TroubleShooting PS9 Display Slowness

Glenn

I don’t know if Adobe have found this card to be incompatible but it is pretty clear from reading a number of reports of slowness that many of the problems are on systems using GeForce4 Ti 4XXX series (or older) cards. You can buy a new graphics card for less than £50. I don’t know what the latest thing in graphics cards is but I’m running a two year old NVIDIA 5600 and am having no problems.

Thanks, Mick…

I think I’ll wait to hear from Adobe, or maybe from someone with the same card, before i go changing what is an otherwise smooth setup.

I just wish i understood more what happens here: what the demands are that PS9 imposes that nothing else does?

Glenn

I have a GF4 4600 (Gainward) and CS2 redraws the image rather slow. Palettes are slow as well. No problems with any other software with this card. CS1 runs perfect.
Maybe Adobe can test this card and tell us more.
Stefan

Re: GeForce4 Ti 4600 card (Asus V8460) with the latest NVIDIA Driver (7.1.8.9) Win 2K SP4

While waiting for someone from Adobe hopefully to get back on the status of this Card/Driver (or someone with the same who perhaps is using it successfully with PS9?) could somebody suggest any further Troubleshooting steps i could try – other than the ones mentioned on my earlier Link?

I have since tried removing everything but the Logo Calibration Loader from all Startup locations; and removed and re-seated the Card – all with no effect at all.

Any other ideas…?

Glenn

Stefan… Are you on Win 2K or XP?

Hi Glenn,
I`m using WinXP Home SP1.
But after reading the other posts in this forum it seems that the problem is not related to nvidia cards. Some of the users have ATIs and the same problem.
Stefan

I’m an ATI user with the same problems (Radeon SDR PCI, Cat5.4 drivers *embarrassed frown*). What I would like to know is what video cards Adobe has in their testing systems. Since they can’t seem to duplicate the problem, that information might be useful for some of us (I’m ordering a new video card in a few weeks anyway) to do more extensive troubleshooting on our own.

Stefan… So, different OS, but same GF 4600 Card Chipset (I don’t know if or how much the Card manufacturer matters). Are you getting each of the symptoms I listed? Have you tried v) the Curves and Info test and does vi) the strip in CameraRAW show up the same?

Josh… Yes, that’s why I thought asking here for further ways to troubleshoot this end would be worth trying…

…. Glenn

Glenn,
I`ll try tomorrow when I`m at my "work-PC".
Stefan

Just in case this may help. I’m running a ati radeon 8500 with non-catylist driver – 6.14.10.6387 from Nov. 2003 and not having the display, refresh, or loading problems on CS or CS2, which is now faster with other tweaks offered here. XP home, sp2, P4 2.2, 1g RAM.

Stefan…
No! You put down the food you’re eating, you turn off the concert on the radio, you get in your car and you drive to work and do these tests this very minute! I’m sure the Adobe folk are working round the clock on this and so must we…!

No, but seriously… I don’t know what your repeating what i did will show us very much anyway (apart from maybe confirming OS doesn’t play a part?). What we need is either Adobe to confirm that this ChipSet does not have a suitable Driver as yet, or (cat amongst the pigeons) to hear from someone with that Card/Chipset who is not experiencing any problems….

I think, unless Adobe can propose further configuration tests, it’s wait-and-see time…

Hello, Kent…

Thanks, but i was hoping we could focus here on just the one card (or series?) to see if we could establish anything common to those with problems, or to those without problems, who are known to be using the same Card…

If that makes sense…?

Understood.

Glenn,
I should have done what you said and gone to work.
My food was terrible and tonight`s TV-program is just boring! Stefan

Josh – we have thousands of cards in house and with beta testers. We cannot possibly list all of them.

But the 4600 and 4400 appear to have a common problem as reported by users.

We’re still trying to location one of these cards in house. If we can’t locate one, we’ll call NVidia and get one.

Glenn,
number V and VI of your test are definitely the same on my system. Number I: Compared to CS1 CS2 is slow in this area.
Number II: ditto (but no transparency as you described)
Number III: Softproof is slow, but due to the slowness of the overall image redraw slowness. Number IV: MY PC is rather slow, so I could see the tile by tile redraw in CS1 as well, just faster.
Chris,
I can remember, that some users in this forum reported NO problems although they had those cards….whatever that means…
Stefan

Stefan…
Well, seems like it’s not OS version specific. Are you sure about others reporting no problem with the 4600? I Searched and came up with just one about the 4400 (#6 in ‘UI Performance…’ thread).

An extra problem here is that NVidia’s site directs all enquiries to the Board manufacturer! I’ve mailed Asus here in UK but not holding my breath…

Chris…
One thing more i thought to try, to test configuration here, was that the Graphics Card shares an IQ number (16) with a USB controller. It didn’t show as a ‘Problem Device’ but I disabled it anyway (it turned out to be where my Wacom lives) but no difference. I don’t know how to re-allocate the IQ’s (in BIOS, is all i know) but if you think it’s worth trying, to set the 4600 to non-sharing – and there’s a non-hazardous procedure – i’ll give it a go….

Glenn

Glenn,
I`m not absolutely sure about the 4600, only the 4400. But as far as I remember, the only difference is GPU speed and memory speed.
Stefan

I am experiencing screen redraw slowness/sluggishness using an ATI Radeon X850 XT 256M AGP card. I also read a post in this forum of a user reporting screen redraw slowdowns with an ATI FireGL card.

— font9a

Just in case anyone else was thinking along the same lines…

To confirm whether or not my System Configuration might be playing any part in these Display plroblems, I temporarily installed WinXP on a fresh drive. And immediately installed PS9 there as the first and only app. No AV, Firewall, Utilities, device drivers, nothing.

Ran PS; still using Windows’ VGA driver, obviously couldn’t assess redraw but the Curves/ArrowKey/Info symptom still showed as before.
Installed latest 7.1.8.9 NVidia Driver. No change. And Redraw symptoms exactly as with original Setup.
Installed the XP SP2 Pack. No change. All usual variables – Reduced Acceleration, bit-depth, Bigger Tile Plugin etc – all tried, as before, with no improvement.

Conclusion: My original Setup (Win2k SP4 with attendant Devices, utilities, fonts etc) had no bearing at all on the problem.

Which (excluding mainboard, RAM, HD etc) does just leave the Card – GeForce4 Ti 4600 card (Asus V8460) – or its latest Driver not handling Photoshop’s new requirements.

Nothing I can do about that. So nothing more i can do to troubleshoot this end…?

Try another video card, nVidia FX 5200, for example. It did not have driver problems in my configuration.

Not to be controntational, but in the other thread, you said that the FX5200 "it improved redrawing a little bit," which seems to me to be different than totally fixing the issue. Also, others are known to have issues with the card ( Alvaro Garcia, "PHOTOSHOP CS2, MORE SLOW THAN EVER !!!" #92, 9 May 2005 6:36 pm </cgi-bin/webx?14/91> ), so maybe it would be good to try something else, if one must spend money to test it.

I have a slow computer, Celeron 1.7 so part of the problem can be a CPU speed. I can’t meassure the difference so, I cann’t tell exactly how much difference it makes. But what I noticed that FX 5200 with latest drivers was working at the same speed as my old ATI Rage Fury Pro with hardware accelearation turned down.
Try something else if you are not sure. I am searching for Matrox G550 to replase my ATI card now. It was also reported to work good with the latest drivers.

Alexander, thanks for the suggestion. I had considered a new card at one point but, as Josh says, it is totally unclear which Card is 100% compatible with PS9.

(Plus, this Card cost me £300+ just a couple of years back so is hardly cheap or ancient; performing perfectly well with everything else, I see no reason why it should not – with the correct interchange between NVidia and Adobe – handle anything PS9 cares to throw at it)…

Your point about knowing how to judge a retarded performance is why I was asking on the other thread for those with no redraw slowness to verify that the Curves/Arrow/Info behaviour is no different in 9 than in 7 or 8. Because I think (but obviously can’t know for sure) that this might be one clear sign of a Graphics Card failing to handle v9 properly.

Glenn

Glenn: "to verify that the Curves/Arrow/Info behaviour is no different in 9 than in 7", it works in v7, but does not work in V9, this is an oversight by the development team (or bug if you prefer).

ID… Thanks very much for clearing that up. You say it with conviction (dare I say authority)? Has that been officially posted somewhere?

I’m a wee bit peeved, cos I’ve spent a fair bit of time and effort troubleshooting stuff here on the basis that that was a display driver fault.

….

FWIW & IMHO there are 2 ways to combat what is perceived as slownes and this may or may not include CS2

1 – overclocking: really to be avoided IMHO
This approach is typified by trying to get the system finely tuned so as to run most effectively under its present configuration. It speeds things up because everything is going that little bit faster.

This probably corresponds to driving the auto under hi-revs at all times.

2 – tweaking
This is usually happening all the time under various OSs as, for example, most used programs, codings & files probably end up in a faster access part of a hard disk with equivalence as least used programs, codings and files ending up in a slower part of the hard disk.

It is probably based on a "most used" basis as recorded by the OS

What may be lacking is an "I prefer" option where users may specify a program that should always reside in fast access mode. The trouble is: unexpected, unforeseen or unknown consequences.

We may forgive OS designers (IMHO) for going for safe & predictable outcomes eg hence this thread.

There are options to consider for local or single program tweaks as detailed in earlier postings elsewhere, such as:
– tweak process priority to High
– use something like Optimize Memory
– don’t apply changes or tweaks to many programs and keep them within limits.

The above are general points. If one is really intrigued and has urgent demands about increasing computer performance of performance of a limited number of super-use programs it is probably better to employ the services of a reliable and qualified and experienced computer engineer of some description (with warranties for work undertaken)

thus spake das deebs

Obviously, if your computer is too slow, it is best to call a consultant instead of doing it yourself.

Obviously, if your digital camera is dusty, send it to Canon/Nikon/Minolta instead of cleaning it yourself.

However, my system has been overclocked slightly (it’s a work system, not for gaming, so nothing extreme), and has just about every safe tweak that’s not urban legend (like the FAT32 being so much better than NTFS type of junk) applied to it, so if that were the only cause, I wouldn’t be seeing it.

There’s certainly something going on in cards or drivers, and I’m just waiting for confirmation on what is known to work before I pull the trigger, er, mouse button, on ordering a replacement video card.

Again and IMHO:

Warranties, support and support & warranty clauses should contain a line that overclocking invalidates the warranty & support and is made for standard configurations (whatever that may be) of the time.

Overclocking of any kind introduces or rather has the propensity to introduce data irregularities leading to all manners of local inconsistensies

Components are rated for a given speed under worst-case conditions (I mean, _really_ bad conditions), so it’s not unreasonable to find that many CPUs have a small (or in some famous cases, a large) amount of headroom for overclocking. The important matter is to test with programs like Prime95 and Memtest86 to ensure stability, rather than using your own important data. With that said, if a system were being used for purely professional purposes, I think just getting a faster processor is much cheaper than spending the time to ensure that a particular amount of overclocking will not introduce instability.

Anyway, it’s not your discouraging of overclocking that I take offense to (I don’t recommend it for professional systems either), but your suggestion that one should call a computer consultant for any and every computer upgrade need, that is totally unnecessary unless one is entirely technologically inept. I know a fellow who is in his 60’s, and is not exactly a world-renown computer expert (I’m trying to be polite ;), but he can replace and upgrade the modems, memory, etc. in his own computers with no trouble. Since I suspect that most of the folks on this forum are slightly above that level of knowledge, I think it is perfectly reasonable to expect that they can do their own computer work just fine.

BTW, users have reported random issues when PS is assigned a high priority; I have not tested this myself, but Windows warns of instability and other problems when one attempts to change process priority in the Task Manager.

No offense meant Josh – I use computers as play tools too but not the one I am using at the moment (not yet anyway 🙂 )

If the same creative energiess put into past behaviour pattern analysis were applied to an "I prefer" option things would be IMHO a bit better.

Maybe even Win XP Pro (Adobe version)?

Now that really is a partnership possibility

Deebs…
I’m sure your thoughts here are really well-meant but we have been trying to keep this specific to the problem we’re sharing here. Which is not to do with tweaking or overclocking: it’s to do with identifying what might be preventing otherwise well-managed systems from handling Photoshop 9.

I for one really think the less clouded we keep this thread the better…

All best

Glenn

Who is "We"?

It seems to me that the "Curves/Arrow/Info behaviour" can only be described as a bug and an indicator that someone’s been tinkering around and maybe didn’t get it back quite right. Curves and the Info Palette have been around for awhile, there is no reason that these should not function as before. Logic would seem to beg the question, what else could have been changed? Is it possible? It seems that we have one known identified bug and it isn’t related to a driver and it is one of the components that seems to cause some problems, namely the Info Palette.

It seems to me that the "Curves/Arrow/Info behaviour" can only be described as a bug

It’s a bug. However, that doesn’t mean it’s not related to the same issue as the problems other are experiencing.

Precisely the point I was trying to make which is, take a look a the obvious things that aren’t working and use it as an indicator of what might be causing problems in these other areas which, in the past, did not cause a problem either.

Glenn: "Curves/Arrow/Info behaviour": I currently use v7 and have the CS2 trial, I checked with both, others have reported this problem and Ian just confirmed it was a bug.

Re: GeForce4 Ti 4600 card (Asus V8460) with the latest NVIDIA Driver (7.1.8.9) Win 2K SP4. RAM: 1024MB 2700 DDR. MB: ABit BD711-RAId. Norfolk UK; Chilly, moderate to strong Nor’Easterly; Light but unbroken cloud; Visibility….?

~ ~ ~

Today I installed a Driver (7300) marked as later (15April 05) than that available at the moment from NVidia and that I have been using without success til now (7189 -01April). The 7300 is apparently a ‘leaked Beta’. Normally I’d not have even approached it but, given no apparent alternatives as yet…

But… no change. Still everything as before. ( HERE <http://www.pellmell.freeserve.co.uk/PS9/SysSpecs.htm>)

Tomorrow I’m hoping to borrow a Matrox 650 to see how that fares. As explained before, I have done a fresh OS and PS9 install on a separate Drive (without any improvement) so, with the Matrox connected to that, I should learn something…

Before then, can somebody point me to a list of known bugs existing in PS9 that might be mistaken for Display Driver Errors, so I don’t make the same mistake as I did before, by presuming the Curves>Arrow>Info phenomenon I was seeing was a symptom of failure my end…?

I now read someone else (not troubled by the slow display syndrome as far as i can tell) also has found the Layer Transparency during Transform hides the layer til the slider stops: again, I’ve been taking that as a problem with my setup, and had it marked a symptom to test for… Is it or isn’t it…?

How can we – those of us suffering retarded display and trying to help identify possible causes – reach any sensible conclusions when we don’t know whether a flaw is our end or in PS itself…?

There must be a list published somewhere…?

Glenn

The biggest symptom is slow redrawing of the palettes.
Hitting tab twice should hide the palettes and then redraw them almost instantly.

now read someone else (not troubled by the slow display syndrome as far as i can tell) also has found that the Layer Transparency during Transform hides the layer til the slider stops: again, I’ve been taking that as a problem with my setup, and had it marked as a probable symptom to test for… Is it or isn’t it…?

that was me… a bit annoying but not a show stopper. haven’t been back to that thread yet to see what other people have said. thanks for reminding me!

dave,

You can highlight the opacity percentage (click on the word "Opacity") and use the mouse wheel to increase/decrease the percentage without the weirdness. The wheel alone increments 1%, and Shift+wheel goes 10%. A workaround until it gets fixed.

Frank

Alas, turns out the Matrox must wait at least another day…

Chris…
I know slow palette redraw is a symptom. I have said as much. That is not what I was asking. Perhaps if I give (just two) specific examples…?

Will I be right in taking as a sign of a fault my end that: a) Ctrl+Y (using a 590kb Custom Printer Profile) takes at least about twice as long to ‘draw’ in PS9 as it does in PS8; and b) that the (aforementioned) 40px strip in CameraRaw does in PS9 draw about 1 to 2 secs after the rest of the image has tiled, when in PS8 there is no such strip or delay?

Perhaps you could also let us know whether you have yet located and tested a GF4XXX series Card? (Obviously you’re miles better placed to accurately assess what part it might be playing here)…

Dave…
I agree it doesn’t (or wouldn’t) amount to a show-stopper. But, if it were your machine in doubt, and it seemed people were attributing any flaw encountered in PS9 to Driver or Configuration fault, would you not assume aha, another sign of a problem my end..?

And would you not agree it would be useful (not to mention more honourable) if the Adobe people were to be a bit more forthcoming about such faults as do exist in this Release of Photoshop?

Apologies if my fraying patience is starting to show…

Glenn

But, if it were your machine in doubt, and it seemed people were attributing any flaw encountered in PS9 to Driver or Configuration fault, would you not assume aha, another sign of a problem my end..?

sure glenn, i would. but as the rest of my system seems to be working well, i thought you’d like to know that that specific problem doesn’t seem to be related to your other problems, helping you narrow the field a bit.

thanks frank, i’ll try that next time i’m working on that machine…

dave

R yews bee r nawrfok maan?

It’s been a bit of a while since I was at the Wash, Walsingham, Upper, Lower and Burnham Market – great meals and fantastic deli’s

Now then, I know I have been harping on about it but have you tried: Intel Booster/Accelerator?
OptimizeMemory
applied increased priority to CS or CS2/Bridge?

On my system all provide a reasonable performance tweak plus another: when about to do some serious Photoshopping I pull the network wire (don’t tell SysDev team though)

To be fair the wire is only a cable modem device but it help to avoid those sluggish encounters that happen from time-to-time just before a notice to say: Firewall, Anti-virus, Anti-Spyware, Anti-adware, Anti-malware program has updated.

All the above seem to work quite well with stability on CS and/or CS2/Bridge

But I do wonder what performance will be like on the 4000+ AMD thingy with ATI all-in-a-tiz 16x graphocs bored. …
(not spelling mistakes 🙂 )

Intel whatever won’t work on my system.

OptimizeMemory won’t do anything, at least it won’t on my system, partly because I’m not installing something that won’t tell me what it does, and I’m _certainly_ not installing anything whatsoever that comes with an "advanced search bar" for IE (in other words, spyware).

Increased priority doesn’t work, I tried. Others have actually had very weird issues come up after increasing the priority, so it’s definitely not a fix for everyone.

The network connection isn’t going to do anything significant unless you’re 1) running theoretical benchmarks or 2) keeping your system busy with sharing files, printers, etc. _and_ other computers are using those services regularly.

"the wire is only a cable modem device" (this is where on another forum I would insert a popular TLA) methinks we have a slight misuse of terms here. Barring the fact that a "cable modem" isn’t really a "modem" at all (at least I think not, given that a DSL modem isn’t a "modem" either), I don’t see how the wire, the adapter that it connects to on your computer, nor the switch that it connects to on the other end in an Ethernet-based network, could be considered a "cable modem device." Please be more precise in your networking terminology usage. Some of us know enough to be dangerous, so try to use the right words so we don’t go off and do something stupid.

And yes, all my anti-this and anti-that stuff is kept up to date semi-automatically (it checks, flashes, I click yes to update – because I’m a control freak). It’s clear that you have a working system, but that doesn’t mean that if everybody goes out and does things the same way you do, that it’s magically going to make our broken hardware (I assume) work correctly. Besides, we’re already doing most of what you do, and the rest is stuff that you’re going to have a tough time convincing me to try (like needless memory "optimization" software cum spyware).

Dave…
Yes, thanks, I was pleased (in a way!) to learn someone with a true-running machine was also seeing that – cos it did, as you say, narrow the field my end…

But… that is just my point: that it should take the chance reading of another post to learn that! I just think an updated checklist of known problems would prevent a lot of misunderstandings and wasted time…

If you get a chance to try those other two points at some time (gather you’re not on that machine right now) I’d be much grateful…

Deebs…
Yes, I be norfolk lad…
Thanks for the ideas, and I’m glad they work for you but I have tried moreless everything I can think of (including them) and think it’s a case now of letting my cat have a go…

Glenn

Glenn

a) possibly, but the performance will depend on the profile contents. This is because we greatly increased the quality of the CMYK preview (and pretty much all color conversions) since CS.

b) No idea

Yes, we have the NVidia cards and are testing them now.

But… that is just my point: that it should take the chance reading of another post to learn that! I just think an updated checklist of known problems would prevent a lot of misunderstandings and wasted time…

seems to me the guys at adobe (chris is one of the most visible here) are working pretty hard to resolve issues. the program’s only been out a short time.

Chris, thanks… (#46)

Will be very interested to hear what you find with the GF Card.

So, can I ask again: a) is this a sign of a faulty setup my end somewhere, something you (or whoever else with a trouble-free setup) are not experiencing?

And, though i understand you are not responsible for Camera Raw, it is still a part of PhotoShop and I wouldn’t have thought it would take too much time for you to open a RAW, zoom in a couple of times, take a note, close it, and let us know how it compares to PS8 on a fault-free machine?

A number of people have reported delayed display symptoms. You have asked us to help you identify any possible causes of this apparent problem. What happens when we try? Do you really think "No idea" was the most helpful thing you could offer?

So, I’ll ask that again, as well – of you or of anyone with a trouble-free system: b) is there, in PS9, a 40px wide strip down the left side of the CameraRAW image window that takes 1 or 2 secs longer to redraw after the rest of the window has finished retiling following a zoom in or out? Or is it a sign of a
fault my end?

That’s all I’m asking: simple, straightforward questions… A bit of cooperation in trying to help cast some light on a problem lots of us are finding hard to understand. Why do I feel – and not for the first time – it’s like squeezing blood from a stone…

Totally exasperated…

Glenn

Glenn,
I have the same strip in ACR as you describe.
GF4600Ti latest driver; Win XP Home SP1.
Stefan

Thanks, Stefan…

I think we really need someone (one of the 99.whatever percent) without a display lag to step up and say what they find – at least with these two specifics …?

What dismays me is that I’m finding exactly the same behaviour, in every way, with this different (2D optimal) Graphics Card.

I still haven’t tried it in my bare XP setup but I strongly doubt I’ll find any difference…

I don’t think what I’m seeing is as extreme as some people are experiencing; but everything I do – simple rectangle mask, open a dialogue, move an adjustment slider, everything is noticeable, where in PS8 it was instantaneous… It’s just too distracting to work properly…

Maybe when Chris gets back on the GF Card we’ll know more…?

Glenn

Glenn – no, that’s on the edge between expected and problematic. It could be a complex (and large) profile, or it could be other display related issues.

Thanks, Chris…

I’m becoming more and more convinced that what I’m seeing is ‘expected’ behaviour after all. That changes made in PS9 are causing a slower display response than in PS8.

I just don’t understand why someone has not come out and said categorically that, on these two issues, they are seeing no degradation at all in screen redraw since upgrading…

My Trial time’s fast running out…

Glenn

Chris, isn`t there any 2D benchmark test, that is using the API Photoshop is using and that can show wether the video card is able to handle that API at a reasonable speed?
I mean this is very confusing. Many users here have top video cards from ATI,Nvidia and Matrox. Some of them are having speed problems while others don`t, although they have absolutely the same card and driver.
Stefan

And could the skin of windows interfere? Using the blue UI vs the classic one? And the rest of the eye candy.

Pierre…

Not in my case. My regular setup is bare-bones anyway, but I’ve tried it all ways.

One thing I did notice, though – with ‘Show Windows Content’ on, it does display the problem someone here posted a screenshot of: whiteout trailing behind a dragged window…

You’ve reminded me I ought to hunt out that poster and mention it…! Now, which thread was it…?

Glenn

Glenn – several people have said they see no noticable difference between CS and CS2.

What you’re seeing is "expected" only in that you seem to have a problematic video card driver.

Stefan – apparently not.

We’re still trying to figure out what’s going on.

Chris…

"…you seem to have a problematic video card driver."

I wish, I wish, I wish I could agree…

I’ve just tried replacing my NVIDIA GeForce4 Ti 4600 with a borrowed (new) Matrox P650. Both with latest drivers. Tried both on my regular Win2k SP4 setup and also on a test setup of fresh installed WinXP SP2 on a bare Drive with nothing but PS9 installed.

The problems of Display lag were consistent throughout.

Logically, how can that possibly be down to Display card driver or corrupt configuration? Both cards, all drivers (I also tried the XP setup first with Windows’ default driver – Matrox 1.4 I think it was)… all exhibiting exactly the same delay in display response.

I have just finished updating the BIOS of my mainboard (Abit BD711-RAID) to eliminate that as a possibility. Sure enuff, no change…!

That’s nothing else left but the CPU and the RAM sticks, is there? Except, of course, Photoshop CS2…

Thanks for your ear. Realise you’re well busy. If you can think of anything else I might try…

Glenn

We are seeing the Matrox 4400 and 4600 as a bit slower than other cards. (yes, we’re testing them)

We have narrowed it down pretty well to video card drivers. Now we just need to determine what API is causing it, and see if we can work around that problem or if we have to depend on driver updates.

Chris,

Thank you for the courageous update. I am really relieved to hear that Adobe is taking us seriously and looking at the cards, the drivers, and the APIs.

I can appreciate your tremendous credibility and dedication to resolving the issue. And I’m sure others here can, too.

Thank you.

Have you compiled a list of card/driver situations reporting poor performance? I’d hate to be left out in the cold…

ATI X850 XT 256MB/catalyst 5.5 drivers

very best regards,

— font9a

Note: Chris meant nVidia 4400 and 4600 in post 59.

Glenn_UK & font9a: I’d like both of you to email me ASAP so I can work directly with you on these display issues.

Either email me at ajerugim at adobe dot com or travlin_adam at yahoo dot com.

We have a few things we’d like you to test to see if it helps rectify the problem.

thanks,
-Adam

Gosh! I wish I had an nVidia graphics card now…

Well done peeps!

I’ve just tried replacing my NVIDIA GeForce4 Ti 4600 with a borrowed >(new) Matrox P650. Both with latest drivers. Tried both on my regular >Win2k SP4 setup and also on a test setup of fresh installed WinXP SP2 >on a bare Drive with nothing but PS9 installed.

The problems of Display lag were consistent throughout.

Gee, I’ve just ordered Matrox G550. Is G550 model free of these problems? Anybody can confirm?

Pierre – yeah, that’s what I meant. Oops.

font9a – other than the NVidia 4400 and 4600, plus the ATI Catalyst drivers, we don’t have a list.

But I think Adam has found some important clues about the issue. Once we’ve confirmed the behavior, that’ll help narrow down the cause (and why some people don’t notice it) and we can work on a fix or workaround.

Alexander,

My 550 is good as gold. Due to the number of variables involved, I don’t think anyone can guarantee you success. However, the odds are in your favor IMHO.

dfshdiuf
HELLO

Chris, are there any news concerning the "display slowness"? I have noticed another strange behaviour:
When I`m using a brush (and my cursor is set to show the brushsize)the cursor changes to the normal "Windows-arrow" for the fraction of a second after I finished the brushstroke and immediately changes back to "brushsize". The arrow is only visible for maybe 0,05sec..
But if I use the brush again before it changed back to "brushsize" (and that seems to be exactely the moment when the action is recorded in the history palette) the arrow stays there as long as the next brushstroke is going on. So if I paint at usual speed making one brushstroke after the other I end up seeing the arrow all the time…..or I have to wait all the time until the brushstroke is recorded in the history palette.
WIN XP Home, P3 730Mhz (I know, I know…..), nvidia GF4600.

Stefan

Same problem here on two different machines both with GeForce cards. One has a lot more ram than the other and isn’t as bad, but both are significantly slower than CS was. I’d love to hear what solution Adam comes up with.

I would like to hear from Adobe about their progress in resolving the issue.

We’re still working on it…

I seem to be having the same problem of slow display with an ATI 8500 AIW also. I have tried the ATI drivers all the way from two years ago all the way to the current Catalyst drivers and still have the same redraw slowness. This card does execelent redraws in games which I normanaly don’t have on this computer but loaded to check out the graphics. I dont have any qualms about reformating the system drive and reloading which I do on a quartly basis anyway. XP Pro SP2 with all MS updates. System is Asus 533C with 512 meg of RDRAM. PS 5.0, 6.0, and 7.0 run fine on this system. 8.0 and 9.0 are slugs.

sjprg – are you having any problems with the menus and palettes drawing white?

And we don’t have any reports of Photoshop CS (8.0) running slow like that.

sjprg, Chris,

I’m running similar system 2.2gig XP_Home_ sp2, with 1 gig RAM and the 6.14.10.6387 drivers on the ati 8500 and no slowness, menu problems or white palettes. CS2 faster than CS1 but I’ve made some changes – high priority on task manager, set my mem allocation to 55%-65% (was 80% on CS1). All the optimization stuff on paging/scratch, etc.

If I open a picture, then open the info tab (the one with the eyedrop RGB – CYMK colors). Then select my paint brush tool, and move my mouse over the picture, my CPU gets maxed out at 90-100%, and the mouse movement lags/jerks.

If I switch from the info tab, to navigator, then my cpu is ok, and the lag goes away.

This is on top of all the other slow down problems, most others are having.

Is anyone else having this problem?

I’m running an ATI 8500 at work and things seem to have been OK except for the brush/info palette issue.

Thanks Paul.

This might be a crazy question, but has anyone tried CS2 WITHOUT any windows service packs installed?

I know CS2 wants SP1 or better installed, but looks like it would work without it.

I only installed SP1 because CS2 wanted it there, but at the same time, CS2 said I could try to install without the SP.

When SP1 first came out, I installed it, and one of my games I used to play started lagging bad. I removed Sp1, and the lag went away.

Not sure if this could be related in anyway or not.

FWIW, Illustrator CS2 seems a little slower than CS (occassionally stalling for a second or two and maxing out the processor), though it isn’t nearly as bad as Photoshop and I haven’t really used it much yet. It seems to stall on specific operations (opening a file or closing one), whereas PS is just plain slow all over.

Any updates on this? Does it look like something that ATI must fix, or will Adobe be releasing an update to work around it?

We’re still trying to figure it out.

So far, ATI can’t reproduce the problem….

Glad to hear it chris.

I like CS2 so much better than CS now. I really hope Nvidia, ATI, and Adobe figure this out.

After reading one of the posts in this forum, I tried to reduce my "Hardware Acceleration" by one, and it does seem to help the screen redraw lag.

It’s not a huge change, but did speed up photoshop slightly.

My thoughts are, photoshop must be needing new demands from video cards, or video card drivers for that matter, in order to run correctly.

I think Chris Cox was right all along. It is the video card drivers I believe.

At first I didn’t believe this was true, but after talking to a friend who is a computer tech, I realize what the problem probably is.

The video card can’t handle the new demands of photoshop due to drivers that don’t support needed features. As a result, the video card then moves what it can’t handle over to the CPU, which in turn, runs at a high load percent, which causes lag.

For those who have duel CPU’s, the video driver insufficiency has little to no impact because they have plenty of CPU to make up the difference.

This is all speculation, but after the new video drivers come out, most of us might see our problems go away.

Unfortunately, it’s not JUST the drivers.
It seems to have something to do with VGA vs. DVI connections, refresh rate, and the phase of the bloody moon.

ATI and Nvidia are having trouble reproducing the problem, and we can’t always reproduce it on current video cards (4 year old cards are easy ;-).

So, fat0n3s – are you using a CRT or LCD?
If LCD, is it a VGA or DVI connection?
What refresh rate?

I know you didn’t ask me, but the LCD vs. CRT thing applies, so here goes:

Athlon XP 1700+, 1GB PC3200 DDR SDRAM, ATI Radeon 7200 with Cat5.5 drivers and VGA connection only. Until earlier today, I was using a 19" CRT () on the analog connection (yes, I am going to use DVI, but can’t until I get a new video card, which won’t be until I get a PCI-E supporting mainboard, which won’t happen until my Athlon X2 gets here later this month). If you want me to try anything else (I’ve already got the latest drivers AFAIK and tested the reduced hardware acceleration thing), I’d be more than willing.

Thanks for all the work you’ve done troubleshooting this already.

Josh – if you separate all the palettes (and I mean ALL) and place them all in the Photoshop workspace, then hit tab twice, how long does it take to redraw? (how slow are we talking about)
How much does it speed up when changing the acceleration?

I am having the same problems that fatON3 has and I see the same problem with my CRT (Sony F520 at 100 hertz refresh) and on my other system with my Samsung 193P with a DVI at 60 hertz. (Two different Motherboards, and different ATI video cards.

Paul – see the questions I asked Josh. And I’d like to get times for both your VGA and DVI systems (and the models of cards used).

CC 4 year old cards are easy…

Could you do a fix for five year old ones. 😉

Thanks for the reply Chris.

I just bought a Nvidia Geforce 5500 256 DDR Ram PCI vid card 1 week ago, and the problems didn’t change.

I just exchanged that card for an ATI Radeon 9250 DDR 256 Ram PCI card today, and the same problems remain; although I think I like this ATI a little better than the Nvidia. =)

Anyways, my current specs are:

17" NEC Multisync FE700+ CRT monitor. I run it at 1024 X 768 85Hz refresh rate.

Are you thinking a LCD would not have the refresh rate problems of a CRT?

I use a CRT because I heard it has a larger color gamut than a LCD. Not sure how true that is though.

fat0n3s, please time the refresh delay of all palettes on screen, as Chris asked.

There was alot of posts while I was typing. =D

Ok, here are my results from Chris’s test.

I opened ALL of the pallets, and seperated every one by itself.

With FULL hardware acceleration, it takes about 2.8 secs to open all of the pallets, after hitting the tab key.

With hardware acceleration reduced by one, the pallets take about 1.8-1.9 secs to open after hitting tab.

And set to zero?

fatOn3s–

What’s interesting is that on almost the same monitor (NEC MultiSync FE750–VGA) It takes less than a second…however, I am using a Matrox G550 card, I even tried it with all of the palettes open in the second monitor and the same thing…almost no delay…

fatOn3s – next experiment, change the refresh rate as high and low as it’ll get and retime it.

And do you mean PCI Express, or really PCI? (as opposed to AGP)

Pierre,

After setting hardware acceleration to none/zero, I get the same delay as reducing by only one.

Christine,

That is interesting. I wish there were stores by my house that sold Matrox video cards. It sounds like that is the card to have when working with graphics. I might have to just order me one. =)

What puzzles me most, is the lack of info on the web about "tweaking" your video card for design. You can find hundreds of sites that talk about tweaking performance for games, but is there any for design, and photo editing?

Chris,

Changing my refresh rate did not have any affect.

From 60Hz – 85Hz, which is the highest for me at 1024×768, the pallet delay is the exact same.

1.8 secs at zero hardware acceleration, and 2.8 secs at full hardware acceleration, no matter the refresh rate.

And yes, it is just a PCI card, not a PCI express. =(
There is no AGP port on this motherboard.

Maybe alot of my problem is my computer is becoming out dated.

I keep telling the wife this, but she never seems to agree. LOL.

That could be a big part of it in your case.

But that doesn’t explain the problems for other users.

Here are my system specs again:

512 MB ram. Which is the max this mobo can handle.
Pentium III 1.0Ghz CPU.
ATI Radeon 9250 256 MB Ram PCI. With latest driver.
2 – 40 GB Maxtor Hard drives. The windows swap file is on the 2nd drive, in its own partition.

Now, CS ran quick for me. Do you think there would be big difference in speed from CS to CS2, with the system I have?

Because CS2 calls one of the graphics APIs a little differently, and apparently not all video card drivers handle that call very quickly.

I opened every palette and on both machine it only takes about 1 second to redraw the workspace.

The ATI card that I use with my CRT (Sony F520 @ 100Hertz) is a Fire GL 8800 128m.

My other systems uses an ATI 8500 128m @ 60 hertz for the DVI out to the Samsung 193P

The problem I’m seeing that is the same as fontOn3 is the brush lag with the info palatte open and the 100% CPU usage when using any of the brush tools with the info palatte open. (if the info palatte is closed everything is normal, it also doesn’t seem to matter which brush shape (from preferences) you chose.

As an aside, I’m currently building a new system to replace my graphics system on the Sony:

ASUS A8N-SLI DELUXE
AMD Athlon 64 4000+ San Diego
Western Digital Raptor WD740GD 73GB 10,000 RPM 8MB Cache Serial ATA150 Western Digital Raptor WD740GD 73GB 10,000 RPM 8MB Cache Serial ATA150 Western Digital Caviar RE WD2500SD 250GB 7200 RPM 8MB Cache Serial ATA150 ATI 100-435513 Radeon X800XL 256MB 256-bit GDDR3 PCI-Express x16 4 Gig RAM
Windows 64

Probably have it up by next weekend and it will be interesting to see if it still maxes out the processor usage with the info palatte open.

The info palette lag is something different (and being seen by fewer people – but thanks for the info on it).

Right now we’re trying to solve the palette drawing issue, it should be almost instantaneous to redraw (as it was in CS – under 0.2 seconds).

Paul – that is some bit of kit

Two Raptors? It’ll be like Jurassic Park when the whole system is up and running with that 250 GB WD coming in as a rather large T Rex?

Well I wouldn’t exactly call it an instantaneous redraw as they come back quickly but individually.

Deebs: Yea, my friends say I can be a bit impatient waiting on apps

Paul

Plus a multi-core CPU in a year or 2?

Does it give a free suntan with wind swept look?

Probably, It’s a disease

Athlon XP 1700+, 1GB (2×512) Corsair XMS PC3200 DDR SDRAM ATI Radeon 7200 (PCI) with Cat5.5 drivers, connected to Samsung 213T by an analog connection Epson printer

I tried several tests; with full acceleration, the redraw time is around 2.18 seconds. At one stop down from full acceleration, the time is around .75 seconds. At no acceleration, the time is slightly over 1 second, but everything in every program appears in an extremely slow "wipe down" fashion.

I also tested on full acceleration with a disabled my antivirus program (Bitdefender 8), and with the client shut off. Neither test case caused any change in the results.

BTW, I can not change the refresh rate with this LCD… I’m stuck at 60Hz no matter what I do.

Josh – thanks, that helps. I’ll have the ATI folks take a look at your config.

And for curiousity’s sake – how long did it take Photoshop CS to do the same redraw?

FWIW, I am having no problems with redraw of the palletes on my ststem. The video card is an ATI 7500 series, it’s an Athlon 3800+ with 2 GB of memory, and the palettes mostly live on a second monitor. I don’t know if this means anything, but I thought I’d post it in case it offers up any answers.

Perhaps the palletes on a second monitor are the answer? Just a moment, I’ll check it out…….

No, it makes no difference if the palletes are on one monitor or two. The redraw is a fraction of a second, I certanly couldn’t time it!

Oh, well, hope it helps!

Chris, I’m sorry I don’t have PSCS around any more to test, but I can assure you that it was an order of magnitude faster. Anyway, thanks again for looking into this issue. 🙂

And BTW, that was "Bitdefender 8" – evidently the forum software converts an 8 and a ) into a smiley of some sort.

I’m seeing as you typed it, no smilies.

I suspect that somewhere there is an option to turn "8" + ")" into a smiley icon.

Hi,

I would like to add I am getting slow redraw, symptoms as reported in this thread. camera raw is also very, very slow, but I figure it is likely machine related since it’s an old system. MSI + Via chipset mobo, dual p3 800, 512mb cas2 ram, two 20GB 7200rpm drives in motherboard determined raid 0 (software based though), one 200mb 7200rpm drive, and ATI Fire GL 8800. Runnin XP Pro, sp2. I have a 19" crt monitor plugged into the dvi slot using the supplied analog->dvi adaptor. 19" dvi lcd on it’s way in.

Takes 2-3 seconds to redraw using tab-test with full hardware accelleration, and might just be placebo affect that I see it a wee bit faster with hardware accel off or a notch down. Vid-card drivers used: one from couple years ago, and latest one off of ati’s website.

Please let me know if I should run any other tests.

Looks like matrox g400, 450, and 550 can be had pretty cheap on ebay, should I just get one of those for now?

Thank you for your time,
jano

I have PS CS2 using a GeForce FX5500 256 MB. The driver is ForceWare 71.89 (latest) on a Samsung 213T using DIV input and analog on the second monitor.
I have none of the problems others have noted.

Someone else said they had tried this card and it didn’t help so I thought I’d add this info.

Rob E.

Robert Enns,

Please post your system specs here.

Are you running PCI, or AGP?

Did you do any video card tweaks? IE, Riva Tuner…etc.

I just returned that exact same card because it had the same problems as my geforce 4 mx 420 with slow redraw, and 100% cpu usage with info pallet open.

I have the same problems with my new ATI Radeon 9250.

BTW, this Radeon 9250 seems to have MUCH better 2D graphics then the two Nvidia cards I mentioned. It seems I can see so much more when editing pictures in photoshop.

Maybe it’s just me.

Fat0n3s,
if you buy a nvidia card and want good 2D quality, you have to buy it from manufacturers, that are known for using high quality components.
The "same" card differs in 2D quality from manufacturer to manufacturer. I can highly recommend Gainward cards, which are known for very good 2D quality. Noname manufactureres often use cheap signal filters.
Stefan

I see.

Thanks for the info Stefan.

Does ATI only have one manufacturer that builds cards?

Ati cards are available from various manufacturers as well, for example www.elsa.com Stefan

Fat0n3s,
Here are my specs.

Adobe Photoshop Version: 9.0 (9.0×196)
Operating System: Windows XP
Version: 5.1 Service Pack 1
System architecture: Intel CPU Family:15, Model:2, Stepping:9 with MMX, SSE Integer, SSE FP, SSE2, HyperThreading
Physical processor count: 1
Logical processor count: 2
Processor speed: 2792 MHz
Built-in memory: 1535 MB
Free memory: 947 MB
Memory available to Photoshop: 1377 MB
Memory used by Photoshop: 55 %
Image cache levels: 2
Serial number: xxx
Application folder: C:\Program Files\Adobe\Adobe Photoshop CS2\ Temporary file path: C:\DOCUME~1\Robert\LOCALS~1\Temp\
Photoshop scratch has async I/O enabled
Scratch volume(s):
D:\, 4.89G, 4.11G free
F:\, 9.84G, 6.91G free
Startup, 111.7G, 69.8G free
Primary Plug-ins folder: C:\Program Files\Adobe\Adobe Photoshop CS2\Plug-Ins\ Additional Plug-ins folder: not set
Installed plug-ins:
ADM 3.10×21
ASDStrm 1.03×2
Accented Edges 9.0
Angled Strokes 9.0
Average 9.0 (9.0×196)
BMP 9.0 (9.0×196)
Bas Relief 9.0
Bigger Tiles 9.0 (9.0×196)
Camera Raw 2.4
Camera Raw 3.0
Chalk & Charcoal 9.0
Charcoal 9.0
Chrome 9.0
Cineon 9.0 (9.0×196)
Clouds 9.0 (9.0×196)
Color Halftone 9.0 (9.0×196)
Colored Pencil 9.0
CompuServe GIF 9.0 (9.0×196)
Conditional Mode Change 9.0 (9.0×196)
Contact Sheet II 9.0 (9.0×196)
Conté Crayon 9.0
Craquelure 9.0
Crop and Straighten Photos 9.0 (9.0×196)
Crop and Straighten Photos Filter 9.0 (9.0×196)
Crosshatch 9.0
Crystallize 9.0 (9.0×196)
Cutout 9.0
Dark Strokes 9.0
De-Interlace 9.0 (9.0×196)
Difference Clouds 9.0 (9.0×196)
Diffuse Glow 9.0
Displace 9.0 (9.0×196)
Dry Brush 9.0
Embed Watermark 1.70.19
Export Transparent Image 9.0 (9.0×196)
Extract 9.0 (9.0×196)
Extrude 9.0 (9.0×196)
FastCore Routines 9.0 (9.0×196)
Fibers 9.0 (9.0×196)
Film Grain 9.0
Filmstrip 9.0 (9.0×196)
Filter Gallery 9.0
Fit Image 9.0 (9.0×196)
Fresco 9.0
Generic EPS 9.0
Glass 9.0
Glowing Edges 9.0
Grain 9.0
Graphic Pen 9.0
HDRMergeAlign 9.0×001
HDRMergeUI 9.0×001
Halftone Pattern 9.0
Ink Outlines 9.0
Lens Blur 9.0
Lens Correction 9.0
Lens Flare 9.0 (9.0×196)
Lighting Effects 9.0 (9.0×196)
Liquify 9.0
MMXCore Routines 9.0 (9.0×196)
Merge to HDR 1.0×001
Mezzotint 9.0 (9.0×196)
Mosaic Tiles 9.0
Multiprocessor Support 9.0 (9.0×196)
NTSC Colors 9.0 (9.0×196)
Neon Glow 9.0
Note Paper 9.0
Ocean Ripple 9.0
OpenEXR 9.0 (9.0×196)
PCX 9.0 (9.0×196)
PNG 9.0 (9.0×196)
PNG Icons 1.22×1
Paint Daubs 9.0
Palette Knife 9.0
Patchwork 9.0
Paths to Illustrator 9.0 (9.0×196)
Pattern Maker 9.0 (9.0×196)
Photo CD 9.0 (9.0×196)
Photocopy 9.0
Photomerge 9.0 (9.0×196)
PhotomergeCylmap 9.0 (9.0×196)
PhotomergeRender 9.0 (9.0×196)
PhotomergeUI 9.0 (9.0×196)
Picture Package 9.0 (9.0×196)
Picture Package Filter 9.0 (9.0×196)
Pinch 9.0 (9.0×196)
Pixar 9.0 (9.0×196)
Plaster 9.0
Plastic Wrap 9.0
Pointillize 9.0 (9.0×196)
Polar Coordinates 9.0 (9.0×196)
Portable Bit Map 9.0 (9.0×196)
Poster Edges 9.0
Radial Blur 9.0 (9.0×196)
Radiance 9.0 (9.0×196)
Read Watermark 1.70.19
Resize Image 9.0 (9.0×196)
Reticulation 9.0
Ripple 9.0 (9.0×196)
Rough Pastels 9.0
Save for Web Version 9.0×012
ScriptingSupport 9.0
Send Video Preview to Device 9.0 (9.0×196)
Shear 9.0 (9.0×196)
Smart Blur 9.0 (9.0×196)
Smudge Stick 9.0
Solarize 9.0 (9.0×196)
Spatter 9.0
Spherize 9.0 (9.0×196)
Sponge 9.0
Sprayed Strokes 9.0
Stained Glass 9.0
Stamp 9.0
Sumi-e 9.0
Targa 9.0 (9.0×196)
Texturizer 9.0
Tiles 9.0 (9.0×196)
Torn Edges 9.0
Twain Acquire 9.0 (9.0×196)
Twain Select 9.0 (9.0×196)
Twirl 9.0 (9.0×196)
Underpainting 9.0
Vanishing Point 9.0
Variations 9.0 (9.0×196)
Video Preview 9.0 (9.0×196)
WIA Support 9.0 (9.0×196)
Water Paper 9.0
Watercolor 9.0
Wave 9.0 (9.0×196)
Web Photo Gallery 9.0 (9.0×196)
Wind 9.0 (9.0×196)
Wireless Bitmap 9.0 (9.0×196)
ZigZag 9.0 (9.0×196)
Plug-ins that failed to load: NONE

Installed TWAIN devices:
WIA-HP Scanjet 4570c/5500c
HP Scanjet 4570c/5500c
Dell Image Expert Supported Camer

I have no video card tweaks.

Rob E.

Thanks alot Robert.

The only difference I see is your system is about 4 times faster than mine. lol.

Chris,

I installed Photoshop CS1, and ran the same pallet test as I did with CS2.

The pallets redraw so fast in CS1, that I couldn’t time them.

Also, with the info pallet open in CS1, my cpu load doesn’t go over 50%, and there is no mouse lag what so ever.

This is with full hardware acceleration.

I tried with the latest and greatest Catalyst 5.6 drivers and the slow redraw/repaint problem of CS2 still exists (I didn’t encounter anything similar in CS).

Are there any follow-ups about a possible solution?

———————————————————— —

Adobe Photoshop Version: 9.0 (9.0×196)
Operating System: Windows XP
Version: 5.1 Service Pack 2
System architecture: AMD CPU Family:6, Model:10, Stepping:0 with MMX, SSE Integer, SSE FP Physical processor count: 1
Processor speed: 1837 MHz
Built-in memory: 1535 MB
Free memory: 1011 MB
Memory available to Photoshop: 1378 MB
Memory used by Photoshop: 90 %
Image cache levels: 6
Serial number:
Application folder: C:\Program Files\Adobe\Adobe Photoshop CS2\ Temporary file path: C:\DOCUME~1\ADMINI~1\LOCALS~1\Temp\ Photoshop scratch has async I/O enabled
Scratch volume(s):
E:\, 149.0G, 6.04G free
Startup, 25.4G, 8.32G free
Installed TWAIN devices: NONE

Hello,

Just wanted to give result of a silly test. Please refer message #110 for my outdated system specs.

– I was able to coax the machine to booting with 1gb ram. No change in performance – I purchased a matrox g550 and nvidia quadro4 nvs200 for very cheap:
i) system wouldn’t boot with nvidia, oh well
ii) no change in tab-tab test with matrox vid card
iii) dropping HW accel to zero made tab-tab with matrox quicker by 1s. However, other XP windowy type things were now poor
iv) matrox display looks very nice 🙂

It’s probably a combination of old system plus drivers not too happy being on via stuff? I can try a few bios tweaks (agp aperture, etc), but doubt those will have any affect. Guess I’ll just deal with this until I can get a new system.

Bye,
jano

"we are making some progress on the problem. More new later" (Chris Cox, June 20th in this thread).
Chris, it is nearly November now and there is still no solution, right? Some day CS3 will be released. If you can`t solve that problem, am I right to assume, that CS3 will still have that same specific problem?
Why can`t Adobe get rid of that problem? CS2 is out since half a year. And I`ve seen that problem on computers of some friends of mine also, that have totally different machines (including video card)
than I have. Can we expect to get a statement ( or even better a patch)from Adobe some day concerning that bug or do we have to wait for CS3 or even longer?
Stefan

A few months ago I wrote the following about the dispaly behaviour: "When I`m using a brush (and my cursor is set to show the brushsize)the cursor changes to the normal "Windows-arrow" for the fraction of a second after I finished the brushstroke and immediately changes back to "brushsize". The arrow is only visible for maybe 0,05sec..
But if I use the brush again before it changed back to "brushsize" (and that seems to be exactely the moment when the action is recorded in the history palette) the arrow stays there as long as the next brushstroke is going on. So if I paint at usual speed making one brushstroke after the other I end up seeing the arrow all the time…..or I have to wait all the time until the brushstroke is recorded in the history palette.
WIN XP Home, P3 730Mhz (I know, I know…..), nvidia GF4600. "

Now I have noticed exactely the same behaviour on different (much faster….about 3Ghz) PCs. It`s the same behaviour on each PC I`ve tried. No matter what video card, CPU etc..
The only difference is, that on a fast PC you normally don`t notice that behaviour, because it happens extremely fast.
The interesting thing is, that the cursor does NOT go back to the normal Windows cursor , if the info palette is closed. Then everything behaves normally.
Chris told us months ago, that Adobe is close to a solution. Now nobody is interested in that bug anymore. Adobe obviousely does not make any progress (at least Chris doesn`t tell us anymore) and the users in this forum obviousely gave up on it.
Stefan

"And the Meek shall inherit the Earth… (if the rest of you don’t mind)…"

6 months now… Certainly beginning to look like our being patient, polite, and obliging isn’t getting us very far, is it…?

It seems it (the range of problems I and others here have reported) has been back-burnered to being nothing but an inconsequential "slow redraw of palettes"… I see little point in reiterating experience to the contrary, but thought instead i’d try again to isolate something my end that might yet help to identify the cause of this slow screen redraw.

I used Tom Fors’ ACR Calibration script on a Canon G2 target shot. This, on my machine, takes 100% CPU throughout its 20+ mins duration. In both PS8 and PS9 I could, naturally, do nothing (within PS) while it ran. The difference appears in what happens when minimising and then restoring the PS window with this Script running. PS8 handled this impeccably: as one would expect, all screen components redrew instantly on Restore. (I’m presuming all those healthy PS9 installs out there behave likewise?) Yet this ScreenShot <http://myweb.tiscali.co.uk/pellmell/tech/CS_ScreenShot.htm> shows my PS9 window with the palettes and active document window remaining un-redrawn – not just momentarily but indefinitely… until, and this is what struck me (and I’ve no tech-knowledge here to do any more than suspect it’s possibly significant?) another activity – outside of PS: eg Click on TaskBar; Alt-Clck to another open App – would, after a 2 – 3 sec pause, wake PS into action and the redraw would then commence.

Two additional points to note: a) PS window going to and from Maximised and ‘Floating’ (middle icon: instead of Minimised) caused no loss of Screen components. b) If, during the 2-3 secs pause before PS wake-up, another window, eg Explorer, is opened (floating) over PS, then the PS redraw does not commence.

No idea if this is of any use at all but, rather than just going on waiting impotently…

Glenn – that’s because of the changed preference for accelerated action playback (doesn’t redraw unless it has to) instead of step-by-step playback (slower). Photoshop is busy when running the action/script – why would it slow down to redraw everything?

Sorry, Chris, don’t understand your answer.

"why would it slow down to redraw everything?"

Why should it need to? When PS8 didn’t?

My layman’s logic… let’s reduce it to this: ignoring what Tom Fors’ script specifically does (calling actions?), it is occupying Photoshop’s full attention, yes? Okay, so I minimise a fully occupied PS8 and, restoring it, the Screen elements are instantly restored. I minimise a similarly occpied PS9 and the screen elements fail to restore – until jogged into life by an activity outside Photoshop. Without such an activity, the palettes etc would remain absent.
So, what calls are being made in PS8 that are not in PS9 on restoring from minimise?

I realise this is really no more than an extension of the original complaint: it just seemed notable that what was visible only as a brief stutter was evidenced here as a perpetual condition (a virtual freeze) – one which responded only to external activity.

Just thought that might help in narrowing down which of the myriad GUI(?) calls you’ve mentioned might be at play… (Wishful thinking, eh?)

Appreciate your time, regardless…

Because drawing takes time, and if you redraw the image and palettes after each step that time really adds up.

CS, by default, drew at every step.
CS2, by default, does not draw at every step, and runs actions and scripts faster because of that.

When you maximize, CS2 is still busy, and it’s not drawing at every step – so it doesn’t draw until it has to.
CS was drawing at every step, so it drew as soon as some step of the action/script completed.

This _is_ a documented preference that you can change…

Yes, i did get your point about the results of Action States being not constantly redrawn. (I had PS8 set to Accelerated also – and most palettes wouldn’t need redrawing at all in the course of this script, anyway).

But my last post tried to remove the nature of the Script itself from the picture (my mistake for introducing it) and let it simply represent any procedure that fully engaged the CPU for 20 minutes.

It remains that the symptom of an non-refreshing screen – palettes, doc window et al – is still present, not just for a fleeting moment, but caught in a state of suspended animation for 20 odd minutes (unless woken from without) and just waiting to be scrutinised, I’d imagined, by some keen debugging analyst…

Chris, can you tell us wether Adobe is still working on that problem at all? Months ago you told us, you`re close to a solution.
The behaviour that I described in my last post is the same on each and every PC with CS2 that I´ve seen until now. For me that prooves, that it is not related to any hardware or driver.
Stefan

We are still working on it. But that’s about all I can say.

The palette redraw behavior is not the same on all PCs: it is much worse on some, and almost absent on others. (of course, it’s almost absent on most of the machines we have here — there is some factor that we still haven’t isolated making it worse on some systems).

The palette redraw maybe different on different machines, but the strange cursor behaviour that I described earlier is the same on all PCs I´ve seen. Only the speed with which it happens depends on the CPU speed. On a fast PC you can hardly see it at all, but it`s there as long as the info palette is visible.
Just try to make one brushstoke after the other as fast as you can with the shortest possible time between the brushstrokes. If you do it fast enough, you will get a cursor arrow instead of a "brushsize".
On a slow PC, you can do your brushstrokes rather slowly and will still get the cursor arrow. That`s totally annoying and you have to close the info palette.
Stefan

On Sun, 6 Nov 2005 03:51:03 -0800, wrote:

The palette redraw maybe different on different machines, but the strange cursor behaviour that I described earlier is the same on all PCs I´ve seen. Only the speed with which it happens depends on the CPU speed. On a fast PC you can hardly see it at all, but it`s there as long as the info palette is visible.
Just try to make one brushstoke after the other as fast as you can with the shortest possible time between the brushstrokes. If you do it fast enough, you will get a cursor arrow instead of a "brushsize".
On a slow PC, you can do your brushstrokes rather slowly and will still get the cursor arrow. That`s totally annoying and you have to close the info palette.
Stefan

After adding a new LCD display to my system, I now see the strange cursor behavior on the LCD display – never saw it on the same system when using my trinitron CRT display. Thought this observation might help troubleshooting of this problem…

Ted

Hello Stefan…

What image/brush size are you finding this happens with? And does/did it not happen similarly on your machine with PS8?

I have tried it and find no cursor switch. I went with a 40+MB image and 900px brush (cursor as size), working as rapid as I could but, apart from the eventual image-redraw delay – well understandable at such a rate – i noticed only the briefest of hourglass symbols, and never got locked out of brush cursor. (That’s all with Info, and History up).

Indeed everything repeated identically in PS8 took a lot longer (to redraw)..! (I’m with 2.6 P4 and 1.5GB RAM).

It’s odd you seeing this behaviour on faster machines… which is why I’m wondering what scale you working on for this to happen?

Glenn

Glenn,
I guess it happens independently of brush size and image size, but most of the time my file sizes are about 22MB (Canon 20D files) and sometimes about 50MB (1DS MK2). As soon as my brushstroke ends, I always see the normal Windows cursor for the fraction of a second (depending on the speed of the PC).Then it changes back to "brushsize". On a fast machine you have to look hard, but it`s there. As I said earlier, if I start the next brushstroke in exactely that moment the Windows cursor stays forever…until I stop brushing.
I will try it tomorrow with CS (1) as soon as I get to work. Stefan

Glenn, the behaviour is the same in CS1 as far as I can tell. But I`d guess it`s happening 10 times as fast and I had to try very hard even on a lame 750MHz machine to get the Windows cursor.
So maybe this behavior is more or less correct, but it`s simply MUCH slower in CS2. Stefan