Adobe Photoshop CS5 Official Announcement

[KevinStecyk]

Originally Posted by KevinStecyk

I am curious if you can replicate the following steps and results, and if you too think this reported behavior is a bug. You should get different behaviors using CS3 vs. CS4.

Using CS4:

1) Open up an image.

2) Insert a point sample on the image (say 3x3) point sample

3) Add an adjustment curve and move the RGB curve so that the image lightens and darkens. You should see the first set of numbers in the info palette showing you the original and the second set showing you the revised values. So far, so good.

4) With your image adjusted (make it a big adjustment), open up your "Blend If" sliders on your curves adjustment layer. When you do this, do you notice that your first and second set of numbers on the info palette have reversed positions?

5) Now move the Blend If sliders so that your image is affected. Do you notice that the values in the info palette remain frozen?

I *believe* if you use CS3, then a) the numbers don't reverse as shown in Step 4; and b) the values do change when you move your blend if sliders around.

I am curious what others experience.

Originally Posted by Jerry Skrocki

Kevin, I get lost in that procedure. I don't know what you mean by insert a 3x3 point sample.

Jerry, I meant to write, "Using the color sampler tool, insert a color sample point, using a 3x3 pixel point sample."

In Step 3, add a grossly exaggerated "Crontrast S Curve." That is, one that lightens the lights and darkens the darks.

In Step 4 and using CS 4, the initial set of color sample values move to the new values location, and the new values move to the initial values location.

Step 5, I am wanting to know if when you move/adjust the "Blend If" sliders, do the RGB values from the point sample change. They should, but on CS4, they are frozen in place.

Please let me know if this explanation is understandable.

[Jerry Skrocki]

Kevin,

Followed the procedure in CS5 and the program behaved as you expected.
The first and second sets of numbers switch positions and moving the sliders has no effect on the numbers.

The tools is working as the blend is occurring, but the numbers do not correspond to the change.

[Noel_Carboni]

Hey Jerry,

Any indication of any incompatibility with pre-CS5 3rd party plugins?

I know you're on Mac and I really want to know about PC compatibility, but knowing the similarities under the covers if you're having no problems I'm pretty sure PC users will have no problems as well.

I'm pre-ordered. Estimated availability for download is 2 weeks-ish.

-Noel

[KevinStecyk]

Originally Posted by Jerry Skrocki

Followed the procedure in CS5 and the program behaved as you expected.
The first and second sets of numbers switch positions and moving the sliders has no effect on the numbers.

The tools is working as the blend is occurring, but the numbers do not correspond to the change.

Thank you for running the test--I appreciate your efforts.

As a beta tester, are you able to submit bugs?

[Jerry Skrocki]

Originally Posted by Noel_Carboni

Hey Jerry,

Any indication of any incompatibility with pre-CS5 3rd party plugins?

I know you're on Mac and I really want to know about PC compatibility, but knowing the similarities under the covers if you're having no problems I'm pretty sure PC users will have no problems as well.

I'm pre-ordered. Estimated availability for download is 2 weeks-ish.

-Noel

Mac users can expect incompatibilities with 3rd party plug-ins because of the platform switch (32-64 bit). With the Mac version you do get the option to start the program in 32bit mode (CMD+I) brings up the info box allowing you to check 32bit. I have not testing any 32bit plug-ins on the Mac version.

I am still bound by the non-disclosure agreement not to divulge information or reports of other beta testers from the Adobe Prerelease Program Forum.

[Jerry Skrocki]

Originally Posted by KevinStecyk

Thank you for running the test--I appreciate your efforts.
As a beta tester, are you able to submit bugs?

Yes Kevin, I have submitted several bug reports. It is somewhat late in the game now to submit a bug report and quite frankly I don't know if your test would be considered a bug.

You have a background layer on which a color sample point is placed and then you have a second curve adjustment layer. Selecting the second layer you further adjust blending by going to fx - blend if.

We would not expect the refinement of the second layer to affect the established color sample point of the first layer.

[KevinStecyk]

Originally Posted by Jerry Skrocki

Yes Kevin, I have submitted several bug reports. It is somewhat late in the game now to submit a bug report and quite frankly I don't know if your test would be considered a bug.

You have a background layer on which a color sample point is placed and then you have a second curve adjustment layer. Selecting the second layer you further adjust blending by going to fx - blend if.

We would not expect the refinement of the second layer to affect the established color sample point of the first layer.

Many consider it a bug. It worked differently in CS3. Moreover, switching positions of the RGB values is at a minimum *weird*. That alone qualifies as a bug.

Secondly, if when you move the Blend If sliders and the values don't update--as they did in CS3--then it is much more difficult to nail down the values. So yes, I and many others do expect the refinement of the second layer to affect the established color sample point of the first layer. Again, bug.

If not, then why are there two sets of RGB values?