JVC Autocal Software V12 calibration for 2019 RS1000/RS2000/RS3000/NX5/NX7/NX9

[AVSMike]

No problem. I've updated the FAQ with the two minor issues I experienced today (one re the f/w update and the USB driver detection, the other with the mis-detection of my model due to the IP address reverting to my rs500).

Overall, especially if you have a USB driver that the JVC likes, the f/w update is a much nicer procedure than with older models. It takes a while, so you have to be patient, but it's straighforward.

The firmware up date time will vary, depending on what previous version you had in your projector. So if your firmware is 1.17, it will take longer to update, than if your firmware version was 1.20.

[Manni]

The firmware up date time will vary, depending on what previous version you had in your projector. So if your firmware is 1.17, it will take longer to update, than if your firmware version was 1.20.

Well, I was updating from the latest V1.20, so if the shortest possible time is that long, people with earlier versions have to be even more patient

It think it also has to do with the USB drive speed. Mine was super old (and slow) as my more recent (and faster) one wasn't detected. There might be quite a few factors at play...

[SirMaster]

Hi Manni,

Thanks for your reply over in the other forum's NX owners thread about my calibration.

So I decided to just try out a full (gamma + color) autocal on my NX5 with my Spyder5.  Doing so actually fixed my high bulb issues completely (at least to my eye).  I believe this really did take me 90% of the way there to a decently calibrated image.

I found that I could most easily see a before/after comparison by moving my manual lens aperture between -4 and -3.  At -4 I was seeing the original un-calibrated image, and then moving it to -3 the color would very noticeably change to my calibrated result.

I am still a little bit confused though on the Individual vs Reflect to All option.  As I understand, on my NX5 only Lamp mode changes and lens aperture changes (in 4 groups) will not apply the autocal result.

So what I did was created a User1 picture mode (HDR color profile, D6500 color temp, 2.2 gamma, low bulb) and then I ran the autocal on that.

Then I created another picture mode User2 (HDR color profile, D6500 color temp, 2.2 gamma, high bulb) and then I ran autocal on that.

I understand I needed to do this because of the lamp mode change.

But now I also have 2 more picture modes, User3 (rec709 color profile, D6500 color temp, 2.2 gamma, low bulb) and User4 (rec709 color profile, D6500 color temp, 2.2 gamma, high bulb).

So from what I understand, these are already calibrated for me (since they dont have any different lens aperture settings) and I can't do anything with autocal to improve them without overwriting my HDR color profile calibration.

But color mode shouldn't necessarily change the calibration results anyways right?

I only use rec709 for when I do things like play YouTube videos, since all my local content is played through madVR and I have selected "this display is calibrated to BT2020", so as I understand, madVR puts my rec709 SDR video into the BT2020 container and it should come out looking correct with my NX5 in HDR color profile for both HDR and SDR video in madVR.

I guess let me know if my assumptions here are correct or if you see sometime that I should be doing differently, thanks!

[Manni]

Hi Sirmaster,

Glad to hear that the JVC Autocal resolved your problem. Your Spyder 5 must be a good one, which is great news

Please read the calibration section (link in post 2) as it will answer most of your questions (assuming you've also read the JVC Autocal manual).

Lens aperture doesn't affect gamma calibration. It only affect color calibration. As you don't have a filter and have already calibrated high lamp and low lamp, yes gamma should be done for your SDR calibrations, and color should be done if the iris range is the same, otherwise you will have to redo them.

Reflect to all only applies to gamma, not to color.

It is possible that things work differently on an N5, please let us know if it was the case.

As I explained, we won't discuss madVR here as it's a complex subject and it would derail the thread. There are other threads I mentioned that will be a better place for support regarding this. I might open a madVR thread after the next public build is released. That would be a better place to discuss this in more details.

Others might have further comments of course, especially other N5 users.

[Manni]

I only use rec709 for when I do things like play YouTube videos, since all my local content is played through madVR and I have selected "this display is calibrated to BT2020", so as I understand, madVR puts my rec709 SDR video into the BT2020 container and it should come out looking correct with my NX5 in HDR color profile for both HDR and SDR video in madVR.

Regarding this point specifically, I misread it earlier and thought you were talking about 3D LUTs. There are two issues with this approach which isn't madVR specific:
1) Although you can ask madVR to convert everything to BT2020, it's not advised with a projector because you are not optimizing calibration for each content. Usually, you will want the SDR rec-709 calibration to hit a peak of around 50-60nits, and the HDR calibration to hit as high as you can, with black floor and contrast always to compromise with peak brightness. If you use your HDR for both, that means that you have far less contrast and a higher black floor than you would get with a separate SDR rec-709 calibration with the iris closed down to reach 50-60nits.
2) This might be the reason why you can't create a separate color calibration for each, because you're using the same iris setting. If you were setting the manual iris to a lower setting in SDR than in HDR (ideally in a different range, so if HDR is at 0 then set the iris at -4 or below, as you have observed this is a different calibration), then you would be able to run your SDR color calibration (iris at -4 or lower) separately from the HDR color calibration (iris open).

[SirMaster]

If the only reason is brightness control then I think I am OK.

My 140" 1.0 gain screen according to my light meter measurement is 19fL (65nit) in low lamp, iris 0 which is what I am using for SDR content.  Maybe a tad bright, but I like a bright image.

Then for HDR content I have madVR kick the bulb into high mode, still with iris at 0 and get 24fL and I have my madVR set to 80nit peak.

[Manni]

If the only reason is brightness control then I think I am OK.

My 140" 1.0 gain screen according to my light meter measurement is 19fL (65nit) in low lamp, iris 0 which is what I am using for SDR content.  Maybe a tad bright, but I like a bright image.

Then for HDR content I have madVR kick the bulb into high mode, still with iris at 0 and get 24fL and I have my madVR set to 80nit peak.

As explained, it's not just for brightness. With the settings above, you are sacrificing black floor (if not using the DI) and native contrast in SDR by choosing a target (19fL) that is too high. The standard is 14-16fL, and as you don't have much headroom for HDR (80nits is OK but not that much), I would personally set the peak in HDR to 12-14fL, no more than 16fL if you like it bright, so that you'll get a better black floor and native contrast in SDR, less DI artifacts if/when using the DI, and more of a difference when watching HDR. Going from 19fL to 24fL isn't a big difference, you're bringing SDR and HDR closer that way, by showing SDR too bright (assuming you have a dedicated room and no ambient light).

If you were setting SDR to reference, going from 12-14Fl to 24fL would not only be more "correct", it would also enhance the difference between the two. You'll get a better SDR (better black floor, better native contrast, less artificact if/when using the DI) and more visible improvements with HDR.

Now of course if it's about preference and you prefer a bright image at the expense of other elements, keep it (too) bright

If you only have these two calibrations, then color and gamma will be using different tables (low lamp / high lamp) so you can optimize them separately with autocal. But if you need to make a difference between two calibrations that use low lamp, using an iris setting fully open and the other at -4 or more would allow you to optimize them separately, though I'm not sure why that would be needed.

I go from 60nits post cal with the iris at -12 in SDR (-13 would be too dim at 40-45nits) and close to 48,000:1 native on/off with pretty much zero DI artifacts to 115nits with the iris fully open (both in low lamp, I could get 150nits in high lamp) in HDR. This 1:2 ratio between SDR and HDR is pretty good and both are (almost) correct/ideal (though I would take more nits in HDR if I could with low lamp, ideally around 150nits).

The main downside of HDR with iris fully open is the raised black floor and the reduced native on/off (30,000:1) as the DI isn't usable (to me) due to the too many artifacts with the iris fully open. So I'm eagerly waiting for JVC to fix the DI, hopefully in the next f/w update as they confirmed they were working on it. I certainly wouldn't use iris fully open in SDR if I didn't have to.

[SirMaster]

As explained, it's not just for brightness. With the settings above, you are sacrificing black floor (if not using the DI) and native contrast in SDR by choosing a target (19fL) that is too high. The standard is 14-16fL, and as you don't have much headroom for HDR (80nits is OK but not that much), I would personally set the peak in HDR to 12-14fL, no more than 16fL if you like it bright, so that you'll get a better black floor and native contrast in SDR, less DI artifacts if/when using the DI, and more of a difference when watching HDR. Going from 19fL to 24fL isn't a big difference, you're bringing SDR and HDR closer that way, by showing SDR too bright (assuming you have a dedicated room and no ambient light).

If you were setting SDR to reference, going from 12-14Fl to 24fL would not only be more "correct", it would also enhance the difference between the two. You'll get a better SDR (better black floor, better native contrast, less artificact if/when using the DI) and more visible improvements with HDR.

Now of course if it's about preference and you prefer a bright image at the expense of other elements, keep it (too) bright

If you only have these two calibrations, then color and gamma will be using different tables (low lamp / high lamp) so you can optimize them separately with autocal. But if you need to make a difference between two calibrations that use low lamp, using an iris setting fully open and the other at -4 or more would allow you to optimize them separately, though I'm not sure why that would be needed.

I go from 60nits post cal with the iris at -12 in SDR (-13 would be too dim at 40-45nits) and close to 48,000:1 native on/off with pretty much zero DI artifacts to 115nits with the iris fully open (both in low lamp, I could get 150nits in high lamp) in HDR. This 1:2 ratio between SDR and HDR is pretty good and both are (almost) correct/ideal (though I would take more nits in HDR if I could with low lamp, ideally around 150nits).

The main downside of HDR with iris fully open is the raised black floor and the reduced native on/off (30,000:1) as the DI isn't usable (to me) due to the too many artifacts with the iris fully open. So I'm eagerly waiting for JVC to fix the DI, hopefully in the next f/w update as they confirmed they were working on it. I certainly wouldn't use iris fully open in SDR if I didn't have to.

Well currently I am seeing DI artifacts (massive blooming and yellowing) at essentially any capped DI level.  I mean maybe if I really took my iris to -10 or less and enabled the DI, but that is just too dim for my preference for SDR (plus it's only going to get dimmer as my bulb ages) so I also kinda of find that unacceptable for me.

I will have to keep playing with the settings, but so far I have them at what I have them at because it looks best to me.  My only concern before was whether there was something wrong about making my SDR content output in BT2020 color space with madVR as opposed to switching madVR to rec709 and the projector to rec709 space when playing SDR.

I certainly don't see a difference there, but I am new to using these different color spaces so I wasn't sure whether that was an appropriate way to run things.

I mean, my HDR is already as bright as it can be at 24 fL with the bulb on high and the iris wide open.  All I can do it seems from your recommendation is to close my iris a little on SDR to bring the 19fL down to closer to 16fL.  I've played with that, but I didn't like how dim it looked to me.

But either way, I don't see myself ever using the DI as the artifacts are way too distracting to me.  I also don't personally hold much hope for JVC "fixing" it with firmware based on their prior firmware update history but we will just have to wait and see what happens.  If they do fix it then I will certainly use it.

[Murray]

Manni I believe the gamma drift change issue using the DI on auto 1/2 on the new JVCs has been corrected. I had to have my NX9 replaced v.1.80. My new NX9 installed yesterday I was cheching out for issues and the bad white titles on black during fadeout/fade-in remained steady (no iris flicker) and the white credits remained total white, no yellowing, it looks very very nice to me!

However the RGB 24 8bit/12bit bug still remains, do they just not understand this is WRONG!?><)(*&^%
I have reported it again

[DLCPhoto]

Manni I believe the gamma drift change issue using the DI on auto 1/2 on the new JVCs has been corrected. I had to have my NX9 replaced v.1.80. My new NX9 installed yesterday I was cheching out for issues and the bad white titles on black during fadeout/fade-in remained steady (no iris flicker) and the white credits remained total white, no yellowing, it looks very very nice to me!

However the RGB 24 8bit/12bit bug still remains, do they just not understand this is WRONG!?><)(*&^%
I have reported it again

Interesting on the DI!

One other criticism Manni and others have reported is that the fade-to-black seemed a bit limited relative to the prior generation of Projectors.  Does it seem like this has improved as well, along with the yellow/gamma shift problem?

And how is it on end credits - very noticeable as on the prior generation, or minimal, as in the current generation with previous firmware?

Thanks.

[Murray]

Its absolutly fine on end credits these days unlike the past models. I have no issues on my NX9 about fade to black, its fantastic, however I cant comment on the NX5 or 7 as I havent seen them.

[Manni]

Manni I believe the gamma drift change issue using the DI on auto 1/2 on the new JVCs has been corrected. I had to have my NX9 replaced v.1.80. My new NX9 installed yesterday I was cheching out for issues and the bad white titles on black during fadeout/fade-in remained steady (no iris flicker) and the white credits remained total white, no yellowing, it looks very very nice to me!

However the RGB 24 8bit/12bit bug still remains, do they just not understand this is WRONG!?><)(*&^%
I have reported it again

Hi Murray,

Thanks for reporting this.

This is great news and it sounds like a step in the right direction.

I'm preparing a report for JVC so that the other issues, especially blooming on highlights against dark background, can be resolved. It would also be great if they could close the iris further for us rs1000/rs2000 users, because the black floor is higher compared to older models. This is why there is less pumping on credits with the new DI implementation, but frankly I don't care about pumping on credits. I want better fade to black, that lasts for more than 1-2 secs. On the rs500, fade to black was good for at least 5 secs.

Did you check the MI:Fallout example I gave in my mini-review of the rs2000 (it's one amongst many)? Look at the car headlights during the scene starting at 00:05:57, with the DI on and the manual iris fully open (what most people use for HDR). Pause, then bring the menu up, and look at the original picture (what you would see without the DI, if the iris was on manual). Bring the menu down, and you'll see how the DI blooms the highlights, especially in HDR with the iris fully open. It's much less noticeable if the iris is closed to -7 or more.

Warning: only do this if you care about keeping the original detail in the picture, and overall picture quality. If you love the DI as it is, don't look for this. Once you've seen it, you'll see it everywhere. That's why I leave the DI off for now with HDR content, but I've used the DI in SDR for ages without any issue, and even in HDR before we got better tonemapping. It took me a while to notice the picture degradation on HDR highlights caused by the DI.

[Murray]

Sorry Manni I dont have that film to check.

[Manni]

Sorry Manni I dont have that film to check.

No problem. Do you have The Meg or 2001 - A Space Odyssey? I plan to list timecodes for these as well.

[Murray]

no sorry Ive just started to collect 4K