| Author |
Message |
redfox001
Joined: 16 Mar 2009
Posts: 2257
Location: The Netherlands
|
| Posted: Sun Aug 24, 2014 2:38 pm Post subject: |
|
|
| stridsvognen wrote: | Nice pattern.
Its clear that you will get a much sharper, dynamic and better contrast running the 720P 64Mhz resolution. As it seems to fit the bandwidth of your video chain.
Can you share that pattern with the rest of us.? |
Yes I will try a few tv series that are 720p and I used to see them on 1080p. I will say if they look better and sharper on 720p.
The pattern is from a Nec generator that I received from Francisco some time ago. Ik is a windows program called 'NEC Pattern Gen'.
http://www.softpedia.com/get/Tweak/Video-Tweak/NEC-Test-Pattern-Generator.shtml
|
|
| Back to top |
|
 |
stridsvognen
Guest
|
| Posted: Sun Aug 24, 2014 3:02 pm Post subject: |
|
|
| redfox001 wrote: | | stridsvognen wrote: | Nice pattern.
Its clear that you will get a much sharper, dynamic and better contrast running the 720P 64Mhz resolution. As it seems to fit the bandwidth of your video chain.
Can you share that pattern with the rest of us.? |
Yes I will try a few tv series that are 720p and I used to see them on 1080p. I will say if they look better and sharper on 720p.
The pattern is from a Nec generator that I received from Francisco some time ago. Ik is a windows program called 'NEC Pattern Gen'.
http://www.softpedia.com/get/Tweak/Video-Tweak/NEC-Test-Pattern-Generator.shtml |
Thanks ill try play with that.
If i understand right the 909 have a menu where you can adjust peaking in steps, you might need to ad a bit of peaking to make a better 1080P performance, just be carefull not to add to much.
As long the horizontal line dont start bloom at the start of the line, its usually better than no peaking.
I just wonder where your limitation is, as its not douing 64Mhz perfect. Maybe others who know the 909 better can say whats normal bandwidth to a 909, with a confirmed hight bandwidth source/ DA converter.
|
|
| Back to top |
|
|
redfox001
Joined: 16 Mar 2009
Posts: 2257
Location: The Netherlands
|
| Posted: Sun Aug 24, 2014 3:13 pm Post subject: |
|
|
My own thoughts on this are that Mike Parker wrote somewhere that the VGA adapters or cards have limiting filters on board. He said that only Moome produced a hdmi to RGB that did it right but as I understand only the internal Sony and Marquee do this at the moment. But Francisco is working on the Barco solution. That will be the first barrier to even 720p I suspect.
I have a modified input board that was modified by Mike Parker so I don't expect that to be the problem but the board was never really tested. The Mike Parker modified RGB driver that I have works only in the 808 series. But I looked at it closely and there is only one new video amplifier on it and I will try placing that video amplifier on my 909 RGB driver. The old video amplifier is at -3db at 800MHz so does not look so bad to me but anyway Mike will have found a better one. But first the input should be right before anything else.
|
|
| Back to top |
|
|
stridsvognen
Guest
|
| Posted: Sun Aug 24, 2014 3:32 pm Post subject: |
|
|
| redfox001 wrote: | My own thoughts on this are that Mike Parker wrote somewhere that the VGA adapters or cards have limiting filters on board. He said that only Moome produced a hdmi to RGB that did it right but as I understand only the internal Sony and Marquee do this at the moment. But Francisco is working on the Barco solution. That will be the first barrier to even 720p I suspect.
I have a modified input board that was modified by Mike Parker so I don't expect that to be the problem but the board was never really tested. The Mike Parker modified RGB driver that I have works only in the 808 series. But I looked at it closely and there is only one new video amplifier on it and I will try placing that video amplifier on my 909 RGB driver. The old video amplifier is at -3db at 800MHz so does not look so bad to me but anyway Mike will have found a better one. But first the input should be right before anything else. |
The filters on the analog output of your DA converter can du whats seen on your shots.
As i understand the 909 will never be able to get the bandwidth to resolve a 1080P 60hz image, but for sure you still dont want the limitation at the 1st stage of your analog video chain.
Actually the only DA converter i know off douing a perfect 1080P 60hz bandwidth is the MP modified Marquee Moome card, a standard moome card will still role off a bit.
But its not always good to pass to high bandwidth to the projector if it cant handle it, so the standard Moome card seems to be a good compromise.
I left Moome some feedback about filtering of the SONY V3 card, Mike did mod one for me, and im just waiting for parts to do a adapter to test it in my Marquee.
So maybe the standard SONY V3 card will in the future pass a perfect 200Mhz.
I also have a Hdfury4 for testing here, and a 2nd one is on its way, ill try if the analog output on that one can be saved, as is its not to good, looks like it also suffers from some bad filtering+ maybe some other problems.
|
|
| Back to top |
|
|
redfox001
Joined: 16 Mar 2009
Posts: 2257
Location: The Netherlands
|
| Posted: Sun Aug 24, 2014 5:06 pm Post subject: |
|
|
Well you are right that if someone with a testpattern generator has tried and not succeeded with the 909 than I would be interested where the limit in bandwith is. Thinking about the whole video chain I think one can win little bits all the way down even at the coax cables used. Might still be that the neckboard is just no good and that we will never get 200MHz but 114MHz is the goal. If we are there maybe two neckboard chips stacked might do something 
Hmm that means I have to double the current bandwith at least....
Oh forgot to mention. I did experiment with peaking but there was no change in the testpicture. However the contrast (light output whole picture) was a little higher with high bandwith peaking.
@Francisco could you check at what bandwith your chain still resolves reasonable? That way we would have some hard numbers to improve upon.
Last edited by redfox001 on Sun Aug 24, 2014 6:04 pm; edited 6 times in total
|
|
| Back to top |
|
|
redfox001
Joined: 16 Mar 2009
Posts: 2257
Location: The Netherlands
|
| Posted: Sun Aug 24, 2014 5:28 pm Post subject: |
|
|
|
What kind of pattern generators do the professional calibrators use?
|
|
| Back to top |
|
|
gjaky
Joined: 05 Jun 2010
Posts: 2802
Location: Budapest, Hungary
|
| Posted: Sun Aug 24, 2014 6:33 pm Post subject: |
|
|
| redfox001 wrote: | | My own thoughts on this are that Mike Parker wrote somewhere that the VGA adapters or cards have limiting filters on board. He said that only Moome produced a hdmi to RGB that did it right but as I understand only the internal Sony and Marquee do this at the moment. But Francisco is working on the Barco solution. That will be the first barrier to even 720p I suspect. |
I've measured a few VGA cards here with my oscilloscope ->(non professional) VGA card bandwidth evaluation The results varied, yet I didn't notice any realworld limitation, for example this picture was made with the intel GMA950 and NEC 6PG xtra (80MHz advertised bandwidth), 720P 120Hz, around 165MHz bandwidth...
Generally I advise for bandwidth tests to use lower resolution but at higher refresh that makes focusing less important in evaluation.
_________________
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: Sun Aug 24, 2014 10:39 pm Post subject: |
|
|
This is what Mike said:
| Quote: |
The Moome should do better than any video out from PC or Mac. The reason being the filters on the PC card that I mentioned earlier a few post back, so I can understand the Moome being sharper. The V3 version of the Moome could have my modification and that would explain why you see the difference.
1920X1080P requires a "specific" filter in its reconstruction network in order to properly pass both bandwidth and speed for 1920X1080P one pixel test. A PC card that would be able to get this right every time would need to have a tunable filter that would change for every change of the resolution. Especially when doing the higher resolutions.
|
It is a little over my head right now 
On the positive side I compared 720p material on a 720p resolution and a 1080p resolution and I liked the 1080p a little better even though it did not fully resolve the bandwidth. Might be the upscaling but everything looked more natural and better polished and I did not mis hard edges But in general the difference was hardly noticable but nice to have. In the 720p pictures I did notice a little more definition.
To be fair I would have to compare 720p@60 with 720p@120 but the Barco starts to make a lot of noise at 120.
|
|
| Back to top |
|
|
redfox001
Joined: 16 Mar 2009
Posts: 2257
Location: The Netherlands
|
| Posted: Sun Aug 24, 2014 11:18 pm Post subject: |
|
|
I am thinking about this calculation. Correct me if I am wrong.
With 720p@60 we get an bandwith of 60 MHz. That is the product of horizontal pixels with vertical pixels and frames per second.
But if all these pixels where changing from white to black and back the wave that carried this signal would have to be a square wave at 60 MHz. If I add a sinus of 60 MHz with a first harmonic at 120 MHz and second at 180 MHz I would get a reasonable square wave not? So if Barco says that the video chaine is 180 MHz they might be right and it resolves 60 MHz fine in a square wave.
But if I want 110MHz I would need a bandwith of 330 MHz for the harmonics not? This would explain why 60 MHz is the limit here and 110 MHz looks like 150 MHz with no harmonics resolved ???
|
|
| Back to top |
|
|
gjaky
Joined: 05 Jun 2010
Posts: 2802
Location: Budapest, Hungary
|
|
| Back to top |
|
|
redfox001
Joined: 16 Mar 2009
Posts: 2257
Location: The Netherlands
|
| Posted: Mon Aug 25, 2014 8:20 am Post subject: |
|
|
Thanks! Lot to think about. Scanning the article I think I was estimating the pixel rate that is 60Mp/s for 720p in my case.
If I want to display these pixels I need a square wave of 30MHz because a black and a white pixel together are one wave so 1/2*60=30MHz.
To fully resolve there is this bandwith equation that says to multiply with 6,55. That gives 197MHz. And that is still close to the specs of 180MHz.
The first halve of the article is more about the details we can see in real video. I am more interested in testpictures Please elaborate on my over simplifying if so.
|
|
| Back to top |
|
|
redfox001
Joined: 16 Mar 2009
Posts: 2257
Location: The Netherlands
|
| Posted: Mon Aug 25, 2014 9:00 am Post subject: |
|
|
One small thing about Barco's video amplifiers on input and RBG driver boards I read this.
| Quote: |
I didn't say they were better than any other, I was just saying that with all the hype about the 909 being the one to beat, they sure used alot of HFA110's, which is flat to only 100Mhz.
|
Now this chip is the one MP has replaced so I guess fellow modders we have to look for this chip on the input too. This might be one bottle neck.
http://www.avsforum.com/forum/16-crt-projectors/368896-what-best-crt-projector-out-there-32.html
|
|
| Back to top |
|
|
redfox001
Joined: 16 Mar 2009
Posts: 2257
Location: The Netherlands
|
| Posted: Mon Aug 25, 2014 2:08 pm Post subject: |
|
|
Ok I modified the RGB driver green path. I could only see the video amplifier being replaced and all surrounding components (R/C) where left the same. So now I have a modified input where these video amps are replaced MP has done that and a modified RGB driver where the green video amp is replaced. It might actually look a little better but I don't know if the bandwidth picture shows much.
I suspect the vga adapter to be the bottleneck. Those neckboards are supposed to be 180MHz. Can not believe they are less than 720p capable.
114MHz
The modified opamps are likely the CLC449 because they have the same pin's. So bad MP does not reply on questions or does someone have a personal email adres for me?
| Description: |
|
| Filesize: |
354.67 KB |
| Viewed: |
5869 Time(s) |
|
Last edited by redfox001 on Mon Aug 25, 2014 2:44 pm; edited 1 time in total
|
|
| Back to top |
|
|
redfox001
Joined: 16 Mar 2009
Posts: 2257
Location: The Netherlands
|
| Posted: Mon Aug 25, 2014 2:27 pm Post subject: |
|
|
About these video opamps in the Barco video chain. They are the HFA1100. They are in the input and on the driver.This is what MP says with regard to these chips:
| Quote: |
Most of the later Marquees use the CLC449 (Comlinear) op-amp. They replaced the CLC409 and H1100. Both of these chips had a roll off problem and could cause smearing and slight distortion in the image when used on certain resolutions. The H1100 may not show a problem, and may not be a problem from testing, so the replacement of the H1100 is an upgrade option. CLC449 is obsolete.
So far I've received very good reports from the chip upgrade on the neck boards, and some have already replaced the chips on the VIM, with very good results
New chip options:
Harris (Intersil) HFA1100's: They're basically the same bandwidth and would work well.
|
So I guess the HFA1100 are not so bad but can be better?
| Quote: |
The MAX4224ESA (Dallas Semiconductor) chips: I've already installed them on a neck board, tested and viewed. So far this chip seems to be a perfect substitute. I could not even notice any difference in the gray scale. This chip seems to be a good replacement for the now obsolete CLC449. The MAX is a much faster chip than the CLC409, so I would suggest that you first try changing the resistors values to around 249 ohm or something near that. 120 ohm feedback/ground resistor is definitely too low of a value for the speed of the Max4224. I would also replace the bypass caps and put one 4.7 mfd and one 0.1 mfd in parallel on each rail (+/-) of the chip. If used on the neck board, replace R12 (261 ohm) with a 331 ohm, but it must be a metal film @ 1 %. R12 is the feedback resistor for U2.
|
http://www.curtpalme.com/Marquee_Maintenance1.shtm
|
|
| Back to top |
|
|
stridsvognen
Guest
|
| Posted: Mon Aug 25, 2014 3:12 pm Post subject: |
|
|
You need something like this at 178Mhz.
Save the tubes focus yokes and lenses, and trow away the rest. 
|
|
| Back to top |
|
|
gjaky
Joined: 05 Jun 2010
Posts: 2802
Location: Budapest, Hungary
|
| Posted: Mon Aug 25, 2014 4:17 pm Post subject: |
|
|
| redfox001 wrote: | Ok I modified the RGB driver green path. I could only see the video amplifier being replaced and all surrounding components (R/C) where left the same. So now I have a modified input where these video amps are replaced MP has done that and a modified RGB driver where the green video amp is replaced. It might actually look a little better but I don't know if the bandwidth picture shows much.
I suspect the vga adapter to be the bottleneck. Those neckboards are supposed to be 180MHz. Can not believe they are less than 720p capable. |
The Barco's 180MHz bandwidth is only a marketing BS, that have been dicussed a few times already.
What I find interesting -and I'm not writing this for the first time- MP and even Greg Eiseman modified the different preamplifier boards in the 909 series, but the break through never really happened in the bandwidth. On the other hand there is no modification for the neckboards, and this forcing me to think that the real bottleneck is the neckboard, I could be wrong of course...
_________________
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 |
|
|
stridsvognen
Guest
|
| Posted: Mon Aug 25, 2014 4:23 pm Post subject: |
|
|
It seems that Greg Eiseman is the only one realy modding the 909 these days, but there is not realy any thing who indicate thet there has been a bandwidth improvement who matters.
It looks that he just continue the marketing BS stating high nr. He have never been interested to show any pictures of the best bandwidth performance he have been able to do.
|
|
| Back to top |
|
|
gjaky
Joined: 05 Jun 2010
Posts: 2802
Location: Budapest, Hungary
|
| Posted: Mon Aug 25, 2014 4:39 pm Post subject: |
|
|
| stridsvognen wrote: | It seems that Greg Eiseman is the only one realy modding the 909 these days, but there is not realy any thing who indicate thet there has been a bandwidth improvement who matters.
It looks that he just continue the marketing BS stating high nr. He have never been interested to show any pictures of the best bandwidth performance he have been able to do. |
Maybe I should get in the 909 modding business too, things needed: a fancy web page, promise double the bandwidth than Greg's do, ask 2/3 the price
Seriously, it would be interesting to see what's in the hybrid IC on the neckboard, one should do backengineer it, that should give use more insight...
_________________
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 |
|
|
stridsvognen
Guest
|
| Posted: Mon Aug 25, 2014 4:49 pm Post subject: |
|
|
| gjaky wrote: | | stridsvognen wrote: | It seems that Greg Eiseman is the only one realy modding the 909 these days, but there is not realy any thing who indicate thet there has been a bandwidth improvement who matters.
It looks that he just continue the marketing BS stating high nr. He have never been interested to show any pictures of the best bandwidth performance he have been able to do. |
Maybe I should get in the 909 modding business too, things needed: a fancy web page, promise double the bandwidth than Greg's do, ask 2/3 the price
Seriously, it would be interesting to see what's in the hybrid IC on the neckboard, one should do backengineer it, that should give use more insight... |
If you join the 909 mod crew, just remember only sell to those who dont test your mods.
|
|
| Back to top |
|
|
redfox001
Joined: 16 Mar 2009
Posts: 2257
Location: The Netherlands
|
| Posted: Mon Aug 25, 2014 5:00 pm Post subject: |
|
|
You might be right. I was thinking the same thing but modding the Barco has mostly been the 808 and 1209 models it seems. But if it was easy Eisemann would have shown pictures so I have little trust that he solved it.
But still I don't want to jump to conclusions as I don't have a good hdmi converter yet and that is absolutely needed.
And I do have the feeling I learned a lot these past days. The neckboard chip is the only mysterie left. It should produce 110V high bandwith. There must be a chip somewhere maybe even in the Marquee world! Does someone have a service manual describing the Marque neckboard? There must be something we can copy ?
|
|
| Back to top |
|
|
|
|
