|
As this forum is rarely used anymore, we've locked it. Feel free to browse and read. Questions? Please reach out to us directly. Cheers! |
|
 |
|
|
| Author |
Message |
stridsvognen
Guest
|
| Posted: Mon Sep 01, 2014 3:08 pm Post subject: |
|
|
| redfox001 wrote: | Ok will look into it. I tried lowering contrast and on 720p one on one became a lttle better possable on low contrast not much to notice 804p nothing to notice but did not look in tube. I also see perfect bladk on the screen when I see the end title of a movie. What i was wondering on and what I do not understand is that I see a mirror reflection of the white words. If the word is far enough from the center. I mean whe the line is at the bottom I see a weak white mirror reflection on the top. I checked if by any chance the coating of the lens is damaged but the coating is near perfect except very little scratches. The anti reflecting coating is there but still the mirroring. I never had liquid coupled bedore so perhaps this is normal? It is not a halo around an object that makes contrast worse. Well it almost keeps me out of my sleep. As I said perhaps it is normal for the curved lens to reflect to the oposite of the screen a little with liquid coupled lenses? The reflection is also perfect white so if it was bad coating than all three are exactly the same bad so I don't think that.
I do not recognise what I see on the screenshot but I will look on a normal screen tomorrow. But it looks sharp whatever it is! |
Its a clip from the HD Basic Blu Ray. And its not sharp, its taken with free hand, the lenses have terrible focus, and most of the image is out of focus as its taken off angle on the screen.
It has very nice detail level, and thats a result of the bandwidth performance and low noise level, who gives a much better pixel to pixel and overall contrast and dynamic performance, and that you might experience as addet sharpness on a distance, but it has nothing to do with optical or magnetical sharpness.
|
|
| Back to top |
|
 |
redfox001
Joined: 16 Mar 2009
Posts: 2257
Location: The Netherlands
|
| Posted: Mon Sep 01, 2014 4:46 pm Post subject: |
|
|
| stridsvognen wrote: | | gjaky wrote: | | stridsvognen wrote: | | I also understood you pointed out they had 7 peakings onboard, that indicates there must be a bandwidth problem, as the old modified boards have 0 peaking addet |
To judge this, you'd have to know what really peaking is.
To illustrate this I circled all the "peaking" components on the old VNB. I am really curious if all those peakings are removed from your "zero peaking" VNB.
If you really that concerned about peaking methods you really should read this: National AN1013, page 16, section 7 - Peaking methods. It wasn't written by Mike Parker himself, yet it has probably based on a true story... Of course all the industry leaders did these methods, and this many company shouldn't get wrong right. |
I cant say how much was removed on the old boards, but there is a lot of components missing.
Im also aware that peeking is needed in some amount in a analog chain like this, but ill say less peaking and more real bandwidth clearly looks better.
And my experience say its ok to peak a bit as long it never do a overshoot on the 1:1 horizontal line at 100% IRE output.
If the bandwidth is so poor that it needs heavy peaking to get the vertical line right, and it then overshoot on the horizontal line, ill prefer some roleoff on the vertical.
I also seen the spec sheet on the AD834 used on the 03VIM, and its describes built in peaking with 0 overshoot.
The old standard Marquee VNB have peaking, and its visible that it do overshoot, and its important to remove that when a high bandwidth VIM is addet.
I also think that the low noise level on the MP mods free up a lot of bandwidth, so its hard to say what do what.
I asked others to shoot some pics of the 1:1 pattern starting with contrast at 0 and slowly turn it up, and display how the pattern behave.
If you can display a perfect horizontal line with no overshoot, and get the vertical perfect out of black ( Use the horizontal as reference) You have sufficiant bandwidth, and perfect matched peaking for the selected resolution.
And what makes the Marquee the best CRT projector ever is its high standard bandwidth, and its potential to be optimized for HD material. |
Ok I have read somewhere that we should look at the two pixel lines. If there is to much peaking than the two pixel lines will become to bright. As long as they are neutral more peaking is allowed.
|
|
| Back to top |
|
|
stridsvognen
Guest
|
| Posted: Mon Sep 01, 2014 9:01 pm Post subject: |
|
|
| redfox001 wrote: | | stridsvognen wrote: | | gjaky wrote: | | stridsvognen wrote: | | I also understood you pointed out they had 7 peakings onboard, that indicates there must be a bandwidth problem, as the old modified boards have 0 peaking addet |
To judge this, you'd have to know what really peaking is.
To illustrate this I circled all the "peaking" components on the old VNB. I am really curious if all those peakings are removed from your "zero peaking" VNB.
If you really that concerned about peaking methods you really should read this: National AN1013, page 16, section 7 - Peaking methods. It wasn't written by Mike Parker himself, yet it has probably based on a true story... Of course all the industry leaders did these methods, and this many company shouldn't get wrong right. |
I cant say how much was removed on the old boards, but there is a lot of components missing.
Im also aware that peeking is needed in some amount in a analog chain like this, but ill say less peaking and more real bandwidth clearly looks better.
And my experience say its ok to peak a bit as long it never do a overshoot on the 1:1 horizontal line at 100% IRE output.
If the bandwidth is so poor that it needs heavy peaking to get the vertical line right, and it then overshoot on the horizontal line, ill prefer some roleoff on the vertical.
I also seen the spec sheet on the AD834 used on the 03VIM, and its describes built in peaking with 0 overshoot.
The old standard Marquee VNB have peaking, and its visible that it do overshoot, and its important to remove that when a high bandwidth VIM is addet.
I also think that the low noise level on the MP mods free up a lot of bandwidth, so its hard to say what do what.
I asked others to shoot some pics of the 1:1 pattern starting with contrast at 0 and slowly turn it up, and display how the pattern behave.
If you can display a perfect horizontal line with no overshoot, and get the vertical perfect out of black ( Use the horizontal as reference) You have sufficiant bandwidth, and perfect matched peaking for the selected resolution.
And what makes the Marquee the best CRT projector ever is its high standard bandwidth, and its potential to be optimized for HD material. |
Ok I have read somewhere that we should look at the two pixel lines. If there is to much peaking than the two pixel lines will become to bright. As long as they are neutral more peaking is allowed. |
You can look wherever you like. And adjust peakin or no peaking to your liking, i aim for perfection, i might not be possible, but ill never accept visible peaking on the pattern as i have seen what it do to the image.
Im also 100% into this for picture quality, to be as close to the source material as possible, no matter if its CRT or Digital, ill go with whatever kind of projector who gives me the best performance and movie experience.
Many has blamed me to look to much at test patterns, and i do, but i might also very well be the one who look the most movies and run the most hours on my projectors.
Testpatterns give me a reference point i cant get in a movie scene, with movie material i get a feeling, with test patterns i get hardcore facts, i like to combine them both as no specifik testpatterns will display all the qualities, or potential problems there is in displaying a image.
Then i try emagine the potential problem if i dont have the bandwidth and i miss the level on my high resolution part, that may be roleoff or overshoot, dont matter. They are both distortions and place the saturation or the color of the specifik pixel at a point where it was never ment to be, and so on.
Anyone care to tell how the 1:1 vertical comes out of black, if you start with a contrast at 0 and slowly turn it up looking into the lens.?
|
|
| Back to top |
|
|
redfox001
Joined: 16 Mar 2009
Posts: 2257
Location: The Netherlands
|
| Posted: Mon Sep 01, 2014 9:38 pm Post subject: |
|
|
This is the source of the information I wrote:
Of course you can increase peaking to improve the higher end of frequencies, but that lets you have excess gain at upper medium frequencies which can make your 2 pixel line field appear brighter than the 3 or 4 pixel wide lines.
http://www.avsforum.com/forum/16-crt-projectors/645727-sanyo-vpa18-180mhz-video-amp.html
Seems to me these are designers and know what they are talking about. Peaking is not bad it is used everywhere in the chaine just don't use it to much.
|
|
| Back to top |
|
|
redfox001
Joined: 16 Mar 2009
Posts: 2257
Location: The Netherlands
|
| Posted: Mon Sep 01, 2014 9:46 pm Post subject: |
|
|
This one might be interesting to some of us too:
What's the minimum bandwidth required for 1080p? I'm guessing the refresh would be 60Hz.
Some info I found is:
The formula to calculate bandwidth is
BW = [(TP x Vt) /2] x 3
Where BW = Bandwidth in Megahertz (MHz)
TP = total pixels (horizontal x vertical)
Vt = picture refresh rate in Hertz (Hz)
The bandwidth, @ 60Hz refresh rate, required for the HDTV signals is as follows.
720p, 1280x720 (83 MHz)
1080i 1920x540 (93 MHz) only half of the 1080 horizontal lines are displayed at once.
1080p, 1920x1080 (186.6 MHz)
I'd assume that ALL components in your video chain (PJ, cables, switcher/scaler, source) would need to have this minimum bandwidth.
What happens, PQ wise, if one of the components has a lower bandwidth than the minimum?
http://www.avsforum.com/forum/16-crt-projectors/653094-bandwidth-1080p.html
As I have a feeling the 909 is using a vpa18 or one with the same specs it should be able to resolve 180p we just have to get the bandwdth to that chip as MP said.
There is a /2 in the formula and that is because you need a black and a white pixel for one wavelength and there is a 3 because you need at least 3 harmonics to the ground wave to get a square wave.
|
|
| Back to top |
|
|
stridsvognen
Guest
|
| Posted: Tue Sep 02, 2014 5:11 am Post subject: |
|
|
|
How do your 1:1 vertical come out of Black? try turn your contrast Down to 0 and turn it up slowly, while you look into the lens.
|
|
| Back to top |
|
|
gjaky
Joined: 05 Jun 2010
Posts: 2802
Location: Budapest, Hungary
|
| Posted: Tue Sep 02, 2014 5:37 am Post subject: |
|
|
| redfox001 wrote: |
The bandwidth, @ 60Hz refresh rate, required for the HDTV signals is as follows.
720p, 1280x720 (83 MHz)
1080i 1920x540 (93 MHz) only half of the 1080 horizontal lines are displayed at once.
1080p, 1920x1080 (186.6 MHz)
I'd assume that ALL components in your video chain (PJ, cables, switcher/scaler, source) would need to have this minimum bandwidth. |
Actually no, the thing is even worse. According to rise time of cascaded blocks if you have a projector and a source which both just have the same 186MHz BW they act up as two filters in series, meaning the output will have lower BW than either, in this case both with 186MHz they will let an overall 131 MHz (!!!) chain, and this is why it matters if you upgrade the opamps in your video chain from 500 to 700MHz, especially if there are a lot of them. For proper 186MHz resolution with two equally limited blocks you'd need 263MHz BW from each blocks.
_________________
projectors in the past : NEC 6-9PG xtra, Electrohome Marquee 6-7500, NEC XG 1351 LC ( with super modified Electrohome VNB neckboard !!!)
current: VDC Marquee 9500LC
The MOD: VNB-DB, VIM-DB
|
|
| Back to top |
|
|
redfox001
Joined: 16 Mar 2009
Posts: 2257
Location: The Netherlands
|
| Posted: Tue Sep 02, 2014 7:24 am Post subject: |
|
|
Thanks for the info!!!
Think this does not calculate with peaking.
When I calculate the effect of 4 times 800MHz on 1 180 MHz I get 164MHz.
Now it becomes easier to design the chain. Also I found my scope goes to 150MHz and that might be enough if the goal is to resolve a 140MHz minimum for 800p.
I have read somewhere that there must be 2,4GHz opamps out there but I don't know if that is bragging  The el5166 can do 800MHz at Av=2 but I see it has to do Av of 5.7 in the driver but there is the peaking circuit right behind it that can not be a coincidence
|
|
| Back to top |
|
|
redfox001
Joined: 16 Mar 2009
Posts: 2257
Location: The Netherlands
|
| Posted: Tue Sep 02, 2014 8:39 am Post subject: |
|
|
There is also some discussion over this multiplying with 3. To resolve a real world wave with samples you need a factor 2 (that is nyquist shannon criteria) If you go to resolving it with a bandwdith 3 times the sample rate than you are not filtering the wrong alias information and I don't know what that would look like to the eye.
Ok I had to rewrite this. The sampling theorem says you can reproduce the original signal with sharp spikes at the samples two. But than you also have the alias info above two times the sampling frequency. That signal has to be filtered for high frequencies afterward. The interpolation that is natural with CRT is a type of upsampling that is not done with digital pixels that only do a sample and hold hoping that your eyes or brain will do the filtering.
However you want to reproduce the height of the original samples as good as possible and for that you need an electronical bandwidth as high as possible including the peaking.
This however is relevant to the 4k discussion. With digital the 4k might give a more natural look. With CRt I don't think it is necessary to upsample to 4k. Original 4k material is another thing and I have to see it if it relay adds something on CRT.
Ok these are my thoughts for this moment. Open to discussion of cause.
|
|
| Back to top |
|
|
redfox001
Joined: 16 Mar 2009
Posts: 2257
Location: The Netherlands
|
| Posted: Tue Sep 02, 2014 9:57 am Post subject: |
|
|
|
Aaaaarch I am still thinking in pixels! With CRT having a to high electronical bandwidth will polute the signal with alias signals. So I guess it is saver to stay between the 2 and 3 times limit and that means slightly higher than the pixel clock.
|
|
| Back to top |
|
|
redfox001
Joined: 16 Mar 2009
Posts: 2257
Location: The Netherlands
|
| Posted: Tue Sep 02, 2014 10:23 am Post subject: |
|
|
When you read this post you'll see that 100MHz will be enough to completely resolve 1080p. It will not show in the one on one testpattern but it will resolve the real world 1080p sampled signal.
He makes a calculation for 1080i and comes to 50HMz required. But I do agree I don't get to the 100MHz with the Barco.
| Quote: |
A number of very relevant comments from several folks here in this thread. Mike (and cmw9) are right on about looking for bandwidth differences in fine background details, which when preserved can bring added depth and dimensionality to an image (if they're there to start with). And Bob's observations about test patterns providing an obvious torture test to expose bandwidth differences are also correct. Lastly, Chiem's extensive and well-explained analysis provides a lot of the technical underpinings to understanding the effects of bandwidth differences.
However, all this leaves out a couple of very important questions: what are you watching, and where are you getting it from?
I.e., if you're looking at test patterns, or computer-generated spreadsheets or text documents, then Chiem's analysis of preserving the higher-order components of the square-wave are appropriate. In this context, you want to maintain as close to a 100% MTF as possible, for sharp, crisp edges. When the bandwidth is exceeded, the MTF drops off, and the full amplitude and detail can no longer be maintained. If you want to put up 1600x1200 desktops @72 Hz, you'll need all the bandwidth you can get in your PJ, and differences will be readily apparent.
But assuming that what we're really interested in watching is HDTV from various HD sources (60 Hz), the context changes significantly. Even 1920x1080i content is sampled at 74.5 MHz, and thus has to be filtered and steeply rolled off below 37.25 MHz to satisfy Mr. Nyquist. There are no square waves at 37 MHz to worry about, no 3rd harmonics at 111 MHz to maintain, and even sinusoids at the limiting frequency have their amplitude severely attenuated. I described this last year, when I wrote in a thread about fixing a BW problem on the JVC 30K D-VHS deck:
Quote:
You missed the point of Cary's comments about Nyquist. The sampling rate (clock) of 74 MHz is already twice the highest frequency required (37 MHz). A 37 MHz signal already contains all the information to reconstruct the full 1920 pixel scan line.
Unfortunately, due to the need for filtering (for anti-aliasing and anti-imaging purposes), we need to roll off the response before we hit that 37 MHz mark, and thus can never quite achieve that full 1920 horizontal resolution, in the analog realm (without some attenuation). But we can get close.
SMPTE specifies that the output response (NOT bandwidth, which is the -3dB point) should be maintained "virtually flat" out to 30 MHz. I.e., +/-0.05 dB. The rolloff should start AFTER that point, and be down at least 12 dB at 37 MHz, and continue to roll off at a steep rate (IIRC 6th order), so that by the next octave, when you hit the sampling frequency, you're down at least 40 dB.
What this means is that image integrity (full detail, with unimpaired MTF) is maintained out to 1560 horizontal (i.e., 30 MHz), and rolls off smoothly beyond that, with the bandwidth (-3dB point) being 34 MHz (equivalent to 1770 horizontal). There can still be content all the way out to 1920, but 1770 is about the effective limit of reproduction (if everything in the source chain is perfect).
As long as the bandwidth of your display device can meet these lesser criteria (and as Greg Rogers has pointed out, a display with a -3 dB bandwidth rating of ~50 MHz can usually maintain flatness out to the 37 MHz required), it will do a good job of reproducing the full HiDef signal. Which is why Person99, secstate, RVonse and others aren't seeing significant differences between PJs that exceed that metric, on HD source material. Now, if you want to double that, to 1080p, that imposes a significant extra burden on the PJ, and differences will become apparent pretty quickly. But that's another kettle of fish.
Also important to keep in mind is that every signal-passing component you insert into the signal chain will cumulatively degrade the image quality. You want to maintain that bandwidth flatness, and if your components are only rated out to a BW of 37 MHz, that means that each will contribute a 3 dB loss at their rated BW. Stack a couple of these, and your PQ is shot. That's the reason why Extron recommends the very high BW figures that they do, and why folks like Mike push the BW of their products way out... so they won't have any negative impact at the lower frequencies of interest, even when cascaded with other components. That doesn't imply though that the PJ itself needs 200+ MHz BW to properly display HD, or other inflated claims that frequently percolate through this Forum.
- Tim
|
http://www.avsforum.com/forum/16-crt-projectors/594155-barco-bandwidth-differences-real-world-opinions.html#post6404463
|
|
| Back to top |
|
|
redfox001
Joined: 16 Mar 2009
Posts: 2257
Location: The Netherlands
|
| Posted: Tue Sep 02, 2014 10:43 am Post subject: |
|
|
|
So my conclusion is that if you can resolve a 100MHz pixelclock lets say a 800p picture with some porches with the one on one testpattern than you are fully resolving 1080p even if the 1080p one on one is not resolved.
|
|
| Back to top |
|
|
stridsvognen
Guest
|
| Posted: Tue Sep 02, 2014 12:11 pm Post subject: |
|
|
| redfox001 wrote: | | So my conclusion is that if you can resolve a 100MHz pixelclock lets say a 800p picture with some porches with the one on one testpattern than you are fully resolving 1080p even if the 1080p one on one is not resolved. |
May i recommend that you try play with a CRT with a pixel clock that you can fully resolve, and i mean on the tube face/ screen without any visible distortions, and do so from black to white.
Try think of the bandwidth as fall and rise time, you will notice that 100Mhz is actually not enought to resolve anything but maybe a SD 480p/ 576p 60hz signal to look decent.
Can your Barco make a perfect 480P 60 hz 1:1.?
|
|
| Back to top |
|
|
jbmeyer13
Joined: 03 Dec 2010
Posts: 1135
|
| Posted: Tue Sep 02, 2014 3:10 pm Post subject: |
|
|
| redfox001 wrote: | | So my conclusion is that if you can resolve a 100MHz pixelclock lets say a 800p picture with some porches with the one on one testpattern than you are fully resolving 1080p even if the 1080p one on one is not resolved. |
What your talking about is the subjective difference in terms of real world viewing and Kurt is talking about objective test results. 100mhz is not enough to FULLY resolve 1920 x 1080p when measuring objectively (using test patterns).
Kurt is one of the few who notices very subtle attributes to the image and he's spent countless hours testing not only the Marquee but also the G90 with multiple modified boards. He sees differences that most of us probably don't and in a way that's a curse 
If you are happy with the image your 909 throws then all of this is largely irrelevant from the subjective stand point but if we are talking about objective test results then it's a different story. Also, if you sit 3x the distance then you will not see many of the finer attributes regardless of which PJ you are using.
_________________
Projector: Modded 9501LC ULtra- MP VIM, Vold VNB, ETECH LVPS, Silver VIM Cables, HD10F's & a V1 case!
|
|
| Back to top |
|
|
gjaky
Joined: 05 Jun 2010
Posts: 2802
Location: Budapest, Hungary
|
| Posted: Tue Sep 02, 2014 3:56 pm Post subject: |
|
|
| redfox001 wrote: |
Now it becomes easier to design the chain. Also I found my scope goes to 150MHz and that might be enough if the goal is to resolve a 140MHz minimum for 800p. |
No.
Remember, when you measure with your oscilloscope it will become a part of your signal chain. Also, the oscilloscope's measuring probes have an own bandwidth too. A 250MHz passive probe (which is very good actually) plus your 150MHz scope will make up 128MHz together. And then there comes the problem of the ground lead, since it acts as an additional inductance which also worsen the bandwidth, so you'll find yourself with a 100MHz capable measuring rig which is too slow to test the bandwidth in detail. I know this from personal experience. Measuring neckboard performance is a thing which I also have interest, and also struggling to do it right. I have a 275MHz oscilloscope, have 250MHz passive probe, but now I also have a 500MHz active FET probe as well.
| redfox001 wrote: |
When you read this post you'll see that 100MHz will be enough to completely resolve 1080p. It will not show in the one on one testpattern but it will resolve the real world 1080p sampled signal. |
I don't quite believe in that. The things explained in that post is probably true if an analog electronic signal is digitized (like a CVBS signal), but think about how are 35mm films are digitized, they are basically scanned with a DIGITAL equipment, like a CCD or CMOS sensor, there is a pixel array that is able to detect light, and ideally it can output a 1:1 pattern without any limitation, this is a sort of electro-optical D/A conversion.
Therefore the bandwidth is only loosely dependant from pixel clock. With 120MHz pixel clock the "fastest" signal is a 60MHz square wave signal. An ideal square wave signal would have infinite (!!) bandwidth, the spectrum of square wave only consist of odd harmonics, for 60MHz these are: 60, 180, 300, ... MHz, while the amplitude of the harmonics are descending but they are doing it only with the ratio of 1/r (r is 1,3,5,....).
So for a proper image reproduction you want as high bandwidth as possible, without introducing too much noise.
| stridsvognen wrote: |
Try think of the bandwidth as fall and rise time, you will notice that 100Mhz is actually not enought to resolve anything but maybe a SD 480p/ 576p 60hz signal to look decent. |
With 100MHz bandwidth you'll able to faithfully reproduce signals a bit more than that. 
_________________
projectors in the past : NEC 6-9PG xtra, Electrohome Marquee 6-7500, NEC XG 1351 LC ( with super modified Electrohome VNB neckboard !!!)
current: VDC Marquee 9500LC
The MOD: VNB-DB, VIM-DB
Last edited by gjaky on Tue Sep 02, 2014 4:10 pm; edited 1 time in total
|
|
| Back to top |
|
|
stridsvognen
Guest
|
| Posted: Tue Sep 02, 2014 4:07 pm Post subject: |
|
|
| jbmeyer13 wrote: | | redfox001 wrote: | | So my conclusion is that if you can resolve a 100MHz pixelclock lets say a 800p picture with some porches with the one on one testpattern than you are fully resolving 1080p even if the 1080p one on one is not resolved. |
What your talking about is the subjective difference in terms of real world viewing and Kurt is talking about objective test results. 100mhz is not enough to FULLY resolve 1920 x 1080p when measuring objectively (using test patterns).
Kurt is one of the few who notices very subtle attributes to the image and he's spent countless hours testing not only the Marquee but also the G90 with multiple modified boards. He sees differences that most of us probably don't and in a way that's a curse
If you are happy with the image your 909 throws then all of this is largely irrelevant from the subjective stand point but if we are talking about objective test results then it's a different story. Also, if you sit 3x the distance then you will not see many of the finer attributes regardless of which PJ you are using. |
Im cursed.. 
Actually i dont think im the only one who will se a difference, i might be one of the few who have the gear, and a shitload of different boards to test with.
I have done a-b test on friends who dont know/ care about whats going on, and they all confirmed my findings without me helping. And always on movie material.
So im sure anyone of you will see what i see if you look at the same screen as i do..
|
|
| Back to top |
|
|
thewolfman
Joined: 28 Mar 2011
Posts: 1311
Location: Sweden
|
| Posted: Tue Sep 02, 2014 4:48 pm Post subject: |
|
|
[quote="stridsvognen"] | redfox001 wrote: | | One question about the focus of the Marquee. The focus coil of the Marquee can be replaced with a 909 focus coil I understand. Is that easy? I remember that there was a Franken yoke or something? With my 8500 I had very early blooming when I put contrast higher. |
| Quote: | | ... Looking into my LCP tubes from a distance, i see scanlines on the tube faces, and thats with standard yokes. I cant say how it will look with 909 yokes, that you need to ask Mike about, think he uses 909 yokes. ... |
I don't think having visible scanlines on the tube face means anything. I believe I always had them, in one combo or another, and never given them some thought until you mention it every now and then. I don't know for sure, but I think a stock Marquee, or any other 9 inch pj for that matter, will show scanlines on the tube face with 1080p. Don't they? I only have 817 showing, but it's the same thing only top and bottom is cut out of the mix.
|
|
| Back to top |
|
|
jbmeyer13
Joined: 03 Dec 2010
Posts: 1135
|
| Posted: Tue Sep 02, 2014 4:52 pm Post subject: |
|
|
| stridsvognen wrote: |
Im cursed..
Actually i dont think im the only one who will se a difference, i might be one of the few who have the gear, and a shitload of different boards to test with.
I have done a-b test on friends who dont know/ care about whats going on, and they all confirmed my findings without me helping. And always on movie material.
So im sure anyone of you will see what i see if you look at the same screen as i do.. |
I'm cursed as well so don't feel bad 
Perhaps if you tell them what to look for but I can guarantee a lot of folks miss things that you or I may see. If they are sitting 3x the distance it will likely be harder to spot certain things as well.
_________________
Projector: Modded 9501LC ULtra- MP VIM, Vold VNB, ETECH LVPS, Silver VIM Cables, HD10F's & a V1 case!
|
|
| Back to top |
|
|
stridsvognen
Guest
|
| Posted: Tue Sep 02, 2014 5:04 pm Post subject: |
|
|
| jbmeyer13 wrote: | | stridsvognen wrote: |
Im cursed..
Actually i dont think im the only one who will se a difference, i might be one of the few who have the gear, and a shitload of different boards to test with.
I have done a-b test on friends who dont know/ care about whats going on, and they all confirmed my findings without me helping. And always on movie material.
So im sure anyone of you will see what i see if you look at the same screen as i do.. |
I'm cursed as well so don't feel bad
Perhaps if you tell them what to look for but I can guarantee a lot of folks miss things that you or I may see. If they are sitting 3x the distance it will likely be harder to spot certain things as well. |
Sure 3x the distance is a bit far, some things you wont see, but others you still see clear, like the improved contrast, better color definition, and much much more dynamic picture.
With CRT i like to sit around 1,5-2x the screen width away, but i have no problems at 1x, other than i think its so close that i need to turn my head when action is crosing the screen.
There is also the audio part, distance to speakers to sit the place where i get max effect from the speakers in the hole frequence range. ( My new speakers is to small to shake the house )
Im also cursed when it comes to audio, maybe even more than with video. 
|
|
| Back to top |
|
|
redfox001
Joined: 16 Mar 2009
Posts: 2257
Location: The Netherlands
|
| Posted: Wed Sep 03, 2014 1:23 am Post subject: |
|
|
Guys think of it this way. When you record with 1920 pixels horizontal this means you make 1920 samples. The bandwidth theorem from Shannon or Nyquist states that you can at most record a sinus wave of 960 Hz this way. You can not record a square wave of 960 Hz because that would need at least three harmonics or three times the 1920 sample rate.
So if you have a black and white pattern of 320 Hz or 320 pixels in width 320 white black parts that should resolve 100 %.
The one on one square wave can not be recorded with a camera. The camera can not distiguish between the sinus and the square wave.
Think of a cd recording. It is done at 44 Khz. The highest sinus in it is 20 Khz. With 20 Khz you can not make a square 20 KHz wave that would need 3 times 44 KHz recording speed at least.
The best square testsound you coul record would be a 7 kHz square wave. It is all physics. Everything above 20 KHz is filtered in the cd player so everything above 100 Mhz is filtered in the video dac for 1080 p.
This is the difference between real world and testpictures. tim is very clear on it when he calls it rediculous high numbers. He says that is why a lot of people don't note a big difference between very high bandwidth crt and lower bandwidth. The only thing the higher bandwidth gives is a flatter response in the area that matters. That might bring out some detail. This does explain why the picture I see looks very good on my Barco and the difference with 720p is very very small.
I will however not beat you guys to dead with this story. Just remember to look at real movies if you see something and probable the flatter response is noticable when you are very sensitive to it. I might notice it myself. However when I am going to modify the video chain I will not strife for high bandwidth at the cost of peaking destroying the flatness or bandwidth introducing video noise. We have to be carfull with that. it is just more complicated to do a MP mod.
Edited some confusing text away.
Last edited by redfox001 on Wed Sep 03, 2014 5:43 am; edited 3 times in total
|
|
| Back to top |
|
|
|
|
|
|
|
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
You cannot attach files in this forum
You can download files in this forum
|
Forum powered by phpBB © phpBB Group
|
|
