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

[Amon]

Re the questions above:

1) You can't run an autocal of the Frame Adapt DTM picture mode, but if you calibrate any mode using the same gamut (not sure if they use DCI-P3 or BT2020, but it might be DCI-P3 which would explain why your previous calibration didn't translate as you most likely used BT2020) and iris, lamp and filter mode. I would try to calibrate a user mode with DCI-P3, iris open, filter and lamp set as in the DTM color profile, and it should automatically calibrate DTM.

It would be nice to get official confirmation that calibration settings carry over. The documentation is not 100% clear: 'Calibration does not work in Frame Adapt HDR mode. Set to another Picture Mode and calibrate using the same Aperture, Lamp Power and Filter settings.' I assume you are reading this as it does carry over but user can't select the new picture mode in the software, but I still want explicit confirmation. It appears we are doing black box guesswork here based on one line in the documentation, no offense. I used 2020 on the original calibration. Why would selecting a different color space affect color calibration for the same iris open, filter on, high lamp? Won't it overwrite the old calibration for high lamp, open iris, filter on? 2020 contains P3, why would this change anything? Are you thinking P3 calibration more accurately reflects the projector's capabilities(is P3 the native gamut?) without relying on transformation math, so it may be a better fit for this new mode? I'm not following your logic here. I don't know how to pass the P3 inside the 2020 container in madVR MPC-BE, so if you have any links handy that would help. I'm not doing any tone mapping in MadVR, HDR is passed through.

3 and 4 sound like you have the wrong levels set in your source or in the PJ. There is no way you'd need that much adjustment on brightness, so I would correct this first.

It doesn't make sense that the new mode requires a brightness change to +16 and the old HDR mode with same iris open, filter on, high lamp would be set to zero. I will confirm that the old mode doesn't clip with brightness=0. If the input is the same, how can the output be different?
Regarding SDR calibration having brightness set to -50, that was definitely not a manual change. It was originally set to 0 before the update and before my color temp adjustment. There's probably a bug here.

5) Switch the DI off in HDR (especially with the iris fully open) if you want to keep highlights or if you want to avoid yellowing issues, unless you have an rs3000. The DI blooms highlights and this is visible in any title where you have bright objects against a dark background, so there are many scenes in 2001 where this is visible. If you still have the issue with the DI off, then try to see if selecting low in the DTM option helps. There is no blooming of highlights here (RS2000) with the DI switched off.

Will switch off the DI and test, but don't remember seeing this in the old HDR mode with Auto 2, open iris, at least not to the same effect. The clipping in highlights is very noticeable when HDR Level = High. I will do a comparison with old HDR autocal Auto 2 vs Frame Adapt HDR Auto 2 vs Frame Adapt HDR DI Off

[Manni]

What the documentation says makes sense, so we're not guessing, we're assuming that until proven wrong the documentation is correct. I don't see any reason why the DTM profile would be isolated from the rest of the PJ, that wouldn't make any sense at all.

You used BT2020 in the original calibration because the PJ was tonemapping using a static PQ curve.

As explained in my former post, it's very possible that the DTM tonemaps to P3.

To get the correct saturations, you need to select the factory color profile that you want to calibrate. If BT2020 was selected in your previous calibration, it might not translate to the DTM.

I'm only suggesting you try this to see if it makes a difference. There has always been a degree of reverse engineering in working out how the Autocal works. I don't have the time at the moment to do detailed tests for the DTM because I use madVR to watch all my content. So you can either try it and see if it makes a difference, or wait until someone else's does.

Of course the user can select a different pictude mode in the software. You simply double click on it and select the one you want. Ideally select one with the same gamut (so P3 or BT2020, we're not sure yet), same lamp, same filter and same iris setting. That should do it.

I don't know what JVC is tonemapping to, but it would certainly make more sense to tonemap to P3 than to BT2020, given the the PJ can't reach wider than P3.

In any case that's none of the user's business. Irrespective of the source, if you want to use the JVC DTM, you have to send HDR in a BT2020 container using PQ from the source, otherwise the PJ will not be able to tonemap it properly.

Consumer content (UHD Bluray, HDR streaming) is always mastered to BT2020 as a container. Some VPs (and possibly this PJ with DTM) can discard the container and tonemap to P3. It's only relevant in case selecting P3 instead of BT2020 in the user mode that you autocal allows you to get better results with DTM. It might or it might not. The iris, filter and lamp mode is certainly more relevant.

Please ask any madVR related question in a madVR support thread. I would suggest you try to get this to work with a standard source, before making sure that you are using the correct settings in madVR, which is off topic here.

Regarding levels, I was suggesting this because I don't experience this at all. I didn't have to change my brightness or contrast settings, either in SDR or HDR. I'm using -2 for both, and that's the correct setting here to get darkest black floor and optimal on/off native contrast. My sources are set to send 16-235, and the PJ is set to HDMI standard. Again, try with a standard source first. There are lots of issues with GPU drivers, OS changes that cause wrong levels depending on the GPU driver, video decoder, renderer, bit depth, etc. This thread isn't a HTPC / madVR debug thread.

Re clipping of the highlights, it varies a bit depending on the source and mostly the manual iris setting, but it's there as long as the DI is switched on if the manual iris is fully open (then it diminishes as you close the iris). Not everyone notices them, but they are there (and have been on previous models as well, except on the rs3000 apparently). So switch the DI off before trying to see if the highlights are crushed, because it's unlikely it's an autocal or a DTM issue, and it's most likely simply the "normal" behaviour of the DI.

[CyberPaul]

Gamut is not relevant for gamma, it is relevant for color.
You need BT2020 when the source sends HDR and you display HDR, because no consumer source will send P3.
But if you tonemap HDR content sent by the source to SDR with a VP (or, in this case, the PJ), you can tonemap to any gamut you want: BT2020, P3 or rec-709. This is what madVR does, what the Radiance Pro does, and possibly what the DTM does. I haven't tested it, but it's not impossible that the DTM tonemaps the the native gamut of the PJ, which is P3, not BT2020.
You were talking about an N7. If an N5, it will not use the filter, but it could still tonemap to anything it wants, rec-709, P3 or BT2020. P3 would still make a lot more sense, because even without the filter these PJs cover more than rec-709.
HDR (whether HDR10 or HDR) means PQ gamma. This doesn't apply to DTM, which tonemaps to SDR, most probably to 2.2 or 2.4.

You said above:
"If you don’t change filter, lamp mode or iris setting, the calibration remains valid even if you change the gamut, gamma or color temp"
If we state that filter, lamp mode and iris are the only discriminators, then we must accept that the gamut can't be relevant. If the gamut was important, when we change color profile with a custom one to correct the inaccuracy of the Spyder 5, then the correction would not work as we would have a new uncalibrated mode.
However, regarding the DTM, I think you are interpreting the color space used as the target after the tone mapping, while I think from a calibration perspective we should consider more the input color space on which the tone mapping works. It is like you are trying to identify an "intermediate" step of the process: input signal -> tone mapping -> internal model. We don't know how exactly is working and we don't need to know it, but for sure we need to feed it with a BT.2020 signal, this is what I meant in my post when I said "I’m pretty sure that DTM is using BT.2020 when filter is set to wide, not DCI"
However, I still believe that the gamut that is used is not important. As Amon reported the manual says:
'Calibration does not work in Frame Adapt HDR mode. Set to another Picture Mode and calibrate using the same Aperture, Lamp Power and Filter settings.'
No mention of color profile...

[Manni]

You said above:
"If you don’t change filter, lamp mode or iris setting, the calibration remains valid even if you change the gamut, gamma or color temp"
If we state that filter, lamp mode and iris are the only discriminators, then we must accept that the gamut can't be relevant. If the gamut was important, when we change color profile with a custom one to correct the inaccuracy of the Spyder 5, then the correction would not work as we would have a new uncalibrated mode.
However, regarding the DTM, I think you are interpreting the color space used as the target after the tone mapping, while I think from a calibration perspective we should consider more the input color space on which the tone mapping works. It is like you are trying to identify an "intermediate" step of the process: input signal -> tone mapping -> internal model. We don't know how exactly is working and we don't need to know it, but for sure we need to feed it with a BT.2020 signal, this is what I meant in my post when I said "I’m pretty sure that DTM is using BT.2020 when filter is set to wide, not DCI"
However, I still believe that the gamut that is used is not important. As Amon reported the manual says:
'Calibration does not work in Frame Adapt HDR mode. Set to another Picture Mode and calibrate using the same Aperture, Lamp Power and Filter settings.'
No mention of color profile...

The input color space is irrelevant for calibration, it's only relevant for the DTM algo, and it should definitely be BT2020 for consumer HDR content. The only gamut that matters when tonemapping re calibration is the output gamut, i.e. the DTM algo target. 

Anyway, I really have no time to argue about theory. I made a suggestion to help solving a problem, people are free to try it or dismiss it.

[estabanj]

How many hours should I wait to do an AutoCal? 10hrs, 20hrs.....

I know the Lamp/PJ doesn't stabilize until ~100hrs and that is the time to wait until you get a professional calibration.

In my case, the PJ is being replaced (blue vertical bars -- yes its still a thing) so will have a brand new PJ.

I saw substantial improvement when I did Autocal in my current NX7 so I am wondering is there a reason to wait before doing an AutoCal with a new PJ, assuming I understand that I will need to do it multiple times before 100hrs.

I am thinking about 10-20 hrs, watch a few movies to make sure that I a good unit (HW infant mortality)

Thoughts?

[Manni]

The 100 hours is to make sure in theory 1) that the bulb stabilizes and 2) that the bulb or PJ isn't a dud. That makes sense if you're hiring a calibrator.

In practice, the bulb keeps changing in the first 300 hours, and then the PJ keeps changing (gamma droop) over the life of the PJ.

So if you're doing it yourself, I would do it right away, maybe wait 10 hours to make sure that the lamp or PJ isn't a dud, but after that it's every 100-500 hours anyway depending on your level of OCD.

I usually check my calibration every 200 hours or so, and only recalibrate if necessary.

These PJs drift anyway, so unless you warm up the PJ/Lamp for at least 45 minutes before calibration and critical watching and your room is temperature controlled (i.e. the room temp doesn't go up as the PJ warms the room), your calibration won't be accurate over a whole film.

Both gamut and RGB balance change significantly as the PJ/room warms up. If it stops warming up at some point, then you can have a repeatable calibration, otherwise I wouldn't be too OCD about this...

[estabanj]

Thanks Manni,

That is what I will do.

BTW if I understand your previous DTM posts correctly its probably worth doing three Picture modes while I have all the equipment out. With User 3 being done in case the DTM algo is actually tone mapping to P3

User1 - SDR709 - Color Profile 709, 6500K, 2.4, low lamp, iris -6 (~60 nits)
User2 - HDR2020 - Color Profile BT2020, HDR10, HDR(PQ), high lamp, iris -0
User3 - HDR-P3 - Color Profile DCI, HDR10, HDR(PQ), high lamp, iris -0

Am I understanding your post correctly?


Thanks again

[grendelrt]

So I read a bunch of people say their black level changed with the new firmware and instead of running -2 or -3 on the brightness they are back to 0. I just checked mine and it looks like I am in the same boat. If I ran autocal last week on gamma with brightness set to -2 do I need to re run autocal at brightness 0. I thought I had seen that answered before but couldn't find the post :O Also just a heads up for people that my brightness did change with the firmware update as some other users had reported (in SDR and HDR modes). 

[CyberPaul]

I was the first to highlight this on AVS forum
Happy you reported the same perception.
Theoretically calibration should be agnostic of level of brightness, but I leave the floor to Manni.

[CyberPaul]

Thanks Manni,

That is what I will do.

BTW if I understand your previous DTM posts correctly its probably worth doing three Picture modes while I have all the equipment out. With User 3 being done in case the DTM algo is actually tone mapping to P3

User1 - SDR709 - Color Profile 709, 6500K, 2.4, low lamp, iris -6 (~60 nits)
User2 - HDR2020 - Color Profile BT2020, HDR10, HDR(PQ), high lamp, iris -0
User3 - HDR-P3 - Color Profile DCI, HDR10, HDR(PQ), high lamp, iris -0

Am I understanding your post correctly?


Thanks again

I keep on thinking that 3rd is not needed and furthermore is going to overwrite the second calibration!
The manual is very clear: the color calibration applies to all profiles with same combination of lamp, Iris and filter. For the gamma just lamp and filter are relevant. In your case, you need to perform gamma calibration twice as well. So color+gamma for 1 and 2. The color space is not considered.

[Manni]

Thanks Manni,

That is what I will do.

BTW if I understand your previous DTM posts correctly its probably worth doing three Picture modes while I have all the equipment out. With User 3 being done in case the DTM algo is actually tone mapping to P3

User1 - SDR709 - Color Profile 709, 6500K, 2.4, low lamp, iris -6 (~60 nits)
User2 - HDR2020 - Color Profile BT2020, HDR10, HDR(PQ), high lamp, iris -0
User3 - HDR-P3 - Color Profile DCI, HDR10, HDR(PQ), high lamp, iris -0

Am I understanding your post correctly?


Thanks again

User 1 and User 2 make sense, User 3 only makes sense if you use a power gamma, though you only have to run a gamma autocal for user 2 or user 3 if the filter/lamp settings are the same. However, you can try to run a color calibration both for user 2 and user 3 if doing just one of these doesn't calibrate your DTM mode.

I have calibrated my SDR rec-709 (gamma+color), SDR BT2020 (color only as it shares its gamma calibration with P3) and SDR P3 (gamma+color) modes and DTM was calibrated fine as far as I could see.

I don't use any HDR modes as I either tonemap with madVR to DCI-P3 or use the DTM for non madVR sources.

[Manni]

So I read a bunch of people say their black level changed with the new firmware and instead of running -2 or -3 on the brightness they are back to 0. I just checked mine and it looks like I am in the same boat. If I ran autocal last week on gamma with brightness set to -2 do I need to re run autocal at brightness 0. I thought I had seen that answered before but couldn't find the post :O Also just a heads up for people that my brightness did change with the firmware update as some other users had reported (in SDR and HDR modes).

I had no such changes and I still need brightness set to -2 in both SDR or HDR in order not to raise the black level and kill on/off contrast. I didn't do a reset to factory settings though.

You shouldn't need to recalibrate after changing the brightness settings. As far as I know, user settings such as brightness/contrast and RGB offsets/gains or gamma settings are only applied for logs, but are not taken into account for calibration. This is why it's a good idea to use a 2.2 gamma and to reset RGB settings before calibration, if you want to have usable logs. Otherwise make sure that you check and adjust all user settings after calibration, as the old ones won't be valid but are still applied.

[grendelrt]

I had no such changes and I still need brightness set to -2 in both SDR or HDR in order not to raise the black level and kill on/off contrast. I didn't do a reset to factory settings though.

You shouldn't need to recalibrate after changing the brightness settings. As far as I know, user settings such as brightness/contrast and RGB offsets/gains or gamma settings are only applied for logs, but are not taken into account for calibration. This is why it's a good idea to use a 2.2 gamma and to reset RGB settings before calibration, if you want to have usable logs. Otherwise make sure that you check and adjust all user settings after calibration, as the old ones won't be valid but are still applied.

Thanks, I didn't see much difference in HCFR, so I think I am good. Yeah I was definitely at -2 since February to get my black floor correct, I had to go back to 0 after the update to get 18 SDR /72 HDR visible blinking again ( still without showing 64 HDR and 16 SDR). I only checked because I saw a couple other people post the same thing, and I thought my image looked too dark during John Wick 3. 

[estabanj]

User 1 and User 2 make sense, User 3 only makes sense if you use a power gamma, though you only have to run a gamma autocal for user 2 or user 3 if the filter/lamp settings are the same. However, you can try to run a color calibration both for user 2 and user 3 if doing just one of these doesn't calibrate your DTM mode.

I have calibrated my SDR rec-709 (gamma+color), SDR BT2020 (color only as it shares its gamma calibration with P3) and SDR P3 (gamma+color) modes and DTM was calibrated fine as far as I could see.

I don't use any HDR modes as I either tonemap with madVR to DCI-P3 or use the DTM for non madVR sources.

Thanks Manni, I will give it a try, when I get the new PJ

[CyberPaul]

User 1 and User 2 make sense, User 3 only makes sense if you use a power gamma, though you only have to run a gamma autocal for user 2 or user 3 if the filter/lamp settings are the same. However, you can try to run a color calibration both for user 2 and user 3 if doing just one of these doesn't calibrate your DTM mode.

I have calibrated my SDR rec-709 (gamma+color), SDR BT2020 (color only as it shares its gamma calibration with P3) and SDR P3 (gamma+color) modes and DTM was calibrated fine as far as I could see.

I don't use any HDR modes as I either tonemap with madVR to DCI-P3 or use the DTM for non madVR sources.

I respectfully disagree.
I'm pretty sure 3rd will nullify 2nd. Again, the color profile selected is not relevant according to the manual. Also in your example, if the two modes share the same combination, you don’t need to calibrate color again.