Owen Posted September 11, 2004 Share Posted September 11, 2004 Recently I have made several posts regarding the inappropriateness of small displays for HD use and suggested that small displays did not need to be anywhere near 1920x1080 resolution, as you just can not see that resolution on a small display at a normal viewing distance. I made comments like “a 76cm TV is a waste of time for HD” and “ most people would be hard pressed to see much difference between a HD and SD Plasma at 3 meters”. This prompted understandably strong responses, with the suggestion that I was probably mad, stupid, arrogant, delusional or all of the previous. Well, I wanted to prove to myself that I am not mad, stupid or delusional. I already know that I am unashamedly arrogant. So I did some reliable and repeatable tests of my ability to see different resolutions from various distances. I decided to us my Notebook PC TFT display for the task, as it is a very high resolution digital fixed pixel display connected via DVI and will therefore provide guaranteed reliable 1:1 pixel resolution without any chance of a VGA or Component connection system affecting the results. The panel measures 30.5cm W by 22.5cm H (4:3 aspect ratio) and has a resolution of 1600x1200. If it where 16:9 aspect ratio is would have a resolution of 2133x1200 for the same pixel size and density and is therefore more then good enough to evaluate 1920x1080 resolution. I used a PC monitor test application called Display Mate to generate vertical and horizontal resolution test patterns at 1600x1200 resolution so as to ensure 1:1 pixel mapping for the display. I also generated a 800x600 pattern. I then tested from what distance I could resolve the test patterns. I did both 1600x1200 and 800x600 horizontal and vertical test patterns several times and took an average. The result was about 800mm average, so that at any greater distance the pattern looked like a gray screen with no discernable lines. From this information it is easy to extrapolate the maximum viewing distance for various size displays at 1920x1080. 76cm screen = approx 66cm x 37cm or 1.6 times bigger then test screen. Max viewing distance to resolve 1920x1080 1.6 x 0.8 =1.28 meters 106cm screen (42”) = approx 92cm x 52cm or 2.3 times test screen. Max viewing distance to resolve 1920x1080 2.3 x 0.8 = 1.84 meters 153cm screen (60”) = approx 133cm x 74cm or 3.3 times test screen. Max viewing distance to resolve 1920x1080 3.3 x 0.8 = 2.64 meters Now, before you all start in about me being blind, I tested my eyes in way that can be easy referenced to other people, as individual eyesight varies. I took a standard news paper (The Daily Telegraph) and tested at what distance I could reliably read the standard small text. For me, that tuned out to be about 1.7-1.8 meters and I would say that my eyesight is about average. You can test your own eyes and adjust my test results accordingly to suit you eyesight. The results of this simple test are surprising, and are even worse then I expected. Those people that think 60” displays are to big should think again, because if you view from any more then about 2.7 meters, you will not be able to see 1920x1080 resolution. This justifies my often used comment that true HD is designed for BIG displays. As for the little 76cm screens, well how many people view one of those at 1.3 meters. Not many I’ll bet. If you view from 2.6 meters, your eyes can only resolve about 1000x500 resolution and if you go to 3 meters you are getting down to about NTSC DVD resolution or 720v480. Even the very popular 42” size displays would need to be viewed from only 1.9 meters for 1920x1080 viewing, if or when such a high resolution device even becomes available. At 4 meters distance a 42” SD Plasma should look just about as good anyway. If you have a PC, I highly recommend you try this test yourself. You will then have a better understanding of how incredibly high a resolution 1920x1080 actually is and how important screen size and viewing distance is in being able to resolve it with the human eye. One should also keep in mind that almost zero source material is full 1920x1080 resolution anyway and even the best 1920x1080 source material has been heavily compressed and no longer contains and real detail that requires 1920x1080 anyway. That being said, it can still be useful for a digital display to have more pixels or resolution then is needed to resolve the source. More pixels can result in a smoother, more natural and less digital look if they are used correctly. I think we should be less preoccupied with the resolution specifications of displays and more interested in other more important and useful aspects or performance, such as solarization, pixilation, motion artifacts, shadow detail and black levels just to name a few. If you think my testing methodology is flawed, let me know and I will retest and post new results as required. Regards to all, Owen P.S. After writing all this crap I thought the maximum viewing distance numbers looked familiar, and they are. I remember seeing a viewing distance calculator for the THX and HDTV standards. Here is the link. http://www.myhometheater.homestead.com/vie...calculator.html The calculator confirms my viewing distance findings almost exactly. So I am not definitely not mad, stupid or delusional, just arrogant and smug. I would say that the above is vital information for anyone considering a HD TV purchase. You could save your self a lot of money by purchasing a SD display if you can’t sit close enough to take full advantage of a HD display. It appears that a 57”-60” screen is a minimum for true THX standard home theater, even in a small room with a 3 meter viewing distance. The problem is that almost all current digital displays of that size do not look good at that distance as they don’t have sufficient pixel density. They NEED to be 1920x1080. A properly set up and calibrated 57" 1080i CRT RPTV works just fine at 3 meters because it has NO pixels to worry about. Link to comment Share on other sites More sharing options...
86rayl Posted September 11, 2004 Share Posted September 11, 2004 Owen: I really appreciate your are being so passionate on doing all these testing for us, This experiment is actually quite interesting. According to your logical explanation and the mathetical calculation, i really agreed to what you have said. Picture Quality we percieved really depend on the human eye sight, and in combination with the viewing distance. If we Watch a Normal CRT TV at a few meters away, we cannot see the very detail part anyway. So it is clear. For my viewing distance on my 42 inches plasma, if i want to resolve the unclear quality, i usually sitting at 2-3 meters for my viewing distance, and that distance almost equal to what u have calculated. So From now on i don't need to worry too much about the panel resolution quality on mine that this has been annoyed me for a while... ( being honest ). you have done well, Owen !! Regard Ray. Link to comment Share on other sites More sharing options...
IanD1503559705 Posted September 12, 2004 Share Posted September 12, 2004 76cm screen = approx 66cm x 37cm or 1.6 times bigger then test screen.Max viewing distance to resolve 1920x1080 1.6 x 0.8 =1.28 meters Owen, Your results are interesting and seem to be similar to Sony's recommended viewing distances for widescreen TVs (which are quite different from 4:3 TVs). IIRC, Sony suggested viewing widescreen displays at 3x the image height for maximum "immersive" experience. This equates to roughly 1.1m for a 76cm diagonal TV, if my calculations are correct, and compares favourably to your 1.28m resolvability figure. Of course, it wasn't until the advent of HDTV that these sorts of viewing distances could be achieved without individual pixel structure being noticeable. I would guess that widescreen was really designed for HDTV for maximum impact. and I imagine the majority of widescreen TV adopters are still viewing the new TV at the same viewing distances as they used for their old 4:3 TV. This is understandable when an SDTV's pixel structure becomes obvious at the recommended viewing distance for widescreen. I wonder how many HDTV adopters are viewing at the recommended distances? Ian Link to comment Share on other sites More sharing options...
Seak Posted September 12, 2004 Share Posted September 12, 2004 BBC has just released a white paper (WHP092) titled: Tests of visual acuity to determine the resolution required of a television transmission system by John Drewery and Richard Salmon Abstract Experiments have been carried out to determine the human eye’s visual acuity, using fragments of pictures containing textures of various scales, and displayed on a cathode ray tube. Care has been taken to ensure that the display properties have not influenced the result. The results confirm the widely held value of 1 minute of arc, and can be used to determine the TV standards needed in the coming large-screen display age. It is concluded that the 1280x720 transmission format will suffice for the foreseeable future for domestic requirements in Europe. Download the 4MB PDF from: http://www.bbc.co.uk/rd/pubs/whp/whp092.html Link to comment Share on other sites More sharing options...
Dave T Posted September 12, 2004 Share Posted September 12, 2004 Recently I have made several posts regarding the inappropriateness of small displays for HD use and suggested that small displays did not need to be anywhere near 1920x1080 resolution, as you just can not see that resolution on a small display at a normal viewing distance.I made comments like “a 76cm TV is a waste of time for HD” and “ most people would be hard pressed to see much difference between a HD and SD Plasma at 3 meters”. This prompted understandably strong responses, with the suggestion that I was probably mad, stupid, arrogant, delusional or all of the previous. Well, I wanted to prove to myself that I am not mad, stupid or delusional. I already know that I am unashamedly arrogant. So I did some reliable and repeatable tests of my ability to see different resolutions from various distances. I decided to us my Notebook PC TFT display for the task, as it is a very high resolution digital fixed pixel display connected via DVI and will therefore provide guaranteed reliable 1:1 pixel resolution without any chance of a VGA or Component connection system affecting the results. The panel measures 30.5cm W by 22.5cm H (4:3 aspect ratio) and has a resolution of 1600x1200. If it where 16:9 aspect ratio is would have a resolution of 2133x1200 for the same pixel size and density and is therefore more then good enough to evaluate 1920x1080 resolution. I used a PC monitor test application called Display Mate to generate vertical and horizontal resolution test patterns at 1600x1200 resolution so as to ensure 1:1 pixel mapping for the display. I also generated a 800x600 pattern. I then tested from what distance I could resolve the test patterns. I did both 1600x1200 and 800x600 horizontal and vertical test patterns several times and took an average. The result was about 800mm average, so that at any greater distance the pattern looked like a gray screen with no discernable lines. From this information it is easy to extrapolate the maximum viewing distance for various size displays at 1920x1080. 76cm screen = approx 66cm x 37cm or 1.6 times bigger then test screen. Max viewing distance to resolve 1920x1080 1.6 x 0.8 =1.28 meters 106cm screen (42”) = approx 92cm x 52cm or 2.3 times test screen. Max viewing distance to resolve 1920x1080 2.3 x 0.8 = 1.84 meters 153cm screen (60”) = approx 133cm x 74cm or 3.3 times test screen. Max viewing distance to resolve 1920x1080 3.3 x 0.8 = 2.64 meters Now, before you all start in about me being blind, I tested my eyes in way that can be easy referenced to other people, as individual eyesight varies. I took a standard news paper (The Daily Telegraph) and tested at what distance I could reliably read the standard small text. For me, that tuned out to be about 1.7-1.8 meters and I would say that my eyesight is about average. You can test your own eyes and adjust my test results accordingly to suit you eyesight. The results of this simple test are surprising, and are even worse then I expected. Those people that think 60” displays are to big should think again, because if you view from any more then about 2.7 meters, you will not be able to see 1920x1080 resolution. This justifies my often used comment that true HD is designed for BIG displays. As for the little 76cm screens, well how many people view one of those at 1.3 meters. Not many I’ll bet. If you view from 2.6 meters, your eyes can only resolve about 1000x500 resolution and if you go to 3 meters you are getting down to about NTSC DVD resolution or 720v480. Even the very popular 42” size displays would need to be viewed from only 1.9 meters for 1920x1080 viewing, if or when such a high resolution device even becomes available. At 4 meters distance a 42” SD Plasma should look just about as good anyway. If you have a PC, I highly recommend you try this test yourself. You will then have a better understanding of how incredibly high a resolution 1920x1080 actually is and how important screen size and viewing distance is in being able to resolve it with the human eye. One should also keep in mind that almost zero source material is full 1920x1080 resolution anyway and even the best 1920x1080 source material has been heavily compressed and no longer contains and real detail that requires 1920x1080 anyway. That being said, it can still be useful for a digital display to have more pixels or resolution then is needed to resolve the source. More pixels can result in a smoother, more natural and less digital look if they are used correctly. I think we should be less preoccupied with the resolution specifications of displays and more interested in other more important and useful aspects or performance, such as solarization, pixilation, motion artifacts, shadow detail and black levels just to name a few. If you think my testing methodology is flawed, let me know and I will retest and post new results as required. Regards to all, Owen P.S. After writing all this crap I thought the maximum viewing distance numbers looked familiar, and they are. I remember seeing a viewing distance calculator for the THX and HDTV standards. Here is the link. http://www.myhometheater.homestead.com/vie...calculator.html The calculator confirms my viewing distance findings almost exactly. So I am not definitely not mad, stupid or delusional, just arrogant and smug. I would say that the above is vital information for anyone considering a HD TV purchase. You could save your self a lot of money by purchasing a SD display if you can’t sit close enough to take full advantage of a HD display. It appears that a 57”-60” screen is a minimum for true THX standard home theater, even in a small room with a 3 meter viewing distance. The problem is that almost all current digital displays of that size do not look good at that distance as they don’t have sufficient pixel density. They NEED to be 1920x1080. A properly set up and calibrated 57" 1080i CRT RPTV works just fine at 3 meters because it has NO pixels to worry about. You don't have a HD Crt do you Owen? If you did you would'nt be playing with your laptop or other things in that region, instead you would be enyoying PQ only dreamed of with other formats! Link to comment Share on other sites More sharing options...
DannyR Posted September 12, 2004 Share Posted September 12, 2004 Very interesting Owen, very intersting indeed. I have been watching the HD vs SD 42" plasma debate very closly, and agree at the normal 3 to 4 meter viewing distance, its hard to tell the differece, as a videophile, I can, and think a HD panel is worth while, but as you say, at 1080p its a waste of time, a 720p panel is plenty good enough. And I can't see screens getting any bigger for the vast majority of consumers. There is a point where huge screens are only good for a specialized home theatre rooms, for which 1080p Blu-Ray movies can be used for source material. I would have to say I am now almost convinced we are wasting out time with 1080i HD transmission, and think we should change to 720p50 and the advantages progressive scan brings. Danny. Link to comment Share on other sites More sharing options...
Vermin Posted September 12, 2004 Share Posted September 12, 2004 Interesting experiment, any idea how this relates to dynamic images (moving/changing pictures)? V. Link to comment Share on other sites More sharing options...
digitaladvisor Posted September 12, 2004 Share Posted September 12, 2004 Very interesting Owen, very intersting indeed.I have been watching the HD vs SD 42" plasma debate very closly, and agree at the normal 3 to 4 meter viewing distance, its hard to tell the differece, as a videophile, I can, and think a HD panel is worth while, but as you say, at 1080p its a waste of time, a 720p panel is plenty good enough. And I can't see screens getting any bigger for the vast majority of consumers. There is a point where huge screens are only good for a specialized home theatre rooms, for which 1080p Blu-Ray movies can be used for source material. I would have to say I am now almost convinced we are wasting out time with 1080i HD transmission, and think we should change to 720p50 and the advantages progressive scan brings. Danny. Yes it does make NOT much sense as all plasmas even those considered HD ALL convert the signal to progressive. In fact all LCDs and RPTV LCDs and DLPs convert to progressive. All video projectors (except the bulky CRT projectors weighing a ton) convert to progressive as well. Even a 720p transmission is easier to translate better, less artifacting and less fast movement inter frame loss. Interlace is an invention derived from old CRT technology. All ground rules change when it comes to digital displays. The screen door effect on digital DLP RPTVs do not apply. Regards DA Link to comment Share on other sites More sharing options...
rochford Posted September 12, 2004 Share Posted September 12, 2004 Some very interesting points. However, I would suggest the key issue is not the ability to distinguish one pixel from another (as a test pattern would demonstrate). The issue is whether the eye can view a higher resolution image in a 'better' way than a lower resolution image - in other words, does the loss of resolution result in 'identical' images to the human eye, or two slightly different 'downsampled' images (I'm assuming HD would be better than SD here). I suspect this is why downsampled HD ostensibly looks better than native SD. You could consider this to be a similar issue to colour depth - the eye cannot distinguish all 16.7 million colours in a 24 bit colour palette when shown together, but a 24 bit colour palette produces better results than a palette that matches the eye's capability (in the low millions of colours). The higher the original quality, the better the downsampling, so again, I would expect two slightly different 'downsampled' images. All of this said, 1080i is very bandwidth intensive, and the networks do not really give it enough to shine; 720p50 would be a much better choice IMO (and I know this debate has been done to death...), as it would be the best compromise between resolution, frame rate and available bandwidth. Or am I completely barking up the wrong tree here? - Miles. Link to comment Share on other sites More sharing options...
betty boop Posted September 12, 2004 Share Posted September 12, 2004 You don't have a HD Crt do you Owen? If you did you would'nt be playing with your laptop or other things in that region, instead you would be enyoying PQ only dreamed of with other formats! No don't think he has a crt. My bet is he has a "A properly set up and calibrated 57" 1080i CRT RPTV" hehehe Actually Owen wish this whole thing was as simple as you make it out to be. Plug in the screen size into the calculator and hey presto thats the room size/viewing distance you should put it in. Or given the room dimensions(viewing distance) so this is the size screen you should get. Doesn't work like that in reality. These are some of other factors that probabaly for most people decide what screen size to get : 1. Is it going into a dedicated HT room or a lounge room - not everyone is going to want a 60" behemoth in their lounge room 2. The screen size being replaced - people don't usually go from a 51cm TV to a 60" plasma, they usually go a couple of sizes up. 3. Finances - not everyone can afford a 60" display 4. Type of material that is going to be watched on this display - if your mostly watching SD (which is the case for the majority) its going to look pretty ordinary on a 60" display at the recommended viewing distance. 5. WAF (maybe I should put this top of the list) - don't think I need to say anymore here 6. Room size/viewing distance possible - not everyone is comfortable sitting 2.4m from a 60" diplay - bit like being at the front row at the cinema 7. Amount of space the display itself is going to occupy - not everyone wants a 60" behemoth taking up valuable space in their home especially if its a block of flats crt RPTV Most people buy a display size they are comfortable with and place it at a viewing distance they are comfortable with. Probably the biggest flaw with the viewing distance argument is the huge discrpancy in recommended viewing distance between SD 720x576 and HD 1920x1080 on the same size display. I bet you do not move your couch up to 2.1-2.4m from your 57"RPTV for watching HD 1920x1080 and then move it out to 6.4m for watching SD 720x576. If I did that I'd need to put my couch on wheels. The other problem with this calculator is it does not provide minimum viewing distances - what distance should I view my display at without seeing individual pixels. I bet you this is something people do consider and they make this judgement based on SD usually (as thats just about all thats on at the moment) - probably to the detrement of their HD viewing, however I think this is a sacrifice many people are willing to make - for instance your not trying to tell me that SD would not look like absolute sh!t at 2.1m on a 60" display even if it was capable of 1920x1080 Guess what I'm saying is yes the viewing distance calculator - good tool, handy to have more information but I think people might just as well choose other wise on display size and viewing distance for a whole lot of other reasons. Link to comment Share on other sites More sharing options...
ritesh Posted September 12, 2004 Share Posted September 12, 2004 I bet you do not move your couch up to 2.1-2.4m from your 57"RPTV for watching HD 1920x1080 and then move it out to 6.4m for watching SD 720x576. If I did that I'd need to put my couch on wheels. Alebonau, good point AND a similar point is the resolution of the Digital Displays (plasma, LCDs etc.). Ideally, we would want the resolution to be different for different types of material (SD, or HD), so as to avoid any scaling artifacts, and mix it up with movement of couches in the lounge rooms.... Rockford, you have raised some interesting points as well, I totally agree with the following: The issue is whether the eye can view a higher resolution image in a 'better' way than a lower resolution image - in other words, does the loss of resolution result in 'identical' images to the human eye, or two slightly different 'downsampled' images (I'm assuming HD would be better than SD here). I suspect this is why downsampled HD ostensibly looks better than native SD. Owen, whilst information about viewing distances and screen sizes are available at a lot of places, it still is an interesting experiment. I hope you had a helper to exactly calculate your viewing distance, regarding "resolving" resolution with your eyes, did you try playing with the color, sharpness, contrast, brightness and other settings? They might have some effect as well, especially when you are extrapolating your results.... Whilst the recommended viewing distance for a screen type is an easy information to obtain, I am NEVER watching anything that is larger than 40'' in size from 1.5 or 2 meters ! That is like getting a front row seat in a cinema. Thus I have different question, IF we are watching a 42'' display from 3-4 meters (which is the usual distance in most households, I imagine). What should be the resolution of the display be?... probably 100x100, eh I agree with you, there are LOTS of other more important criteria to PQ than the mere numbers.... cheers, Ritesh Link to comment Share on other sites More sharing options...
Owen Posted September 12, 2004 Author Share Posted September 12, 2004 Interesting experiment, any idea how this relates to dynamic images (moving/changing pictures)?V. Vermin, Its more difficult for the eye to discern fine detail in movement. Combine that with the dramatic loss of detail caused by Mpeg2 compression on motion scenes and you wind up with much lower resolution requirements for motion. Regards, Owen Link to comment Share on other sites More sharing options...
Santa1503559644 Posted September 12, 2004 Share Posted September 12, 2004 Hey! I've ranted about most of the subjects covered above, in various threads over quite some time. Firstly, to correct digitaladvisor, most respectable CRT projectors can and do output progressive at high resolution. (While the baseline 7" models may be pushing it at 1080p, anything above this can easily cut it - and the modern 9" CRTs can LAUGH at ~2000p) WRT the resolution and distance arguments, I've been harping on about these for over a decade to anyone who'd listen! In the end, though, it all boils down to two things: 1) What YOU prefer to view. (And your eyesight/eye-brain complex) 2) How immersive an experience you want. Personally, I've gone for the "real thing" - the recommended ~33 degree viewing angle afforded by a 112" screen viewed from aprox 4m. I can hereby confirm that this size requires really good SD to look good - but it CAN look really good on SD if the quality is there. HD = icing on the cake (provided it is decent bitrate). The point of HD was to enable truly big screen displays (NOT 42" plasmas, which most people think=home cinema). If you want a pearler of an image, you project a 1920x1080 image onto a 90" screen, and view from ~4m. If you want a little more immersion, you bump up the size to 110-120". However, there is no problem with putting a higher definition signal onto a smaller display - you're not going to get a worse picture from over-doing the resolution for any screen-size, as opposed to going to far the other way (too big screen)*. ________ *Certain exclusions apply. See in-store for details. Link to comment Share on other sites More sharing options...
Owen Posted September 12, 2004 Author Share Posted September 12, 2004 QUOTE rochford "Some very interesting points. However, I would suggest the key issue is not the ability to distinguish one pixel from another (as a test pattern would demonstrate). The issue is whether the eye can view a higher resolution image in a 'better' way than a lower resolution image - in other words, does the loss of resolution result in 'identical' images to the human eye, or two slightly different 'downsampled' images (I'm assuming HD would be better than SD here). I suspect this is why downsampled HD ostensibly looks better than native SD. You could consider this to be a similar issue to colour depth - the eye cannot distinguish all 16.7 million colours in a 24 bit colour palette when shown together, but a 24 bit colour palette produces better results than a palette that matches the eye's capability (in the low millions of colours). The higher the original quality, the better the downsampling, so again, I would expect two slightly different 'downsampled' images." That’s good point Miles, and there may be something to it. A good test would be to see if we can distinguish a single different colored pixel out of contrasting colored background. For example, a white, red, green or blue pixel on a black, white or contrasting colored background. An example of this is a defective pixel on a digital display. If someone has a digital display with a defective-stuck pixel, can they tell use the size and resolution of the display and at what distance the defective pixel becomes visible. The main reason that HD video downscaled and displayed on a SD display looks good, is because the downscaled HD video is taking full advantage of the maximum resolution of the SD display. The normal SD transmission is almost never full PAL resolution. The vertical resolution is fixed at 576 lines for the PAL system, but that does not mean that every line carries different information. A PAL VHS VCR is only good for about 280 lines of resolution but always outputs 576 lines to the TV. Horizontal resolution also varies wildly, and is rarely full 768 due to poor source material or Mpeg compression. Another good example of this is Foxtel-Austar digital. To be PAL compliant, video output must be 576 lines total but only about 500-525 lines are normally visible due to deliberate overscan on the display. The resolution however is nowhere near 768x576 due to the very heavy compression used for transmission. The free to air digital SD transmissions are also Mpeg2 compressed and are never full 768x576 resolution, although they are much better then Foxtel due to a much higher data rate. PAL DVDs are 720x576 pixel resolution but they almost never reach that screen resolution due to Mpeg compression. As you can see, Mpeg compression is the enemy of quality, and a major limitation with all digital video systems. The standard analogue TV transmission is actually capable of greater resolution then digital as it is uncompressed, but it suffers from interference problems and is still only as good as the signal going in at the transmission end. This explains why HD downscaled to SD and displayed on an SD TV looks better then the SD transmission. As for 720p/50 using less bandwidth, that is unfortunately not correct. 720p at 50 frames per second used MORE bandwidth then 1080i 25 frames per second. Maybe you where thinking of 720p/25. Regards, Owen Link to comment Share on other sites More sharing options...
rochford Posted September 12, 2004 Share Posted September 12, 2004 As for 720p/50 using less bandwidth, that is unfortunately not correct. 720p at 50 frames per second used MORE bandwidth then 1080i 25 frames per second. You are absolutely right. My point was really that it is a better compromise in terms of using the available bandwidth - although it too can suffer if not given enough bits. In short, I'd rather have 50fps (even 25p?) progressive scan video, even if it is technically a lower vertical resolution than 1080i. It's better bang for bucks. - Miles. Link to comment Share on other sites More sharing options...
Owen Posted September 12, 2004 Author Share Posted September 12, 2004 alebonau, The point of this thread was to make people aware of the limitation of there vision and also make them aware of the importance of viewing distance when choosing a display. They need to know that purchasing a small HD display to be viewed at a considerable distance may be waste of money. It was not my intension to tell people what size display they should buy, just make them aware of the limitations. As for screen size, I owned a 109cm 4:3 TV for 4 years and viewed it from 2.5 meters. The 109cm had a screen height not much smaller then my current 57” 16:9 TV ( yes I do now have “A properly set up and calibrated 57" 1080i CRT RPTV” how intuitive of you to figure that out ) and I got used to the screen size very quickly. I soon leaned that for very wide screen DVD’s, like the 2.35:1 aspect ratio ones, I needed a display about double the size that I had. I understand that most people are put off by the idea of a large screen, but once you have lived with one for a short time you would never want a small screen again. Flat panel displays take up almost no room and can be hung on a wall, so they can easily be used in small rooms. Just because they are expensive does not make them a bad idea. My RPTV is bulky, but it is placed in the corner of a small room at 45deg to the walls so that all I can see of it is the front panel. The depth and bulk are hidden very effectively. I have seen plenty of entertainment units bigger then my TV. All that aside, the only thing that is important to me is picture quality and the viewing experience. I can say that viewing a 57” TV from 3 meters with a 26 deg viewing angle is definitely not like being in the front row at the cinema. In fact, it’s like being in the back row. If you read the information on the page I linked it explains about recommended viewing angles for theaters, and states that a 26deg viewing angle is the MINIMUM requirement for the back row seating for THX certification and recommends a much larger 36deg viewing angle. I should add that DVD’s upscaled and processed by my HTPC look outstandingly good if they are of good quality, and most are these days. So good if fact, that most people think they are HD. Foxtel “digital” is very poor quality, and looks fairly ordinary at 3 meters but is perfectly watch able. There is no doubt that big screens are not for everyone for various reasons, but I do get the impression that many people buy smaller TV’s because that is what they are used to seeing. Not just because of space or monetary constraints. They don’t know what they are missing. Regards, Owen Link to comment Share on other sites More sharing options...
ChrisM Posted September 12, 2004 Share Posted September 12, 2004 Another issue in relation to screen size is how the original material was shot to be viewed. Cinema is shot quite differently to TV. TV programs use a lot of close-ups and action often jumps around the screen. This is because the director and camera operator are aware that, on an average TV, people will not be able to see fine detail in wide shots. Lots of close ups are inter-cut into the action for this reason. Material shot for the cinema doesn't use close-ups in the same way as TV, because the big screen allows the audience to see more detail. Movie directors and cinematographers are also aware that action needs to be carefully staged so that the audience don't have to rapidly look back and forth across a big screen on edits. To do this they maintain screen position on cuts and use camera cranes, tracks etc. to provide more fluid movement. Rapid cuts and altering screen position are not such a big deal on TV, because your eye's don't have to move as far on a small screen. Watching a large TV image, even if it's HD, is hard work if the image was not shot with the size in mind. On many "live" type programs, like sport, you'll probably be looking at a lot of soft shots too, as it is really hard to hold focus on HD in some situations. I presently have an SD 16:9 CRT display for general TV viewing and a PJ for DVD movies and some HDTV, and I find it's a good combination. Ideally I'd like a 576p+ flat panel TV and a 1920x1080P PJ. Link to comment Share on other sites More sharing options...
Owen Posted September 12, 2004 Author Share Posted September 12, 2004 Good info Chris, You sounds like your have a background in TV production. Regards, Owen Link to comment Share on other sites More sharing options...
kootaberra Posted September 13, 2004 Share Posted September 13, 2004 As for 720p/50 using less bandwidth, that is unfortunately not correct. 720p at 50 frames per second used MORE bandwidth then 1080i 25 frames per second. You are absolutely right. My point was really that it is a better compromise in terms of using the available bandwidth - although it too can suffer if not given enough bits. In short, I'd rather have 50fps (even 25p?) progressive scan video, even if it is technically a lower vertical resolution than 1080i. It's better bang for bucks. - Miles. There seems to be a lot of varried views about this subject. For those of you who have seen full bandwidth 1080i on a professional HD monitor (CRT), you should know the doemstic HD displays just don't do HD justice. What the TV stations and production people know is that at the moment there is only 2 standards to use at the moment. These are 1080i/50 or 60 and 1080p/24 or 25. There will be 1080/50p or 60p shortly, but as noted the bandwidth is doulbled. From these, the emission format can be down con-vertered to 720p with minimum loss. As for the 25p or 50p question, we can go back to film to get the answers. The optimum frame rate to make the flicker seem to stop, is about 60 frames per sec. This is what IMAX use. Now the normal old film industry is stuck with 24 frames per second. They do how ever try and reduce the flicker by using 2 shutters or even 3 shutters rotating in front of each frame as it is projected. What TV 24p does is give us 24 frames per sec, but the 100hz screens try and minimise the flicker. At 50p the flicker will be about as good as it gets. The trouble is no cameras have 50p standard (yet). Most have 24, 25 or even 30p. Other reason 24/25p is popular as appossed to 50/60p, is a 24p tape can be played straight to air from a 1080i or 1080p/25 VTR with a 4% speed change. When we can shoot and produce in 1080p/50 or 60 and transmitt in 1080p or down-convert to 720p/50 or 60, then you should see the difference. In Japan they are aready looking to twice the resolution of 1080p, refered to as 4K resolution. Link to comment Share on other sites More sharing options...
ChrisM Posted September 13, 2004 Share Posted September 13, 2004 Imax is actually 24fps but a format called Showscan was 70mm film (Like Lawrence of Arabia) at 60fps. (This next bit's a bit off topic) I saw a demo of Showscan in LA. In the demo they rolled the projector and then, (shock, horror), the film stopped in the gate and the film melted - leaving a white screen. The projector lamp was shut down, the screen went black and there was an apology over the PA. The lights went on behind the screen and I could see a guy walking around between the speaker boxes etc. He walked up to the screen and apologised for the delay. It was only after about 30 seconds I realised that the "behind the screen" image was actually the projected film image - it was that realistic! After that trick they went into the actual demo reel. 70mm at 60fps is the best PQ I've ever seen! Link to comment Share on other sites More sharing options...
Dik Posted September 15, 2004 Share Posted September 15, 2004 If someone has a digital display with a defective-stuck pixel, can they tell use the size and resolution of the display and at what distance the defective pixel becomes visible OK I admit it. My Panny PT-AE500 isn't perfect. It has a stuck pixel, so I have done the experiment. Lets get a bit of background information out of the way first. My eyesight isn't as good as it once was and I use reading glasses but not for TV viewing. I have had a Panny 10A 76cm HD CRT for nearly 3 years and would swear blind that 1080iHD native is better than 1080iHD station upconvert is better than 1080iHD STB upconvert is better than 576i (is better than 576p) My viewing distance is around 4 metres and this is the same for both the 76cm CRT and 230cm screen (112 cm screen height). Resolution for the PJ is 720x1280 pixels The "lazy" pixel does not show up under all conditions/colours. The pixel is cyan when it is visible, but is totaly invisible on black or blue and faint on red. It shows best on colours with a high level of cyan but also contrasting well with it. Browns and greens are best to find it on. Under normal circumstances I do not see it because I am watching the whole picture not the pixel, but if I go looking for it I can clearly see it at 4 metres (3.6 x screen height) if the background colour is right. So the experiment. Pause a picture with wood panneling on the portion of the screen where the pixel is. Focus on the pixel and start moving back. At 6 metres (5.3 x screen height) in a darkened room the pixel is still clear. Open the door and go outside. I finaly lost the pixel at 9.5 metres. To find it again I was back to around 5 metres (4.5 x screen height), but once I had focused on it I could move back to 9.5 metres (8.5 x screen height) keeping it in view again. So, the ability to actually focus on an individual pixel when you really try is quite remarkable, but it isn't the pixel you are looking at it's the picture. And a good picture is about it being smooth and seamless not the pixel structure being hard to focus on. As we all know downconverted pictures always seem to look beter than upconverted ones. This is because of redundency. By viewing a small HD display we should not be judging it on whether there is an ability to see the pixel structure but the percentage of redundency in the picture as viewed. Surely a picture with 50% redundency (ie 200% better than the eye can see) is going to look better than one with 0% redundency or even 10% redundency. When we look at things in the real world we never see all the detail, but that doesn't make a rose or a mountain any the less beautifull when viewed at a distance just because you can't see every petal or every rock. Best of luck, Dik Link to comment Share on other sites More sharing options...
Santa1503559644 Posted September 15, 2004 Share Posted September 15, 2004 Rolf! Nice experiment ... 9.5 meters! Luckily you dont have a busy road outside! (ala this week's Six Feet Under) Link to comment Share on other sites More sharing options...
alanh Posted September 15, 2004 Share Posted September 15, 2004 See my new post in this strand on testing colour vision AlanH Link to comment Share on other sites More sharing options...
IanD1503559705 Posted September 16, 2004 Share Posted September 16, 2004 From these, the emission format can be down con-vertered to 720p with minimum loss. As for the 25p or 50p question, we can go back to film to get the answers. The optimum frame rate to make the flicker seem to stop, is about 60 frames per sec. This is what IMAX use. Now the normal old film industry is stuck with 24 frames per second. They do how ever try and reduce the flicker by using 2 shutters or even 3 shutters rotating in front of each frame as it is projected. With motion estimation and interpolation available now, is 24fps from film really a limitation? It shouldn't be too difficult to create any desired framerate with smooth motion: it just requires the will and effort to actually do it. I'm disappointed that not enough effort is put into source authoring to ensure relatively simple playback techniques are required to achieve excellent progressive results: sophisticated de-interlacers are only required because of shortcuts taken in authoring (including not using motion interpolation to convert from NTSC to PAL) IMO. It seems to me that the cost of excellent results is being shifted onto the consumer instead of being handled at the source where it would be more cost effective overall. Ian Link to comment Share on other sites More sharing options...
BEST1503559813 Posted September 24, 2004 Share Posted September 24, 2004 Interesting Comments Owen. I currently own a Tosh 32SW9UA, Tosh HD25 did own a cheap NEC LCD projector but recently sold it waiting for projectors to be released with HDMI connectivety. My viewing distance is 2.5M. I have viewed various sources from and can tell the difference. From best to worst. 1080I HD TV (actualy prefer 720P for footy etc, change output of Tosh box) 576P HD TV 576D DVD 576I DVD PAL SD Foxtel via SVIDEO Laserdisc via SVIDEO (remember this old format, looks aweful but in mid 90s was the best) VHS There are lot of clues when watching different sources such as digital artifacts, blackness, colour saturation. I bought the TOSH box as it output a signal to SVIDEO even in HD mode so I can record on my DVD recorder. LG did not offer this feature. Did not buy DVD or STB with HDD as with burnt DVDs you can lend out if other people who missed the show. Normally kick wife out of loungeroom to familyroom to watch recorded stuff. Some shows look aweful in HD eg Friends (is definitely upscaled) regards BEST (Better Engineered Sound Today) Equipment: Tosh 76CM TV 32SW9UA Tosh SD25 HD STB Tosh DR1 DVD Recorder Pioneer Laserdisc CLD-D790 Yamaha DVD 5 Disc C920 Denon AVR2803 Denon AVD2000 (used as AC3RF switch box for Laserdisc player) Foxtel Digital STB (Pace) 2 X Aaron Sub120 (subwoofers) 6 X Aaron SS120 1 X Aaron C120 (centre speaker) Link to comment Share on other sites More sharing options...
Recommended Posts