1. Join Now

    AVForums.com uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.

CRT sweetspot and HCPC's

Discussion in 'Projectors, Screens & Video Processors' started by JohnAd, Dec 5, 2001.

Thread Status:
Not open for further replies.
  1. JohnAd

    JohnAd
    Active Member

    Joined:
    Mar 14, 2001
    Messages:
    362
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    18
    Location:
    London
    Ratings:
    +4
    I've noticed a few threads recently where people are talking about the best CRT resolutions and wanted to check that I have this right in my head before I dare pronounce on the subject. And also to help me get ready for my own CRT baptism next year.

    To me this is a two part problem.

    1) You should make sure that the scan lines just overlap. This means that for each projector there is usually 1 perfect number of horziontal lines for 4:3 and one for 16:9 in a given setup.

    2) The horizontal resolution on the other hand is different. CRT is an analogue device and is able to resolve detail up to a certain frequency. Now if we talk about DVD only then the maximum frequency of the actual data in the source is 5.5Mhz or 360 distinct lines. PC scaling doesn't increase this resolution it only gives a better (or worse) approximation of it using pixels which have hard edges rather than a nice analogue waveform.

    If the above is true then to get the best picture you need to first find your perfect number of horizontal lines. Then you need to put as many pixels as you can in each line to get the truest representation of the real information in the 720 pixels as you can. This means that you should use the higest from 1080, 1440, 1800, 2160 that you can for video even if your projector doesn't fully resolve the detail in the desktop pixels at that refresh rate. You will resolve the actual detail in the DVD but not put in any pixellation or scaling based artefacts if you do this.

    In fact if you can see the pixel edges then you need to goto a higher resolution if any of this is true.

    Note that this is assuming you only want to watch video at this resolution. Quick experiments on a CRT monitor seem to confirm for me that video scaled to 1800 horizonatal pixels looks better than 720 even though the windows desktop looks very blurry.

    Am I mad? Or does this sound right to people? The consensus seems to be that you want to be able to see the pixel structure but this seems wrong to me.

    John
     
  2. Jenz

    Jenz
    Guest

    Products Owned:
    0
    Products Wanted:
    0
    Ratings:
    +0
    No John, this is not strictly the case.

    What you need to do is watch the Movie not the Pixels :D

    My preference is that the Scan Lines do NOT overlap. Overlapping to me means that the detail of one line is blocking the detail of the next line.

    Your overlapping and ability to see scanlines will be different depending on 4:3 and 16:9. Personally on my Projector it is 960p and 720p respectively. But it also depends on you yourself.

    Your theory on Horizontal Pixels also stacks up to, but in practice I believe there is little to no visible difference in picture quality. In fact the more number of horizontal pixels the harder the projector has to work to draw it.

    I prefer the theory that a pure number of multiples (ie a whole number) in both the horizontal and vertical directions gives a better picture. This would result in resolutions of 1440x720, 1440x960 etc.

    Ultimately your CRT will be the deciding factor.

    As an aside if you ever fancy putting it to the test pop up to me in Milton Keynes one Sunday afternoon! We could have a mini-CRT-Resolution Event...

    Regards Neil.
     
  3. JohnAd

    JohnAd
    Active Member

    Joined:
    Mar 14, 2001
    Messages:
    362
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    18
    Location:
    London
    Ratings:
    +4
    Neil

    Thanks a lot for the offer, may well take you up on that in the new year :D

    You're right about the overlap thing I suppose I meant touch rather than overlap.

    Not sure about your comment about the projector doing more work with a greater number of pixels. I would think the projector does less work at you approach a true analogue represention of the input. This is because square waves are the hardest thing for any analogue device to handle. Since we're using HCPC's we can't help but use pixels but we can help to minimize that effect.

    John
     
  4. Roland @ B4

    Roland @ B4
    Active Member

    Joined:
    Apr 11, 2001
    Messages:
    2,238
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    51
    Location:
    Reading.
    Ratings:
    +32
    Think of the Phosphor of the tube surface as blotting paper. and the beam of electons as an old fashoned ink pen.

    (I dare say the vast majority of forum members are now asking their Dad, Whats an ink pen?)

    If you draw a line slowly the ink will flow and make the aparent line fatter as the ink flows out, the faster you draw the sharper the line becomes because there is less ink speading.

    Same is true with the CRT. Faster you go sharper it becomes but less light. but you can draw more lines. More lines gives more light so it sort of evens out.

    If the lines touch or overlap they will not decay quickly enough for the next line to draw onto unexcited phospher and you can end up with blooming as ajacent phopher areas excite.
    What you are trying to create is a very fine black line of non excited phophor between each line of video information.

    This is why so much attention is put on the vertical resolution of the image.

    I haven't done a true comparision yet but the Horizontal makes very little differnce to the projector. the PC will of course find it easier to be scalling up in whole numbers.

    My experience so far is that 1024, 1152, 1280 and 1440 look very similar on the same CRT device.

    Each projector is differnt so I can't give exact figures for each machine. but if you can't see scan lines then the resoloution is too high.


    As Jenz says the idea is not to watch the pixels, but to watch the film.
     
  5. JohnAd

    JohnAd
    Active Member

    Joined:
    Mar 14, 2001
    Messages:
    362
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    18
    Location:
    London
    Ratings:
    +4
    Roland

    Thanks for the explaination. If I understand you correctly you are saying the optimal vertical resolution will be different depending of the refresh rate as the spot size is function of the scanning rate.

    Also that you don't want the scanlines to touch at all but just want to make them as close together as you can with them staying distinct.

    On the number of horizontal pixels I'm sure you're right that it is almost impossible to tell the difference but I also want to understand why people choose one number over another.

    I'm looking forward to being able to get my own setup rather than just daydreaming about it, but in the mean time I just want to make sure I will have everything I need ready so that the install will go smoothly and I can go straight into a movie watching marathon once the pj is up :D .

    I promise to watch the films and not look for artefacts the whole time although it will be a hard habit to break. Still three months to go, not sure how I'm going to cope....

    John
     
  6. Roland @ B4

    Roland @ B4
    Active Member

    Joined:
    Apr 11, 2001
    Messages:
    2,238
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    51
    Location:
    Reading.
    Ratings:
    +32
    John

    Spot on

    (pun intended)
    :D
     
  7. JohnAd

    JohnAd
    Active Member

    Joined:
    Mar 14, 2001
    Messages:
    362
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    18
    Location:
    London
    Ratings:
    +4
  8. squid

    squid
    Standard Member

    Joined:
    Feb 26, 2001
    Messages:
    546
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    19
    Location:
    the big uk
    Ratings:
    +0
    the thing i think is best is to first find the vert sweet spot of the projector . then match the horisontal res with your screen aspect ratio.

    this is what i did and it works best for me :)
     
  9. JohnAd

    JohnAd
    Active Member

    Joined:
    Mar 14, 2001
    Messages:
    362
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    18
    Location:
    London
    Ratings:
    +4
    Simon

    You're right that that way is very simple and will get a great picture. It will also help to make any use other than windows easier as the pixels are square. However much recent video software removes the requirement for sqaure pixels and frees us to choose a horizontal resolution indepentantly of the vertical resolution.

    With this software it ought to be possible to improve the picture quality slightly by choosing a "good" width like 1440 whatever the vertical resolution.

    I'm just trying to work out which resolutions to store in memory and have Roland do accurate convergance for.

    Is anybody using a 1440 width?

    John
     
  10. Jenz

    Jenz
    Guest

    Products Owned:
    0
    Products Wanted:
    0
    Ratings:
    +0
    John

    I use 1440x720 & 1280x720 as I experiment between the two.

    I see no difference at this time between the two.

    Regards Jenz.
     
  11. Roland @ B4

    Roland @ B4
    Active Member

    Joined:
    Apr 11, 2001
    Messages:
    2,238
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    51
    Location:
    Reading.
    Ratings:
    +32
    Jenz what if you go lower to say 1024 or 800 x 720.

    When does it start to notice?
     
  12. JohnAd

    JohnAd
    Active Member

    Joined:
    Mar 14, 2001
    Messages:
    362
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    18
    Location:
    London
    Ratings:
    +4
    The more I think about this the harder it is to come up with good example material to see an effect with.

    I suppose the best test pattern is that one on that has thick lines at one end going down to thin lines at the other. Now the correct result of this should not be that all the pixels are either black or white, but that the transition should be smooth and the peaks and throughs should be separated. Also there should be no obvious area where there is a new moire pattern created.

    The static S&W zone plate sort of has this as well. Again I suppose we're looking for nice clean circles and no messy moire fringes. This pattern is good because it stresses the vertical resolution as well.

    John
     
  13. Jenz

    Jenz
    Guest

    Products Owned:
    0
    Products Wanted:
    0
    Ratings:
    +0
    Personally I see a difference (a negative one at 1024).

    Neil.
     
  14. squid

    squid
    Standard Member

    Joined:
    Feb 26, 2001
    Messages:
    546
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    19
    Location:
    the big uk
    Ratings:
    +0
    john i know your right .

    i just like to keep things simple in my cinema . i'm a bit thick you see :D

    plus i only have a 7inch pj on a realtivly small screen so 1024x768 looks great to me . i did try to take the hor res up a bit but did'nt realy notice any benefit so decided to just keep it as simple as pos . now if i had bigger tubes like one of the 8" barcos it could be different

    maybe one day when i win the lottery i will be able to see for myself . as thing's stand i can't even afford any dvds :( . i'm one of those you see looking though the bargain bins
     
  15. JohnAd

    JohnAd
    Active Member

    Joined:
    Mar 14, 2001
    Messages:
    362
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    18
    Location:
    London
    Ratings:
    +4
    Simon

    I'm also getting a 7 inch pj (a BD708s I think) so I'm in the same boat as you. We shouldn't be able to resolve pixels on our puny pjs at 1440 but we still should get clearer video at this resolution.

    Mostly I was just questioning the view that part of the test of finding your sweetspot is the ability to be able to resolve pixels on a windows desktop.

    John
     
  16. squid

    squid
    Standard Member

    Joined:
    Feb 26, 2001
    Messages:
    546
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    19
    Location:
    the big uk
    Ratings:
    +0
    the other big problem i have is that i am still using a hollywood+ mpg card for dvd . it likes to be set at the right res for the aspect to get the best results .

    just upgraded the pc . the next step is to get a better sound card (my spdif out dont work). i have windvd that i got with my geforce so when i get a better soundcrd i may do a bit of experimenting with the hor res a bit:)
     
  17. JohnAd

    JohnAd
    Active Member

    Joined:
    Mar 14, 2001
    Messages:
    362
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    18
    Location:
    London
    Ratings:
    +4
    Simon

    You have a treat in store when you see what DVDs are like from a GeForce or a Radeon :D

    John
     
  18. Rob.Screene

    Rob.Screene
    Active Member

    Joined:
    Sep 29, 2001
    Messages:
    1,124
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    51
    Location:
    Berks, England
    Ratings:
    +45
    Jenz and Roland,
    I don't follow the desire to see scanlines on up-scaled video? If it were not interpolated, then yes, otherwise no. :confused:

    I see the desire as being able to reproduce the film contours as close as possible, not represent quantisation or upscaling noise?

    I think you me the phrase would be: "Don't watch the pixel edge noise, then see the film"

    For me it's balancing detail v.s source chain noise (telecine, quantisation, mpeg enciding, mpeg decoding, scaling, output dac/bandwidth, cable, beam focus).

    regards,
    Rob.
     
  19. JohnAd

    JohnAd
    Active Member

    Joined:
    Mar 14, 2001
    Messages:
    362
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    18
    Location:
    London
    Ratings:
    +4
    Rob

    I think what Roland was saying is that if the pixels overlap you will get an artefact due to the phosphor getting excited twice in very quick succession. This will have the effect of reducing real detail but may increase apparent detail due to the new false sharp edges.

    John
     
  20. Roland @ B4

    Roland @ B4
    Active Member

    Joined:
    Apr 11, 2001
    Messages:
    2,238
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    51
    Location:
    Reading.
    Ratings:
    +32
    The americans are going beserk about 48Khz

    I haven't played yet and I know you will get some flicker on bright images. but it won't stress the projector as much and should give a brighter image.

    Any one played yet?
     
  21. Jenz

    Jenz
    Guest

    Products Owned:
    0
    Products Wanted:
    0
    Ratings:
    +0
    Yes, Gordon & I had a look last night.

    We did not note any additional detail and pans were about the same. In fact the panning at 48khz was slightly worse on Insurrection.

    Flicker was bad.
     
  22. Gordon @ Convergent AV

    Gordon @ Convergent AV
    Distinguished Member AVForums Sponsor

    Joined:
    Jul 13, 2000
    Messages:
    13,982
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    166
    Location:
    Living in Surrey, covering UK!
    Ratings:
    +2,786
    Can I just add a wee bit here.

    We did this last thing at night and it was very very late. It wasn't defintative. 72Hz isn't 72 actually it's a weirdo number of 71.*****. The device we were watching had been very carefully tweaked for judder free replay at his refresh rate.

    We changed to 48Hz, not 48.*** to check more about detail and dynamics than panning. The first time throught the scene it wa smooth until a short section where it fell down and jjjjjuuuuuddddeeeerred. Very briefly though. I believe playing about with refresh rate tweaking would fix this but Neil's the man to ask about that.

    The desktop looked VERY flickery at 48Hz (almost unwatchable), especially after 72......However, once programme material was running it was very watchable and you didn't really see "flicker" as such...it was very strange to see this huge apparent difference....

    Detail didn't seem to be vastly improved but we didn't do a nutter convergence at 48Hz so that's not conclusive. Dynamically colours may have appeared slightly more saturated at 48 but I wouldn't have bet my life on it.

    So from what I saw last night 72 was better but I'd like to spend longer seeing what 48 could do...guess I need to get one of those new fangled HCPC's........:)

    Gordon
     
  23. JohnAd

    JohnAd
    Active Member

    Joined:
    Mar 14, 2001
    Messages:
    362
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    18
    Location:
    London
    Ratings:
    +4
    Gordon

    The rates you want are all multiples of 59.94 the NTSC rate multipled by 2/5 which is 23.984

    This makes the correct rates
    48Hz = 47.968
    72Hz = 71.952
    96Hz = 95.936
    120Hz = 119.92

    You should have these burnt in by now :D

    I'm not sure I'm fully convinced on the 48Hz thing myself, on display and hold devices you can see an improvement to colour, definition etc but this is mostly due to poor cable bandwidth, on a setup with quality cables and BNC's I would expect the difference to be much lower.

    Also when there are the enevitable corrections when the timing resync you will notice a much smaller effect at 72 than at 48

    At 48 you will get either 2 2 2 1 2 2 2 or 2 2 2 3 2 2 2
    At 72 you will get either 3 3 3 2 3 3 3 or 3 3 3 4 3 3 3

    the effect of the adjustment should be smaller.

    John
     
Thread Status:
Not open for further replies.

Share This Page

Loading...