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

[Manni]

@DCLPhoto
Glad to hear you’ve resolved your issue.
It might help those who don’t (re)read the basic section or the FAQ when they have an issue.
It’s point 9 in the preparation section
As I suggested earlier, you could run a color autocal and display a grescale ramp before deciding if you want to save the calibration or not. You can also save it and test, then restore the latest backup (made after your gamma autocal) if you’re not happy with the results.
The only way to test for sure is with a known good meter and a third party software, or by using an i1pro2 or clone for color.

[DLCPhoto]

However, I’m still not sure why the first (January 19) results were so different from second set.

As for gamma droop, my NX7 also showed a relatively straight line at 400 hours in the Autocal gamma plot, but the blue colour had severe droop, which is visible only in the Color Temp plot (or if you manually measure the RGB response).  Unfortunately that plot only shows the After results, so one would not see it when running gamma auto-cal.

On January 19, the problem was, as Manni correctly identified, that I was too close.  I was only within the 'correct' position within the software, because the 'tilt' was not optimal.  To get it within the box, at the right, I was incorrectly using tilt.

Your suggestion of adjusting tilt while viewing the position in the software, is a great way to find that optimal position (especially since with the DCR Lens in place, I couldn't see any reflection from the Spyder5).  Forward and backward can then be adjusted, along with tilt, in an iterative fashion, since one affects the other, until pretty much perfect placement can be achieved.

Perhaps Manni can add this tip to the Preparation steps, which really comes in handy for those of us doing this with the DCR Lens in place.

In the situation where you found significant Gamma Droop for Blue, what would be the visible impact?  Would it be a shift in the image in a warm (or cool?) direction?  I ask, because I like to think I'm fairly aware of Color Temp (and other aspects of an image) because of my years in Photoshop, addressing these issues in my nature photography with a fairly high-end DSLR).  My NX7 throws what appears to my eye to be an accurate image (in terms of skin tone, sky, grass, etc.), just as I thought the overall tonality was accurate (perhaps confirmed by my Gamma Autocal results).

@DCLPhoto
Glad to hear you’ve resolved your issue.
It might help those who don’t (re)read the basic section or the FAQ when they have an issue.
It’s point 9 in the preparation section 
As I suggested earlier, you could run a color autocal and display a grescale ramp before deciding if you want to save the calibration or not. You can also save it and test, then restore the latest backup (made after your gamma autocal) if you’re not happy with the results.
The only way to test for sure is with a known good meter and a third party software, or by using an i1pro2 or clone for color.

In my case, I had read, and re-read those sections, but simply forgot I had made those Gamma corrections several months before, in a Picture Mode I no longer use.  There's no fix for stupid!

To both of you:

It is straightforward what the 'correct' Gamma graphic should look like after a successful Autocal.  But as I'm still not that well versed in Calibration procedures, what would the 'correct' or 'accurate' Graphics look like for both the Color And Color Temp sections that Autocal provides?  I understand I can visually assess the result using a Gray Ramp, but would like to be better able to understand and interpret what those Graphics are telling me.

Thanks!

[Manni]

It is straightforward what the 'correct' Gamma graphic should look like after a successful Autocal.  But as I'm still not that well versed in Calibration procedures, what would the 'correct' or 'accurate' Graphics look like for both the Color And Color Temp sections that Autocal provides?  I understand I can visually assess the result using a Gray Ramp, but would like to be better able to understand and interpret what those Graphics are telling me.

The only way to know if a gamma autocal went well (according to the Spyder) is to select a gamma of 2.2 and reset all the gamma user settings before the autocal (as advised in the preparation). If you do this, the "after" line should be flat, identical to the target one. If you don't do this, there is no way to assess the result without a known good meter and a third party software, because the target line ALWAYS shows 2.2, even if you select a different gamma in your settings.

Regarding the color autocal, provided you have also reset all the color temp user settings in the calibrated mode, the log after autocal should show all the measurements in the graph close to the 6500K target, and the actual measurements close to 6500K, with a small dE.

If your measurements are far from 6500K, something is not going correctly with the Autocal.

If your after cal log is close to 6500K, this only means that according to this specific spyder unit, the autocal went well. It could still be completely off if the spyder itself is off.

This is why, if you don't have a known good meter to check the results in a third-party software, the only way to check your color autocal is to check visually.

As I said, put up a greyscale ramp before the autocal, look for any non-white/grey shades, run the color autocal, and check the same ramp (without saving). If the greyscale looks better, you likely have a good spyder, so you can save the autocal. If it looks worse, you have a poor one, just discard it.

You have good eyes, I'm sure you'll be able to tell.

[Manni]

I forgot, if you use the Spyder X facing the screen for a color Autocal, make sure that you DON'T select a screen type and that screen type is disabled in the JVC menus.

If you use the Spyder 5 facing the lens for a color autocal, make sure that you DO select the correct screen type in the JVC menus after the autocal, before assessing the greyscale ramp or making any manual adjstments, otherwise your calibration might be off depending on the screen (or you'll lose a significant amount of brightness if you make manual adjustments). My Carada is fairly color neutral, so the screen type doesn't make a big difference, but with some screens the difference can be significant.

This is another indication of a bad spyder / by autocal by the way: excessive loss of brightness due to over-correction.

Make sure that you measure brightness before and after the color autocal (using HCFR for example).

Peak brightness shouldn't change after a gamma autocal, but it will most likely change after a color autocal, sometimes dramatically if the specific Spyder unit is off.

[DLCPhoto]

Regarding the color autocal, provided you have also reset all the color temp user settings in the calibrated mode, the log after autocal should show all the measurements in the graph close to the 6500K target, and the actual measurements close to 6500K, with a small dE.

If your measurements are far from 6500K, something is not going correctly with the Autocal.

If your after cal log is close to 6500K, this only means that according to this specific spyder unit, the autocal went well. It could still be completely off if the spyder itself is off.

How close is close to 6500K in this context:   +/- 5, 10, 50, 100, 300, 500?

And regarding dE - forgive my ignorance, but is this the same as what is termed delta u'v' in Autocal, or is it found somewhere else?  And what would be considered small?

Is there somewhere I can read about the x, y, L data provided for the 7 sets of Color parameters in the Graphics, to better understand what they're telling me?

Finally, if somebody could post screen captures of excellent results for both the Color and Color Temp graphic (and perhaps corresponding ones for what's poor, so a range could be established), that would be helpful - a picture is worth a thousand words, as they say!

I know this is basic stuff for most of you, and I hope I'm not the only one who would benefit from this discussion!

[Manni]

How close is close to 6500K in this context:  +/- 5, 10, 50, 100, 300, 500?

And regarding dE - forgive my ignorance, but is this the same as what is termed delta u'v' in Autocal, or is it found somewhere else?  And what would be considered small?

Is there somewhere I can read about the x, y, L data provided for the 7 sets of Color parameters in the Graphics, to better understand what they're telling me?

Finally, if somebody could post screen captures of excellent results for both the Color and Color Temp graphic (and perhaps corresponding ones for what's poor, so a range could be established), that would be helpful - a picture is worth a thousand words, as they say!

I know this is basic stuff for most of you, and I hope I'm not the only one who would benefit from this discussion!

You are overcomplicating this and not reading the information already provided.

Just look at the pictures I posted in the basic section (section calibration), they show you how things should look like and the kind of errors that are to be expected.

My color autocal showed was around 6300K, and I didn't really mind as I run a 3D LUT after that.

Again, please re-read the basic section, most of the answers to your questions are already there.

The autocal gets you in the ballpark (3-5dE 2000 in my experience, with an i1pro2 as a reference). You first assess if the Autocal goes well, then you assess the results visually.

If you want more than that, you need a better meter and a third-party software so you can fine-tune manually and/or apply a 3D LUT.

[SirMaster]

Hey Manni, I didn't know any other way to get you this info so I'm putting it here.

I just thought you would like to know.

My friend recently re-bought another NX9 and also still an NX5.

I spent some time at his place the other day calibrating and comparing them and I spend some time comparing the DI.


I can confirm without a doubt now that the NX9 does not have the same DI clipping (or yellowing) in highlights that the NX5 or 7 do.

My theory for why this is, is that if you take a look at the actual size of the iris on the NX9, it is significantly larger than the NX5/7.

When it's fully open it's a much larger round circle and when it closes down, it's still a much larger area and even a different shape.

In the Greatest Showman Jenny Lind scene when it's zoomed way out and she is tiny on stage, I see both iris close down the same %, but the NX5 iris is essentially a square (in a diamond orientation) with the height and width corner to corner in the iris the same distance.  On the NX9 however in the same scene, the iris shape is not a square.  The distance from the top and bottom corners is much longer than the left and right corners of the iris opening.

Also when the NX9 iris is closed down in this scene, the total area of the opening seems to be about as big as the NX5 iris is when it's full open.

So my theory is that the whole iris mechanism is vastly different on the NX9 and that the iris opening at every level is larger than the NX5/7.  And of course it's larger because the whole lens is much larger on the NX9.

Based on this I do not expect JVC to be able to "fix" the iris on the NX5/7 unfortunately.


Hope you find this info useful.

I am going back to his place on Wednesday if you have any questions.

[Manni]

Thanks for this confirmation. The different mechanical iris has been the main theory/explanation as to why the issue is present on the N5/N7 and not the N9 (this has been confirmed by Kris Deering and others for a while).

Personally, I don’t think it can be fixed and I am not expecting it to be, especially as the same issue is present in the earlier models (it’s there in my rs500 ), so it’s nothing new.

This is why I have been requesting for a while now an auto3 mode with no gamma manipulation or iris adjustments at all, except during fade to black. The rest of the time, I don’t really need a DI with around 115nits in low lamp. I only would like to lower the black floor during long fade to black.

I have no idea if this is possible, or if JVC is considering implementing this. I certainly hope they are, because the DI is switched off here, it’s just not usable in HDR with the iris fully open.

Not related to the JVC Autocal though, so let’s not start a long debate on this here.

[jj-34]

Hi Manni, I think there are missing replies in this thread ?
The last reply is dated 23rd january, have you pruned it ?
JJ.

[Manni]

Hi Manni, I think there are missing replies in this thread ?
The last reply is dated 23rd january, have you pruned it ?
JJ.

Yes I did delete a couple of my own posts.  They didn't seem to generate much interest and I would need to spend more time to investigate, as my tests were pretty inconclusive. Unfortunately this isn't going to happen for a while, so rather than leaving a few posts hanging I decided to clean up the thread.

[estabanj]

Manni that's a shame, as I also have been seeing some strange results for FA HDR (big inbalance of Red and Blue at the the low-end) which can't be tamed by 2 point RGB adjustments. I suspect they could be tamed by gamma adjustment but of course you can't adjust gamma with with FA-HDR.

I completely understand and respect that you don't want to feel obliged to investigate something so time consuming and   outside your general usage.

I for one don't expect you investigate every aberration seen by you, or anyone else, just because you started the thread.

Thanks for everything you have already provided and feel free to delete this post if you don't think it helpful. I just wanted to give you the feedback I should have when when you originally posted about FA-HDR and give general support for your efforts.

Cheers Stephen

[dcchan2]

I calibrated the grey scale in HDR Picture Mode, but found that FA HDR grey scale does not quite follow it, in spite of what JVC said. 

[estabanj]

Dominic,

Yes, I am confident of being able to post calibrate/tweak the HDR Picture mode like I did for SDR but have consumed most of my time trying to understand FA-HDR. Clearly the setting high, medium and low changes the  gamma and in my currently preferred setting "high" I can't really tame the grey scale much usinging just the 2 point RGB.

Stephen

[dcchan2]

In my case I used gamma Autocal in the HDR10 mode and got good grey scale, but as mentioned in the FA HDR mode the R/G/B tracking is quite a bit off.

[estabanj]

Yes I just checked my HDR profile, this was after autocal with Spyder5Pro gamma and i1pro2 color.
Tweaking the 2 point RGB and the 3 point gamma I was able to get a grey scale deltaE of 2 pretty easily and could likely do better with more time. 

For FA-HDR no amount of tweaking the 2 point RGB could get a grey scale sweep deltaE of under 3.6 - The gamma (which can't be adjusted for FA-HDR) was substantially different (lumpy and below target) from the HDR Profile gamma which was pretty linear and above target, even before tweaking. Which probably makes sense given FA-HDR uses gamma manipulation.

I will continue to test as I get time.

BTW I was able to get a grey scale sweep deltaE of 1.3 for SDR after autocal.