Re: EyeOne min patch size in strip mode
Re: EyeOne min patch size in strip mode
- Subject: Re: EyeOne min patch size in strip mode
- From: Terence Wyse <email@hidden>
- Date: Tue, 01 Dec 2009 21:25:19 -0500
On Dec 1, 2009, at 6:45 PM, Graeme Gill wrote:
> Terence Wyse wrote:
>> increased the value in .5mm increments. I found that it would "agree" to measure the
>> 5x5mm patches but not reliably. I started getting reliable measurements around
>> 5.5x5.5mm. I measured a relatively small chart (letter size or so) so I've no clue
>> whether it will measure this patch size reliably up to around 24".
>
> Note that there is a difference between reliably recognizing the patches,
> and getting accurate readings. The instruments have a finite aperture,
> so this works as a convolution filter as you scan over the patches,
> "fuzzying" up the boundaries and mixing the colors together as
> it crosses each one. The instruments also have a finite sample
> rate (and a signal to noise ratio that gets worse as you sample faster),
> so the end result as you decrease the patch size is that you get
> fewer and fewer samples for each patch, and a greater chance
> of patch boundary samples being included in the reading, or
> becoming a larger proportion of the reading.
I agree Graeme.....my next battery of tests was to measure the same target created with different size patches and then compare the data. Assuming, for example, 9x9mm patches is a large enough sample to be used as a "reference", I would compare smaller patch sizes until I started seeing evidence of significant deviation compare against the large patch reference.
My initial goal was to see if there was a way to create smaller size testcharts for use in offset press profiling where sheet "real estate" is somewhat limited. It's difficult as it is to get good profile data from a press sheet.....having the option using multiple smaller testcharts or a single larger testchart to get better sampling/averaging across a press sheet. If my testing proves that the smaller patch sizes compromise the measuring accuracy/reliability of the instrument, then it's simpler another dumb idea relegated to the dust bin. :-)
One might assume that the manufacturer knows the minimum acceptable limits for their instruments but I'm banking on the fact that they might be perhaps a bit conservative in their specifications. No harm in trying and I'm certainly not going to recommend my findings to others.
Regards,
Terry
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Colorsync-users mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden