That is they turn on and off real quick. Light up a pixel and within a few milli seconds it has faded to black. An advantage when things are moving across the screen. Alot of digitals will leave the pixel on for a full frame (16 milli seconds) or more. That leaves a retention in a person's sight that can cause a blurred image to be seen. The following pics show the responce of Sony CRTs in an Ampro 4200. Green Panasonic CRTs generally have shorter persistance.

Scott

More pics....

Red, green, and blue responses. These pics are at 2 milli seconds per horizontal division. The vertical scale is the same for each. Notice the red tapers to zero after about two divisions or 4 mS. The green about four divisions or 8mS. The blue is very fast and decays in less than a mS. It's amplitude is much higher than the other colors so as to produce enough light in it's short time.

Scott

Nice post,

Just a warning... if a bunch of Digital owners come into this thread to try and derail it... they can expect an immediate suspension.

Carry on,

Kyser

Nice post,

Just a warning... if a bunch of Digital owners come into this thread to try and derail it... they can expect an immediate suspension.

Carry on,

Kyser

I have absolutely no dog in this fight at all, but doesn't the very nature of this thread call for some kind of debate?

I have absolutely no dog in this fight at all, but doesn't the very nature of this thread call for some kind of debate?

He didn't say you couldn't post data about the response time of panels, or o-scopes of digitals and/or discuss the various impacts on image quality of this, etc, etc. He just said that you could not "derail" it.

Dave

He didn't say you couldn't post data about the response time of panels, or o-scopes of digitals and/or discuss the various impacts on image quality of this, etc, etc. He just said that you could not "derail" it.

Dave

Fair enough. Sorry to have done so.

This remids me of a humorous story.
Henry Kloss had his own factory to build the CRT tubes that were part of the sets he manufacturered. In some of his early sets he choose the phospher based on color and not persistance of the phospher, and consequently they were very slow decay.
Well need less to say, when there was action moving across the screen there were always these trails that followed.
Guess it was just a sign of the times.....lol

Terry

He didn't say you couldn't post data about the response time of panels, or o-scopes of digitals and/or discuss the various impacts on image quality of this, etc, etc. He just said that you could not "derail" it.

Dave

Bingo!

Scott, it would be interesting to track the response time of different technologies (and implementations of those technologies).

I would think the DLP is even faster than CRT since those little mirrors a flickering an insane number of times during even 1 frame (16 ms). So, they have to have the fastest response time since they articulate in less than 1 ms.

LCoS and LCD are the slower ones. But, the smaller the pixel the better and the panels for PJs have been getting smaller. I've always like the look of the JVC LCoS better than the Sony LCoS, but it seems they their response times are slower. Not sure about the new crop though.

Dave

This remids me of a humorous story.

Well need less to say, when there was action moving across the screen there were always these trails that followed.
Guess it was just a sign of the times.....lol

Terry

You mean I was paying good money to see something that was already there? :eek:

Guys!

Standard green is laggy if used in fast refresh stereoscopic apps hence the use of faster P43 green for stereo. 104 frames a second is not unusual in that.

Nice post,

Just a warning... if a bunch of Digital owners come into this thread to try and derail it... they can expect an immediate suspension.

Carry on,

Kyser

As a CRT owner, I have no problem with others adding their commentary if it's constructive or even humorous. I am not satisfied with your attitude, I mean these threads are part of the foundation of the forums.

I am just registering my unhappiness with regards to the threat to close this thread which is, IMO, an unecessarily pre-emption. Relax.

Anyway, there was a recent thread here about this in the digital forum which many may find interesting:

http://www.avsforum.com/avs-vb/showthread.php?t=802850&highlight=hold

See also: http://www.microsoft.com/whdc/archive/TempRate.mspx

As a CRT owner, I have no problem with others adding their commentary if it's constructive or even humorous. I am not satisfied with your attitude, I mean these threads are part of the foundation of the forums.

I am just registering my unhappiness with regards to the threat to close this thread which is, IMO, an unecessarily pre-emption. Relax.

Anyway, there was a recent thread here about this in the digital forum which many may find interesting:

http://www.avsforum.com/avs-vb/showthread.php?t=802850&highlight=hold

See also: http://www.microsoft.com/whdc/archive/TempRate.mspx
Sorry, the last thread that I had to close was derailed and I wanted to put out a message so that it wouldn't happen again.

Disagree all you want, just PM me and keep it out of this thread.

Cheers,

Kyser

Kyser... wouldn't it be more constructive to just delete the offending posts rather than to close the thread entirely?

.... Alot of digitals will leave the pixel on for a full frame (16 milli seconds) or more. ...
DLPs have a response time of < 1 ms.
SXRDs and the new DILAs are ~ 4 ms
LCDs are ~ 8 to 12 ms

Kyser... wouldn't it be more constructive to just delete the offending posts rather than to close the thread entirely?

If I have the time... yes. When I come across a thread filled with them... I'll close it if I don't have the time to clean house. If it has gone too far off course, I won't bother at all.

Let's get back on topic. I can be reached via PM for any further questions.
Anything OT from here on in will be deleted.

Kyser

DLPs have a response time of < 1 ms.
SXRDs and the new DILAs are ~ 4 ms
LCDs are ~ 8 to 12 ms

DLP does seem to be unique. AFAIK LCD and LCOS have response time like you say but I believe they are stating the turn on plus turn off time. Once a pixel is turned on it stays in that state for at least the full frame and only changes when a different level comes along. So it is on a minimum of 16mS. The CRT turns on and is off long before the next time that it is scanned. The CRT picture would look like a horizontal bar moving down the screen if our eye/brain didn't have it's image retention.

Scott

this is the main reason i HAD to stay with crt ,i tried a benq 8720 that was better than crt at this but the rainbow effect killed me . I'm a gamer and i like fps's and to turn real fast is a must and it must stay sharp. i also tried the canon which used the old jvc's panels ,and blur happen on games and tv.So i'm hoping the new jvc's panels will solve this or i'll have to get a 3 chip dlp in a couple of years when my g70 needs new tubes.

Scott,

Do you know anything about what film does here? I know they flash frames, but I don't know what their duty cycle between showing a frame and showing nothing is. As mentioned elsewhere, a digital could put up a blank image between frames, but that would decrease the light output (and might hurt the CR, depending on how it was done). If real film has the screen blank 10% of the time (just an example) with 2 to 3 cycles per frame, then it seems like a mode could be added for digitals (at least for DLPs) to do the same thing (and not do it for cases where people want more light output instead).

--Darin

I don't want to get into a fistfight or anything, but the fast rise-fall time of CRTs obviously has its disadvantages as well. I personally prefer the sample-hold look of LCD/LCOS in terms of creating a solid image. The flicker and rainbow issues that occur because of phosphor decay are, in my view, negatives.

Film varies depending on the projector, there are two blade, 3 blade five blade shutters etc. You can look at the shutter itself (a big rotating disc with pieces cut out of it) and work out the duty cycle. It is typically 2 blade or 3 blade for 35mm projection (or was back when I used to do it).
You could hook up a stepper motor and a shutter in front of a digital and try it pretty easily, you could use the sync pulse to trigger the stepper.

How many of you see rainbows in CRTs? I do... to a lesser (and tolerable) extent compared to DLPs (I have not seen any 1-chip DLP that I find tolerable).

I see rainbows on CRTs, but I am very susceptible to rainbow-like artifacts. I am sure that many people do not. Visibility depends on many things including refresh rate and viewing environment. I use 72hz to minimize rainbows as much as possible and also for no judder.

Isn't rainbow like artifact from CRT caused by out of phase tube's cut-off?
If so in some CRT's there are parameters that can tweak that. I have never seen RBE with my XG and I can readily see DLP's RBE.

Well this all brings up an interesting observation.
Just how fast can the decay of P43 phosphor be that it has been long held that the P43's couldn't be used in HT applications.......else you would see flicker.
I have personally seen a few higher end 9" CRT's w/ P43 phosphor and have never seen flicker. These were set up at 60 or 72hz refresh.

Terry

I see rainbows on DLP PJs, but I've never seen them on a CRT PJ. I guess I'm lucky.

Well this all brings up an interesting observation.
Just how fast can the decay of P43 phosphor be that it has been long held that the P43's couldn't be used in HT applications.......else you would see flicker.
I have personally seen a few higher end 9" CRT's w/ P43 phosphor and have never seen flicker. These were set up at 60 or 72hz refresh.

Terry


I have attached a PDF called "Measurement of Phosphor Persistence".
It deals with P20 vs. P43 phosphors in X-ray image intensifiers.

It is taken from the SPring-8 Annual Report 1998 http://www.spring8.or.jp/en/support/download/publication/annual_report/html/AR98.html

Maybe the plots in there can be of help

Well this all brings up an interesting observation.
Just how fast can the decay of P43 phosphor be that it has been long held that the P43's couldn't be used in HT applications.......else you would see flicker.
I have personally seen a few higher end 9" CRT's w/ P43 phosphor and have never seen flicker. These were set up at 60 or 72hz refresh.

Terry

I agreed with Terry . I would think the fast decay using P43 phosphor would better match that of Red and Blue or am I missing something ?

I agreed with Terry . I would think the fast decay using P43 phosphor would better match that of Red and Blue or am I missing something ?

Could it be that when the persistenece of all tubes are closer together you notice a flicker however when one is further out of phase(green) then it is not noticable.

Makes me wonder how much of the streaking many see on CRT projectors can be attributed to the eightfold decay rate of the Green CRT as compared to Blue ?

But streaking is a spatial problem. that is to say it's an electronic thing, decay doesn't affect neighboring phosphor or anything like that, it's completely different from streaking.

Could it be that when the persistenece of all tubes are closer together you notice a flicker however when one is further out of phase(green) then it is not noticable.

But here is the thing, I have never seen flicker when isung P43 phosphers at 60 hz and above. And I would think that the highr the refresh rate, the more desidable the P43 would be.
I also think that P43 produces a smaller dot when compared to standard G phospher. My sample isn't large enough to say that is absolutely the case but from the few I have seen it was small.
Therefore I have concluded that for refresh rates above 60 hz in Home Theater applications, there is nothing lost by using P43 and maybe something gained.
"P43 phospher no good for HT" ****BUSTED*****

Terry

I saw rainbows on DLP's to the point where I got a headache from eye strain when looking at bright crisp images. This was with a few large rear projection sets. I've never seen rainbows on my CRT which is an XG110LC. And frankly, I like the softer film-like image I see.

Craigo

One possible disadvantage to P43 MIGHT be less light output because of the shorter duration. If it has higher efficiency than standard green then it could have similar output.

Information on the different phosphors seems to be somewhat elusive. I can find specs like MS for medium short but what exactly does that mean?

Scott

I have tried P43 tubes in a Marquee 8500 machine and noticed no flicker at 60Hz.
I did notice the colors were off quite a bit and lack of screen punch, I did not spend too much time adjusting colors or grayscale and just figured the color problem was due to P43 tubes. Does anybody know if P43 can be color corrected to work on HT application? I always thought it was the color and not flicker that prevented us from using them in HT use. Knowing that a P43 could be used in HT would be excellent really excellent for me.

Thanks any info appreciated.

I have tried P43 tubes in a Marquee 8500 machine and noticed no flicker at 60Hz.
I did notice the colors were off quite a bit and lack of screen punch, I did not spend too much time adjusting colors or grayscale and just figured the color problem was due to P43 tubes. Does anybody know if P43 can be color corrected to work on HT application? I always thought it was the color and not flicker that prevented us from using them in HT use. Knowing that a P43 could be used in HT would be excellent really excellent for me.

Thanks any info appreciated.

Hello

P43 might be tolerable if filtered because it is somewhat pastel (read reddish) and darkens if overdriven so smaller screens would be the preferred choice, siz foot width or less.

Yes I would call the green P43 color almost pastel, the green alone looked lighter in tint than a standard one. I thought about using green glycol as filtering but would desire a color meter to really try and give it a go. So what you are confirming to me is the P43 colors will be off unless you filter it in someway and that the adjustments on the machine G2, Bias, cutoff, ect will NOT get you to the correct color unless you do something more like filter?

I know you haven’t tried filtering a P43 but have you ever achieved good colors and grayscale by adjustment alone on a PJ with one installed.

Thanks again
Pat

Concerning refresh rates of digitals, DLP derives its grey scale by rapidly modulating the mirrors. It works via the same principle as a single bit DAC, sort of. The mirrors are modulated at MHz range speeds so the response time of DLP is theoretically capable of
supporting refresh rates into the thousands of frames per second. But I'm not sure they really have the bandwidth to do the job at that level.

What I don't know is, on a DLP (or any other digital, for that matter) is, what's the refresh sequence? Well, I sort of know what the refresh sequence is on single chip DLPs including
those with spiral color wheels, but what about 3-chip DLPs?

It would be kind of interesting to find out. Especially if a super-fast digital movie camera were used to visualize the refresh process in super slow motion.

CJ

I have a high speed camera at work that can do 2000 frames per second. Would that do any good? I'll look up the model.

Concerning refresh rates of digitals, DLP derives its grey scale by rapidly modulating the mirrors. It works via the same principle as a single bit DAC, sort of. The mirrors are modulated at MHz range speeds so the response time of DLP is theoretically capable of
supporting refresh rates into the thousands of frames per second. But I'm not sure they really have the bandwidth to do the job at that level.

The problem you run into though, is that the colors that are being produced depend on being integrated over a period of time since a DLP can only create 4 colors at any given moment in time for a single chip(depending on the colorwheel segments, could be more), or 8 colors for a 3-chip. Creating all the other colors come from combinations of these colors over a period of time which are perceived as colors between all these instant possibilities. As you shrink that time down far enough, you begin to decrease the effective bit-depth of the image as it struggles to keep up with creating a unique color but doesn't have long enough to do it.

This was one of the complaints about particularly early 1-chip DLPs that they would begin to posterize and essentially their bit-depth would begin to suffer with motion because it wasn't enough time to create subtle shades of color.

So at a certain point if you tried to push the framerate higher and higher, you begin to sort of run out of colors that you can really address over such a short period of time.

I see rainbows on CRTs, but I am very susceptible to rainbow-like artifacts.

I posted a thread here and the consensus was I was either crazy or I was driving the bandwidth too high. What I see are flashes of yellow every once in a while. I can see it on 60 hz material but more frequently on 24p material as 72 hz. Movies get darker more and you can see these flashes if only part of the screen is bright and the rest dark. I'm glad to see I'm not crazy that there's sentential color artifacts with CRTs and just not DLP.

One thing I've never been quite clear on is the way that DLP handles the mirror modulation on a per-pixel basis, dynamically. I'm sure it's not massively parallel, but I would be surprised if each mirror were serially addressed, as well. My GUESS is that it's modulated
in parallel, one row or column at a time, as it would represent a good tradeoff between
bandwidth and image quality.

If you had a true zero persistence camera system which is phenemonally fast, and were to use it to watch a single frame drawn with a DLP engine at a high enough rate to capture EVERY single flip of individual pixel mirrors, what would you see when you played
it back in slow motion?

That's what I don't know.

CJ

Seems it could work like CRT- one pixel after another. This is how I imagine the signal is sent.

One possible disadvantage to P43 MIGHT be less light output because of the shorter duration. If it has higher efficiency than standard green then it could have similar output.

Information on the different phosphors seems to be somewhat elusive. I can find specs like MS for medium short but what exactly does that mean?

Scott

I initially bought a 9517 LC ultra in 2005 with had a P43 green CRT.

light output was at least 30% lower than what I have now with the P19 LUG green.

resolution and color tracking also was not satisfatory.

I still have the P43 green tube if anyone is interested (1200hrs. on it)

Michael

I posted a thread here and the consensus was I was either crazy or I was driving the bandwidth too high. What I see are flashes of yellow every once in a while. I can see it on 60 hz material but more frequently on 24p material as 72 hz. Movies get darker more and you can see these flashes if only part of the screen is bright and the rest dark. I'm glad to see I'm not crazy that there's sentential color artifacts with CRTs and just not DLP.

No you're not crazy at all. I'm surprised you see it more frequenly at 72hz, it should be less at a faster refresh rate. It begins to bother me significantly at 60hz compared to 72 which is not bad unless I'm watching something like Sin City.

You can see a similar effect on plasmas as well, a reason I am not a big fan of plasmas in general.

I've had some longer threads in the past discussing this. It's not common on consumer CRTs that have slower phosphors, instead you usually get green trails/ghosting because the decay is so slow. But on CRT projectors which have faster phosphors designed for faster refresh rates and progressive scanning, you can see more the yellowish 'rainbows' that you're talking about.

72 which is not bad unless I'm watching something like Sin City.


I've been watching Dexter lately which is super dark, maybe more so than Sin City. It uses 24p source. I notice it a lot on that series which I output using 71.93 hz. I can't imagine watching that show on anything but a CRT. You'll miss half of the show otherwise.

Hilarious if you know Dexter:

http://www.youtube.com/watch?v=6hI8QHZZPBU

I've been watching Dexter lately which is super dark, maybe more so than Sin City. It uses 24p source. I notice it a lot on that series which I output using 71.93 hz. I can't imagine watching that show on anything but a CRT. You'll miss half of the show otherwise.

Hilarious if you know Dexter:

http://www.youtube.com/watch?v=6hI8QHZZPBU

If it is a TV series are you sure it was filmed at a film rate and not a video rate? It could be the conversion from a 30 frame video rate to the 24(72) film rate your watching it at.

Athanasios

If it is a TV series are you sure it was filmed at a film rate and not a video rate? It could be the conversion from a 30 frame video rate to the 24(72) film rate your watching it at.

Athanasios

Most brand new TV series done in HD that I know of are filmed at 24p. But anyway:

http://www.avsforum.com/avs-vb/showpost.php?p=15490210&postcount=1243

Title ...
Movie ....
Dexter: Season 1 Disc 1

VIDEO:

Codec Bitrate Description
----- ------- -----------
MPEG-4 AVC Video 22939 kbps 1080p / 23.976 fps / 16:9 / High Profile 4.1

If it is a TV series are you sure it was filmed at a film rate and not a video rate? It could be the conversion from a 30 frame video rate to the 24(72) film rate your watching it at.

Athanasios


Athan,

95% of all TV series in the States are filmed on film (24fps)

have a look at the end credits and it normally says something like

Panavision Cameras and Lenses.

there might be a shift to HD Cameras in the near future



Michael

And here I thought I was seeing things too. I see what are rainbow like flashes from time to time on my Zenith1200x, or should I describe them as huge convergence misalignments skews which last a split second. DLPs and rainbows = headaches within minutes.