Interfaces: PC and MAC and the screenGamma property...

[Chipp Walters]

Dar,

You are correct, the original image in the demo stack was set at 
128,128,128 but the one I sent you was different. I left the original 
one at work. Let me know if you'd like another 128,128,128 PNG from a 
PC and I'll fire you one.

--Chipp

On Jan 3, 2004, at 1:15 PM, Dar Scott wrote:

>
> On Saturday, January 3, 2004, at 12:38 AM, Chipp Walters wrote:
>
>> Dar, perhaps you can tell me what the internal gamma setting is for a 
>> Pshop
>> PC generated PNG vs a PShop Mac generated PNG...my guess it they are 
>> the
>> same, as I couldn't get a PNG and trans GIF to ever work correctly on 
>> my Mac
>> unless setting the gamma to the PC value.
>
> Your guess turns out better than mine.
>
> The one you sent me has a gamma of 2.222222.
>
> The one I created with my Photoshop on OS X has a gamma of 2.200026, 
> and it also has a primary chromaticities chunk that might override 
> some of that.  Maybe I don't have my photoshop set up right.
>
> Setting screenGamma to 2.2 still might not be a solution.  I saved my 
> test stack last night with your test PNG image on it, the one with PNG 
> written in the corner.  When I loaded it this morning I found the 
> gamma in both the text property and the export showed 1.7.  Based on 
> what I saw yesterday and what I see with the test box you sent, this 
> _was_ 2.2.  Maybe, setting screenGamma to 2.2 is still OK because, I 
> suspect the pixel data is still for 2.2.
>
> I set my test image to 80,80,80 (hex) grey in Photoshop.  I don't know 
> what values were actually stored in the PNG file.  When I imported it 
> into Revolution the pixels became 69,69,69 hex; well, they were 69 as 
> soon as I could read them.  Why I saw 86,86,86 (hex) on your png (the 
> other one with the png label) beats me.  I thought it was created as 
> 50% grey, also.  Oh, maybe those were 68s.  Rats.  Sorry.
>
> I made a rectangle with a 50% background and it looks a lot lighter.
>
> put (baseConvert("69",16,10)^(1/2.2))/(255^(1/2.2)) && 
> (128^(1/1.7))/(255^(1/1.7))
> ==>  0.668099 0.666689
>
> I _think_ this means that x69 on 2.2 will look the same as x80 on 1.7.
>
> Maybe Rev is not converting the 69s to 80s as it should.  Well, I 
> don't know enough to say should.
>
> One experiment might be to fiddle with gamma on images and see if Rev 
> respects those.
>
> The more I think on this, the less I know.
>
> Dar Scott
>
> _______________________________________________
> use-revolution mailing list
> use-revolution at lists.runrev.com
> http://lists.runrev.com/mailman/listinfo/use-revolution
>