or Connect
AVS › AVS Forum › Display Devices › Digital Hi-End Projectors - $3,000+ USD MSRP › Sony Ruby Service Mode
New Posts  All Forums:Forum Nav:

Sony Ruby Service Mode - Page 3  

post #61 of 141
don't forget this article as well, but you will need to convert from German

http://www.cine4home.de/tests/projek...ubyPreview.htm
post #62 of 141
Thread Starter 
Quote:
Originally Posted by Paul Carleton
Wow, since when did AVS get so down on Tweaking?
Paul,

There are many of us here who access service/factory parameters when they provide a useful benefit. Like yourself, I have been tweaking my displays for decades.

My business is not ISF, so I have never gotten into the private club of shared information.

Many AVS members willing share discoveries for fine tuning various displays. Some do not. I started this thread because I have my eye on the Ruby, and I plan to follow the tweaking discussions, just as I did before I bought the HS-51.

I look forward to hearing your observations, and hope the Ruby service/factory menus are a little more useful than the HS-51.
post #63 of 141
Thread Starter 
Quote:
Originally Posted by tycoondog2
don't forget this article as well, but you will need to convert from German

http://www.cine4home.de/tests/projek...ubyPreview.htm

The website seems to talk alot about gamma near the end. I link the image below as a possible indication of what's happening with BC.

http://www.cine4home.de/tests/projek...mmaDynamik.jpg

Unfortunately, google does not translate the entire page. If anyone knows a better translator that might tackle the whole page, please let me know.
post #64 of 141
copy and paste the text that is further down rather then doing the whole URL
post #65 of 141
Quote:
Originally Posted by Paul Carleton
Darin, not yet but I will. The reason is I have not decided on what screen I am going with yet. I was planning on diving into this when I can adjust it to the screen I settle on and after I get a few hours on the bulb.

I did notice some iris adjustment paramenters under the "Other" directory in the Factory Settings. Here are the numbers for evaluation only.

43 Iris/Open Reg: 319
44 Iris/Close Reg: 579
45 Iris/Open Hall: 694
46 Iris/Close Hall: 694
47-49 N/A
50 Iris/On_Reg: 522
51 Iris/Max_Correct_Value: 512
Quote:
Originally Posted by Bytehoven
The bottom two numbers 50/51 leave alone. They react to the source image an fluctuate.

The top four numbers 43/44/45/46 control the way the iris reacts from both ends of the luminance spectrum. This is just like the HS-51, except the HS-51 only has 256 values, while the Ruby seems to have a greater range than 256 8 bit steps. Maybe it's still 8 bit, but with a different number range
Thanks. I ended up changing #43 and #44 and am trying it that way. When I went in there the values were 295 and 563. The value was the same for #45 and #46 at 612, but I noticed those would change on their own (maybe as I changed the other two). I used HDMI from my 5U to HDMI input on the projector with the 1080i DVE tape and ended up trying 70 for contrast and 52 for brightness.

Then I found that by lowering #43 I could get the iris to stay more open on mid APL scenes. By raising #44 I could get the iris to close down further on all video black or below. I'm sure this will come at a price, but at one point I measured close to 30k:1 on/off CR in the center of the screen (my corners are about twice as bright with all video black). I tried 250 for #43 and 700 for #44 (I watched the iris close and the 700 was about where it got to the minimum). I then watched one of the Miracle Planet shows from DiscoveryHD. I ended up seeing a fair amount of what looked like white crushing or brightness compression, but when I would pause it and go switch between iris auto and iris off or on I found that it most cases it was in the source. There were a few cases where there was some brightness compression but they weren't major IMO and were not making the problem much worse than it already was from the crushed whites in the source (I've seen this quite a bit with DiscoveryHD in the past). But there was one place on a ladies face where I thought the brightness compression was reasonably distracting. So, I hooked up my PC with the Image Director software and made a custom gamma for Gamma3 with a little bit of white expansion just below the top of the gamma curve. I tried this and it got rid of just about all of the brightness compression on that one image.

I still wasn't able to get rid of the blotchy skin problem on Nicole Kidman in one spot in "Moulin Rouge" with the auto iris, but I'm guessing that this is an error in Sony's algorithms that won't get fixed unless Sony provides a fix. If somebody could get them to look at the particular scene (17:30 from the beginning of the D-Theater tape version, so probably more like 16:00 into the movie), that would probably be our best bet for a solution to that one.

I still need to go watch some things with my new settings (although I haven't done a color calibration and don't feel like spending the time for that right now) and see if going for the extra CR is going to bother me in some way.

I think the #43 and #44 might be useful for people with some different goals. One possibility being to be able to use the auto iris with the lumens of iris "off", but the CR of iris "on". In other words, maybe it would be possible to get 600 lumens and 6000:1 on/off CR without much brightness compression (since the iris wouldn't need to close much to double the on/off CR from 3000:1) for people who would prefer that. It would seem like people who understand what they are doing could tweak these to their own liking to some extent.

--Darin
post #66 of 141
Thanks Darin, nice post. Thanks for spearheading this. :cool:
post #67 of 141
Great work guys.
post #68 of 141
Quote:
Originally Posted by darinp2
One possibility being to be able to use the auto iris with the lumens of iris "off", but the CR of iris "on". In other words, maybe it would be possible to get 600 lumens and 6000:1 on/off CR without much brightness compression (since the iris wouldn't need to close much to double the on/off CR from 3000:1) for people who would prefer that. It would seem like people who understand what they are doing could tweak these to their own liking to some extent.

--Darin
That's basically what I was trying to convey in one of my above posts. I think this is achievable without going into the svc. menu, to some extent (lower the contrast in the "auto" mode).

I am very interested in your continued experimentation into the svc menu, and the effects that changing specific parameters provide. Please keep us posted on your reslults !

Thanks,

- Chris
post #69 of 141
Regarding convergence, I haven't seen the Ruby, but on the HS50 there is a "frame" of unused pixels (about 2-3 pixels) all around the image, even using the full "official" panel resolution, which is 1280x720 on the HS50 (I'm using 1:1 pixel mapping with an HTPC). This is only visible very up close to the screen.

I suppose this could be used to tweak the convergence, as it doesn't seem to make sense to have those extra pixels on the panels for any other reason.

Also, a person from Sony confirmed that electronic convergence adjustments are possible on the HS50 in 1-pixel steps (which makes sense), though he said it required special software, only available at Sony service centers. I'm not sure, though, if this is actually true or just misinformation (probably), nevertheless I'm just writing what the guy said.

Finally, as most of you guys already found out, there doesn't seem to exist any controls for convergence on the service menu, though there are "missing entries" (skipping numbers) in some of the menus. This "missing entries" can be just "by design", but there maybe a "hidden" way of accessing them if they exist.

Cheers,
Dave.
post #70 of 141
Thanks to everyone for the pathfinding here, this can be scary stuff indeed sometimes.

And wouldn't it be nice if the imagedirector program could download/restore ALL of the particular unit's settings...
post #71 of 141
Quote:
Originally Posted by Iamjcl
That's basically what I was trying to convey in one of my above posts. I think this is achievable without going into the svc. menu, to some extent (lower the contrast in the "auto" mode).
It seems like lowering the contrast would give lumens closer to the "On" mode, where I was thinking of getting the lumens of the "Off" mode above. If you are okay with the lumens that you get after lowering the contrast in "Auto" mode, you might be able to raise #44 to get the iris to close further for video black. While this might give more brightness compression, if for a full white screen the iris is in the same position as for iris "On" (that is closed down somewhat) then you would be starting with 6000:1 and then raising #44 to close the iris down further to give half the light output could give something like 14,000:1 (I'm guessing at 7000:1 with the iris closed down) with 320 lumens and never going below 6000:1 static on/off CR. That higher starting static on/off CR could be an advantage and might be better than having the iris go from full open to half closed with a neutral density filter for some people.

I watched quite a few scenes last night with a neutral density filter and my settings of 140 for #43, 700 for #44, contrast set to 70, and my custom Gamma 3 where I put a little bit of white expansion near the top. I need to check whether the lower contrast reduced the white level and the CR. I might have been getting the close to 30k:1 that I had measured earlier, but I'm not sure because of the contrast setting and will have to measure again. I looked for brightness compression, but most of the time it seemed like it wasn't an issue. I found things with crushed whites already where the iris would dim the bright stuff, but as far as losing many details I had to look reasonably hard. I found a place in "Lost" where some detail on a guy's shoulder disappeared with the "Auto" mode, but it was a case where some detail was already gone and the "Auto" mode just made it somewhat worse. There was also an SD commercial where I think a couple's faces had the details compressed, but weren't gone. I wasn't seeing the iris at work even with the higher dynamic on/off CR, so that was good.

One thing that did strike me is how bright 30k:1 in the center and probably 15k:1 in the corners still looked in a lot of cases. I didn't measure, but I was probably watching at around 18 ft-lamberts for white, so this is likely one reason. But I think another reason is that for the dark scenes the iris has already shut down and then if it goes to a blackout from there my eyes have already adjusted to the lower level somewhat (although it does have the gamma curve adjustment).

One thing that is continuing to bother me are the brighter corners (kind of a continuous increase from the center to the corners). I hope we can figure out a way to fix this (or that Sony will do something). Even at this higher dynamic on/off CR they are kind of distracting. On my first Ruby they were about 3 times as bright as the center and on this one the upper left corner was also in this range, but I think I got it down to about 2x in each corner with the more closed iris for video black. With 3x using the Firehawk (to try to create some hotspotting to counteract this) didn't really help much. With 2x and sitting closer maybe it will and I'll have to try it later (I was using the High Power for the above). But anything like this would decrease my uniformity for other scenes.

--Darin
post #72 of 141
Thread Starter 
Darin...

Please note I revised my last post a little.

These are used for making adjustments as you noted.

43 Iris/Open Reg: 319
44 Iris/Close Reg: 579

These two should indicate values as the source image luminance chances.

45 Iris/Open Hall: 694
46 Iris/Close Hall: 694

The HS-51 does not have these, maybe they also have a benefit on the iris.

50 Iris/On_Reg: 522
51 Iris/Max_Correct_Value: 512
post #73 of 141
Quote:
Originally Posted by tycoondog2
William I think you will get alot of call when people manage to mess up the projector
I will probably just tell them to call Sony. :rolleyes:

Paul, It sounds like you are being careful enough, it's not you I worry about. I will bet you though that there will be people reading this and giving it a try who will not be so careful.

The first thing needed is a comprehensive backup tool. That's not a trivial task! JVC provides one to all that attend their factory service training. I wish Sony did.

About convergence: I do not know how the VW100's panels are assembled yet, but I would guess that they are aligned on a test jig and then cemented in place. Early JVC panels were soldered, but with ROHS they had to switch to cement, the panels could not take the higher solder temps. Panel alignment is a 3d puzzle, not just x,y but also z for size and focus. The convergence errors many of you are describing are a combination of all three. For example, if red matches green in the center but is inside green at left and right, the red panel is closer to the optical block than green and thus has a smaller size. Making an optical block that has post assembly adjustments *and* won't shift in shipping is very expensive. I know of only one LCOS projector that has this, the JVC QXGA. It would be a pleasant surprise to learn that Sony figured out a way to make an adjustable LCOS optical block at this price point but I seriously doubt it.

At this point my only interest in this projector is to port over the Gamma and Shading tools, and then license them. I really do not have time to do anything else.
post #74 of 141
Quote:
Originally Posted by wm
At this point my only interest in this projector is to port over the Gamma and Shading tools, and then license them. I really do not have time to do anything else.
By 'port' I take it you mean from the 004.....
post #75 of 141
Thread Starter 
Quote:
Originally Posted by wm
Panel alignment is a 3d puzzle, not just x,y but also z for size and focus. The convergence errors many of you are describing are a combination of all three.
Although an electronic convergence control for each of the panels would not address all of the 3D convergences challenges you mention, being able to individually shift each of the RGB panels a pixel or two, would help solve or minimize some convergence issues.

Were you ever able to access the HS-51 in the same manner as the Qualia with your tools?
post #76 of 141
A couple more notes. While I'm sure we can screw something up in the service or factory menu, there is a save to memory feature, so even the changes I made yesterday to #43 and #44 for the iris disappeared and I had to go make them again. I did find some stuff called something like, "DE/H. Shift R" with vertical and horizontal for each color. I believe the legal values were 0 to 3 and I figured I might be able to use them to fix a uniform panel misconvergence. But I didn't see anything change in the images when I changed them.

I went back to my 140 for #43 and 700 for #44 with my custom gamma. I believe that the 140 keeps the iris open all the way down to a full screen 25 IRE, or somewhere in there. I also switched to the Firehawk from the High Power and watched another episode of "Miracle Planet" from DiscoveryHD. I found that by using the Firehawk and going to 700 for #44 (which makes the iris close down a lot), I could hardly tell the corners were brighter even just on credits with a black background. I need to go watch an actual movie, but I'm feeling really good about these settings at the moment. Maybe somebody else would notice extra transitions because I've made it keep the iris open so far down in APL (Average-Picture-Level I believe) and then transition so far closed for the darkest stuff, but I'm not and it looks like it is working well for me at the moment. I believe one tradeoff is that in some mixed images the blacks aren't as dark as some other people get with the iris on "Auto" but I also don't get as much brightness compression in those and still get to go very dark (even darker than default auto mode) when it is called for. I'll report back after I watch some film stuff on the Firehawk like this.

--Darin
post #77 of 141
Quote:
Originally Posted by wm
About convergence: I do not know how the VW100's panels are assembled yet, but I would guess that they are aligned on a test jig and then cemented in place. Early JVC panels were soldered, but with ROHS they had to switch to cement, the panels could not take the higher solder temps. Panel alignment is a 3d puzzle, not just x,y but also z for size and focus. The convergence errors many of you are describing are a combination of all three. For example, if red matches green in the center but is inside green at left and right, the red panel is closer to the optical block than green and thus has a smaller size.
Yes, of course! I know that convergence is much more complex than simple 2D alignments. And electronic physical convergence control is simply too much to ask (too complex and too expensive to implement, no point in having it from a manufacture’s perspective). But for my HS50 (soon to be replaced, as it has a scratch - long story...), I have an exact 1-pixel shift on the green panel which is the same from left to right, top to bottom (I checked this thoroughly with the HTPC). In cases like this it would be perfect to have an electronic pixel-shift control, because all else is perfectly aligned. And I can see those unused pixels on the borders which are probably there for this.

BTW, you could check if these extra pixels are also present on the Ruby. On the HS50, the easiest way to see them is with an HTPC, an all-black picture with some dot's on the corners (to get the frame limits) and iris off (or else it's more difficult to see them).

Cheers,
Dave.
post #78 of 141
Which service menu item turns off vertical outlining (from sharpening)? And has one been found now that turns off horizontal as well?
post #79 of 141
Great work, all, on begining to decipher the svc. menu. I think the next week or so should be interesting as more of it begins to take shape.

Darin, how far out of convergence is your "keeper" PJ (and was it any different than your other one)?

Thanks for the updates.

- Chris
post #80 of 141
Quote:
Originally Posted by mhafner
Which service menu item turns off vertical outlining (from sharpening)? And has one been found now that turns off horizontal as well?
By Vertical outlining to you mean "Ringing"?
post #81 of 141
Quote:
Originally Posted by Paul Carleton
By Vertical outlining to you mean "Ringing"?
See WSR review of the Ruby.
post #82 of 141
Well, that didn't exactly answer my question to you. I am assuming then that the review meant ringing.

Anyway I have not read the WSR article. I don't have a subscription.

If we are talking about Ringing then from what I understand, in that review had a pre-production unit and ringing is not an issue anymore. I am not getting ringing on my Ruby on the Component or DVI outs. I am not sure on the S-Video which I am using for on screen display for my AV receiver. That does have ringing but I have made no adjustments to that input.

There are a number of edge enhancement parameters that I noticed in the Factory settings. If your concern has not been addressed then possibly it can be fixed in the factory service mode. The other option is to check your sharpness setting to make sure it is not over sharpening the picture to cause ringing.
post #83 of 141
While having a quick look in the factory settings I noticed some panel settings that look like they may affect convergence, anyone tried these yet?
post #84 of 141
Quote:
Originally Posted by JeffY
While having a quick look in the factory settings I noticed some panel settings that look like they may affect convergence, anyone tried these yet?
Yep and they don't seem to make any visible difference :(
post #85 of 141
Is misconvergence a common problem with the Ruby? And if so, how much is it noticed in during normal viewing?

Mark
post #86 of 141
Hi Mark how are you :)

The units vary some are pretty much bang on other may have a pixel row out. However the pixels are so small that IMO at normal viewing distance you will never see it. And certainly not when you are watching normal material rather than test patterns. :D
post #87 of 141
Quote:
Originally Posted by Elliot
Hi Mark how are you :)
Great thanks :D Looking forward to inviting you over to see the new cinema once complete!

Quote:
The units vary some are pretty much bang on other may have a pixel row out. However the pixels are so small that IMO at normal viewing distance you will never see it. And certainly not when you are watching normal material rather than test patterns. :D
Any comment on Sony's position on this? Is one pixel misconvergence within their accepted tolerance or will they take a unit back for tweaking until perfect?

Mark
post #88 of 141
It's not uncommon for sources or video processors to have 1 pixel chroma shift and we are talking SD resolutions here so the same as around 2.5 pixels on the Ruby.
post #89 of 141
At this stage Sony are quoting me up to 3/4 pixels out is acceptable, before all the us guys cry blue murder about that should not be allowed in wouldn't be over here etc (i agree btw) :D

You should bear in mind Sony Uk are a very different company to Sony US and believe it or not do not seem to understand that people expect things to be right.....

I am in discusssion with them now about allowing us to adjust the units in house....
post #90 of 141
Quote:
Originally Posted by Mark_H
Great thanks :D Looking forward to inviting you over to see the new cinema once complete!
So when are you coming to see me about the genelecs then :p you really should spend the time and come have a listen! and see a Ruby + Qualia while you are up ;)
New Posts  All Forums:Forum Nav:
  Return Home
This thread is locked  
AVS › AVS Forum › Display Devices › Digital Hi-End Projectors - $3,000+ USD MSRP › Sony Ruby Service Mode