You know those moments of calm near the end of the movie? You finally relax. Christine the psychopathic car has been fed into the crusher and is just a harmless cube of scrap metal, or the Balrog is tumbling down into the abyss, and the hobbits are safe. While the credits haven’t yet begun to roll, you’re already thinking about the best way back to the car, maybe so you can nip into Bar Italia or back past that interesting-looking pub you noticed earlier in the evening. Maybe you will be allowed that pint after all? A nice cosy feeling, isn’t it? And then, all of a sudden the beast springs back to life. Well, with Apple barely able to give Aperture away at half price, its market share evaporating with the deafening silence about its future, suddenly it’s on your tail, growling furiously. Did you spill your popcorn?
I was a bit surprised, certainly by the timing, but only the week before it was released I had been mentally drafting a post asking where – if anywhere – Aperture 3 might go. My guess was that it might look for space upmarket by offering multi-user capability. After all, Aperture and Lightroom users struggle when more than one person needs to work on a shoot. Alternatively I wondered if they might try to break out of the ghetto and attack Adobe on both fronts by going cross-platform. Even to a “Mac user through and through” like Richard Earney, being limited to Mac is a disadvantage for Aperture users. Being the minority program and then limited to one computer brand means the third party ecosystem is always going to be less extensive – fewer learning resources, fewer plug-ins, fewer friends to learn from, but most of all, fewer sales for Apple. So fundamental changes – going multi-user or cross-platform – were how I thought Apple would try to recover all that lost ground, if they thought there was still money in it.
Neither speculation was right. Instead “one more push” seems to be the strategy – pouring effort into matching Lightroom’s localised adjustments while bolting on a couple of prominent features from elsewhere.
Instability – take the fanboys’ word for it
One thing that is very obvious is that Aperture 3 has been rushed to market and has stability problems. My own experience has not been too bad – my single crash occurred 10 minutes after installing the program – but I’ve heard reports of much worse from Mac-using friends. But rather than rely on hearsay, even a self-confessed Aperture fanboy such as Scott Bourne is “averaging one total Aperture crash every 90 minutes”, as he writes in Very Cool But NOT Ready For Prime Time. Openly admitting his “rooting for Apple and Aperture” bias, Bourne “can’t advise serious photographers to trust Aperture 3.0″. Also see Rob Boyer The Best And The Worst. Again, a source biassed towards Aperture and Apple has stability problems.
In the “Lightroomsphere” such guys would be beta testers and would have given the program a tough workover well before its release, but it’s obvious that hasn’t happened. It would no doubt run counter to Apple’s culture, but maybe they need beta testers who aren’t yes men?
Adjustments – the big catch-up
The big new feature that any photographer must welcome is that Aperture 3 has now got localised adjustment brushes. They seem to work well, neither better nor worse than those in Lightroom 2. Their naming seems more task-oriented, so Aperture has a “soften skin” brush while the Lightroom user can get the same effect by choosing a negative Clarity value or by choosing a brush preset – called “soften skin”. So Aperture 3 has caught up – it needed to.
A curious aspect of how Aperture 3 implements local adjustments was brought to my attention by Gilles from Utiliser Lightroom. Aperture creates a hidden TIF file for each local adjustment that you apply to an image. Here I applied two brush adjustments to Billy, skin smoothing around the face, and sharpening just around his eyes, so two of these hidden mask files were built. Amounting to 250kb for just two adjustments to one image, these TIFs could soon add up. But the main reason for mentioning them is that they may also explain why Aperture seems to allow only a single application of each brush adjustment on a certain spot, unlike Lightroom where you can brush on the same adjustment again and again – four or five skin smoothing adjustments on the same wrinkles, for example.
There are lots of other minor tweaks in the adjustment area. Presets mean that gullible Aperture users can now be duped into paying for fake film and other special effects. Black and white is still perfunctory, with no ability to split tone, and there’s nothing as intuitive as Lightroom’s wonderful targeted adjustment tool (you can’t help but feel Apple would have found a catchier name for that) for painting global adjustments while keeping your eyes on the picture.
Old problems remain. Richard Earney also points out Apple’s slowness in supporting new cameras, while Gilles tells me that Aperture’s new rendering engine only supports 7 camera bodies.
And Aperture still trails badly in simultaneously making identical adjustments to multiple images. It’s still stuck in the “lift and stamp” or “copy and paste” days – 5 keystrokes compared to the single one needed when you’re working in Lightroom AutoSync’s mode. So in terms of adjusting your pictures, Aperture 3 has barely caught up in terms of what you can achieve, and still lags in efficiency.
Places – yes
Of the other two major new features, the GPS feature “Places” is certainly the more polished. Just like 2 or 3 years ago Microsoft added Virtual Earth to Expression Media, essentially Apple have dropped in Google maps and then written a bit of code around it.
So it recognises GPS data that’s already been added to images and it displays them on a map, while also allowing you to drag images onto a map and storing the co-ordinates. Unfortunately, while GPS data is written into any files exported from the system, it doesn’t let you update the originals’ EXIF information. In many ways it’s much better to geotag in something like HoudahGeo so your GPS data isn’t totally dependent on Aperture but is available to any other program. But overall, I like this feature. I just doubt that it will make anyone move to Aperture – if GPS data is important for your photography, you’ve probably covered that base long ago.
Faces – right idea, cheap-looking implementation
The other big new feature, Faces, has a curious sandy-gravel background which doesn’t match Aperture’s professional appearance and readily betrays its origin in the consumer-oriented iPhoto given away with every Mac. This background shouldn’t even be a preference.
There is significant value in having a tool dedicated to letting the user record who is in a photo and find photos containing that person. iView/ExMedia recognised this need and provided its People field, and it’s a shame that Lightroom 2 provides no other mechanism other than using keywords (what if you don’t want your kids’ names to be in images that may get online?). So I’ve no doubt recording people’s names is a plus for Aperture.
Unfortunately, apart from the gravel, what makes Faces seem such a trivial feature is that it is strapped up to facial recognition. As far as I can tell, you have to let Aperture go through trying to detect faces – you can’t just say “there’s Charlie the Cavalier King Charles Spaniel”, tag those images and move on. But I did like how it auto-completes names by referring to your Mac’s address book or Mail (not sure which, but it’s looking up something from wholly outside Aperture).
In my test the face recognition works well enough, so once I had added Billy Bragg’s name to one photo, it detected him in 7 of the 8 other frames from that day. However, typing in his name would have been a lot faster. Equally, while I also liked how it displayed other faces that it had detected, effectively offering me a to-do list, in practice this was less helpful when a set of pictures showed a crowd and lots of different faces were in-focus. It also managed to suggest that someone I identified in a wedding picture was also in a picture of a horse – so there’s a lot of hit or miss here.
So, while recording names in the catalogue is certainly a worthwhile addition, the facial recognition aspect seems more of a distraction – as well as a processing hog. It also dazzles the unwary user into overlooking that the identification data is stored only in the catalogue and can’t be saved back to the files or to XMP sidecars. Apple (and McCreate) have failed to notice that almost a year ago the IPTC introduced the Person Shown field and that in any case the XMP concept allows anyone to add their own metadata. So face data is only available so long as you keep using Aperture – from a DAM perspective, that’s not very good practice.
Metadata – finally it reads from sidecar files
It was always odd that Aperture couldn’t read sidecar XMP files, yet was always able to write them when you exported duplicates of your originals. This meant it was easy to abandon Aperture and take your keywords and other metadata with you, but much harder to import existing sidecar-based metadata if you were coming the other way. Aperture 3 has rightly corrected this imbalance and can now read metadata from sidecar files. Good move.
What ever happened to never touching your raw files?
Another interesting detail is that it looks like Apple have gained access to camera makers’ proprietary file information. This certainly has benefits – I like View > Show Focus Points.
But the surprise is they have also used this access to write IPTC metadata directly back into your raw files. That’s a brave decision, as Nikon’s amazingly-unreliable SDK used to be a great way to corrupt your pictures. And it’s also quite a big about turn – remember how loudly Aperture once promised to leave your raw files untouched?
Also see David Rieck’s important article issues with how Apple Aperture 3 writes metadata, to which I linked yesterday. It highlights major weaknesses in how Aperture preserves metadata that’s already in your pictures – when it writes back metadata entered in Aperture, that existing metadata can be lost.
A similar problem affects star ratings that have been applied in other programs – Aperture 3 fails to read them! Moving to Aperture would mean doing them all again. For a while there was a support note indicating this failure is in fact a design decision because ratings and labels aren’t part of XMP 1.0. But when you think about it, that’s very weak. Aperture is finally reading sidecars so that users can bring in existing metadata that has been created in third party applications, so it should be allowing for the “non-standard” ways that ratings are used in third party apps. It’s a bit like me refusing to read a post because the author can’t spell or clearly can’t use apostrophes properly. The fact is, in the real world, just like folk don’t know their “it’s” from their “its”, ratings and labels are stored in non-standard ways. Fair enough for Apple to write those fields in the standard manner, but it’s lame to try to cover their failures by hiding behind standards. It’s obviously a stock excuse they’ve just dug up – and again shows the weakness of yes-man user testing. Update: And guess what, despite there being no standard, that support note was soon suppressed and reading star ratings was one of Aperture 3′s first bug fixes.
In each case, the underlying fault is hiding behind a narrow interpretation of standards and a too-limited view of what to do when there are inconsistencies in how the standards have been applied in practice. This is a tricky area – one wouldn’t expect anyone to get it right first time – but let’s not pretend there isn’t a nasty problem.
And my conclusion? It’s a bit like you think you’ve finally got someone out of your life, you’ve got some beer in, and you were happily moving on. But then your errant lover shows up on your door, says she/he has cleaned up her/his ways and has a couple of new features, a cheap boob/moob job. Not really enough to turn back the clock….
PS I’m aware of the McCreate article that rolls out Apple’s party line. While I have no problem with it taking quotes from this post, I have not given permission to use my text in full or to use any images and am surprised at the McCreate author David Schloss’s loose attitude to copyright infringement.