AVS Forum banner

1261 - 1280 of 2609 Posts

·
Premium Member
Joined
·
1,177 Posts
Since the LK970 does not have a wider colour space than Rec709, is it right to assume that the perceived comparisional differences in HDR are primarily due to increased in lumens between this and the JVC N5/N7 or the Sony 760/870?
This is not correct. The LK970 does have a color mode with wider gamut than rec 709. In Cinema mode, the gamut is larger. I know it doesn't cover full DCI-P3, but I think it's a relatively high percentage of P3. Dave Harper has measured his and can chime in here if he wants.
 

·
Registered
Joined
·
483 Posts
This is not correct. The LK970 does have a color mode with wider gamut than rec 709. In Cinema mode, the gamut is larger. I know it doesn't cover full DCI-P3, but I think it's a relatively high percentage of P3. Dave Harper has measured his and can chime in here if he wants.
Thanks for the info. I must be mistaken. Can someone who has measured share the data on P3 coverage?

Sent from my ALP-L29 using Tapatalk
 

·
Premium Member
Joined
·
1,177 Posts

·
aka jfinnie
Joined
·
4,242 Posts
No, it doesn't. Dave and I had a long (slightly acrimonious!) back and forth about that diagram, and it transpired a few posts in that this diagram was the result of him sticking a filter in front of the lens to expand the native colourspace. It is not representative of the native capabilities of this projector.
 

·
Registered
Joined
·
28,840 Posts
No, it doesn't. Dave and I had a long (slightly acrimonious!) back and forth about that diagram, and it transpired a few posts in that this diagram was the result of him sticking a filter in front of the lens to expand the native colourspace. It is not representative of the native capabilities of this projector.
Yep, natively the projector can't even get to Rec709. Even BenQ only lists 92% of Rec709.
https://business-display.benq.com/en-us/findproduct/projector/installation-projectors/lk990.html
 

·
Registered
Joined
·
28,840 Posts
This thread is on the LK970, not the LK990. What is your basis for saying the LK970 doesn't do rec709?
My mistake. Thought I was in the 990 thread. Though I will say, do you think BenQ would go backwards in color space, going from 970 to 990 and then add HDR support? That would not make sense.
 

·
aka jfinnie
Joined
·
4,242 Posts
FYI @Citation4444

You can read the green primary point off BENQ's own infographic on their website for the LK970, which "helpfully" doesn't show a reference gamut (only a gamut for some "lamp-based projector" of their own creation)



I make it out to be only at around or just under the REC709 position on the chart (looks on their chart to be x 0.33 y 0.64). P3 would be at (x 0.265 y 0.690) which it is no-where close to.

I repeat I'm 99.9% certain that image from @Dave Harper was an experiment in adding filters, and does not represent the native behaviour of the LK970. I'm sure he can confirm or deny.
 

·
Registered
Joined
·
765 Posts
That is reference gamut they show for their blue core laser projectors. It is the engineering design basis for the LK970. Same light engine and color wheel. The 990 has a unique graph and 1000 more lumens with brilliant color so I would venture the native gamut could be different from the 970.

If you are calculating color volume it’s 109 % of Lamp based Red triangle and 91% DCI P3. This seems possible but I have no way to confirm other than qualitative viewing. I would be really surprised if it didn’t cover rec709 after many hours with the unit over the past 6 months.


FYI @Citation4444

You can read the green primary point off BENQ's own infographic on their website for the LK970, which "helpfully" doesn't show a reference gamut (only a gamut for some "typical" projector of their own creation)



I make it out to be only at around or just under the REC709 position on the chart (x 0.33 y 0.64).

I repeat I'm 99.9% certain that image from @Dave Harper was an experiment in adding filters, and does not represent the native behaviour of the LK970. I'm sure he can confirm or deny.
 

·
aka jfinnie
Joined
·
4,242 Posts
That is reference gamut they show for their blue core laser projectors. It is the engineering design basis for the LK970. Same light engine and color wheel. The 990 has a unique graph and 1000 more lumens with brilliant color so I would venture the native gamut could be different from the 970.

If you are calculating color volume it’s 109 % of REC709 and 91% DCI P3. This seems possible but I have no way to confirm other than qualitative viewing. I would be really surprised if it didn’t cover rec709 after many hours with the unit over the past 6 months.
Where did those numbers come from out of interest?
 

·
Registered
Joined
·
765 Posts
I took the color points, plotted them and calculated the volume. Then compared that calculation to the same method for the LS 10000, Optoma UHZ65 and Sony VW1100. The volume calculation matched the review measured data in those 3 cases closely within 1% ,so I figured it was in that range. The color volume percentage is relative to the DCI P3 D65 color point volume as the basis.

I did this for the red triangle as well, but that is a little short in green of rec 709. So I would say what benQ posted is a high percentage of 709 but not full 709.


That is reference gamut they show for their blue core laser projectors. It is the engineering design basis for the LK970. Same light engine and color wheel. The 990 has a unique graph and 1000 more lumens with brilliant color so I would venture the native gamut could be different from the 970.

If you are calculating color volume it’s 109 % of REC709 and 91% DCI P3. This seems possible but I have no way to confirm other than qualitative viewing. I would be really surprised if it didn’t cover rec709 after many hours with the unit over the past 6 months.
Where did those numbers come from out of interest?
 

·
Banned
Joined
·
6,554 Posts
Since the LK970 does not have a wider colour space than Rec709, is it right to assume that the perceived comparisional differences in HDR are primarily due to increased in lumens between this and the JVC N5/N7 or the Sony 760/870?

In short, since there is no real spec for HDR on projectors and some form of tone mapping would be required the projector than can go brighter even though the colour space is limited would seem to be more impactful when viewing side by side.

Also is it right to assume that in the tweaks that are necessary in HDR there would be a need to crush black on the LK970 (meaning not clipping at 64 but more at say 80? - i am assuming as I am unsure how the black floor was improved by woofer and Dave).

For SDR, since most projectors are set to hit a target of about 16fl , does one need to clip above digital black of 17 (say 24) to have better perceived blacks. DLPs by nature are pretty good at clipping at 16 when the brightness is set to 0 but at this stage the black floor is usually much higher than LCD/LCoS/SXRD/LCoQ.

I'm curious how much has changed as the last time I compared a well known DLP with a JVC/Sony it was apparent that the black floor was much higher that the other technologies.

Sent from my ALP-L29 using Tapatalk

I’d say you’re in the right track. I was able to get the saturation points out further than rec709, but the main thing that gives the colors the pop that you’re alluding to is the color brightness.

You’re on track with the black floor statement too. On HDR patterns you do seem to crush some black (brightness at about 47 is correct, but 43 is best in images) when only using the black bars, but in actual images you need to go lower for it to look right. You can go right up and look and see there’s no actual black crush though, even when compared to my Calibrated flat panel. I think this is a result of the Automatic Power Control feature which does something similar on the high end when showing a full white pattern, per @12GAGE.

This is a perfect example of numbers not telling the whole story.

Thanks for the info. I must be mistaken. Can someone who has measured share the data on P3 coverage?

Sent from my ALP-L29 using Tapatalk
No, it doesn't. Dave and I had a long (slightly acrimonious!) back and forth about that diagram, and it transpired a few posts in that this diagram was the result of him sticking a filter in front of the lens to expand the native colourspace. It is not representative of the native capabilities of this projector.

Yeah that was fun, huh?

Yes that gamut chart was from when I tried some color filters. The color saturations got better but it wasn’t worth the expense of light and color brightness.






My mistake. Thought I was in the 990 thread. Though I will say, do you think BenQ would go backwards in color space, going from 970 to 990 and then add HDR support? That would not make sense.

They sacrificed color gamut for brightness, which is more important in the commercial projector world. I think you can Calibrate it back to what it needs to be though. Hopefully I’ll know soon.

That is reference gamut they show for their blue core laser projectors. It is the engineering design basis for the LK970. Same light engine and color wheel. The 990 has a unique graph and 1000 more lumens with brilliant color so I would venture the native gamut could be different from the 970.

If you are calculating color volume it’s 109 % of Lamp based Red triangle and 91% DCI P3. This seems possible but I have no way to confirm other than qualitative viewing. I would be really surprised if it didn’t cover rec709 after many hours with the unit over the past 6 months.

Yes that’s pretty much how I recall figuring it out. We had a small discussion on this iirc. It’s been awhile since then though.
 

·
Banned
Joined
·
6,554 Posts
Since the LK970 does not have a wider colour space than Rec709, is it right to assume that the perceived comparisional differences in HDR are primarily due to increased in lumens between this and the JVC N5/N7 or the Sony 760/870?

In short, since there is no real spec for HDR on projectors and some form of tone mapping would be required the projector than can go brighter even though the colour space is limited would seem to be more impactful when viewing side by side.

Also is it right to assume that in the tweaks that are necessary in HDR there would be a need to crush black on the LK970 (meaning not clipping at 64 but more at say 80? - i am assuming as I am unsure how the black floor was improved by woofer and Dave).

For SDR, since most projectors are set to hit a target of about 16fl , does one need to clip above digital black of 17 (say 24) to have better perceived blacks. DLPs by nature are pretty good at clipping at 16 when the brightness is set to 0 but at this stage the black floor is usually much higher than LCD/LCoS/SXRD/LCoQ.

I'm curious how much has changed as the last time I compared a well known DLP with a JVC/Sony it was apparent that the black floor was much higher that the other technologies.

Sent from my ALP-L29 using Tapatalk

I’d say you’re in the right track. I was able to get the saturation points out further than rec709, but the main thing that gives the colors the pop that you’re alluding to is the color brightness.

You’re on track with the black floor statement too. On HDR patterns you do seem to crush some black (brightness at about 47 is correct, but 43 is best in images) when only using the black bars, but in actual images you need to go lower for it to look right. You can go right up and look and see there’s no actual black crush though, even when compared to my Calibrated flat panel. I think this is a result of the Automatic Power Control feature which does something similar on the high end when showing a full white pattern, per @12GAGE.

This is a perfect example of numbers not telling the whole story.

Thanks for the info. I must be mistaken. Can someone who has measured share the data on P3 coverage?

Sent from my ALP-L29 using Tapatalk
No, it doesn't. Dave and I had a long (slightly acrimonious!) back and forth about that diagram, and it transpired a few posts in that this diagram was the result of him sticking a filter in front of the lens to expand the native colourspace. It is not representative of the native capabilities of this projector.

Yeah that was fun, huh?

Yes that gamut chart was from when I tried some color filters. The color saturations got better but it wasn’t worth the expense of light and color brightness.






My mistake. Thought I was in the 990 thread. Though I will say, do you think BenQ would go backwards in color space, going from 970 to 990 and then add HDR support? That would not make sense.

They sacrificed color gamut for brightness, which is more important in the commercial projector world. I think you can Calibrate it back to what it needs to be though. Hopefully I’ll know soon.

That is reference gamut they show for their blue core laser projectors. It is the engineering design basis for the LK970. Same light engine and color wheel. The 990 has a unique graph and 1000 more lumens with brilliant color so I would venture the native gamut could be different from the 970.

If you are calculating color volume it’s 109 % of Lamp based Red triangle and 91% DCI P3. This seems possible but I have no way to confirm other than qualitative viewing. I would be really surprised if it didn’t cover rec709 after many hours with the unit over the past 6 months.

Yes that’s pretty much how I recall figuring it out. We had a small discussion on this iirc. It’s been awhile since then though.
 

·
aka jfinnie
Joined
·
4,242 Posts
I took the color points, plotted them and calculated the volume. Then compared that calculation to the same method for the LS 10000, Optoma UHZ65 and Sony VW1100. The volume calculation matched the review measured data in those 3 cases closely within 1% ,so I figured it was in that range. The color volume percentage is relative to the DCI P3 D65 color point volume as the basis.

I did this for the red triangle as well, but that is a little short in green of rec 709. So I would say what benQ posted is a high percentage of 709 but not full 709.
Hrmph... I think your method is bogus if I'm honest...

Just because the triangle has a bigger area than REC709 does you can't claim >100% REC709 if it doesn't also include all of REC709! :) You actually have to be able to display all the colours bounded by the REC709 triangle to have >100% REC709. Once you have all the colours in REC709 covered, and you have a bigger triangle, then you have >REC709 gamut coverage. Otherwise you could claim >REC709 gamut coverage with some truly ridiculously unusable gamuts.

If you plot the proper true REC709 green point onto the BENQ graph (x 0.3 y 0.6) you will see that there is an area to the left side of the blue triangle which does not get covered. This is because the green primary in the LK970 image is too far to the right for how high it is. Because you can only mix colours at the points between the green and blue primary along the line joining them at that edge you will never be able to make those green / cyan / blue colours bounded by that triangle slither.

I've added it to the graph and attached so you can see clearly hopefully.



At the end of the day, there is a reason this image shows some nebulous "lamp based projector" and NOT the REC709 and DCI-P3 standard gamuts, it is the same old marketing nonsense. It is the same reason there isn't mention of gamut coverage in the marketing material in anything other than woolly terms.

Please don't get me wrong, I'm not saying this makes it a bad projector, but there is some science here and some of the numbers mentioned appear to be coming out of a random number generator! :)
 

Attachments

·
Registered
Joined
·
765 Posts
I would agree gamut coverage and color volume aren’t necessarily the same thing. Color volume most definitely can be calculated mathematically based on the points and chart given. What I have provided references specifically color volume. At the end of day, personally what matters is how the unit looks with content. I get your point to a degree, but without having any experience with unit criticizing its gamut coverage is a bit pointless. If you said on x movie the unit couldn’t render a scene effectively because of the gamut that is different. I don’t think many if any owners have had gamut coverage issues with current uhd content. P3 inside rec2020 I believe.

I took the color points, plotted them and calculated the volume. Then compared that calculation to the same method for the LS 10000, Optoma UHZ65 and Sony VW1100. The volume calculation matched the review measured data in those 3 cases closely within 1% ,so I figured it was in that range. The color volume percentage is relative to the DCI P3 D65 color point volume as the basis.

I did this for the red triangle as well, but that is a little short in green of rec 709. So I would say what benQ posted is a high percentage of 709 but not full 709.
Hrmph... I think your method is bogus if I'm honest...

Just because the triangle has a bigger area than REC709 does you can't claim >100% REC709 if it doesn't also include all of REC709! /forum/images/smilies/smile.gif You actually have to be able to display all the colours bounded by the REC709 triangle to have >100% REC709. Once you have all the colours in REC709 covered, and you have a bigger triangle, then you have >REC709 gamut coverage. Otherwise you could claim >REC709 gamut coverage with some truly ridiculously unusable gamuts.

If you plot the proper true REC709 green point onto the BENQ graph (x 0.3 y 0.6) you will see that there is an area to the left side of the blue triangle which does not get covered. This is because the laser green primary in the LK970 image is too far to the right for how high it is. Because you can only mix colours at the points between the green and blue primary along the line joining them at that edge you will never be able to make those green / cyan / blue colours bounded by that triangle slither.

I've added it to the graph and attached so you can see clearly hopefully.



At the end of the day, there is a reason this image shows some nebulous "lamp based projector" and NOT the REC709 and DCI-P3 standard gamuts, it is the same old marketing nonsense. It is the same reason there isn't mention of gamut coverage in the marketing material in anything other than woolly terms.

Please don't get me wrong, I'm not saying this makes it a bad projector, but there is some science here and some of the numbers mentioned appear to be coming out of a random number generator! /forum/images/smilies/smile.gif
 

·
aka jfinnie
Joined
·
4,242 Posts
I would agree gamut coverage and color volume aren’t necessarily the same thing. Color volume most definitely can be calculated mathematically based on the points and chart given. What I have provided references specifically color volume. At the end of day, personally what matters is how the unit looks with content. I get your point to a degree, but without having any experience with unit criticizing its gamut coverage is a bit pointless. If you said on x movie the unit couldn’t render a scene effectively because of the gamut that is different. I don’t think many if any owners have had gamut coverage issues with current uhd content. P3 inside rec2020 I believe.
You are putting words into my mouth. At no point have I criticised the gamut coverage in this thread of conversation. I have pointed out mistaken use of an irrelevant graph and shown that by BENQ's own graph there isn't 100% coverage of REC709 gamut, let alone "very high percentage of P3" which was the original statement. I'm not making a judgement on how the projector looks, but if we're going to talk about these things in numbers we should be using accurate numbers, which was my objection on both counts.

If you'd seen some of my other posts you would note I've also pointed out that it is quite common for some lamp based JVC units to equally not be able to hit this corner of the REC709 gamut - my own DLA-X7900 needs the colour filter engaging in order to reach the REC709 primary.

This is all about accurate reporting of quality numbers and nothing else.

(just a small point on colour volume - you can't get to a colour volume from a CIE diagram as a colour volume is a 3 dimensional shape with luminance - which is the hidden dimension on the CIE diagram, it is only a "slice" through the colour volume. Colour volumes can either be self-relative to the display's own peak luminance, or relative to a hypothetical display luminance. There is a pretty good description at the start of this: http://www.spectracal.com/Documents/QSGs/Color_Volume_Analysis_SetupGuide.pdf).
 

·
Registered
Joined
·
819 Posts
in chromapure 3 i can get 108% rec709 if i tweak the color/color enhancer and just measure it, i m not sure if that qualify but at that point the picture is oversturated
 
  • Like
Reactions: Jue Liang and bobof
1261 - 1280 of 2609 Posts
Top