| Author | Message | ||
     Marcin New member Username: Marcin Post Number: 12 Registered: 09-2002 |
Useful local contrast enhancement technique (http://www.luminous-landscape.com/tutorials/contrast-enhancement.shtml) requires quite high Radius to be applied. PWP's USM Range limits it to 40. Do you have any plans to increase available range? | ||
| Jonathan Sachs New member Username: Jsachs Post Number: 203 Registered: 08-2002 |
I take issue with the article in question and do not recommend this technique. Jonathan Sachs Digital Light & Color | ||
| Elie Dinur New member Username: Elied Post Number: 35 Registered: 07-2003 |
Jonathan - Can you elaborate on why you think "local enhancement" is not good? I often use USM/20%/R40 at the beginning of the editing flow and Advanced Sharpening/200%/R0.3-0.4 at the end and the combo seems to work quite well. Elie | ||
| Jonathan Sachs New member Username: Jsachs Post Number: 207 Registered: 08-2002 |
It is not a problem with blurring (e.g. Noise Reduction phase of AS), but more with sharpening. If you use a radius that is too large, you can get broad halos around objects. Jonathan Sachs Digital Light & Color | ||
| David Griffel New member Username: Madhg Post Number: 20 Registered: 02-2003 |
This topic was discussed a year or 2 ago, and I posted some pictures on which Jonathan pointed out subtle halos which I hadn't noticed. I now use less "amount" as a result of that discussion. Jonathan, OK, you don't like to use too large a radius. You have very good judgement, and that's why PWP is so excellent. But in this case ... is your view on radius >40 strong enough for you to deny PWP users the facility to use it? | ||
| Jonathan Sachs New member Username: Jsachs Post Number: 211 Registered: 08-2002 |
I suppose it is - I'm really just trying to keep people from shooting themselves in the foot. The subtle halos can, after tonal modification become more obvious. Also they are easy to overlook and then when you make your final print after doing a whole sequence of operations it may not be at all obvious where they came from. Jonathan Sachs Digital Light & Color | ||
| Anonymous |
Thanks for protecting us from subtle halos, Jonathan. What about other, more damaging things we can do to our images, like spiral, kaleidoscope, emboss, high contrast, and the like?  | ||
| Maciej Tomczak New member Username: Tomczak Post Number: 27 Registered: 08-2004 |
Here is a test image that shows halos that Jonathan dislikes produced by USM r=40/25%. From top to bottom: 1) original; 2) USM r=40/25% applied; 3) the absolute difference between the two with black and white points stretched to 0-100%. The middle checker boards which received the most contrast boost are made of 40/60% gray, the outer checker boards are black and white. The halos and various artifacts depend quite a bit on the contrast, geometry and size of adjacent shapes. Note that the lower gray gradient band has been affected by USM much more than the identical (reversed) upper gray gradient. This is mostly because of its contrast with the black and white blocks below it, whose thickneses are comparable with the USM blurring radius. The banding of halos is probably the result of using 8bit file.  | ||
| den Unregistered guest |
Maciej ..... a very effective illustration. There may be some justification for the ability to have a larger USM radius value than 40 because as the radius increases, the halos become more diffuse......  For illustration purposes only, the Amt=100 was used. For my own images that do not have significant clipped shadows and/or highlights [LCE will increase the number of clipped pixels], I use for LCE = USM [R=40, Amt=20 or less] followed with general over image sharpening of USM [R=2, Amt=25]. This is on camera downloaded images of 2272x1704 pixels, superfine jpeg compression. | ||
| den Unregistered guest |
I should add that I am not advocating a change in PWP's USM for being able to select a larger radius for LCE.... if the radius selection was increased, then the amount would have to be adjusted accordingly for the same current effect. Also for the most part, my LCE = USM [R=20, Amt=20 or less] is just my starting point. More often than not.... the final amount is around 15. This effectively reduces the impact of the halos at R=40, to provide a very subtle contrast improvement, especially when followed with the suggested USM [R=2, Amt=25]. | ||
| den Unregistered guest |
the mind is quicker than the hand..... correct the previous post to read..... "....Also for the most part, my LCE = USM [R=40, Amt=20 or less]......" | ||
| Maciej Tomczak New member Username: Tomczak Post Number: 28 Registered: 08-2004 |
I wouldn't be surprised if Jonathan said that decreasing the halos by increasing the USM radius or reducing its amount is at the limit equivalent to increasing the midtone contrast globaly, which is what he originally suggested... |
