Mmm, icons
Jun. 4th, 2004 10:32 am![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png)
Last night, I uploaded my fiftieth icon <points at icon>. Well, it's actually my more-than-fiftieth, because I've had others that I've abandoned, but it fills my last icon slot, which is a sad thing in a way, although there are a few of my existing icons that I don't much like, and will be happy to ditch when the time comes.
But this rambling sort of has a point. I made this icon on my laptop (flat screen, Mac), on which the colours and brightness look fine. Now, on my work puter (CRT, PC), it's far too dark, so I need to know how it looks to you guys...
[Poll #303301]
Hmmm. Some people use polls all the time, and seem to consider them the killer app for paid accounts. This is probably only my second or third poll ever - icons are the killer app for me, all the way ;-)
But this rambling sort of has a point. I made this icon on my laptop (flat screen, Mac), on which the colours and brightness look fine. Now, on my work puter (CRT, PC), it's far too dark, so I need to know how it looks to you guys...
[Poll #303301]
Hmmm. Some people use polls all the time, and seem to consider them the killer app for paid accounts. This is probably only my second or third poll ever - icons are the killer app for me, all the way ;-)
no subject
Date: 2004-06-04 10:31 am (UTC)There are also different standards at work here - Apple QuickDraw applies its own built-in gamma correction (of about 1.45 according to some bloke on the interweb). Evaluating on a Mac what will look OK on a PC is a mug's game, because even monitors/screens with comparable gammas will come out wildly different. I don't know whether Quartz does the same trick.
Unless your image manipulation software knows the total gamma of your screen, and it is correctly saving this desired gamma in the image, and the PC displaying it knows the total gamma of its monitor, and correctly applies a correction to get from the gamma recorded in the image to the gamma of the system displaying it, then there's no reason the brightnesses would match.
no subject
Date: 2004-06-04 10:36 am (UTC)no subject
Date: 2004-06-04 10:47 am (UTC)1.45 is actually quite a lot of difference. CRTs typically range from 2.2 to 2.5, and it's all multiplicative, so 1.45 represents about triple the difference you'd normally see between different monitors. Since your image is composed entirely of midtones, it's exactly the kind of thing that is most sensitive to errors of gamma correction, because shifting midtones destroys the contrast.
Try creating an image with several very pale grey shades - if it then looks *more* contrasty on your PC, then it's the gamma rather than the contrast making the difference.
Anyhoo, this is just a suggestion - if you can find an "assumed gamma" setting anywhere in your OS or image software, you might be able to deal with the problem without having to mess about changing the brightness of your laptop (and thereby makeing everything other than this icon too bright).
no subject
Date: 2004-06-04 10:53 am (UTC)Gamma
To my way of thinking, there is a unique correct way for a monitor to display a picture with a certain set of RGB parameters (for a fixed position of the monitor's twiddly knobs). Gamma corrections on a given platform are there to fix this, not to get creative with the settings.
The whole point of all this being that you should be able to save a picture on one system and have it look pretty much exactly the same on another. (And when this fails, at least one system has its gamma set wrong.)
Ian always claimed otherwise, but was never able to satisfactorily explain to me why not.
Re: Gamma
Date: 2004-06-06 06:51 pm (UTC)So yes, it basically sucks in that printers have a gamma of 1, so a monitor with anything other than that makes things tricky. And woe betide if your monitor has different gamma values for different phosphor colours. But since most graphics setups can't apply a variable correction, we're basically stuck with it.