View Full Version : Lumagen Radiance now Supports 72 Hz!
CIR-Engineering 05-29-09, 08:49 AM :)
I am so happy about this I had to post. After working with Lumagen on this for some time, I got to try their new FW yesterday for 817x1920 @ 72 Hz. My impressions were really good so they went ahead and gave it a general release yesterday evening. The new FW seems to work perfectly at extracting 24 Hz from 60 Hz sources, and for tripling 24 Hz to 72 Hz.
The Radiance processors I feel give the best image I have ever seen on my G90 especially at 72 Hz. But until yesterday, the Radiance had troubles with falling out of film mode when running 72 and the screen would start tearing. Because of this, I have been stuck at 60 Hz since the Radiance was released... Well not anymore!
Also, another Radiance tip for 9" Marquee and 9" Sony CRTs... The Radiance has a full CMS and can fully correct primary and secondary colors. However, I don't generally like the way a digital CMS works with a CRT. That being said, the 9" Sony and Marquee projectors have a supper duper over saturated red primary. What I generally do with a Radiance on these CRT's is to partially correct the red primary. This brings in the red a little bit without applying full correction, and I think it looks absolutely fantastic this way.
To partially correct the red primary go into the CMS menu and then go into the CMS matrix. On the red, reduce the red from 1024 to 1015 and add between 60~75 of blue. This will bring red a bit closer to reference without sacrificing image punch. And those numbers should apply to all 9" Marquee and Sony projectors.
After you do it, all the colors will feel more natural and seem clearer.
craigr
Citation4444 05-29-09, 09:11 AM This is fantastic news and thanks for helping them with this. I have been driving my 2 Cine9's from the RadianceXE outputs with limited success. I've got a moome EXTHD on one output and a HDFury2 on the other output. They both work at the standard timings of 1080p, but the HDFury2 doesn't work at the custom timings I need to use. Maybe this new firmware will make the RadianceXE more friendly to CRT users. I can't test my Fury2 as I have sent it back for some changes, but I expect it back soon.
I generally don't use the Radiance CMS with the Cine9's. G and B are very close to standards (especially G) and R is only slightly over-saturated and not worth correcting. Both pj's are the same; virtually perfect primaries and color decoding.
Way to go Lumagen!
Bob
Alan Gouger 05-29-09, 09:39 AM This is good news. Previously I gave up using the Radiance with CRT. The timings just would not work so I have been using the VP50pro. Ill up date the the software and give this a try.
Craig has the 817x1920 @ 72 Hz been added to the preset list or does it have to be created as a custom resolution and does it show up as no scaling to avoid any NR being added ?
Thanks for the alert.
VPH-G90 05-29-09, 09:46 AM Without questioning the quality of radience XE I think the best way to correct the primary colors on a CRT projector is to change the C-element or color glycol.
I've got crystalio2 and G90, I spent a lot of hours to test different configurations including the 1920x817p 72 (that is my resolution to use for some time). I calibrated my best I tested the G90 and the CMS of Crystalio 2 to correct the primary. The result was good on each primary and 100% saturation, but the intermediates are very bad ! Colorimetry these large deviations. Impossible to correct. May lumagen is better?
But in principle it is better for the electronics of the CRT and the result to obtain the coordinates of the red primary with only one tube, red tube!
I'm curious to have graphics on the results obtained with the radiance XE with modified primers.
Thank you. (sorry for my bad English).
mp20748 05-29-09, 09:56 AM Without questioning the quality of radience XE I think the best way to correct the primary colors on a CRT projector is to change the C-element or color glycol
Yep, and regardless of any claims, that should be the only way get correct primary colors from CRT.
Alan Gouger 05-29-09, 11:44 AM Thanks for the heards up on this Craig.
Just swapped out the VP50Pro for the Radiance with the new update set to 817x1920 @ 72 Hz and it locked right on:)
Ive had some more time to play with this and it is working great. Thanks Lumagen.
Thanks for the heards up on this Craig.
Just swapped out the VP50Pro for the Radiance with the new update set to 817x1920 @ 72 Hz and it locked right on:)
Ive had some more time to play with this and it is working great. Thanks Lumagen.
Alan,
you didn't have to edit any timings for the Marquee ?
Michael
Alan Gouger 05-29-09, 06:06 PM Alan,
you didn't have to edit any timings for the Marquee ?
Michael
Switching from the VP50Pro to the Lumagen was a breeze with no additional editing required. I was not expecting this:)
Of course if you have never setup a memory for this resolution before then you
would have to go through a full setup.
Switching from the VP50Pro to the Lumagen was a breeze with no additional editing required. I was not expecting this:)
Of course if you have never setup a memory for this resolution before then you
would have to go through a full setup.
interesting, it was reported that the Marquee does not like most fixed
setups of VP's. Guess Lumagen modeled their setup for Marquee use.
Alan, can you please post the detailed settings of this setup?
Thanks
Michael
CIR-Engineering 05-29-09, 07:02 PM This is good news. Previously I gave up using the Radiance with CRT. The timings just would not work so I have been using the VP50pro. Ill up date the the software and give this a try.
Craig has the 817x1920 @ 72 Hz been added to the preset list or does it have to be created as a custom resolution and does it show up as no scaling to avoid any NR being added ?
Thanks for the alert.
It is the very first resolution at the top of the presets :) Called 817p.
You have to switch the output aspect ratio in the config to 2.35 and set your input aspect to 2.35 to avoid ALL scaling and get perfect 1:1 mapping.
Also, there is a bug that I found yesterday and Lumagen will correct it in the next FW update. You must keep your zoom out % steps set to the default of 15%. If you switch your zoom step to 5% the Radiance will scale 1920 to 1920 when using 2.35 as the output aspect and you will loose 1:1 mapping. To work around for the moment just keep your zoom steps at 15%.
craigr
CIR-Engineering 05-29-09, 07:04 PM This is fantastic news and thanks for helping them with this. I have been driving my 2 Cine9's from the RadianceXE outputs with limited success. I've got a moome EXTHD on one output and a HDFury2 on the other output. They both work at the standard timings of 1080p, but the HDFury2 doesn't work at the custom timings I need to use. Maybe this new firmware will make the RadianceXE more friendly to CRT users. I can't test my Fury2 as I have sent it back for some changes, but I expect it back soon.
I generally don't use the Radiance CMS with the Cine9's. G and B are very close to standards (especially G) and R is only slightly over-saturated and not worth correcting. Both pj's are the same; virtually perfect primaries and color decoding.
Way to go Lumagen!
Bob
I think we have about the same primaries don't we. I run a G90 here with a Marquee green C element and an LUG tube. My green and blue are darn near perfect, but red is well outside the rec 709 gamut. If your red is outside the gamut you might want to give it a try. I don't think the red should be fully corrected, but I think it helps to dial it back a little.
craigr
CIR-Engineering 05-29-09, 07:10 PM Without questioning the quality of radience XE I think the best way to correct the primary colors on a CRT projector is to change the C-element or color glycol.
I've got crystalio2 and G90, I spent a lot of hours to test different configurations including the 1920x817p 72 (that is my resolution to use for some time). I calibrated my best I tested the G90 and the CMS of Crystalio 2 to correct the primary. The result was good on each primary and 100% saturation, but the intermediates are very bad ! Colorimetry these large deviations. Impossible to correct. May lumagen is better?
But in principle it is better for the electronics of the CRT and the result to obtain the coordinates of the red primary with only one tube, red tube!
I'm curious to have graphics on the results obtained with the radiance XE with modified primers.
Thank you. (sorry for my bad English).
The Radiance is unique in that it has a full CMS for all three primary and secondary colors. It is capable of fully correcting all the colors techinically as long as your are over saturated to start with. No amount of digital processing can correct under saturated colors.
Generally, most CRT's have under saturated colors to start so the Radiance can not correct for this. However, some of the 9" projectors do have over saturated primaries so you could use the radiance.
Inversely, the G90 Sony C element has an under saturated green. You can not use the Radiance to correct it. But once you put a Marquee green C element in a G90 the green becomes slightly over saturated.
All that being said, I personally do not like to use a digital CMS to fully correct primaries on a CRT. Numerically with test patterns the image looks better. However, holistically when I look at the real world picture when using the CMS I don't like it. All I suggest is that I like the look of a slight partial correction to reduce the saturation of red slightly. With the values I posted, the red is still well outside the rec 709 gamut, it is just moved a little closer.
craigr
CIR-Engineering 05-29-09, 07:12 PM Yep, and regardless of any claims, that should be the only way get correct primary colors from CRT.
I know Mike fully understands this because him and I spoke of it on the phone yesterday. So this is just for the record and to reiterate what I said. Most CRT's are under saturated to start so a digital CMS can't help them. This is especially the case with green on many projectors. However, there are some CRT's that do have one or more primary outside the rec 709 gamut so a CMS technically could be implemented to improve the colors. That being said, just because you can technically improve the colors doesn't mean that the actual image quality will improve and it may get worse.
craigr
CIR-Engineering 05-29-09, 07:19 PM interesting, it was reported that the Marquee does not like most fixed
setups of VP's. Guess Lumagen modeled their setup for Marquee use.
Alan, can you please post the detailed settings of this setup?
Thanks
Michael
I do modify the timings on the Marquee because it reduces the raster ringing on the left side. These are the timings I use on both the Marquee and the G90 with the Lumagen Vision and Radiance processors.
818x1920 72 Hz
Vtotal = 855
Vactive = 818
Vfront = 4
Vsync = 5
Vrate = 71.93
Htotal = 2200
Hactive = 1920
Hfront = 90
Hsync = 88
Clock = 135296703
1080p 60 Hz
Vtotal = 1125
Vactive = 1080
Vfront = 4
Vsync = 5
Vrate = 59.94
Htotal = 2200
Hactive = 1920
Hfront = 90
Hsync = 88
Also, if you want to do custom resolutions on the Radiance you have to enter the clock rate by hand. Here is a link to a spread sheet I put on my web page to calculate clock rates CLOCK CAL LINK. (http://www.cir-engineering.com/bin/clock_calc.xls)
craigr
Alan Gouger 05-29-09, 08:12 PM Craig is it possible to change the refresh from the preset 817 x1920 72 to 48?
Thx
overclkr 05-29-09, 08:37 PM Craig is it possible to change the refresh from the preset 817 x1920 72 to 48?
Thx
Do you like 48 Alan? I found it to be too "flickery".
Cliff
CIR-Engineering 05-29-09, 09:51 PM Craig is it possible to change the refresh from the preset 817 x1920 72 to 48?
Thx
Yes. For 817x1920 @ 48 Hz you just have to calculate the clock rate and enter it. Download the spread sheet for clock calculation off my web page HERE (link). (http://www.cir-engineering.com/bin/clock_calc.xls) Go to the Lumagen output timings menu and check your total V and H lines and enter them into the spread sheet. Enter 48 into the "Frequency Hz" space on the spread sheet. The clock will automatically be calculated by the spread sheet.
Enter the clock number from the spread sheet into the Lumagen and the refresh rate in the Lumagen will automatically change to 47.95 Hz.
If you have any trouble just ask here and I can help. All of the red test in the spread sheet can be changed to your own custom timings to calculate the clock rate for the Lumagen.
craigr
Alan Gouger 05-29-09, 11:14 PM Do you like 48 Alan? I found it to be too "flickery".
Cliff
Hey Cliff
As long as you do not drive the contrast to hard you will rarely see the flicker
but in the brightest of scenes.
Coming from Film I am use to 48fps flicker so I feel right at home:)
Yes. For 817x1920 @ 48 Hz you just have to calculate the clock rate and enter it. Download the spread sheet for clock calculation off my web page HERE (link). (http://www.cir-engineering.com/bin/clock_calc.xls) Go to the Lumagen output timings menu and check your total V and H lines and enter them into the spread sheet. Enter 48 into the "Frequency Hz" space on the spread sheet. The clock will automatically be calculated by the spread sheet.
Enter the clock number from the spread sheet into the Lumagen and the refresh rate in the Lumagen will automatically change to 47.95 Hz.
If you have any trouble just ask here and I can help. All of the red test in the spread sheet can be changed to your own custom timings to calculate the clock rate for the Lumagen.
craigr
Thanks Craig Ill ck it out. Id love to have both 48 and 72 in memories.
Thanks for the spread sheet.
Alan Gouger 05-30-09, 12:49 AM Craig not sure if I am doing this correct. I entered the appropriate numbers 1920 817 and 48 and I ended up with 75219500 for the clock. I hit enter in Test mode and the projector did not like that. That number seams low, did do this correct, thanks.
CIR-Engineering 05-30-09, 01:16 AM Craig not sure if I am doing this correct. I entered the appropriate numbers 1920 817 and 48 and I ended up with 75219500 for the clock. I hit enter in Test mode and the projector did not like that. That number seams low, did do this correct, thanks.
You must enter the TOTAL pixels. For 1920 the default is 2200, for 817 you need to look at what the timings menu says for Vtotal and use that (probably around 875 by default).
craigr
THE_COW_IS_OK 05-30-09, 04:38 AM Hey Cliff
As long as you do not drive the contrast to hard you will rarely see the flicker
but in the brightest of scenes.
Coming from Film I am use to 48fps flicker so I feel right at home:)
Ditto. Even with the brightest scene, it will only show on the brightest parts of the scene. I rarely notice it.... I lived with 50Hz pal all my life, that must count into something.....
Sometimes I think its worth it with all the bandwidth it save.
Anyways, thanks for the heads up! Will update my firmware tonight.
VPH-G90 05-30-09, 06:35 AM The Radiance is unique in that it has a full CMS for all three primary and secondary colors. It is capable of fully correcting all the colors techinically as long as your are over saturated to start with. No amount of digital processing can correct under saturated colors.
Generally, most CRT's have under saturated colors to start so the Radiance can not correct for this. However, some of the 9" projectors do have over saturated primaries so you could use the radiance.
Inversely, the G90 Sony C element has an under saturated green. You can not use the Radiance to correct it. But once you put a Marquee green C element in a G90 the green becomes slightly over saturated.
All that being said, I personally do not like to use a digital CMS to fully correct primaries on a CRT. Numerically with test patterns the image looks better. However, holistically when I look at the real world picture when using the CMS I don't like it. All I suggest is that I like the look of a slight partial correction to reduce the saturation of red slightly. With the values I posted, the red is still well outside the rec 709 gamut, it is just moved a little closer.
craigr
Craigr why not to do SMPTE-C gammut ? Most blu ray have the SMPTE-C gamut, if you want to be in the standard must be based on this gamut.
Here is my data with the Crystalio to try with lumagen
817x1920 72 Hz
Vtotal = 832
Vactive = 817
Vfront = 4
V back porch = 6
Vsync = 5
Htotal = 2140
Hactive = 1920
Hfront = 88
Hsync = 24
H back porch = 108
refresh rate = 71.93
Pixel clock 128.07 MHz
We must set the shift and the convergence of the CRT projector
With this data I have very good result with my G90.
I noticed that when you exceed 130-132 MHz pixel clock, the image of the G90 was less precise and became softer despite optimal settings.
To 128 MHz pixel clock the sharpness of the image is very close to the sharpness of the image 1920x817p48, but with the advantages of 72 Hz, there was no flicker :)
Alan Gouger 05-30-09, 07:42 AM You must enter the TOTAL pixels. For 1920 the default is 2200, for 817 you need to look at what the timings menu says for Vtotal and use that (probably around 875 by default).
craigr
Ahhh.."total" I missed that part. Thats what I get for taking the Evelyn speed reading course:)
2200 & 875 are the magic numbers.
Edited: 817 to 875:)
CIR-Engineering 05-30-09, 10:24 AM Ahhh.."total" I missed that part. Thats what I get for taking the Evelyn speed reading course:)
2200 & 817 are the magic numbers.
Well, 2200 is correct, but 817 is still not the TOTAL vertical pixels. Be sure to use total pixels for both or your clock rate will be too high and not be a multiple of 24Hz!!!
craigr
CIR-Engineering 05-30-09, 10:28 AM Craigr why not to do SMPTE-C gammut ? Most blu ray have the SMPTE-C gamut, if you want to be in the standard must be based on this gamut.
Here is my data with the Crystalio to try with lumagen
817x1920 72 Hz
Vtotal = 832
Vactive = 817
Vfront = 4
V back porch = 6
Vsync = 5
Htotal = 2140
Hactive = 1920
Hfront = 88
Hsync = 24
H back porch = 108
refresh rate = 71.93
Pixel clock 128.07 MHz
We must set the shift and the convergence of the CRT projector
With this data I have very good result with my G90.
I noticed that when you exceed 130-132 MHz pixel clock, the image of the G90 was less precise and became softer despite optimal settings.
To 128 MHz pixel clock the sharpness of the image is very close to the sharpness of the image 1920x817p48, but with the advantages of 72 Hz, there was no flicker :)
Those timings are really similar to some that I have used in the past. I just find that with less than 2200 H pixels I start to see the raster ringing. The image is sharper with less as you observe, but I don't like the raster ringing on the left.
The gamut for HD is rec709 and depending on wheather or not the BD is mastered properly will dictate what gamut is it using. HD should really always use 709. That being said, you can't change the gamut of a CRT without changing the tubes, C-elements, or glycol coloring. If you run SMTPE-C your projector will create primaries that are even further outside the gamut. The G90 is closest to rec 709 with the Marquee green C-element than it is to SMTPE-C. No matter what gamut you feed the projector to start, the xy coordinates for red, green, and blue will be the same coming out of the projector ;)
craigr
overclkr 05-30-09, 11:37 AM Hey Cliff
As long as you do not drive the contrast to hard you will rarely see the flicker
but in the brightest of scenes.
Coming from Film I am use to 48fps flicker so I feel right at home:)
I'm gonna have to try it again big dog. It's been a while. :)
Alan Gouger 05-30-09, 11:59 AM I'm gonna have to try it again big dog. It's been a while. :)
Your setup is unique in that you have so much light output I think its going to be a no go with your stack:)
overclkr 05-30-09, 11:42 PM Your setup is unique in that you have so much light output I think its going to be a no go with your stack:)
Hopefully I can get Ken here next friday to hang out for a triple stack pow wow. :cool:
That mystery guest from the Triple meet is coming along with another good friend who I have not seen for a while. :)
Should be an interesting comparison between 48 and 60. I'm kind of bummed because Art's G90 is giving me an elevated green raster right now with lines. I'm hoping to "bump" the tube to get it to stop. We'll see what happens.
Either way, I think they will be mucho impressed. ;)
Cliffy
Alan Gouger 05-30-09, 11:53 PM That mystery guest from the Triple meet is coming along with another good friend who I have not seen for a while.
Who could that or they be?
Either way, I think they will be mucho impressed.
You just may mucho spoil them:)
VPH-G90 06-04-09, 01:16 PM Those timings are really similar to some that I have used in the past. I just find that with less than 2200 H pixels I start to see the raster ringing. The image is sharper with less as you observe, but I don't like the raster ringing on the left.
...
craigr
Hi craig,
Strange, i have no raster ringing on the left at 128 MHz !
video image is very stable
Mark_A_W 06-05-09, 04:33 AM Do you like 48 Alan? I found it to be too "flickery".
Cliff
48 or 50hz is like watching a strobe light to me.
And no, I don't like the flicker at the movies either...
Alan Gouger 06-20-09, 12:34 AM Using the Lumagen at 1920 x817 at 47.95 I get rap around. Im missing a few inches off the image. The clock rate im using is 92307692.
Changing the retrace does not help it makes it worse.
How do you correct for this using the lumagen. Changing the hfront or hsync does not help. I have to change hactive from 1920 to 1895 to fit everything in and now the processor is scaling.
CIR-Engineering 06-24-09, 11:21 AM Have you used the G90 "SHIFT" control to move the image inside the raster? The "CENTER" control moves the image and raster, but the "SHIFT" control moves adjusts the image placement inside the raster...
craigr
nashou66 06-24-09, 11:36 AM Craig he has a Marquee.
Athanasios
CIR-Engineering 06-24-09, 05:13 PM Craig he has a Marquee.
Athanasios
Ah... never mind on the shift then :p
But has the raster/image orientation been checked?
craigr
|