i1Pro strip scanning speed and accuracy

[tony22]

I've been using both ArgyllCMS and i1Match to develop printer profiles. In both cases it seems like the rule of thumb for the strip scanning speed is something like "it should take about 5 seconds to scan an individual strip". I've been trying to keep to this recommendation, but does anyone know at which point the i1Pro begins to lose the ability to calculate an accurate comparison for a given color patch? I assume the timeline is to allow for a given number of samples to be taken off each patch, but how many samples is enough? Is it possible to take too many samples (i.e. take 15 seconds to scan one strip)?

[Scott Martin]

Yes there's such a thing as too slow and too fast. Get to know the "sweet spot" and enjoy. The RevA EyeOnePro's sweet spot was much slower than newer EyeOnePros so take that under consideration. Also, the software you measure with can effect the "too slow" point.

[tony22]

Well, I have a Rev D, which I understand takes more samples per unit time than the Rev A. But I guess what I'm wondering is whether the "single beep confirmation" that arises when a strip is scanned "correctly" really means that it can't be scanned any more accurately. Or is there still room to get better data?

[PhilipCummins]

But I guess what I'm wondering is whether the "single beep confirmation" that arises when a strip is scanned "correctly" really means that it can't be scanned any more accurately. Or is there still room to get better data?

I think there's always room for getting more data - I'd say the single beep just means that the data has been at least scanned once successfully. If you did say multiple strip scans and then averaged the results together of all the samples per patch that would be more accurate than just doing it once.

[Alan Goldhammer]

I've done repeated scans at some various speeds using Argyll to see what the difference might be and it's actually very little.  The important factor is to use a constant rate of scan and I try to aim for about five seconds or so for each row of 21 patches.  I read each patch set twice going in the opposite direction the second time (Argyll allows you to set it for either bidirectional or unidirectional scanning) just in case the direction might make a difference (and I don't think it does) and then average the patch sets.  If you use a robust number of patches and I use 1848 (four letter size pages), and difference will be minimized.  Profiles are good and I've also done Epson ABW profiles using a 51 step gray scale and QTR.