Hi all and happy new year.

I was just wondering if there would be a way to have 48 hz on each PJ stacked but with one delayed in time in order to have 96 hz on the screen ?
This is not very clear but I am sure you get what I mean... synch two low refresh rates to obtain a higher one.
Do you think this is possible?
I have asked on the cinetson forum and someone answered that I should get interlaced artifacts... and I don't see how two progressive would get interlaced.

What do you think? I can't get this out of my head.
Thx
Rom

Interesting idea. I wonder if it's as simple as setting one projetor to positive Vsync and the other to negative.

I doubt that even the best set-up stack could be aligned as well as what you are describing would require - even Ken Whitcomb would have trouble ;)

Thx Jerry and Oliverg.

I don't see why the alignment would more a problem than in a classic stack, can you explain please Oliverg?

Thx Jerry and Oliverg.

I don't see why the alignment would more a problem than in a classic stack, can you explain please Oliverg?

If I'm reading you right, you're talking about aligning up a whole bunch of very very tiny long alternating lines - how many zones do your PJs have?

Overlaying (stacking) of identical images is hard enough - IMHO - doing what you're describing would be painstaking - you're trying to manually do something that would be much easier by just "plain old stacking"

Also - motion artefacts would be introduced

A 'normal' stack requires a bit of work to maintain as it is. The setup you are proposing would probably be a lot more prone to drift.

But... having said all that - I'd love to try it out. If you do, make sure you let us all know how it went :)

Kind regards

Hello

48hz is 48hz regardless how many tubes are shooting at the screen, and a delay between two projectors would blur any motion.

I built some hardware that does something similar about a year and a half ago. My idea was to use two 1292 projectors to produce a 1080p 60 image. Projector 1 is fed only the odd lines of the frame, and projector 2 the even lines of the frame. This means each projector is running at half the horizontal frequency. The main benefit is of course, that the scan velocity is only half of what it would be if each projector were producing all lines in the frame. All else being equal, lower scan velocity = higher horizontal resolution. Yes, there were issues with projector alignment and raster geometry, but nothing that could not be overcome with careful alignment and color calibration. My setup consists of a vertical stack in rear-projection configuration (a must for 1292s).

While I was able to prove the concept, I soon realized I was working on the wrong problem. The source I want to be able to display really well is blu-ray disc. Judder is the enemy here. The current project is a frame doubler to produce 1920x1080p48 from 1920x1080p24. If the flicker isn’t too bad at 48Hz, the picture should be incredible. There may be scalers available that can double / triple 1080p framerates, but that would take all the fun out of it:)

48hz flicker is too distracting. I've tried it a couple times and still have my stack setup to use that frequency.

Cliff

48hz flicker is too distracting. I've tried it a couple times and still have my stack setup to use that frequency.

Cliff

I don't mind 48hz flicker, but I run 72 to appease people who come to watch movies. I kind of like 48hz; on my 808s it looks almost exactly like cinema flicker (only noticeable with lotsa-white screens) but there's no sharpness reason to run it as the bandwidth on the graphics s is so high anyway.

One of my friends was complaining about the 48hz flicker, and I said, 'Why, it's the same as in the theater!' and he replied, 'Exactly... can't stand it there either!'

No real response to that one. :)

One thing I wonder is if lower refreshes marginally increase brightness output; if I understand correctly the lowered brightness with brighter images is due to HV falling down. Presumably lower refreshes lower the average load, and thus increase perceived light output as your eyes are making up for the dark bits with persistence-of-vision?

This would suggest that running just at the threshold of visible flicker would result in the best perceived brightness. But the real effect could be very small... I might try some long exposures to see if my camera can tell the difference, though!

As far as the original question, if you stacked two PJs and ran each at 48hz, you probably would perceive less flicker. But it wouldn't HAVE to be aligned precisely time-wise, even if that was possible. The only problem would come if you had very long beat frequencies between the two PJs' refreshes, and then you'd end up with the image looking like 96hz and then 'fading' to looking like 48hz and back at whatever the beat frequency is. Now that I think about it, this is probably what you'd get - just like a faster version of watching two turn signals at once.

To the guys saying you'd get blur or motion artifacts: This is incorrect; even if you had perfect timing and 'aligned' the frames time-wise to go 1 2 1 2 1 2 the delay is so small it's a non-issue. And because the original framerate is 24fps, it'd be impossible to see. To wit:


Projectors are 1 and 2

Frames:

PJ1 at 48hz:

1 1 1 1 1 1 1

PJ2 at 48hz:

12 2 2 2 2 2

Combined:

1212121212

Movie frame number, along with projected frames:

1---.2---3---.4...
1212121212121...

No blur or artifacts; neither projector affects the other and each real frame of data is displayed twice by each PJ so this isn't possible. However, as I said before, this alignment is likely not possible, meaning you'd end up with a beat frequency like so (now the Arial is helpful!)

Frames:

PJ1:----------------
PJ2:~~~~~~~~~~

There, PJ1 has displayed 16 frames; PJ2 has only displayed 10. So at the beginning, both are firing a frame at exactly the same time; toward the end they're almost alternating. At a low enough frequency this would start to be visible. At a high enough frequency it would start to be pointless. :)

For what it's worth, if you assume a beat frequency you would, in one out of four frames during the peak of the completely stacked frames in the beat frequency, get a 50/50 blur between two different frames. This effect, however, is so ridiculously small that only the terminally pedantic would complain, particularly given that the beat frequency itself is a bigger issue. In fact, I'm somewhat ashamed for even thinking of it.