ColourSpace CMS - Next Generation Calibration... Thread - Page 23 - AVS Forum | Home Theater Discussions And Reviews
Forum Jump: 
 905Likes
Reply
 
Thread Tools
post #661 of 764 Old 06-28-2020, 12:50 PM
aka jfinnie
 
Join Date: Aug 2015
Location: Norwich, UK
Posts: 4,208
Mentioned: 97 Post(s)
Tagged: 0 Thread(s)
Quoted: 3359 Post(s)
Liked: 2606
Quote:
Originally Posted by Light Illusion View Post
But as with many of our approaches, there are no set rules.
For example, when performing MPVM probe matching, a patch set with an inverse gamma applied (so there are less dark patches) can be advantageous.
I did help some friends generate such sets - using the tool linked a patch power of around 0.65 seemed pretty useful for that purpose, and avoided having to manually cull patches with the inherent annoyance / chance of mis-step. Maybe you might consider a simple patch power adjustment option for the cube patch list.

Quote:
Originally Posted by Light Illusion View Post
And we define the patch colours in 8 bit, we do not round down.
Fair enough, I don't know exactly how your code works, though given the end result is the same I'm not sure it much matters (I worked out your method by analysing a few patch sets and coming up with an algorithm that came up with the same result.). I think I did the "rounding" by casting a float to an int.

Quote:
Originally Posted by Light Illusion View Post
In future iterations of ColourSpace we will add 10 bit, and likely 12 bit native values for patch generation.
Cool, look forward to see it. Lumagen it seems only do 8 bit patterns, but I recall my Accupel does 10 bit.
bobof is online now  
Sponsored Links
Advertisement
 
post #662 of 764 Old 06-28-2020, 12:54 PM
AVS Forum Special Member
 
Light Illusion's Avatar
 
Join Date: Aug 2010
Posts: 2,186
Mentioned: 38 Post(s)
Tagged: 0 Thread(s)
Quoted: 1005 Post(s)
Liked: 1612
Quote:
Originally Posted by bobof View Post
Maybe you might consider a simple patch power adjustment option for the cube patch list.
You can do that already...
Just generate a 1D LUT with an inverse Gamma of the desired value, and export the patch set through it.

Steve
bobof likes this.

Steve Shaw
LIGHT ILLUSION

Light Illusion is offline  
post #663 of 764 Old 06-29-2020, 03:11 PM
Advanced Member
 
liberator72's Avatar
 
Join Date: Mar 2018
Location: Sunny (when it's not raining) Devon, U.K.
Posts: 830
Mentioned: 53 Post(s)
Tagged: 0 Thread(s)
Quoted: 609 Post(s)
Liked: 1147
Quote:
Originally Posted by liberator72 View Post
There is no “evil” hidden anywhere. I was openly honest in the post you refer to and say explicitly that it was just quick test for my own benefit with no intention of posting it here and should only be used as a “rough idea” to gauge the benefits!!!

I also clearly state they are exact same bpd’s used for the FCVM in ColourSpace directly exported/imported to LightSpace so that I could see how the resulting FCMM would perform using that same correction data, but that they are two separate profiles. They should not be used as a fair comparison but more of a rough guide. So again, there is no “evil” hidden anywhere.

But to answer your request, yes I can send you the bcs and bpd files later when I get the chance
Following up on this from yesterday. I accept that there was something maybe "lost in translation" but I must say that at the time I did feel like I was being called out as being dishonest even though it now clear to me that was obviously not intended.

So, I have sent the files as requested to ebr9999, but I have also done myself what he wants to do, which was to take the FCMM profile from LightSpace and invert the correction from it to provide uncorrected profile data (with the help of Anger.miki). That uncorrected profile data has then been imported to ColourSpace where FCVM has been applied to it using the same original BPD files.

Note. The FCVM profile has 1 extra point within it as I removed Black from the Patch Sequence to speed up Jeti read times (it can take 2 minutes to read black). LightSpace does not add this point automatically, whereas ColourSpace adds Black if it is missing upon saving a profile.

First, the original FCMM profile compared to a Jeti Profile


Then, that EXACT same profile data, but with FCVM applied and compared to the EXACT same Jeti Profile


Once again, the original post was intended to be used as a "rough guide" as to the benefits for i1d3 users. This can be considered as "real world usage" benefits to i1d3 users

LG OLED65C9MLB**Denon AVR-X3500H**Focal Sib Evo Atmos 5.1.2**Sky Q UHD**Xbox One X**Panasonic DP-UB820**ATV 4K
Klein K10-A**Jeti 1501**i1D3 OEM**ColourSpace HTX**LightSpace HTP**CalMAN Home for LG**DVDO AVLab TPG**HD Fury Integral 2
Ted's Disk & Media Files**S & M HD & UHD HDR Benchmark Disks

Last edited by liberator72; 06-29-2020 at 03:18 PM.
liberator72 is online now  
Sponsored Links
Advertisement
 
post #664 of 764 Old 06-29-2020, 08:12 PM
Advanced Member
 
Vishwa Somayaji's Avatar
 
Join Date: Jun 2001
Location: Edmonton,Alberta Canada
Posts: 613
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 136 Post(s)
Liked: 126
Thank you very much for doing this. I believed what you said the first time around! I don't have access to the expensive spectro or colorimeter. But, it feels good to know how close just the iD3 comes in.
mrtickleuk likes this.
Vishwa Somayaji is offline  
post #665 of 764 Old 06-30-2020, 12:08 AM
AVS Forum Special Member
 
Manni01's Avatar
 
Join Date: Sep 2008
Posts: 9,747
Mentioned: 384 Post(s)
Tagged: 0 Thread(s)
Quoted: 5925 Post(s)
Liked: 6508
@liberator72 Thank you for sharing your results, your posts were perceived as intended the first time. And apologies for adding to the noise by asking you to confirm something that I had missed in your initial post.

I have done a quick test with the Discus and I measure a standard anisometric 10x10x10 in 28 minutes vs your 20 minutes (in my slowest SDR calibration with a peak white of 50nits). If this translated to LCOS, this means that the i1d3 would be about 30% faster, while being less accurate than the Discus in low light. I assume that the K10a would be about 50% faster (15 minutes?) while being more accurate in low light than the Discus.

The main question for me is how accurate would the i1d3 be with a 21points LUT, when reading 5% stim RGBW on LCOS. When I evaluated the i1d3 a few years back, it was not only unable to read black reliably (like the Discus is under certain conditions), which makes it impossible to use for BT1886 or contrast measurements, but it was unable to read 5% white off the screen with my JVC projector at the time.

In the low end, with your "turbo" settings, how low do you expect the i1d3 to read reliably? Is it able to read the darkest patches of a 21pts LUT reliably with these settings? What is your peak white? What is your 5% white Y?

Currently, with the Discus, it takes around 2.5 hours to run a 17pts LUT for HDR tonemapped content (120 peak nits, 5% white around 0.17nits) without low light averaging, and a lot longer with my calibration for SDR content (currently 54nits peak white, 5% white around 0.05nits after calibration using BT1886, would be lower with power 2.4), up to five hours with low light averaging enabled. I currently use an optimised patch sequence of under 4,000 points that runs under 2 hours for my HDR calibration with the Discus profiled to the i1pro2 (I haven't had a chance to try FCVM yet but on LCOS FCCM already worked very well, so I'm not expecting huge improvements) and provides excellent results with CS.

If your results translated to a 30% speed increase here, it would allow me to read 5,000 patches in a bit more than 1.5 hours with the i1d3, which makes a 21pts LUT just about doable at around 3 hours, but only if the results are usable, especially in the low end.

Have you tried measuring a 21pts LUT with your turbo settings, and is your i1d3 reliable reading the darkest patches? Have you tried a calibration with a peak white as low as 50nits (which is the standard for SDR with projectors in a dedicated room) and patches going down as low at 0.05nits for white?

Thank you again for sharing all your results, I don't own an i1d3 currently but your results with turbo settings make me think about getting one to run large LUTs faster than I can with my Discus.

I'd love to not have to upgrade to a K10a, but despite the significant speed improvements that you report with your "turbo" settings in your setup (I assume targeting 100nits in SDR), I'm not sure that an i1d3 would work reliably here with 21pts LUTs, especially with my SDR calibration (50nits peak white, 0.05nits for 5% white).

That's the reason why I had to go with the Discus rather than the i1d3, and I'm concerned that this limitation might still stand given the limited ability of the i1d3 in low light.

Thoughts?
mrtickleuk likes this.

Batch Utility V4.02 May 16 2019 to automate measurements files for madVR with support for BD Folders
JVC Macro feature on Vertex/Vertex2/Integral2/Maestro/Diva

Last edited by Manni01; 06-30-2020 at 01:33 AM.
Manni01 is offline  
post #666 of 764 Old 06-30-2020, 12:58 AM
Member
 
Join Date: Jan 2017
Posts: 61
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 36 Post(s)
Liked: 3
Hi guys, I have a question for AWP

AWP was obtained by matching with reference systems such as Pioneer Kuro, but then the xy values were purchased with high quality Spectroradiometers, but who owns only an EyeOne Pro2 Spectrophotometer in theory will be very far from the correct XY values or the distance is acceptable for which the AWP indicated is still valid?

I am interested in understanding this AWP issue because I find ColourSpace exceptional and the new FCVM system has allowed me to eliminate the Y problem on the Red (which also created problems in the evaluation of Magenta and Yellow) now even a short Lut of 1000 points has not worsened the skin color unlike what happened before (albeit in a very subtle measure), now if I solve the doubt AWP vs my modest EyeOne pro2 / iD3 I can try a wider Lut 3D.
mikigio is offline  
post #667 of 764 Old 06-30-2020, 01:04 AM
AVS Forum Special Member
 
Manni01's Avatar
 
Join Date: Sep 2008
Posts: 9,747
Mentioned: 384 Post(s)
Tagged: 0 Thread(s)
Quoted: 5925 Post(s)
Liked: 6508
@Iron Mike Thanks, but there is no need to be insulting to anyone sharing info here, and you haven't read my post properly.

I am not being "fooled", and I have no doubt that a tweaked i1d3 can give great results (both fast and accurate) on some displays.

I evaluated 2 or 3 i1d3s a few years back and rejected them due to their poor low light ability on LCOS projectors. As explained in my previous post, this is why I bought the Discus at the time. I'm very happy with the Discus performance, especially in low light. It's just a bit slow with large LUTs.

I have no doubts about @liberator72 results in his set-up, I'm only questioning whether a current i1d3 would allow me to run large LUTs in mine, hence my questions.

As explained, I'm using optimized patches already (around 4,000 patches) that give me better results than a 5,000 17pts cube, but the time needed with the Discus (2 hours+) and the remaining errors (using the native, uncorrected gamut) make me consider upgrading my Discus because I need more points than that to lower the few spikes I'm still getting, or to run the same size profile in less time.

Yes a Klein K10a has been on the list for years (after the discus, I want LED/laser aiming, so that rules the CR100 out), but if I was able to keep measuring contrast with the Discus and could measure larger LUTs (most probably optimised, I was only using 21pts as a frame of reference) faster yet still reliably, that could be an interesting compromise as long as I don't hit the same limitations in low light I hit with my previous evaluations of the i1d3.

Apart from using a custom patch set, I'm not after any "laborious ways" of improving results. There are enough steps involved in generating a valid LUT for madVR/Envy in CS, I don't want to have to add more filters or inject any data. Once I have found my custom patch sequence, I want to profile, calculate the LUT, apply a video filter to make it valid and upload it to madVR/Envy.
mrtickleuk likes this.

Batch Utility V4.02 May 16 2019 to automate measurements files for madVR with support for BD Folders
JVC Macro feature on Vertex/Vertex2/Integral2/Maestro/Diva
Manni01 is offline  
post #668 of 764 Old 06-30-2020, 01:20 AM
AVS Forum Special Member
 
Iron Mike's Avatar
 
Join Date: Oct 2006
Posts: 1,081
Mentioned: 10 Post(s)
Tagged: 0 Thread(s)
Quoted: 190 Post(s)
Liked: 504
@Manni01 :

if u don't wanna use laborious workarounds, just buy a proper meter.

or live with the result u get from whatever u have.

in any case: u can't and won't have it both ways no matter what u think or some claim here.

also, just a pro tip, I'd look into the specifics of the color engine of ur cal solution b/c depending on what that is, more patches won't help so don't fool urself again.

good luck !

calibration & profiling solutions: DCT v2, Lightspace, Spaceman, Calman, Argyll, ColorNavigator, basICColor
profiling & calibration workflow tools: Display Calibration Tools, Lightspace & Custom Color Patch Set & Gamma Calibration on Panasonic 65VT60
meter: CR-300, Klein K-10 A, i1Pro 2, i1D3
Iron Mike is offline  
post #669 of 764 Old 06-30-2020, 01:27 AM
AVS Forum Special Member
 
Manni01's Avatar
 
Join Date: Sep 2008
Posts: 9,747
Mentioned: 384 Post(s)
Tagged: 0 Thread(s)
Quoted: 5925 Post(s)
Liked: 6508
Quote:
Originally Posted by Iron Mike View Post
@Manni01 :

if u don't wanna use laborious workarounds, just buy a proper meter.

or live with the result u get from whatever u have.

in any case: u can't and won't have it both ways no matter what u think or some claim here.

also, just a pro tip, I'd look into the specifics of the color engine of ur cal solution b/c depending on what that is, more patches won't help so don't fool urself again.

good luck !
If you genuinely want to be useful, please answer the specific questions I asked @liberator72 in my post. You have a K10A and an i1d3, like he does. They are VERY specific questions and have nothing to do with the color engine of the software solution, and everything to do with the meter limitations in low light.

If you have any experience with using an i1d3 specifically on LCOS, specific information would be welcome. The last time I tried one was ages ago, and I had not tried to optimise any speed settings. With the default settings, it was about as fast as the Discus, but it was simply not good enough to be used on LCOS, with any calibration software.

Otherwise, your posts are just unneeded, unhelful and patronizing noise. @liberator72 is providing some useful data. You are providing none.
mrtickleuk likes this.

Batch Utility V4.02 May 16 2019 to automate measurements files for madVR with support for BD Folders
JVC Macro feature on Vertex/Vertex2/Integral2/Maestro/Diva

Last edited by Manni01; 06-30-2020 at 02:28 AM.
Manni01 is offline  
post #670 of 764 Old 06-30-2020, 02:10 AM
AVS Forum Special Member
 
Iron Mike's Avatar
 
Join Date: Oct 2006
Posts: 1,081
Mentioned: 10 Post(s)
Tagged: 0 Thread(s)
Quoted: 190 Post(s)
Liked: 504
Quote:
Originally Posted by Manni01 View Post
If you genuinely want to be useful, please answer the specific questions I asked @liberator72 in my post. You have a K10A and an i1d3, like he does. They are VERY specific questions and have nothing to do with the color engine of the software solution, and everything to do with the meter limitations in low light.
If you have any experience with using an i1d3 specifically on LCOS, specific information would be welcome. The last time I tried one was ages ago, and I had not tried to optimise any speed settings. With the default settings, it was simply not good enough to be used on LCOS, with any calibration software.
Otherwise, that's just unneeded, unhelful and patronizing noise.

following this very grateful post, I can only advise u to start listening to audio books of movies, instead of trying to watch movies on calibrated screens.

calibration & profiling solutions: DCT v2, Lightspace, Spaceman, Calman, Argyll, ColorNavigator, basICColor
profiling & calibration workflow tools: Display Calibration Tools, Lightspace & Custom Color Patch Set & Gamma Calibration on Panasonic 65VT60
meter: CR-300, Klein K-10 A, i1Pro 2, i1D3
Iron Mike is offline  
post #671 of 764 Old 06-30-2020, 02:21 AM
AVS Forum Special Member
 
Manni01's Avatar
 
Join Date: Sep 2008
Posts: 9,747
Mentioned: 384 Post(s)
Tagged: 0 Thread(s)
Quoted: 5925 Post(s)
Liked: 6508
Quote:
Originally Posted by Iron Mike View Post
following this very grateful post, I can only advise u to start listening to audio books of movies, instead of trying to watch movies on calibrated screens.
What should I be grateful about exactly? I have been calibrating JVC projectors for more than 10 years, and I am exchanging information with pro-calibrators, software publishers and hardware manufacturers who find my advice useful in some specific areas.

You are providing zero information, are bad-mouthing other people who are providing some, are giving unsollicited patronizing advice and are not answering any questions (it looks you you haven't even read them).

I can only advise you start posting in the "chit chat" forums, instead of posting in a calibration thread.

I'm posting here for information, nothing else. Provide some, like @liberator72 does, and I'll be grateful.
mrtickleuk likes this.

Batch Utility V4.02 May 16 2019 to automate measurements files for madVR with support for BD Folders
JVC Macro feature on Vertex/Vertex2/Integral2/Maestro/Diva

Last edited by Manni01; 06-30-2020 at 02:26 AM.
Manni01 is offline  
post #672 of 764 Old 06-30-2020, 02:30 AM
AVS Forum Special Member
 
Light Illusion's Avatar
 
Join Date: Aug 2010
Posts: 2,186
Mentioned: 38 Post(s)
Tagged: 0 Thread(s)
Quoted: 1005 Post(s)
Liked: 1612
Quote:
Originally Posted by mikigio View Post
Hi guys, I have a question for AWP

AWP was obtained by matching with reference systems such as Pioneer Kuro, but then the xy values were purchased with high quality Spectroradiometers, but who owns only an EyeOne Pro2 Spectrophotometer in theory will be very far from the correct XY values or the distance is acceptable for which the AWP indicated is still valid?

I am interested in understanding this AWP issue because I find ColourSpace exceptional and the new FCVM system has allowed me to eliminate the Y problem on the Red (which also created problems in the evaluation of Magenta and Yellow) now even a short Lut of 1000 points has not worsened the skin color unlike what happened before (albeit in a very subtle measure), now if I solve the doubt AWP vs my modest EyeOne pro2 / iD3 I can try a wider Lut 3D.
The concept of an alternate white point is really based on Perceptual Matching, so a high-end spectro is not needed.
This is something we developed with FSI (Flanders Scientific Inc) for their high-end displays.

See: https://www.lightspace.lightillusion...our_match.html

And the reason the new FCVM process helps with the previous red error when using FCMM is the use of measured luma from the RGB patches, as well as the White patch.

We found the same improvement using FCMM, but with the luma values for RGB included in the calculation.
(But FCVM is better in other ways, so we have dropped FCMM totally.)

Steve
omarank likes this.

Steve Shaw
LIGHT ILLUSION

Light Illusion is offline  
post #673 of 764 Old 06-30-2020, 02:38 AM
AVS Forum Special Member
 
Light Illusion's Avatar
 
Join Date: Aug 2010
Posts: 2,186
Mentioned: 38 Post(s)
Tagged: 0 Thread(s)
Quoted: 1005 Post(s)
Liked: 1612
Manni - the biggest issue with the Discus is really its lack of a Sync capability.
That means it has to use a longer integration time to generate the same, stable, result you will get with an i1D3 on the same patch.
That alone can make the i1D3 a lot faster, especially as the i1D3 can switch Sync mode from period to frequency automatically using AIO mode.
Was AIO mode available when you last tested the probe?

Combining that with the Intelligent Integration option in ColourSpace/LightSpace will give the fasted possible profiling speed.
Unfortunately, this is not possible with the Discus.

Steve

Steve Shaw
LIGHT ILLUSION

Light Illusion is offline  
post #674 of 764 Old 06-30-2020, 03:00 AM
AVS Forum Special Member
 
Manni01's Avatar
 
Join Date: Sep 2008
Posts: 9,747
Mentioned: 384 Post(s)
Tagged: 0 Thread(s)
Quoted: 5925 Post(s)
Liked: 6508
Quote:
Originally Posted by Light Illusion View Post
Manni - the biggest issue with the Discus is really its lack of a Sync capability.
That means it has to use a longer integration time to generate the same, stable, result you will get with an i1D3 on the same patch.
That alone can make the i1D3 a lot faster, especially as the i1D3 can switch Sync mode from period to frequency automatically using AIO mode.
Was AIO mode available when you last tested the probe?

Combining that with the Intelligent Integration option in ColourSpace/LightSpace will give the fasted possible profiling speed.
Unfortunately, this is not possible with the Discus.

Steve
Hi Steve,

Thanks for this reply.

No, the AIO mode wasn't available when I evaluated the i1d3, it was soon after the initial release. That's why I'm ready to re-open the assessment.

However, I'm only interested in performance on LCOS, within the tolerance I have (I'm using an i1pro2 as a spectro, which shows that I'm only after near-reference, not reference results). The Discus is a bit slow in fastest mode, but accurate enough for me once profiled to the i1pro2. I have no complaints re accuracy / repeatability, at least on LCOS, within the meter's limit (for example, black can't be measured reliably more than a few minutes after a dark initialisation, as very near black drifts very quickly).

The main limitation I found on LCOS with the i1d3 at the time was regarding low light performance. Before evaluating one, I'm therefore trying to assess whether this main limitation is gone.

I've asked very precise questions in my post to @liberator72 in order to try to assess that.

If I was to run a 21points LUT with the i1d3 and optimal speed/integration settings, would it be able to read all the patches including the darkest ones reliably, given that my peak white for an SDR calibration is around 50nits and 5% white after a BT1886 calibration is around 0.05nits (would be lower with gamma 2.4).

I think it's a reasonable question to ask, because the Discus can do this without any problem, and none of the i1d3s I evaluated at the time was able to do that, which is why I went for the Discus in the first place.

I don't mind using the Discus for contrast measurements (as the i1d3 is definitely unable to read black), but if I can't use the i1d3 for large LUTs and 21pts (or more) greyscale measurements, there is no point using one.

So if anyone is using an i1d3 and is getting reliable results near black with a JVC projector targetting 50nits or so for an SDR calibration with at least 50,000:1 native on/off contrast), I'd love to get precise, actual data on near black measurements.

Otherwise I don't really want to waste the time (both mine and yours) trying one, as I ve done that a few years ago and it really wasn't up to the task.

Batch Utility V4.02 May 16 2019 to automate measurements files for madVR with support for BD Folders
JVC Macro feature on Vertex/Vertex2/Integral2/Maestro/Diva

Last edited by Manni01; 06-30-2020 at 03:07 AM.
Manni01 is offline  
post #675 of 764 Old 06-30-2020, 03:06 AM
aka jfinnie
 
Join Date: Aug 2015
Location: Norwich, UK
Posts: 4,208
Mentioned: 97 Post(s)
Tagged: 0 Thread(s)
Quoted: 3359 Post(s)
Liked: 2606
I think partnered with a Spectro the i1d3 can be a pretty good meter for projection use. Point the spectro at the screen and use the i1d3 facing the lens with the diffuser on; done like this it is quite possible to measure the black levels of the JVC with a Discus to a pretty high level of accuracy and can be driven very fast (position the i1d3 so that uncorrected you get >1000nits for peak white).

Of course, folk looking at TVs don't get to do that "trick" so to an extent have to chase better and better probes (though for the black king OLED displays the Discus has other issues due to poor matching to CIE standard observer of its filters).
Manni01 and mrtickleuk like this.
bobof is online now  
post #676 of 764 Old 06-30-2020, 03:14 AM
AVS Forum Special Member
 
Manni01's Avatar
 
Join Date: Sep 2008
Posts: 9,747
Mentioned: 384 Post(s)
Tagged: 0 Thread(s)
Quoted: 5925 Post(s)
Liked: 6508
Quote:
Originally Posted by bobof View Post
I think partnered with a Spectro the i1d3 can be a pretty good meter for projection use. Point the spectro at the screen and use the i1d3 facing the lens with the diffuser on; done like this it is quite possible to measure the black levels of the JVC with a Discus to a pretty high level of accuracy and can be driven very fast (position the i1d3 so that uncorrected you get >1000nits for peak white).

Of course, folk looking at TVs don't get to do that "trick" so to an extent have to chase better and better probes (though for the black king OLED displays the Discus has other issues due to poor matching to CIE standard observer of its filters).
Thanks, I used to do this with the C5 at the time, but I don't like the idea of having two tripods etc. Set-up time takes too long, I'd have to reprofile every time (with the Discus it's always at the same position, aiming at the exact same point on the screen, so I can re-use the same meter profile). Once I've checked it if I was to use one, I plan to use the i1d3 without profiling to the i1pro2 (as it's impossible to know which one is closer to reference on LCOS using the projector profile), but whichever meter replacing my Discus will be facing the screen, and ideally with laser/LED, because it's one of the very convenient features I appreciate in the Discus. So it looks like a K10A rather than an i1d3, but given the price difference, I'd like to rule out the i1d3 as a possible compromise.

Batch Utility V4.02 May 16 2019 to automate measurements files for madVR with support for BD Folders
JVC Macro feature on Vertex/Vertex2/Integral2/Maestro/Diva
Manni01 is offline  
post #677 of 764 Old 06-30-2020, 03:41 AM
aka jfinnie
 
Join Date: Aug 2015
Location: Norwich, UK
Posts: 4,208
Mentioned: 97 Post(s)
Tagged: 0 Thread(s)
Quoted: 3359 Post(s)
Liked: 2606
Quote:
Originally Posted by Manni01 View Post
Thanks, I used to do this with the C5 at the time, but I don't like the idea of having two tripods etc. Set-up time takes too long, I'd have to reprofile every time (with the Discus it's always at the same position, aiming at the exact same point on the screen, so I can re-use the same meter profile). Once I've checked it if I was to use one, I plan to use the i1d3 without profiling to the i1pro2 (as it's impossible to know which one is closer to reference on LCOS using the projector profile), but whichever meter replacing my Discus will be facing the screen, and ideally with laser/LED, because it's one of the very convenient features I appreciate in the Discus. So it looks like a K10A rather than an i1d3, but given the price difference, I'd like to rule out the i1d3 as a possible compromise.
I must say I keep going round the getting better probes thing myself, and am not sure with the availability of such techniques it is merited for projection.

One trick that can be useful is to pick a position for the i1d3 that works for the luma ranges and shrink the test pattern window to just cover the diffuser. You can quite easily find basically the same spot then (though on the JVC units not sure how much of a big deal this really is as uniformity in the centre of the screen is very good indeed). This should still give a reasonable size target to hit on the screen. On the Radiance Pro with LS / CS in my room the "Small" pattern works well for this.

But yes, the two tripods is a bit of a faff - I think our rooms are similar sizes and it certainly clutters up the place, I've taken to setting up and leaving the cinema for the comfort of the lounge with the laptop (I have a NUC in the cinema that I run the calibration software on via Teamviewer). I have a couple of photographic tripods with the same quick release plates I use - a tall and a short(er) Manfrotto unit. Is the faff worth the cost of a Klein?... Hrmph...
Manni01 likes this.
bobof is online now  
post #678 of 764 Old 06-30-2020, 03:59 AM
AVS Forum Special Member
 
Manni01's Avatar
 
Join Date: Sep 2008
Posts: 9,747
Mentioned: 384 Post(s)
Tagged: 0 Thread(s)
Quoted: 5925 Post(s)
Liked: 6508
Quote:
Originally Posted by bobof View Post
I must say I keep going round the getting better probes thing myself, and am not sure with the availability of such techniques it is merited for projection.

One trick that can be useful is to pick a position for the i1d3 that works for the luma ranges and shrink the test pattern window to just cover the diffuser. You can quite easily find basically the same spot then (though on the JVC units not sure how much of a big deal this really is as uniformity in the centre of the screen is very good indeed). This should still give a reasonable size target to hit on the screen. On the Radiance Pro with LS / CS in my room the "Small" pattern works well for this.

But yes, the two tripods is a bit of a faff - I think our rooms are similar sizes and it certainly clutters up the place, I've taken to setting up and leaving the cinema for the comfort of the lounge with the laptop (I have a NUC in the cinema that I run the calibration software on via Teamviewer). I have a couple of photographic tripods with the same quick release plates I use - a tall and a short(er) Manfrotto unit. Is the faff worth the cost of a Klein?... Hrmph...
Yeah, I also use the 1% size pattern trick to aim the i1pro (and the Discus), but the two tripods (I have them, with quick release and I use TeamViewer or Remote Desktop too) isn't an option for me. I had to do it with the JVC Autocal at the time and I'm glad I don't have to do it anymore (both because of the Spyder X and because I'm not using the JVC Autocal anymore). My seats are where the tripod should be, and yes, for me I'd rather pay for a K10A than keep faffing around with one meter facing the screen and one facing the PJ. I'd get reliable contrast measurements, laser aiming, excellent low light accuracy, increased speed, no dark initialisation, etc. I've been resisting for as long as I could, but I'm getting to a stage where I'm more and more tempted, especially since I'm now using profile off with the filter, which provides great results but needs larger LUTs than when using presets (with the known downsides). If, as I suspect, measuring with the diffuser facing the PJ is the only way to use the i1d3 on LCOS reliably, then it's not an option for me. I could have done that then (as I said I used to do this with the C5 and the i1pro), but I went for the Discus instead. I'm not going to do it now
bobof likes this.

Batch Utility V4.02 May 16 2019 to automate measurements files for madVR with support for BD Folders
JVC Macro feature on Vertex/Vertex2/Integral2/Maestro/Diva

Last edited by Manni01; 06-30-2020 at 04:05 AM.
Manni01 is offline  
post #679 of 764 Old 06-30-2020, 04:11 AM
AVS Forum Special Member
 
Iron Mike's Avatar
 
Join Date: Oct 2006
Posts: 1,081
Mentioned: 10 Post(s)
Tagged: 0 Thread(s)
Quoted: 190 Post(s)
Liked: 504
Quote:
Originally Posted by Manni01 View Post
if you actually read what people write

don't worry, I've read the posts. my answers stand exactly as such. There's no need to re-post if I said it already three times, three different ways. There's a few other folks following in here, they deserve proper info.

u however can believe whatever u want, though.

Quote:
Originally Posted by Manni01 View Post
My projector is drifting so quickly that I need to recalibrate it every 200 hours, at the very least. Why on earth would I want to get a 0.5 dE accuracy when within a few hours the drift of the projector itself will be far beyond that? Still, I want a good calibration, without artifacts, and I'm exploring various ways to reach that.
u contradict urself. Next, I only know calibrators that always want the best result in each cal session, so that even w/ fast drift dE stays as low as possible.

Quote:
Originally Posted by Manni01 View Post
Anyone laughing at my remark about color engine hasn't read my posts either, or doesn't know that they are talking about. My questions are specifically about the meter capability in low light. They would apply when measuring a 21pts greyscale. There is nothing about a patch sequence here, or a LUT engine here.
it's not a "LUT engine" and patch sets and grid sequences aren't "LUTs" either, which is what u posted various times today... (?)

understand the minimum of this: all of these things are connected, and weakest link in the chain ultimately limits the final result. no matter how much the other parts in the chain overcompensate, it won't change any of this.

Quote:
Originally Posted by Manni01 View Post
this means that the i1d3 would be about 30% faster, while being less accurate than the Discus in low light. I assume that the K10a would be about 50% faster (15 minutes?) while being more accurate in low light than the Discus. The main question for me is how accurate would the i1d3 be with a 21points LUT, when reading 5% stim RGBW on LCOS.
as I explained before: none of the colorimeters are accurate, so ur entire hierarchy is pointless. understand this:

(1) re accuracy: it doesn't matter which colorimeter u use
(2) u need proper ref offset data for all of the meters w/ the best offset variant (that is available in ur solution)
(3) re speed: if that is of concern, K10A is ur choice
(4) re repeatability: that is the most important aspect, and (1), (2) and (3) do NOT affect or change that.... K10A followed by CR100 are king here, which is why Pros have these (besides low light and speed).

if u do not have (4), then (2) is way less effective. which is exactly what I explained before.

some folks here claim they can bring the i1D3 close to K10A levels re (4). I ask for stats, they have none.

What they post are synthetic calculations that do not reflect the real world, it only shows that the offset variant works on a static data set - a one time data capture.

but in the real world where real calibrators live, the i1D3 will not provide the same data repeatably as the K10A does (unless u severely sacrifice speed and re-read multiple times), hence what was captured in the probe offsets is obviously not the same reads u will do w/ the i1D3 during a long profile (hence the offsets are off, and therefore the math will be off)..... read this again, understand this.

doing synthetic calcs on static data sets only proves the math of the offsets variant is somewhat accurate. it does NOT in any way shape or form prove repeatability, which is why I've been asking multiple times for data that proves this, until it was clear it's clown business statements.

like I said: u and nobody else can have it both ways. but w/ extra work and smart patch sets u can scrape a lot of useful data from the i1D3.


Another thing u should seriously consider is (assuming u can only afford one high-end meter) i1D3 + CR250/Jeti. Both of ur current meters are compromises in Pro standards.
ht_pete likes this.

calibration & profiling solutions: DCT v2, Lightspace, Spaceman, Calman, Argyll, ColorNavigator, basICColor
profiling & calibration workflow tools: Display Calibration Tools, Lightspace & Custom Color Patch Set & Gamma Calibration on Panasonic 65VT60
meter: CR-300, Klein K-10 A, i1Pro 2, i1D3

Last edited by Iron Mike; 06-30-2020 at 04:26 AM.
Iron Mike is offline  
post #680 of 764 Old 06-30-2020, 04:40 AM
AVS Forum Special Member
 
Manni01's Avatar
 
Join Date: Sep 2008
Posts: 9,747
Mentioned: 384 Post(s)
Tagged: 0 Thread(s)
Quoted: 5925 Post(s)
Liked: 6508
Quote:
Originally Posted by Iron Mike View Post
don't worry, I've read the posts. my answers stand exactly as such. There's no need to re-post if I said it already three times, three different ways. There's a few other folks following in here, they deserve proper info.

u however can believe whatever u want, though.

u contradict urself. Next, I only know calibrators that always want the best result in each cal session, so that even w/ fast drift dE stays as low as possible.

it's not a "LUT engine" and patch sets and grid sequences aren't "LUTs" either, which is what u posted various times today... (?)

understand the minimum of this: all of these things are connected, and weakest chain in the link ultimately limits the final result. no matter how much the other parts in the chain overcompensate won't change any of this.

as I explained before: none of the colorimeters are accurate, so ur entire hierarchy is pointless. understand this:

(1) re accuracy: it doesn't matter which colorimeter u use
(2) u need proper ref offset data for all of the meters w/ the best available offset variant (in ur solution)
(3) re speed: if that is of concern, K10A is ur choice
(4) re repeatability: that is the most important aspect, and (1), (2) and (3) do NOT affect or change that.... K10A followed by CR100 are king here, which is why Pros have these (besides low light and speed).

if u do not have (4), then (2) is way less effective. which is exactly what I explained before.

some folks here claim they can bring the i1D3 close to K10A levels re (4). I ask for stats, they have none.

What they post are synthetic calculations that do not reflect the real world, it only shows that the offset variant works on a static data set - a one time data capture.

but in the real world where real calibrators live, the i1D3 will not provide the same data repeatably as the K10A does, hence what was captured in the probe offsets is obviously not the same reads u will do w/ the i1D3 during a long profile..... read this again, understand this.

doing synthetic calcs on static data sets only proves the math of the offsets variant is somewhat accurate. it does NOT in any way shape or form prove repeatability, which is why I've been asking for data that proves this multiple times, until it was clear it's clown business statements.

like I said: u and nobody else can have it both ways. but w/ extra work and smart patch sets u can scrape a lot of useful data from the i1D3.


Another thing u should seriously consider is (assuming u can only afford one high-end meter) i1D3 + CR250/Jeti. Both of ur current meters are compromises in Pro standards.
Again, zero useful information. Only the ABC in calibration: Wow, you need a spectro to make a colorimeter more accurate! Wow, the i1pro2 isn't a reference spectro! Wow, the K10a and the CR100 are more accurate and provide more repeatable measurements! Amazing, brand new information, that I'm sure benefits a lot of people in this thread... Seriously, you need to learn how to read.

Please tell me something that I don't know regarding the current version of the i1d3, which is how low it can read reliably on LCOS when facing the screen with optimised speed settings in Colourspace (if you can provide these as well, that's only better). That's one simple, factual question. I want information. Once I have this information, I'll make my own decision, thank you very much. I am not expecting the i1d3 to be as fast, accurate or to provide measurements as repeatable as the K10a. Just as I'm not expecting the i1pro2 to be as accurate as a CR250. I'm only trying to assess if the i1d3 would be usable to me, and would provide usable results in my set-up and improve on my Discus results (which I'd need to keep for contrast measurements), for about 5% of the cost of a K10a.

The rest, I knew years ago when I sent back the i1d3s that I evaluated at the time. Again, I don't need your basic unsollicited "advice". There is no information in your post. Zero. Zilch. Nada. Rien.

I said many times that I'm happy with a near-reference spectro such as the i1pro2. I would even be happy with an uncorrected i1d3 if it measures close enough to my i1pro2 given that there is no way to know on LCOS which one is closer to reference (unless you compare them to a reference meter, obviously). The i1pro2 (or a good i1d3) is "good enough" for me. I have zero interest in a CR250. I'm after a better, faster colorimeter. I don't need to be any closer to reference than what my i1pro2 (or a good 1id3) provides on LCOS.

Your "advice" in this discussion is what anyone who knows anything about calibration already knows. I'm asking for SPECIFIC information. If you don't want or are unable to provide it, why are you wasting bandwidth (and my time)?

Batch Utility V4.02 May 16 2019 to automate measurements files for madVR with support for BD Folders
JVC Macro feature on Vertex/Vertex2/Integral2/Maestro/Diva

Last edited by Manni01; 06-30-2020 at 05:00 AM.
Manni01 is offline  
post #681 of 764 Old 06-30-2020, 04:47 AM
aka jfinnie
 
Join Date: Aug 2015
Location: Norwich, UK
Posts: 4,208
Mentioned: 97 Post(s)
Tagged: 0 Thread(s)
Quoted: 3359 Post(s)
Liked: 2606
Quote:
Originally Posted by Manni01 View Post
Yeah, I also use the 1% size pattern trick to aim the i1pro (and the Discus), but the two tripods (I have them, with quick release and I use TeamViewer or Remote Desktop too) isn't an option for me. I had to do it with the JVC Autocal at the time and I'm glad I don't have to do it anymore (both because of the Spyder X and because I'm not using the JVC Autocal anymore). My seats are where the tripod should be, and yes, for me I'd rather pay for a K10A than keep faffing around with one meter facing the screen and one facing the PJ. I'd get reliable contrast measurements, laser aiming, excellent low light accuracy, increased speed, no dark initialisation, etc. I've been resisting for as long as I could, but I'm getting to a stage where I'm more and more tempted, especially since I'm now using profile off with the filter, which provides great results but needs larger LUTs than when using presets (with the known downsides). If, as I suspect, measuring with the diffuser facing the PJ is the only way to use the i1d3 on LCOS reliably, then it's not an option for me. I could have done that then (as I said I used to do this with the C5 and the i1pro), but I went for the Discus instead. I'm not going to do it now
Yup, very similar sized rooms. My couch is equally in the wrong place for this to be much "fun". Luckily in my case the two tripod legs work out just right and nestle into the crevice between the couch base and couch back nicely.
Manni01 likes this.
bobof is online now  
post #682 of 764 Old 06-30-2020, 05:42 AM
AVS Forum Addicted Member
Moderator
 
rboster's Avatar
 
Join Date: Jul 2000
Location: USA
Posts: 26,100
Mentioned: 78 Post(s)
Tagged: 0 Thread(s)
Quoted: 1841 Post(s)
Liked: 2793
Bickering posts/attacks removed. I would suggest using the ignore feature.

FS: RUIPRO 8K Fiber DisplayPort Cable 33 feet : https://www.avsforum.com/forum/211-a...d-support.html
rboster is online now  
post #683 of 764 Old 06-30-2020, 06:44 AM
Senior Member
 
ebr9999's Avatar
 
Join Date: Jun 2013
Posts: 464
Mentioned: 18 Post(s)
Tagged: 0 Thread(s)
Quoted: 262 Post(s)
Liked: 178
Quote:
Originally Posted by liberator72 View Post
Following up on this from yesterday. I accept that there was something maybe "lost in translation" but I must say that at the time I did feel like I was being called out as being dishonest even though it now clear to me that was obviously not intended.

So, I have sent the files as requested to ebr9999, but I have also done myself what he wants to do, which was to take the FCMM profile from LightSpace and invert the correction from it to provide uncorrected profile data (with the help of Anger.miki). That uncorrected profile data has then been imported to ColourSpace where FCVM has been applied to it using the same original BPD files.

Note. The FCVM profile has 1 extra point within it as I removed Black from the Patch Sequence to speed up Jeti read times (it can take 2 minutes to read black). LightSpace does not add this point automatically, whereas ColourSpace adds Black if it is missing upon saving a profile.

First, the original FCMM profile compared to a Jeti Profile


Then, that EXACT same profile data, but with FCVM applied and compared to the EXACT same Jeti Profile


Once again, the original post was intended to be used as a "rough guide" as to the benefits for i1d3 users. This can be considered as "real world usage" benefits to i1d3 users
Many thanks, Leon for having shared your measurements data, and I agree, I can see a lot of benefit into your I1D3 accuracy.

Here my analysis:
I have first given a look to your FCMM correction. If you correct target using FCMM, chroma is obviously perfect, but that I1D3 FCMM correction (matched on W), has the following Luma Error on RGB: 4.34%,-2.23%,-0.03%.
I have given a look to my one, and typically I see only around 2% on R. I would expect even better outcomes in my case.

Anyhow,
I have computed for each measured patch dx, dy, dxy, and %Y relative to FCMM and FCVM. Then I have computed the relevant difference between FCVM and FCMM for dxy and %Y. I have them coloured according to NIST requirements. Keep in mind that they are very strict, and, for example, Klein use more relaxed requirements (0.002 and 3% on RGB)
The first thing you note is that there is no dispute about Luma. As somehow expected FCVM is much better than FCMM.
Looking at primaries you may not be so impressed:

but miracles are not possible. dY%'s are correct and dx and dy are kept.
What has impressed me are secondaries:

Beyond dY%, now also chromas appear pretty good.
I have then tried to isolate critical patches (where I think a little devil may destroy everything). I have selected the patches whose FCMM chroma correction is better than 0.001:

The only noticeable patch (105,23,13) shows some misalignment (essentially a worse dy), but surely in the overall balance not so important.


PS: Also the excel file is attached. The selector I usually use when checking those kind of file are there for your use.
Attached Thumbnails
Click image for larger version

Name:	Leon I1D3 - Primary.png
Views:	176
Size:	42.5 KB
ID:	2751104   Click image for larger version

Name:	Leon I1D3 - secondaries.png
Views:	174
Size:	86.7 KB
ID:	2751106   Click image for larger version

Name:	Leon I1D3 - other out.png
Views:	174
Size:	48.4 KB
ID:	2751108  
Attached Files
File Type: zip Template 10x10x10 4 FCMM & FCVM Leon data.zip (668.8 KB, 6 views)
Manni01, jrref and mrtickleuk like this.

TV: LG 55E8PLA, Pana TX-L37ET5E BD/DVR: Panasonic BMR-BWT735 Media player: Teufel Connector 2, HT Rec: ANTHEM MRX 510,
Front: Woofer&Tweeter, Center: 2Woofer&1Tweeter, Surround: Bose acoustimass 5 series II, Sub: SVS SB2000
ebr9999 is offline  
post #684 of 764 Old 06-30-2020, 09:14 AM
Advanced Member
 
liberator72's Avatar
 
Join Date: Mar 2018
Location: Sunny (when it's not raining) Devon, U.K.
Posts: 830
Mentioned: 53 Post(s)
Tagged: 0 Thread(s)
Quoted: 609 Post(s)
Liked: 1147
Quote:
Originally Posted by Manni01 View Post
@liberator72 Thank you for sharing your results, your posts were perceived as intended the first time. And apologies for adding to the noise by asking you to confirm something that I had missed in your initial post.

I have done a quick test with the Discus and I measure a standard anisometric 10x10x10 in 28 minutes vs your 20 minutes (in my slowest SDR calibration with a peak white of 50nits). If this translated to LCOS, this means that the i1d3 would be about 30% faster, while being less accurate than the Discus in low light. I assume that the K10a would be about 50% faster (15 minutes?) while being more accurate in low light than the Discus.

The main question for me is how accurate would the i1d3 be with a 21points LUT, when reading 5% stim RGBW on LCOS. When I evaluated the i1d3 a few years back, it was not only unable to read black reliably (like the Discus is under certain conditions), which makes it impossible to use for BT1886 or contrast measurements, but it was unable to read 5% white off the screen with my JVC projector at the time.

In the low end, with your "turbo" settings, how low do you expect the i1d3 to read reliably? Is it able to read the darkest patches of a 21pts LUT reliably with these settings? What is your peak white? What is your 5% white Y?

Currently, with the Discus, it takes around 2.5 hours to run a 17pts LUT for HDR tonemapped content (120 peak nits, 5% white around 0.17nits) without low light averaging, and a lot longer with my calibration for SDR content (currently 54nits peak white, 5% white around 0.05nits after calibration using BT1886, would be lower with power 2.4), up to five hours with low light averaging enabled. I currently use an optimised patch sequence of under 4,000 points that runs under 2 hours for my HDR calibration with the Discus profiled to the i1pro2 (I haven't had a chance to try FCVM yet but on LCOS FCCM already worked very well, so I'm not expecting huge improvements) and provides excellent results with CS.

If your results translated to a 30% speed increase here, it would allow me to read 5,000 patches in a bit more than 1.5 hours with the i1d3, which makes a 21pts LUT just about doable at around 3 hours, but only if the results are usable, especially in the low end.

Have you tried measuring a 21pts LUT with your turbo settings, and is your i1d3 reliable reading the darkest patches? Have you tried a calibration with a peak white as low as 50nits (which is the standard for SDR with projectors in a dedicated room) and patches going down as low at 0.05nits for white?

Thank you again for sharing all your results, I don't own an i1d3 currently but your results with turbo settings make me think about getting one to run large LUTs faster than I can with my Discus.

I'd love to not have to upgrade to a K10a, but despite the significant speed improvements that you report with your "turbo" settings in your setup (I assume targeting 100nits in SDR), I'm not sure that an i1d3 would work reliably here with 21pts LUTs, especially with my SDR calibration (50nits peak white, 0.05nits for 5% white).

That's the reason why I had to go with the Discus rather than the i1d3, and I'm concerned that this limitation might still stand given the limited ability of the i1d3 in low light.

Thoughts?
Okay so this is absolutely not being directed towards you as it is clear you actually read and understood what I said. But the rest of what has been said today and over the past couple days is getting pretty boring and is completely unwarranted.

I never once said the i1d3 can be turned into a Klein with some cooked up settings. I have repeatedly said that it is just tweaking of the normal settings that are available to anyone with an i1d3 Rev.B meter. They are there, right in the settings menu for all to see. They have always been there. Anyone can do what they want with them and if they work, then great it can make you i1d3 much faster and in fact can make it comparable to a Klein over the 1000 points that I showed. It doesn't magically make the i1d3 able to read any lower than it already could. It doesn't make it any more accurate than it already is. It doesn't make it any more repeatable than it already is. It doesn't do anything other than shorten the read time from whatever Integration time you usually set it at (generic default advice is 0.75 or 750ms) to 0.25 or 250ms. That is quite a bit of gain when measuring over a substantial amount of patch points (note, when used in conjunction with AIO Mode). Then, to ensure that the low light readings it is capable of measuring are stable, just set a threshold for where Intelligent Integration kicks in. Whether that be 2 nits, 5 nits or whatever.

In LightSpace there is a Measure and Log feature (not available yet in CS, but it was one of the first things I requested on WIBNI list during Alpha testing) where you can measure and record whatever you want and check for repeatability of measurements in a CSV log file. On my WOLED panel, with my i1d3, in my environment with my setup, setting Intelligent Int at 2 nit means it will take reads until they are repeatable (up to the meters maximum Integration time) and then move on to the next patch. Very simple.

With my setup, everything measured above 1.5 nits is fine, repeatable and stable with the shortened Integration time of 250ms, whereas below 1.5 nits it is not as repeatable. So to be safe, I set it at 2 nits. Again, very simple.

That is all there is to it and I absolutely for the life of me cannot understand how some people cannot see that, and then go off on some wild mission to twist what was said into something it isn't. I have no interest in justifying my purchase of the i1d3, if anything I would want to justify my purchase of the Klein and say that the i1d3 is rubbish and all should ditch theirs and get a Pro meter so that I'd feel my expensive purchase was not made in vain. My initial first post and subsequent reply to another user was made due to the fact that many/most here use the i1d3 and that the new FCVM probe match method in ColourSpace works very, very well and happened to mention that if you tweak your i1d3 settings you can make it much faster without losing any of its accuracy. It really is that simple, but maybe too simple for some to understand

One final thing on that subject is that they are not "my" turbo settings. This was all discussed a long time ago in the LightSpace thread where people were trying to get things to work faster. Several tried it and it worked for most/all of them IIRC. It is OLD NEWS!

Anyway, sorry for the rant. And again, not directed towards you as it is clear you understood what I said, or at least attempted to and didn't try and twist anything.

To answer some of your questions.

Yes, I have used the i1d3 to do a 21p cube several times for some tests and experiments I was doing a while back just for fun. If I recall, it takes about 30-35 minutes longer than the Klein when using the i1d3 settings above. It was before the Start/End time box and patch counter etc was added in ColourSpace so I timed it roughly and I am sure (with my hazy recollection) it was 30 (ish) minutes longer over 9261 points. Not bad at all I thought!!!

It cannot, and will not ever read the darkest patches any more reliably than it already does no matter what settings yo use. It is hardware. You cannot change the physical characteristics of it. It does not matter if you set Integration at 0.25 or 6.0. It measures what it measures at all times. It will take measurements based on the time you set your Integration value at and with an Intelligent Int value set, it will slow down and "Intelligently" take reads until it is stable and move on as explained above. The key is to set the correct Integration time in the first place. The faster the better obviously, but it needs to be good reads so you can either make repeatable measurements and check for consistency, or use LightSpace Measure and Log function.

I have never calibrated a Projector and do not foresee myself ever having to. I have one SDR picture mode at 80 nits (night viewing when I am tired), one at 100 nits (general normal dark room viewing) and one at 120 nits (day mode). I have done all three of these multiple times with the i1d3 before I got my Klein and achieved great results when using either a 17p or a 21p cube.

In your case, measuring off screen for a projector, I doubt (although I do not know) if you would be able to reduce the Integration time that low, so no, I do not think (nor do I know) if you would see any benefit from using an i1d3 in the way I have described. But this is why I said repeatedly that users would have to test for themselves with their own setup in their own environment, and there are very, very simple ways to do this as I have said above.

If you can justify the cost of a Klein (or CR probe) then absolutely, get one. It is fantastic and I do not regret for one moment buying mine. But I absolutely love the (two) i1d3 probes I have and they are more than worth their money (IMO) and are excellent little probes. Whether the i1d3 would be of any benefit to you, I really could not honestly give an opinion.

Sorry for the late reply, I have been at work all day, but I really hope this clarifies everything fully so as to prevent any more of totally unwarranted, obnoxious and somewhat offensive posts over the past few days.
Manni01, wl1, sat4all and 2 others like this.

LG OLED65C9MLB**Denon AVR-X3500H**Focal Sib Evo Atmos 5.1.2**Sky Q UHD**Xbox One X**Panasonic DP-UB820**ATV 4K
Klein K10-A**Jeti 1501**i1D3 OEM**ColourSpace HTX**LightSpace HTP**CalMAN Home for LG**DVDO AVLab TPG**HD Fury Integral 2
Ted's Disk & Media Files**S & M HD & UHD HDR Benchmark Disks
liberator72 is online now  
post #685 of 764 Old 06-30-2020, 09:26 AM
AVS Forum Special Member
 
Manni01's Avatar
 
Join Date: Sep 2008
Posts: 9,747
Mentioned: 384 Post(s)
Tagged: 0 Thread(s)
Quoted: 5925 Post(s)
Liked: 6508
@liberator72 : That's perfect, thanks for taking the time to answer my question, and I regret as much as you do the obnoxious tangent taken by some. Your posts were perfectly clear from the start.

Just to clarify my own query, I never thought that any setting or matrix would change the physical limitations of the meter. It's just that the last time I evaluated an i1d3, it didn't meet my requirements regarding low light measurements, and quite a few things have changed since (new meter revision, AIO mode, optimised settings for speed etc) so I was just interested in finding out if I should give it another try. It looks like I should (even if only for my HDR content calibration which is 120nits and the one I care about the most), though the odds that it will meet my specific needs for my SDR calibration remain limited.

I'm definitely considering the K10A (and have been for years), but I'm not there yet.

Thanks again for sharing your results and for taking the time to answer my question to the best of your ability and for your suggestions. It's very useful and I am grateful.
liberator72 likes this.

Batch Utility V4.02 May 16 2019 to automate measurements files for madVR with support for BD Folders
JVC Macro feature on Vertex/Vertex2/Integral2/Maestro/Diva

Last edited by Manni01; 06-30-2020 at 09:36 AM.
Manni01 is offline  
post #686 of 764 Old 06-30-2020, 09:36 AM
Advanced Member
 
liberator72's Avatar
 
Join Date: Mar 2018
Location: Sunny (when it's not raining) Devon, U.K.
Posts: 830
Mentioned: 53 Post(s)
Tagged: 0 Thread(s)
Quoted: 609 Post(s)
Liked: 1147
Quote:
Originally Posted by Manni01 View Post
@liberator72 : That's perfect, thanks for taking the time to answer my question, and I regret as much as you do the obnoxious tangent taken by some. Your posts were perfectly clear from the start.

Just to clarify, I never thought that any setting or matrix would change the physical limitations of the meter. It's just that the last time I evaluated an i1d3, it didn't meet my requirements, and quite a few things have changed since (new meter revision, AIO mode, etc) so I was just interested in finding out if I should give it another try. It looks I should, though the odds that it will meet my specific needs remain limited.

I'm definitely considering the K10A (and have been for years), but I'm not there yet.

Thanks again for sharing your results and for taking the time to answer my question to the best of your ability. It's very useful and I am grateful.
Yes, I fully realise that you did not think at all that anything I said meant the physical limitations of the meter could be changed, as it seems you are fully capable of reading, understanding and writing the English language using full sentences and complete words. But I am sure you understand why I had to make that absolutely clear as it appears some people here cannot.

In my absolute honest opinion, for the price of an i1d3 Rev.B, it should probably be considered as an essential piece of equipment in any persons toolkit, be they pro's or hobbyists. Even if you only ever use it once in a blue moon, it is extremely handy to be able to pull it out and know it just works.
SamuriHL, Manni01 and mrtickleuk like this.

LG OLED65C9MLB**Denon AVR-X3500H**Focal Sib Evo Atmos 5.1.2**Sky Q UHD**Xbox One X**Panasonic DP-UB820**ATV 4K
Klein K10-A**Jeti 1501**i1D3 OEM**ColourSpace HTX**LightSpace HTP**CalMAN Home for LG**DVDO AVLab TPG**HD Fury Integral 2
Ted's Disk & Media Files**S & M HD & UHD HDR Benchmark Disks
liberator72 is online now  
post #687 of 764 Old 06-30-2020, 09:39 AM
AVS Forum Special Member
 
Manni01's Avatar
 
Join Date: Sep 2008
Posts: 9,747
Mentioned: 384 Post(s)
Tagged: 0 Thread(s)
Quoted: 5925 Post(s)
Liked: 6508
Quote:
Originally Posted by liberator72 View Post
Yes, I fully realise that you did not think at all that anything I said meant the physical limitations of the meter could be changed, as it seems you are fully capable of reading, understanding and writing the English language using full sentences and complete words. But I am sure you understand why I had to make that absolutely clear as it appears some people here cannot.

In my absolute honest opinion, for the price of an i1d3 Rev.B, it should probably be considered as an essential piece of equipment in any persons toolkit, be they pro's or hobbyists. Even if you only ever use it once in a blue moon, it is extremely handy to be able to pull it out and know it just works.
No worries, I absolutely understand.

I'm not snobbish about the tools I use, I had to buy Spyders 4, 5 and X which are horrible meters but were extremely useful with the JVC Autocal (when I used that method) to get the best possible baseline, even if I would do my main calibration with a Discus trained to the i1pro2. I will consider keeping the i1d3 if I can make use of it.

I'll post back my findings when I've had a chance to run some tests with one.

Batch Utility V4.02 May 16 2019 to automate measurements files for madVR with support for BD Folders
JVC Macro feature on Vertex/Vertex2/Integral2/Maestro/Diva

Last edited by Manni01; 06-30-2020 at 09:50 AM.
Manni01 is offline  
post #688 of 764 Old 06-30-2020, 09:58 AM
Advanced Member
 
liberator72's Avatar
 
Join Date: Mar 2018
Location: Sunny (when it's not raining) Devon, U.K.
Posts: 830
Mentioned: 53 Post(s)
Tagged: 0 Thread(s)
Quoted: 609 Post(s)
Liked: 1147
Quote:
Originally Posted by ebr9999 View Post
Many thanks, Leon for having shared your measurements data, and I agree, I can see a lot of benefit into your I1D3 accuracy.

Here my analysis:
I have first given a look to your FCMM correction. If you correct target using FCMM, chroma is obviously perfect, but that I1D3 FCMM correction (matched on W), has the following Luma Error on RGB: 4.34%,-2.23%,-0.03%.
I have given a look to my one, and typically I see only around 2% on R. I would expect even better outcomes in my case.
One thing that should be noted about this for full disclosure.

Usually, when performing a probe match you obviously validate the match. Previously, when using FCMM with the i1d3 on a WOLED, you can sometimes fight like hell to get a NIST valid match, and sometimes you never will on a WOLED in Wide gamut. xy is usually always correct with RGBW, and Y is always within NIST for GBW, whereas Red can struggle with Y. Sometimes I can get it NIST valid but usually, as you state above, it averages around 2%.

But...The correction used above was done very quickly with FCVM and as you can see, RGB are all perfectly well within NIST. However, when that correction data is used with the traditional FCMM method, the errors are much larger than they usually are.

What I'm trying to say is, with FCMM on WOLED, it can be a struggle with the i1d3 to get a good (NIST valid) probe match on WOLED, whereas with FCVM it easy, straightforward and works.

If I get the time in the near future and if other have an interest in me doing it, I will perform the FCMM in LightSpace so as to get that match as good as I possibly can, and with that BPD data, use it for FCVM too. As Enrico says above, it could potentially lead to even further benefits!
Manni01, jrref, omarank and 1 others like this.

LG OLED65C9MLB**Denon AVR-X3500H**Focal Sib Evo Atmos 5.1.2**Sky Q UHD**Xbox One X**Panasonic DP-UB820**ATV 4K
Klein K10-A**Jeti 1501**i1D3 OEM**ColourSpace HTX**LightSpace HTP**CalMAN Home for LG**DVDO AVLab TPG**HD Fury Integral 2
Ted's Disk & Media Files**S & M HD & UHD HDR Benchmark Disks
liberator72 is online now  
post #689 of 764 Old 06-30-2020, 10:24 AM
aka jfinnie
 
Join Date: Aug 2015
Location: Norwich, UK
Posts: 4,208
Mentioned: 97 Post(s)
Tagged: 0 Thread(s)
Quoted: 3359 Post(s)
Liked: 2606
Quote:
Originally Posted by Manni01 View Post
@liberator72 : That's perfect, thanks for taking the time to answer my question, and I regret as much as you do the obnoxious tangent taken by some. Your posts were perfectly clear from the start.

Just to clarify my own query, I never thought that any setting or matrix would change the physical limitations of the meter. It's just that the last time I evaluated an i1d3, it didn't meet my requirements regarding low light measurements, and quite a few things have changed since (new meter revision, AIO mode, optimised settings for speed etc) so I was just interested in finding out if I should give it another try. It looks like I should (even if only for my HDR content calibration which is 120nits and the one I care about the most), though the odds that it will meet my specific needs for my SDR calibration remain limited.

I'm definitely considering the K10A (and have been for years), but I'm not there yet.

Thanks again for sharing your results and for taking the time to answer my question to the best of your ability and for your suggestions. It's very useful and I am grateful.
For what it is worth the sample of i1d3 rev B units I've had through my hands (2x retail, 1x OEM with 2000 nits capability) have been slightly less sensitive than both of the i1d3 Rev As I've seen (one retail, one OEM). from screen the Rev As were relatively happy returning a (junk for chroma) reading down to 0.003 nits, but the Rev B units struggle to get down to there. However I still have only in the end kept my most recent OEM rev B as for reading from lens having some confidence of the top end is more valuable than a still mediocre or unnavailable dark reading from screen at the bottom, and this current i1d3 seems to have filters that were most aligned to CIE observer for my projector vs the Jeti 1201 I had.

Reading "JVC dark" from screen on an i1d3 certainly hasn't got any better, so don't waste time / money on it if you have an expectation to get a result that is worth anything there.
Manni01 likes this.
bobof is online now  
post #690 of 764 Old 06-30-2020, 10:25 AM
AVS Forum Special Member
 
Join Date: Nov 2014
Location: New York
Posts: 6,451
Mentioned: 242 Post(s)
Tagged: 0 Thread(s)
Quoted: 5347 Post(s)
Liked: 6554
Quote:
Originally Posted by liberator72 View Post
One thing that should be noted about this for full disclosure.

Usually, when performing a probe match you obviously validate the match. Previously, when using FCMM with the i1d3 on a WOLED, you can sometimes fight like hell to get a NIST valid match, and sometimes you never will on a WOLED in Wide gamut. xy is usually always correct with RGBW, and Y is always within NIST for GBW, whereas Red can struggle with Y. Sometimes I can get it NIST valid but usually, as you state above, it averages around 2%.

But...The correction used above was done very quickly with FCVM and as you can see, RGB are all perfectly well within NIST. However, when that correction data is used with the traditional FCMM method, the errors are much larger than they usually are.

What I'm trying to say is, with FCMM on WOLED, it can be a struggle with the i1d3 to get a good (NIST valid) probe match on WOLED, whereas with FCVM it easy, straightforward and works.

If I get the time in the near future and if other have an interest in me doing it, I will perform the FCMM in LightSpace so as to get that match as good as I possibly can, and with that BPD data, use it for FCVM too. As Enrico says above, it could potentially lead to even further benefits!
Even with my Klein, without FCVM i was getting a consistent elevation in Red. With FCVM, no problem works fast and great!

John
Sony 55A1E, A9F, LG CX
Marantz 7012, Ohm Walsh Speakers
Klein K10-A, Jeti 1501, Murideo Six-G Gen2
Calman Ultimate, LightSpace Pro, ISF Level III Certified
jrref is online now  
Sponsored Links
Advertisement
 
Reply Display Calibration

Thread Tools
Show Printable Version Show Printable Version
Email this Page Email this Page


Forum Jump: 

Posting Rules  
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off