AVS Forum banner

2661 - 2680 of 4389 Posts

·
Premium Member
Joined
·
9,824 Posts
Discussion Starter #2,661
Everest is another title with the black issue. Seems it was only a handful of initial Universal titles that had the black issue.
Thanks I'll check it out.

I remember that Divergent (Lionsgate) and/or one of the Hunger Games also had raised blacks, but I'm not sure they were 0.005 nits titles. Need to look at these too.
 

·
Registered
Joined
·
8,830 Posts
For rs420 x5500 series can the autocal with spyder5 do hdr calibration?
Autocal will calibrate the BT2020 colour profile, but in most cases you would need a better EOTF (gamma) curve than the one provided by JVC.
 

·
Registered
Joined
·
23,130 Posts
Alright all, so I screwed up my last autocal, lost probably at least 10% of my Lumens, not exactly sure how much as I didn't note the before/after.

So now the question is, how do I get it back. I know I've seen Manni reference losing brightness after autocal, do I have to reload my INIT file and start over?

I don't recall seeing much discussion on correcting a "bad" autocal.
 

·
Premium Member
Joined
·
9,824 Posts
Discussion Starter #2,665 (Edited)
Alright all, so I screwed up my last autocal, lost probably at least 10% of my Lumens, not exactly sure how much as I didn't note the before/after.

So now the question is, how do I get it back. I know I've seen Manni reference losing brightness after autocal, do I have to reload my INIT file and start over?

I don't recall seeing much discussion on correcting a "bad" autocal.
There is no need to restore the INIT file, as this will reset all your autocals, not just the last one you "screwed", provided you run more than one during each session (one for different modes, i.e. SDR, HDR, etc). If one Autocal isn't satisfactory, just run another one over it without changing the iris/cmd/filter settings until you get good results and keep going. If you want to get back to your last autocal results, it's much better to restore the last backup (the one saved after your last autocal of your last mode/combination of settings at the end of your autocal session). That means you'll get back to what you had just before your "screwed" autocal session, for all your calibrations. You won't get back to what you had after your last successful one, due to lamp, panel, filters drift etc.

Each backup file, saved when you save the Autocal, saves the PJ state just after the Autocal session just saved, for ALL the calibrations. The INIT file is exactly the same, except it's BEFORE you've run any calibration. It's just the state of the PJ for ALL calibrations BEFORE you run the first one. Due to lamp/panel/filter drift, it's only exceptionally useful to go back there and almost never a good idea to restore it, except when the PJ is new/almost new.

First thing to check is the meter. If it has drifted it could be the reason for the brigthness loss. The Spyders have plastic filters so they do drift after a while if you don't keep them in Peli-case with dessicant, which slows down the process but doesn't entirely prevent it.

Is your brightness loss right after the autocal, or after you correct the gains at 100% white?
Have you selected the correct screen (especially if your screen isn't color neutral)?
Have you corrected the errors of the Spyder with a matrix using a reference meter?
Have you checked during the autocal that the log didn't show any zero gamma measurement near black?
Which measurement is wrong? The P3 filter and the panels can drift too, so there are things against which there is little you can do: poor saturation tracking in BT2020 due to filter drift, gamut shrinking due to panel drift, brightness drop due to lamp aging.

If you tell us what is "screwed" it will be easier to diagnose if it's the meter, the user, the PJ, the lamp or a combination these. Hopefully, you measured your "before" state with another software and a better/reference meter before running the Autocal, so you know what's caused by the Autocal and what's not. If you haven't, it's highly recommended to do so to be able to identify possible issues due to PJ drift (lamp/filter/panel) that might not be autocal related even if you only notice them after the autocal.
 

·
Registered
Joined
·
23,130 Posts
First thing to check is the meter. If it has drifted it could be the reason for the brigthness loss. The Spyders have plastic filters so they do drift after a while if you don't keep them in Peli-case with dessicant, which slows down the process but doesn't entirely prevent it.
I'll probably have to do that.

Is your brightness loss right after the autocal, or after you correct the gains at 100% white?
Yeah, right after the autocal (color only), I didn't mess with anything else.

Have you selected the correct screen (especially if your screen isn't color neutral)?
I haven't actually been using that, my screen appears to be pretty color neutral.

Have you corrected the errors of the Spyder with a matrix using a reference meter?
Do you mean a custom color profile to account for the spyder errors? If so, yes, otherwise I'm not sure, I was just measuring fc with my Lux meter.

Have you checked during the autocal that the log didn't show any zero gamma measurement near black?
I didn't run a full gamma autocal, but there were no zero's in the log either before or after.

Which measurement is wrong? The P3 filter and the panels can drift too, so there are things against which there is little you can do: poor saturation tracking in BT2020 due to filter drift, gamut shrinking due to panel drift, brightness drop due to lamp aging.
So I didn't have a chance to go remeasure with my i1D3/Calman, what I did was:
Setup a new user memory referencing my new custom 2020NF profile.
Measured fc with my light meter, compared it to my other user memory with a custom 2020F profile
Ran color only autocal on that user memory
Measured fc with my light meter, and noticed a 10% (like I said I didn't actually write down the numbers, shame on me) drop vs pre-autocal for peak white.

If you tell us what is "screwed" it will be easier to diagnose if it's the meter, the user, the PJ, the lamp or a combination these. Hopefully, you measured your "before" state with another software and a better/reference meter before running the Autocal, so you know what's caused by the Autocal and what's not. If you haven't, it's highly recommended to do so to be able to identify possible issues due to PJ drift (lamp/filter/panel) that might not be autocal related even if you only notice them after the autocal.
Yeah, I was kind of in a hurry, I really just wanted to try out a 2020NF profile to see what a few more nits look like, so I wasn't as thorough as I should have been. But I'd had pretty good luck with autocal (or so I thought) so far when I have double checked with Calman, I figured I could just do a "quick" cal of my new profile/memory.

I probably really should just set aside a couple hours, restore my INIT and just start from scratch.

Maybe I should try and find a used i1 Pro or something.
 

·
Registered
Joined
·
349 Posts
Hey guys,..working with Ryan Masciola to get his test pattern files to work on ATV4. He sent me the black clipping 2 file to test and it works on ATV4 in 4k as it's saved in my itunes library!!!! See attached, for the first few seconds, the 81 and higher flash grey then black, grey then black and then after the first few seconds, it seems all of them start flashing as in the pic below. Meaning at 64 you can see a very very faint hint of great but definitely can see at 77. Assuming people have used this same file burnt on 4k DVD, do you wait until all the bars are flashing (like after a few seconds) and bbo to make 77 not flash (black)?

Sorry in advance if the file is HUGE...can't figure out how to shrink it down....
 

Attachments

·
Premium Member
Joined
·
9,824 Posts
Discussion Starter #2,669
I'll probably have to do that.



Yeah, right after the autocal (color only), I didn't mess with anything else.



I haven't actually been using that, my screen appears to be pretty color neutral.



Do you mean a custom color profile to account for the spyder errors? If so, yes, otherwise I'm not sure, I was just measuring fc with my Lux meter.



I didn't run a full gamma autocal, but there were no zero's in the log either before or after.



So I didn't have a chance to go remeasure with my i1D3/Calman, what I did was:
Setup a new user memory referencing my new custom 2020NF profile.
Measured fc with my light meter, compared it to my other user memory with a custom 2020F profile
Ran color only autocal on that user memory
Measured fc with my light meter, and noticed a 10% (like I said I didn't actually write down the numbers, shame on me) drop vs pre-autocal for peak white.



Yeah, I was kind of in a hurry, I really just wanted to try out a 2020NF profile to see what a few more nits look like, so I wasn't as thorough as I should have been. But I'd had pretty good luck with autocal (or so I thought) so far when I have double checked with Calman, I figured I could just do a "quick" cal of my new profile/memory.

I probably really should just set aside a couple hours, restore my INIT and just start from scratch.

Maybe I should try and find a used i1 Pro or something.
Yes I meant a custom colour profile correcting the Spyder's errors.

You can't simply use BT-2020NF and run the color autocal because whatever gamma you have was calculated either with the right gamut but with a filter (reference) or with the wrong gamut but without a filter (standard). There is no reference-NF preset, so we can't run an autocal for BT-2020NF. I need to run some tests and do measurements to see what the best factory color profile should be used to achieve the best possible calibration with BT2020-NF. It's something I ran into a while ago but completely forgot about, and recently Asharma brought it up (see recent posts) and we said he had to wait until I could find the time. So either you have the time to run all the experiments, or you'll have to wait too. At this stage, I can't make any recommendation regarding how to calibrate with BT-2020NF.

If you've wrecked your HDR calibration, it's safer to redo it (using reference for gamma+color), or even simply to reload your previous corrected profile if you have it. Once you correct a profile there is no need to re-run the colour autocal on the corrected profile. You can try to interpolate the correction to BT2020NF from BT2020-F, but there is no guarantee it will not make it worse as the corrction to apply might not be the same with and without the filter.

There is no point in buying an i1pro if you have an i2d3. The accuracy is about the same with LCOS (provided yours is decent, and most of them are). If you get a spectro, get an i1pro2, at least that model is supported by the new models/new versions of Autocal. But as the i1pro2 isn't usable on its own for gamma without training it to a colorimeter (I use my Discus) I have no idea how that works.

If it's only to create your custom colour profile correcting the errors of the Spyder, an i1pro will give you very little over your i1d3.
 

·
Registered
Joined
·
23,130 Posts
You can't simply use BT-2020NF and run the color autocal because whatever gamma you have was calculated either with the right gamut but with a filter (reference) or with the wrong gamut but without a filter (standard). There is no reference-NF preset, so we can't run an autocal for BT-2020NF. I need to run some tests and do measurements to see what the best factory color profile should be used to achieve the best possible calibration with BT2020-NF. It's something I ran into a while ago but completely forgot about, and recently Asharma brought it up (see recent posts) and we said he had to wait until I could find the time. So either you have the time to run all the experiments, or you'll have to wait too. At this stage, I can't make any recommendation regarding how to calibrate with BT-2020NF.
I recalled the recent discussion, but I misinterpreted what you were telling Asharma.

If you've wrecked your HDR calibration, it's safer to redo it (using reference for gamma+color), or even simply to reload your previous corrected profile if you have it. Once you correct a profile there is no need to re-run the colour autocal on the corrected profile. You can try to interpolate the correction to BT2020NF from BT2020-F, but there is no guarantee it will not make it worse as the corrction to apply might not be the same with and without the filter.
Yeah, I remember that now, it's just been a long time. I'd forgotten that when I made my original "S5" (Spyder 5) profiles, I'd just done a regular autocal, ran measurements between my Spyder and i1D3 in Calman (or probably HCFR since Calman didn't support the Spyder at the time) computed the new profile, and uploaded it.

I also need to go download Chad's profile creator spreadsheet from the Calman forums, I just made my own, I'm sure his is more accurate.

There is no point in buying an i1pro if you have an i2d3. The accuracy is about the same with LCOS (provided yours is decent, and most of them are). If you get a spectro, get an i1pro2, at least that model is supported by the new models/new versions of Autocal. But as the i1pro2 isn't usable on its own for gamma without training it to a colorimeter (I use my Discus) I have no idea how that works.
Doh, I guess that used one I saw wasn't an i1pro2. But that's the rub isn't it, how do you know if your i1d3 is good. I assume mine is, but it's an assumption. Of course the reasons above is why I'd never go buy a new one for $1000+. I thought I found one on Ebay for $375, but apparently it's an older one, not the 2.

If it's only to create your custom colour profile correcting the errors of the Spyder, an i1pro will give you very little over your i1d3.
Good to know. We'll see when I get back to this, I may be able to "fix" my cal by just reverting to a calibration before yesterday. Things seemed decent before I (errantly) ran autocal when I shouldn't have last night. Maybe I'll take the time to upload my prior autocal and do a quick run through with Calman and my i1D3 and see how things look. My cal has actually been looking pretty good in Calman, though my secondaries were off a decent amount last time I checked.
 

·
Premium Member
Joined
·
9,824 Posts
Discussion Starter #2,671
I recalled the recent discussion, but I misinterpreted what you were telling Asharma.

Yeah, I remember that now, it's just been a long time. I'd forgotten that when I made my original "S5" (Spyder 5) profiles, I'd just done a regular autocal, ran measurements between my Spyder and i1D3 in Calman (or probably HCFR since Calman didn't support the Spyder at the time) computed the new profile, and uploaded it.

I also need to go download Chad's profile creator spreadsheet from the Calman forums, I just made my own, I'm sure his is more accurate.

Doh, I guess that used one I saw wasn't an i1pro2. But that's the rub isn't it, how do you know if your i1d3 is good. I assume mine is, but it's an assumption. Of course the reasons above is why I'd never go buy a new one for $1000+. I thought I found one on Ebay for $375, but apparently it's an older one, not the 2.

Good to know. We'll see when I get back to this, I may be able to "fix" my cal by just reverting to a calibration before yesterday. Things seemed decent before I (errantly) ran autocal when I shouldn't have last night. Maybe I'll take the time to upload my prior autocal and do a quick run through with Calman and my i1D3 and see how things look. My cal has actually been looking pretty good in Calman, though my secondaries were off a decent amount last time I checked.
I checked two or three i1d3 (before returning them for other reasons) and they were all so close to my i1pro2 (with the right profile for LCOS) that I wouldn't have profiled them to my i1pro2 as I woudn't have know which was more correct. The meter survey I posted (probably linked to in the first post) also shows very little variance and great accuracy for the i1d3, unlike the Spyders. So I would assume yours is good unless something seems off with it. The main problem of the i1d3 is its accuracy in low light. It's very poor and it can't read black, so you can't use it for BT1886 gamma calibrations or to read on/off contrast. That's why I sent it back and got a Discus instead (i1pro2-Discus is the best combo you'll find unless you go for true reference meters and silly prices). But above 5% white, the i1d3 is fine on LCOS, for both gamut and gamma.

You don't need to get to the Calman forum for Chad's spreadsheet, both his original method and my modified ones are linked to in the first post, as well as my own spreadsheet and method to calibrate HDR-BT2020 (as Chad's method extrapolates from rec-709).

In fact I'm working on a way to improve this method because I've ran into issues when the native gamut is undersaturated as I initially feared would be the case, but I don't have the time to discuss now or document.
 

·
Registered
Joined
·
8,830 Posts
The main problem of the i1d3 is its accuracy in low light. It's very poor and it can't read black, so you can't use it for BT1886 gamma calibrations or to read on/off contrast.
To calibrate gamma I have the i1D3 face the projector lens, with the diffuser on. It may still have problems when the Auto Iris is on, but I don't calibrate that way.
 

·
Premium Member
Joined
·
9,824 Posts
Discussion Starter #2,673
To calibrate gamma I have the i1D3 face the projector lens, with the diffuser on. It may still have problems when the Auto Iris is on, but I don't calibrate that way.
Yes, that's exactly what I'm saying, you can't use the i1d3 to calibrate BT1886 gamma off the screen, which is the way it should be done to take the screen into account.

With the Discus I can measure fairly reliably my on/off contrast facing the screen, even with the DI on. I don't calibrate with the DI, but it's still useful to measure its impact.
 

·
Registered
Joined
·
8,830 Posts
Yes, that's exactly what I'm saying, you can't use the i1d3 to calibrate BT1886 gamma off the screen, which is the way it should be done to take the screen into account.
The screen reflectance is linear, so I don't see any way the gamma can be affected by the screen, unlike grey scale, luminance etc.
 

·
Registered
Joined
·
2,804 Posts
I'll probably have to do that.



Yeah, right after the autocal (color only), I didn't mess with anything else.



I haven't actually been using that, my screen appears to be pretty color neutral.



Do you mean a custom color profile to account for the spyder errors? If so, yes, otherwise I'm not sure, I was just measuring fc with my Lux meter.



I didn't run a full gamma autocal, but there were no zero's in the log either before or after.



So I didn't have a chance to go remeasure with my i1D3/Calman, what I did was:
Setup a new user memory referencing my new custom 2020NF profile.
Measured fc with my light meter, compared it to my other user memory with a custom 2020F profile
Ran color only autocal on that user memory
Measured fc with my light meter, and noticed a 10% (like I said I didn't actually write down the numbers, shame on me) drop vs pre-autocal for peak white.



Yeah, I was kind of in a hurry, I really just wanted to try out a 2020NF profile to see what a few more nits look like, so I wasn't as thorough as I should have been. But I'd had pretty good luck with autocal (or so I thought) so far when I have double checked with Calman, I figured I could just do a "quick" cal of my new profile/memory.

I probably really should just set aside a couple hours, restore my INIT and just start from scratch.

Maybe I should try and find a used i1 Pro or something.
I was looking to run an Autocal on bt2020nf last week but Manni advised not to as there are potential issues in doing so...I was also looking for increased nits but did not run Autocal as Manni advised against it...
 

·
Registered
Joined
·
2,700 Posts
Manni said above (bold is mine):

You can't simply use BT-2020NF and run the color autocal because whatever gamma you have was calculated either with the right gamut but with a filter (reference) or with the wrong gamut but without a filter (standard). There is no reference-NF preset, so we can't run an autocal for BT-2020NF. I need to run some tests and do measurements to see what the best factory color profile should be used to achieve the best possible calibration with BT2020-NF. It's something I ran into a while ago but completely forgot about, and recently Asharma brought it up (see recent posts) and we said he had to wait until I could find the time. So either you have the time to run all the experiments, or you'll have to wait too. At this stage, I can't make any recommendation regarding how to calibrate with BT-2020NF.
I suspect I'm missing something here, as per Posts 9 and 10 in this thread, providing detailed recommendations for JVC Autocalibration, a Color only Autocal was advised for BT2020-NF, following the Gamma+color Autocalibration using a standard Color Profile and 'normal' Gamma (for each combination of lamp, iris, CMD state, etc.).

This seemed to be an established procedure. Is this now being called into question?

I didn't quote Manni in the usual way, as I don't mean to take him away from whatever he's working on, but there are others here who seem to understand these procedures quite well.

Am I missing something fundamental here, or is a change in the basic Autocalibration procedure for the Color Profile aspect of HDR now underway?
 

·
Premium Member
Joined
·
9,824 Posts
Discussion Starter #2,678 (Edited)
Manni said above (bold is mine):



I suspect I'm missing something here, as per Posts 9 and 10 in this thread, providing detailed recommendations for JVC Autocalibration, a Color only Autocal was advised for BT2020-NF, following the Gamma+color Autocalibration using a standard Color Profile and 'normal' Gamma (for each combination of lamp, iris, CMD state, etc.).

This seemed to be an established procedure. Is this now being called into question?

I didn't quote Manni in the usual way, as I don't mean to take him away from whatever he's working on, but there are others here who seem to understand these procedures quite well.

Am I missing something fundamental here, or is a change in the basic Autocalibration procedure for the Color Profile aspect of HDR now underway?
Thanks for not quoting me, appreciated :)

The initial recommendation was correct at the time (probably still the best way to use BT2020NF, although I haven't double-checked that) but that doesn't work with the enhanced method provided by Chad initially and adapted for HDR-BT2020 later, because we don't have the equivalent of a reference-nf profile to correct the meter errors and to calibrate gamma with a gamut that is not entirely different. I guess Chad's original method, that interpolates the corrections from the rec-709 measurements could be used, but I'm not making any recommendations because I need to run some tests to define the best procedure and haven't had the time to do so. I'll do it as soon as I can but right now, I don't want to say anything because the last time I looked at BT2020-NF was a long time ago and our best practice recommendations have changed since, at least if you have a better meter than the Spyder to correct its errors.

If you only use the Spyder, the procedure described in posts 9 and 10 is likely the best way, but I can't guarantee it as I haven't revisited it in a long time.

Of course, if anyone has run the measurements with a reference meter and is able and willing to provide some info in the meantime, they are welcome to do so.
 

·
Registered
Joined
·
349 Posts
I don't know the answer to your question but it is WAY awesome if we are going to see patterns compatible with the ATV4k. I have the soft copy files and the UDH Disk, and this would help me ensure that my ATK4K is outputting the same signal as the UHD player. Thanks!
Ya, I ordered the package and took some time to download, extract and place in my itunes library on main computer. When I did that, all titles were created in the itunes library but only 64 would show on Apple TV in my library. Was stumped so worked back and forth with Ryan and he finally figured out what was wrong....that is the great news. Bad news is it's a whole rewrite of his pattern files for it to work on ATV4....for the time being I just wanted the Black Clipping and that is what I have now.

Still stumped on why the first few seconds only above 81 flashed grey then for the remainder of the track all above 64 show a bit of grey. Someone on here must have used this same file to dial in bbo so was hoping I could get some input before I do that.
 

·
Registered
Joined
·
5,049 Posts
First thing to check is the meter. If it has drifted it could be the reason for the brigthness loss. The Spyders have plastic filters so they do drift after a while if you don't keep them in Peli-case with dessicant, which slows down the process but doesn't entirely prevent it.
Is this a fact or educated guess? I have no idea how plastic filters work on Spyders, but is it possible a dessicant can make it worse? My thought is that as plastic ages it dries out and a dessicant should make the air in the case drier, which could accelerate the aging of the plastic filter?

I have no idea, just thinking out loud.

Also, are we sure there's no outgassing of the the dessicant that could also harm the Spyder?

??
 
2661 - 2680 of 4389 Posts
Top