TVI basics

[Dario]

Hi all!
I have a customer who gives me press proofs where there are, for each ink, 3 patches: 25-40-80%.
Obviously, if I print the same patches I have to get the exact measurements that I read on his prints (it's the only way).
... now ... I'm confused ...
1) I print the first test without any TVI curve,
2) then I measure the 3 patches,
...and then what?
I no longer remember how I should use the data from the different readings to create a TVI curve in my software!
For example:
If I measure the 40% patch and read the same value on his and my printout, then I know the curve is already okay.
But if I read 35% on his patch and 45% on mine, how should I use that data in my program?
Excuse me, but I'm really lost, and I could use some help because I only get to do this every 2 or 3 years!

 

[gordo]

Explained here: The principle of dot gain compensation plate curves

 

[Dario]

Thanks Gordon, but my problem is that I can't print the whole range (10-90%) of patches and then search for the one whose measurement is the same as my customer's patch (as I would normally do).
I can only print the same 3 patches.
So here I am getting lost...
If his 40% patch measures 53% and mine measures 61% ... I don't know how to use this information to create the curve, so that when I next print out my 40% patch also measures 53%.

 

[SteveSuffRIT]

53 - 61 = -8.
At the 40% adjust your curve to be -8%.

 

[Dario]

53 - 61 = -8.
At the 40% adjust your curve to be -8%.

I do not think that an 8% measured on paper has the same value as an 8% on file - but I am certainly wrong.

 

[gordo]

With one data point:

It's basically a guessing game.
The "At the 40% adjust your curve to be -8%" Steve offered is unlikely to give you what you want but might make you feel that at least you're trying something.

You could also try with generic curves - his in RED and yours in BLUE (use your actual curve if you have it) (green line shows method as per the link to my blog).:

So, you would set your 40% to 31%. You can use this method if you have the data from your client for the 25% and 8% as well. A more accurate guess (because of the unknown actual curve) than Steve's but still a guess.

 

[chriscozi]

With one data point:

LOL - Perfect!
Thanks for the chuckle Gordo.

 

[Dario]

thanks to everyone

 

[xmoles]

You could try pressyng and its predefined curves

 

[Dario]

btw...
this is the answer I found, for my (and anyone's) future reference:

 

[danremaley]

See if this helps. . .