1. Join Now

    AVForums.com uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.

SpyderTV - calibrate your projector to D65

Discussion in 'Projectors, Screens & Video Processors' started by Gary Lightfoot, Sep 17, 2005.

  1. Gary Lightfoot

    Gary Lightfoot
    Distinguished Member

    Joined:
    Apr 6, 2001
    Messages:
    12,437
    Products Owned:
    1
    Products Wanted:
    1
    Trophy Points:
    166
    Location:
    Surrey. UK.
    Ratings:
    +2,099
    PaulB loaned me his SpyderTV calibration tool, and it comprises a colorimiter and software to calibrate your direct view display (which includes CRTs, LCDs and DLPs, as well as plasmas).

    The software is wizard driven and will choose the best settings on your display to what it considers optimal. That's fine as it stands, but you don't know what these settings are in relation to D65 - it might still be innacurate as far as D65 is concerned, but optimal for what controls the tv has available.

    However, there are some other Colorvision software packages available, as well as a 'back door' option with the SpyderTV software that allows you to see the raw colourspace data in the form of x,y,Y co-ordinates. They relate to the CIE chart and black body curve and indicate how far you are from your selected white point (D65).

    I tested the STV colorimiter alongside Colorfacts with the Trichromat-1 colorimiter, and it is possible to get very close results that are comparable, so it is possible to use this to calibrate to D65, but with some extra legwork involved (since the software isn't as helpful as Colorfacts is).

    You can put the values into a spreadsheet here:

    http://www.keohi.com/keohihdtv/isf/michaeltlv/isf_m-tlv_calibrationreport_michaeltlv_v3_page1.html

    Because the co-ordinates that are derived from the STV software are only accurate to -2% to -4%, the results you put into the spreadsheet (downloadable from page 2) will mean your settings will be innacurate by that amount. However, if you change the target values to the values that I measured whilst reading my DLP @D65, the results will then be closer to 100% accurate.

    In the spreadshet it has target values of x=0.313 and y=0.329. Depending on the software you decide to use, will depend on what values you will put there, but the supplied STV software will require something like x=0.301 and y=0.322. You will have to use a division to get the luminence value closer to an accurate figure so that is there too.

    I can/will supply more info but feel free to ask questions or point out any errors etc (Gordon et al) so we can make sure we get better results and as close to D65 as possible. For £200, this could be Colorfacts Hobby rather than Colorfacts Pro. :)

    I've enclosed an Excell spreadsheet of all the data I took so you can see what Colorfacts read and what the other software read. I can supply links for the other software and eithe PaulB or myself can help point you to other threads over on avsforum if you want more info.

    I have measured my CRT monitor too, and with one particular software, the data was about as accurate as it needs to be, so if a CRTs projected light spectrum is the same as a CRT monitor, then with the right software it should be possible to get accurate results straight from the software without any adjusting etc.

    Gary.
     

    Attached Files:

  2. Gordon @ Convergent AV

    Gordon @ Convergent AV
    Distinguished Member AVForums Sponsor

    Joined:
    Jul 13, 2000
    Messages:
    13,999
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    166
    Location:
    Living in Surrey, covering UK!
    Ratings:
    +2,803
    Gary,

    Good work. However. It should be pointed out that any filter based product (ie all tri- stimulus devices) cannot be accurate compared to spectroradiometers. You are essentially biasing the errors in SpyderTV to the errors in your own measurement tool. You are making the Spyder as Innacurate/Accurate as your own tool. Undoubtedly using this tool, with Garry's co-ordinates will get you closer to accurate. You should not believe though that what you end up with is the most accurate you can have. you should also not presume that D65 is all there is to accurate calibration. I used to use tri-stmulus device, one which even used offsets like you have created to get accurate results but I ditched it to get correct results by using a Spectroradiometer.

    I am sure Garry understands all this but just want to make sure everyone else does. Please do not think I am trying to stop folk from doing calibrations by themselves as I am not. More folk who understand calibration means more Lumagen dealers as only Lumagen support multi point gamma......expect everyone else to follow as usual.......

    Gordon
     
  3. gandley

    gandley
    Well-known Member

    Joined:
    Aug 31, 2003
    Messages:
    5,024
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    106
    Location:
    Paradise Lost
    Ratings:
    +269
    OK i should know this by now :suicide: but how the heck do you go about calibrating ones projector to D65 ?. I mean i can tell my colours are a little off as i just use the user settings for R.G.B by eye to what i feel looks right, so that probaly means im nowhere near D65 knowing my luck, I have seen alot about D65 calibration but have not yet set about embracing this method of calibration as im not 100% clear on how to go about it.
    It took me long enough to get white and black level right :rolleyes:
     
  4. Gordon @ Convergent AV

    Gordon @ Convergent AV
    Distinguished Member AVForums Sponsor

    Joined:
    Jul 13, 2000
    Messages:
    13,999
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    166
    Location:
    Living in Surrey, covering UK!
    Ratings:
    +2,803
    Dustin,

    The problem is that it is very hard to calibrate to D65 without a reference or a tool that is accurate in it's measurement. Our eye has an auto white balance feature so when faced with an image we believe to be colourless we auto compensate. This creates issues if you are trying to do greyscale by eye. Look at a blue image long enough and you believe it's grey. The when you look at a D65 image you believe it's really red....then look at a really red 3500K image for a while you think it's grey then when you see a D65 image it's BLUE....doh!


    The only truth is that when you measure and calibrate you see what the DOP wanted you to. There is a reason why we invented measurement tools.

    Gordon
     
  5. Gary Lightfoot

    Gary Lightfoot
    Distinguished Member

    Joined:
    Apr 6, 2001
    Messages:
    12,437
    Products Owned:
    1
    Products Wanted:
    1
    Trophy Points:
    166
    Location:
    Surrey. UK.
    Ratings:
    +2,099
    Hi Gordon,

    I agree the spectroradiometers are more accurate than the tri-stimulus devises (Specs are accurate to +or- 0.001 in the x,y axis IIRC except at low light levels), but recent comparisons of the new tri-stimulus devices are pretty close IIRC. I'll try to find the comparison charts if I can, but here's a link which says the the Trichromat is more than accurate to SMPTE standards, so at least we'll be in the ball park:

    http://www.milori.com/products/professional/faq.asp

    Dustin,

    I'll get together a process for those that want it, but it can be relatively straight forward. You set the white and black levels, and then run this program. You know what the x and y targets are, so you play a 30% grey field from Avia or DVE, and get the co-ordinates from the SpyderTV software, adjust accordingly and then do 80IRE. Go back and forth between them until they remain stable. Your "x" value corresponds to how much blue or red you have, and the "y" value is how much green you have. If your "x" is too high, increase blue or decrease red. If your "y" is too high, cut green, if it is too low, then raise green (you can counterbalance green with a combination of red and blue, but for a starter, you may want to stick to this method until you get a better feel for what's going on). Ideally once you have set the white and black levels, you don't want to touch green, as setting the levels is primarily setting green, but there is room for manoever. After RGB adjustment you should re-check your white and black level settings which may have changed if green was adjusted in the process. With DLPs (and maybe LCDs), after setting 30 and 80IRE, the rest will often fall into place as well, so what you can then do is run all grey fields from black to white and see how cloese to being the same they are. Plotting gamma is something else but we can cover that later as well.

    Another consideration is finding the maximum levels for your colours, and normaly red runs out before blue and green (due to the properties of the lamp), so if you can find the maximum for red and not exceed that value, that will ensure a flat greyscale throughout, without red dipping above 80/90IRE as it runs out. Calibrating at 100IRE and increasing red until it doesn't alter the x y values is one way to do it, but there are other ways (suitable test patterns - I made my own but may try to make a better one).

    Having the tools is one thing, and it's a great way to learn how things work (which is one of the reasons why I bought CF), but knowing how to use them is also useful and how to determine what ios causing errors etc and how to fix them. That's why Gordon is good at his job and I ask him lots of questions. :)

    Gary.
     
  6. gandley

    gandley
    Well-known Member

    Joined:
    Aug 31, 2003
    Messages:
    5,024
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    106
    Location:
    Paradise Lost
    Ratings:
    +269
    Right..... :eek: , i think i will need to read this again in the morning after a nice strong arabic coffee.

    That just went straight over my noggin. I think it sounds worse than it is.
    So i need the tools for the job. will i need colourfacts to do a proper job?
     
  7. Gary Lightfoot

    Gary Lightfoot
    Distinguished Member

    Joined:
    Apr 6, 2001
    Messages:
    12,437
    Products Owned:
    1
    Products Wanted:
    1
    Trophy Points:
    166
    Location:
    Surrey. UK.
    Ratings:
    +2,099
    Colorfacts is a combination of software and hardware - the software interprets the xyY data from the hardware and makes it easy to understand visually on your lap top with hints, tips and wizards, such as the three colour bars of red green and blue for instance - when they are level, you are at D65. SpyderTV doesn't do that, so you have to adjust the colours until the numbers are close to x=313 and y=329. Not as pretty to look at but the result is the same. The CF software alone costs around $2000 normally IIRC, and the hardware is on top of that: $650 for the TR-1 and the Eye One is now up from $850 to $1400, but can be had cheaper else where. You can see why SpyderTV is so attractive at just £200 (or $299). You just need to use a spread sheet and put the numbers in which isn't as easy but a lot cheaper. :)

    It's late now but I will try to post more info over the next few days, but this link is a start:

    http://www.avsforum.com/avs-vb/showthread.php?t=572098
     
  8. Joe Fernand

    Joe Fernand
    Distinguished Member AVForums Sponsor

    Joined:
    Jan 20, 2002
    Messages:
    24,284
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    166
    Location:
    The Borders
    Ratings:
    +2,573
    Hello all

    Dustin as others have said you need to fully understand what your trying to achieve, know your kit inside out and then have the correct tools.

    The SpyderTV unit is quite exciting on a few fronts - its designed to fill a basic requirement for lots of folk with Plasma, LCD and CRT Displays/TV's who have a wish to ensure they are somewhere in the ball park as far as achieving a decent video image.

    For a lot of enthusiasts who want to go a step further it gets more exciting when you see what Gary and other inquisitive minds can get up to using what is a relatively low cost calibration tool.

    Up until now you've had folk investing time and energy into the full on kit (such as Gordon uses) and then working away to try and work out how to do things within a DIY budget - the SpyderTV + a third party spreadsheets now answers that need if you want to do a bit of DIY!

    Having learned how to manipulate your kit using the likes of SpyderTV and Spyder2PRO, low cost devices, you could then go the other way at some point and buy into the tools the Pros use; that sounds a far more realistic plan to me as laying out a couple of hundred pounds seems far more reasonable than laying out a couple of thousand pounds to then find its all too much (even after the decent coffee:))

    Alternatively for a lot of folk the simplicity of calling in an experienced professional (Gordon, Piers etc) is always going to be the best option.

    Best regards

    Joe
     
  9. Joe Fernand

    Joe Fernand
    Distinguished Member AVForums Sponsor

    Joined:
    Jan 20, 2002
    Messages:
    24,284
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    166
    Location:
    The Borders
    Ratings:
    +2,573
    Hello all

    For those attending the Scottish Sound & Vision show next month we'll have SpyderTV on Display - not sure if we'll have the UK (PAL) release as yet; it may still be the NTSC disc we have.

    We'll hopefully be demonstrating it in its 'intended' form on an LCD or Plasma Display - we wont have the time or facilities for any in-depth assessment of its abilities as a Projector calibration tool along the lines of what Gary has been up to but you'll get to see the basics.

    Best regards

    Joe
     
  10. Gary Lightfoot

    Gary Lightfoot
    Distinguished Member

    Joined:
    Apr 6, 2001
    Messages:
    12,437
    Products Owned:
    1
    Products Wanted:
    1
    Trophy Points:
    166
    Location:
    Surrey. UK.
    Ratings:
    +2,099
    The basics are that you use your DVD player to generate an 80IRE grey field and adjust the red green and blue contrasts (sometimes found in advanced menus) until the co-ordinates equal the target values. Then you do the same at 30IRE using the RGB brightness controls. Then go back and fourth between them until they are close to the target and have settled down from the adjusting (i.e the RGB brightness settings haven't hugely changed the RGB contrast settings and vice versa).

    Then you can take readings from black (OIRE) to white (100IRE) and plot your greyscale. You can use the Y value to plot gamma too, but I'll find out more about that as we go.

    Gary.
     
  11. gandley

    gandley
    Well-known Member

    Joined:
    Aug 31, 2003
    Messages:
    5,024
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    106
    Location:
    Paradise Lost
    Ratings:
    +269
    Ok gary just read your last post again and that has done the trick, now i fully get it.

    i always wondered why there is a RGB colour contast setting and an ordinay RGB colour setting.
    what exactly is the difference between the two?
     
  12. Gary Lightfoot

    Gary Lightfoot
    Distinguished Member

    Joined:
    Apr 6, 2001
    Messages:
    12,437
    Products Owned:
    1
    Products Wanted:
    1
    Trophy Points:
    166
    Location:
    Surrey. UK.
    Ratings:
    +2,099
    There should be RGB brightness for the low IREs and RGB contrast for the high IREs. Not sure what RGB colour controls will do unless it's to move the primaries to attain a better gamut within the CIE chart. The Excel spread sheet shows the CIE (colour 'shark fin'), and each standard (PAL, NTSC etc) will have primary points which form a triangle. The aim is for the display to accurately show all the colours of the standard so when you measure the individual red green and blue, the points should co-incide with the points of that particular standard (PAL, NTSC). Sometimes the primaries may not be in quite the right place but being able to move them along the triangle can help.

    Gary.
     
  13. ROne

    ROne
    Active Member

    Joined:
    Dec 21, 2000
    Messages:
    778
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    18
    Ratings:
    +7
    Best place to purchase this? Is it US only?
     
  14. Jeff

    Jeff
    Well-known Member

    Joined:
    Jul 13, 2000
    Messages:
    5,489
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    106
    Location:
    Basingstoke
    Ratings:
    +256
    Dabs do it for £182
     
  15. Gary Lightfoot

    Gary Lightfoot
    Distinguished Member

    Joined:
    Apr 6, 2001
    Messages:
    12,437
    Products Owned:
    1
    Products Wanted:
    1
    Trophy Points:
    166
    Location:
    Surrey. UK.
    Ratings:
    +2,099
    I didn't ask where PaulB got his one from or how much it was.

    There was some issue of it only being an NTSC version originally due to the test pattern DVD that comes with it, but if you have Avia, DVE or Peter Finzel disk you won't need the supplied disk. I'm not even sure the supplied disk has all the patterns we'd use on it to be honest. £182 seems a good price though. Joe Fernand of The Media Factory sells them as well I believe though I don't know how much for.

    A guy in the USA is devising a Colorfacts style spreadsheet that should model the data in a similar way, and be more useful then the KeohiTV one, though there will be a small fee for it. I'll post more info here when I have it.

    Apparently Mark Hunter of Milori and Colorfacts and now Colorvison will be doing a hardware shoot-out between different colorimiters spectroradiometers etc, so that will be interesting to see how the Spyder compares in the overall scheme of things.

    Gary.
     
  16. ROne

    ROne
    Active Member

    Joined:
    Dec 21, 2000
    Messages:
    778
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    18
    Ratings:
    +7
    Will the spyder2pro allow me to do the same thing? (allow me access to these co-ordinates?)

    It's just that it's available in a local store!
     
  17. ROne

    ROne
    Active Member

    Joined:
    Dec 21, 2000
    Messages:
    778
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    18
    Ratings:
    +7
    I've had a look around, and indeed the spyder two pro V2.0 should be able to do all of this and model a profile with the graphics card to do greyscale.

    I am going to get one.
     
  18. NicolasB

    NicolasB
    Well-known Member

    Joined:
    Oct 3, 2002
    Messages:
    6,046
    Products Owned:
    1
    Products Wanted:
    0
    Trophy Points:
    136
    Location:
    Emily's Shop
    Ratings:
    +664
    Someone really should organise a Spyder powerbuy.
     
  19. ROne

    ROne
    Active Member

    Joined:
    Dec 21, 2000
    Messages:
    778
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    18
    Ratings:
    +7
    I got the spyder two pro this afternoon. Be a couple of days before I report.
     
  20. retrof

    retrof
    Active Member

    Joined:
    Feb 28, 2001
    Messages:
    155
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    18
    Ratings:
    +14
    Will be great to hear how this works out for you ROne. I have finally decided to ditch my old Z1, and have a Z3 on the way from nexnix (along with an Oppo 971 on the way from the US, to replace my Yamakawa 365).
    I am looking forward to working with the settings which you originally posted for your Z3. Also interested in getting a Spyder if the results are worthwhile...

    Thanks,
    -rf
     
  21. Gary Lightfoot

    Gary Lightfoot
    Distinguished Member

    Joined:
    Apr 6, 2001
    Messages:
    12,437
    Products Owned:
    1
    Products Wanted:
    1
    Trophy Points:
    166
    Location:
    Surrey. UK.
    Ratings:
    +2,099
    Hi Nic,

    Joe Fernand may be able to organise a Power Buy as it was him who brought it to the forums attention IIRC and I believe he was going to be stocking it.

    ROne,

    Let us know how you get on. The KeohiTV spreadsheet should do as a start, and let me know what the image looks like if you use the numbers the colorimiter goves you directly. If they look slightly out, try the alternative x,y co-ordinates from my comparison chart and see how they look. It should be interesting to see how the results you get compare with SMART.

    Spyder2One has a colorimiter option under Tools, and this will give you the co-ordinates you want.

    Gary.
     
  22. ROne

    ROne
    Active Member

    Joined:
    Dec 21, 2000
    Messages:
    778
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    18
    Ratings:
    +7
    Will do. Ta.

    How much care is needed in setting up the spyder? Are there guidelines for alignment on a projector?
     
  23. Gary Lightfoot

    Gary Lightfoot
    Distinguished Member

    Joined:
    Apr 6, 2001
    Messages:
    12,437
    Products Owned:
    1
    Products Wanted:
    1
    Trophy Points:
    166
    Location:
    Surrey. UK.
    Ratings:
    +2,099
    ROne,

    I placed the sensor on a tripod pointing at the pj approx 4ft away from it. You can possibly use the luminance (Y value) to get it set up for the best reception by changing the angle etc until you get a maximum Y value.

    I've enclosed the calculation spreadsheet with the results from my projector at D65 using my Trichromat sensor, and put those readings in the r/h side. Assuming mine is accurate, I placed the readings of the Spyder2One readings in the left hand column, and I think the comparison can suggest that even if you used the readings directly from the Spyder2One Colorimiter tool without any correction for inaccuracy compared to the TR-1 (i.e, you aim for x0.313 y0.329), your De values will be just within acceptable limits. If you change the target to the new D65 values that I put into my comparison spreadsheet then you should be pretty close to the Colorfatcs Pro Trichromat-1 sensore I'm using.

    I've looked at the readings another chap with a Gretag Mcbeth Eye One Beamer took, and compared them with my TR-1 readings at similar SpyderTV readings, and my sensor is reading the same or only marginaly different, so I'd feel the readings are very close to accurate.

    I think that UV values relate to perceptible levels of accuracy, so anything greater than dE10 will be noticable with regards to colour deviation.

    Gary
     

    Attached Files:

  24. ROne

    ROne
    Active Member

    Joined:
    Dec 21, 2000
    Messages:
    778
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    18
    Ratings:
    +7
    Okay I had a little mess last night with the spyder2pro studio V2.0 (projector support).

    I first tried my CRT Monitor and the spyder was able to bring this in using its in built wizard system to a reasonable tolerance. It then creates an ICC profile which embeds itself in the graphics card colour management tab, so that your graphics card is being calibrated using its own RGB profiles.

    I don't know if this will relate to the projector as such, or am I right in thinking if I use the wizard to calibrate the HTPC graphics card to projector @ 6500 it will give me a 6500 Desktop.

    Would the ICC profile that is created for the desktop drive the VMR output at the same value? Or will there be colour-space errors between the desktop and the DVD output of Media-Center.

    (I've yet to move onto the CIE plotting - just trying to save myself some legwork first!)

    On the menu system you have several none wizard tools for plotting numbers, and you can also access a RGB realtime adjustment system for user-definable colour temperatures. Whether I can swap between the Studio's interface to take a reading from TheaterTek or not, I'm not sure.

    I was thinking about hooking the unit upto a seperate laptop and reading the other target HTPC rather than using the HTPC for the actual readings, that way I could get numbers from AVIA etc.
     
  25. Gary Lightfoot

    Gary Lightfoot
    Distinguished Member

    Joined:
    Apr 6, 2001
    Messages:
    12,437
    Products Owned:
    1
    Products Wanted:
    1
    Trophy Points:
    166
    Location:
    Surrey. UK.
    Ratings:
    +2,099
    Hi ROne,

    It may work to a point if you first max out the RGB gains without crushing before allowing the ICC profile to be created, but I don't know if that software allows you any projector RGB adjustment, so the profiling is probably not going to be ideal. As you know, PC levels are 0 to 255 and video is 16 to 235, so it may be setting the levels incorectly (i.e the desktop). I personaly wouldn't trust it as you have no way of knowing how accurate for D65 it is.

    What I would do is do a run using a laptop as the calibration PC and the colorimiter option under the tools menu. That will give you your xyY data which you can plug into the spreadsheet above (I labeled it Spyder and TR-1 but it is a basic spreadsheet to help model your readings and will show you how far from D65 it is, including dE values). This should help you with which colour to adjust for the xy values:

    If MCE doesn't have the full grey fields from 0IRE to 100IRE, then you can rip or play AVIA and use those instead. The software player will then be using the correct video settings and you should be good to go to take readings and put them into the spreadsheet. Using a laptop with Avia/DVE and the colorimiter tool means you can calibrate any pj using any source with teh correct grey fields.

    Gary.
     
  26. ROne

    ROne
    Active Member

    Joined:
    Dec 21, 2000
    Messages:
    778
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    18
    Ratings:
    +7
    Thanks Gary, just a one more query; the software with studio2proV2 gives you access to a final colour temperature derived from the xyY data, is there any reason I wouldn't use this as a guide? It's just that on your chart the spyderTV says "calculated" for the temp, so I'm assuming the STV doesn't have a colour TEMP readout?

    Also the proV2.0 software comes with a menu you can access that gives you real time "pots" of RGB colour as well as xyY data available under one of the wizard settings which would allow real-time tune up.

    I did one run last night from one of my SMARTED filter based set-ups and it was more or less 6800 across the board so I thought that was pretty good.
     
  27. Gary Lightfoot

    Gary Lightfoot
    Distinguished Member

    Joined:
    Apr 6, 2001
    Messages:
    12,437
    Products Owned:
    1
    Products Wanted:
    1
    Trophy Points:
    166
    Location:
    Surrey. UK.
    Ratings:
    +2,099
    That's exactly right - the SpyderTV doesn't give you the colour temp, just the xyY info so I used a chromacity calculator to get the CT.

    I seem to remember the colour pots you mention but didn't see them in action - if they move up and down in real time as you adjust the RGBs, and getting them flat means D65 then that's exactly how Colorfacts does it.

    6800K sounds pretty good, so it would be interesting to see how much closer you can get it with the Spyder. One of the main advantages of the colorimter is that it sees lighht in the same way as the human eye whereas SMART doesn't, so the colour temp with SMART is not in relation to D65 but the Spyder results will be (in fact, the colour temp graph in Colorfacts isn't relative to D65 either apparently, but the target to D65 is). Did you plug the co-ords into the spreadsheet and see what your dE figure was?

    I spoke with Mark Hunter (via PM) the other day, and he says the Spyder is pretty accurate to xy 0.003 (the EyeOne is xy 0.001), so the Spyder may actually be more accurate than my Trichromat! He will also be doing a sensor shoot out very shortly over on avs and we should be able to see how accurate they are comparatively. I'll keep an eye out for it and post the results or a link here.

    Gary.
     
  28. ROne

    ROne
    Active Member

    Joined:
    Dec 21, 2000
    Messages:
    778
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    18
    Ratings:
    +7
    Gary - what a chore! Obviously the software is lacking quite a lot from what you have, with just colour temperature and xy data to make decisions from, as well as odd responses from the Z3s RGB, offset & Gain makes it quite difficult to calibrate.

    I have now got it reasonably close though 6500 +/- 100 across the range save 10 IRE, which I conclude will always be way out.

    The picture looks nice, and shames my previous acceptence of what color balance should look like especially in the upper IRE. This seems due to the Z3s initial picture presets, such as Powerful, Creative Cinema etc which effect the way the picture responds to a signal in terms of Gamma & Gain. With the Powerful preset it has artificially been boosting contrast right at the top of IRE, meaning no mater how I played with RGB I couldn't get a decent reading close to 6500. Basically contrast was too high with far too much green. I have had to lower contrast by 25 steps to get close, which has knocked the punch albeit artificial punch out of the picture.

    Whether I can get a better CR with another preset I'm not sure. I am starting to believe once calibrated to 6500 across the range, and with C and B set as per AVIA that will be it, irrespective of the initial preset.
     
  29. theritz

    theritz
    Active Member

    Joined:
    Aug 17, 2002
    Messages:
    2,451
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    51
    Location:
    Ireland
    Ratings:
    +27
    Rone,

    Are you calibrating the Z3 using filters or not ? Any chance that you'd publish your settings for others to have a look ??

    Following with great interest, thanks.


    Sean.
     
  30. ROne

    ROne
    Active Member

    Joined:
    Dec 21, 2000
    Messages:
    778
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    18
    Ratings:
    +7
    Sean, see over on the z3 smart thread, most recent post.

    I've put the latest settings with filters there.

    http://www.avforums.com/forums/showthread.php?p=1983285#post1983285

    ROne.
     

Share This Page

Loading...