iScan DUO vs. Lumagen XD for CMS? - AVS Forum
Video Processors > iScan DUO vs. Lumagen XD for CMS?
gremmy's Avatar gremmy 10:31 AM 02-28-2012
Is the CMS on the Lumagen more advanced than that on the iScan DUO?

I see many reviews using the term CMS to describe the color adjustment functionality in the DUO, but I don't see any indication that the luminance of the primaries/secondaries can actually be adjusted. Saturation, yes. Brightness, No. Or is this incorrect?

And can the Lumagen do both?

P.S.

I amost snagged a used Radiance XD off Ebay for $680 this weekend, but came in just short in the auction.

sceptre-lcd's Avatar sceptre-lcd 02:00 PM 02-28-2012
i only have experience with lumagen for past 5 years - it has the best easiest to use and well supported CMS implementation from everything i've read over the past 3 years on these forums.
Gordon Fraser's Avatar Gordon Fraser 03:54 PM 02-28-2012
I cannot give any info about dvdo...i can say that the lumagens cms works correctly in linear RGB colourspace and can adjust hue, saturation and brightness of primary and secondary colours.
TomHuffman's Avatar TomHuffman 11:43 PM 02-28-2012
The CMS for each is similar. The Lumagen has some advantages. You have much more flexibility when it comes to test patterns and it provides 3D. It also provides the option for grayscale/gamma in 5% increments. They also provide better support and regular updates. But the Duo is cheaper and I think easier to use.
blee0120's Avatar blee0120 01:46 AM 03-01-2012
Are both of these video processors be able to calibrate 4K devices or will they need to be upgraded? I want to be sure when I get a 4k projector that I won't need to buy a new processor.
Gordon Fraser's Avatar Gordon Fraser 07:32 AM 03-01-2012
Neither can output 4k but both can calibrate current sources being fed to a 4K display...with the caveat that only the lumagen will do this for 3d content i believe.
gtgray's Avatar gtgray 05:59 PM 03-05-2012
Along these same lines, the CMS on the DUO can only adjust up to 100 percent simulus. Do the Lumagens also have provisions to adjust at 109% as well?
rfpublic's Avatar rfpublic 11:34 AM 03-06-2012
The Radiance video processors have 75% and 100% patterns for adjusting CMS by color. You can also input your own external 109% color patterns to use for CMS calibration.

Press "Menu, Output, CMS's, [CMS], Color Gamut, By Color, [color], Ok". Press "6" to toggle 75/100 IRE patterns. Press "4" to toggle between different size window patterns and an external pattern. Pressing the "Help" button displays a list of all the controls available.

We don't recommend using a 109% pattern for CMS adjustment. This is a reserved value and no real world video uses this high a value. Also many displays don't have a linear response at this value. If you make adjustments using a 109% pattern make sure that you do lots of testing at lower values to make sure that the display has a linear response. It is likely that you won't get good calibration results using this method.

Randy Freeman
gtgray's Avatar gtgray 12:07 AM 03-08-2012
Quote:
Originally Posted by rfpublic View Post

The Radiance video processors have 75% and 100% patterns for adjusting CMS by color. You can also input your own external 109% color patterns to use for CMS calibration.

Press "Menu, Output, CMS's, [CMS], Color Gamut, By Color, [color], Ok". Press "6" to toggle 75/100 IRE patterns. Press "4" to toggle between different size window patterns and an external pattern. Pressing the "Help" button displays a list of all the controls available.

We don't recommend using a 109% pattern for CMS adjustment. This is a reserved value and no real world video uses this high a value. Also many displays don't have a linear response at this value. If you make adjustments using a 109% pattern make sure that you do lots of testing at lower values to make sure that the display has a linear response. It is likely that you won't get good calibration results using this method.

Randy Freeman

I was really talking about grayscale at 109 percent and it is the lack of linearity that I was attempting to go after. The displays gamma curve causes the color shift a bit more on each step past 238-240. One can certainly argue that what happens past 235 does not matter that much but it is something I can see... the picture quality is clearly better if the contrast/gamma stays flat out to 109 and there is no color shift. I can fake this on the DUO by reducing contrast on the input via picture controls at the cost of max brightness at 100 percent stimulus.Otherwise I have to take a major hit to max brightness using the displays built in controls and just lowering contrast to where things stay linear to 109. I am not sure how many ft lamberts are lost but it is significant.

If the general rule is no eye fatigue, no clipping and no color shift when setting contrast and the above white area is not ajustable you kind of limit dynamic range.. Now perhaps RGB grayscale tracking can't align as you go up but it seems to me if you can put up a 109 percent pattern and using the 100 percent zone controls get RGB level on the 109 percent gray pattern, having another zone control or two above 100 percent would be useful?

It would also so allow me to use more of the dynamic range of the set with out messing up higlights.
Naiera's Avatar Naiera 12:47 PM 04-12-2012
So the Radiance XD would be just as good as a Duo, or even slightly better, for getting correct colour reproduction on my future JVC X30 projector?
Gordon Fraser's Avatar Gordon Fraser 01:29 AM 04-13-2012
Quote:
Originally Posted by Naiera View Post

So the Radiance XD would be just as good as a Duo, or even slightly better, for getting correct colour reproduction on my future JVC X30 projector?

yes
jrp's Avatar jrp 09:05 PM 04-14-2012
I hesitate to bring the math into the discussion, but I guess I will...

Pretty much everyone in the consumer space - with the notable exception of Lumagen - uses the HLS (or HSL if you prefer. Hue Lightness Saturation) color space for what they call CMS calibration. The HLS color space is a non-linear color space and is NOT the correct color space to do CMS in. If you calibrate in HLS and can manage to get the calibration points correct (no easy task in many cases) all the other points will still be wrong –- to some degree. I say to some degree as how much depends on how far out the display was to start.

Lumagen uses a linear Gamma 3D color cube for CMS since it is mathematically the correct color space for CMS. When you calibrate CMS with a Lumagen you get linear results and so the rest of the colors are correct –- assuming the display has a linear response mode to calibrate. Most high Quality displays and projector now have a mode or two that have linear response, so this is likely to work out well.

I had the opportunity to chat with Charles Poynton, a world renowned color mathematician and he agreed with me on this. Go to www.poynton.com for more from him on the subject (not sure if this is covered or not but there is a lot of good info and pointers to his book, and upcoming book, on the subject).

So if you have a display that is close, you can mitigate the CMS with other CMS systems, but to get it right you need a Radiance Video Processor.

BTW: We are looking to add to the Radiance CMS system in the near future – stay tuned.
Naiera's Avatar Naiera 05:37 AM 04-15-2012
I trust my AV guy above pretty much all else, and he was fine with using a Duo to correct the X30's slight colour issue. But now I have a RadianceXD on the way, so it doesn't matter much now anyway.
Ronomy's Avatar Ronomy 08:50 AM 04-15-2012
I owned the Lumagen HDQ for a couple years. The menu system took a while to learn. Once you get the hang of it it does work well.

However when it came time to upgrade I ended up with the DVDO Duo. It is way easier to use and works great. Even my RS1 with way over saturated green looks amazing with the Duo. All the other colors and levels look fine when I measured them after Cal. Its the best bang for the buck. I needed more than 4 HDMI inputs and analog video inputs. Also two video outputs came in handy. Today I wish I had the ability to CMS calibrate 3D so a Radiance would be nice to have. I just can't see spending so much more just to add 3D CMS capabilities.

That's my two cents worth!

Ron
CharlesF's Avatar CharlesF 01:17 PM 04-15-2012
Does the Duo actually do the calculations in HSL space? I know it uses and HSL interface to the CMS, but can't it still do the calulations in linear-gamma RGB space? Thanks.
jrp's Avatar jrp 01:56 PM 04-15-2012
Quote:
Originally Posted by CharlesF View Post

Does the Duo actually do the calculations in HSL space? I know it uses and HSL interface to the CMS, but can't it still do the calulations in linear-gamma RGB space? Thanks.

Based on the non-linear results it gets that I have been told about by calibrators, and the way its controls work, I would have to say that yes, it is using HLS color space. Based on all if have heard its CMS is certainly not in linear-Gamma RGB space.
Ronomy's Avatar Ronomy 04:34 PM 04-15-2012
I am surprised Tom Huffman would support the Duo if it wasn't linear. I have never seen any problems with my readings.
Ronomy's Avatar Ronomy 10:09 PM 04-16-2012
Folks if you read Tom Huffmans review of the Duo you will find out that early versions for the CMS was nonlinear and that DVDO re-engineered the CMS to function in a linear color space.
jrp's Avatar jrp 11:33 AM 04-17-2012
Quote:
Originally Posted by Ronomy View Post

I am surprised Tom Huffman would support the Duo if it wasn't linear. I have never seen any problems with my readings.

The largest errors are inside the cube and not on the cube corners and at this time neither Chromapure or Calman can measure these points. So your calibrator would not have been able to measure them.

This might change in the future.

----

I don't think the Duo changed to linear-gamma RGB, and so it would still be doing CMS in the wrong color space. Someone from DVDO who actually knows for sure can chime in and correct me if I am wrong.
TomHuffman's Avatar TomHuffman 04:34 PM 04-17-2012
Quote:
Originally Posted by jrp View Post

The largest errors are inside the cube and not on the cube corners and at this time neither Chromapure or Calman can measure these points. So your calibrator would not have been able to measure them.

Well, ChromaPure can measure them. We have an advanced color management feature just for this purpose. We just can't do much about what is measured until a tool becomes available for adjusting color within the color space.

Quote:
Originally Posted by jrp View Post

This might change in the future.

I hope so. This is a huge hole in current consumer color management. Many (not all) of the CMS tools offered by display manufacturers do a fair job of adjusting the color at the gamut boundary, but often screw up performance inside the color space.

Quote:
Originally Posted by jrp View Post

I don't think the Duo changed to linear-gamma RGB, and so it would still be doing CMS in the wrong color space. Someone from DVDO who actually knows for sure can chime in and correct me if I am wrong.

As I recall, the Duo team initially applied correction in a gamma-weighted color space, and then later changed this to making changes in a linear color space, and then reapplying gamma prior to output.

From a user's perspective this helped at very low luminance levels when adjusting grayscale. Prior to the change, adjustments would initially seem to have no effect and then suddenly a huge change in RGB balance would appear.

For the record, the Lumagen does a better job in this respect. Even now, using the Lumagen to make adjustments at low levels of stimulus are much smoother and more predictable than with the Duo. This issue doesn't really arise for CMS adjustments because these are all done at 75% or 100% stimulus in any case. However, if we ever get a tool that adjusts throughout the color space, instead of just at the boundary, this issue will be important for CMS as well as grayscale.
noah katz's Avatar noah katz 11:47 AM 04-18-2012
So is it possible to say how the Duo's CMS shortcomings, and when used with Calman AutoCal, would manifest visually on a JVC RS10?

Also, does AutoCal set to the user's chosen gamma, or is that done independently?
cinema mad's Avatar cinema mad 09:29 PM 04-18-2012
In the lower luminance levels IE:0-40% stimulus for grayscale the steps get larger as you go lower with regards to adjustment, the more points the finer adjustment within reason: Duo has 11point Radiance 21point..
noah katz's Avatar noah katz 10:43 PM 04-18-2012
Quote:
Originally Posted by cinema mad View Post

In the lower luminance levels IE:0-40% stimulus for grayscale the steps get larger as you go lower with regards to adjustment, the more points the finer adjustment within reason: Duo has 11point Radiance 21point..

How does that translate to what you see on the screen?
Ronomy's Avatar Ronomy 11:15 PM 04-18-2012
Quote:
Originally Posted by noah katz View Post

So is it possible to say how the Duo's CMS shortcomings, and when used with Calman AutoCal, would manifest visually on a JVC RS10?

Also, does AutoCal set to the user's chosen gamma, or is that done independently?

The Duo does a nice job on my RS1 so it should work great on your RS10.
wildchild22's Avatar wildchild22 04:41 PM 04-19-2012
I have lumagen (HDP and XD) and dvdo products (HAD VP50 now have DUO).

I have found even the lumagen HDP scales better for 480i to 1080p compared to any dvdo product even the duo.
The duo has excessive haloing on childrens tv shows were the lumagens have none.

But for HDTV or anything higher than 480i I do not notice any haloing. I also find the lumagen easier to use for setting up for NTSC and PAL where my tv only supports 23.976hz,24hz and 59.94hz and 60hz.
The lumagen is heads and tails above for customizing. And more fun to play with.

But for pic quality and inputs you cannot beat the duo except the 480i isnt good in my opinion.
Ronomy's Avatar Ronomy 08:25 PM 04-19-2012
I agree on the 480i conversions! I like Lumangen but just didn't like the added cost at the time.

I rarely watch 480i TV these days and if its a DVD I still have a Toshiba XA2 for that and it continues to kill any upconversion I have seen. I do remember enjoying my old HDQ upconversion of 480i TV on directv. It really cleaned up the image. But the HDQ reduced HD picture quality compared to my Duo. Big time difference. I haven't tried using the Duo processing to make 480i look better.
Up
Mobile  Desktop