Hi cowanrg,
I do actually know about the 23.976 and 29.97 fps stuff. I didn't want to confuse anybody reading the posts with extra details.
The thing about the interlacing stuff goes like this:
When a film source is converted to DVD, the studios generally take the 24fps film and scan it into progressive frames at the raw 24fps of the film itself. This is encoded directly into a 24fps MPEG2 video stream, with the addition of special "3:2 pulldown" flags and "SMTPE Drop" flags.
These pull-down flags tell the DVD player (ALL players) to do what they call 3:2 pulldown. This involves interlacing the progressive images and literally inserting additional half frames into the stream.
This 3:2 pulldown and the additional half frames pushes the raw frame rate up to 30fps. The NTSC signal requires 29.97 fps, and so, the DVD player reading the SMTPE Drop flags, literally drops every 1001th frame. This brings the overall frame rate down to the 29.97 required for NTSC.
If you try to Inverse Telecine this, you end up getting 23.976 frames per second due to the dropped frame. You also lose some frame quality because the dropping of the frame interferes with the sequence of inserted half frames.
Most, if not all hardware players, slavishly follow this scheme. If they output progressive, then they generally DO use a simple line doubler to re-create progressive frames from the 29.97 source. Any that don't will be VERY rare - even at $1k - due to the memory and processing requirements needed to properly detect the pull-downs and dropped frames.
They do it like this because in an attempt to stop piracy (among other things) the film studios generally interleave fully interlaced sequences in amongst the 24fps progressive stream. This causes all sorts of fun when you are ripping...
Also, as said, if the source material was originally interlaced, the pull-down/SMTPE-DROP never happens, and the source is pure 29.97 interlaced. To make the progressive stuff work within a reasonable price range in real-time, they simply rely on a line doubler after the output stage. Some are better than others, but often not by much!
Bill.
I do actually know about the 23.976 and 29.97 fps stuff. I didn't want to confuse anybody reading the posts with extra details.
The thing about the interlacing stuff goes like this:
When a film source is converted to DVD, the studios generally take the 24fps film and scan it into progressive frames at the raw 24fps of the film itself. This is encoded directly into a 24fps MPEG2 video stream, with the addition of special "3:2 pulldown" flags and "SMTPE Drop" flags.
These pull-down flags tell the DVD player (ALL players) to do what they call 3:2 pulldown. This involves interlacing the progressive images and literally inserting additional half frames into the stream.
This 3:2 pulldown and the additional half frames pushes the raw frame rate up to 30fps. The NTSC signal requires 29.97 fps, and so, the DVD player reading the SMTPE Drop flags, literally drops every 1001th frame. This brings the overall frame rate down to the 29.97 required for NTSC.
If you try to Inverse Telecine this, you end up getting 23.976 frames per second due to the dropped frame. You also lose some frame quality because the dropping of the frame interferes with the sequence of inserted half frames.
Most, if not all hardware players, slavishly follow this scheme. If they output progressive, then they generally DO use a simple line doubler to re-create progressive frames from the 29.97 source. Any that don't will be VERY rare - even at $1k - due to the memory and processing requirements needed to properly detect the pull-downs and dropped frames.
They do it like this because in an attempt to stop piracy (among other things) the film studios generally interleave fully interlaced sequences in amongst the 24fps progressive stream. This causes all sorts of fun when you are ripping...
Also, as said, if the source material was originally interlaced, the pull-down/SMTPE-DROP never happens, and the source is pure 29.97 interlaced. To make the progressive stuff work within a reasonable price range in real-time, they simply rely on a line doubler after the output stage. Some are better than others, but often not by much!
Bill.
ok, so you know your stuff!
but all in all, a GOOD dvd player will work VERY well. but, a computer still owns. (however, a computer doesnt really do as well in the audio realm. but thats for another thread).
as far as the prisms go. i really just havent gotten back into it. its a lot of work for a not so great result. sure, it can look good, but you are still limited to 150:1 contrast ratio, unless you go out and get a NICE flat panel monitor and tear it apart. then, you still have 250:1 or so.
commerical projectors are insanely high. like 1000:1. we just cant compete with that. and considering a lcd panel with high res and high contrast ratio will set you back $400-$500 (at least), its hard.
and usually, those panels arent small. passing a sufficient amount of light through a 15" panel is really tough. then, you get into hot spot issues. the edges might not look so great.
so, im just getting a better job at an audio/video store, and gonna buy my sharp progressive scan projector sure, its $4K, but employee discounts rule.
but all in all, a GOOD dvd player will work VERY well. but, a computer still owns. (however, a computer doesnt really do as well in the audio realm. but thats for another thread).
as far as the prisms go. i really just havent gotten back into it. its a lot of work for a not so great result. sure, it can look good, but you are still limited to 150:1 contrast ratio, unless you go out and get a NICE flat panel monitor and tear it apart. then, you still have 250:1 or so.
commerical projectors are insanely high. like 1000:1. we just cant compete with that. and considering a lcd panel with high res and high contrast ratio will set you back $400-$500 (at least), its hard.
and usually, those panels arent small. passing a sufficient amount of light through a 15" panel is really tough. then, you get into hot spot issues. the edges might not look so great.
so, im just getting a better job at an audio/video store, and gonna buy my sharp progressive scan projector
sure, its $4K, but employee discounts rule.Agreed about the audio!
One thing to watch for on the projector front - though the DLP projectors can give exceptional contrast ratios, it would appear that for video work the LCD ones still rule - your first-hand experiences would be welcome.
Incidentally, I think that putting a second identical LCD panel directly above the first one (assuming a parallel light path through the LCD), you can DRAMATICALLY increase contrast without reducing brightness by too much.
The major losses in an LCD are through polarisation (50%) and the fact that each of the R, G and B shutters absorb 67% of the remaining light (green shutter absorbing blue and red light etc.). leaving a theoretical maximum of 16.5% of the original light to get through. Not good!
However, after the light passes through the first LCD, it is already polarised, and if properly aligned, the 67% absorbtion will not happen in the second panel because only the correct coloured light should reach the second layer of shutters.
Thus, for a minimal additional light loss, there would be a drastic increase in contrast. Interesting... I'm waiting for a second cheap twin for one of my panels to appear...
Bill.
One thing to watch for on the projector front - though the DLP projectors can give exceptional contrast ratios, it would appear that for video work the LCD ones still rule - your first-hand experiences would be welcome.
Incidentally, I think that putting a second identical LCD panel directly above the first one (assuming a parallel light path through the LCD), you can DRAMATICALLY increase contrast without reducing brightness by too much.
The major losses in an LCD are through polarisation (50%) and the fact that each of the R, G and B shutters absorb 67% of the remaining light (green shutter absorbing blue and red light etc.). leaving a theoretical maximum of 16.5% of the original light to get through. Not good!
However, after the light passes through the first LCD, it is already polarised, and if properly aligned, the 67% absorbtion will not happen in the second panel because only the correct coloured light should reach the second layer of shutters.
Thus, for a minimal additional light loss, there would be a drastic increase in contrast. Interesting... I'm waiting for a second cheap twin for one of my panels to appear...
Bill.
TWO PANELS.
I think there is one problem to using two lcds in sequence.
depending on the type of lcd, the two polarizers on the panel could be facing different directions. I can't remember specifics, but I think that tft lcd use the twisting nature of the crystal to change the polarization of the light passing through them. That is why some panels are clear when deactivated and some are opaque.
It would only work if you used two panels that were "clear" when not active.
I'm probably not making any sense. I don't even know what I just said!
I think there is one problem to using two lcds in sequence.
depending on the type of lcd, the two polarizers on the panel could be facing different directions. I can't remember specifics, but I think that tft lcd use the twisting nature of the crystal to change the polarization of the light passing through them. That is why some panels are clear when deactivated and some are opaque.
It would only work if you used two panels that were "clear" when not active.
I'm probably not making any sense. I don't even know what I just said!
Re: TWO PANELS.
well, you would HAVE to use to exact same panels... that is a given. i dunno about it though because even if they were touching, you may have some kind of focus issue, because after the fresnel, the light is being "seen" by the lens. so the light is coming from the panels to the lens like this: /\
so, the lens would not align the top and bottom panel's pixels right. so the picture would be fuzzy. the bottom panel would have to be just a little bigger, to compensate.
its hard to explain, maybe someone else could help me out... but, its like this... take a regular OHP. put a transparency on it... now, put ANOTHER one on top of it, with the same stuff... just raise it like 1/4" or 1/2" above the bottom... ONE will be out of focus...
well, you would HAVE to use to exact same panels... that is a given. i dunno about it though because even if they were touching, you may have some kind of focus issue, because after the fresnel, the light is being "seen" by the lens. so the light is coming from the panels to the lens like this: /\
so, the lens would not align the top and bottom panel's pixels right. so the picture would be fuzzy. the bottom panel would have to be just a little bigger, to compensate.
its hard to explain, maybe someone else could help me out... but, its like this... take a regular OHP. put a transparency on it... now, put ANOTHER one on top of it, with the same stuff... just raise it like 1/4" or 1/2" above the bottom... ONE will be out of focus...
tech head said:
Mario007,
If your fresnel is projecting an image, then you have one of two problems:
1) the rays coming through the LCD are already converging, having been focussed already;
2) the fresnel is further away from the LCD than its focal point.
If you place your fresnel in front of a laptop screen, very close to it (1cm), it will magnify the image and create a virtual image of the screen further away from your eye than the real screen. It is acting as a magnifying glass, and as such, will not project a real image.
If, however, the fresnel is moved further away from the lcd than its focal point, it will then project a real image.
On your projetion lcd, if the rays coming through the lcd are already converging then the fresnel will probably see the lcd as being further away than it really is (a factor of the converging rays). It may even appear further away than the fresnel's focal length - causing an image to be projected. The rays should be close to parallel if you are going to put your fresnel after the LCD.
Eebasist,
You WOULD need to closely align the panels, but if their polarisation characteristics are compatible and the light passing through them parallel, then it wouldn't be too bad - nothing that couldn't be done by hand in a couple of minutes. You also wouldn't need to strip them down - just place them as close as is possible.
That way any divergence of the light through each panel would have minimal effects.
Bill.
If your fresnel is projecting an image, then you have one of two problems:
1) the rays coming through the LCD are already converging, having been focussed already;
2) the fresnel is further away from the LCD than its focal point.
If you place your fresnel in front of a laptop screen, very close to it (1cm), it will magnify the image and create a virtual image of the screen further away from your eye than the real screen. It is acting as a magnifying glass, and as such, will not project a real image.
If, however, the fresnel is moved further away from the lcd than its focal point, it will then project a real image.
On your projetion lcd, if the rays coming through the lcd are already converging then the fresnel will probably see the lcd as being further away than it really is (a factor of the converging rays). It may even appear further away than the fresnel's focal length - causing an image to be projected. The rays should be close to parallel if you are going to put your fresnel after the LCD.
Eebasist,
You WOULD need to closely align the panels, but if their polarisation characteristics are compatible and the light passing through them parallel, then it wouldn't be too bad - nothing that couldn't be done by hand in a couple of minutes. You also wouldn't need to strip them down - just place them as close as is possible.
That way any divergence of the light through each panel would have minimal effects.
Bill.
Just to chirp in with other useful info. Regular Dish Network channels are broadcast at 480x480 while their PPV channels are 720x480. I can't remember all the specifics, but supposedly some of the 480x480 channels are broadcast at different bit-rates.
As for DVD's, I think MPEG2 is a sad format. I would much rather watch uncompressed 4:2:2 480p if they had to down convert to so low a resolution... Heck, maybe even compressed in a 4:2:2 50Mb DV codec...
As for DVD's, I think MPEG2 is a sad format. I would much rather watch uncompressed 4:2:2 480p if they had to down convert to so low a resolution... Heck, maybe even compressed in a 4:2:2 50Mb DV codec...
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.