AVS Forum banner

14201 - 14220 of 14270 Posts

·
Registered
Joined
·
29,705 Posts
I am also talking about calibration, please.

The underlying foundation of a good calibration result is how you understand the display/projectors capabilities and limits prior to actually starting your calibration.

I don't think you understand.

The way I understand all this is, all or most colors extending beyond Rec709 on these projectors are an extension of the already mapped out Rec709 colors, the wider gamut is not native and is tricked in place.

Have I got it all wrong, can you not shed more light on this, please?
A good calibrator would apply a 3D LUT if needed.
 

·
Registered
Joined
·
19 Posts
A good calibrator would apply a 3D LUT if needed.
Yes, but a 3D LUT cannot be any better than what the display/device/projector is able to reproduce.

You cannot accurately create a 3D LUT for these projectors larger than Rec709, from what I understand.

Everything beyond Rec709 will be just as unreliable as the projector itself when it comes to a wider color gamut.


I'm not making any sense at all to you, am I?
I'm surprised you guys don't seem to be aware of this issue, you are big users of this enthusiast forum.
 

·
Registered
Joined
·
29,705 Posts
Yes, but a 3D LUT cannot be any better than what the display/device/projector is able to reproduce.

You cannot accurately create a 3D LUT for these projectors larger than Rec709, from what I understand.

Everything beyond Rec709 will be just as unreliable as the projector itself when it comes to a wider color gamut.


I'm not making any sense at all to you, am I?
I'm surprised you guys don't seem to be aware of this issue, you are big users of this enthusiast forum.
That is not correct.
 

·
aka jfinnie
Joined
·
4,449 Posts
Yes, but a 3D LUT cannot be any better than what the display/device/projector is able to reproduce.

You cannot accurately create a 3D LUT for these projectors larger than Rec709, from what I understand.

Everything beyond Rec709 will be just as unreliable as the projector itself when it comes to a wider color gamut.
The above is incorrect.

There are plenty of us who are aware of the limitations of these projectors with respect to their gamut coverage and the behaviour of their colour profiles. As it is they have probably one of the widest gamuts in the market of lamp based projectors, probably only really exceeded by the N7 and NX9 models which are very close to 100% P3.

I have a 3DLUT which is able to cover around 92-94% of P3 (wider in high lamp). By engaging profile off with the filter manually pressed into service using some IP commands it is possible to have access to the native panel gamut and gamma of the projector in a linear fashion (without gamut edge foldback). This is, however, a bit of a hack. All the other profiles have gamut edge foldback to some extent or other, and makes generating a 3DLUT a bit more difficult. There are some workarounds in calibration SW, but they're not very pretty.

My X7900 is deficient in green to reach full P3 even with the filter in, which is pretty common, but it does significantly exceed REC709 with the cinema filter engaged. Without the cinema filter, though the gamut is larger than REC709, on my unit it doesn't quite cover the whole of REC709 (gamut size and gamut coverage being two different and oft confused things).

I'm not making any sense at all to you, am I?
I'm surprised you guys don't seem to be aware of this issue, you are big users of this enthusiast forum.
With this kind of attitude you're not really encouraging folk who do know to share what they know. Remember, you're coming here asking for a favour...

You may be happier with a Sony projector, which has a much smaller gamut than any of these JVC units... ;)
 

·
Registered
Joined
·
9,309 Posts
Yes, but a 3D LUT cannot be any better than what the display/device/projector is able to reproduce.

You cannot accurately create a 3D LUT for these projectors larger than Rec709, from what I understand.

Everything beyond Rec709 will be just as unreliable as the projector itself when it comes to a wider color gamut.


I'm not making any sense at all to you, am I?
I'm surprised you guys don't seem to be aware of this issue, you are big users of this enthusiast forum.
There was a lengthy discussion a while back in the LightSpace thread on this topic. LightSpace was (is?) not able to produce good 3D-LUT for JVC projectors, and blame the projector for that. Apparently DisplacCAL/madVR have no problems.
See for example
and many others.
 

·
Registered
Joined
·
19 Posts
I have a 3DLUT which is able to cover around 92-94% of P3 (wider in high lamp). By engaging profile off with the filter manually pressed into service using some IP commands it is possible to have access to the native panel gamut and gamma of the projector in a linear fashion (without gamut edge foldback). This is, however, a bit of a hack. All the other profiles have gamut edge foldback to some extent or other, and makes generating a 3DLUT a bit more difficult. There are some workarounds in calibration SW, but they're not very pretty.
Can you explain more about this hack, link me to a guide that explains it all?
I think this is what I'm looking for.
 

·
aka jfinnie
Joined
·
4,449 Posts
Can you explain more about this hack, link me to a guide that explains it all?
I think this is what I'm looking for.
and

The latest as yet unreleased firmware for the N7 etc has a proper profile off filter on mode, at last. With an X7900 you'll have to settle for a hack. Works for me, your projector may implode or catch fire. Good luck.
 

·
Registered
Joined
·
9,309 Posts
Turns out you can turn on the filter while in Profile Off mode. Did you know this? Minor annoyance is that it retracts on an HDMI resync (I guess because it re-sets all the correct state for Profile Off mode, which triggers the filter to retract).
I believe Manni also added something related to this in the Vertex2, but I never got into the details of that.
 

·
aka jfinnie
Joined
·
4,449 Posts
I believe Manni also added something related to this in the Vertex2, but I never got into the details of that.
Neat, I guess they added the option to trigger the filter command when a resync happens, so you don't have to rely on it "staying where it is" like I do.

I find the triggering of modes and related in the JVC over IP to be a bit annoying as the projector often ignores the IP interface for a period of time while it is dealing with the mode change, so it seems best to do it as little as possible if possible. I do use the IP stuff in my own control of my X7900 to switch user modes, and it's often a significant amount of time before the projector is in a state to receive and act on commands. Because of this I think if you can contrive your system to have the filter in place always then it is probably a better experience.

Not to mention it sounds like someone just got shot when it rams the filter home...! Honestly the quality of the electromechanical setup in these units is pretty far from giving a high-end feel; between the lamp iris that sounds like a chimp dragging a suitcase on concrete, the lens motors with all the positive action of a spoon made of jello, and the filter BB gun...
 

·
Registered
Joined
·
19 Posts
Am I correct if there was a workaround that has been mentioned regarding getting the i1 Display Pro OEM to work and act as a Spyder with the JVC calibration software?

I know it officially supports the Spyder and the i1 Pro 2, but the latter is maybe not the best meter to measure blacks, goes down to 0.2cd/m2 or 0.2nits, where the i1 Display Pro OEM goes down to 0.1cd/m2 or 0.1nits.

Slighly better, but still far off what the JVC actually can do, Kris Deering mentions a black level reading of 0.0034nits here:

You don't need to measure absolute black?

What are the correction points for the 12pt gamma in the JVC calibration software?
10-100 with 5 and 95%?

5% would be higher nit readings than both 0.1 and 0.2 from both the X-Rites, so maybe no issues?
 

·
Registered
Joined
·
9,309 Posts
Am I correct if there was a workaround that has been mentioned regarding getting the i1 Display Pro OEM to work and act as a Spyder with the JVC calibration software?
See this thread:
 

·
aka jfinnie
Joined
·
4,449 Posts
Am I correct if there was a workaround that has been mentioned regarding getting the i1 Display Pro OEM to work and act as a Spyder with the JVC calibration software?

I know it officially supports the Spyder and the i1 Pro 2, but the latter is maybe not the best meter to measure blacks, goes down to 0.2cd/m2 or 0.2nits, where the i1 Display Pro OEM goes down to 0.1cd/m2 or 0.1nits.

Slighly better, but still far off what the JVC actually can do, Kris Deering mentions a black level reading of 0.0034nits here:

You don't need to measure absolute black?

What are the correction points for the 12pt gamma in the JVC calibration software?
10-100 with 5 and 95%?

5% would be higher nit readings than both 0.1 and 0.2 from both the X-Rites, so maybe no issues?
If you are 3D LUT'ing using Profile Off there is no point using the JVC autocal tool for anything, as profile off bypasses both the colour calibration and the gamma calibration (on the X7900 - I think someone noted that gamma cal does still come into play on the new series).

If you are concerned about the black level measurements with i1d3, then use the meter facing the lens, diffuser down, with the distance set so you are getting approx 1000 nits for peak white. Then you'll find that the best black measure you'll get is still around 0.01nits if you've got iris really closed down for 100K:1. You can reference the meter using a four colour matrix to measurements made off screen (preferably with a spectro, but alternatively with a colorimeter in a pinch) to get the benefit of scaled luminance values and the offset to the actual colour at the screen.
 

·
Registered
Joined
·
19 Posts
If you are 3D LUT'ing using Profile Off there is no point using the JVC autocal tool for anything, as profile off bypasses both the colour calibration and the gamma calibration (on the X7900 - I think someone noted that gamma cal does still come into play on the new series).

If you are concerned about the black level measurements with i1d3, then use the meter facing the lens, diffuser down, with the distance set so you are getting approx 1000 nits for peak white. Then you'll find that the best black measure you'll get is still around 0.01nits if you've got iris really closed down for 100K:1. You can reference the meter using a four colour matrix to measurements made off screen (preferably with a spectro, but alternatively with a colorimeter in a pinch) to get the benefit of scaled luminance values and the offset to the actual colour at the screen.
I don't have the Lumagen, and I buy discs, so no computer or madVR is possible.
I will be doing manual calibration, so assume even no need for Profile Off as gamut foldback only affects 3D LUT.

If the JVC calibration software starts at 5% with the gamma correction, then we don't need to discuss further.
Do you know if it's 5%?

The software doesn't run without having connected the meter and the projector, I have both sitting in their boxes.
 

·
Registered
Joined
·
9,309 Posts
If the JVC calibration software starts at 5% with the gamma correction, then we don't need to discuss further.
Do you know if it's 5%?
It depends on the “quality” setting in the software. You can select 33, 20, or 9 points (not sure about the last one). Also, you can adjust the zoom to give you the smallest (brightest) image when running gamma Autocal; the zoom does not affect the gamma much.
 

·
aka jfinnie
Joined
·
4,449 Posts
I don't have the Lumagen, and I buy discs, so no computer or madVR is possible.
I will be doing manual calibration, so assume even no need for Profile Off as gamut foldback only affects 3D LUT.

If the JVC calibration software starts at 5% with the gamma correction, then we don't need to discuss further.
Do you know if it's 5%?

The software doesn't run without having connected the meter and the projector, I have both sitting in their boxes.
Ok, I figured as you were referencing and seemed concerned by something that is only an issue with 3DLUT, that you were planning on using 3D LUT...

Re: the other Q; can't help as it's ages since I used autocal as... I do it all in 3DLUT! :)
 

·
Registered
Joined
·
19 Posts
It depends on the “quality” setting in the software. You can select 33, 20, or 9 points (not sure about the last one). Also, you can adjust the zoom to give you the smallest (brightest) image when running gamma Autocal; the zoom does not affect the gamma much.
Very valuable information, thank you.
Yeah, been seeing that there's a Speed(9), Normal(20), Quality(33) step setting in Settings.

Speed(9) should then start at 10%, Normal(20) at 5%, Quality(33) at around 3%?

16-235 video range
235/9=26.1/235=0.11= ~10% steps (10% being lowest reading, eyeballing black)
235/20=11.75/235=0.05= 5% steps (5% lowest)
235/33=7.12/235=0.03= 3% steps (3% lowest)

EDIT: actually 235-16=219, not 235, but gives the same results...

More simply:
100/9 or 100/20 or 100/33, same results

So for X-rites, specifically the i1 Pro 2 which JVC is listing as OK with their software:
Say you calibrate for 22fL, 22*0.03=0.66fL = 2.26 nits, so 0.26 nits above i1 Pro 2 lowest nit range (0.2 - 1200).
So basically as long as I don't aim for less than 20fL, homerun.

Am I not correct? i1 Pro 2 will work.

But, 33p is rather unrealistic, maybe a 20p is more than fine even if you're meticulous.
With 20p I can go much lower in brightness, according to my math.
 

·
Registered
Joined
·
9,309 Posts
It’s not the % input that matters; it’s the nits which follow a gamma 2.2 curve, so at 5% it would be 0.14 nits, for 100 nits white. That’s why it’s best to calibrate gamma with maximum brightness.
 
14201 - 14220 of 14270 Posts
Top