That's one of the caveats of beta software ... it allows for the testing of the feature sets, though beta software usually isn't optimized for the best possible performance as that optimization is the last process before finishing up the final release. That process can be pretty intensive and most developers don't care to go through the process multiple times.
This is one of the reasons developers also don't recommend betas for production work ... but for testing and experimentation at your leisure to offer feedback to the developer.
Same for me, very slow.
Unusable for real use, but at least we can get a glimpse of the new features.
I was using the adjustment brush on -100 saturation to desaturate a whole part of a D700 dng image, my Win7 machine with 3go ram would freeze and not complete the task. Tried it on Mac OS X Lion too, not as slow but still far behind LR 3.
There is quite a big lag when moving sliders and clicking buttons and it seems to increase after a moment using it. After 1 hour I had to click several times on some items to activate them when building a book for instance
The most fundamental aspects of the LR workflow haven't changed between LR3 and LR4b. The only exception might be the addition of a step for geo-tagging in LR4 (if that is pertinent to your work). Thus, in terms of high-volume production work, there's no reason to think that LR4b would be all that much different from LR3... there's not really any need to "test" LR4b in the context of high-volume production work.
The only things worth testing, in that case, are the various new enhancements... and that can be done on a low-volume, "sample" basis.
As someone who uses Lightroom 3 every day, I can tell you without any shadow of doubt that Lightroom 3 is considerable faster than Lightroom 4 in it's present form. The only way to test a program is to use it as you would normally, not by choosing the odd file file and seeing how that works, that will tell you nothing about how the software operates in a work situation. I personally think that Adobe will fix the speed issue during the beta testing period. It is offererd as a public beta in the first place, for us the public to give Adobe feedback about how the software operates in all situations!
The problem we are encountering with Lightroom 4, apart from the speed issue, is the same that is faced by all software manufacturers when they try to make a software program that is all things to all people. I would happily forgo the geo-tagging and video capabilities of Lightroom 4 in favour of a well sorted base software with options to add those as bolt on's at purchase or at a later stage when required.
Agreed... LR4b is not as speedy as LR3. In my own use, thus far, I haven't noticed too many enormous speed deficiencies... but I concur that it is slower overall.
My only point was that, given the assumption that the final release of LR4 will be refined to be mostly comparable in speed to LR3, there's no other wildcards that would affect LR's performance as high-volume post-processing software. Thus, there's no reason to fear that LR4 would be any less equipped for high-volume work than LR3. The workflow is still essentially the same... the only variable is the software's processing performance. If you "personally think that Adobe will fix the speed issue", then there's nothing to be worried about and no necessity to put the beta version through the paces with high-volume processing.
The whole point of the beta is to allow us to sample the new enhancements and provide feedback, not to put what is arguably "prototype" software into a rigorous production scenario. It's not refined for that scale of use yet. If the fundamental workflow of LR4 represented a massive diversion from that of LR3, it would be a different story... but the two versions are fundamentally the same and there's no reason that our workflow would take significantly longer with the final release of LR4.
As far as wanting a "well sorted base software" and foregoing geo-tagging video and what not, I'm not really sure what you mean. Forego geo-tagging for what? I don't get it. But if your only hang-up with LR4b is the performance of the beta version (which is certain to improve by the final release), then it would seem that you like LR4, right?
I do like most of the software that is correct. But I also think you should fully test software in order to give good feedback to the manufacturer, that is not really possible with this version taking the speed limitations into account. If Adobe get that sorted quickly then we can really see how good LR4 is and what neeeds to be addressed in the final version. I personally do not need the geo-tagging or video capabilities within LR4. I would like to see them as optional add on's.
Andrew Rodney wrote:
Ken McKay wrote:
Point taken, but how can you test a product fully without putting it through real production situations?
You wait until it is released. Beta software is NOT optimized for speed. There are bugs. Working production on a beta is unsafe.
And I'm near certain there will be a 30 day trial, as there always has been.
Now that LR4 has been RELEASED I downloaded the trial.
Compared to LR3 I do feel it's a lot less responsive. Moving sliders like clarity produces so much lack that it's kind of impossible to get a feel for subtle changes.
I'm on OS 10.7.3 on a 8-core 3GHz Mac Pro with 24GB RAM, so I don't think performance is due to a lack of system resources. The only thing that might make it run less than optimal is the fact that my somewhate aged Mac does not run OS natively in 64 bits, but LR should.
I like LR3 for the immediate response of controls. That feeling is gone with LR4.
The development algorithm is clearly different from LR3. It brings out a lot more levels in deep shadows.
It's hard for me to say if that's a good thing, yet, but the fact that everything is so sluggish makes it hard to have fun testing the impact.
Is anybody out there who can say if performace has improved over beta as so many posts have predicted?
What's your feeling towards usabilty?
When I look at the CPU load on the activity meter, something as simple as changing the black level in LR4 produces a massive spike of activity on all cores. A similar manipulation in LR3 is hardly noticable. No wonder the sliders don't really follow small changes.