[Pecker]

How good is a DIY calibration (using sensors, etc, not just DVE) compared to ISF?

How close will DIY get you?

Also, how close should a calibration be, be it ISF or DIY? What I mean is, I'm sure no display, no matter how well it is calibrated, will be 100% spot on - the blue will still be out, even if it's only by 0.001%, or whatever.

So what's the 'standard'? What's the maximum it can be 'out' before you can say it hasn't been calibrated properly?

Cheers.

Steve W

[Gordon @ Convergent AV]

You calibrate to get it as close to the standards as is possible within the physical limitations of the device. Once that is acheived then you have "calibrated" the display. How close that is to the standards when you are finished is dependant on the display and the tools you have available in it.

If you have an understanding of what you are trying to achieve and you have the tools and the patience then I would imagine that you could do a very good job yourself. The forums are littered with folk who do not appear to have the correct tools or understanding as to what they are trying to achieve....but my best customer does his own calibrations, is self taught and has all the tools and test pattern generators you could ever want and I expect he does a fine job on his tv's and projector.

Gordon

[KelvinS1965]

Originally Posted by Gordon @ Convergent AV

.........If you have an understanding of what you are trying to achieve and you have the tools and the patience then I would imagine that you could do a very good job yourself.
Gordon

My biggest issue now, having read and absorbed as much as I could on here and other websites, is that I bought a sensor that was commonly recommended as being accurate (for it's cost); the I1-LT. I managed to acheive delta E on the greyscale on my TV of less than 2 from 10 to 100IRE (using my Lumagen's 11 point greyscale adjustment). I watched a few nights of TV with these settings and felt the picture looked better, with more shadow detail and depth to the image. I felt elated that I'd managed this after only a month or so of getting the sensor, then I had two issues:

I looked at a greyscale ramp test pattern and it resembled sticks of pastel shades of rock (slight pink here, green there, etc).

I re-ran my greyscale calibration check and it was out at many IREs by a delta of 5 or even 10, yet I'd made no adjustments, just did it another day and got different results again. Basically the runs were not repeatable, I had no idea which was the 'correct' one, so if I adjusted again it might still measure 'out' when I rechecked later (even later the same night). This is an LCD TV, so no excuse that the lamp had dimmed drastically between measures.

I did do a very quick greyscale run on my HD350/HDQ as well, but who knows whether I've made it worse or better as again it wasn't repeatable. In the end I cleared the HDQ's greyscale adjustments to default and simply set up the 'main' controls as best I could (contrast and brightness mainly).

I figured I'd put some more hours on the HD350 and then get a professional calibration done. TBH I've kind of mellowed to the idea and just left the colour 'mitigation' settings applied in the HDQ (seemed to reduce the 'neon green' effect and did measure closer to rec709 than without, but again, I don't really know for sure how much closer). I think if I do revisit calibration I'll get a pro in, because I just can't justify the cost of something like an I1-Pro only to find that it too isn't stable/repeatable enough, when a pro will have something much better and all the experience to get the best out of it and your equipment. I'm not sure what they will make of using a HDQ for colour mitigation rather than a 'proper' CMS, but I'll ask before commiting myself.

So in short, I tried it, was convinced by it (briefly) and then became disallusioned by DIY (and still no sign of the AVForums course that was mooted a while ago). At least I'm just enjoying my PJ (and TV) which no doubt pleases SWMBO as I'm not constantly fiddling about with the settings.

[Pecker]

Kelvin, your experiences mirror my own concerns.

I suppose the standard of calibration is dependent on the standard of the calibrator and the standard of the equipment used to calibrate.

This is why I ask the question - what standards are set for equipment? How accurate is accurate?

I mean if we were talking about the Olympic Javelin competition, I don't doubt for a second that the measuring equipment used must be accurate to within 1cm (or 1mm, or 0.1mm, or whatever).

If blue saturation is too low on a projector, and I simply 'turn it up', then is that calibrated? I doubt it very much!

So how near to 'right' must it be?

How accurate must the measuring equipment be?

ISF are 'standards based' - surely there must be a standard as to how close they get to the standard, otherwise it becomes a bit of a farce.

Steve W

[KelvinS1965]

I'd wait a more detailed and accurate answer from Gordon or one of our esteamed ISF members, but I can at least make these observations:

When calibrating greyscale and colour gamut there are the various standards that can be measured to such as (but not limited to) achieving a greyscale of 6500 Kelvin or how close to rec709 the colour gamut measures. How closely these measurements meet the specifications can be expressed in terms of Delta E. It is my understanding that for greyscale a Delta E of less than 3* (*might be 2) is the point at which our naked eye can't tell any difference. There is a similar limit for colour, so this provides a 'target' for the calibrator to aim for.

It might be that for a particular display it is not possible to acheive delta Es within the above limits, for example rec709 with my HD350 on it's own. In this case the calibrator can only work within the limits of the display's capabilities...whether you consider the end result 'calibrated' is a matter for debate I suppose.

With a display capable of adjusting it's greyscale and colour gamut fully (perhaps a HD750 as an example, though there are others no doubt), then it may well be possible for a good calibrator to achieve Delta Es of less than 3*. The accuracy of this reading with a very good quality sensor might mean a tolerance of 0.5 Delta E (just a guess on this point) and that, I suggest, is what we are paying for when a professional does the calibration. Basically I would expect a proffesional's equipment to be accurate enough that the tolerance for a measured 'perfect' greyscale/colour gamut to remain with the Delta 3* limit mentioned above.

With my I1-LT I got a 'tolerance' of over 5 between runs on the same display, which could take a 'perfect' greyscale to over Delta 5 off, which is above the point at which we can see the difference. I would expect an ISF calibrator to have a high quality sensor(s) and equally have an in date certificate for their test equipment (I have to before I can even touch my customer's equipment if I am to calibrate it, though not AV equipment I have to add).

I hope Gordon will reply at some point and dig me out of the hole I seem to be getting myself in.

[Neil Davidson]

Back in the old days (2 years ago) we used to teach 4dE as the target accuracy during a calibration. These days there are displays out there where you can get 2dE or less - but note that this is still almost impossible across every control with the colour management systems the worst offenders.

[Neil Davidson]

Originally Posted by KelvinS1965

I would expect an ISF calibrator to have a high quality sensor(s) and equally have an in date certificate for their test equipment

My i1Pro needs a new Xenon lamp (for reflectance measurements) and recalibration - £170 and shipped to Switzerland for the work to be done

[Pecker]

Originally Posted by KelvinS1965

I'd wait a more detailed and accurate answer from Gordon or one of our esteamed ISF members, but I can at least make these observations:

When calibrating greyscale and colour gamut there are the various standards that can be measured to such as (but not limited to) achieving a greyscale of 6500 Kelvin or how close to rec709 the colour gamut measures. How closely these measurements meet the specifications can be expressed in terms of Delta E. It is my understanding that for greyscale a Delta E of less than 3* (*might be 2) is the point at which our naked eye can't tell any difference. There is a similar limit for colour, so this provides a 'target' for the calibrator to aim for.

It might be that for a particular display it is not possible to acheive delta Es within the above limits, for example rec709 with my HD350 on it's own. In this case the calibrator can only work within the limits of the display's capabilities...whether you consider the end result 'calibrated' is a matter for debate I suppose.

With a display capable of adjusting it's greyscale and colour gamut fully (perhaps a HD750 as an example, though there are others no doubt), then it may well be possible for a good calibrator to achieve Delta Es of less than 3*. The accuracy of this reading with a very good quality sensor might mean a tolerance of 0.5 Delta E (just a guess on this point) and that, I suggest, is what we are paying for when a professional does the calibration. Basically I would expect a proffesional's equipment to be accurate enough that the tolerance for a measured 'perfect' greyscale/colour gamut to remain with the Delta 3* limit mentioned above.

With my I1-LT I got a 'tolerance' of over 5 between runs on the same display, which could take a 'perfect' greyscale to over Delta 5 off, which is above the point at which we can see the difference. I would expect an ISF calibrator to have a high quality sensor(s) and equally have an in date certificate for their test equipment (I have to before I can even touch my customer's equipment if I am to calibrate it, though not AV equipment I have to add).

I hope Gordon will reply at some point and dig me out of the hole I seem to be getting myself in.

Cheers, that looks like a pretty comprehensive answer to me.

Calibrating beyond the point where the human eye can see a difference is probably a tad excessive.

So, of the calibration sensors on the market, are there any which do not achieve this accuracy? Which ones do?

In short, what's the cheapest that can get you that accurate? I ask this in the grounds that spending less means you won't be calibrating your equipment properly, whilst spending more won't get you any greater accuracy than you can see, and as such would be an inappropriate use of funds (hope that makes sense).

Steve W

[Neil Davidson]

No colorimeter device will be able to ensure accuracy of 2dE or less on all display types at this time.

The i1 Pro should be able to offer this accuracy on most display types with a couple of possible exceptions.

Exceptions are actually based on the spectra of the light being produced by the light source in the display. Plasma, Mercury and Xenon projection, and all non LED LCD display light sources should be fine.

LED or laser based light sources can produce problems due to very narrow spikes in their light spectrum. The way colorimeters work make it difficult for them to deal with very narrow spikes in the spectrum and, particularly with LED based LCD you can (but don't always) get some larger than expected errors.

The i1Pro has a resolution down to 3.5nm over a range of 350-750nm so you can see the spikes do need to be pretty narrow not to be picked up. I think that the argument for 2nm or beter resolution only stacks up if you are regularly handling commercial digital cinema calibration where a 1 or 2% improvement can be appreciated (maybe!) due to the scale of the screen.

[Neil Davidson]

Reading between the lines here...

The CalMan Spyder package will work very well for most people on almost any display. This is the cheapest way to get a consistently good level of performance IMO.

The CalMan i1Pro package works as good as pretty much anyone would need at this time on most displays. The only reason to spend more would be if you were doing commercial calibrations but even then the price jump to something like the Orb Optronix at £7k or so would be hard (very hard I can tell you) to justify.

Of course if you are minted and just like the best toys, can I interest sir in a PR655....

[Pecker]

Originally Posted by Neil Davidson

Reading between the lines here...

The CalMan Spyder package will work very well for most people on almost any display. This is the cheapest way to get a consistently good level of performance IMO.

The CalMan i1Pro package works as good as pretty much anyone would need at this time on most displays. The only reason to spend more would be if you were doing commercial calibrations but even then the price jump to something like the Orb Optronix at £7k or so would be hard (very hard I can tell you) to justify.

Of course if you are minted and just like the best toys, can I interest sir in a PR655....

Neil, I think this is exactly the information I was looking for. Thanks.

So, CalMan Spyder package/CalMan i1Pro package. Where from? How much?

Steve W

[KelvinS1965]

Originally Posted by Neil Davidson

My i1Pro needs a new Xenon lamp (for reflectance measurements) and recalibration - £170 and shipped to Switzerland for the work to be done

TBH, that doesn't seem excessive considering the work that you could get out of it before the next calibration, as I said above I would expect this of a proffesional anyway as a matter of course. I would feel more sympathetic if it comes back and the 'as found' results were out of spec; would you have to go back and redo any work that might have been done with an out of spec sensor? I used to perform 6 monthly calibrations in the pharmaceutical industry: If the 'as found' was out of spec they had to evaluate the possible impact on any work they had done since the previous 6 monthly visit, as they couldn't pin point exactly when the reading had gone out.....nightmare.

I was more surprised by the use of an i1Pro per se as I thought they fell into more of an 'advanced DIYer' catagory than a professional tool, but from what you say that isn't the case. I had considered buying one myself, but figured the cost of one would allow me maybe two proffesional calibrations as I'd assumed that they would be using something 'better' and with all the experience that would come with it, maybe I'll reconsider...my only worry is storage of the sensor causing it to drift with age. I could recalibrate my PJ and two TVs while it is new so I'd get some mileage out of it, I may even be buying a new TV for the living room, so that would be a good time to commit to buying a better sensor. If it's only £170 (compared to maybe twice that for another ISF) to have it recalibrated it would be a reasonable cost to have done maybe once a year, or perhaps when the PJ lamp needs replacing. It's got me thinking anyway.

Originally Posted by Neil Davidson

Reading between the lines here...

The CalMan Spyder package will work very well for most people on almost any display. This is the cheapest way to get a consistently good level of performance IMO.

I had seen some mention of the Spyder before I bought my i1-LT and the general recommendation seemed to be that the Spyder was/is better at lower light levels, but less accurate in general than the i1-LT. Given my dissapointing results with the 'better' i1-LT, I'm not so sure about the Spyder myself, perhaps I'm overly cynical given my poor experience.

Originally Posted by Pecker

Neil, I think this is exactly the information I was looking for. Thanks.

So, CalMan Spyder package/CalMan i1Pro package. Where from? How much?

Steve W

Useful thread here, which I think I may have read before buying my i1 as it seems to speak highly of it. i1Display2 vs i1Pro AVSForums

I found the Calman packages available here: CalMAN - Purchase

I'm sure Manni01 posted a link or maybe it was a PM that I've since deleted for a UK supplier, but I can't find it ATM. I'll post back if I do.

[Neil Davidson]

Originally Posted by KelvinS1965

I was more surprised by the use of an i1Pro per se as I thought they fell into more of an 'advanced DIYer' catagory than a professional tool, but from what you say that isn't the case.

I also have a Sencore OTC1000 with a NIST certificate along with a bunch of other probes. As the old adage goes it is not the price of the probe but what you do with it that counts!

Somewhat more seriously an accurate measuring device is only one part of the equation. Experience and knowledge are the other parts that combine to separate a good calibration from a bad one.

Originally Posted by KelvinS1965

I had seen some mention of the Spyder before I bought my i1-LT and the general recommendation seemed to be that the Spyder was/is better at lower light levels, but less accurate in general than the i1-LT. Given my dissapointing results with the 'better' i1-LT, I'm not so sure about the Spyder myself, perhaps I'm overly cynical given my poor experience.

To clarify, I was referring to the new Spyder3 package from the guys at CalMAN. This system includes a correction file for the Spyder3 probe that is included. Each probe is individually analysed before shipping to ensure accuracy. This is a pretty nice way to do things I think.

[Pecker]

Originally Posted by Neil Davidson

To clarify, I was referring to the new Spyder3 package from the guys at CalMAN. This system includes a correction file for the Spyder3 probe that is included. Each probe is individually analysed before shipping to ensure accuracy. This is a pretty nice way to do things I think.

It looks astonishingly cheap, too.

$280, I think. Can this be right?

Followingthe 'Calibration for Dummies' (seemed appropriate for me ) link it seemed to say that you plug in the sensor, place it next to the screen, then it tells you what to do (turn red gain up, or whatever), and it's as simple as that. Alternately, you bring up the bullseye and move one of your sliders (again, say 'red gain') and if it moves closer to the centre keep going, and it it goes further away, move it the opposite way.

Is it really that simple?

Steve W

[Manni01]

Originally Posted by KelvinS1965

I'm sure Manni01 posted a link or maybe it was a PM that I've since deleted for a UK supplier, but I can't find it ATM. I'll post back if I do.

Hi All,

It wasn't a link for the Spyder + Calman package but for the i1 Basic (i1pro with a nice case and a few accesories but without unnecessary stuff esp software).

The Spyder3 + Calman can either be bought from Spectracal direct (developpers of Calman) or CurtPalme (sometimes better price).

The reseller for the i1Pro (and before that my d2) is Native Digital.

I don't think I'm allowed to post links, but google is your friend. I think the offer is until the end of July until the price goes UP.

Briefly, I can just confirm that from my experience, the d2 is cheap but pretty inaccurate, for all the reasons mentionned by Kelvin. The i1Pro is much more accurate for gamut work, but is less accurate for low light greyscale (20 IRE and below).

So I ended up keeping both, I use the i1pro to work on the gamut, or the greyscale limited to color temp adjustments (using 30IRE for offsets and 70IRE for gains) and I train the d2 to the i1pro to do some finetuning on the greyscale/gamma. I find this combination great, but I wouldn't be happy with one of these on its own.

I have no idea about the Spyder/Calman combination. I personally had stability problems with Calman so I reverted to HCFR (Spectracal is a great company to deal with by the way, no complaints whatsoever). I also heard that the Chroma 5 was a good probe, but I don't have any first hand experience.

If you do buy in the US from Curtpalme/Spectracal, beware of the added costs for shipping, VAT and taxes.

[Manni01]

Originally Posted by Pecker

It looks astonishingly cheap, too.

$280, I think. Can this be right?

Followingthe 'Calibration for Dummies' (seemed appropriate for me ) link it seemed to say that you plug in the sensor, place it next to the screen, then it tells you what to do (turn red gain up, or whatever), and it's as simple as that. Alternately, you bring up the bullseye and move one of your sliders (again, say 'red gain') and if it moves closer to the centre keep going, and it it goes further away, move it the opposite way.

Is it really that simple?

Steve W

No, it's not that simple unfortunately.

You can get many things wrong with the meter placement, the parameters, etc, although it's true Calman does a great job to guide the novice, and its tutorial is fantastic (you don't need to buy the software to test it, you can download a demo from the Spectracal website to get a taste of the software and access the tutorial). The Calman team is a great source of help though, their support is great and they answer questions quickly and accurately. This is a big plus compared to HCFR (no official support, so you have to rely on fellow forum members).

The bull's eye thing is good to set x,y, but you have to look to a separate graph to adjust luminance (Y). HCFR gives exactly the same tool (target window).

If Calman works on your computer, it's great (and the Spyder 3 package is a great deal). I personally tried on 4 different laptops/desktops, fresh installs of Vista/XP, and had crashes on low light readings (with my d2 at the time) on all computers except the desktop (which I can't use for calibration). I'm sure it's bad luck, but there is no way to know until you've bought it (the simulated probe in the demo never crashes).

I personally think Calman is great for the pros, especially because you can use the pattern generator to do complete runs automatically, it can send IR commands to your DVD for patterns, etc, and it can drive professional gear and generate professional looking reports. But for the non pro, I find HCFR just as complete and accurate (with the limitation of luminance correction, which still requires to use an external tool like Greg Rogers calculator or a separate spreadsheet, unless you know by heart the luminance level of each color for each standard).

[KelvinS1965]

Thanks to Manni and Neil for their replies, I didn't mean my previous post to look like I was having a go at ISF or Neil in particular (probably not as well written as I could have done). I'm sure a large part of the calibration is down to the experience and knowledge of the calibrator, but I did feel that using the I1-LT (display2) wasn't doing me any favours and wasn't aware of the Spyder 3.

At least it looks like Pecker has had some useful answers, which will help when he starts his new CIH cinema room (eventually ). It's got me rethinking about whether to ISF or get an I1-Pro myself. Thanks to Neil re the recalibration pricing for the I1-Pro (I'd kind of presumed that it was more a 'throw away' device that couldn't be recalibrated once it had aged).

[Pecker]

Originally Posted by KelvinS1965

At least it looks like Pecker has had some useful answers, which will help when he starts his new CIH cinema room (eventually ).

OY! DON'T START!



I just want to echo those comments...well, the first bit. I've decided to go for an ISF calibration in the short term, and maybe go DIY later on.

I'll let you know how I get on.

Steve W

[Gordon @ Convergent AV]

Originally Posted by KelvinS1965

I'd wait a more detailed and accurate answer
When calibrating greyscale and colour gamut there are the various standards that can be measured to such as (but not limited to) achieving a greyscale of 6500 Kelvin or how close to rec709 the colour gamut measures. How closely these measurements meet the specifications can be expressed in terms of Delta E. It is my understanding that for greyscale a Delta E of less than 3* (*might be 2) is the point at which our naked eye can't tell any difference. There is a similar limit for colour, so this provides a 'target' for the calibrator to aim for.

Only thing to add is that there are also different formula's available for calculating delta errors. 1 unit of delta error is generally regarded as the smallest perceptible difference humans can see between two shades of colour when they are both stationary and within the same field of view. If they are moving around beside each other then the error has to go above three units to be visible. So ONE is the holy grail and THREE or under is a very good result anyway.....

When you measure colour gamut you are usually measuring a couple of levels, ie 75% and 100%saturation....but the final colour we see is an addition of colour and luminance so I try to get colour gamut errors down to under three if possible and then try to get greyscale errors under 1. You want to use a forumula for delta error measurement that takes account of luminance imho....it is after all perfetly possible to get x,y of a primary spot on but have the luminance massively out...with one delta error measurement you could have around 1 unit out and with another a magnitude greater....

[gorman]

Is the Chroma 5 buyable online anywhere in Europe (EC, I'm in Italy)?

[KelvinS1965]

Just to revive this thread rather than start another one: I'm considering getting an ISF calibration done now that we're into the dark night PJ viewing season. I've had a quick check of my greyscale now using my Eyeone LT & HFCR, I've got to 180 hours (I did a very quick adjustment a bit early at 30 hours), but again I'm back to worries of inconsistant measurements. I also have an issue with gamma being too high at low IRE readings and then too low higher up. I may have another attempt at adjusting it (the gamma is different depending on my source; DVD or separate BD player), but I think a pro might be the way forward.

In light of the above I have a couple of questions directed to the ISF proffesionals on here:

I'd want to use my HDQ as much as possible to calibrate my HD350 using my BD player, with only minor RGB, contrast, brightness adjusted in the HD350. In other words the HDQ could be used to 'match' inputs with the HD350 staying on the same setting regardless of source. Would there be any issues doing it this way as it would at least give me a baseline to adjust my other less important sources using spare inputs on the HDQ?

Given that the HDQ doesn't have a full CMS, would they be able to dial in the colour gamut to some degree using the more basic controls available? I have done this myself already, but of course I'm not confident that my sensor is accurate enough anyway.

Would they come at night/evening to carry this out as during the day time some light still gets into the room...not so difficult now as the nights draw in, but I wondered if this is typical? Measuring the low IREs, gamma adjustment, etc would be even more difficult it there is any ambient light (I always watch after dark so it's never an issue for me).

Any recommended ISF professionals in the Berkshire area, or is there a list somewhere I can look through?

[KelvinS1965]

Probably not expecting much response here, given my last post above doesn't seem to have provided any Berkshire based ISFrs, but I have a question regarding the eyeone LT: Given my varied results measuring greyscale, how useful would it be to measure gamma? My results from last weekend showed too high gamma at low IREs and too low gamma above approx 40 IRE. Is it worth me attempting to adjust the gamma using my I1 or is it likely to be out/inconsistent with this too?

[Gordon @ Convergent AV]

Any of the ISF guys who frequent this forum are likely to cover Berkshire, inclusing myself.

I have never used an i1 so cannot comment on it's accuracy.

I carry a roll of bin bags and masking tape to block out all ambient light for calibrations during the day. The only time this hasn't been possible was the chap who had his cinema in a conservatory...I had to do that in the winter, on the evening....

[KelvinS1965]

I suppose that makes sense. When I had my original home demo of the HD350 & HD750 my AV dealer brought some black out lining that we stuck to the windows and patio door as the curtains still allow some light peak through the edges. I thought you were based in Scotland, so didn't think you'd be up for a visit this far South. Unless you are down here already doing a few jobs.

Going to have another try tonight setting up levels and then rechecking gamma from both my sources, I'll see how I get on, then I'll consider the professional option. I presume it's the done thing to PM for prices as I don't recall seeing anything other than a very ballpark figure discussed on the forums.

[Gordon @ Convergent AV]

I live in Surrey.....

[KelvinS1965]

Originally Posted by Gordon @ Convergent AV

I live in Surrey.....

Dunno where I got Scotland from then. At least I know where 'infinity and beyond is' now. I may well be in touch then.

[Gordon @ Convergent AV]

I'm from Glasgow...moved south again about 4yrs ago.