Lightroom 3.3 Performance Feedback

[Tom Hogarty]

Please use this discussion topic for your feedback on Lightroom 3.3 RC and the final Lightroom 3.3 release when it becomes available.  The Lightroom team has tried very hard to extract useful feedback from the following discussion topic but due to the length and amount of chatter we need to start a new, more focused thread.  Please post specifics about your experience and be sure to include information about your hardware configuration.

Regards,

Tom Hogarty

Lightroom Product Manager

[Sean McCormack]

So Dan Tull posting here last week doesn't count as Adobe staff? Ironic because he is. Julie Kmoch was also in the Facebook thread the other day. Eric Chan is a regular poster too. Of course, these are users to user forums, so there's no indication that staff will ever be here. But they are.

As to bug reports you won't get a reply unless it's something really serious that's hard to reproduce their end. If they can reproduce it, they've no need for further contact. becs often posts in threads for more information.

[Shocker2000]

I agree. Dan was looking for feedback on 3.4 RC to see if improvements had been made. I was more than happy to provide feedback and answer questions. Heck, I'd even drive my laptop over to them to look at if it helped improve the product.

[hamish niven]

and I nearly downloaded th 3.4RC....

very glad I did not,

going to look at Aperture 3.1 and see what that does, but I'm really hoping LR gets fired up properly 3.4 proper

[Guest]

Seán McCormack wrote:

So Dan Tull posting here last week doesn't count as Adobe staff? Ironic because he is. Julie Kmoch was also in the Facebook thread the other day. Eric Chan is a regular poster too. Of course, these are users to user forums, so there's no indication that staff will ever be here. But they are.

As to bug reports you won't get a reply unless it's something really serious that's hard to reproduce their end. If they can reproduce it, they've no need for further contact. becs often posts in threads for more information.

I think you make it a bit easy for the Adobe staff. While I agree that we may not expect that every single bug report is answered, I think it is not appropriate that bugs are acknowledged and then nothing is heard about it again and nothing is done at all, waiting release for release.

If you defend such a policy, it may be fine with you, but in general I have the feeling that paying and loyal Adobe customers deserve something else - see the thread about bug fixing policy. I am not talking about hard to reproduce performance issues, I refer to clearly reproducable and acknowledged bugs, such as the keywords with spaces issue.

Even if this is a "user to user" forum, a minimum of hints, what they gonna do with such long standing bugs, would not harm them and would not cost a lot of their "precious" time.

If Adobe things that a "user to user" forum from the company is not a place for also informing about bug fixing intentions (because it is only a "user to user" forum, they should rather close this forum and leave the operation of such a platform to independant institutions. I am not saying that Adobe staff does not participate here, but the information policy about bug fixing progress isn't where it should be.

[Sean McCormack]

Look around at any larger company and you'll find the policies to be the same. Take a fruity computer company: much worse and very closed. Negative threads and discussions get deleted there, but not here.

You've seen the bug fixing policy thread, so you know how it gets approached. Time spent answering emails is time not spent working on things. I'd prefer the latter over the former any day.

[GaryRowe-gcvt07]

Not where I worked!     OK the s/w it produced wasn't for image editting - is it a special case for them?

Gary

[Guest]

I think also that spending more time on answering emails than on bug fixing is not what we want, but on the other hand we don't see the results that they spend more time on bug fixing. I don't think that we should take comapnies, which are even worse in information policy as examples.

Seán McCormack wrote:

Look around at any larger company and you'll find the policies to be the same. Take a fruity computer company: much worse and very closed. Negative threads and discussions get deleted there, but not here.

You've seen the bug fixing policy thread, so you know how it gets approached. Time spent answering emails is time not spent working on things. I'd prefer the latter over the former any day.

[Samoreen]

Hi,

tgutgu wrote:

I think also that spending more time on answering emails than on bug fixing is not what we want, but on the other hand we don't see the results that they spend more time on bug fixing.

Agreed. Given the big amount of time spared by not answering so many bug reports or messages in the forum, there should not be any long standing (or persisting) bug in LR. However, some bugs reported in version 1.0 of the product are still there in version 3. So how did they use this spare time? I'm afraid, it was used to produce new features justifying an upgrade to the latest version. That's the problem: some users always want more features while others give the priority to a more stable product. But fixing old bugs doesn't bring cash.

I have been an independent developer and sofware vendor for many years and I always tried to fix all the problems that were reported, provided I could reproduce them. For sure, I gave higher priority to the bugs that had the most impact on the user's experience but I fixed everything anyway. And I never asked money for fixing bugs : that is, I always tried to fix all reported bugs before releasing a new major version instead of telling the user that such or such fix would only appear in the next major release.

[Guest]

Samoreen wrote:

Hi,

tgutgu wrote:

I think also that spending more time on answering emails than on bug fixing is not what we want, but on the other hand we don't see the results that they spend more time on bug fixing.

Agreed. Given the big amount of time spared by not answering so many bug reports or messages in the forum, there should not be any long standing (or persisting) bug in LR. However, some bugs reported in version 1.0 of the product are still there in version 3. So how did they use this spare time? I'm afraid, it was used to produce new features justifying an upgrade to the latest version. That's the problem: some users always want more features while others give the priority to a more stable product. But fixing old bugs doesn't bring cash.

I have been an independent developer and sofware vendor for many years and I always tried to fix all the problems that were reported, provided I could reproduce them. For sure, I gave higher priority to the bugs that had the most impact on the user's experience but I fixed everything anyway. And I never asked money for fixing bugs : that is, I always tried to fix all reported bugs before releasing a new major version instead of telling the user that such or such fix would only appear in the next major release.

Well, I think we have to see both. To release a new version with new features (besides regularly adding new camera support) is critical to stay competitive or even to set new trends. At the same time, a product such as Lightroom has to keep a good maintenance level and communication about it. The latter two apparently get neglected quite a bit.

[TK2142]

Seán McCormack wrote:

Take a fruity computer company: much worse and very closed.

Pointing to something even worse is not a good way to justify behaviour.

It really only matters what the customers think of the company.

Take tgutgu, for example. It is my impression that he was enthusiastically defending Adobe when he joined this forum. I believe he would have recommended LR to other people. See how annoyed he is now? I guess he might add the words "at your own peril" to a recommendation now.

Speaking for myself, I've been very impressed with the LR 3 beta (my first contact with LR), of course with the assumption that the sluggishness will disappear once the real deal (advertised as a quantum leap in performance) has become available. Well, since then I learned that some performance problems existed since version 1.0 and have never been addressed since. LR 3.3. works OK-ish on my system but has nowhere near the performance it could have (e.g., with many local adjustments). I'd rate it as an enthusiast tool but would expect much more from software intended to be used by professionals. I cannot shake off the feeling that some things are left as they are to give people a reason to look into Photoshop ($$$). It appears that Adobe keeps the LR team cheap to run because they lack the competition from other vendors and earn the bigger $ with Photoshop. I know that these thoughts are not popular among some forum members here, but what are we supposed to think? That it is impossible to remove the many remaining problems in LR?

I believe there would be a lot less complaints about Adobe staff not showing their faces on the forum sufficiently often and/or bug reports apparently disappearing into black holes, if the "time saved by not dealing with customers" were really effectively used to kill bugs. I'd be busy recommending LR to other people instead of whining on the forum, if I had the feeling that the code quality were going somewhere. Look at 3.4RC. How is it possible with Matt Kloskowski calls "well-tested" that there is an obvious, easy to find *new* bug with keyboard control for the adjustment brush? Maybe it was introduced when they dealt with mouse handler code to make it less inefficient and that had an undesired impact? These things happen, but one tests before one releases, even if it is just a release candidate, right? IIRC, there was a similar issue with the HSL panel with LR 3.3RC, i.e., a new bug was introduced compared to 3.2.

I could be someone installing 3.4RC and providing feedback. I'd like to see what changed and I'd like to help improve LR. Past experience, however, made me reluctant to replace a slow but working version with one that has new bugs. So unless someone reports a performance improvement for an RC version, I'd rather sit it out and thus get spared from the most blatant new bugs.

P.S.: I do like LR a lot and the ACR code used is fantastic (maybe with the exceptions of local adjustments, if ACR has the same performance issues). I'm only frustrated because I'd like LR to be the product it could be if only the code quality were better.

[Jeff Schewe]

TK2142 wrote:

I cannot shake off the feeling that some things are left as they are to give people a reason to look into Photoshop ($$$). It appears that Adobe keeps the LR team cheap to run because they lack the competition from other vendors and earn the bigger $ with Photoshop. I know that these thoughts are not popular among some forum members here, but what are we supposed to think? That it is impossible to remove the many remaining problems in LR?

It would be useful if you understood the marketplace. As is stands, the market segment that represents "photographers" is a bit less than 10% of the total Photoshop user base. So, Adobe isn't spending too much time thinking of ways to maximize the buy into Photoshop by photographers. In the grand scheme of things, photographers really don't matter much to Photoshop sales (although traditionally, photographers tend to have a high degree of impact regarding features).

As far as the size of the team, Photoshop's is very large relative to Lightroom or Camera Raw (whose primary engineers can be measured on a single hand). Lightroom's engineering team is small when compared to Photoshop's. Yes...why is that? Well, part of it is the relative size of the marketplace. As much as photographers like to think of themselves as really important, they simply aren't. A recent market survey indicates there may be 70K+ "professional photographers" in the US. Compared to Photoshop's 4+ million users, photographers are a drop in the bucket. Sorry, the marketplace simply doesn't support an engineering staff to equal Photoshop's staff.

Part of Lightroom's underlying issues is the fact that LR is designed as a database driven application. Lightroom deals with thousands of images...Photoshop really is designed to deal with a single image (people who open multiple images in Photoshop often complain about Photoshop's performance because they don't understand how to work in Photoshop). So considering Lightroom's mandate, it actually does deal with thousands of image rather well.

Clearly, there is a limit to the parametric edits done to a single image while maintaining efficient speed. Adding hundreds of spot healing or 10+ Adjustments Brushes and multiple Gradient Filters is gonna slow Lightroom (or ACR) down. Sorry, that's the reality of parametric editing. If you are adding more than 10-20 spot healing spots and more than 3-4 local adjustments, you are trying to force parametric editing into the realm of pixel editing better (and more efficiently) done in Photoshop. Just because you CAN do something in LR/ACR doesn't mean it's smart to do so.

And no, I really don't think there's a conscience effort on Adobe's part to segment parametric/pixel editing and some sort of attempt at driving photographers to buy Photoshop in addition to Lightroom. If you have half a clue about the fundamentals, it's pretty simple; if you need to do a lot of specific local retouching and manipulations, you are better off doing so in Photoshop. If you need to adjust hundreds (or thousands) of images, you are better off doing it in ACR/LR. You use the tools provided for the purposes they are designed to do.

And just to be perfectly clears, the Adobe User to User Forums are designed to have users help users. It's simply ignorant to presume that the inclusion or lack of posting by actual Adobe reps on these forums means anything in particular. Some of the engineers and product managers do come into the forums (essentially on their own) to try to drill down on specific issues they may be working on. This has been the way the forums have been set up and designed to run–even the beta versions of the various forums over the years. If you don't like it, you are welcome to quit visiting the forums but the odds that your complaints will change things are very slight. Deal with the reality instead of baying at the moon. You'll be better off in the long run.

[dougryan]

Jeff Schewe wrote:

Part of Lightroom's underlying issues is the fact that LR is designed as a database driven application. Lightroom deals with thousands of images...Photoshop really is designed to deal with a single image (people who open multiple images in Photoshop often complain about Photoshop's performance because they don't understand how to work in Photoshop). So considering Lightroom's mandate, it actually does deal with thousands of image rather well.

Clearly, there is a limit to the parametric edits done to a single image while maintaining efficient speed. Adding hundreds of spot healing or 10+ Adjustments Brushes and multiple Gradient Filters is gonna slow Lightroom (or ACR) down. Sorry, that's the reality of parametric editing. If you are adding more than 10-20 spot healing spots and more than 3-4 local adjustments, you are trying to force parametric editing into the realm of pixel editing better (and more efficiently) done in Photoshop. Just because you CAN do something in LR/ACR doesn't mean it's smart to do so.

You made a sound and logical argument IMO, up until this part. While I agree that photographers make up a small part of the market for Adobe's consumer base, it would make very little business sense to develop and continue to support an independent project such as LR, if there wasn't a big enough market to earn a profit from it. You're making it seem as if there are literally just a couple of people working in an lab, addressing everything related to LR, which while I don't have definitive proof of, I'm sure is not true.

To address your statement above, you're right on one account. LR IS designed as a database driven app. But this isn't the underlying "issue"... The fact that Adobe took the time to develop this app as such, means that they had every intention of it running smoothly with thousands of images in tow. And I really don't get the logic of your saying that just because you can do something, it doesn't mean it's smart to do so?  If the Adobe engineers aren't thinking like a photographer when working on LR, then they have no business putting it on the market in the first place. I don't know how other photographers work, but I think that your example of x amount of healing adjustments and x amount of gradient filters is perhaps both exaggerated (would imagine that would be your intention to make a point) and irrelevant, given that most people here are likely experiencing slow downs with the absolute bare minimal of these adjustments. I know that I certainly am.

Adobe is filling a gap for us small time beans, and while we all appreciate having a tool such as LR to work with vs PS which is far more expensive and overkill for a photographer, I think we've also earned the right to expect a $300 product, to have better quality control invested in it from its developers. The only real competition at its level right now, is Aperture which is about a quarter the price of LR. And thus far, some of the real differences I'm seeing between the two, presently, is that Lightroom's sharpening and noise reduction algorithms are better. The latter of those two being much better IMO. But I'll tell you what... Aperture is a lot faster than it used to be, and has surpassed LR in terms of processing speed, regardless of what kind of tools I throw in the mix when processing (brushes etc. )

For whatever it's worth, I'm not giving up on LR until Adobe does. But the point is that if they continue down this path of adding new frivolous features just to stay "competitive", (which in and of its self is a joke since being competitive also means squashing bugs and keeping code tidy to ensure maximum efficiency)  then I'm afraid that LR will end up being where Aperture was not too long ago. And I don't know about other people, but the main reason I chose LR over Aperture when I began this trip is because of how much more efficient LR was. And now, that's all changing. I'm learning more and more about Aperture now, and enjoying the hell out of some of its features, but I'll maintain my library with LR.

I'm just glad that I know both applications. Just in case. Most people who have years invested in Adobe wouldn't even think about trying to change their workflow, and I can't say I blame them. But there comes a point when enough is enough I guess. I just hope I don't have to say that about Adobe and LR.

All that said: I'm sorry I didn't see what this thread was supposed to be for in the first place. I searched these forums because I am having issues, and in my search, I was brought here. But this thread was created so that we could provide feedback on the actual product, and not just vent or rant, since the last time that happened, the result was the creation of this thread. So again, sorry to the Adobe staff who meant for this thread to be purely for information purposes. The least I can do is give that feedback, and will do so when I gather more data.

I just want to make sure that I'm being accurate and take all conditions into account.

Doug

[Jeff Schewe]

dougryan wrote:

To address your statement above, you're right on one account. LR IS designed as a database driven app. But this isn't the underlying "issue"... The fact that Adobe took the time to develop this app as such, means that they had every intention of it running smoothly with thousands of images in tow. And I really don't get the logic of your saying that just because you can do something, it doesn't mean it's smart to do so?  If the Adobe engineers aren't thinking like a photographer when working on LR, then they have no business putting it on the market in the first place. I don't know how other photographers work, but I think that your example of x amount of healing adjustments and x amount of gradient filters is perhaps both exaggerated (would imagine that would be your intention to make a point) and irrelevant, given that most people here are likely experiencing slow downs with the absolute bare minimal of these adjustments. I know that I certainly am.

Not exaggerated, the slow downs are from personal experience when adding too many spot healing and Adjustments Brush adjustments...LR/ACR is NOT designed to be a "final retouching and editing environment" but to address 80/20 of the edits photographers usually need to make. As you add spot healing and Adjustment Brush/Gradient Filters up to a certain point you go through the threshold of what's reasonable in parametric vs pixel editing...

If you doubt my understanding of how and why Lightroom was developed, maybe you should do a little reading regarding Lightroom's initial development, read: THE SHADOWLAND/LIGHTROOM DEVELOPMENT STORY and see just how close Lightroom was to not being released...ironically, Mark Hamburg who was the founding engineer for Lightroom (AKA Shadowland) left Adobe and went to MSFT but came back to Adobe last year. It's not real clear what role Mark will have with future development of LR but he's at least back and involved (to a certain extent).

[dougryan]

I'm not saying that you're exaggerating your OWN experiences, rather that those specific values don't necessarily reflect the bulk of other people's experiences out there. I used myself as an example, stating that I don't even need to use that many spot healing adjustments, or gradient filters in order to have LR get all "hot and bothered" and start misbehaving.  Beyond that, I don't think it's reasonable for any of us to have to start guessing about what is a reasonable amount of parametric editing or not. I want to use the tool I bought for its intended purposes.

I don't doubt, nor do I care (no offense- that's not meant to sound nasty) about your understanding of how or why LR was developed, because it has nothing to do with the price of tea in China. The point NOW is that it WAS developed, and it has a fairly decent user base, and making money for Adobe. I hope you realize that not every photographer wants or needs to use PS as a "final" solution for retouching. LR is in fact, more than enough for anyone whom doesn't have to conform to the standards of a big name publisher. Sure, we all know that you can't get away with printing a fashion shot and such, without having it pass through to an editing professional... but for others whom don't have such a luxury, it's imperitive that they use the best tools they can afford. But I don't even think we even have to go that far up on the food chain in order to site examples of why LR is getting hammered.

Nowhere does Adobe make claims about LR which state such things as :  "Is really useful... up to a point"  or,  "please refrain from making TOO many adjustments, we didn't design LR to handle them".   But of course they wouldn't, because then, they wouldn't sell a damned thing.  But none of this is the point either, Jeff, because most of us have been harping on the fact that LR wasn't giving us these problems before V3. That is something you can't just make excuses for, no matter how much logic, or how many articles you throw at us. Everything you've said really, is rendered moot due to that very simple fact. LR was fantastic at 2.7. What happened ? I don't know. You don't know. Only the devs know, and they ain't talkin'. But that's ok, I'll live with it and hope that they'll figure it out and get it back to what it was. And that I suppose, is the real point of this thread... unless of course, it's all just to keep us preoccupied and distracted, while they figure out how to implement Yatzy! in 3.4

Doug

[JayS_In_CT]

Jeff Schewe wrote:

dougryan wrote:

To address your statement above, you're right on one account. LR IS designed as a database driven app. But this isn't the underlying "issue"... The fact that Adobe took the time to develop this app as such, means that they had every intention of it running smoothly with thousands of images in tow. And I really don't get the logic of your saying that just because you can do something, it doesn't mean it's smart to do so?  If the Adobe engineers aren't thinking like a photographer when working on LR, then they have no business putting it on the market in the first place. I don't know how other photographers work, but I think that your example of x amount of healing adjustments and x amount of gradient filters is perhaps both exaggerated (would imagine that would be your intention to make a point) and irrelevant, given that most people here are likely experiencing slow downs with the absolute bare minimal of these adjustments. I know that I certainly am.

Not exaggerated, the slow downs are from personal experience when adding too many spot healing and Adjustments Brush adjustments...LR/ACR is NOT designed to be a "final retouching and editing environment" but to address 80/20 of the edits photographers usually need to make. As you add spot healing and Adjustment Brush/Gradient Filters up to a certain point you go through the threshold of what's reasonable in parametric vs pixel editing...

If you doubt my understanding of how and why Lightroom was developed, maybe you should do a little reading regarding Lightroom's initial development, read: THE SHADOWLAND/LIGHTROOM DEVELOPMENT STORY and see just how close Lightroom was to not being released...ironically, Mark Hamburg who was the founding engineer for Lightroom (AKA Shadowland) left Adobe and went to MSFT but came back to Adobe last year. It's not real clear what role Mark will have with future development of LR but he's at least back and involved (to a certain extent).

Interesting history.. At what point did Raw Shooter Pro enter in..  I remember the day it was announced as we waited for LR 1..

Jay S.

[Jeff Schewe]

JayS In CT wrote:

Interesting history.. At what point did Raw Shooter Pro enter in..  I remember the day it was announced as we waited for LR 1..

Adobe bought Rawshooter in the summer of 2006. They bought it not so much to get the tech (although several features have been incorporated although rewritten from the ground up). But Adobe didn't really want the company Pixmantec so much as getting the main engineer Michael Johnson who ended up moving from Copenhagen to San Jose. Michael was a former engineer at Phase One working with their raw processing engine till leaving and starting Pixmantec.

Ironically, another Phase One engineer, Christian Poulsen left Phase One to make the Imacon scanner and somehow ended up as CEO of Hasselblad (although he is gone now).

[hillrg]

Jeff Schewe wrote:

JayS In CT wrote:

Interesting history.. At what point did Raw Shooter Pro enter in..  I remember the day it was announced as we waited for LR 1..

Adobe bought Rawshooter in the summer of 2006. They bought it not so much to get the tech (although several features have been incorporated although rewritten from the ground up). But Adobe didn't really want the company Pixmantec so much as getting the main engineer Michael Johnson who ended up moving from Copenhagen to San Jose. Michael was a former engineer at Phase One working with their raw processing engine till leaving and starting Pixmantec.

Ironically, another Phase One engineer, Christian Poulsen left Phase One to make the Imacon scanner and somehow ended up as CEO of Hasselblad (although he is gone now).

While I'm sure what Jeff says here is true I seriously doubt it was the main reason for the purchase.   At the time of purchase Rawshooter was polling around 30% market share, which adobe effectively bought.  If adobe was not looking at this (market share) then one would have to seriously question their business acumen.

However, Michael did not stick around that long working for adobe, which is unfortunate.

[Samoreen]

Hi,

Jeff Schewe wrote:

Deal with the reality instead of baying at the moon. You'll be better off in the long run.

Let me give you an idea about something "real": when I can bring LR3 to its knee on a rather powerful system just by creating a single local adjustment and when, oddly enough, the problem totally disappears by removing the capture date and time overlay from the display options, that's a real bug, not an inadequate use of the product.

So I don't know who is "baying at the moon" but it's time to consider fixing LR3 instead of telling the customers that they didn't understand how to use it.

Can't believe it.

[TK2142]

Jeff Schewe wrote:

As is stands, the market segment that represents "photographers" is a bit less than 10% of the total Photoshop user base. So, Adobe isn't spending too much time thinking of ways to maximize the buy into Photoshop by photographers.

How much is 10% of the total Photoshop user base in revenue? Maybe it is enough to care about? Also, your argument really works in favour of my statement. The fewer photographers use Photoshop, the more potential is there to turn them into Photoshop customers, right? Wouldn't it be great for Adobe if every LR customer also bought Photoshop?

Jeff Schewe wrote:

Sorry, the marketplace simply doesn't support an engineering staff to equal Photoshop's staff.

Not exactly news to Lightroom users, is it? The lack of performance and less than ideal quality control speaks for itself. It would have been a surprise to learn that a huge budget is behind this. The frustrating thing is that many aspects of LR are world-class but are let down by the sub-par programming in other areas. Your  "70K+ professional photographers" number sounds low but surely the world-wide sales including amateurs exceed this by a big margin.

Jeff Schewe wrote:

Part of Lightroom's underlying issues is the fact that LR is designed as a database driven application. Lightroom deals with thousands of images...

So what? The SQLite database used under the bonnet is capable of performing well with ten thousands of entries. Lightroom's problems are not caused by large catalogs.

Jeff Schewe wrote:

Clearly, there is a limit to the parametric edits done to a single image while maintaining efficient speed. Adding hundreds of spot healing or 10+ Adjustments Brushes and multiple Gradient Filters is gonna slow Lightroom (or ACR) down. Sorry, that's the reality of parametric editing.

That's just plain wrong. You've been told that enough times to take it on board. Of course a naive approach where you replay *all* editing actions for each new change won't scale. But that would be terrible programming. Simple caching techniques can address the issue you are alluding to. It is simply not the reality of parametric editing that it cannot scale with multiple edits. There is one point in time when there is no way to get around paying the price for excessive edits: When the final rendering has to be regenerated from scratch using the RAW file. Once that has happened (until that time you show the user a good enough preview that has been stored before) every further edit can build on the previous rendering. There is no good reason for editing to become progressively slower with more edits.

Jeff Schewe wrote:

And no, I really don't think there's a conscience effort on Adobe's part to segment parametric/pixel editing and some sort of attempt at driving photographers to buy Photoshop in addition to Lightroom.

Wouldn't it be good business thinking if they did? Why only sell LR to a customer if you can give them very good reasons to splurge considerably more money on Photoshop as well? If you are right then we must see better healing/cloning/retouching support in LR in version 4, right? Clearly many are driven to use Photoshop for this (and are guided this way by many evangelists who emphasise that almost every of their image edits involves a trip to Photoshop).To preempt your "Photoshop-like retouching is not possible with a parametric editing approach"-argument: It is possible. Just think about it like this: One could turn Photoshop into a completely non-destructive editing environment by recording every action and mouse movement the user makes. If the user discovers that they want to revert some decision they could start with the original image again, replaying all actions and mouse movements they wanted and leaving out others. That's basically (dumbed down) how LR works. Of course LR has to "push pixels" too. Of Photoshop is partially non-destructive too. Are you saying you can only have so many adjustment layers in Photoshop before performance goes down the drain and that Photoshop is not designed to deal with many adjustment layers? Destructive edits are great for saving the time to re-create the final rendering and saving space for previews that are otherwise needed but apart from that both Photoshop and LR can benefit from the same caching techniques during edits.

Implementing "lens correction" in the presence of local edit operations was much more of a performance challenge than providing better healing/cloning/retouching support is. So I don't buy the argument that technical reasons prevent LR to reach Photoshop levels of retouching. Yes, renderings will take longer in the presence of heavy editing operations but there is no reason why editing needs to become progressively slower.

Jeff Schewe wrote:

If you have half a clue about the fundamentals, it's pretty simple; if you need to do a lot of specific local retouching and manipulations, you are better off doing so in Photoshop.

I'm only better off doing it in Photoshop because LR is not fit for heavy editing at the moment, but that's just a reality and not proof of a technical impossibility.

Jeff Schewe wrote:

If you doubt my understanding of how and why Lightroom was
developed, maybe you should do a little reading regarding Lightroom's
initial development

I've read the "Shadowland" article and I listened to many of the George Jardine podcast interviews with Adobe staff and yourself. It made me wish that Thomas Knoll and Phil Clevenger were driving LR. I was surprised how overtly some questionable technical decisions by other people were discussed during these interviews.

[imajez]

Jeff Schewe wrote:
As far as the size of the team, Photoshop's is very large relative to Lightroom or Camera Raw (whose primary engineers can be measured on a single hand). Lightroom's engineering team is small when compared to Photoshop's. Yes...why is that? Well, part of it is the relative size of the marketplace. As much as photographers like to think of themselves as really important, they simply aren't. A recent market survey indicates there may be 70K+ "professional photographers" in the US. Compared to Photoshop's 4+ million users, photographers are a drop in the bucket. Sorry, the marketplace simply doesn't support an engineering staff to equal Photoshop's staff.

That's a dubious bit of logic and maths there. LR is not solely bought by Pros, just like most high end camera kit isn't bought buy pros. I've heard enough pros joke about how to tell the amateurs from pros - 'they're are the ones with the best/most expensive kit'.  Hasselbald and Phase would probably go under if they relied only on sales to just professional photographers.

Part of Lightroom's underlying issues is the fact that LR is designed as a database driven application. Lightroom deals with thousands of images...Photoshop really is designed to deal with a single image (people who open multiple images in Photoshop often complain about Photoshop's performance because they don't understand how to work in Photoshop). So considering Lightroom's mandate, it actually does deal with thousands of image rather well.

That's not really the issue. It's how LR chokes on individual files as well as crashing and general flailing around.

Clearly, there is a limit to the parametric edits done to a single image while maintaining efficient speed. Adding hundreds of spot healing or 10+ Adjustments Brushes and multiple Gradient Filters is gonna slow Lightroom (or ACR) down. Sorry, that's the reality of parametric editing. If you are adding more than 10-20 spot healing spots and more than 3-4 local adjustments, you are trying to force parametric editing into the realm of pixel editing better (and more efficiently) done in Photoshop. Just because you CAN do something in LR/ACR doesn't mean it's smart to do so.

Well that's a really dumb comment.

Here have this software with these great features, but don't use them as it cannot cope. A particularly daft thing to say as it contradicts your assertions that PS isn't really used by photographers that much anymore. So should all those people who buy LR also buy PS if they want to do local adjustments? Local adjustments are the specific feature that for most people makes PS redundant. Not to mention local adjustments in LR are actually much better than doing fiddly local adjustments in many ways than in PS. I love using PS, but LR simply frees me up to do the things PS is not as good at and PS then lets me do things LR is weaker at like compositing, layer blending etc.

Now what completely undermines your argument is that a file that LR really struggles with and can crash or lock up often programme, works just fine in ACR - which is the same rendering/development engine.

And no, I really don't think there's a conscience effort on Adobe's part to segment parametric/pixel editing and some sort of attempt at driving photographers to buy Photoshop in addition to Lightroom. If you have half a clue about the fundamentals, it's pretty simple; if you need to do a lot of specific local retouching and manipulations, you are better off doing so in Photoshop. If you need to adjust hundreds (or thousands) of images, you are better off doing it in ACR/LR. You use the tools provided for the purposes they are designed to do.

Yet as it happens local adjusting is something I tend to do in LR more than in PS nowadays, as a lot of the time as it's easier than faffing around with masking and selecting. LR is in fact designed for local adjusting if it has tools that can do local adjusting, wouldn't you say? And it's by people pushing programmes capabilities that drive development forwards. PS would not be the dominant programme it is today if  people hadn't pushed the boundaries of its capability.

And just to be perfectly clears, the Adobe User to User Forums are designed to have users help users. It's simply ignorant to presume that the inclusion or lack of posting by actual Adobe reps on these forums means anything in particular. Some of the engineers and product managers do come into the forums (essentially on their own) to try to drill down on specific issues they may be working on. This has been the way the forums have been set up and designed to run–even the beta versions of the various forums over the years. If you don't like it, you are welcome to quit visiting the forums but the odds that your complaints will change things are very slight. Deal with the reality instead of baying at the moon. You'll be better off in the long run.

So you're effectively saying these forums are a waste of time? So why not suggest to Adobe that they close them and save some money, if they are indeed so pointless.

[thewhitedog]

@ Seán: Whatever Adobe's policies may be, the fact remains that Lightroom 3.4 RC not only did not fix the performance issues so many users reported here, it introduced a serious bug that makes the program unusable for me. Should I not be upset over this state of affairs, after so long trying to maintain a reasonable tone in my posts?

I have no way of knowing, but it could be that this bug is only in the Mac version of Lightroom, which would explain why few others have reported experiencing the problem. Like the folks at Adobe, though, you haven't seen fit to address to meat of the issue but rather have diverted attention away from the substance by blaming the victim, in this instance, me.

It's possible, Seán, that you don't use the up and down arrow keys to adjust settings in the Adjustment Brush and the Graduated Filter (or that you use Windows and the bug doesn't exist there), but the loss of this functionality, which is generally available throughout the rest of the program, is hardly, in my opinion, a minor oversight. I dare say, though, that if you did use these features you would be as upset as I am that Adobe broke them. It's easy to remain calm and detached when the problems others have don't affect you. Try doing so when your own ox is being gored.

As I reread the list of RC fixes, there is mention of the up/down arrow keys using improper increments in the Graduated Filter (it didn't for me so I'm not clear what they are talking about) so it may be that in trying to fix that increment bug they broke the whole thing. Not that it makes them look any better for doing so, but this could be the origin of the problem I'm having.

The real issue isn't Adobe's policies about posting to user forums and responding to bug reports. That's a side show and a distraction. The real issue is that the fixes they made in Lightroom 3.4 RC are distinctly minor and relevant only to a small minority of users. At the same time they left the big problems regarding performance unresolved. And, to add insult to injury, they broke a primary feature of the application. The question isn't about policies, it's about performance. And Lightroom 3.4 RC does not perform. For months I've tried to give the Adobe engineers the benefit of the doubt regarding their efforts and their intentions, not to mention their fundamental competence. Lightroom 3.4 RC calls all those assumptions into question and raises doubts I'm no longer willing to hide. It's one thing if they don't have time to monitor and post informative replies to this forum. It's quite another when the fruits of their labor prove to be inadequate, as is clearly evident in the 3.4 RC.

Regardless of my personal issues, I'm sure I'm not the only one disappointed that performance problems were not substantively addressed in this RC. And if the Lightroom team is indeed understaffed, as some suppose, then the only way to get Adobe to pay more attention to the project and staff it up is for users to raise a stink. If we stop recommending Lightroom to our friends and associates, that will hit Adobe where it really hurts, in the bottom line. Adobe has traditionally relied on its user base to evangelize its products. They can ill afford for that base to turn sour on Lightroom.

[Samoreen]

Hi,

thewhitedog wrote:

If we stop recommending Lightroom to our friends and associates, that will hit Adobe where it really hurts, in the bottom line. Adobe has traditionally relied on its user base to evangelize its products. They can ill afford for that base to turn sour on Lightroom.

I read again the statements made above by Jeff Schewe : they obviously don't care that much. I have exchanged informations with a member of the development staff and honestly, I can't say that they are not trying to solve the performance problems and fix the related bugs. However, they are doing what they can or what they are allowed to do. Which again leads the discussion to what the company is ready to do (that is, invest) to have these bugs fixed. Jeff Schewe statements give us an idea.

When I was a Microsoft MVP, I reported bugs in some of their development tools that they could not reproduce. Since others reported the same bugs/issues, they sent an engineer to me and we spent some time on my systems working together on these problems. This helped a lot solve these issues. So the problem is not what the LR developers can do but what Adobe are wanting to do.

[imajez]

Samoreen wrote:

I read again the statements made above by Jeff Schewe : they obviously don't care that much.

Jeff has said on several occasions he is not an Adobe employee, though he certainly does advise and give them useful feedback. And when he is not letting off steam and being somewhat tactless, he can be extremely helpful at times to users of this and other forums.

[Samoreen]

HI,

imajez wrote:

Jeff has said on several occasions he is not an Adobe employee, though he certainly does advise and give them useful feedback. And when he is not letting off steam and being somewhat tactless, he can be extremely helpful at times to users of this and other forums.

Correct. Let's say he has tight relations with them.