strange performance part 2-with measurements

[Chad B]

I did an Elite 111 today and got the same result- compression of the gamma curve. It's as if ISF mode is turning on a power save function, but I triple checked that, along with the room light sensor and all enhancements (all off).
I had the instructions open while calibrating and followed them step for step. I was using 1080i HDMI, pure cinema 1 (standard).
Please check out these graphs/data and help me figure this out!
Ist run is in Pure mode contrast at 35. Luminance is 112 Y with beautiful gamma.
2nd is in ISF mode, contrast 35, white balance numbers identical to pure mode. Luminance is higher, around 127, and gamma compression is horrible.
3rd run is not shown, because I forgot to save it. I reduced ISF mode contrast to around 27 to match pure mode's luminance at 112 Y. Gamma compression is still horrible, almost as bad as #2.
4th run (3rd attachment), I had to reduce ISF mode contrast to 15 to eliminate the gamma compression. Gamma is still worse than Pure mode. Luminance is now down the toilet.

[Chad B]

This is my 1st day of a 2 week tour full of Elites whose owners want ISF modes done.
In fact, I've got another tomorrow. If I can't get this figured out, I don't know what's going to happen.
I can't do the multipoint gamma if there's compression, because the compression is window size related. So I would have no choice but to reduce contrast to 15, do the multipoint gamma, then maybe bump contrast back up some. If I didn't bump it back up I'd have a display putting out only about 26 ft-l, when it's capable of maybe 30-50% more.

[Turbe]

Are you changing the 9-Point Gamma Controls (any channel on any point)?

You may want to leave the 9-Point Controls at defaults (0), many are doing that.

Did you study these settings?

http://www.controlcal.com/forum/showthread.php?t=96
http://www.controlcal.com/forum/showthread.php?t=77

Can you post your settings and which memory it is for (isf Day/isf Night/isf Auto)?

[Saladshooter]

I am nearly certain this is a setting problem. If you post all of your settings for the ISF mode that is resulting in the error then we can help. Particularly suspicious are your luminance values. In my limited experience, the luminance chart for ISF should mirror your original Pure mode, color by color if you approximate the same 100% white Y value.

You should absolutely not need to touch the 9 point gamma. The problem is originating somewhere else and the gamma will snap in line when you figure it out.

[Saladshooter]

These are my settings based on Turbe's description of controls. Check it against yours for possible variations.

PURE CINEMA
0 - Off
1 - Standard
2 - Advance
3 - Smooth

CTI
0 - Off
1 - On

COLOR TEMP (This Control is also available in Part/Page #1 and #2)
0 -
1 - Low
2 - Mid-Low
3 - Mid
4 - Mid-High
5 - High
6 - Manual

COLOR SPACE
0 -
1 - 1
2 - 2

BLACK LEVEL
0 - Off
1 - On

ACL
0 - Off
1 - On

3DYC
0 - Off
1 - Low
2 - Mid
3 - High

I-P MODE
0 -
1 - 1
2 - 2
3 - 3

TEXT OPTIMISATION
0 - Off
1 - On

INTELLIGENT MODE
0 - Off
1 - On

DRE PICTURE
0 - Off
1 - Low
2 - Mid
3 - High

ENHANCER MODE
1 - 1
2 - 2
3 - 3

BLOCK NR
0 - Off
1 - On

3DNR
0 - Off
1 - Low
2 - Mid
3 - High

FIELD NR
0 - Off
1 - Low
2 - Mid
3 - High

MOSQUITO NR
0 - Off
1 - On

ZOOM
0 - Dot by Dot (FHD Only)
1 - 4:3
2 - Full
3 - Zoom
4 - Cinema
5 - Wide or Wide1
6 - Full 14:9 (All Models except in North America and Japan)
7 - Cinema 14:9 (All Models except in North America and Japan)
8 - Full 2 (Models in Japan Only)
9 - NOT USED
10 - NOT USED
11 - Auto*
12 - Wide 2 (Europe SD Models with Digital Inputs Only - Regions2)

[Chad B]

Thanks. The next opportunity I get I will post my settings.
I had not touched the 9 point gamma adjustment either time.

[scooper750]

Don't mean to thread jack, but I have a couple stupid questions.
I noticed that Power Save Mode should be OFF and Orbiter Mode should be 2.
At what point are those set. Are they set in the user menu after the ISFccc modes are activated? I ask because I noticed they are not settings in the ISFccc panel.

[Turbe]

Hi :hiya:

Inside the ISFccc Interface, there are no controls for those two options (cannot be locked in a ISF Memory). They are set in the User Menus (and can be changed later by the Owner).

I would set the Power Save option first and set the Orbiter after (not sure if others have it off during calibration - hopefully they will comment in regards to how they handle the Orbiter option during calibration).

[D-Nice]

Are you sure you did not allow the ISFccc connection to time out? If you do not initiate a command after an extended period of time, some very weird things will happen......an example would be the panel switching to standard mode without you knowing (if you hit the A/V mode button on the remote you will see that you are actually in Standard mode :eek:). Your luminance curve represents what you would see in Standard mode. I recommend that you hit the save button every 2 minutes to ensure you are not booted out.


I did an Elite 111 today and got the same result- compression of the gamma curve. It's as if ISF mode is turning on a power save function, but I triple checked that, along with the room light sensor and all enhancements (all off).
I had the instructions open while calibrating and followed them step for step. I was using 1080i HDMI, pure cinema 1 (standard).
Please check out these graphs/data and help me figure this out!
Ist run is in Pure mode contrast at 35. Luminance is 112 Y with beautiful gamma.
2nd is in ISF mode, contrast 35, white balance numbers identical to pure mode. Luminance is higher, around 127, and gamma compression is horrible.
3rd run is not shown, because I forgot to save it. I reduced ISF mode contrast to around 27 to match pure mode's luminance at 112 Y. Gamma compression is still horrible, almost as bad as #2.
4th run (3rd attachment), I had to reduce ISF mode contrast to 15 to eliminate the gamma compression. Gamma is still worse than Pure mode. Luminance is now down the toilet.

[zaks]

I must admit that I have experienced something similar on my 151: if I increase contrast beyond 21 in isfccc mode I get the same kind of gamma run out as seen in Chad B's second picture. With contrast at 21 I get a spot on 2.2 gamma curve. That translates to about ~100nits (~29ft-l) maximum brightness for the 100% window, which seems fine to me for night time viewing.

I'm calibrating with my HTPC as the source BTW, and CalMAN set to digital levels (0-255)...

Turbe: I also noticed that when I used RESET once for one of the HDMI inputs it put all 9-point gamma settings at -5... is that normal?

[Turbe]

Turbe: I also noticed that when I used RESET once for one of the HDMI inputs it put all 9-point gamma settings at -5... is that normal?

Well, -5 (-5%) is actually 0 but the default is 0 (0%) which is actually 5.

That is odd that it's not resetting it to the default, but to 0.

I'll have to check my Display.

[JavierS]

Turbe: I also noticed that when I used RESET once for one of the HDMI inputs it put all 9-point gamma settings at -5... is that normal?

Same happened to my with my 5090H

[PioFreak]

When I set gamma to 1, it says S06 on the display. I am guessing this means a value of 6. So maybe mine is defaulting to 5 as well.

[Turbe]

Default is fine, a little odd that reset is setting to 0/-5.

After Reset, you will want to manually select each Point and Set each Channel to 0 (mid point on the Slider). Don't forget to press SAVE after you go through all Points. :)

[Chad B]

Are you sure you did not allow the ISFccc connection to time out? If you do not initiate a command after an extended period of time, some very weird things will happen......an example would be the panel switching to standard mode without you knowing (if you hit the A/V mode button on the remote you will see that you are actually in Standard mode :eek:). Your luminance curve represents what you would see in Standard mode. I recommend that you hit the save button every 2 minutes to ensure you are not booted out.

Thanks for the reply. Yes, I'm pretty sure I didn't let it time out. I kept it busy.
My road trip plans were modified, so I did not have another opportunity to calibrate a Kuro until I did another 151 and 150 yesterday. The 151 showed a similar limiting problem, but the 150 had a weak output anyway (going from memory I think it was showing gamma compression at 26 ft-l in Pure mode) so I didn't even attempt to reduce the output to before compression on that set. I am now convinced that the ISF mode on the 9G Elites is activating some sort of power saving or output limiting circuit. What else can I think when every 9G Elite I've tried has done it vs. Pure mode? Of course, I am far from infallible, and I know no one else has reported the problem, so I have my doubts.
Here are my ISF settings for the 151. When analyzing them, one thing to note is that while I have never before boosted the brightness or black levels on Kuros, this particular 151 did show a need for +1 black level. I used 1080i HDMI (pure cinema standard) for the cable box input and 1080p/24 (pure cinema advanced) for the BD input, both Y Cb Cr color space. My generator (Quantum Data 802BT controlled by CalMAN) produces 1/4 raster area windows, which are larger than DVE and many other generators (Sencore, Accupel). I've thought that maybe I should request CalMAN use a smaller window size than the QuartBox pattern. Regardless, the window size will effect the gamma compression, but it should not cause more gamma compression in ISF mode than in Pure mode, which is what I'm experiencing.
I know window pattern compression is not that big a deal in real life because of the huge differences with window sizes. I calibrated ISF day mode with contrast at 34 and ignored the compression, and it looked great. I will continue to use the ISF modes. My problem is why is there more compression in ISF mode than in Pure mode, which suggests that ISF mode is weaker or purposefully limited? I can't make pretty graphs for the customer in ISF day mode with all that compression.
Cable box input:
ISF night
page 1
pic 16 (anything higher compressed the gamma curve, and this resulted in a 100% window reading 87 Y)
bright 1
sharp -15
color 3
tint 0
color management all 0
color temp 6 (2,0,6,0,0,-1)
page 2
0,0,2,0,1,0,1,0,0,1,0,6,0,0,0,2,0,0
I did not mess with the multipoint gamma.
On the other hand, in Pure mode, I could put pic contrast much higher before compression (35 if I remember right), which resulted in a white window reading 115 Y. All other controls in Pure mode were set similarly to the ISF mode.
If anyone would like to see the graphs and data like I showed in post #1, I can. The ISF mode limited my compression free output to 87 Y, while Pure mode allowed 115 Y.

[D-Nice]

Hi Chad,

I did three Elites between last night and today and none of them produced the problems you are having. If this has happened to most/all of the 9G Elites you have encountered, I would start looking at your equipment, specifcally your pattern generator. Try to use a difference pattern source and see if that helps.
Thanks for the reply. Yes, I'm pretty sure I didn't let it time out. I kept it busy.
My road trip plans were modified, so I did not have another opportunity to calibrate a Kuro until I did another 151 and 150 yesterday. The 151 showed a similar limiting problem, but the 150 had a weak output anyway (going from memory I think it was showing gamma compression at 26 ft-l in Pure mode) so I didn't even attempt to reduce the output to before compression on that set. I am now convinced that the ISF mode on the 9G Elites is activating some sort of power saving or output limiting circuit. What else can I think when every 9G Elite I've tried has done it vs. Pure mode? Of course, I am far from infallible, and I know no one else has reported the problem, so I have my doubts.
Here are my ISF settings for the 151. When analyzing them, one thing to note is that while I have never before boosted the brightness or black levels on Kuros, this particular 151 did show a need for +1 black level. I used 1080i HDMI (pure cinema standard) for the cable box input and 1080p/24 (pure cinema advanced) for the BD input, both Y Cb Cr color space. My generator (Quantum Data 802BT controlled by CalMAN) produces 1/4 raster area windows, which are larger than DVE and many other generators (Sencore, Accupel). I've thought that maybe I should request CalMAN use a smaller window size than the QuartBox pattern. Regardless, the window size will effect the gamma compression, but it should not cause more gamma compression in ISF mode than in Pure mode, which is what I'm experiencing.
I know window pattern compression is not that big a deal in real life because of the huge differences with window sizes. I calibrated ISF day mode with contrast at 34 and ignored the compression, and it looked great. I will continue to use the ISF modes. My problem is why is there more compression in ISF mode than in Pure mode, which suggests that ISF mode is weaker or purposefully limited? I can't make pretty graphs for the customer in ISF day mode with all that compression.
Cable box input:
ISF night
page 1
pic 16 (anything higher compressed the gamma curve, and this resulted in a 100% window reading 87 Y)
bright 1
sharp -15
color 3
tint 0
color management all 0
color temp 6 (2,0,6,0,0,-1)
page 2
0,0,2,0,1,0,1,0,0,1,0,6,0,0,0,2,0,0
I did not mess with the multipoint gamma.
On the other hand, in Pure mode, I could put pic contrast much higher before compression (35 if I remember right), which resulted in a white window reading 115 Y. All other controls in Pure mode were set similarly to the ISF mode.
If anyone would like to see the graphs and data like I showed in post #1, I can. The ISF mode limited my compression free output to 87 Y, while Pure mode allowed 115 Y.

[Saladshooter]

The ISF mode limited my compression free output to 87 Y, while Pure mode allowed 115 Y.

I'm only using AVS and a Bluray player as a generator but I am not seeing any gamma compression even at much higher Y levels on Pure or ISF modes. Definitely keep looking at conflicts in your settings.

[scooper750]

I've only done only one 9G Elite, but like D-Nice I have not run into the issues that Chad have run into. I need to see if I can get my hands on another 9G.

[zaks]

As I mentioned before, I have the same issue with my 151 as Chad B reported: if I go beyond a certain contrast level in isfccc the point gamma curve takes a dramatic dip at the high end (exactly as shown in Chad B's second run picture).

I double checked all settings on both pages - I even RESET, then set everything by hand again: no difference. Power save mode is set to off. Could this be a defect of the set, or a "feature" of my particular firmware revision?

[Turbe]

Zak, is this with the I1 Pro? If so, same with the Chroma 5?

Also, please post your settings and charts/graphs... :please:



I also noticed with Chad's settings, he didn't touch the CMS where D-Nice is on his posted settings. Color is +3 for both though...

[D-Nice]

As I mentioned before, I have the same issue with my 151 as Chad B reported: if I go beyond a certain contrast level in isfccc the point gamma curve takes a dramatic dip at the high end (exactly as shown in Chad B's second run picture).

I double checked all settings on both pages - I even RESET, then set everything by hand again: no difference. Power save mode is set to off. Could this be a defect of the set, or a "feature" of my particular firmware revision?Have you attempted you use another pattern source besides your HTPC? Also, the ISFccc modes expect studio level signals....not PC spec'ed.

[Chad B]

Thanks everyone. The pattern generator is not faulty, but the larger window size may be showing up the problem. The Sencore and Accupel windows are only about half the area of these quarter raster windows. That might explain why only one other person has experienced it; my largish windows are putting the APL over some threshold that smaller windows do not. I did another 151 yesterday with exactly the same results.
Trouble is, the ISF mode looks good, even with the compression. The main issue seems to be that I cannot get good looking "after" calibration graphs. Also, while the picture does look good, this is revealing some sort of problem with the ISF mode. I will post the graphs and data, as well as my settings.

[Chad B]

BTW, I am using Y Cb Cr colorspace with studio levels.

[Chad B]

1. Pure mode- notice the luminance at 100%: 120 Y. Contrast was set at 37.
2. ISF Day- notice the luminance at 100%: 118 Y. Contrast was set at 31.
3. ISF night- I reduced the contrast until the compression disappeared. Luminance is way down and contrast had to be reduced to 18. At contrast 19 there was a hint of compression.
4. Final ISF day- I decided to ignore the compression. Contrast is at 37. Notice the luminance is 128 Y.

These results were the same whether I was in the calibration mode or after when I was totally out of any adjustment modes. I tested this because some TV's act differently in the service menu than out.

[Chad B]

Sorry, it doesn't look like the attachments made it through. I have to resize them.

[D-Nice]

Exactly how large are your test patterns?
Are they taking up 40-50% of the screen if so, your pattern generator is triggering the APL circuitry in the Kuro....all of which is completely normal.
Thanks everyone. The pattern generator is not faulty, but the larger window size may be showing up the problem. The Sencore and Accupel windows are only about half the area of these quarter raster windows. That might explain why only one other person has experienced it; my largish windows are putting the APL over some threshold that smaller windows do not. I did another 151 yesterday with exactly the same results.
Trouble is, the ISF mode looks good, even with the compression. The main issue seems to be that I cannot get good looking "after" calibration graphs. Also, while the picture does look good, this is revealing some sort of problem with the ISF mode. I will post the graphs and data, as well as my settings.

[Chad B]

They are 1/4 raster area.
With 1080P resolution, they are 960x540, which works out to 25% of the picture area. You could fit 4 of these windows, in 2 columns and 2 rows, in the picture area.
Yes, it may be triggering APL circuitry, but this is revealing that if that is the case, the APL circuitry is more aggressive in the ISF mode than in Pure mode.

[Chad B]

Here are the latest graphs and data.

[D-Nice]

They are 1/4 raster area.
With 1080P resolution, they are 960x540, which works out to 25% of the picture area. You could fit 4 of these windows, in 2 columns and 2 rows, in the picture area.
Yes, it may be triggering APL circuitry, but this is revealing that if that is the case, the APL circuitry is more aggressive in the ISF mode than in Pure mode.The ISFccc and Pure are completely different on the 9G Kuros. The APL circutry is more agressive in the ISFccc modes as you can get 50% more light output.

[Turbe]

But that still doesn't explain why you (D-Nice) and Stephen aren't getting this compression issue. I asked David Abrams to check this Thread and comment but I know we would have heard something from him (or Kevin, Jeff or Gregg) if he/they were experiencing this.

:confused:

[Chad B]

But that still doesn't explain why you (D-Nice) and Stephen aren't getting this compression issue. I asked David Abrams to check this Thread and comment but I know we would have heard something from him (or Kevin, Jeff or Gregg) if he/they were experiencing this.

:confused:

Any way we could get an official word from either Pioneer or the ISF?

[Turbe]

Well, I'm not sure exactly what to ask yet.

How much larger are those windowed patterns from your generator vs an Accupel/Sencore? EDIT: NVM, you posted this above.

I wish you could try another pattern source... even do a simply test with AVS HD709 with the customer's BD player...

[Chad B]

Well, I'm not sure exactly what to ask yet.

How much larger are those windowed patterns from your generator vs an Accupel/Sencore?

I wish you could try another pattern source... even do a simply test with AVS HD709 with the customer's BD player...

I used to have a Sencore VP403, but I just sold it a month ago. I always thought it's windows were pretty small. Going from memory, I'd say they were only about half the area of the Quantum Data's. I have an Accupel HDG2000, but it's component only and only has 25% steps. It's windows are small, too. I have a DVDO VP30 that I use as a backup, and it's windows look about the same as the Quantum Data's, or maybe even a tiny bit bigger.
I could try DVE HD-DVD or DVE Basics Blu Ray... Trouble is, I don't have any more 9G's on the calendar for a while.

[Chad B]

Well, I'm not sure exactly what to ask yet.

How much larger are those windowed patterns from your generator vs an Accupel/Sencore? EDIT: NVM, you posted this above.

I wish you could try another pattern source... even do a simply test with AVS HD709 with the customer's BD player...
I would like to know if indeed the ISF mode has more aggressive APL circuitry, and if so, why?
Strong light output is not the enemy! Watching a TV that puts out a meager 26 ft-l is no fun. It's passable in a totally dark room, yes, but it's horrible in a typical living room.
Could it be possible that the comments like "ISF mode is like Pure mode on steroids" is due in part to the limiting, which is boosting or limiting depending on the raster size? Kind of like FM radio, where they compress the audio's dynamic range even more than it was compressed during mastering because it makes it sound louder and "better"?
I know if it were a straight (non window size related) compression it would look bad, probably imparting a clayface look. But with it being window size related, maybe it's only boosting/compressing certain images in a euphonic way and therefore making it look much brighter without obviously bad looking side effects.
I want to clarify, though, that I am merely speculating, and I do not want to become some crusader for the "evils" of ISF mode. On the contrary, I do like ISF mode, and as I said before I like the picture even when I measure all that compression. I thought it looked brighter and punchier than Pure mode. I'm just curious why I like it so much more; is it an inaccuracy that I like?

[Wolfy]

But that still doesn't explain why you (D-Nice) and Stephen aren't getting this compression issue. I asked David Abrams to check this Thread and comment but I know we would have heard something from him (or Kevin, Jeff or Gregg) if he/they were experiencing this.

:confused:Am I missing something. Are D-Nice et al who doesn't notice this compression issue using the same pattern generator as Chad B?

If not, then the size of the patterns may well be the source of this issue. IIRC, DVE uses 18% window size, maybe an increase to 25% is enough to trigger the APL limiter when in ISF mode?

It does seem poor design on Pioneer's part if, which seems very likely, the APL limiter kicks in that much more aggressively at 25% and not 18%. Then again this should manifest itself as quite noticeable floating whites when in ISF mode and watching higher APL scenes.

[Wolfy]

I would like to know if indeed the ISF mode has more aggressive APL circuitry, and if so, why?Well, if ISF Mode yields more light output then a more aggressive APL limiter is simply put in place as a safe guard not to overdrive the display's electronics, but I find it strange that this manifests itself at 25% window size and not at for example 18% window size. Had your window size been 40% then it would have been more understandable.

[scooper750]

Just a little clarification on my experience with Pro 151FD. My signal generator is the Sencore V403C. Anyway, in regards to gamma compression I did not have any issues in ISF Day mode. I yielded a about 43 ftl of output and had a flat gamma. In ISF Night, I calibrated and yielded about 35 ftl but I did experience a bit of gamma compression. I made what I would consider fairly drastic changes to the 10 point gamma to achieve a fairly flat gamma at about 2.15. When I calibrated ISF night and took my initial readings I remember saying .... WTF.... Especially because I never did experience that nuance in ISF Day. At that point I shifted gears into gamma adjustment mode. Thank goodness for 10 pt gamma adjust.

[Chad B]

Well, if ISF Mode yields more light output then a more aggressive APL limiter is simply put in place as a safe guard not to overdrive the display's electronics, but I find it strange that this manifests itself at 25% window size and not at for example 18% window size. Had your window size been 40% then it would have been more understandable.

I am curious; is anyone else experiencing higher light output for the same contrast numbers in ISF mode like I am? For example, in Pure mode contrast setting of 37 yeilds me 120Y, but in ISF mode it yields 128 Y with compression.

[D-Nice]

The reason it's not happening to me is because the test patterns I use do not take up 1/4 of the screen.

But that still doesn't explain why you (D-Nice) and Stephen aren't getting this compression issue. I asked David Abrams to check this Thread and comment but I know we would have heard something from him (or Kevin, Jeff or Gregg) if he/they were experiencing this.

:confused:

[Chad B]

Now that I have my window size to suspect as triggering the problem, it will be easy for me to test. I can either request CalMAN use a smaller window (the Quantum Data has hundreds of patterns, and I would think there would be a smaller useable pattern in there), or I can make my own smaller windows with it's VGM software and when it comes time to make the run I will just control the generator manually like in the good ol' days. Of course, I could also use DVE as I said earlier.
I wish I had another 151 or 111 coming up in the next few days. The idea that window size was triggering it was only starting to take shape with me when I did the last one, and I didn't think to experiment with different sizes.

[D-Nice]

I am curious; is anyone else experiencing higher light output for the same contrast numbers in ISF mode like I am? For example, in Pure mode contrast setting of 37 yeilds me 120Y, but in ISF mode it yields 128 Y with compression.Chad, Pure mode is not the same as the ISFccc modes. Pure mode has a capped light output of around 42fL (+-1fL) before clipping. That's a Contrast setting of 41-42 (varies per set). The ISFccc modes do not have this cap. They peak around 58fL without clipping at the same contrast setting. On all of the 111FD/151FDs I've calibrated, a contrast setting of 37 in Pure mode would yeild 36-37fL and the same contrast setting in the ISFccc modes would yield 48-52fL.

[Turbe]

Am I missing something. Are D-Nice et al who doesn't notice this compression issue using the same pattern generator as Chad B?

Correct. Stephen (scooper750) and Kevin Miller have not experienced this either. I'm sure there would have been comments/Posts from David Abrams and Jeff (umr) if they saw this compression issue too.

The reason I said I don't know what to ask yet is because only Chad has reported this (Zak too, but he's using an HTPC source and I think he needs to try another pattern source too).

This is one of the changes in the 9G ISFccc Interface vs PURE vs 8G ISFccc Interface (as discovered by D-Nice - he has or has had the benefit of owning several Kuros :clap: and can spend a lot of time investigating their performance potential).

[Turbe]

The reason it's not happening to me is because the test patterns I use do not take up 1/4 of the screen.

You posted the explanation in the same Post I replied to (I just read right over it). :devil:


The APL circuitry is more aggressive in the ISFccc modes

[zaks]

Ok, I can now confirm that the gamma compression is indeed dependent on the size of the window pattern.

So far I had used as my source the CalMAN Pattern Generator which, by default, uses 50% as the size of the window. Doing a run with the size at 25% completely eliminated the compression and I get the ruler flat point gamma graph D-Nice promised. ;)

The attached images are based on the settings I had been using so far, with the only difference being contrast increased from 21 to 35 - detailed settings and measurements will follow once I calibrate with this new found knowledge.

Thanks everyone for pointing me in the right direction! :thumbup:

Cheers,
zaks

[Chad B]

I will make some smaller window patterns for use next time and consider the issue resolved; because while I'm not happy about the more aggressive APL, I realize it is more a measurement issue than anything else. Thanks everyone for your help!

[D-Nice]

Hi Chad,

I was actually able to reproduce your gamma compression while using the power save modes in conjunction with ISF-Day. Very weird results if I might say.

[zaks]

I've been giving this some thought, and I'm still a bit puzzled about what is happening with larger window sizes... if this were truly related to the APL, why are the luminance values for a particular step increasing? If the circuitry was truly already trying to limit the power consumption I would expect the luminance to be lower, not higher...

Have a look at the pictures I posted: the data grid at the bottom lists the actual and target luminance (Y and T Y respectively) for each step. For the 50% window Y is much higher than what it should be...

Which brings me to another question: even if we calibrate with a smaller window, wouldn't signals with a higher APL later screw up the gamma curve? I think it would result in highlights getting crushed - I'll have to check that with actual viewing.

Anyways, something to think about...

[zaks]

Actually, forget what I said... I think it makes sense: we increase contrast and the TV complies, only to cut off when the limit is reached. So we are seeing the correct luminance values for a particular value, only the limited value at 100% (or already lower, depending on how far we've already cranked the contrast) screws up the graph in CalMAN because it scales it according to the highest value read and the gamma function in use.

Feel free to correct me if I'm wrong... :)

The question about white crush still applies though...

[vega]

Ok, I can now confirm that the gamma compression is indeed dependent on the size of the window pattern.

So far I had used as my source the CalMAN Pattern Generator which, by default, uses 50% as the size of the window. Doing a run with the size at 25% completely eliminated the compression and I get the ruler flat point gamma graph D-Nice promised. ;)

The attached images are based on the settings I had been using so far, with the only difference being contrast increased from 21 to 35 - detailed settings and measurements will follow once I calibrate with this new found knowledge.

Thanks everyone for pointing me in the right direction! :thumbup:

Cheers,
zaks

how did you get the calman generator to change its pattern size? no matter where I set the size it produces a 50% widow pattern using it as the pattern generator in CalMan. If I produce each pattern manually the sizes are correct.

thanks

[Turbe]

Info:
We also have an option in CalMAN to use 15% windows for just this problem with Plasma.
Derek,
how do I access the 15% windows?
For pattern generators that support window size. Our HTPC, Quantum Data and the Lumigen Radiance. Under Addtional Test patterns you will see the toggle to define what window size to use.
The 15% is what we came up with from testing. AccuPel only has a single window size at this time.