AVS Forum banner

3761 - 3780 of 8142 Posts

·
Registered
Joined
·
582 Posts
He chased me out of the room, so I didn't get to see precisely what he was up to. I'm guessing it was a mix of direct work on both the JVC RS-500 and Lumagen Pro. He's got his own mix of CalMAN driven steps. I believe for HDR the activities were based on prior and rather lengthy calibrations and results from those using Lumagen models as a basis.



For me it's the results that count. Given that he's been at this since the turn of the century, I'd say he's mastered his craft.
 

·
aka jfinnie
Joined
·
4,233 Posts
Maybe just me because this is a hobby, but I would not hire any calibrator that didn’t allow me to stay in the room and ask some questions. They have plenty of time especially when Calman is running measurements. SJ
Can you imagine the uproar though from the animal rights lobby when it is revealed that pro calibrators round the globe are cleaning the lens with freshly squeezed unicorn tears?
 

·
Registered
Joined
·
535 Posts
2u units are removed from the lumagen website. Are they now discontinued?
 

·
Premium Member
Joined
·
3,902 Posts
The picture was changed to a 1U unit. But the description still has all of them. Would not be enough inputs for my setup unless they change it. SJ
 

·
Registered
Joined
·
18,139 Posts
Maybe just me because this is a hobby, but I would not hire any calibrator that didn’t allow me to stay in the room and ask some questions. They have plenty of time especially when Calman is running measurements. SJ
I stayed in the room when Chad calibrated my projector. You need to give input time to time. Wasn't a problem. Mostly I observed.
 

·
Registered
Joined
·
2,734 Posts
Did Chad use the Lumagen to calibrate?


Yes, I had the Lumagen do the tone mapping.

I answer all my customer’s questions. I just don’t like to feel like I’m being watched while I work especially if I am dealing with a particularly difficult or unusual situation; it just adds to the stress level. In those cases, if I am not asked, I say something along the lines of I prefer working by myself. I think most people would want their calibrator to be as relaxed and stress free as possible when he’s making judgement calls about HDR tone mapping.


Sent from my iPhone using Tapatalk
 

·
Registered
Joined
·
1,626 Posts
2u units are removed from the lumagen website. Are they now discontinued?
The 2U version of the Radiance Pro 444X is now housed in a 1U case. We switched all Radiance Pro products to 1U for manufacturing reasons.

However, note there is an exception in that the Radiance Pro 424X is available in a small, non-rack-mount "compact" case (424x-C). This is intended for calibrators and/or instructors who need to carry a unit with them. We have also sold a few to people with limited space for the Radiance Pro.

We updated the Lumagen website front page, but still need to remove the 2U from some other pages.
 

·
Registered
Joined
·
1,626 Posts
The Radiance Pro initial release of dynamic tone mapping is getting closer but we have been off on a bug hunt concerning Image Based Auto Aspect which seems to have been getting more use of late. We have always prioritized bug fixes higher than new features and this is no exception. Patrick should be back working on dynamic tone mapping soon.

====

Please note that if you have a bug report, you should email Lumagen support. If it is urgent, or hard to express in an email, you can call Lumagen support. Usually email is best. Please make sure you are on the latest release first, and include as much information as possible including how to create the bug, sources TV/projector, audio processor, connections, and (very important) brand, model, and length of HDMI cables and where they are connected. We also like to get pictures of the four "Info pages" (press OK on the remote, take picture, repeat for all four info pages).

Certainly fine if you post that you have found the bug and are contacting Lumagen, but please do not expect a response to your post from Lumagen. You might get a Lumagen response of course, but might not. Posting is helpful as it will let others know about the potential bug and some may have a comment or resolution.
 

·
aka jfinnie
Joined
·
4,233 Posts
The Radiance Pro initial release of dynamic tone mapping is getting closer but we have been off on a bug hunt concerning Image Based Auto Aspect which seems to have been getting more use of late. We have always prioritized bug fixes higher than new features and this is no exception. Patrick should be back working on dynamic tone mapping soon.
I for one am extremely grateful for the bug fixes to auto aspect. I've been using it to "family friendly" correct masking and calibration in my home theatre - it really gets my goat having a masking screen that I'm the only one to get the benefit of if it is driven manually (these things cost way too much for what they are as it is). It is working really well now it seems with most content; just seen some reluctance with AppleTV HDR content.

----------

For anyone who is wondering; I've just re-written what was a horrid shell script running on a Raspberry Pi in something a little nicer - python - though the script is still a bit quick and dirty. I'll probably iterate over this and make it do a bit more in terms of parsing the Lumagen status outputs.

There really isn't much required. This script waits for serial lines to appear, then works out what the command received is, and if it is a V2 report (you have to enable them in comms menu) it parses it to get the aspect reported. It then grabs a particular URL based on what the aspect is - that is how I set the aspect of the screen in my home control system. It just sits in a loop forever waiting. This script has the added advantage of working with both the output of the auto aspect logic and also the manually selected aspect ratios, so it is quite useful even if you're not using auto aspect and have a mask screen.

Took about 90 mins to work out how to do this, much of that being due to limited knowledge of Python! I'm not a huge fan, the use of whitespace as an active code element and case sensitivity of Python drives me a little up the wall :)
Code:
rakoip = "192.168.xxx.xxx"

screencommands = {
        "178": "3",
        "185": "4",
        "200": "5",
        "220": "5",
        "235": "6",
        "240": "1",
        }

rakocommand = rakoip + "/rako.cgi?room=10&ch=0&com="

import serial
import urllib2

ser = serial.Serial('/dev/ttyUSB0',9600 )  # open serial port
print "serial port: " + ser.name         # check which port was really used

while True:
        print "--------------------------------------"
        line = ser.readline()     # read a line
        print "Lumagen line:" + line # print line
        command = ""
        arguments = ""
        foundcmd = False
        endcmd = False
        for char in line:
                if endcmd:
                        arguments = arguments + char
                if foundcmd and not endcmd:
                        if char not in ',':
                                command = command + char
                        else:
                                endcmd = True
                if not foundcmd:
                        if char in '!':
                                foundcmd = True
        print "Lumagen command: " + command
        print "Lumagen arguments: " + arguments
        if command == "I22":
                print "found v2 status report"
                argnum = 7
                currentarg = 1
                argument = ""
                for char in arguments:
                        if char not in ',' and currentarg == argnum:
                                argument = argument + char
                        if char in ',':
                                currentarg = currentarg + 1
                print "aspect reported: " + argument
                url = "http://" + rakocommand + screencommands.get (argument)
                print "URL to get: " + url
                request = urllib2.Request(url)
                contents = urllib2.urlopen(request).read()
ser.close()             # close port
 

·
Registered
Joined
·
237 Posts
I for one am extremely grateful for the bug fixes to auto aspect. I've been using it to "family friendly" correct masking and calibration in my home theatre - it really gets my goat having a masking screen that I'm the only one to get the benefit of if it is driven manually (these things cost way too much for what they are as it is). It is working really well now it seems with most content; just seen some reluctance with AppleTV HDR content.

For anyone who is wondering; I've just re-written what was a horrid shell script running on a Raspberry Pi in something a little nicer - python - though the script is still a bit quick and dirty. I'll probably iterate over this and make it do a bit more in terms of parsing the Lumagen status outputs.

There really isn't much required. This script waits for serial lines to appear, then works out what the command received is, and if it is a V2 report (you have to enable them in comms menu) it parses it to get the aspect reported. It then grabs a particular URL based on what the aspect is - that is how I set the aspect of the screen in my home control system. It just sits in a loop forever waiting. This script has the added advantage of working with both the output of the auto aspect logic and also the manually selected aspect ratios, so it is quite useful even if you're not using auto aspect and have a mask screen.

Took about 90 mins to work out how to do this, much of that being due to limited knowledge of Python! I'm not a huge fan, the use of whitespace as an active code element and case sensitivity of Python drives me a little up the wall :)
Code:
rakoip = "192.168.xxx.xxx"

screencommands = {
        "178": "3",
        "185": "4",
        "200": "5",
        "220": "5",
        "235": "6",
        "240": "1",
        }

rakocommand = rakoip + "/rako.cgi?room=10&ch=0&com="

import serial
import urllib2

ser = serial.Serial('/dev/ttyUSB0',9600 )  # open serial port
print "serial port: " + ser.name         # check which port was really used

while True:
        print "--------------------------------------"
        line = ser.readline()     # read a line
        print "Lumagen line:" + line # print line
        command = ""
        arguments = ""
        foundcmd = False
        endcmd = False
        for char in line:
                if endcmd:
                        arguments = arguments + char
                if foundcmd and not endcmd:
                        if char not in ',':
                                command = command + char
                        else:
                                endcmd = True
                if not foundcmd:
                        if char in '!':
                                foundcmd = True
        print "Lumagen command: " + command
        print "Lumagen arguments: " + arguments
        if command == "I22":
                print "found v2 status report"
                argnum = 7
                currentarg = 1
                argument = ""
                for char in arguments:
                        if char not in ',' and currentarg == argnum:
                                argument = argument + char
                        if char in ',':
                                currentarg = currentarg + 1
                print "aspect reported: " + argument
                url = "http://" + rakocommand + screencommands.get (argument)
                print "URL to get: " + url
                request = urllib2.Request(url)
                contents = urllib2.urlopen(request).read()
ser.close()             # close port
Same logic I used for an application I wrote based on .NET framework and it works but as I've reported to Lumagen and as reported from some former there's still something in Radiance to be fixed up in order to reach 2.35 frame recognition. Sometime it jumps from 2.35 to 1.78 and back.
In Lumagen bug fixing is a priority and their support is second to none in my experience. With next firmware I'm sure it'll be working perfectly.
 

·
aka jfinnie
Joined
·
4,233 Posts
Same logic I used for an application I wrote based on .NET framework and it works but as I've reported to Lumagen and as reported from some former there's still something in Radiance to be fixed up in order to reach 2.35 frame recognition. Sometime it jumps from 2.35 to 1.78 and back.
In Lumagen bug fixing is a priority and their support is second to none in my experience. With next firmware I'm sure it'll be working perfectly.
Fingers crossed. I've seen it steadily become much, much better. I'm sure the remaining issues are "baked in" to the content / source black level fidelity - yesterday I saw the same HDR content on AppleTV not work well (DV converted to "HDR10_ish_lol"), but work fine when played via my Sony UHD player in HDR10.

It's obviously a fine line between detecting quickly correctly and staying locked in, and not having false detections. From what I could see the only issues left look like these "marginal" cases. Of course to the end user they don't look hard at all as we can't see what is going on in the black bars usually.

For me auto-aspect is a real stand-out feature in the product and really adds both "wow" feeling to the theatre and enhances PQ of the presentation. Arguably with the quality of image being put out by most projectors these days via their internal controls (at least for SDR) there is more overall PQ benefit to a correctly masked image than there is to 3DLUT calibration. Of course if you have a Kaleidescape system you have instant aspect information available, and with an HTPC you could similarly implement 100% accurate aspect info; but increasingly with content reaching us through streaming means via a multitude of services and products something like the Lumagen Pro is the only way to achieve it for all sources (in fact, I think it is the only "consumer" way to do this - I've seen something advertised for a particular Barco projector if I recall but I think that was $$$$$).
 

·
Registered
Joined
·
1,958 Posts
So, can someone explain to a newbie what this all means?

If I buy the lumagen, what do I do then? do I buy a lightmeter? does it come with software? or do I buy that seperate? Do I have to be a trained professional to figure it out? or is there a walk thru tutorial. ?

Do I just leave the projector on out of the box settings and the lumagen changes evertyhing it needs on the input signal? or does the software tell me what to change on the projector? Is this close to or the same as an ISF calibration?

Or do I still have to hire someone to come and do that?

Depenmding on the initial accuracy of the displays rec709 or rec2020 or P3 colour settings you could just plug and play a Pro and set up some intial parameters to give you, most likely, superior HDR playback performance on many displays. Using a basic probe like a DisplayPro3 probe and some simple software like LightSpace you could, with a little learning, quickly learn how to do basic set up and calibratrion on your display followed by additional LUT based calibrations using the Lumagen as the colour processor.



It was a year ago that i said i'd do a basic video walk through of this but i have been very busy. I really should try to get on to that
 

·
Registered
Joined
·
1,047 Posts
I don't own the Pro but have owned the XS and previous models for a number of years. I hope my question will be applicable to Pro users.


Streaming from Amazon and Netflix now includes a number of foreign shows. These seem to be shown at 1080p25 or 1080p30. When I upscale to 4K with my Oppo 203 is there a consensus whether l080p25 should be shown at 4K25p or 4K50p? Similarly for 1080p30, 4K30p or 4K60p?


Since 1080p24 is upscaled to 4K24p and not 4K48p it seems we should stay at 25p and 30p.
 

·
.NET Solution Architect
Joined
·
944 Posts
I’m thinking I may invest in a Lumagen to avail of the dynamic tone mapping which is incoming.



I can’t spare the time to mess around with Mad VR and so on, life’s too short, and time is too precious.



Just wondering what all you owners consider to be the best feature, most used and most beneficial?
I use Lumagen for about 4-5 years can't remember exactly. What the best thing I love in Lumagen is the calibration abilities to your display, especially the 3d Lut calibration. If done in proper hands with mind thinking you will get absolutely superb stunning result. On the other hand Lumagen is the only one video processor, which works with many different inputs as many as you want and applies tone mapping as required, also soon should be dynamic tone mapping be present. From those one of the mains features, there are tons of others starting from scaling and ending with easier setups with anamorphic lens with NLS stretches and etc. It is a great video processor to the current time and is the best in the world. Every penny I've spent on it is absolutely worth it, do don't even doubt if you have budget just spend on that item and then you'll find yourself that you cannot live without it :)

Sent from my ONEPLUS A6003 using Tapatalk
 

·
Registered
Joined
·
3,399 Posts
I've been using it for a couple of years now. Cut my teeth just doing SDR LUTs, then hacked around doing HDR LUTs for my old X30 JVC with the 2143 (using a Linker to force output of the WCG2020 data at 1080p to the 2143), and now got the Pro. The LS docs are pretty good and you have the benefit that Steve is here in our timezone and appears to not sleep, so will reply it seems at most times within hours. There are a few things about the interface and graphs which are a bit idiosyncratic, and it isn't a great tool for manual (interactive) calibration, but for 3DLUT I don't think it really has any competition. I did own CP3 and really didn't get on with that at all, I've seen CM being used and don't like the restrictive workflow.

Once you get your head round profiling and LUT generation being separate things you do it is actually pretty straightforward - especially if you're already across the technical reqs, which is where a lot of CM users I think fall down because Calman significantly lowers the bar in terms of what you need to understand to be able to get some kind of calibration done.

It is pretty much:

Set up patch generation in the Lumagen or an external patch gen (oddly, under Upload). Upload a NULL cube to the currently active CMS by checking the box.
Go to Calibration interface - it will ask you to connect to the meter. Spend a little time taking some manual readings and quick profiles to figure out how you're going to set up the projector in what modes for what you are trying to achieve.
Go to Display Characterisation and set up the cube or patch sequence you are going to profile. Run the profile.
Use Convert Colorspace to choose the target colourspace (top section) and the measured profile (middle section); setting options / LUT algorithm (bottom section).
Upload the cube by unchecking the upload NULL box, which will upload the currently selected LUT to the currently active CMS. It uploads a calculated 1D and 3D lut together as a pair (they're both generated mathematically from the one profile pass).

I've had my Lumagen for a good 9 months now and finally started delving into CP3 as I've been extremely happy with JVC autocal to date (as limited as it is) with my 4500. I pretty quickly found I really didn't like CP3 much at all and now keen for an alternative. Souinds like LS could be the go now that I am on the thread reading the recent feedback here.
 

·
aka jfinnie
Joined
·
4,233 Posts
I've had my Lumagen for a good 9 months now and finally started delving into CP3 as I've been extremely happy with JVC autocal to date (as limited as it is) with my 4500. I pretty quickly found I really didn't like CP3 much at all and now keen for an alternative. Souinds like LS could be the go now that I am on the thread reading the recent feedback here.
I had CP3 and was unimpressed with the LUT generation. From what I understand from watching it in action it basically iterated over the LUT points trying to lower their dE, and stopped once it got to a certain dE threshold. It's an annoying way to do it:
1) why limit how good you could get
2) it seemed to result in quite wonky greyscales, I guess because it was just lowering the dE, rather than doing math to try and keep all the errors the same side of grey. I'd often end up with a grey then a pink then a green (exagerated) in the low end greyscale.
3) you had to set the target you were adjusting to before starting the LUT process, and couldn't re-target or adjust after the fact,

Lightspace is nice. It very much feels like a no-nonsense no-frills pro tool, which is exactly what it is. From what I've seen it appears their 3DLUT math is the best in the business; I've not used Calman but friends have and haven't been impressed (your mileage may vary). Lightspace has been constantly improved significantly over the period I've owned it, and I've not had to buy any updates in that time as they've had a "no updates charge" policy forever. The Lumagen integration works well.

Being critical my gripes would be:
a) manual calibration; the interface isn't that helpful to the task in hand (free HCFR does a better job on the usability for what you need to achieve during manual cal. But for setting up the pre-conditions for LUT in the display it is more than adequate
b) one of the charts is a bit non-standard for HT use (gamma difference instead of absolute gamma as it both appears upside down to what you might expect, is difficult to scale and doesn't show the actual target gamma). This one is a bit of an idealogical war still with Light Illusion it seems... :p I doubt it will ever change.
c) the license cost is tied to some features and some meter options - you get more meters and LUT math options with HTP. HTL (the least expensive option) supports entry level meters like i1d3, spyder, i1pro. You have to buy the much more reasuringly HTP level at least to support high end meters like Kleins / Jetis / Colorimetry Research / Minolta etc. There is a bit of an anomally there I think in that even the Discus (a £700 colorimeter) is included in the HTP level but not HTL. Which meant I had to get HTP. As it happens I have a Jeti now so I'd have to have bought HTP anyway (eventually).

Steve is very responsive to support queries and clearly very knowledgeable, support is first rate.

Anyway, worth giving a go. If you can afford a long calibration weekend you can buy a time limited 3 day license to play with it.
 

·
Registered
Joined
·
221 Posts
How smart is the auto aspect detection? If I switch my cable tv channel from a channel playing a 1.78:1 movie to one playing a 2.35:1 movie will it detect it?
 

·
aka jfinnie
Joined
·
4,233 Posts
How smart is the auto aspect detection? If I switch my cable tv channel from a channel playing a 1.78:1 movie to one playing a 2.35:1 movie will it detect it?
Yes* it will. I love this feature as I have it automating the electric masks on my screen which open and close automatically in time with the content.

*At the moment there are some titles auto-aspect is reluctant to detect on, or detects and flip-flops in. It seems from discussions with Lumagen support these are titles where there is noise in the bars / and / or they're not pure video black. Anyway, Lumagen are working on it.
 

·
Registered
Joined
·
3,399 Posts
I had CP3 and was unimpressed with the LUT generation. From what I understand from watching it in action it basically iterated over the LUT points trying to lower their dE, and stopped once it got to a certain dE threshold. It's an annoying way to do it:
1) why limit how good you could get
2) it seemed to result in quite wonky greyscales, I guess because it was just lowering the dE, rather than doing math to try and keep all the errors the same side of grey. I'd often end up with a grey then a pink then a green (exagerated) in the low end greyscale.
3) you had to set the target you were adjusting to before starting the LUT process, and couldn't re-target or adjust after the fact,

Lightspace is nice. It very much feels like a no-nonsense no-frills pro tool, which is exactly what it is. From what I've seen it appears their 3DLUT math is the best in the business; I've not used Calman but friends have and haven't been impressed (your mileage may vary). Lightspace has been constantly improved significantly over the period I've owned it, and I've not had to buy any updates in that time as they've had a "no updates charge" policy forever. The Lumagen integration works well.

Being critical my gripes would be:
a) manual calibration; the interface isn't that helpful to the task in hand (free HCFR does a better job on the usability for what you need to achieve during manual cal. But for setting up the pre-conditions for LUT in the display it is more than adequate
b) one of the charts is a bit non-standard for HT use (gamma difference instead of absolute gamma as it both appears upside down to what you might expect, is difficult to scale and doesn't show the actual target gamma). This one is a bit of an idealogical war still with Light Illusion it seems... :p I doubt it will ever change.
c) the license cost is tied to some features and some meter options - you get more meters and LUT math options with HTP. HTL (the least expensive option) supports entry level meters like i1d3, spyder, i1pro. You have to buy the much more reasuringly HTP level at least to support high end meters like Kleins / Jetis / Colorimetry Research / Minolta etc. There is a bit of an anomally there I think in that even the Discus (a £700 colorimeter) is included in the HTP level but not HTL. Which meant I had to get HTP. As it happens I have a Jeti now so I'd have to have bought HTP anyway (eventually).

Steve is very responsive to support queries and clearly very knowledgeable, support is first rate.

Anyway, worth giving a go. If you can afford a long calibration weekend you can buy a time limited 3 day license to play with it.

Thanks bobof, great feedback there. I'll definitely look into LightSpace!
 
3761 - 3780 of 8142 Posts
Top