Pages: [1]   Go Down

Author Topic: Jeff? RSP Processing Bias Parameters & LR Match-up  (Read 3401 times)

JayS

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 64
Jeff? RSP Processing Bias Parameters & LR Match-up
« on: September 13, 2007, 06:01:46 pm »

Jeff (and all),
     Is there any co-relation one can take yet (since RSP was picked up by Adobe) with the Processing Bias Parameters and LightRoom.  As an example, a set of parameters were developed for the Canon 20D that based on ISO would set the following for ISO 1600 (all ISOs were available)..

Sharpening Bias = 5
Detail Extraction vs. Noise Reduction = -20
Noise Suppression Bias = 33
Color Noise Suppression Bias = 27
Hot Pixel / Pattern Noise Bias = 0
 
   These were just plain wicked good in RSP for starting you at with a relatively noise free starting point.  Within RSP, as you may recall, you could still adjust the sliders on top of these, but very rarely did I have to.  I am trying to match up the above and their counterparts in LR, since with 1.1 we can ascribe certain things to happen based on ISO.  What I can't resolve is something like "Detail Extraction vs. Noise Suppression" being a -20.  First, there isn't a matching category in LR, and none of the sliders there go negative..  I truly would like to get this in some fashion into LR and would appreciate any help and thoughts.

Jay S.
Logged

madmanchan

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2115
    • Web
Jeff? RSP Processing Bias Parameters & LR Match-up
« Reply #1 on: September 14, 2007, 08:14:32 am »

Check out this article here:

http://www.photoshopnews.com/2007/05/31/about-camera-raw-41/

It explains how to use the new sharpening controls in ACR 4.1 (and now ACR 4.2).

Ignore the issue with some programs having "negative" values for sliders and other programs not. Numeric values are arbitrary -- the developers could have renumbered the range from [-50,50] to [879,3201] and the effect could be the same.

Since the algorithms ACR 4.1/4.2 use are most likely different than the algorithms used by RSP, you may not be able to get the exact same result. But you ought to be able to get reasonably close by fiddling with the Detail tab sliders. The article above ought to give you enough of an understanding of how the controls work, so as to help get you in the ballpark for the "look" that you want to achieve.

Once you arrive at settings you like, save them as a preset so you can quickly apply them to all future images.
Logged
Eric Chan

JayS

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 64
Jeff? RSP Processing Bias Parameters & LR Match-up
« Reply #2 on: September 14, 2007, 12:15:16 pm »

Quote
Check out this article here:

http://www.photoshopnews.com/2007/05/31/about-camera-raw-41/

It explains how to use the new sharpening controls in ACR 4.1 (and now ACR 4.2).

Ignore the issue with some programs having "negative" values for sliders and other programs not. Numeric values are arbitrary -- the developers could have renumbered the range from [-50,50] to [879,3201] and the effect could be the same.

Since the algorithms ACR 4.1/4.2 use are most likely different than the algorithms used by RSP, you may not be able to get the exact same result. But you ought to be able to get reasonably close by fiddling with the Detail tab sliders. The article above ought to give you enough of an understanding of how the controls work, so as to help get you in the ballpark for the "look" that you want to achieve.

Once you arrive at settings you like, save them as a preset so you can quickly apply them to all future images.
[a href=\"index.php?act=findpost&pid=139367\"][{POST_SNAPBACK}][/a]

Eric,
    Thanks.  Some of that content is in the LR Video Tutorial from Jeff and Michael (recommend if you haven't gotten it as you can't beat the price), but a good read none the less.  I'll go through the article.  I'm guessing the algorithms are different, so don't know that matching the numbers (even in the categories that seem similar) would net the same results, as you indicdate.  I don't remember who it was that went through the effort to produce those excellent Pre-Processing parameters.  Also looks like some update in LR 1.2 around noise handling based on one of the posts here.  Will find out since it appears 1.2 was released now.

Jay S.
Logged

madmanchan

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2115
    • Web
Jeff? RSP Processing Bias Parameters & LR Match-up
« Reply #3 on: September 14, 2007, 02:26:21 pm »

I did buy the video (despite having a lot of hands-on experience in the image processing and printing world) and thoroughly enjoyed it. Sharpening was indeed covered, but I think you'll find the article more directly helpful here because of the examples provided.  (Obviously they didn't have time to go into lots of detail in the tutorial.)

In general, the message here is that different RAW converters all have their own sets of controls. This can be challenging if/when you migrate from one to the other, because there isn't necessarily a direct correspondence in quality settings between two different converters. The key is to learn the controls well enough so that you can achieve the desired results, then save time in the future by creating presets / saved configurations of your favorite settings.
Logged
Eric Chan

JayS

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 64
Jeff? RSP Processing Bias Parameters & LR Match-up
« Reply #4 on: September 15, 2007, 01:07:20 am »

Quote
I did buy the video (despite having a lot of hands-on experience in the image processing and printing world) and thoroughly enjoyed it. Sharpening was indeed covered, but I think you'll find the article more directly helpful here because of the examples provided.  (Obviously they didn't have time to go into lots of detail in the tutorial.)

In general, the message here is that different RAW converters all have their own sets of controls. This can be challenging if/when you migrate from one to the other, because there isn't necessarily a direct correspondence in quality settings between two different converters. The key is to learn the controls well enough so that you can achieve the desired results, then save time in the future by creating presets / saved configurations of your favorite settings.
[a href=\"index.php?act=findpost&pid=139455\"][{POST_SNAPBACK}][/a]


Eric,
    Agreed on all your points.  The only thing I was hoping for here was since Adobe had bought Raw Shooter Pro, that some of the way it handled noise would be incorporated close enough to make the comparison.  I had to go back on a job I was doing with ISO 1600 RAW files.  The Luminence setting jumped a fair chunk after applying the 1.2 update.

Jay S.
Logged

madmanchan

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2115
    • Web
Jeff? RSP Processing Bias Parameters & LR Match-up
« Reply #5 on: September 15, 2007, 08:15:48 am »

Ah I see what you're getting at. Yeah, the numbers are essentially meaningless when discussing/comparing different converters. One converter's "zero" is another converter's "25" ...  
Logged
Eric Chan

JayS

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 64
Jeff? RSP Processing Bias Parameters & LR Match-up
« Reply #6 on: September 15, 2007, 02:50:58 pm »

Quote
Ah I see what you're getting at. Yeah, the numbers are essentially meaningless when discussing/comparing different converters. One converter's "zero" is another converter's "25" ... 
[a href=\"index.php?act=findpost&pid=139557\"][{POST_SNAPBACK}][/a]

Eric,
    It would appear so.  I tried the algorithm that maybe the detail vs. noise = -20 meant, whatever you come up with for noise, subtract 20 from detail.    Which was obviously just an absolute guess.  I'm curious what you've seen with any change in setting with the 1.2 update.  I was using a luminance setting of about 35 for a Canon 20D ISO 1600 as a starting point (other considerations as necessary of course).  After the 1.2 update, I am now up closer to 70 for the same effect.  That is a substantial jump.  That said, perhaps the update in 1.2 made the new starting point less aggressive to begin with (they said you would see more noise) and that the higher number is now more in line with the higher ISO ratings...


Jay S.
Logged
Pages: [1]   Go Up