AVS Forum banner

21 - 40 of 47 Posts

·
Registered
Joined
·
176 Posts
Other than native 4k, I think you are describing the currently available Radiance.
Either way I'd be interested in such a device described by the OP, but not at the price point posted in above quote.
Yes, I'm close on the current Radiance.

I use the 2143 and I'm very satisfied with! That's why I named it "midrange" device, positioned between the above discussed "LUT holder only" and the big "Pro". So I'd welcome a LUT holder having my listed features again - even with 4K out and in.

And to be honest, I can't believe, that a 4K capable "LUT holder only" box would be available under $ 1500 - as Jim wrote on Lumagen forum, already the FPGA costs more than an entire Radiance Mini-3D.

http://www.convergent-av.co.uk/forum/index.php?t=msg&th=2803&goto=22074&rid=0&S=6628df6afbb78ead1505ddae6f7336cb#msg_22074
 

·
Registered
Joined
·
1,081 Posts
Discussion Starter #22
Hi guys,

okay in regards to a straight 4K LUT box (only) - let's assume no pattern generator, and a few LUT slots - HOW MUCH would you be willing to pay ?

and please let's not forget this is 4K and is very expensive to process... my personal limit would be US$1.5-2K, and that is IMO already high, but there is no other option on the market... unless eeColor decides to do one, which may end up cheaper, but they are not considering doing one atm...

what would be your max budget for a 4K LUT box ?
 

·
Registered
Joined
·
533 Posts
I think Lumagen would be better served as a company updating the mini 3D to a 2160 HDMI 2.x pipeline with 17^3 or greater LUT.


I'd possibly drop the darbee for this product, but keep some form of scaling as the vast majority of peoples content will still be 1080.


To me this would be a product that would have greater sales potential and thus feasible lower cost. You have to keep in mind that low turn over product is always more expensive and can become easily unaffordable.
 

·
Registered
Joined
·
1,081 Posts
Discussion Starter #25
I think Lumagen would be better served as a company updating the mini 3D to a 2160 HDMI 2.x pipeline with 17^3 or greater LUT.


I'd possibly drop the darbee for this product, but keep some form of scaling as the vast majority of peoples content will still be 1080.


To me this would be a product that would have greater sales potential and thus feasible lower cost. You have to keep in mind that low turn over product is always more expensive and can become easily unaffordable.
I don't think that is possible... from what I heard from Jim, the components to handle 4K are very expensive... if u now support 2160p on the Mini with all the original VP features and scaling u would literally need to remove all components and install new ones... so u can just buy the new box, which prob also has updated airflow, and/or fans...
 

·
Registered
Joined
·
5,163 Posts
Any current pattern generator would work - they are just colour patterns - resolution is irrelevant.

BUT that does raise an interesting point - a Lumagen LUT with a built in pattern generator would be very nice indeed :)
I hope you are correct.
Reading PCmag review on the LG's 65" OLED . They are saying "With a flat white screen provided over a computer connection at 1080p, the screen produced a disappointing 82.78 cd/m2 peak brightness, similar to plasma HDTV with its screen completely lit and slightly dimmer than the 55EC9300. However, our 4K-capable signal generator using a similar test pattern got the screen to reach a far better 292.01 cd/m2, "

ss
 

·
Hello, World!
Joined
·
2,448 Posts
Think about it, what is a "flat white field"?

It's just an RGB signal at a set level (eg, 255,255,255)... if their computer pattern gave a different "white" to their 4K pattern generator then they made a mistake in their chain....
 

·
Registered
Joined
·
431 Posts
Hi guys,

as you may know Lumagen is releasing the "Pro" units later this year that can handle native 4K UHD and most importantly add 4K LUT processing, probably something like a 17^3 (my guess).

Since these Pro units will be on the higher end price wise, I inquired with Jim from Lumagen if he would consider building a straight 4K LUT box that does not have all of the Video Processing features that the general Lumagen units include, but simply offers 4K LUT processing.

The price would be cheaper than the "Pro" units but ultimately it depends on demand and interest to determine how low the price can get, the components are very expensive and it takes quite a few to process and handle 4K.

This box could have either HDMI 2.0 out or 4 x SDI 3G, or a combintaion of both. HDMI 2.0 is 18 Ghz so it can handle higher bit-depth (10 / 12 bit) and full chroma (444).

24 fps and 30 fps would be standard (in up to 10 or 12-bit), maybe higher frame rates with chroma subsampling or a drop in bit-depth.

To my knwoledge, no other vendor has announced a 4K LUT box, it seems that eeColor is also currently not working on one. Please update this thread if u know otherwise.


Who would be interested in a 4K LUT box ?


- M
Hello Mike,

Thank you for this post.

I would like such 3DLUT box from Lumagen. Without scaling, without enhancement, only 1:1 with support from full HD to 4K.

As for me, the pattern generator is a mandatory feature.

Size for 3DLUT should be at least 21^3 points but more is not a constraint :).

I think 10 memory slots should be enough.

With USB and RS232 port for firmware upgrade and integration.

An external power supply is better, it's easier to replace it when it is defective.
 

·
Registered
Joined
·
4,411 Posts

·
Registered
Joined
·
1,906 Posts
10bit 422 minimum, 10bit 444 better. Also the ability to store/select more than one LUT. 17x3 ok, 21x3 or higher better.

No other frills.

A simple unit like this would also be popular with the online-edit and color-grading pro markets, not just the high-end home market.
 

·
Registered
Joined
·
3,842 Posts
Keep it under $1000 and I'm in.
 

·
Registered
Joined
·
528 Posts
Reviving this thread in case there's any new news on this. I would love to have a CMS-only box with 3D LUT. If there are no commercial options, I am wondering if one could be built with a FPGA board with HDMI in/out. Would make for a nice programming project.
 

·
aka jfinnie
Joined
·
4,245 Posts
Reviving this thread in case there's any new news on this. I would love to have a CMS-only box with 3D LUT. If there are no commercial options, I am wondering if one could be built with a FPGA board with HDMI in/out. Would make for a nice programming project.
Not a very useful one, because you won't have HDCP keys (unless you harvest parts from other devices to make your box).
 

·
Registered
Joined
·
528 Posts
bobof, I plan on trying to harvest parts for that. I have a hdmi 2.0 splitter on the way. It may require collecting a few to find the right chips. I also may consider HDfury Integral.
 

·
Registered
Joined
·
8,918 Posts
bobof, I plan on trying to harvest parts for that. I have a hdmi 2.0 splitter on the way. It may require collecting a few to find the right chips. I also may consider HDfury Integral.
Hi, what display are you interest to calibrate for 4K/HDR with 3D LUT? To perform 3D LUT for HDR you have to disable the internal display tone/gamut mapping, this is not possible with any display in consumer world, so makes no sense to find a way to create a 3D LUT Box for HDR/4K even if the required hardware/developement cost 1$ d.

2160p input/output 3D LUT Box is only useful for displays to calibrate 3D LUT for SDR 2160p input, or for projection (where the no standard for HDR) where you handle the signal as a normal 10bit and you apply the 3D LUT.

Even if displays will have a button to disable their internal tone/gamut mapping and provide the same peak output they have in HDR mode, at least OLED's I have checked are very unstable for any kind of 3D LUT calibration.

See there some drift plots of LightSpace, to see how an OLED display is drifting in SDR of various peak levels and at HDR mode, during 3D LUT measurements: http://www.avsforum.com/forum/139-display-calibration/2853954-lg-2017-oled-calibration-thread-settings-44.html#post55316570
 

·
Registered
Joined
·
528 Posts
Hi, what display are you interest to calibrate for 4K/HDR with 3D LUT? To perform 3D LUT for HDR you have to disable the internal display tone/gamut mapping, this is not possible with any display in consumer world, so makes no sense to find a way to create a 3D LUT Box for HDR/4K even if the required hardware/developement cost 1$ d.
The display I have is the JVC x790 projector

2160p input/output 3D LUT Box is only useful for displays to calibrate 3D LUT for SDR 2160p input, or for projection (where the no standard for HDR) where you handle the signal as a normal 10bit and you apply the 3D LUT.
This would be for projection. I would prefer to leave the tone mapping to the projector. I would, however, would want to preserve the signal as HDR while adjusting for color correction. Is it not possible to do this?

Even if displays will have a button to disable their internal tone/gamut mapping and provide the same peak output they have in HDR mode, at least OLED's I have checked are very unstable for any kind of 3D LUT calibration.
How are people using Lumagen Pro's to do color correction on HDR signals?


After some more HW research, the costs to put together a eval board FPGA + hdmi 2.0 IO boards is probably in the $2300 range. It's quite ridiculous what these things cost in eval board form. I think I will be taking a few steps back and play with hdmi 1.3 or 1.4 using something like this: https://store.digilentinc.com/zybo-z7-zynq-7000-arm-fpga-soc-development-board/
It's certainly not hdmi 2.0 or HDR, but could be interesting as an intro to HDMI processing, perhaps a CMS for SDR.

For HDMI2/HDCP, I am not sure if it's viable unless one pays for an HDCP license, as you can't get the documentation for the chipset without the license. You could harvest the chips, but left to figure out for yourself how to make them work.
 

·
Registered
Joined
·
8,918 Posts
The display I have is the JVC x790 projector

This would be for projection. I would prefer to leave the tone mapping to the projector. I would, however, would want to preserve the signal as HDR while adjusting for color correction. Is it not possible to do this?

How are people using Lumagen Pro's to do color correction on HDR signals?
When you want to calibrate a projector for HDR, it's not possible to do in with any projector tone mapping active, so you don't send metadata to projector.

The projector need to provide a wide gamut (close to DCI-P3 primaries) to make sense and calibrate for HDR but not all projectors can give you a wide gamut and high peak output without entering to their HDR mode (where the projector tone mapping is active).

It's not possible to do any 3D LUT for HDR when you have any display/projector tone mapping active.

The generated 3D LUT for HDR will provide a fixed tone mapping your device will have, for example using LightSpace where it has options to make any tone-mapping shape/roll-off, you can create a correction for movies mastered for 1000 nits and a different shape for movies mastered for 4000 nits (or one for all)



There tools, like the the 'Multiplier' for projection HDR which re-scales the EOTF nits value, based on the Multiplier value used. For example, a value of x10 will effectively change an EOTF peak nits value of 97 nits into a perceived level of 970 nits. (A value of x1, as shown above will have no effect on the ST2084 scaling.)

An example application would be for a projector that has a peak nits value of 56,5 nits. To 'mimic' HDR operation a x10 multiplier could be used, which will re-scale the peak EOTF value to 565 nits, as shown here:




So it makes no sense any more what metadata each movie has, the projector will have a fixed performance, for that reason you don't send any metadata to projector (which will enable display tone mapping so it will destroy the calibration you have performed)

With Lumagen PRO, you disable the output of metadata to projector and you 3D LUT the signal like a normal 10bit signal. (HDR10 signal is a 10bit signal with a static metadata.)

There users where they 3D LUT with gamma 2.4 and then use Lumagen's PRO intensity mapping tools.

For HDR for projection, soft-roll off, tone mapping, see there:

https://www.lightillusion.com/hdr_calibration.html

https://www.lightillusion.com/lut_manual.html

After some more HW research, the costs to put together a eval board FPGA + hdmi 2.0 IO boards is probably in the $2300 range. It's quite ridiculous what these things cost in eval board form. I think I will be taking a few steps back and play with hdmi 1.3 or 1.4 using something like this: https://store.digilentinc.com/zybo-z7-zynq-7000-arm-fpga-soc-development-board/
It's certainly not hdmi 2.0 or HDR, but could be interesting as an intro to HDMI processing, perhaps a CMS for SDR.

For HDMI2/HDCP, I am not sure if it's viable unless one pays for an HDCP license, as you can't get the documentation for the chipset without the license. You could harvest the chips, but left to figure out for yourself how to make them work.
About HDCP 1.4 license, it's about $20000 per year the annual fees, another $20000 for HDMI 1.x, in case you want to release something in market, I don't think HDMI.org will give any license to DIY's.

Probably the prices will be higher for HDCP 2.2 and HDMI 2.0a annual fees.
 

·
aka jfinnie
Joined
·
4,245 Posts
The display I have is the JVC x790 projector


This would be for projection. I would prefer to leave the tone mapping to the projector. I would, however, would want to preserve the signal as HDR while adjusting for color correction. Is it not possible to do this?


How are people using Lumagen Pro's to do color correction on HDR signals?


After some more HW research, the costs to put together a eval board FPGA + hdmi 2.0 IO boards is probably in the $2300 range. It's quite ridiculous what these things cost in eval board form. I think I will be taking a few steps back and play with hdmi 1.3 or 1.4 using something like this: https://store.digilentinc.com/zybo-z7-zynq-7000-arm-fpga-soc-development-board/
It's certainly not hdmi 2.0 or HDR, but could be interesting as an intro to HDMI processing, perhaps a CMS for SDR.

For HDMI2/HDCP, I am not sure if it's viable unless one pays for an HDCP license, as you can't get the documentation for the chipset without the license. You could harvest the chips, but left to figure out for yourself how to make them work.
Sounds like you've talked yourself into building an EEcolor, which you could pick up for ~$300. You must be really short of stuff to do in the real world...! :)
 
21 - 40 of 47 Posts
Top