Luminous Landscape Forum
Raw & Post Processing, Printing => Adobe Lightroom Q&A => Topic started by: dreed on January 17, 2012, 06:11:27 am
-
The Windows download of Lightroom 3.5 was about 223MB for both 32bit and 64bit versions.
The Windows download of Lightroom 4 beta is a 411MB .zip file for only the 64bit version.
So I factor that in as LR4 is 4 times the size of LR3.5.
What was that new feature list again?
-
I noticed that too but will reserve judgement until Final is released.
At this point, I'm sure the code is not optimized and it may have a bunch of debug code or symbols in it that would cause the bloat.
-
I don't care about 200MB more or less. After all, that's just about 12 raw files from my camera.
-
Do you have the numbers for LR3 Beta? It might have been a good bit larger than the final optimized version. I won't panic until I see the size of the production version of LR4.
Eric
-
Who cares about the size of the application DL? I do care about the efficiency/speed of running the program and there is no direct correlation.
-
It's a beta build so I would expect it to have more debugging info built in. All this stuff is usually striped on final builds.
-
Refigure......LR4 was only 1% zipped. 424 megs opened.
-
For those who have been into pc's for some time there is an element of "lean and mean" or rather efficient code.
I fear LR has probably gone a bit beyond what I'd like it to do (into areas less important) thus has grown in size.
The install size isn't a massive concern, however the running performance of LR has never been as optimal as it IMO should have been
-
The install size isn't a massive concern, however the running performance of LR has never been as optimal as it IMO should have been
Absolutely.
-
Featuritis causes bloat - and we can only blame ourselves for that, with all the endless feature requests on this very board.
-
there is the old saying 'all applications will expand until they have email functionality' (or some paraphrase close to that).
I keep on reading these feature requests and people are not going to be happy until LR becomes Photoshop (or at least just the parts we tend to use). Ain't gonna happen. LR is NOT PS.
I suspect a chunk of the performance problem also lies w/ the choice of implementation language (Lua).
-
I fully agree. On the other hand I dont see why Lightroom should support Blurb and GPS information. There are much better ways. InDesign from Adobe is a powerfull tool and supports Blurb's PDF to Book procedure. Geosetter is a powerfull tool to handle GPS informations and is free. I do not see what path Adobe is following with Lightroom 4. A bit of everything ? In addition it will not work under Windows XP. Frankly speeeking, I'm not impressed.
-
Geosetter is a powerfull tool to handle GPS informations and is free.
Doesn't work on Macs, though; and the Mac market, fortunately for many of us, is within Adobe's contemplation, even if not yours.
Jeremy
-
You can't please all of the people all of the time. To make LR more attractive they have to add features. They won't take any out for sure so it will become more bloated as more versions are created. I agree that some want it to be akin to PS but not at the same price? If you have a computer with good specs it isn't a problem but I fear some have older computers that can't run it efficiently?
-
How true ! I'm using now Lightroom 3.x and what impressed me most is the fast workflow from camera to disk and the nice developing module good enough for most of the pictures. Camera Raw and PS is used for the more advanced stuff. Adding more and more moduls is not going to be any help for me and fortunately I have no need to change and can save the cost for upgrading.
-
How true ! I'm using now Lightroom 3.x and what impressed me most is the fast workflow from camera to disk and the nice developing module good enough for most of the pictures. Camera Raw and PS is used for the more advanced stuff. Adding more and more moduls is not going to be any help for me and fortunately I have no need to change and can save the cost for upgrading.
Too bad we can't have a series of check boxes for which modules we might want installed. ;) I don't have any interest in video processing.
-
The difference in size may be due to the increase in features, the % of compression for the file or even maybe the fact that's still a beta.
-
Too bad we can't have a series of check boxes for which modules we might want installed. ;) I don't have any interest in video processing.
I agree.
But I can well imagine at some future date that there will be a whole set of Lightroom programs, such as LR Basic, LR Video, LR Web, LR Interior Design, etc., that can all be purchased together as "LightRoom Suite."
Eric
-
Use what you need and don't worry about what you don't need.
I think we should worry about feature bloat.
Adding unwanted features can slow the whole program down, make it unnecessarily more complex to use and that it can restrict it's usefulness(eg not being able to use LR4 on XP because of the video features).
-
I'm not so crazy about the basic video editing, although to be able to color correct the video is a nice thing, but for video editing I guess a lot of us go to another software, the geolocation that's a good one.
-
But, without new features there would be no need to update from LR1 to LR4.
In reality, one persons bloat is another persons must have new feature.
I suspect that many of us who now see a particular feature as not necessary, will discover that it is really useful when they actually use it.
Les
-
I fully agree. On the other hand I dont see why Lightroom should support Blurb and GPS information. There are much better ways. InDesign from Adobe is a powerfull tool and supports Blurb's PDF to Book procedure. Geosetter is a powerfull tool to handle GPS informations and is free. I do not see what path Adobe is following with Lightroom 4. A bit of everything ? In addition it will not work under Windows XP. Frankly speeeking, I'm not impressed.
GPS integration is a very welcome addition. Geosetter is an extremely clunky implementation.
Blurb integration is a tool for Adobe to get a cut of Blurb's revenue (or maybe Blurb just paid for the tab), and for Blurb to expand their potential customer base. If people actually use it, expect to see more such tie-ins in Adobe's future products. Fortunately it is a good feature with benefits for photographers, unfortunately it's just one service provider out of many.
LR will continue to bloat thanks to the never-ending feature requests here and elsewhere, and for Adobe's necessity to roll out updates periodically ($$$).
-
Too bad we can't have a series of check boxes for which modules we might want installed. ;) I don't have any interest in video processing.
Once the software is developed, the cost for shipping you the megabytes is close to 0 for everyone.
While feature bloat can cause a cluttered user interface and a strain on development time/money/people, I think that adding video editing to Lightroom is not what make its still-image development slower.
-h
-
So I'm assuming you are the final voice as to what should and should not be included? See that's the issue. It's not about you or me individually, it's not about what you want or don't want in the software, it's what works well for the total user base.
I don't see my opinion as the final word, but it's just as important as anyone else's.
Maybe my opinion IS more important to Adobe than some of the other voices. If Adobe want to keep their customers upgrading they need to listen to the existing customers that aren't upgrading because they're the people that can build their business on, or loose it. The fanbois that just stump up the cash for every upgrade regardless are cash cows they don't need to worry about.
I guess the most amazing thing Microsoft ever did was create an operating system that lasted a decade. Perhaps that's why they simply don't inspire innovation anymore.
Curious how maturity and stability isn't valued in software. No one suggests throwing away Leicas because they're old.
I can only assume that the voices of LR4 frustration stem from amateurs and hobbyists,
I'm sure you're correct, as I'm also sure that they are the largest sector of users of LR.
Ask for features you would benefit from and you may receive if enough people also require such a feature, but simply saying it's bloated is of no value to the development team.
I've contributed to official feature requests on each beta and release. It's nice to see some requests being implemented eg book creation, soft proofing etc. However it's frustrating when those features then arrive incomplete, eg not being able to soft proof for book creation or not having custom layout capability like Blurb's free software has.
It's also worth pointing out when some additions to the program aren't wanted. They need to know where to stop adding unwanted features too.
-
I suppose the key here is that it's not the download size that is the problem but rather the impact on the installed software.
It would be nice if the Lightroom installer let you select which parts of the package you wanted so that you could (for example) exclude the video stuff from installation.
I suppose the question boils down to is the increase in disk space significant?
Given that the installation of the application is typically on the same partition as that used for caching images (and the library), the more space that the application takes up the less space it will have to use for data.
-
I suppose the key here is that it's not the download size that is the problem but rather the impact on the installed software.
It would be nice if the Lightroom installer let you select which parts of the package you wanted so that you could (for example) exclude the video stuff from installation.
I suppose the question boils down to is the increase in disk space significant?
Given that the installation of the application is typically on the same partition as that used for caching images (and the library), the more space that the application takes up the less space it will have to use for data.
400MB (if that is installed size) is 20 raw files from my DSLR. I have 30000 images in my library. My guess is that those 400MB is mostly image files used in the GUI/help and possible sound files. Adding or removing a function in itself may not impact filesize that much (you can do _a lot_ in 100KB of pure functional code)
Actually, I have my windows, lightroom and image cache on a 128GB SSD, while image library is on a spinning disk. Even then, 0.5GB more or less is insignificant.
If these new functions don't get in my way, allows Adobe to sell more copies, be more profitable and secure the funds in order to continue development of core still-image quality enhancements, I have no problem with it.
-h
-
I'm sure you're correct, as I'm also sure that they are the largest sector of users of LR.
Amateurs (like me) may be the largest sector of users, but not the largest sector of customers, if you know what I mean. ::)
Adobe don't have to listen to non-upgrading users but to non-upgrading customers.
I'll upgrade Lr 99% for every and each version; not because I'm a 'fanboy', but because it's probably the software I use the most on my machine.
I wan't it to be as up to date as possible.
An OS primitive must be fast and stable: a digital photography workflow assistant should be cutting edge.
Anyway, FWIW, I -do- like the new features and I'm not a photo professional. So what?
As others told better before me, what may not work for you could be a huge time saver for others.
I don't care at all the MB count of the binaries. 200MB or 600MB means nothing to me. 20GB would be critical. My attention lies somewhere else.
My SSD drive (which is meant for Apps and other important files) can accomodate way more than that :)
-
It's also worth pointing out when some additions to the program aren't wanted. They need to know where to stop adding unwanted features too.
I seriously doubt whether, given their competing priorities, Adobe ever have added, or ever will add, a feature that isn't wanted. Unwanted by you, sure, unwanted by others on a thread, probably, but they do their homework in ways that you or I never could (unless you happen to own a consumer research business) and wouldn't waste their time on features that aren't wanted. Video is a case in point. I have read, over the years, lots of old-fart (or young-fart) harrumphing on forums from people who never use it. Meanwhile, maybe, out there, plenty of people who use cameras that take quality stills and videos and want to manage the assets together.
-
Not much of interest in the size growth :
227MB = 2064 JPG files
86.5MB = 777 DCP (camera profiles)
177MB = 617 LCP (lens profiles)
197MB = support subdirectory (optionally used files)
EXE and DLL files in Lightroom directory itself only grown from 58 files at 51MB to 88 files at 100MB.
Not bad for the new features and performance of the software seems comparable or better to 3.6 (at a first glance, no measurements).
-
Some of you need to get out and shoot more.
-
But, without new features there would be no need to update from LR1 to LR4.
I beg to differ. This view does not take into account the significant improvements in the processing engine from LR1 to LR4. Those improvements alone almost make a good case for upgrading on their own (IMHO of course).
I'm very glad to get these improvements in core functionality as they directly contribute to higher quality output. As for additional features, I don't care about some of the new things that were added, but I have no problem ignoring them as I don't think they impact the usability/performance of the program noticeably (final proof being the final release of LR4, of course).
Furthermore, I don't think that all of the new stuff are 'additional features' but are fleshing out of functionality already present in the very first LR version. Case in point being soft proofing, nicely complementing and enhancing the printing functionality of LR. These kind of things are certainly most welcome.
Christian
-
A lot of the new size comes from libraries needed to support the new functionality (like books, video, etc.). But like others have said, its footprint is probably tiny compared to your image collection ...
-
I think we should worry about feature bloat.
Adding unwanted features can slow the whole program down, make it unnecessarily more complex to use and that it can restrict it's usefulness(eg not being able to use LR4 on XP because of the video features).
Bloat:Definition - Things added to a programme that the complainer does not want or need. The fact that these features may be incredibly important to many other users is never considered or taken into account.
BTW I've been asking for video inclusion since LR first appeared as having to manage my digitals assets in multiple programmes undermines the point of the Digital Asset Manager that LR purports to be.
Besides software is not necessarily more complex if it adds more abilities that work in the way you would expect from the programme. And adds zero complexity if you don't need to use them. ;)
Having said that LR's faffy way of processing video is well faffy. There's a sync feature in LR Dev module to match images that would do exactly what you need for applying a look to your video files from the sample jpeg, without the daft making presets fudge that is currently being used and this would allow for a much faster workflow too.
-
LR's faffy way of processing video is well faffy.
Define "faffy"...is that the same as fuggly? Or FUBAR? (I'm assuming it's not a positive adjective :~)
-
Hi,
I like to add some short videos to my slide shows, but going to editor, create new project, edit, share for a 5 second clip is to much mess. So I really like that feature in LR4.
The problem is that whatever the feature is, there is someone out there who would not live without it.
Best regards
Erik
I'm not so crazy about the basic video editing, although to be able to color correct the video is a nice thing, but for video editing I guess a lot of us go to another software, the geolocation that's a good one.
-
Would it not be better if things like video was an optional plug in?
-
Would it not be better if things like video was an optional plug in?
;
In LR4 it is optional. You need never see the module again. Adobe is taking care of those who want it, and of those who don't. .
-
I keep on examining these function needs and individuals are not going to be satisfied until LR becomes Illustrator (or at least just the components we usually use). Ain't want to occur. LR is NOT PS.
-
In LR3.5 I already get rid of features I don't want. I just remove the Web and Slideshow modules from the application folder, and they don't load.
So I have Library - Develop - Print
That's all I want, need and use. Even with my big Hass 3FR files it runs briskly and gets the job done.
John
-
Define "faffy"...is that the same as fuggly? Or FUBAR? (I'm assuming it's not a positive adjective :~)
My wife uses the word "faff" regularly. In her vocabulary it means "to fiddle with something in an unnecessarily complicated, unproductive and annoying way". As in "stop faffing around with that camera and start cooking the dinner before we all die of starvation" .
-
Define "faffy"...is that the same as fuggly? Or FUBAR? (I'm assuming it's not a positive adjective :~)
You Americans, not understanding the Queen'sEnglish! ;-)
Faff verb - (UK) To waste time on an unproductive activity.
"I decided to stop faffing about and get some work done."
Faff - adj - (UK) fiddly, fussy.
"He was able to get the job done, but it was a bit of a faff."
Actually as it turns out LR4 isn't a such a faff to use when syncing videos, as you can indeed sync in the dev module as I thought you should be able to. It's just all the instructions I've seen on how to use LR4 videos chose to use a faffy workaround fudge of way of doing things.
-
Faff - adj - (UK) fiddly, fussy.
"He was able to get the job done, but it was a bit of a faff."
Ah, hmmmm . . .
I would point out that this is not an adjectival usage of faff, but demonstrates its use as a noun.
Just to set the record straight ;)
John
-
The final LR4 download for Windows now weighs in at 718MB.
This is three times the size of 3.5.
What happened there?
-
It's both 32 and 64 bit versions and it has a extra modules and a lot more functionality.
-
For me the most important changes from LR3.6 are in the develop module rather than in adding new features, the bulk of which I don't care about.
Adding the healing tools from PS would be nice.
It's far from clear that the existence of other modules slows anything down.
Disk space is so cheap these days it's hard to see why anyone cares about file size.
-
For me the most important changes from LR3.6 are in the develop module rather than in adding new features, the bulk of which I don't care about.
Adding the healing tools from PS would be nice.
It's far from clear that the existence of other modules slows anything down.
Disk space is so cheap these days it's hard to see why anyone cares about file size.
There is some (probably irrational) perception about filesize. When my HP printer tells me to download a 120 MB driver in order to print a document, my initial thought is "incompetent developers and/or bloatware", and it turns out that I am right on both accounts. When the foobar2000 sound player offers playback of all kinds of sound files in 3 MB, my gut-feeling is that those guys must be making a solid application, and it turns out that I am right.
-h
-
I have no problem with bloat as long as it doesn't affect anything else, but the problem is, it always does.
Always. There are no exceptions.
I'm a still photographer, and LR started out as a way to post-process and file still photographs. They sold me that idea, and I bought into it, and now I have a large LR database that would be a little complicated to get out of. Now I have to pay for video capability and book making and other stuff. I have no use for it, but I have to pay for it to keep my database going in a current version. So why not just skip the new version? Because I do want the updates and increased functionality of the product that was originally sold to me -- that is, better data-base functionality, better ACR, more camera coverage, better still-photo post processing. But to get that, I have to pay for all these other new functions that originally were not part of the program.
And that's the difference between updates and bloat. Updates make the old program better. Bloat adds stuff that a lot of the original buyers don't want, but have to pay for.
I think Adobe would have done better is they had a core LR program with lots of optional plugins -- not for small features, like a plug-in for sharpening, but the big ones, like video processing or book production. Sell them separately.
-
And that's the difference between updates and bloat. Updates make the old program better. Bloat adds stuff that a lot of the original buyers don't want, but have to pay for.
All that bloat has resulted in LR4 costing half of what they charged for LR3
-
John - the price has been reduced. I don't think you have too much to complain about paying for the bloat - remember, the revenue generated by sales to people who do want those features are also funding upgrades to the features that you care about.
-
The final LR4 download for Windows now weighs in at 718MB.
This is three times the size of 3.5.
What happened there?
Don't complain: the Mac version is 909MB vs 101MB for LR3.6! What's a little more troubling is that it seems to be taking more than three times as much RAM doing the same tasks.
-
Using or allocating? Are you running out? Is it slower?
-
I have no problem with bloat as long as it doesn't affect anything else, but the problem is, it always does.
Always. There are no exceptions.
Sure there are exceptions. If version 2 of a product is delivered with 1GB of PDF documentation, while version 1 is not, you might say that the distributable file is bloated. It does not affect the gui experience, does not make the application run any slower, and can in fact safely be deleted by the user so as to reclaim hd space.
I think that the correct thing to say is that "bloat" affects anything else in subtle ways that are very hard to estimate by end-users. It might make that application slower, more buggy, etc, to a large degree, to a small degree or to exactly zero degree.
-h
-
Using or allocating? Are you running out? Is it slower?
For me it's using more memory, but I'm not running out, and it is not slower. Not a problem now, could be if it increases proportionally to the way version 3 can. In the very limited testing I've done, 3.6 was using up to about 800MB while 4.0 was up to 2.4GB. I've seen version 3 use over 2GB, if 4 went to 6GB that might be a problem. That said, I'm really liking how responsive version 4 seems, so it seems that perhaps the increased memory usage is being put to good use.
-
Ah, hmmmm . . .
I would point out that this is not an adjectival usage of faff, but demonstrates its use as a noun.
Just to set the record straight ;)
John
The usage in this phrase seems adjectival to me, especially in the context of an idiomatic expression. That is, "bit of faff"= fussy, fiddly or hard. :-/
-
Ah, David, but you have misquoted. I was referring to "a bit of a faff".
When we use 'a' in front of something, that something is usually a noun.
And no, fellow Forum members, this does not have much relevance to Lightroom or the rest of this thread. My apologies.
John
-
Hi folks, with Lr 4 we have some more advanced caching systems in place (esp. in Develop) which can increase memory usage some ... on the other hand, you should experience overall improved responsiveness. This is of course a delicate balance and tradeoff.
-
Ah, David, but you have misquoted. I was referring to "a bit of a faff".
When we use 'a' in front of something, that something is usually a noun.
And no, fellow Forum members, this does not have much relevance to Lightroom or the rest of this thread. My apologies.
John
Let's start a rumor that version 4.1 of LR will have a de-faffing slider. :D
-
you should experience overall improved responsiveness.
Compared to what ? the beta ?
I don't think I've read one comment anywhere yet saying responsiveness of LR4 is better than LR3, but I've read loads of complaints about the sluggishness of the UI in LR4. Adobe have some serious code optimisation to do before Lightroom becomes a slick tool again, assuming that's possible with the new process at all.
-
Ah, hmmmm . . .
I would point out that this is not an adjectival usage of faff, but demonstrates its use as a noun.
Just to set the record straight ;)
John
Duh! Of course it is. Not sure how that got through quality control. :o
-
I'm a still photographer, and LR started out as a way to post-process and file still photographs.
They sold me that idea, and I bought into it, and now I have a large LR database that would be a little complicated to get out of.
No, LR was designed as a tool for photographers. And astonishingly other photographers have different requirements from you. :o
LR 1.0 also came with web, slideshow and print modules, not just a file and process application.
You could always use Bridge and ACR.
Now I have to pay for video capability and book making and other stuff. I have no use for it, but I have to pay for it to keep my database going in a current version.
Except this upgrade is less than previous upgrades.
So why not just skip the new version? Because I do want the updates and increased functionality of the product that was originally sold to me -- that is, better data-base functionality, better ACR, more camera coverage, better still-photo post processing. But to get that, I have to pay for all these other new functions that originally were not part of the program.
And that's the difference between updates and bloat. Updates make the old program better. Bloat adds stuff that a lot of the original buyers don't want, but have to pay for.
Translation of bloat - "Stuff I do not need." Other people other than yourself use LR and what they require from LR is not the same as you. So chill out and do not begrudge others who now have the tools they need.
I think Adobe would have done better is they had a core LR program with lots of optional plugins -- not for small features, like a plug-in for sharpening, but the big ones, like video processing or book production. Sell them separately.
Although I suggested a similar idea many years back for making LR more adaptable in a slightly different context, seeing as the upgrade price is only £50, why even bother at such low prices. And as suggested above, if you do not need certain module, just move them out of LR folder.
-
Compared to what ? the beta ?
I don't think I've read one comment anywhere yet saying responsiveness of LR4 is better than LR3, but I've read loads of complaints about the sluggishness of the UI in LR4. Adobe have some serious code optimisation to do before Lightroom becomes a slick tool again, assuming that's possible with the new process at all.
That's because people rarely bother to come to the web and tell everyone how happy they are or how something is not a problem.
I'm not finding any performance issues with LR4 on a relatively old machine (3 year old PC). I'm not a heavy user with massive files and my system is nicely optimised, but I'd be surprised that there was really any particularly performance hit for anyone based on my experience.
-
That's because people rarely bother to come to the web and tell everyone how happy they are or how something is not a problem.
Not quite true, there are enough people posting "awesome". I've posted already how wll some aspects have worked too.
I'm not finding any performance issues with LR4 on a relatively old machine (3 year old PC). I'm not a heavy user with massive files.....
Well you're either lucky, or you haven't done enough with it to spot the problems yet. No one seems to be posting 'wow it runs so much faster than LR3' messages.
What's disappointing is how little speed and responsiveness is valued in software design generally these days. It's all features, features, features; it runs too slow ? just buy a faster computer.
Adobe can do it. The Mercury playback engine in Premiere Pro 5.5 is a fine example.
-
Hi folks, with Lr 4 we have some more advanced caching systems in place (esp. in Develop) which can increase memory usage some ... on the other hand, you should experience overall improved responsiveness. This is of course a delicate balance and tradeoff.
Is it related to the new option under DNG convert settings to store that extra data or is the speed improvement there anyway when working with proprietary raw files?
-
I'm not finding any performance issues with LR4 on a relatively old machine (3 year old PC). I'm not a heavy user with massive files.....
Well you're either lucky, or you haven't done enough with it to spot the problems yet. No one seems to be posting 'wow it runs so much faster than LR3' messages.
What's disappointing is how little speed and responsiveness is valued in software design generally these days. It's all features, features, features; it runs too slow ? just buy a faster computer.
Adobe can do it. The Mercury playback engine in Premiere Pro 5.5 is a fine example.
I don't think it runs faster, overall, but I'm not hitting any sudden bottlenecks compared to LR3. And the idea that software take advantage of faster hardware is not unreasonable and it's wrong to suggest that it's just badly written or bloated etc because they don't have to optimise due to new hardware.
I also suspect that a point release will probably see some further refinement - that's usually the way.
-
What's disappointing is how little speed and responsiveness is valued in software design generally these days. It's all features, features, features; it runs too slow ? just buy a faster computer.
Adobe can do it. The Mercury playback engine in Premiere Pro 5.5 is a fine example.
How much image quality loss are you willing to see for a 2x speed improvement? I am sure that demosaic, noise reduction, tonemapping etc can be done in a simpler, faster fashion.
-h
-
How much image quality loss are you willing to see for a 2x speed improvement? I am sure that demosaic, noise reduction, tonemapping etc can be done in a simpler, faster fashion.
And I'm sure that if that were indeed possible, Adobe would have done just that. As being able to advertise a doubling in speed would be a great selling point.
The last two versions of LR introduces much more powerful developing features, this sort of thing tends to need more grunt power. Maybe in the next version the process engine will not be altered and instead it will be sped up.
-
What's disappointing is how little speed and responsiveness is valued in software design generally these days. It's all features, features, features; it runs too slow ? just buy a faster computer.
Adobe can do it. The Mercury playback engine in Premiere Pro 5.5 is a fine example.
Except it is just the opposite, that's a hardware solution, one only available to PC users that use specific video cards. (http://blogs.nvidia.com/2010/02/nvidia-quadro-driving-adobes-new-mercury-playback-engine/)
-
And I'm sure that if that were indeed possible, Adobe would have done just that. As being able to advertise a doubling in speed would be a great selling point.
I am confident that it is possible to do raw development faster than Lightroom by sacrificing image quality, and that Adobe would sell less if they did.
-h
-
And I'm sure that if that were indeed possible, Adobe would have done just that.
I'm afraid that's a rather naive point of view.
In the past we've seen early versions of LR heavily criticised for performance issues and fixes for them have been rolled out later. The issues do get eventually fixed and resolved, but they don't help sell to the initial upgraders.
I strongly suspect that the marketing side of Adobe put dead lines for release that are absolutely adhered to pretty much regardless of the state of the software. I'd guess the deadlines are for the release of the public beta and a fixed release date that would only be missed if a very major problem was uncovered with the beta program. After it's released there's the frantic period of fixing all the bugs that have come to light and trying to make good all the features that they planned, but didn't have time to add to the .0 version. An example of that might be how no other book publisher's settings made it into 4.0, but will arrive in a future update.
It must be grim to be an Adobe software engineer knowing a product is being released before it's fully sorted and not reaching it's full potential, but of course no can ever admit that publicly.
-
How much image quality loss are you willing to see for a 2x speed improvement?
There's always a balancing act between quality and speed.
The problem at the moment is that the sluggishness of the UI in develop makes it too awkward to get the absolute best from the software.
There have been many great imaging software products that have failed to achieve their potential because of poor interfaces and usability issues, despite ultimately superior results.
Lightroom has a brilliant UI and workflow, but when it becomes too unresponsive those UI benefits become lost.
-
Bear in mind that each successive version of Lightroom will have been developed on the assumption that users will be running it on progressively higher-resource machines. (Forget the laughable "minimum system requirements" that Adobe quote for all their products - that is merely a sales ploy).
When I upgraded from LR2 to LR3 I was, at that time, using an old PC with only a dual core 2.8GHz processor and 4Gb of RAM (only 3.2Gb of which could be used by 32-bit Windows XP). Speed was just tolerable although, if I was running Lightroom, CS5 and, say, Color Efex Pro together, it became decidedly creaky.
Just before Christmas I upgraded my PC to a quad core i7 processor and 16Gb of RAM running Windows 7 in 64-bit version. At the same time I upgraded LR to the 64-bit version 3.6. The increase in speed was fantastic.
Now that I have upgraded LR to v.4, I notice no reduction in speed compared to v.3.6 but I hate to think what it might have been like on my old obsolete machine.
It really is horses for courses. You can't expect to run 2012 software on, say, a 2010 computer and get reasonable results.
I guess the real test, for me, will be in the next few weeks when I start processing 70Mb Raw files from the Nikon D800
-
Bear in mind that each successive version of Lightroom will have been developed on the assumption that users will be running it on progressively higher-resource machines. (Forget the laughable "minimum system requirements" that Adobe quote for all their products - that is merely a sales ploy).
Not a sales ploy...it's simply the lowest hardware that Adobe tests against and officially supports. More is always better.
So, in my experience, here's the breakdown of LR bottlenecks...it's similar to Photoshop but importantly different.
Top of the ladder is processor speed...
The second is multi cores (although that tops out quickly, 2 cores is almost 2x better than single core but 4 cores is not nearly 4x, more like 2.5-3x at best).
The third is 32 vs 64 bit processing which 64 bit being at least 20-30 % better depending on ram. If you are under 4 gigs then the difference is more like 10-15%.
Ram...but past a certain point it doesn't matter. This is really where Photoshop and Lightroom are fundamentally different. As long as you have 8-16 gigs of ram on a 64-bit machine, that's the best you can expect for Lightroom. Adding more won't help. But regarding Photoshop on 64 bit machines, more ram is indeed important. I have 32 gigs and wish for 64 gigs on my main workstation...particularly when I'm working on pans from my IQ 180 files :~(
The last bottleneck is HD speed. Lightroom is constantly reading and writing lots and lots of little files to the HD. If you drive is optimized for large block writes (like you might do for Photoshop) it may actually slow down small block reads/writes for Lightroom. Faster HDs can and do make Lightroom seem much faster.
Aside from increasing the speed of your CPU or adding cores, the three things you can do to speed up LR is go 64 bit, add ram to at least 8 gigs (16 is better if you need to run LR and Photoshop) and get really fast HDs. It's ideal if you can keep your catalog and previews on a different physical drive than your images. And it's ideal if your catalog and previews are on a really fast non-boot drive (if the OS starts paging it really slows down disk access).
SSDs can make Lightroom fly, but are expensive...stripped arrays can boost speed (as long as you back up cause stripped arrays are fragile).
Lightroom is not really bloated...it's doing a ton of little things and a few big things. The fact the programing is done in Lua is part of the performance issue...the other part is that in reality, LR is a really heavy database app with tones of UI and interface performance requirements. But I'll tell you that will not change...
So there you have it, my best suggestions for looking at speeding LR up!
-
Not a sales ploy...it's simply the lowest hardware that Adobe tests against and officially supports. More is always better.
So, in my experience, here's the breakdown of LR bottlenecks...it's similar to Photoshop but importantly different.
Top of the ladder is processor speed...
The second is multi cores (although that tops out quickly, 2 cores is almost 2x better than single core but 4 cores is not nearly 4x, more like 2.5-3x at best).
The third is 32 vs 64 bit processing which 64 bit being at least 20-30 % better depending on ram. If you are under 4 gigs then the difference is more like 10-15%.
Ram...but past a certain point it doesn't matter. This is really where Photoshop and Lightroom are fundamentally different. As long as you have 8-16 gigs of ram on a 64-bit machine, that's the best you can expect for Lightroom. Adding more won't help. But regarding Photoshop on 64 bit machines, more ram is indeed important. I have 32 gigs and wish for 64 gigs on my main workstation...particularly when I'm working on pans from my IQ 180 files :~(
...
In there somewhere needs to be bus speed of the system - that's the speed at which data is sent to/from RAM to the CPU.
This does not always increase with CPU speed.
If you were to look at this page:
http://en.wikipedia.org/wiki/List_of_Intel_Core_i7_microprocessors
Under the column "Memory", you will see entries such as "3 × DDR3-1066" or "2 × DDR3-1333".
If you were to compare two systems that each had the same amount CPU speed, you should expect that the one with the "DDR3-1333" will be about 33% faster than the one with "DDR3-1066".
The CPU can only work as fast as data can be delivered to it.
-
If you were to compare two systems that each had the same amount CPU speed, you should expect that the one with the "DDR3-1333" will be about 33% faster than the one with "DDR3-1066".
The CPU can only work as fast as data can be delivered to it.
True but I have no specific knowledge about bus speed other than I'm currently using a Mac RAID card for super fast disk speeds...
But don't forget that LR isn't nearly as ram hungry as Photoshop is and depends on ram to a lessor degree.
-
You can't expect to run 2012 software on, say, a 2010 computer and get reasonable results.
That's a fundamentally bad attitude on many levels.
Really good coding is making software faster and more reliable on the same platform. Letting things effectively run slower and slower just isn't elegant coding.
For the end user why should we have to spend an additional £500+ on new hardware every time there's a software upgrade ?
Are the requested features really so resource hungry that has to necessary ? I doubt it. It's especially galling when really simple things get ignored that really can't cost anything in software complexity to add eg a keyboard short cut for the purple label, how hard is that ?
Why should we expect to have to junk expensive working hardware so often ? It's not exactly great for the environment.
-
You don't have to junk anything - you don't have to upgrade the software.
It's unreasonable to tell developers that they can't leverage more powerful hardware with their new software.
-
That's a fundamentally bad attitude on many levels.
Really good coding is making software faster and more reliable on the same platform. Letting things effectively run slower and slower just isn't elegant coding.
For the end user why should we have to spend an additional £500+ on new hardware every time there's a software upgrade ?
Are the requested features really so resource hungry that has to necessary ? I doubt it. It's especially galling when really simple things get ignored that really can't cost anything in software complexity to add eg a keyboard short cut for the purple label, how hard is that ?
Why should we expect to have to junk expensive working hardware so often ? It's not exactly great for the environment.
No one is forcing you to upgrade if you have reasonably recent equipment and software. Can always try out the software before buying, to see if it works with your system. LR3/PS5 is pretty damn powerful, high quality software. However, one thing I learned from my experience with Lightroom 3 is that the functionality with hardware that is not the latest and greatest can be improved. Until LR 3.5, LR3 was very slow with my system (iMac 2-core, 2.66 ghz, 8gb RAM) when using the lens correction features. Now it works fine. LR4 is again seeming a bit unresponsive after many edits and using the lens correction. The clone tool especially. However, still not nearly as bad as early LR3. Hoping they fix LR4 for this as well. By the way, does the Mac OS have any effect here? I am on 10.6.8 and have so far been reluctant to move to 10.7 since everything seemed okay and I really didn't need what OSX 10.7 had to offer.
-
If you were to compare two systems that each had the same amount CPU speed, you should expect that the one with the "DDR3-1333" will be about 33% faster than the one with "DDR3-1066".
This is not generally true. Will your car go twice as fast if you reduce its wind drag to 1/2?
-h
-
I suspect that it wouldn't matter how fast Adobe could make it some wouldn't be happy? Even if it took off out of your monitor and took to the skies some would want a repeat so they could time it. ::)
-
I don't really think diet is a good word.
My situation: I installed Lightroom 4 (trial), and converted my catalog (18000+ RAWs 10MP). I regenerated standard previews. I enabled only library, develop and print modules.
After some moderate use (just occasional 1-2 gradients on image, some brush adjustments), I noticed significant slowdowns (non-responsive sliders, delayed refresh or no image generation at all). On most pictures I adjust exposure, brightness, contrast, clarity, saturation and vibrance.
I checked memory usage - 1GB, max. 1.5GB. LR 3.6 was never a speed champ on my machine, but it was much better than LR4. Stopping and restarting the program didn't help, restarting the PC helped a little (apparently).
I like 2012 Process Version and workflow, but LR4 needs significant performance improvements.
My configuration: Win7 64bit, 6GB RAM, Intel Core2Duo E8400.
-
I am confident that it is possible to do raw development faster than Lightroom by sacrificing image quality, and that Adobe would sell less if they did.
-h
I am confident that it is possible to do raw development faster than Lightroom by sacrificing image quality, and that Adobe would sell less if they did.
Buy LR1, quality is not as good as more recent versions and it ran faster.
-
My configuration: Win7 64bit, 6GB RAM, Intel Core2Duo E8400.
The reality is, that's an older processor now and the associated bus and chipset etc is slower than is currently avialable, even at modest prices.
However, what you haven't indicated is your hard drive setup - that will also have a significant impact.
It again comes down to should developers just ignore new hardware to make sure things run fast on old hardware or should they introduce new technology that can take advantage of new hardware and still work on older hardware albeit slower?
-
For the end user why should we have to spend an additional £500+ on new hardware every time there's a software upgrade ?
Are the requested features really so resource hungry that has to necessary ? I doubt it. It's especially galling when really simple things get ignored that really can't cost anything in software complexity to add eg a keyboard short cut for the purple label, how hard is that ?
Why should we expect to have to junk expensive working hardware so often ? It's not exactly great for the environment.
What an odd attitude.
You don't have to take every software update. The old version will continue to do everything it did before.
You don't have to junk expensive hardware. It is your choice.
But you do have to run the latest software on the latest hardware to get the optimum results from it.
Again it is totally your own choice - you decide what performance and facilities you want/need and you pay for both the hardware and the software to deliver it.
I can still run Maziacs and Hungry Horace on my Sinclair Spectrum every bit as fast and efficiently as I could in 1984. My Canon A1 still takes as good photographs as it did in the same year. The fact that I now use a Nikon D3s and run LR4 on a PC with an i7 processor and 16Gb of RAM is totally my choice. No point girning about the cost.
-
However, what you haven't indicated is your hard drive setup - that will also have a significant impact.
I have two harddisks. My configuration was Windows & LR installed on one partition and the catalog (incl. previews) on another partition, on the same disk.
Since previous report, I moved the catalog (and previews) on a different disk. There is visible speed improvement, but slower than LR3.
-
Tis slow. My computer is up to spec: I7 2600k, 16 GB ram, program on SSD, catalog on another SSD. I can imagine the hurt on a lesser spec machine.
Larry
Opps wrong thread, next one down 'LR 4 slow'. On this topic - who cares. An extra $0.50 cost of drive space.
I don't really notice any speed increase with the SSD's.