JVC Calibration Software V6&V7 For 2015 Models (X9000,X7000,X5000,RS400,RS500,RS600)
As the new version of the software only works with the new models, I decided to create a new thread. I will try to organize it better later as I'm running out of time right now.
DISCLAIMER: When you save the first calibration, the software creates two files: one is a backup file before any changes are made, the other is a your first calibration. Make sure that you save the backup file (with INIT in the name) in a safe place, as restoring that file is the only way to go back to factory settings. A factory reset in the service menu doesn't reverse the autocal changes. These changes are PERMANENT. The only way to go back is to restore a previous calibration file, and restoring the init file is the only way to go back to factory settings. I know that I say to read the manual, but if you don't and corrupt your projector, please don't complain here! This software makes some LOW LEVEL changes, everything in this thread is done under your responsibility. If you don't want to take the risk, hire a professional or don't run this software.
IMPORTANT NOTE: please don't PM me about any of this, post in the thread, there are many other contributors to this thread, you will get a faster answer and it will benefit everyone. Also, please read all the documents linked above (user manual, original thread for V5) and at least the recap posts below as well before asking a question. Chance is that one of these will contain the answer. Thanks!
At the moment (21 December 2015), the only supporter meters are the Spyder 4 Pro and Elite. A user confirmed that the Spyder TV also worked, at least for the V6 version of the software. JVC have announced they intention to support the Spyder 5, but we don't know when this will happen, or which models will be supported. [EDIT: New version 7 supports the Spyder 5 (Elite/Pro/Express), in addition to the Spyder 4 (Elite/Pro) with V6, link below updated].
Here are a few useful links:
If you're looking for the JVC Autocal Software and calibration thread for the new native JVC 4K models (RS1000/2000/3000/NX5/NX7/NX9), please google "JVC Autocal Software V11" or "JVC projectors Autocal Software V11" and click "show more results" at the bottom of the page if the page on the JVC Autocal V11 doesn't show up.
See posts 9 and 10 for HDR and SDR BT2020 settings and calibration, and this post for an updated procedure: https://www.avsforum.com/forum/24-di...l#post53734897
See posts 11 and 12 for a comparison between one Spyder 4 Pro and one Spyder 5 Pro https://www.avsforum.com/forum/24-di...l#post39729554
Info on HDR settings, official BT2020 colour profile and f/w update: http://www3.jvckenwood.com/english/d...rs400_uhd.html
JVC download page for V6 and V7 (2015 models) software & user manual NEW SOFTWARE VERSION V7 with Spyder 5 support, V6.01 still has to be used with Spyder 4: http://www3.jvckenwood.com/english/d...ationsoft.html
JVC download page for V5 (2014 models) software & user manual: http://www3.jvckenwood.com/english/d...ationsoft.html
Screen adjustment mode table: http://www3.jvckenwood.com/english/projector/screen/
Initial thread for V5 / 2014 models (lots of useful info as the software is very similar): https://www.avsforum.com/forum/24-di...libration.html
Some specific posts in that thread:
Harry's recap: https://www.avsforum.com/forum/24-di...l#post24486923
Manni's recap: https://www.avsforum.com/forum/24-di...l#post35130186
How to set-up the HD Fury Linker to disable HDR auto-detection in the JVCs and have the option to enable the dynamic iris in HDR (also without having the PJ forcing Gamma D or a specific user preset when HDR is detected)
Decoding the HDR metadata information for each title using the Linker or the Integral:
Using the HDFury Vertex to switch automatically between SDR and different HDR calibrations on 2015/2017/2018 models (initial release V1.0)
Using the HDFury Vertex to switch automatically between 3D, x.v.color, SDR, HLG and different HDR calibrations on 2015/2017/2018 models (latest release V3.1)
https://www.avsforum.com/forum/24-di...l#post55408090 (includes latest Dolby Cinema Emulation and HLG curves)
[EDIT 02-02-18: this is also where to get the latest Dolby Cinema Emulation curves, as well as a brighter and a darker set and a separate version implementing BT2390 in the low end for brighter shadow details.]
Quick How To to control the JVC Projectors from MadVR:
Other links and FAQ:
Cine4home JVC Autocal review: http://translate.google.com/translat...erung_X500.htm
What is the INIT backup file, what does it save and what are the other backup files:
JVC Autocal with Spyder vs external 3D LUT (eecolor, radiance) and i1d3: https://www.avsforum.com/forum/24-di...l#post40044458
How to check the accuracy of a Spyder followed by discussion about meter accuracy:
@Chad B's method and tools for creating a custom gamut to correct the errors of your specific Spyder unit and improve the Autocal results (needs a more accurate meter, at least an i1d3/C6, preferably an i1pro2 or better spectro):
Slightly modified Chad B method to Autocal SDR Rec-709 and HDR BT2020:
Modified Chad B method to Autocal HDR BT2020 and create custom targets for BT2020:
Resolving the mystery of the fluctuating brightness after running an Autocal, and how to optimize brightness:
Why you need 45mn warm-up time before calibrating:
How to position the meter in pictures (thanks to @Dr. Spankenstein ):
Adjusting color temp using the manual gamma controls in the PJ (3-bands) or with the JVC software (11 points): https://www.avsforum.com/forum/24-di...l#post42145897
Using the manual "old style" 11 point parametric gamma controls in the JVC software vs using the 3-band gamma controls in the projector after an autocal: https://www.avsforum.com/forum/24-di...l#post43480434
Saving the gamma curve after modification using the JVC software: https://www.avsforum.com/forum/24-di...l#post43587882
HDR GAMMA / BT2020
HDR 101: Quick Summary on the state of HDR on the JVCs by Stevenjw and Stranger89 as of Feb 21 17:
Do you need to rerun the gamma autocal for the BT2020 profile?
Do I need to rerun a gamma calibration for gamma D?
More info on the Rec-2020F/NF and Rec-709F/NF custom profiles (see below as well):
https://www.avsforum.com/forum/24-di...l#post40573498 and https://www.avsforum.com/forum/24-di...l#post45042946 as well as https://www.avsforum.com/forum/24-di...l#post39729530
PQ GAMMA Calibration
[EDIT 02-02-18: most of the information below is obsolete. The only curves of mine that I recommend using are those included in the JVC Macro Pack for the Vertex. Of course feel free to use curves provided by others if they work better for you]
HDR Calibration on the JVCs using a multiplier to scale the charts:
Creating a custom gamma calibration for PQ Gamma to replace Gamma D: https://www.avsforum.com/forum/24-di...l#post50609353
How to import a custom gamma profile:
Screenshot of layout used to calibrate a custom PQ Gamma in Calman and calibration report:
Explanation of the various options to deal with Gamma in HDR and report on what's needed from JVC (as of 02/11/17): https://www.avsforum.com/forum/24-di...l#post50666657
More details on shaping the PQ Curve and explanation of the custom gamma file name convention:
Details about the new curves in low lamp and quick calibration steps (how to create a custom curve):
Details on adjusting the roll-off
A details on highlights clipping and adjusting the roll-off
Wookii's quick start guide on using Arve's tool (master branch):
New parameters for Arve's tool and three new curves:
Quick explanation of some parameters of Arve's tool:
Zombie10K's screenshots to explain the parameters in the wip branch (thanks Jason):
Latest thoughts on latest curves and low-res version for 130-135nits peakY (800nits max brightness virtual):
Setting brightness in HDR, avoiding a raised black floor and resolving the DI mystery
Using the clipping highlights feature in Arve's tool (wip branch) (thanks @lovingdvd !)
More details on setting brightness in HDR and details regarding the requested testing
Javs' screenshot comparison SDR BT2020 vs Standard HDR vs Custom Gamma
Dolby Vision Cinema Emulation curves (peakY at 107nits for those who can target 95-115nits peak brightness) included in the Vertex JVC Macro pack (the Vertex is only necessary to switch between the curves automatically according to content, you can use the curves by themselves
Projector black floor vs content black floor in HDR
@lovingdvd 's setup guide to create a custom curve (wip branch) Thanks Ric!
Using the new bbi/bbo parameters (wip branch only) and parameters for a slightly updated 800nits curve:
Hard clipping at 4000nits vs hard clipping at 2750nits: what difference?
Changes and explaining the plot info in the threaded branch (supersedes the wip branch):
HDR High Lamp vs HDR Low Lamp calibrations and impact on color volume
Custom Gamma Curves to download:
EDIT 02-02-18: I've deleted all the old curves as they were really old. The only curves I still prove are those in the Vertex JVC Macro pack.
I'm using the jvc tool to create custom curves provided here by Arve (thanks a million Arve, you're a star!). Details in this post: https://www.avsforum.com/forum/24-di...l#post50819409
Anyone who wants proper custom curves, learn to use Arve's tool or ask a competent calibrator to do it for you.
Wookii has made a great quick start guide on using Arve's tool: https://www.avsforum.com/forum/24-di...l#post50960289
@Dominic Chan has generated manual curves follwoing BT2390. They are targeting 86nits and 130nits. Very good shadow detail.
To install the provided custom curves you need to use the import/export option from the main menu of the JVC Autocal, select gamma data (*.jgd), select the custom gamma profile you want to import the file into, click on "import", select the file you want to load and once the file is loaded, select start
To enable them, select the same custom gamma profile you imported the custom gamma file into, then select "import" as correction value. All the usual gamma settings (picture tone, dark gamma, bright gamma) are locked and greyed out. This is normal. It's a limitation of the custom gamma curves.
It is highly recommended to run a JVC Autocal (gamma+color) at the lamp/iris setting you are going to use so that the RGB balance is done for your HDR mode. This won't correct your custom gamma curve, but as the custom curves below only change white, the RGB balance remains untouched.
PLEASE DO NOT PM me about all this, I don't have the time to answer. Read all the posts linked in the section, and if you still have questions ask them in this thread (I don't participate anymore but there are lots of competent users to help) or in the JVC Owners thread for the RS500 (Which I follow when I can). Others will help there as well.
You've done your homework, read the manuals, the first posts and all the links above, but still don't get good results. Here are a few tips that might help: https://www.avsforum.com/forum/24-di...l#post46040009
POTENTIAL ISSUES WITH AUTOCAL
Potential raised levels issue near black after a JVC Autocal when targeting a power gamma curve: detailed report here https://www.avsforum.com/forum/24-di...l#post46081417
Second report and solutions for the potential raised levels near black with a power gamma calibration: https://www.avsforum.com/forum/24-di...l#post46137081
Why I stopped contributing to this thread as much as I used to, courtesy of Andreas21: https://www.avsforum.com/forum/24-di...l#post46172857
Color profiles to download:
[EDIT: note that JVC have now provided an official BT2020 profile, identical to the Rec2020F below. Check the links for the new software at the top of the post to download the official BT2020 profile].
[EDIT 12/09/17: I've attached all the color profiles in a ZIP at the end of this post as the Dropbox links are dead.]
1) Rec-709NF (the profile I've used for my first calibration below):
This is a standard rec-709 profile (with standard coordinates and without the filter), but you get much better results calibrating this custom profile once uploaded to the projector than calibrating one of the existing presets, even fairly good ones like natural.
For example, standard after a JVC autocal will still be oversaturated and won't track perfectly.
This rec-709 profile, once uploaded into a custom profile slot and after an autocal should be very close and should also track very well.
You only need to create custom colour profiles with custom targets to compensate for possible errors in the meter. Otherwise a custom profile with standard targets is all you need.
Here is a version with the filter enabled for those who would like to try to see if the filter allows them to get better black levels:
2) DCI/P3: these profiles were made using the standard coordinates provided for DCI in the JVC Software. When you import it in a custom profile slot, make sure you select the filter version if you have one on your model. It also works with a model without a filter, although of course it doesn't reach as far. [EDIT: there is no consumer content using DCI. All UHD consumer content use the BT2020 container therefore should use a BT2020 profile.]
The profile with the filter will engage the filter automatically when a user profile with this colour profile is called.
DCI-P3F (with filter):
DCI-P3NF (no filter):
On my X500, I get around 85% of P3 with the DCI-P3NF custom preset, and it tracks very well up to 75% sat.
Like the rec-709 profiles, this profile tracks better if you autocal it after importing it.
3) Rec2020: these profiles were created using standard rec-2020 coordinates, as rec2020, unlike DCI-P3, isn't an option in the software. [EDIT: there is now an official BT2020 profile from JVC, see link to HDR settings and f/w update at the top of the post. It should be identical to mine but I haven't checked yet.] [EDIT2: confirmed, they are identical]
Same, pick up the version you want to use depending on whether your model has a filter or not. You can also select a version without the filter to use in a model with the filter if you want to try to reach as far as possible without losing too much brightness.
REC-2020F (with filter):
REC-2020NF (no filter):
I haven't done extensive tests with this one but it seems to be working fine.
Like the rec-709 profiles, this profile tracks better if you autocal it after importing it.
I'll put here a detailed and up-to-date guide for V6 when I find the time, meanwhile here is a quick and dirty updated version of my post for the V5:
[EDIT 21/10/18: there is an updated/simplified method in post #10, make sure you read it before deciding which one works best for you]
This is my updated instructions for use with the V6 of the JVC Autocal software. It assumes you've fully read the user manual and this thread entirely and have installed the drivers and software. Make sure no Spyder software (including their little taskbar app) is running or it will conflict with the Autocal. I tested V6 with an X7000 (RS500) and my golden Spyder 4 Pro, from a Window 10 x64 Pro VM in Parallels 11 on my MBP 13 late 2012 running the latest El Capitan (would work fine in Bootcamp too):
Main changes from V5:
- 12 points gamma manual control are accessible in the software, it's the third option from the left in the main menu
- No environment menu anymore, just a screen adjust setting in the installation menu
- Other than that, it's basically the same software.
- It seems a bit touchy while doing long communication processing, like uploading a profile or saving a gamma calibration. I advise not to do anything else during these operations.
1) A Datacolor Elite isn't necessary and can be as good, as bad or worse as any Pro unit. The only difference is software, and as we don't need the Datacolor software, it really doesn't matter. If you get a bad pro (when checking with a more reliable, ideally reference meter and another software like Calman or CP), just send it back for a replacement and get another one until you get a good one. I tried to get an Elite and it measured worse than the Pro. So save your $$$. The current (2015) JVC models still need the Spyder 4 Pro, Elite or TV. The Spyder 5 should be supported in the near future.
2) (didn't re-check this as they were already installed). After you've installed the Datacolor software to get the drivers for the Spyder meter, make sure the Datacolor utility isn't loaded in memory, it interacts with the JVC software. You only need to install the Datacolor software for the drivers. You don't even need to activate the software. Also make sure you don't have any IP remote running on an iPhone/iPad/android like iRule, as it will interfere too. Make sure for example that iRule isn't loaded in any of the devices connected to the network.
3) Make sure that screen adjust settings in the JVC installation menu is disabled. The JVC Autocal doesn't work if it is enabled. You can re-enable them after the autocal to take your screen into account (TBC).
4) Once the JVC software is running, especially once you've entered the calibration section, only use the software to control the PJ. If you use the remote, the changes you make with the remote won't be taken into account by the software which will lead to issues.
5) Network: switch eco mode off while calibrating. You can re-enable it later if you don't use any LAN remote like iRule, otherwise keep it off. For the same reason, make sure nothing else can access the JVC on the network during the calibration. For example, make sure iRule isn't running on your iPad or your phone...
1) Unless you only want to autocal one user mode with specific settings, I suggest you set the iris fully open in the JVC and when positioning the meter, make sure that you put it as close to the PJ as possible, without getting outside of the "safety rectangle" shown by the software before calibration. That way you'll make dark readings easier, and you won't saturate the meter. The shadow of the meter should roughly in the center of the screen.
1) It looks like the standard colour profile is quite oversaturated, so it should be a good starting point for a 3D LUT. Cinema 2 (which engages the filter) tracks best and seems to cover P3. Reference is the widest gamut (bar profile off, which is the native gamut but disables some features like Clear Black) and seems to try to get as close as possible to rec2020, but it doesn't track as linearily as cinema2.
Custom profiles creation (optional, Newbies you definitely want to skip this step!)
If you want to create a custom gamut with the JVC software, for example a rec-709 one, a DCI/P3 one and/or a rec-2020 one (you will need to create one with custom coordinates as they are not in the software, see here for the rec-2020 details, you need to enter X,Y for each primary in the CIE xy chromaticity diagram, so the one to the left of each table http://www.tftcentral.co.uk/articles...#_Toc379132064), do so in the create profile option (third from the left on the main menu). This is where you'll need to adjust the coordinates to compensate for the error in the Spyder method, but I won't detail this here as it's quite advanced. Save the files in a place where you'll be able to find them again. Those able to do it will understand what I'm saying, other should just ignore this and go to the next step, or re-read the thread. Then import the custom profiles into a custom gamut location on the PJ, six of them are available (import/export option, 4th in the main menu when the PJ is connected, select color profile and the name of the file you saved when you created it), but note that this custom gamut can only be calibrated for color (gamut). Any gamma and color temp calibration will be ignored, and the autocal will show a "-" next to gamma or color temp during the calibration to make this clear.
I created six in total: first, rec-709 (HDTV preset), P3 (DCI preset, with the filter) and rec-2020 (custom targets, with the filter). It can take up to 7-8mn per profile (on a gigabit network) to upload these to the PJ, so be patient.
Then I created three optional ones, one for each gamut: P3 without the filter and rec-2020 without the filter, so that those without a filter in their PJ (X5000/RS400) can still see what they get, and those with a filter in their PJ (X7000/9000 and RS 500/600) but who want to privildge brightness over colour accuracy can also go that way. And finally I created a rec-709 with the filter, so that you can try to get a rec-709 that tracks well with the filter in place (for exemple to get better black levels in rec-709 if you have brightness to spare).
I haven't tested yet which gives better results for which kind of content, but they are all here. My plan is to use rec2020 with the filter for all my content with MadVR, that way MadVR can switch to the correct 3D LUT automatically depending on the content being played without having to control the projector, and I need to profile the widest gamut initially (so rec2020 with the filter). Then the filter will remain in place for the P3, Rec709, PAL and NTSC 3D LUT which MadVR will select automatically. I have the brightness to spare to do this. For all my non-MadVR content, I'll use a rec-709 profile without the filter, so that I get my brightness (and possibly my linearity without a 3D LUT) back.
If you want to fully calibrate the projector, here is the process I recommend:
PHASE 1: Gamma calibration (and color calibration for first iris/CMD position), total time around 20mn:
0) In the settings (last option in the main menu), select gamma + color. Select how detailed you want the gamma calibration to be. I select High for 33 steps, as this only needs to be run ONCE (per calibration, i.e. factory or user mode, calibrating all the range for the whole projector would take literally DAYS!).
0a) Go to calibrate, the first option in the main menu
1) Select a standard color profile, like natural, whichever you plan to use/want to calibrate. If you use user modes, this is the profile you will select in your user mode. But to calibrate, select it directly, don't select a user mode or a custom gamut or gamma and/or color temp calibration will be skipped and the software will show a "-" next to the skipped steps to indicate this. Make sure you set the iris to manual, autocal won't work otherwise. I also recommend to switch all processing off during calibration, so things like Clear Black and Motion Enhance.
2) Select gamma normal (all the gamma presets are calibrated, so better take the most standard one). [EDIT: including gamma D for those who are not sure about the meaning of the word ALL. More info on HDR calibration in posts 9 and 10].
3) Select 6500K color temp (all the color temp presets are calibrated, so better select the most standard one).
4) Check that the iris is fully open (best position to calibrate gamma) and that CMD is off (you'll have to calibrate with it on later). Set filter to off as well unless you want to calibrate a DCI/P3/Rec2020 custom colour profile.
5) Run the calibration. It should take around 10mn if you have selected 33 steps for gamma, it's faster otherwise.
6) Once it's done, check that there is no "-" after either the gamma, color temp or gamut. If a "-" is present, this means this part of the calibration has been skipped and won't be corrected. This usually happens if you have selected a custom color profile or if the environment settings are enabled in the 2014 models. Check that your gamma line is a straight line, that the color temp points are near D65. Gamut won't be properly corrected for most standard profiles, which is why you'll probably need to create a custom colour profile and calibrate it later. So you can ignore if gamut looks oversaturated, which is a good thing if you want to use a 3D LUT aftwerwards. Most important at this stage is the gamma correction: the "after" line should be straight.
7) Save the calibration. This will take a good 5-10mn, so be patient. Make sure that you save the backup file (with INIT in the name) in a safe place, as restoring that file is the only way to go back to factory settings. A factory reset in the service menu Doesn't reverse the autocal changes. These changes are PERMANENT. The only way to go back is to restore a previous calibration file, and restoring the init file is the only way to go back to factory settings.
PHASE 2: Calibration of the various iris and CMD settings (total time less than 5mn per mode/setting)
Because the JVC Autocal only corrects one iris position range and CMD setting at a time, you need to either calibrate the settings you use, or be systematic and calibrate the whole range of AP value, with and without CMD if you use CMD.
0) Go to settings and select color. Gamma only needs to be run once, and it takes much longer when it's selected.
1) Go back to calibrate, and as we selected iris fully open (0) and no CMD, go down the range without CMD and back up the range with CMD. The effective ranges are slightly different but I use 0, -5, -10 and -15 as they are easy to remember and cover all the positions that have to be calibrated. If you don't do that, when you change the iris settings the greyscale will be wrong. Alternatively, you can calibrate only the iris setting you plan to use.
2) As iris open (0) / no CMD was done in PHASE 1, I then select the following settings for Lens AP:
-5 and calibrate (the process is much faster as only colour is selected, it's the gamma that takes the longer, especially if 33 steps/high was selected). Then I save.
-10 (calibrate) (save)
-15 (calibrate) (save)
3) I then switch to CMD On, calibrate and save (without changing Lens AP)
4) Then I go back up the lens AP settings, keeping CMD On:
-10 (calibrate) (save)
PHASE 3: color calibration for custom color profiles
Once you've done this, you have calibrated the standard color profile you had selected (say natural for rec-709 or cinema for DCI-P3) for all lens AP settings, with or without CMD, for all gamma presets and all color temp presets.
Also, user modes using this standard colour profile will be calibrated too.
So, if you select user 1 with standard, a custom gamma using the 2.4 preset and a custom color temp using the 7000K preset, it will be calibrated.
However, what you need to do is run a color only calibration, just once, for each standard or custom color profile you use.
I suggest you select the lens AP and CMD setting you are most likely to use for extra accuracy with these settings, but you only need to do this once.
For example, say you have created a rec709 custom color profile (or imported my rec-709NF for No Filter) and uploaded it to custom1, you simply have to select custom1 as the color profile and run the calibration.
If, in a user2 preset, you select custom1 colour profile, gamma 2.6 and color temp 6500K, it should be calibrated already due to the previous steps.
From this baseline, I get near reference calibration for all my sources (all dEs under 2.8 with colorchecker SG, average dE under 1.5).
I use a BT1886 target gamma and only have to change the dark gamma control (2 notches) to get a near perfect BT1886 gamma curve.
NEW: the 12-point gamma controls that used to be in the projector menus until the 2014 models are now back in the JVC Software (V6 only). It's nice to see them back, it can help fine-tuning a BT-1886 curve a bit further for example, but they are mostly unnecessary once you've done a gamma autocal unless you are not aiming for a standard curve, have done something wrong with the autocal, or are more OCD than me about these things. EDIT: in fact the gamma results for V6 are not as good as what I used to get with V5. They are still very good, but the 12points control might get handy if I can't improve on that.
If I then run a lightning LUT for MadVR with Calman, in 5-10 minutes (it only needs 101 points) I get a reference calibration for my HTPC with a max dE2000 under 1.3 and an average dE2000 under 0.5 for colorchecker SG. I double checked with a custom Colorchecker XXL with more than 1000 points and the whole cube is at reference level, same as with my X500. With a 105 points / 5-10mn LUT, instead of the 17x17x17 (4913 points / 2-4 hours) that I needed with my old X30 to get results not as good.
I hope this helps. Please let me know if I got any of the former is wrong or if you experience a different behaviour, but that's what I observed when I used the Autocal V6 with the X7000/RS500.
The most important point is that trying to select a custom gamut would not allow to correct gamma. A standard color profile has to be selected for gamma to be corrected.
Will revise and clarify when I have more time....
[EDIT: added HDR and SDR BT2020 settings and calibration info in post 9 and 10.]
Results from JVC Autocal using Natural, colour profile standard, color temp 6500K, iris fully open, CMD off, filter off. This is basically what it should look like if everything goes well...
As you can see, the standard color profile, even after the Autocal, is still oversaturated. It's good if you use a 3D LUT calibration afterwards, because it means it's fully correctable, but otherwise we can get closer, see next post.
And this is what you get if you import a custom rec-709 profile (see first post for download link) with standard coordinates and run an autocal (color only) with it. Much better!
I also attach the results for DCI-P3 following the same method, which should also track better than the cinema2 preset.
Now time to do some verification with Calman Ultimate using my Discus trained to my i1pro2. The only change I've made is I've set 100% white to D65.
This is exactly what you get from the autocal, no external correction yet.
Gamma isn't as good as it used to be with V5, but it might be improvable.
Gamut and colorchecker SG are similar.
This is now what I get if I apply a Lightning LUT . This is a very small LUT (101 points only). It's all that is necessary as the Autocal gives a very good baseline. It's very fast: it takes about 5-10mn with the Discus in fast mode and around 15-20mn in high average, long integration mode. This is what I used here to profile the display and generate the LUT, I went back to fast mode to measure:
As you can see, hard to get it any better than that (for a visible difference).
Forgot to give you a baseline, which is what the natural profile looked like OOTB before any calibration (just 100% white set to D65, gamma 2.3 and 6500K selected):
I also wanted to check that we still needed to run a new color autocal if the iris setting changes into a different range or if we apply CMD, and unfortunately we still do.
The next screenshot show the near perfect gamma after the JVC Autocal and the MadVR LUT (as a baseline), done with the iris at -12, to use as a baseline as it makes it easier to see variations from this.
Then I move the iris to fully open, and the greyscale is completely wrong (I had only calibrated the -12 setting to test this).
Then I enabled CMD (on low) and it changes again.
Then I moved the iris back down to -12, the setting that had been autocaled, but left CMD on low, still wrong.
I then disabled CMD and enabled Motion Enhance (on low). It's slightly different, but the difference is much less large than with CMD, so probably fine to use it that way.
I then disabled Motion Enhance to go back to the settings that had been autocaled (iris -12, no CMD, no Motion Enhance) for verification, and all is back to near perfect.
So overall, if you change the iris settings or the CMD mode, you need to autocal the new settings. If you just change the Motion Enhance, you're probably fine.
BT2020 and HDR calibration with V6 1.01 and new f/w
I've used the Autocal to calibrate my new RS500 (X7000) and have a few interesting findings to report.
I don't have the time to rewrite the instructions, so I'm going to concentrate on the differences / new points and assume the reader is already familiar with the software and existing guide.
I don't have much time so this will be a bit random, but hopefully better than nothing.
First of all, I've noticed that when you first import a profile which uses the filter, the filter isn't enabled until you power cycle the PJ. You have to switch it off (just once) after the import to make sure that the filter is enabled when you switch to the profile.
I imported the following profiles:
Rec-709NF (no filter) in custom1 for HDTV and bluray
BT.2020 (JVCs version, with filter) in Custom2 for UHD Bluray
DCI/P3F (filter) in custom3 (optional, for experiments)
BT2020F in Custom4 (with filter, only to compare with the JVC BT.2020)
BT2020NF (no filter, to see how much gamut we lose and how much we brightness we gain when we try to get BT2020 without the filter and its associated brightness loss.
Then I've measured the profiles. Reference isn't BT2020, so shouldn't be used except for the autocal (more later).
JVC's BT.2020 and my BT2020F are identical. This makes sense, because I created the BT2020 profiles using the standard BT2020 targets, so it would have been surprising if they had used non-standard ones.
Bottom line is, if you're using the BT2020F profile, it should look identical to the JVC BT2020.
If you're not using it, use the JVC one, for peace of mind.
Regarding brightness loss with the filter, it's minimal but so is benefit on the gamut. The filter as expected has an effect mostly on cyan and green. The brightness loss in high lamp with the iris fully open is around 13%.
With the filter, the PJ covers 70.54% of BT2020 (or 97.78% of DCI-P3). Peak white is 208.22 nits.
Without the filter, the PJ covers 66.76% of BT2020 (or 91.96% of DCI-P3). Peak white is 239.54 nits.
So using the BT2020NF gives you around 15% more brightness and you lose a bit of cyan and green cover.
If you're desperate for more brightness, it's a good trade-off.
However, assuming you use the same lamp/iris settings, it also raises the black levels.
So if you don't need more brightness, I suggest to stick to the BT.2020 profile.
User settings, new Autocal method and calibration results
Then I've created the following user modes, with the following settings for my puny Carada BW 88" diag 16/9:
NOTE: all measurements taken with my Discus trained to my i1pro2. Contrast numbers are approximate but fairly reliable, I haven't had the time - or the inclination - to get the Minotla T-10 out.
SDR Rec-709 in user1 (for HDTV and bluray)
Colour profile: Rec-709NF (you can use standard otherwise)
Color Temp: User1, with the 6500K preset
Gamma: User1, with the 2.4 preset and +2 on dark gamma for BT1886
Iris: -14 to get 47nits (-13 jumped to more than 70nits!) and around 100000:1 native on/off calibrated to D65
HDR BT2020 in user2 (for UHD Bluray)
Colour profile: BT.2020 (JVC's)
Color Temp: User2, with the 6500K preset
Gamma: User2, with the 2.4 preset and +2 on dark gamma for BT1886 (selected by default for non HDR UHD Blurays) and with the Gamma D preset with +10 on picture tone, +5 on dark gamma and +5 on bright gamma. Set these two gamma presets so that the correct one will be applied depending on the way the disc is mastered. All current UHD Blurays use HDR so Gamma D should be selected automatically, but if you set up the PJ in SDR mode you need to select the gamma that will be applied in SDR.
Iris: fully open to get 240nits [EDIT: I now use high lamp and iris set to -7 to get better on/off and black levels in HDR, this gives me 150nits peak white and I use the Sony patterns to clip content at 1500nits, see the main JVC thread for more details on this step as it's not Autocal related].
I'm still toying with the best way to play HDR, I might settle on low lamp, or high lamp with a closer iris to get more native contrast, will decide later...
I then created the following (optional) profiles:
3D in user3 (for 3D blurays)
Colour profile: Rec-709NF (you can use standard otherwise)
Color Temp: User3, with the 6500K preset and appropriate gains
Gamma: User3, with the 2.3 preset and +2 on dark gamma for BT1886
Iris: fully open
Min4K in user4 for Mastered in 4K blurays
Colour profile: x.v.color
Color Temp: n/a
SDR DCI-P3 in User5 for DCI-P3 experiments
Colour profile: DCI-P3F (with filter)
Color Temp: User2, with the 6500K preset
Gamma: User2, with the 2.4 preset and +2 on dark gamma for BT1886
Iris: fully open to get 240nits
SDR BT2020 (to play UHD Blurays in SDR with the wide colour gamut using the HD Fury Integral to strip the HDR capability from the display)
Colour profile: BT.2020 (JVC's)
Color Temp: User2, with the 6500K preset
Gamma: User2, with the 2.4 preset and +2 on dark gamma for BT1886
Iris: -14 to get 47nits
If you're using an external VP to calibrate, it's probably better to use standard for rec-709 as it's nicely oversaturated so should be fully correctable, but if you want a more accurate internal calibration I highly recommend you use the REC-709NF profile. It's more accurate and tracks near perfection after the Autocal. All the profiles (including BT.2020) benefit from an autocal.
Regarding the autocal itself, I have done a lot of tests and I have reached the following conclusions:
After a gamma+color autocal in SDR Rec-709 iris -14 low lamp (no filter, no CMD), I measured the differences at 100% white with different iris settings, and with/without CMD.
While there is a difference as you open the iris and change range, the biggest one is between -14 and -13 (big jump). Same with CMD. When enabled, the calibration isn't as accurate, but it's not the end of the world.
On the other hand, you do have to run a new gamma calibration for each lamp mode, and for filter / no filter.
So my suggestion, unless you want to spend ages every 200 hours recalibrating everything, is to set the user modes you want to calibrate at the right iris, CMD and lamp settings, and only calibrate once per user mode.
Then as and when you need to open the iris further (especially if you jump from -14 to -13), run just one gamma autocal at the new setting.
So here is what my new procedure is:
I select my SDR Rec-709 user mode, at the iris, lamp and CMD setting (so -14, low, no CMD) I plan to use over the next few hundred hours, and I select the standard colour profile (otherwise gamma isn't calibrated). I run a gamma (33 steps) + colour calibration.
Once this is done, I select my custom colour profile and run a color calibration, once.
I then select Natural user mode for sports (-13, low, CMD low), select the standard colour profile, and I run the gamma+color autocal, once. This is needed because I use CMD in this mode (HDTV) and not in my SDR rec-709 profile (bluray). If I didn't use CMD, a colour only would do (same lamp mode, same filter mode).
I then select HDR BT2020, select the reference profile (needs to be a standard profile for gamma to work and a profile that uses the filter for accuracy) and run a gamma+color calibration with a power gamma preset (normal or 2.4). This is needed because I use a different lamp mode (high) and filter mode (wide). I then select BT.2020 profile and run a colour only. [EDIT: it's not possible to calibrate Gamma D as it would need the projector to be switched to HDR mode / use HDR patterns. ALL the gamma presets, including Gamma D, are calibrated when you run the gamma autocal. This is why you set a standard power gamma when you run the gamma autocal in your HDR user mode with the reference profile selected, and select gamma D manually just once while playing HDR content. That way, when SDR content is played and your HDR user mode is selected, the correct gamma curve (not gamma D) will be selected. When playing HDR content, the correct Gamma D will automatically be selected, and both will be calibrated.]
I then select SDR BT2020, select the reference profile (same reason) and run another gamma+color calibration. This is needed because I use a different lamp mode from HDR BT2020 (low lamp) and filter mode from SDR Rec-709 (wide).
I find this faster and more efficient than having to calibrate for each range, in each mode, but this is still needed if you want to be able to change each setting and still have the most accurate calibration.
When I need to open the iris in any mode, I'll just re-run a gamma+color calibration for that mode.
There are just too many modes now to do it differently unless you want to spend more time calibrating than watching movies :)
One last point: I tried the environment settings in the Autocal which suggests gamma corrections after having measured the residual light with the meter pointing slightly outside the screen, and the results, in my bat loft, were absolute rubbish, as expected. So if anyone managed to get any improvements from this, please post a detailed method and before/after measurements. As far as I'm concerned, I would not recommend using them in a fully dedicated, light controlled room. In my case, it simply ruins the gamma calibration, when 2 clicks on dark gamma is all I need to get a near perfect BT1886 from the 2.4 preset after the gamma autocal.
IMPORTANT NOTE: when looking at the BT2020 gamut linearity screenshots before and after autocal, keep in mind that current content doesn't go wider than P3, which is about 70% of BT2020. There is no content displayed outside of 70% of BT2020. So as long as the saturation tracks correctly up to around 70% (which is the case for the most part), the rest can be ignored and no content is (currently) displayed there. If/when this happens, content outside of P3 will be clipped unless the player or VP does some gamut mapping to avoid clipping and display the wider gamut within the native gamut.
Comparison and measurements Spyder 4 Pro and Spyder 5 Pro
I've received a Spyder 5 Pro and decided to compare the results between my "golden" Spyder 4 Pro and the new Spyder 5 Pro, before letting the Spyder 4 Pro go.
Lamp on the X7000 has 50 hours on.
All measurements in Calman done with my Discus trained to my i1pro2. PJ warmed-up and all meters plugged-in at least an hour before the first measurements, stable temp in the room.
Talking of which, the Spyder 5 isn't supported by Calman (yet). It's sad because Lightspace, Chromapure and HCFR/Argyll CMS all support both the Chroma 4 and 5. Not a big deal for me as I wouldn't use it to calibrate with Calman and I have Lighspace if necessary anyway, but for someone would would like to buy just one meter and be able to use it with the JVC autocal and another calibration software, it will have to be the Spyder 4 for Calman, or if it's the Spyder 5 it will have to be another software until Calman supports it.
I attach the pre-calibration measurements (I have done an Autocal not long ago, so it's not the same as OOTB) for reference.
Then I attach the results of a Gamma + Color Autocal of the standard profile with my normal setting (low lamp, iris -14 which gives me 50nits and 120000:1 native on/off) and my custom rec-709 colour profile (loaded in custom1) with the Spyder 4.
I attach the results of the JVC Autocal as seen by the software and the Spyder, then I measured gamma and gamut linearity with Calman Ultimate and my Discus trained to the i1pro2 for verification.
As you can see, near reference results from the Spyder 4 Pro. All dEs under 2, average dE under 1. I would get a few odd points if I checked the full cube, but I have already done that in the past and when it tracks this well for saturations and luminance, it's very, very good. All the measurements are done without any color temp, gamut, gamma correction post gamma autocal, except my usual +2 on dark gamma to get close to BT-1886 from the 2.4 preset.
These results could be improved slightly if setting 100% white to D65 exactly and possibly fine-tuning the gamma and color temp settings a bit, but I usually do a MadVR 3D LUT on top of this so I didn't waste time with this as I knew I would delete these with the Spyder 5 Pro Autocal and measurements. You can always look at my other attempts if you want that kind of information. Here I just wanted to show how good the calibration can be with a good spyder right after the Autocal, without any manual tuning or 3D LUT.
Then I tried one more time to use the menu that suggests environment settings, following to the letter JVC's instructions. They are completely wrong, at least in my environment (bat loft, all walls, ceiling, carpet entirely black, no ambiant light at all). I need zero adjustment after the Autocal with the Spyder 4 to get near-reference results (just +2 on dark gamma to get BT1886 instead of 2.4), and using the suggested environment settings (+2, -2, -7, 0) completely destroys gamma, as expected (see the Calman measurements after the "recommended" settings are applied, it's scary). Those who keep claiming that the environment settings are working great, please post a detailed how-to and before / after measurements to show how they improve things. Also please give us full details on the environment in which they work so well.
In a dedicated room, I would highly recommend against using them, at least until we get conclusive data showing in which configuration they improve things.
(see next post for Spyder 5 results)
Then I did exactly the same thing with the brand new Spyder 5 Pro (using the V7.0 of the software, I used V6.01 with the Spyder 4 Pro).
It's not completely awful (I've seen worse), but it's poor compared to my golden Spyder 4 Pro
The blue filter is completely off which has a detrimental effect on both gamma (poor tracking of blue) and gamut (undersaturated blue).
The second issue could be dealt with with a custom colour profite using custom targets, but there is no solution for the first one, as using another color temp preset wouldn't solve this. When settings 100% white to D65, you get semi-decent results, but you lose both in brightness (53nits before setting 100% to D65, 47nits afterwards as you need around -20 on both red and green gains to balance the RGB, vs the Spyder4 giving 50nits right away with a near perfect RGB balance at 100% and zero adjustment on the gains) and in contrast.
This is not acceptable to me, so the Spyder 5 Pro is going back and I'm trying another one tomorrow. I'm picky, many people would be happy with this (especially if they can't compare to reference or near reference and don't care about maximizing brightness, contrast and gamut coverage). Not sure if this would be considered defective by Datacolor, probably not, I guess it's within their specs, but for me it's not acceptable to end up with such a poor greyscale and an undersaturated gamut.
I'm used to getting near-reference results from my Spyder 4 Pro which means I don't really need a 3D LUT on top, I'm not going for an upgrade which is, in fact a downgrade, especially when we have no solution yet to calibrate HDR on projectors. I need as good as possible out of the autocal, at least until we can calibrate HDR10 properly. Hopefully the second Spyder 5 will be better.
Although a sample of one isn't representative (it could be a fluke), this just shows what a lottery it still is to get a good Spyder. In my experience, there is no such variance with the i1display pro, which is more expensive but is probably worth it if you want an entry-level meter with some OOTB consistency. Sadly, not an option for the JVC Autocal software.
These poor results with the Spyder 5 aren't the JVC software's fault by the way, I tried both the Spyder 4 and the Spyder 5 a few times, and the results were consistent: near-reference for my good old Spyder 4 Pro, not so great for this brand new Spyder 5 Pro.
Just to clarify, this comparison is only valid for this specific Spyder 4 (arguably a golden unit, which is why I protected it from day one in a peli-case with dessicant, and it hasn't drifted one bit) and this specific Spyder 5 (not a very good one). This doesn't mean anything about each model. You could buy a Spyder 4 which will be worse than this specific 5 (easily, the Spyder 4 Elite I tried was worse than this Spyder 5 Pro), and you could buy a 5 better than my 4 (more difficult, but at least as good is I'm sure possible).
I think any Spyder will get you 80% there. But if you want as much of the remaining 20%, better find a way to check them and get them replaced until you get a good one.
By the way, I also tried the environment settings with V7 and the Spyder 5 Pro, it gave exactly the same results: same suggested settings, same awful results.
I also read claims that the Spyder 5 was much faster than the Spyder 4, but I wasn't able to reproduce this in my environment. Gamma+Color autocal took around 8mn, color only took around 2-3mn with both meters. If any one gets faster results with the Spyder 5 compared to the Spyder 4, please could you post detailed timings for gamma+color and color? Thanks!
This will be invaluable once I get my RS500.
Awesome info, Manni! I really appreciate all of the work you've put into this over time. I look forward to diving into it soon!
Wow, Manni, that's some impressive work. Thank you for making this seem less intimidating. One question: prior to commencing the JVC auto-cal procedure, should we still do the basic brightness and contrast settings (like when setting up for a chromapure auto-cal)?
Yes you should, but I don't think it's taken into account, like if you have some settings already in custom color temps etc.
^^^^^^ thank you Manni, that's good to know --- and interesting.
Okay, finished posting most of the results, and I've also posted a link to the standard rec-709 colour profile I've used to get the results above.
Note that while this is a profile to upload into one of the custom colour profiles slots, it uses standard targets.
However, it gets much better results once autocaled that using one of the standard presets, even after an autocal.
See first post for a link and more info.
Amazing Manni! Fantastic work and contribution!
Manni, you are a most honorable and valued member of the community. Thanks for putting this together!
I went to the JVC download page, clicked on the 'Software license agreement' and entered the serial # of my Spyder4 Pro and I keep getting this message and can't get any farther. "Please accept the Software License Agreement and enter the serial number of the unit you use"
Anyone have any ideas or am I doing something wrong.
I THINK it wants your JVC pj serial number.
Yes, I am sure that's correct. Thanks. That was a stupid question :).
this is important folks !! :)
Make sure that you save the backup file (with INIT in the name) in a safe place, as restoring that file is the only way to go back to factory settings. A factory reset in the service menu Doesn't reverse the autocal changes. These changes are PERMANENT. The only way to go back is to restore a previous calibration file, and restoring the init file is the only way to go back to factory settings.
Manni, thank you for spending the time to post this info.
A couple of questions:
1) Where is the serial number on the RS500? I have just installed mine in a projector alcove and spent time adjusting its height, zoom, focus etc. and it will be a PITA to bring it down again just to read the serial number.
2) Is there any performance difference between the Spyder 4 and 5? Apart from the marketing blub of 55% low luminance accuracy (which should help), is it worth waiting for the JVC software to be updated for Spyder 5 support? Seems that way.
I have a few other questions but I'll read the older thread first.
I've just received an spyder 4 that I got from ebay. After reading's Manni's post, it seems that if you get for some reason a defective meter and you autocal then you just blew it. I guess it is just safer to use calman 5 and a c6 meter and create different profiles. I use your initial settings and did a two point grayscale calibration and got similar results. The unit tracks very well gray and has a great image. The CMS is not great. Green was a bit under saturated but managed to lower the deltae from 3.1 to 2.8.
So few questions from a pj noob. Running RS400.
1) So from the sounds of it, Spyder5 isn't yet supported. I haven't searched eBay yet, but it looks like the 4s are out of stock on Amazon etc.
Any time frame or guess for 5 support?
2) Also, I'm a mac user. Short of a win vm, I'm guessing I'm SOL on running the actual JVC SW...Or borrow or buy a cheap Win machine. True? Has JVC typically not supported Mac?
3) This one will likely draw a few laughs. How necessary is the calibration? After spending tens of thousands on a dedicated HT, it might seem insane not to do so. But how much is gained from running an in-depth calibration vs a more manual Spears&Munsil type manual adjustment? Am I missing more to the story/benefits here? I suppose that one is like someone asking me if they really need to run REQ on their prepro.....
Thanks for the patience guys! I know more about audio and LCDs etc. Trying to come up to speed on my RS400 investment.
1) Serial number is at the back of the PJ and also on the box, on one of the labels. It's an 8 digit number starting with 150 or 160.
2) Read the threads :)
I added a disclaimer in the first post, in bold :)
Fantastic results Manni. Bravo!
|All times are GMT -7. The time now is 10:27 PM.|
Powered by vBulletin® Copyright ©2000 - 2019, Jelsoft Enterprises Ltd.
vBulletin Security provided by vBSecurity (Pro) - vBulletin Mods & Addons Copyright © 2019 DragonByte Technologies Ltd.
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2019 DragonByte Technologies Ltd.
User Alert System provided by Advanced User Tagging (Pro) - vBulletin Mods & Addons Copyright © 2019 DragonByte Technologies Ltd.