or Connect
AVS › AVS Forum › Display Devices › Display Calibration › Free calibration soft (non-excel)
New Posts  All Forums:Forum Nav:

Free calibration soft (non-excel) - Page 8

post #211 of 3254
yowza, is that typical of the Spyder2?
post #212 of 3254
Not too surprising ... HD Holic, our purpose is not to compare really... That require some work and could depend conditions (CRT, LCD, DLPs....) but as anyone can imagine an SpectroMeter would be a more accurate than usual colormeters... (by design)... etc...

Also keep in mind that sub 30IRE are pretty hard to measure.

This is a whole subject in itself and not the purpose of that very thread...

--Patrice
post #213 of 3254
The results you posted seem to indicate poor low IRE readings. In other words, you should be measuring out of the projector for accurate low IRE readings, instead of on the screen. You use a screen RGB(?) offset to keep the PJ and screen calibrated as a system. This way, if the screen pushes a certain color, you still get accurate results when calibrating out of the projector.

See here:
http://www.avsforum.com/avs-vb/showt...&&#post8605066

Any chance we could get this feature added? An option to input screen offset so we can measure directly out of the projector?

More here, on how to add a feature to calculate screen offset:
http://www.avsforum.com/avs-vb/showt...&&#post8197084

I would love to see both of these features in your software. Excellent contribution to the community, by the way. I'd feel compelled to donate some too, if you were to set up a PayPal account.
post #214 of 3254
Joe, your caps are fines.

I understand it is pretty frustrating, this probe is usually no brainer to build, I have made about twenty DIY version (prototypes) and then we've build more than 500 of them without a single problem...

I'd suggest you go back to basics, checks wiring, short circuits, measure DC lines, etc... Schematic will help you (you can download pdf here : http://www.homecinema-fr.com/colorim...USB_Schema.pdf

I can propose to look at your probes but they'll have to travel a bit

--Patrice
post #215 of 3254
I would really like to find the issue so others may avoid the same. I am beginning to wonder about specific variations between components avaliable to you(France) and to me(NZ)(Could be the France vs All black effect ).
Although not with standing my build quality, I did source the farnell referenced parts from farnell, plus Switches 947-1707, 4mhz quartz 971-2283, bc547 1097287.


anyway, in my fiddling the second unit has suddenly come to life, well the green led is going and the software recognized that this one has no IR programming yet. So it is communicating atleast.

I can ask the sensors for version which returns 5.1 for both units.

At first I could calibrate the first unit, well without reported error and any measurement could have been wrong. Probably was as per your comment about results.

Now when I try to calibrate both sensors they have addivity errors. Typically.

Bad addivity for Red: 9323 instead of 12149(-23.3%)

I would really like to find root cause, but before I send you the 2 probes for examination I would like to exaust any tests you could suggest I could try.

Could a small self test programme be writtern to look for issues like low voltage or irregular voltage and chip function?

Could you examine a working sensor and give me test points of typical voltages.
I tested for main voltage after usb connection and returned 5v.


Regards
Smokey Joe

PS I had the 2nd probe die again, found if it can die unplugging then replugging probe 2, where probe 1 can be pluged unpluged without problems.
I held the reset whilst plugging in the 2nd probe and that reset the probe.
I also got the RGB to respond again in probe 1 by doing the same reset option.
I can get tracking by running "measures" in real time in both probes and can deviate the signal by altering what is on the screen, but cannot get either to calibrate without addivity errors. Errors can sometimes go RED green or Blue depending on range from LCD screen. Generally Red though.

Could this be noise with to high a tolerence capacitors, 10% and 20%.
I might alter one with the CMS 1206 100nf (which i sourced today) across the TAOS sensors.
What do you think?
post #216 of 3254
Question
Looking at Preferences>Reference panel, there is RGB Levels 0-255 or 16-235. What's the difference, when would one or the other be used?
post #217 of 3254
Wow... just came across this thread... can't wait to try it. I have CalMan and SpyderTVPro... has anyone compared results between them?
post #218 of 3254
Quote:
Originally Posted by HDholic View Post

Question
Looking at Preferences>Reference panel, there is RGB Levels 0-255 or 16-235. What's the difference, when would one or the other be used?


Not totally certain, but I think that refers to setting true black as either 0 IRE or 7.5 IRE. If so, I think most calibrations are done at 7.5
post #219 of 3254
Quote:
Originally Posted by PJ Moore View Post

I asked Colorvision about the filter onthe Spyder2. Here is their response:

LCD's = ON
CRT's = OFF
Bob Cocciolillo
ColorVision Customer Support - USA


Lee - I assume you are saying to leave the filter ON even for CRTs based on your own experience calibrating CRT systems, and not because some other thread said so. What did you observe that provides better calibration if the filter is left ON during calibration of a CRT monitor or projector?

I am not one to take any OEM recommendation as written in stone, , since many of them do not think of the home theater market as their target market, but I am also painfully aware that some of the advise freely given in this forum is not based on personal experience, and is at best "theoretical" or cloned from some other opinion. I want to learn from those with firsthand experience, if possible, so please pardon me for asking about your personnal experience in this manner.

With Spyder2Pro or Spyder2Express, the software will tell you to take the filter off when calibrating a CRT display. SpyderTV does not. From what I've read over at the CALMAN forum about the spyder2 probe, when you turn on CRT(direct view, not RPTV), the sensor will sync to the scan rate of the display. I don't know why. The readings will be different with and without the filter.

I have used SpyderTV since it first came out, and it always sets white too high, and black too low. Tint and Color seem to work out fine though. I've been using CALMAN to calibrate my Hitachi CRT RPTV with the Spyder2 probe, and it's been worth the time to use it. When I tried it on a direct view TV, I thought the results were more accurate without the filter on, even though SpyderTV never says to remove it for any display type.

The HCFR software looks great so far, I've just not had time to sit down and try it out. Maybe over the long Thanksgiving weekend.

I think that if there is to be any further support of the Spyder2 unit, the writers may have to contact Colorvision and possibly sign an NDA to get the hooks they need into cvspyder.dll.
post #220 of 3254
Laric,
To calibrate projectors DLP with Spyder2,
what Calibration mode I have to select, LCD or PROJ?
post #221 of 3254
Hi HDholic
Quote:
Originally Posted by HDholic View Post

Question
Looking at Preferences>Reference panel, there is RGB Levels 0-255 or 16-235. What's the difference, when would one or the other be used?

This reference is used when displaying patterns. The RGB values sent to PC graphic card are either 16-235 or 0-255. That's useful in many cases. When the PC is your usual source, depending on your settings (software/overlay, etc), you must calibrate using either 0-255 or 16-235, to match actual settings. If you calibrate with the wrong setting, your image will be really poor.
Another example, when calibrating an HDMI input, for another source than the PC (DVD player), if this source sends only 16-235, you may want to calibrate it with the PC linked to the projector instead of the DVD player, to make calibration process easier and faster. The PC must be configured to send the same RGB scale than the DVD player.

Regards
Georges
post #222 of 3254
Hi fpr
Quote:
Originally Posted by fpr View Post

Laric,
To calibrate projectors DLP with Spyder2,
what Calibration mode I have to select, LCD or PROJ?

For DLP projector, you should use the LCD mode. DLP spectrum is closer to LCD spectrum than CRT spectrum (that's logical, they use the same technology for lamps )

Georges
post #223 of 3254
Yes

And don't forget to use 10% or 20% or small (I don't reckon how we translate) patterns when using a CRT Projector...

--Patrice
post #224 of 3254
Quote:
Originally Posted by laric View Post

Yes

And don't forget to use 10% or 20% or small (I don't reckon how we translate) patterns when using a CRT Projector...

--Patrice

Hmm.. I always use full field on my CRT pj.. The difference in luminance is very small, about 2%...

// Lyckman
post #225 of 3254
Hi Laric, any comments about my 226 post?
post #226 of 3254
Hi Joe,

I really don't think your have any components troubles, it is pretty unlikely.

Software cannot be done to check the hardware, but a good indicator is the green (or whatever color you choose) LED... When lighted, it means the CPU (PIC) works fine, the USB driver and pc link is ok and obviously minimal wirring is ok.
Then, using software, version checking of prob. also fully validate link.

The only remaining are the IR leds, we don't care they are not vital (only use to instruct a DVD player), you can even remove them and associated components if you have any doubt... Second, remaining "bit" are the two sensors...

The only way to test them is to measure something, idealy RGB (gammut)... you have a debug mode that fill in the "decoder.log" file in app dir to help (us) debuging.

There is no problem in plugging/unplugging many times (done it thousands of times here ).
If the led did not light (in fact intermitent) when plugged, please reset the proble (lower button)... That wil "reboot" the pic and should sort out connection problem (assuming driver is there). Also have in mind driver installation is required on each USB port you plug the probe (each have their own signature) but XP should warn you anyway.

When the probe lights, start software and chose to not use any profile but do your calibration... after RGB, white and black, acknoledge the additivity and do RGB + YMC measures. (prior, go in probe setup and activate the debug)

I'd suggest you try this and send me your .chc and decoder log files...

--Patrice
post #227 of 3254
Hi Laric, thanks for response.

I will create a chc and log file for each probe tonight.

Although I have basically fumbled my way to the final part you have stated to calibrate the probes against my PC LCD screen. In this case I cannot calibrate as both report the Red addivity errors as typically "Bad addivity for Red: 9323 instead of 12149(-23.3%)"

I can't imagion I overheated the second probe sensors as I tried to balance the melting of solder and connection time to the lowest possible. Although it is still possible.

I played around with runs grey scale and colors without calibration to see if sensors respond to changes and it appears they do.

Could this be all IR interference?
Out of interest I happen to have a 880nm narrow band IR emitter which I use for head tracking hardware and pointed that at the sensors and they respond.
I'll also create a "measure" log which will track the realtime response to this.

I will be watching movies on Friday so will do a run using the AE900E calibration file aswell. I might be able to borrow a correction filter by then aswell.


Regards
Smokey Joe
post #228 of 3254
Patrice

Tkx a lot for the software.
Though I get some very slight different reading using this program over Spyder2 colorimeter. They are fairly insignificant. The user interface is more friendly then all other apps I tried and gets me faster to the point. I Love it. If there are features I can think of that can imporove the value of the app, they would be: addition of screen offset, longer reading time for Spyder Pod, and support for Eye1Pro

Re color accuracy of spyder pod in the low IRE, my experience reflect your graphs. When measuring colors with Y<12, I get incorrect x,y values. And this can be easily seen visually without even using reference spectroradiometer. on my pod for instance, I get noticable color shift towards red after calibration when doing measurements with Y <12. Hence the important need of Screen offset and emissive mode.
Yet for gamma tracking, The pod Y sensitivity succesfully go much lower then 12 and the 0.02 advertised value is possible.
post #229 of 3254
Thanks The Cow...

We have same experience on probes And that's a reason we work hard on our one to try doing better on sub 30IRE reading... (and acheive more than expected )

Joe, with IR filter, our probe (well the sensors) can measure IR and that obviously impact results... Some projectors bulbs do have large IR emition and then perturb measure.
I'd encourage you to use an IR filter in such case.

--Patrice
post #230 of 3254
Patrice,

Are you folks working on an English Help file? I'd really appreciate that and hopefully others might as well. It looks like your Help section is very good but unfortunately I can't read it. I figure it will be faster for you to translate it than for me to learn how to read French!

Wayne
post #231 of 3254
Hi Wayne

We are working on it. But it takes time, and our time is precious And we also have quite much work on the next release

Anyway, English help file is on the grill, but I can't give you a firm release date for it.

Regards
Georges
post #232 of 3254
Georges,

Did you have a chance to compare the accuracy of your DYI probe with Spyder2?

I think if people here knew that it offers better calibration than Spyder probe, then more of them would be made here.
I know we are all very excited about the software, but no one has really discussed the pros and cons of the actual DYI probe. Your personal opinion would be greatly appreciated.
post #233 of 3254
Hi Laric, George.

I can't say more how fantastic this DIY software and hardware is. You guys deserve much credit for something which alot of ppl would turn into big bucks for their pockets; you guys have kept it in the DIY user developed world. Absolutely fantastic.

Now back to business. Developing a probe for myself.

This is runs on 1.11 software with my second probe I made.

For some reason the probe has difficulty starting once plugged in, as I was
checking for voltage on the curcuit lines and I touched the green LED it booted into life.
Either it takes a while to build up voltage supply or there is a poor cucuit some where. So far all curcuits prove to be ok for conductivity. I haven't seen any dry joints or poor contacts so far.
I am using a 16v 100uf electro, maybe this does need to be 10v type. I havent found any locally though.
Once running it has no trouble to restart if unplugged, replugged. But next day operation it has issues. Does the 16v cap hold charge that disipates, but maybe slow to charge up?

I ran runs on each sensor in order, sensor 0 then sensor 1 and then combined.
I found unless I set addivity error to atleast 30~35% it fails as each channel is around -20 to -30% addivity.

You will note a separate decoder log for each setup.

Since the default is -15% is this too high?

The sensor it contained in a case with ports like your case,it was placed central

and 10cm from an PC LCD monitor.
Phillips 170B6
Monitor set to D65 and on a DVI cable.
Video Card is a nVidia 6600GT 128

Run at night time with no external light source conditions.

I had a go at calibration using the combined sensors 0 and 1 thc calibration, I made some improvements overall, with some work to go.

You will note that each run I ran the measures table where I set the scale to 50 and for 5 points starting from left I measured the screen then brought up the test color in order of black, white, Red, green, Blue in front of the sensor to see how the probe responded.

Is this probe performing aswell as other probes made?
If not, why?
If the capacitor brand and tolerance values are to high allowing to much interferance noise on the sensors, if so maybe suggest farnell references which are world wide avaliable.

Overall some progress.

I would like to add to your probe manufacturing website though.
Maybe add a brief tutorial in french/english for probe install, and how to do a calibration without a thc reference file.
With that, a brief discription on each value of probe setup.
Addivity faults or out of range may be an indicator of poor performance of the probe. Maybe have typical range values you should see on a good probe.

Of course this may be in the help file you are working on for english, which I'm sure you will complete at a wecolme pace.

Regards
--Nigel
aka Smokey Joe

 

Probe_B_Files.zip 54.029296875k . file
post #234 of 3254
Is there any pre-packaged kit that is available for the DIY probe? I don't know where to begin tracking parts to build it.
post #235 of 3254
Quote:
Originally Posted by <^..^>Smokey Joe View Post

Hi Laric, George.

I can't say more how fantastic this DIY software and hardware is. You guys deserve much credit for something which alot of ppl would turn into big bucks for their pockets; you guys have kept it in the DIY user developed world. Absolutely fantastic.

Thanks.

Quote:
Originally Posted by <^..^>Smokey Joe View Post

I am using a 16v 100uf electro, maybe this does need to be 10v type. I havent found any locally though.
Once running it has no trouble to restart if unplugged, replugged. But next day operation it has issues. Does the 16v cap hold charge that disipates, but maybe slow to charge up?

This is not a your problem, at least I doubt it, you can even simply remove that cap if you have a doubt... The probe don't "need" it to work. (I won't go in detail, but trust me )

Quote:
Originally Posted by <^..^>Smokey Joe View Post

Since the default is -15% is this too high?

No, the usual results are about 1% (even less) additivity... There must be something wrong on red, and I doubt it is IR as your LCD quite likely don't emit any IR.
Basicaly what is additivity ? on each color (RGB) the measured level plus black should equal white...

Quote:
Originally Posted by <^..^>Smokey Joe View Post

and 10cm from an PC LCD monitor.

You should pace the probe straight close to the screen, it must touch it. (this is to increase dynamic)

Quote:
Originally Posted by <^..^>Smokey Joe View Post

Run at night time with no external light source conditions.

Good.

Quote:
Originally Posted by <^..^>Smokey Joe View Post

Of course this may be in the help file you are working on for english, which I'm sure you will complete at a wecolme pace.

Of course it is (although it is pretty difficult to spoke about typical value as that could be way different...).
There will be an english doc (and likely a German one)... Just dunno when

Nigel, I will now look at your files...

--Patrice
post #236 of 3254
Nigel, first overview of your file doesn't highlight big problems.

I'm chatting with me coleague to try having a good answer for additivity issue...

--Patrice
post #237 of 3254
Georges,

Thanks for the reply about the English Help file. I fully understand. Looking forward to it when you get a chance to complete it.

Wayne
post #238 of 3254
Thanks Patrice,

Out of interest I ran the software1.11 in "simulated sensor" mode and got the same addivity error. Could the error be software induced?

btw I was scrounging up an IR filter from our video and photography dept at the university where i work and low and behold i spot a spyder2pro!!. I now have it in my hot wee hands to test along side my DIY sensors.

I see some late long hours ahead

I happened to notice the cover attachment for LCD monitors is a cyan filter, hmmmmm, the cm500 that the taos sensor recommends is a cyan correction filter.hmmmm hmmmmm

I noticed that the pal iso is version 1.1.1.5 now, I assume some of the errors I mentioned earlier have been fixed?
post #239 of 3254
This is fantastic sw, but since I'm an newbie on measuring stuff so I will ask couple of quick questions.

I have installed version 1.11 but about still says it is 1.1? Maybe this is just an oversight and my sw is correct?

When I installed this sw it had a lot of projector profiles listed, including my HS60, what and how are these used? I cannot find them anywhere and as I don't understand french Help does not want to help me

I measured gray scale with 10 values and primary and secondary colors. There are a lot histograms available, but there is not data all of them. Shift histogram, luminance historgram and combined histogram for free measures are empty. Is this how it should be and how values are entered here?

I would really like to see my Delta E results, but is there a histogram for that?
post #240 of 3254
The v1.11 was just an update to the Spyder interface DLL, the main app did not change, hence it says v1.1.

The profile list is when using our Probe... I assume you don't have it, so don't bother this...
What probe are you using ?

Delta E automatically appear on RGB diagram as soon as you have done measures...

--Patrice
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Display Calibration
AVS › AVS Forum › Display Devices › Display Calibration › Free calibration soft (non-excel)