That’s good to hear Russell: good to hear that Adobe is on the case and good to know that you’re not going crazy. 😉
Compliments on your perseverance. 🙂
good to know that you’re not going crazy. 😉
Crazy-er.. 😉
great job russell (and adam!) 🙂
I have to say I tried it again at , 16 bit RGB and began placing small images on the canvas. The only hesitation I had was trying to move before the earlier move was written, ie the hour glass was running while I tried the next move. I had a total of 5 images, and never did the scratch even reach 3G, even after running all the images around.
Strange, very strange.
Athlon 3200+, 64 bit single core
2G ram,
PCI video card (my PCI-E slot has a problem),
Asus M2V board
I really want to hear what is going on. How can I replicate his problem on my system?
Note: Russell, I am assuming you are running a PCI-E video card. Can you find a standard PCI video card and try it again?
I’ve also done it on a P4 (3G processor – Dual Core – not Core2 or Quad) using a AGP videocard (ATI Radeon model?) and the problem was there too. I was never working in 16bit, only 8 bit.
Right now I’m assuming that it’s an Intel issue as other AMD users such as yourself report that everything is working fine. I’m not going to waste any more time with this now that Adobe has acknowledged that there is an issue that is being worked on.
Thanks, Russell. Certainly, 16 bit should exacerbate the problem.
A big difference between Intel and AMD is FSB. That would be consistent between Core 2 and the P4. If that is the problem, there may not be much of a software cure, but then, those guys can work miracles at times!
I did FSB validation work for Intel. I’m going to touch bases with one of the engineers and see if he will set up a test.
A big difference between Intel and AMD is FSB. That would be consistent between Core 2 and the P4. If that is the problem, there may not be much of a software cure, but then, those guys can work miracles at times!
Might be an FSB issue. But I still wonder why CS2 performance is ‘in general’ better than I’ve experienced with CS3 on the same hardware. They obviously changed their memory management and it’s not working for me.
I first noticed a problem with CS3 on my AMD Athlon 64 X2 3800+ with 2GB RAM and then discovered the same behavior on my new Intel Core2 Quad Q6600 with 4GB RAM. CS2 was fine on my original AMD. My observation is that it has the same problem on both platforms.
Then it may come down to multi-core since I have a single core AMD Athlon 64.
If it is FSB, then it becomes a problem of bus congestion. Why this might be so is perplexing as the bus speeds continue to increase, so that’s why I considered discussing it with one of the engineers.
I suppose I could pungle up the ~$80 for a dual core and see what happens.
Ok, my bad. I forgot Russell was using "Place". I went back and did it again using "Place", and it did slow down as well as building up the scratch to 12G, but it was a matter of 8 to 15 seconds using the PS timer, to move things. If I grabbed a file before the previous had been re-written, it took longer and screwed up the timing.
But no 20 to 40 seconds.
Using "Open" to find and drop in an image to the canvas was far faster and much less scratch. But an interesting difference is that the visual image size of each image on the canvas was different using "Place" than using "Open", even though the files I chose for "Place" and "Open" were the same files, and the canvas size identical.
So, I’ll wait to see what Adobe finds out before proceeding further.
I won’t be using "Place"!
suppose I could pungle up the ~$80 for a dual core and see what happens.
Now why would you want to go spending ~$80 just to look for trouble you don’t have? 😉
At this point the ball’s in Adobe’s court so there’s been ‘nuf time and $$ wasted on this issue.
I guess I just can’t help it!:D
Anyway, that’s what Intel hires me for: troubleshooting, data collection and roll up.
I won’t be spending the money.
As you can see, I did forget the "Place" as the method to acquire the images. To a first approximation, Intel and AMD are about the same, except for recovery.
So, it’s in Adobe’s court.
By for now. 🙂
I forgot Russell was using "Place".
ROTFL!! Well at least now I’m comforted in knowing that AMD systems are not immune 😛
The power of using "Place" is that it imports the images as Smart Objects so you can change their size as much as you wish while doing layouts (I do a lot of collages) and only rasterize them once everything is pretty well sized and in place. This saves a LOT of time and avoids doing a ‘for-position-only’ layout 1st and then go back and replace all the mucked up images that were damaged from the numerous re-sizings which occurred while playing around with the layout.
Russell Brown even has a **very** kewl script inside his "Dr. Brown’s Services 1.9.4" called "Place-A Matic" that allows me to select a number of images in Bridge and automatically ‘place’ them as Smart Object layers in a single document and this saved LOTS of time.
I recently downloaded his "Services". I saw that title but haven’t tried it yet. Nor have I had the need as you do for what Place does, so forgive my ignorance of it’s operation. I was trying to replicate your problem.
I did try Place the first time around but got something wrong. So this time when I retried it, I let it run it’s course, then I recognized what you were doing and how it actually runs.
Learn something every day. Now, I’ll have to figure out how it can enhance my work with Photoshop.
Gad, I still haven’t nailed down perspective cloning!
Anyway, glad it brought you a bit of mirth! 😀
Russell
If you do a lot of collages you might consider InDesign. It makes life easy compared with Photoshop.
I don’t use inDesign though my graphic designer wife does on her G5. But afaik inDesign is a page layout program that competes with Quark.
I do soft edge layer masks on all the layered elements of the collage (usually groups of people at social events) and then add linked curve adjustments to those layers that need colour correction and local colour adjusting on elements within those layers as well as retouching individual faces that need it (lots of shiny straight flash shots). Individual layers are also sharpened if the photographer’s focus was less than ideal in those shots. Once flattened the image (sometimes as large as 8.x x 17 inches @300ppi) is converted to the CMYK spec of the printing press it will be printed on and then the CMYK numbers are rechecked and tweaked as necessary. A final sharpening is then done on the whole thing (more on the black plate) and then the sharpening is removed in areas of faces where it exaggerated wrinkles and skin textures.
I can’t see how I could get this level of control using inDesign which is not really designed to be a image editing/retouching program?
You’re not supposed to do all that in InDesign, you go back and forth between the two, it updates as you go. No problem.
Indeed, that’s the way I’d do it too. In fact, you can do the feathered edges in InDesign keeping the original graphic intact. Just use the edit original command in InDesign to open the graphic in Photoshop.
You’ll have far more control this way.
Bob
Indeed, that’s the way I’d do it too…You’ll have far more control this way.
Whatever way suits you. But if it ain’t broke… (CS2 works fine)
Russell 🙂
