It's been years since I played with HCFR and my Optoma HD80 but I've got my Spyder II sensor out to have a go at calibrating. What prompted me to do it were comments in the HD80 thread that a gamma of -2 in the projector menu would give the desired 2.2 gamma levels.

So I've changed gamma from 1 to -2 and made a few adjustments to Red, Blue and Green gains. The results can be seen from the attachment. The trouble is I'm not really sure what I should be doing next or how to interpret the graphs. May be some kind soul could advise?

Greetings

Looks like your contrast is way too high. The near whites should look pink to you.

regards

hi
Graph 1 shows that your contrast is up too high....as the red starts to diverge with the G and B at above 80.
Graph 2 is useless
Graph 3 needs to be redone after you correct graph 1 (reduce the max light output to match your current 80 ire output measurement).

Thanks guys. I'm setting contrast with Avia and I thought I was following their instructions correctly. :confused:

The second graph - I don't know what's happening. Every time I redo the measurements the red line is always like that.

Graph 3 - Greg how do I reduce the light output to match the current 80 IRE? Do you mean close down my iris on the HD80?

Greetings

Contrast is too high ... so fix it. Turning the iris down won't affect the contrast so it does nothing to fix this.

regards

Okay. Will fix that.

I have another question before I fire up the pj later today... Does it matter that I have my Spyder II situated about foot from and orientated towards the pj (so that it's shadow falls on the windowed IRE areas)? I read elsewhere that it should be sitting a few inches from and directed towards the screen.

Greetings

Point at projector if you are sure the screen does nothing to the image. (Nothing does that)

Sort of like adjusting the stereo without accounting for what the speakers might do to the sound.

regards

Okay, I've had a tinker. What do you think of the settings now? Gamma looks to be tracking under 2.2. How do I raise that? Is there anything else I need to do?

I've tinkered again. Latest results below.

Greetings

Your contrast is still too high. You need to lower contrast until the rgb at the top end converge. That is the maximum usable contrast of the unit.

regards

Wow. I already reduced it from 50 to 42 and other guys in the HD80 threads have theirs around 44. Actually I'm having trouble setting contrast and brightness because when I use my DVE HD-DVD I can't see the below black or 2% above black bars. I can only see the 4% above black. I'll have another go later tonight. The image I have at present doesn't look very punchy at all but that may be because I've always had contrast out of whack.

Greetings

Check that the DVD player is outputting the correct signal type ... check the projector to see if it has some enhanced black hdmi feature that will make the BTB show up.

So even if you pump up brightness the 2% bar still is not there? Never there?

Please note that calibration is not about punchy ... whatever the heck that means. What is punchy to you is not punchy to someone else.

When it is calibrated ... it is calibrated. Set up correctly.

If the goal is "punchy" then it is time to put down the test discs and instruments and just adjust until it pleases your eyes.

You need to be clear on what your goal actually is.

regards

That's right, whatever I do the 2% is never there - which is why I used the AVIA DVD with its moving bars, but I guess that doesn't use the same hi-def settings. I'll recheck my HD-DVD player and projector. Would it matter that the signal is routed via my Denon 1909 receiver and an HDMI switch?

BTW, I want an accurate image! And many thanks for your assistance!

Greetings

Normally that should not matter, but you should check if your receiver also has its own set of contrast and brightness and other controls. (Look for B/W Extension control in the projector)

Check to see if the BD player has these controls ... they may be messed up. Panasonic for instance often crushes and clips their signals coming out of the box.

regards

Hi, I fixed the clipping issue. Don't know how but it now works okay - all I did was change the HD-DVD player's resolution from upto 1080p to 1080p/24.

I have a new set of results if you could be so kind as to check them over.

Thanks

looks like you are clipping red at around 90 % stimulus now.

Hi Greg, what should I do? Is this still a contrast issue? The image seems quite dark compared to what I'm used to, but then it may be just a case of getting used to how things ought to look! How do my gamma levels stack up?

Greetings

When we say clipping ... it always means contrast is too high.

Crushing is a reference to black level set too low.

Reduce contrast until you no longer clip red. The balance of RGB should be about the same through 100% ...

Set it correct first ... and if you want more light outout ... open up the iris. Even more beyond that ... then you bought the wrong projector.

Regards

Thanks again. My iris is set at mid way so I have more to spare if necessary. However, I was wondering about the gamma setting in general. I'm not sure I understand that graph at all. Green and Blue seem to be short of 2.2 while Red also begins below it but ends up going above it. What does this mean? I'm presuming that 2.2 is desired, and if so what can I do about it?

Greetings

Gamma more often than naught takes care of itself when you set the brightness and contrast correctly first.

Since you haven't done that yet ... obsessing over gamma is getting ahead of yourself.

Sort of like wondering why step 90 is not right when you haven't gotten steps 1 to 89 right yet.

regards

In addition to what Michael has said, since you're using HCFR, it might help others if you attached your .chc file so people familiar with that program could look at all your data.

cheers,


--tom

The following is what I do for SXRD contrast, but I think DLP clips in a similar manner.

One way to find where your display is hard-clipping is to run continuous measurements and see how the measures change as you adjust contrast. To do this in HCFR under "view" you can choose "free measures" and then measure a 100% gray pattern. As you adjust contrast watch the R, G, and B measures in the "selected color" "data" area. If you want you can also look at how the primary colors compare against each other by observing the red green and blue bars (same idea as RGB graph), but the main idea is just to watch the R value in the "data" column with different contrast settings.

To start out I would suggest setting contrast very low. Each time you increase contrast the R, G, and B numbers should increase slightly, and generally the RGB levels bars will stay in about the same ratios (percentages). What you will find is that at some point the R number will no longer increase when you turn up contrast, while the G and B numbers will still increase as you turn up contrast. The point where the R number stops increasing is what you might as well consider your absolute maximum available contrast.

You can also do the same thing in reverse. Set contrast very high and watch the R number in the "data" column as you lower contrast. Since you are clipping red on a digital display the R number will likely remain about the same as you turn down contrast from a very high setting, and the RGB levels will generally come closer together. In turning down contrast watch for where the R number begins dropping. The setting where R starts dropping is again basically the maximum contrast setting you can at all use for a calibrated display.

The whole idea of watching the "R" number under the "Selected Color" "Data" area in HCFR while adjusting contrast is just a quick way to find where you are completely running out of Red, or hard-clipping red. Going any higher with contrast will only cause problems with the grayscale. It's possible that a display may still soft-clip and you may need to set contrast a bit lower than where red completely runs out, but at least watching 100% measures will get you in the ballpark for maximum contrast without taking multiple measurement runs.

Fantastic advice. Thanks once again! I used to use Avia to set contrast but for some reason - probably because I wasn't using it correctly - I'd set contrast in the mid fifties on my HD80. I presently have 37 but that's obviously still too high. Just out of interest I trawled through the HD80 thread and in all the years it's been going I can only find 1 posted setting with contrast under 40 - from Guitarman. He posted 32. Obviously every pj is different but perhaps most people actually run with contrast too high?

Greetings

Likely because they don't know any better. 2+2=9 ... until someone points out that the answer is actually "4" ... but 4 is so small ... so that makes "9" right?

regards

Hi, I've tamed contrast. Please could someone give me feedback on my latest settings?

Greetings

See what we were saying about how Gamma tends to take care if itself when you set the brightness and contrast correctly?

The dark end seems off. Is that the best you can do to balance the RGB at 30%

regards

Yes I see now!

I'm trying to get the bottom end better. I've reduced red bias (which I believe affects the bottom end of the scale) which has reduced 30 a lot but at the expense of red from 40 onwards. As you can see from the attachment. I guess I should increase red gain to counter this?

Greetings

The rule of thumb on the dark end is to also use your eyes. Gains = top end ... Bias = dark end.

Our eyes are sensitive to color errors in the red and green on the dark end. Not so much to the blue.

If you have to compromise ... and you always do ... life is one big one afterall ... compromise to blue. Make sure you do not have Red Space in Star Wars or Green Space. Blue space is preferable since our eyes are more forgiving.

regards

Okay, not sure how much more of this I can do to improve things. My latest settings are attached. I still have a red 'sag' in the middle. Don't know how to fix that.

Gamma is reported as 2.66. Is HCFR accurate in this respect? Is 2.66 brighter or darker than 2.22? What can I do to alter that or is it worth bothering with?

Thanks again :)

Okay, not sure how much more of this I can do to improve things. My latest settings are attached. I still have a red 'sag' in the middle. Don't know how to fix that.

Gamma is reported as 2.66. Is HCFR accurate in this respect? Is 2.66 brighter or darker than 2.22? What can I do to alter that or is it worth bothering with?

Thanks again :)

On the gamma graph use the RGB's to be on the white box for 30 and 80% if you can 20,Then set it the RGB's to 100%.This should help with the clipping on both ends.On 20 ire it's too low which could be what's making you not see 2% black.

From the manual (http://marketing.optomausa.com/PDFs/usermanuals/Optoma_HD80_Manual.PDF), it appears you have a Gamma control with different curves, and the Degamma control gives a choice of four predefined tables. You could try different values to see what the effect is.

Alternatively, as you are already adjusting the gains and biases for the grayscale, you can affect the level of the gamma directly with these. At the bottom of the HCFR grayscale grid, it gives the target Y values in the line labeled gamma Y. You can set your target gamma as the reference gamma in the Advanced -> Preferences -> References dialog. Then adjust the RGB gains together such that the overall Y value at 80% is near that calculated by HCFR, and do the same for the biases at 30%. Hopefully, the curve between the two keeps generally the same shape they don't get too far out of balance, but you will probably need to tweak the balance again slightly.

Thanks. Just to make sure I understand - is a higher gamma figure a brighter picture or a darker one? HCFR reports an average gamma of 2.66 for my picture at the moment so if I head for 2.22 will I be making the image darker?

Thanks. Just to make sure I understand - is a higher gamma figure a brighter picture or a darker one? HCFR reports an average gamma of 2.66 for my picture at the moment so if I head for 2.22 will I be making the image darker?

In general, 2.66 means that the picture is darker but more importantly is the shape of the gamma curve. If HCFR is reporting an average gamma of 2.66 it really doesn't tell you any meaningful information.

As mentioned previously, you may want to zip and attach your HCFR .chc file so all the data and graphs are available for folks to look at. Just looking at individual graphs typically doesn't tell the whole story.

cheers,


--tom

Thanks. Just to make sure I understand - is a higher gamma figure a brighter picture or a darker one? HCFR reports an average gamma of 2.66 for my picture at the moment so if I head for 2.22 will I be making the image darker?

You can confirm what Thomas wrote by looking at the luminance graph you posted most recently as gamma7.png (http://www.avsforum.com/avs-vb/attachment.php?attachmentid=147303&d=1246991894). With the dotted white line, HCFR is showing you the predicted luminance for the reference gamma, by default 2.22, at each level of stimulus. So, throughout the middle range you can see that your picture is darker than it would be with the lower gamma.

As indicated by Thomas, the gamma graph is more informative than the luminance graph, as it shows the measured gamma at each point.

Thanks again!

This is a big learning experience!