AVS Forum banner

Official JVC RS20 / HD750 Calibration and CMS thread (NEW FIRMWARE V1.1)

218K views 2K replies 137 participants last post by  rollon1980 
#1 ·
[EDIT 06-05-09: I have edited this first post to add some info from the first calibration thread and to make it a useful table of content to the new calibration and CMS thread]


This thread is a continuation of the original official RS20/hd750 calibration and CMS thread here http://www.avsforum.com/avs-vb/showt...8#post15314888 .


As the old calibration thread is mostly obsolete regarding CMS calibration with the arrival of the new firmware, we have decided to start afresh!


First things first, you need to upgrade the firmware before using the tips available in this thread.


A - NEW FIRMWARE

Where to get the firmware update (and common installation issues): http://www.avsforum.com/avs-vb/showt...8#post16355768

CAUTION: the new firmware erases ALL settings, so if you have general settings or calibration settings you wish to keep, make sure you make note of them before upgrading.


There is a wealth of knowledge regarding greyscale and gamma calibration in the original calibration thread. So here is a list of the most important posts which are still mostly relevant with the new firmware:


B - GREYSCALE: (from original thread, but info still relevant)

Before we get into CMS settings, the first thing to do is to get the best possible greyscale using only color temps adjustments, as this will make a huge difference with THX (and other presets). As already reported, some excellent tips are available here to achieve this from Mark Petersen Official JVC RS20 / HD750 Calibration and CMS thread - Page 22 - AVS Forum and also from Googer and others in the rest of the thread, here is one of Googer's great posts about gamma: http://www.avsforum.com/avs-vb/showt...#post15522847]


C - TWEAKED THX (from original thread, but info still relevant)

Once you have a custom color temp that gives you as perfect a greyscale as possible, the next step is to select the THX preset and enter the service mode (up/down/right/left/enter as fast as possible, either on the remote or the pj). [DISCLAIMER: be very careful when in this mode, you enter it at your own risks, if you make a mistake, don't blame me!]. You can then select a different color temp (for example the custom one you've just tweaked
), and this will be the color temp used by THX. You can now exit the service mode. This will make a HUGE difference to the ootb THX preset, as it will use a greyscale which is likely to be pretty good from 30 IRE up to 100 IRE, instead of a greyscale which is quite flat but off badly over the whole range, whether in normal or high lamp. More info about achieving this here Official JVC RS20 / HD750 Calibration and CMS thread - Page 29 - AVS Forum


If you like this "tweaked THX" preset, which is very close to rec709 - although slightly undersaturated for some - with a standard gamma, you can stop here and enjoy your PJ. If you do find it undersaturated, you can add anything from 1 to 12 - depending on your taste - to the general color control (I personally add 1-5 depending on source). That's it. Job done.


If you want to go one step further, and be able to use a custom gamma, sharpness and detail enhancement controls and adjust the gamut to Rec709, SMPTE-C or to your own taste, then have fun with the next section which is mostly about custom CMS settings...


D - CUSTOM SETTINGS, REPORTS and CALIBRATION TIPS with the NEW FIRMWARE


As members post reports/settings and calibration tips in the thread, I'll try to update this section with the most significant contributions (in chronological order).

Please let me know if I have forgotten something or if I get sloppy updating...


Reports from the pros:

GregR : http://www.avsforum.com/avs-vb/showt...7#post16360087

Tom Huffman: http://www.avsforum.com/avs-vb/showt...3#post16383023

Tom Huffman#2: http://www.avsforum.com/avs-vb/showt...9#post16431819

Settings:

Manni01 (1.4 gain screen): http://www.avsforum.com/avs-vb/showt...7#post16365507

JeffY: http://www.avsforum.com/avs-vb/showt...3#post16371803

Karrih: http://www.avsforum.com/avs-vb/showt...4#post16380924

008: http://www.avsforum.com/avs-vb/showt...9#post16383359

Nelson4u: http://www.avsforum.com/avs-vb/showt...7#post16388557

Ignace: http://www.avsforum.com/avs-vb/showt...7#post16389657

LovingDVD: http://www.avsforum.com/avs-vb/showt...6#post16482286

Lawguy: http://www.avsforum.com/avs-vb/showt...2#post16428562

Manni01 (1.2 gain screen): http://www.avsforum.com/avs-vb/showt...1#post16431351

MarcelW: http://www.avsforum.com/avs-vb/showt...9#post16433669

Lawguy LT vs i1Pro http://www.avsforum.com/avs-vb/showt...8#post16438408

RickS (HDMI Enhanced): http://www.avsforum.com/avs-vb/showt...4#post16630004

Manni01: back to 1.1 gain/new lamp, and HDMI Enhanced, settings for Rec709 and SMPTE-C http://www.avsforum.com/avs-vb/showt...0#post16662220

Manni01: updated Rec709 and SMPTE-C calibrations, along with a PAL calibration: http://www.avforums.com/forums/dlp-l...ml#post9767111

Calibration Tips:

Classic beginner's mistake and link to Tom Huffman's CMS calibration tutorial and GregR's calculator thread: http://www.avsforum.com/avs-vb/showt...3#post16370193

How to use filters to check brightness (LovingDVD): http://www.avsforum.com/avs-vb/showt...7#post16372917

When/why NOT to use filters (GregR & Tom Huffman): http://www.avsforum.com/avs-vb/showt...7#post16373717 and http://www.avsforum.com/avs-vb/showt...1#post16380651

dE discussion: http://www.avsforum.com/avs-vb/showt...5#post16374865

Tom Huffman's spreadsheet for CIELUV: http://www.avsforum.com/avs-vb/showt...7#post16378927

GregR's DisplayCalibration Calculator: http://www.avsforum.com/avs-vb/showt...2#post16382042

Get a screenshot in HCFR: http://www.avsforum.com/avs-vb/showt...8#post16383268

Which USB cable for the firmware upgrade?: http://www.avsforum.com/avs-vb/showt...7#post16383437

Training a meter to another in HCFR: http://www.avsforum.com/avs-vb/showt...4#post16384494

Adjusting greyscale (GregR): http://www.avsforum.com/avs-vb/showt...0#post16387960

Link to Calibration for Dummies: http://www.avsforum.com/avs-vb/showt...9#post16402839

Use of filter and calibration tips (Darinp2): http://www.avsforum.com/avs-vb/showt...9#post16431819

Saturation levels explained (Tom Huffman): http://www.avsforum.com/avs-vb/showt...2#post16439382

Sharpness/focus adjustment tips (GregR / LeDahu / Deanbob): http://www.avsforum.com/avs-vb/showt...2#post16689752

Great tips/tools for gamma calibration (LovingDVD / LeDahu): http://www.avsforum.com/avs-vb/showt...8#post17017488

Gamma tweaking V1.2 (english translation and VERY IMPORTANT SPECIAL PROCEDURE): http://www.avsforum.com/avs-vb/showt...9#post17045339

Gamma tweaking V2 (Le Dahu): http://www.avsforum.com/avs-vb/showt...6#post17060116

Contrast and RGB (Tom Huffman et al): http://www.avsforum.com/avs-vb/showt...8#post17150298

REST OF INITIAL POST:


I have done a first attempt with the new firmware, and the news is mostly good!


- The range for the controls has been extended from -60 to 60 (instead of -30 to 30), which gives enough range to dial the gamut perfectly without touching the general color control. Well done JVC!


- The linearity is good but not perfect. Most colors stay where they should, but green seems to be a bit oversaturated at 75% when calibrating at 100%. Nothing compared to the first firmware, but still not perfect. [edit: this was due to an error on my side, there is NO LINEARITY problem with the new firmware].


- I had only time for a quick calibration today, so I'll post more later. I attach my HCFR files for those who want to have a look at the details (EDIT: I used a brand new i1pro to calibrate).


If you want to try my settings (only after upgrading the firmware), here they are:


[EDIT: please do not use these settings, I and others have posted updated settings, see links above].


Contrast=0

Brightness=0

Color=0

Tint=0


Color temp (new lamp, probably not useful for most of you)

Gain R=-24, G=0, B=-54

Offset R=-1 G=-3 B=0


CMS (H,S,B)

Red -4 -23 7

Yellow 22 -44 38

Green -5 -39 44

Cyan -3 -48 39

Blue 37 -8 -4

Magenta -6 0 -3


I didn't have time to do any fine-tuning on gamma etc...

 

New firmware.zip 153.3935546875k . file



 

Attachments

See less See more
7
#377 ·

Quote:
Originally Posted by Ignace /forum/post/16397645


Damn! He's keeping it secret


Do you want to visit Belgium for a calibration
?

I used to work in Europe in my engineering days, but I only go there for vacation after I retired. I was in Belgium in 2007 on vacation.


Unfortunately my PR-670 is restricted to the US by Photo Research.
 
#378 ·

Quote:
Originally Posted by Ignace /forum/post/16397645


Damn! He's keeping it secret....

Even if I told you exactly what I was doing you would never spend the money to do it. It would cost you a bunch of time and money to replicate it. I have custom calibration software that I use with a programmable video generator and a $24,000 spectro integrated with a rig that can measure accurately to 1 IRE on these products. It would only benefit my competitors if I was to divulge exactly what I do.
 
#379 ·

Quote:
Originally Posted by umr /forum/post/16397671


I used to work in Europe in my engineering days, but I only go there for vacation after I retired. I was in Belgium in 2007 on vacation.


Unfortunately my PR-670 is restricted to the US by Photo Research.

Maybe a last question to get back on track of the purpose of the forum
. What do personally think of the choice to use HSB a space for the CMS? Would RGB not be better, and more linear?


Ignace.
 
#380 ·

Quote:
Originally Posted by Ignace /forum/post/16397748


Maybe a last question to get back on track of the purpose of the forum
. What do personally think of the choice to use HSB a space for the CMS? Would RGB not be better, and more linear?


Ignace.

I have not spent much time thinking about that. I only worry about things I can control. If I had responsibility for the code in these then I would have an opinion.


Thinking about it though HSB seems more likely to have issues than RGB like that found in Samsung products. It would be easy to get in a spot with HSB that was not logical given a specific set of display capabilities. I have seen similar problems with color management systems based on xy coordinate systems. I personally just deal with the hand I am dealt and make the best out of it I can.
 
#381 ·

Quote:
Originally Posted by umr /forum/post/16397720


Even if I told you exactly what I was doing you would never spend the money to do it. It would cost you a bunch of time and money to replicate it. I have custom calibration software that I use with a programmable video generator and a $24,000 spectro integrated with a rig that can measure accurately to 1 IRE on these products. It would only benefit my competitors if I was to divulge exactly what I do.

I still wonder what this Kung Fu
can solve on a limited CMS; 6 HSB points in only 10-bit space (if it is even 10-bit) reducing the gammut so much to the Rec. 709?


Ignace.
 
#382 ·
So you gray marketed it from the U.S.? Sweet!


To the boys. If you have a cheap meter, you can play. It fun. Will it be right? Not a chance but it may not be that wrong. If you are happy with the results and you think you have made it better. That's all that counts. Me and my customers get the best of the calibrators who have the best equipment. I'd rather watch the results then spend my leisure hours trying to do it myself. And I am ISF certified, have calman pro, a radiance, and at one time or another had or have almost all of the cheaper meters discussed here.
 
#383 ·

Quote:
Originally Posted by Ignace /forum/post/16397813


I still wonder what this Kung Fu
can solve on a limited CMS; 6 HSB points in only 10-bit space (if it is even 10-bit) reducing the gammut so much to the Rec. 709?


Ignace.

Who said Rec. 709 is the goal for film? Film is not Rec. 709. The cool thing about these units is that they can restore part of the lost film gamut while not distorting key colors.


We can hit Rec. 709, but that will not restore the look to what we see in a reference theater.
 
#384 ·

Quote:
Originally Posted by mark haflich /forum/post/16397825


So you gray marketed it from the U.S.? Sweet!


To the boys. If you have a cheap meter, you can play. It fun. Will it be right? Not a chance but it may not be that wrong. If you are happy with the results and you think you have made it better. That's all that counts. Me and my customers get the best of the calibrators who have the best equipment. I'd rather watch the results then spend my leisure hours trying to do it myself. And I am ISF certified, have calman pro, a radiance, and at one time or another had or have almost all of the cheaper meters discussed here.

Correct, that is why I am concerned about the color contouring artifacts I have seen in the movie "Hard Candy". Because I've seen if you change the CMS points, it does not seem to affect the issue. Any measurement tool will probably give the same artifact? I can only say I don't see it in THX, and I don't trust completely the user tunable CMS.


Ignace.
 
#385 ·

Quote:
Originally Posted by Ignace /forum/post/16397886


Correct, that is why I am concerned about the color contouring artifacts I have seen in the movie "Hard Candy". Because I've seen if you change the CMS points, it does not seem to affect the issue. Any measurement tool will probably give the same artifact? I can only say I don't see it in THX, and I don't trust completely the user tunable CMS.


Ignace.

I have not looked at Hard Candy, but a properly tuned CMS will not add artifacts. Poorly set ones definitely do. I saw one case with a Radiance where the Pixar logo was full of purple blocks mixed with the cyan background.


What you would find with better tools are that display gamuts are frequently not as wrong as cheaper instruments indicate. Because of this people are running out of range and creating artifacts.
 
#386 ·

Quote:
Originally Posted by umr /forum/post/16397841


Who said Rec. 709 is the goal for film? Film is not Rec. 709. The cool thing about these units is that they can restore part of the lost film gamut while not distorting key colors.


We can hit Rec. 709, but that will not restore the look to what we see in a reference theater.

Correct, movie is not Rec. 709. But the transfer to for example BluRay is converted to Rec. 709, mapping too saturated colors to less ones. How can you restore afterwards what is lost? It is like trying to de-interlace a video signal, there are always situations where it doesn't work.


Ignace.
 
#387 ·

Quote:
Originally Posted by Ignace /forum/post/16397920


Correct, movie is not Rec. 709. But the transfer to for example BluRay is converted to Rec. 709, mapping too saturated colors to less ones. How can you restore afterwards what is lost? It is like trying to de-interlace a video signal, there are always situations where it doesn't work.


Ignace.

The key is to do it subtly with the colors that we are most sensitive to that differ from film the most while not not ruining the most important colors. Basically you must reverse the process used to convert the film. JVC has the best shot at this of anything I have seen.
 
#388 ·

Quote:
Originally Posted by umr /forum/post/16397900


I have not looked at Hard Candy, but a properly tuned CMS will not add artifacts. Poorly set ones definitely do. I saw one case with a Radiance where the Pixar logo was full of purple blocks mixed with the cyan background.


What you would find with better tools are that display gamuts are frequently not as wrong as cheaper instruments indicate. Because of this people are running out of range and creating artifacts.

You should get the movie on PAL DVD, and I buy you a Belgium beer if
you can get rid of the artifacts on the man's face when he stand on the roof, as well on the cyan churgery clothes of the girl.


Ignace.
 
#390 ·
If they shot the girl without any clothes, the problem wouldn't be there and even if it were we wouldn't notice.



Many of the pro people here continually warn about the pitfalls of DIY calibrations. Calibration is part science and part art. Tou have to make small corrections and do them in the right order and at the right points. For example, calibrating at 20 IRE and forcing the point to be right on and be damned what it does to the lower IREs which the cheap meters really have no chance of measuring that accurately. Better to have a small error at say 30 to reduce the errors at lower and then trying for force corrections when the lower measurements are not accurate. Its almost like trying to force a large correction on an audio resonance point with a parametric equalizer. Far better to lower the peak a bit and raise the others a bit. You gotta be smooth and gentle if you want good results. Know the limitations of your measurement device and your correcting system.
 
#391 ·

Quote:
Originally Posted by umr /forum/post/16397900


What you would find with better tools are that display gamuts are frequently not as wrong as cheaper instruments indicate. Because of this people are running out of range and creating artifacts.

So true!
 
#392 ·

Quote:
Originally Posted by umr /forum/post/16397928


The key is to do it subtly with the colors that we are most sensitive to that differ from film the most while not not ruining the most important colors. Basically you must reverse the process used to convert the film. JVC has the best shot at this of anything I have seen.

This is worth exploring further. I assume its like an audio compressor/expander, if you know the original compression curve you can fairly accurately restore the original dynamic range. As the JVC has a larger color gamut it can support a higher than REC709 color "dynamic" range but we have to reduce it so that it matches the source.


If we knew how the colors were compressed in the color space originally, it should be possible to recover some of the original colors. After all, isn't this why JVC designed this projector with such a wide gamut? As a manufacturer, wouldn't it have made sense to set it up for just the REC709 color space so that 99.9% of users would not have a problem with oversaturation. But they chose to build this projector with an expanded gamut - so there must have been a good reason for it. Maybe the reason is so that professional calibrators like UMR can restore a more film like look to the projector, so that the expanded gamut becomes an advantage for the JVC projector rather than a disadvantage - and the basis of a well kept secret (calibrators secret sauce). If we could recover some of the film color space outside Rec709 that would be a distinct advantage of having the PJ calibrated.


I'm speculating on the above, however a good topic to explore further....


Now I have to go back to my HT to complete my DIY JVC projector calibration, where I have +- 1 dE error on grayscale except for 10% where the error is more like 12 dE too green and very obvious. Not sure how I can tackle that, and I certainly did not have this type of error when I calibrated the bulb after 50 hours (its now 150 hours). I also punched in lovingDVD's settings and possibly because of the 10dE error, the image now looks significantly worse than my original calibration effort.


Can I confirm that you want to set HCFR to ""use gamma reference to compute greyscale dE"?? Also that you want to get your target gamma set before fine tuning the grayscale? I can see some interaction between the 2.
 
#393 ·

Quote:
Originally Posted by deandob /forum/post/16400276


This is worth exploring further. I assume its like an audio compressor/expander, if you know the original compression curve you can fairly accurately restore the original dynamic range. As the JVC has a larger color gamut it can support a higher than REC709 color "dynamic" range but we have to reduce it so that it matches the source.


If we knew how the colors were compressed in the color space originally, it should be possible to recover some of the original colors. After all, isn't this why JVC designed this projector with such a wide gamut? As a manufacturer, wouldn't it have made sense to set it up for just the REC709 color space so that 99.9% of users would not have a problem with oversaturation. But they chose to build this projector with an expanded gamut - so there must have been a good reason for it. Maybe the reason is so that professional calibrators like UMR can restore a more film like look to the projector, so that the expanded gamut becomes an advantage for the JVC projector rather than a disadvantage - and the basis of a well kept secret (calibrators secret sauce). If we could recover some of the film color space outside Rec709 that would be a distinct advantage of having the PJ calibrated.


I'm speculating on the above, however a good topic to explore further....


Now I have to go back to my HT to complete my DIY JVC projector calibration, where I have +- 1 dE error on grayscale except for 10% where the error is more like 12 dE too green and very obvious. Not sure how I can tackle that, and I certainly did not have this type of error when I calibrated the bulb after 50 hours (its now 150 hours). I also punched in lovingDVD's settings and possibly because of the 10dE error, the image now looks significantly worse than my original calibration effort.


Can I confirm that you want to set HCFR to ""use gamma reference to compute greyscale dE"?? Also that you want to get your target gamma set before fine tuning the grayscale? I can see some interaction between the 2.

Restoring the original gammut is only possible if they have applied a certain non-linear curve (saturation wise) to it, if they just kept everything linear and just limited the gammut we can never reproduce the gammut. On the other hand if they applied a non-linear curve we probably are not able to reconstruct because the CMS of the JVC does work linear according saturation? Or is the last not true? I didn't measure myself yet. Then we need an external "expander" to do the job? No?


Two things that I can come up with for the choice of an extended gammut are:

- Marketing, for impressing the people with highly saturated colors, like all consumer products.

- Restore the movie extended color gammut by means of external converters, since to me it is very difficult to extend the gammut through the built in CMS without screwing up the so important skin tones?


Ignace.
 
#394 ·
When I setup my projector, I'm looking to try to reproduce the look of the original film rather than do a rigid rec 709 calibration. I have an idea of how it should look but I don't have the tools so I'm mostly bumbling in the dark. Based on what UMR has been saying I think his philosophy matches mines but with vastly better tools and skills. Pity he isn't based in the UK.
 
#395 ·

Quote:
Originally Posted by Ignace /forum/post/16400313


Restoring the original gammut is only possible if they have applied a certain non-linear curve (saturation wise) to it, if they just kept everything linear and just limited the gammut we can never reproduce the gammut. On the other hand if they applied a non-linear curve we probably are not able to reconstruct because the CMS of the JVC does work linear according saturation? Or is the last not true? I didn't measure myself yet. Then we need an external "expander" to do the job? No?


Two things that I can come up with for the choice of an extended gammut are:

- Marketing, for impressing the people with highly saturated colors, like all consumer products.

- Restore the movie extended color gammut by means of external converters, since to me it is very difficult to extend the gammut through the built in CMS without screwing up the so important skin tones?


Ignace.

I could be wrong but I think the biggest difference over rec 709 would be in the hues. Rec 709 hues are based on the limitations of legacy display technology. I see no point in limiting a display like the RS20 in this way.
 
#396 ·

Quote:
Originally Posted by umr /forum/post/16397841


Who said Rec. 709 is the goal for film? Film is not Rec. 709. The cool thing about these units is that they can restore part of the lost film gamut while not distorting key colors.


We can hit Rec. 709, but that will not restore the look to what we see in a reference theater.

Okay. I see your approach now. You are not calibrating to REC 709 or any other standard. That is fine by me so long as your clients understand that they are not getting a conventional video calibration. There can be lots of debate about how good of an idea this is. Certainly many people think it is a good idea and we have seen lots of people on these forums creating custom gamuts.


Purists will be horrified but the proof is in the pudding. I am skeptical that you can consistently "recreate" the actual colors that were used in given films but I can see how doing this might create a more pleasing experience overall.
 
#397 ·

Quote:
Originally Posted by Lawguy /forum/post/16400423


Okay. I see your approach now. You are not calibrating to REC 709 or any other standard. That is fine by me so long as your clients understand that they are not getting a conventional video calibration. There can be lots of debate about how good of an idea this is. Certainly many people think it is a good idea and we have seen lots of people on these forums creating custom gamuts.


Purists will be horrified but the proof is in the pudding. I am skeptical that you can consistently "recreate" the actual colors that were used in given films but I can see how doing this might create a more pleasing experience overall.

I give people what they want. If someone wants Rec 709 that is what I give them if possible. If they want what I like I give them that if possible. You also cannot deliver what a display cannot do. I am discussing what I personnely like here and what this product is best at. If I want Rec 709 I watch my 151FD which is upstairs.


I personally am horrified by what I see as calibrated displays in almost all cases. Little regard is given to any colors than RGBCMYW and those are hosed by using poor tools. It also appears that very few know what D65 looks like so they make things horribly wrong.
 
#398 ·
Jeff, really curious about the release of your concerns by them I could be agree with you


Infact, my aims is to achieve 2 banks, one pure REC 709 and the other custom extended gamut to "restore" some colors and get the best film-like experience pushing on the DNA of JVC machines



If you plan to come to Italy, feel free to contact me, wish to try your approach
 
#399 ·
umr

Interesting!

Can you offer calibration for both film gamut and hdtv gamut? Does RS20 have multiple memories that can handle two gamuts?

Is your film gamut white point less bright than hdtv gamut white point?


First came original aspect ratio.

Now displays are soon capable of original color gamut!
 
#400 ·

Quote:
Originally Posted by umr /forum/post/16400836


I give people what they want. If someone wants Rec 709 that is what I give them if possible. If they want what I like I give them that if possible. You also cannot deliver what a display cannot do. I am discussing what I personnely like here and what this product is best at. If I want Rec 709 I watch my 151FD which is upstairs.

Fair enough.


I agree that the accuracy of many of the meters that people use is an issue. This is probably especially true if they are have been sitting around for a while. I also think that amatuers like me are more likely to make fundamental mistakes in calibrating a display. A good calibrator will have developed good habits and procedures.


Still, color error is measured in dE. A determined hobbyist with a good meter and a good CMS can get very good results even if your better equipment produces a result that measures slightly better. If such a hobbyist and you calibrated to the same standard with different equipment, I bet the results would compare very favorably.
 
#401 ·

Quote:
Originally Posted by Ohlson /forum/post/16401191


umr

Interesting!

Can you offer calibration for both film gamut and hdtv gamut? Does RS20 have multiple memories that can handle two gamuts?

Is your film gamut white point less bright than hdtv gamut white point?


First came original aspect ratio.

Now displays are soon capable of original color gamut!

The RS20 can handle multiple gamuts.


You can make the light output what you like. I tend to like SMPTE light levels, but you could setup multiple different ones along with the gamut. It is a very flexible product assuming you are not trying to toss up too big an image.


I temporarily crank open my iris when guests are over for football and run a little light in the room.
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top