three smaller holes secure the probe, and the larger central hole frames the sensor itself. Of course, the screw on the bottom is for the tripod mount.


[This message has been edited by milori (edited 04-30-2001).]
 
#8 ·
Here's a question...


you said to perform you calibrations in a completely light free environment(except the projector of course).


Wouldn't you want to calibrate the projector to the exact same environment you usually watch movies under?




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

Tryg Hoff
 
#9 ·
Quote:
Originally posted by milori:
Hi Chuck,


I've decided to take an ID Software approach on the delivery date this time (ID Software created Doom, Duke Nukem, Quake, etc.)..."The delivery date will be when everything is done".
Actually 3DRealms created Duke Nukem. Hey, I work in the industry, gotta give credit where credit is due. http://www.avsforum.com/ubb/smile.gif


Say, anyone know if the new colorimiters will let me exactly match paint colors? I spent quite a while this weekend trying to find the exact matching swatch. I'd pay mid-3-digits just to not do that again. And if does projector calibration too, so much the better. http://www.avsforum.com/ubb/wink.gif

 
#12 ·
Why not focus our Dila systems to a "close" screen and drape the

whole enclosure in black velvet for the calibration. I don't know

if we could focus down to 4' or not, but if so, a simple black velvet

booth could make a simple cheap black studio. This might be too close

for the colorimeter though.


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

JVC G1000 DILA

Stewart Greyhawk

Sony: SAT-B3, DTC-100,STR-DA777ES Receiver

Toshiba: SD-3109 DVD

Proscan 27" TV

B&W: DM601, CC6, DS6B, 800ASW speakers
 
#13 ·
Chris, I did a big job for Maxmeyer America, now owned by PPG for

matching automotive paint colors. I know a great deal on this subject.

This probe is probably not suitable for that task. At a minimum special

lighting requirements are necessary for reflective finishes. Devices

do exist for matching paint for print applications, as well as for

paint chip matching. The best in the field are X-rite and Gretag

Macbeth. I kind of like the latter unit. The price of the Macbeth

unit is around $3000 i think.


Mark Hoy, you have a good idea, for the g1000 series this

technique is recommended in fact. The probe works up to

about 230 footlamberts. My g15 generates about 35 footlamberts

on a 100 inch diagonal 4:3 screen.


[This message has been edited by kevin gilmore (edited 04-30-2001).]


[This message has been edited by kevin gilmore (edited 04-30-2001).]
 
#14 ·
OK, it seems apparent that many of you guys are gamers.. Cool..


Check out my Game Server's Web page, and if you play Counter-Strike, Tribes or Unreal Tournament stop by my servers and play sometime!!

www.CainsLair.com


My Tribes server is the busiest in the world, my UT server is a top 5 server, and the CS server is new, but only holds 22 players so I can't compete with the 32 player servers that run Dust 24/7 (not that I'd want to!) but it stays very busy also..


Oh and BTW regarding the topic of this thread!!


WOW!! Awesome, amazing fantastic!!


-- Cain

 
#15 ·
Kevin,


Nice stuff.


As usual, your detail is a plethora of info that exercises the mind and tantalizes the senses.


You have the ability of putting simplicity into complicated tech talk that I, and I'm sure others, really appreciate.


Correct me if I'm wrong, but I read into your post as a lead in for the upcoming wizards, and how it might be possible to use an available tool, either through renting or purchasing, to calibrate our DILAs color scales in our own homes. http://www.avsforum.com/ubb/eek.gif


If so, can I be the first to reserve the rental unit. http://www.avsforum.com/ubb/smile.gif


Chris
 
#16 ·
Mark,


The project sounds very cool.


I would like to point out one thing. Your product would work extremely well for 1-chip DLPs. Since they have more uniform color and more even grey scale tracking. You could use solid colored fields for testing and get accurate results for the whole screen all at once.


An automated test program could run through just a few colors and then make gamma and color corrections as necessary.


With a D-ila your color corrections will have to be done before and after several iterations of shading adjustments. I am sure with less than perfect users doing the shading work.


This isn't a knock on the d-ila by any means simply another road for a similar product you could create for the 1-chip or 3-chip DLP's (with some knowledge of their service menus of coarse)


-Mr. Wigggles


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

The Mothership is now boarding.
 
#17 ·
Sweet Kevin and MArk

This Dila get's better and better but my wife keeps asking me when will it end.

Good thing I got that hidden Stash... you know the one, us marry guys have.


Hey Cain thanks Im a Tribes player.

God help us when we run out of tweaks then what will we do....

Buy a new type of projector and start the cycle again?


Hugo
 
#18 ·
Mark,


I love this section of your webpage:


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


Dilard D-ILA Automation, Modification

and Calibration Software

Version 2.0 (CD)


This CD contains:

Dilard Software

Dilard DILA Projector Automation Capability (*.exe)

Projector Detector Wizard (*.exe)

Image Geometery Wizard (*.exe)

Pixel Alignment Wizard (*.exe)

Tracking / Phase Wizard (*.exe)

Black Level Wizard Wizard (*.exe)

Sample Script Macros (*.dls)

Sample Mods (*.dlm)

Two Automation Files, One for Pronto Remotes (*.dld)

Sample Pronto File for projector control through HTPC (*.ccf)

Users Manual (*.pdf)

Help File (*.chm)

License for Use on a Single Projector (*.lic)

License Agreement (*.txt)
Purchased individually, these features would cost over $1200.00.


Our Price Still $ 350.00


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


If we call in the next 15 minutes will we get some Ginsu knives as well?


-Mr. Wigggles




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

The Mothership is now boarding.
 
#19 ·
A new colorimeter


Times change. 10 years ago a computer with the power of a single pentium III/1000 would have been a 20 processor (or more) sequent with a price exceeding $1.5M. Today you can get this for <$1200.


10 years ago a spectrometer (colorimeter) was large and heavy, used

diffraction gratings or prism's and photomultiplier tubes and cost

tens of thousands of $. A brand new perkin-elmer instrument these

days is built the same as it was 10 and probably 20 years ago, and

starts at $80k. Not really practical for home calibration.

(unless you also own a fork-lift)


Over the past 5 years a number of hand held instruments have appeared

to do the spectrometer function. They fall into two different categorys.


The first category is instruments like the photoresearch pr640, the

mcmahan lightspex, pritchard pr-100, minolta cs200a and others.

The first two use multipoint silicon detectors with resolution ranging

from 64 to 128 points. The last two use photomultipliers and have

resolution from 256 to 1024 points. All of these devices share a

common characteristic, their price. The cheapest of the bunch is

$4500. A little much for people to spend for home use.


The second category consists of devices like the minolta mc7 and ca-100, the optimate I and II, the xrite-dtp92, the macbeth spectrolino, and the sequel chroma4. All are tri-stimulus devices which means

that they have one sensor for red, one for green and one for blue.

Some of these sensors have integrated optical filters, some use plastic glue on's. These devices are great for their real intended use,calibration of direct view crt's. As such their filters try to

match the phosphor characteristics. Great for crt's, lousy for

lcd monitors, and even worse for front (or rear) projection lcd

(including dila) and worst yet single chip dlp projectors because

the spinning wheel drives the detector nuts unless it is properly

sync'd. These devices have market pricing ranging from $249 to

$1295 (not usually including software). Attempts to make these

devices do what they were not designed to do has resulted in numerous

problems.


Enter a new class of colorimeters. The device is priced for home

use. It has 7 color sensors instead of 3. (not counting the additional

sensor used to determine refresh rate if any). It contains sharp

optical filters that lose less than 1db in the passband as compared

to 4 to 6 db for the tri-stimulus sensors. (precision diffraction grating) Since there are more sensors across the 400nm to 700nm optical band the necessity to adjust the readings after they are taken from the sensor for different display devices disappears.


Sensitivity and color accuracy improve. A factor of 5 or so over

older tri-stimulus designs. Color temperature accuracy is better

than 100 degrees at .1 footlamberts. Price of the detector with

some really excellent software is <=???.

(note, 3 digits, middle 3 digits)

(mark asked me to delete the price)


This is the sensor that dilard will be recommending for home use.

Although we will also support the sequel chroma 4 and the minolta

units. We will also be supporting a 500 point unit that i am

building personally. It has gone up in price a bit because

the hamamatsu r636 GaAs detector i am using is more than i wanted

to spend and hamamatsu seems unwilling to rebuild this device

into the smaller package with one or 2 less dynodes. But it is

absolutely amazingly flat across the entire visible spectrum and

beyond in both directions...
 
#20 ·
One person asked how big reference spectrometers are.

Here is my answer

http://gilmore.chem.nwu.edu/ja1.jpg

(remember i did say forklift)

http://gilmore.chem.nwu.edu/ja2.jpg

(just a little hernia generator)


but wait, there is more, it slices it dices

(the competition that is)


Yes this device works correctly with spinning wheel things,

but that is in the future.


Hugo:

God help us when we run out of tweaks then what will we do....

Buy a new type of projector and start the cycle again?


ABSOLUTELY!!!!!!!!!!


I want a 2048 x 1536 dila projector. NOW. I don't care how

much it costs, all it has to do is fit in the room.


 
Status
Not open for further replies.
You have insufficient privileges to reply here.
Top