mathlawguy

This is a really interesting idea. If you think of Lightroom edits as a program, what imagen-ai is doing is synthesizing a meta-program in response to data from the underlying image. One could imagine the same occurring fairly soon with video editing. And then, perhaps, on to text space. The training AI observes what edits a human text editor makes in response to various documents and then tries to develop a program that, when you feed it similar documents, produces mimicking edits. This latter task will be harder because the space of human language edits is even larger than for Lightroom photo edits, but conceptually the two tasks seem similar. The future will be ... interesting.

I appreciate that the photos in this gallery have some artistic flair. Maybe I'm less picky than some, but to me the lens looks awfully good.

This A9 owner is excited.

Seems to me the goals for this project are too modest. Why just photography? Given the expansion of AI into so many domains, there will be almost endless opportunities to train those stupid humans via algorithmically determined mild electric shocks. </sarcasm>

Could this technology be used to permit cameras with big pixel sites and thus higher dynamic range to take photos that can then be upscaled with some fidelity? Think of a Sony A7SII that, if one wanted to make whatever sacrifices are entailed by the machine learning algorithm -- particularly once it gets even better -- can upscale to the resolution of an A7RII. (Or think of a small camera that was willing to sacrifice megapixels for large pixel sites and could make the same trade off). I think that would be a very interesting option. Also, for those scared about Google abusing the technology, please recognize that others should be able to do this too and that it may be able to run locally on your own computer or camera.

Hey, putting aside the issue of the lens for a moment, I'm delighted to see dpreview has an excellent photographer. These are really well done!

This is fine. And if you need to be in the woods for weeks and think carrying a lot of batteries is worse than carrying heavier equipment, I understand that choice. Keep your DSLR and enjoy it. There are indeed some advantages. But why would Ming Thein assume that his tradeoffs (and yours) are universally superior? Evidently, an awful lot of people are deciding that the burden of switching batteries after a few hundred shots just isn't such a big deal. And they like other features that mirrorless alone can provide. .

I completely agree (except the part about the Canon 5DS, about which I do not have an opinion). But, if this is really the official position of dpreview as opposed to Richard Butler's inappropriate use of the Royal We, it makes me question Dpreview's credibility and objectivity in reviews, and certainly those of Mr. Butler. I think the readers deserve an apology, not from Mr. Thein -- he is entitled to his opinion -- but from DPreview for its thoughtless and little qualified endorsement or a very unbalanced article.

DPreview has alienated many readers, including me, by an enthusiastic and little-qualified endorsement of a very one-sided blog entry without any apparent recognition of the issues that have persisted for a long time with DSLR cameras. It insults them because many of dpreview's very knowledgeable readers have decided that the insoluble problems of 50 years of DSLRs are outweighed by the huge advantages of mirrorless, notwithstanding their residual flaws. Unfortunately, some good points in the blog entry have their credibility overwhelmed by a polemical tone suggesting that design choices in mirrorless are "wrong" because the author simply has different priorities than many consumers. Yes, the Sony A7 series, for example, does not have the battery life of DSLRs; as a user, I too wish it had better. But this is a tradeoff. As a non-professional, I prefer inserting little batteries more frequently to leaving my camera or a lens at home because it does not fit in a carryon.

Outstanding article. Really makes you appreciate how much work goes in to making a lens.

Hey folks, my apologies for a rude remark. I try to stay constructive on these sites and I regret forgetting my manners here.

Hey folks, my apologies for a rude remark. I try to stay constructive on these sites and I regret forgetting my manners here.

Fill flash?, HDR?

Fill flash?, HDR?

A new low in DPreview sample photos?

A new low in DPreview sample photos?

Music is awful and distracting. Cuts are too fast. Seems very teenage in its appeal. But the shots themselves are impressive. Maybe they need a remix for people over 25?

They really should have filed a Chapter 11 petition along with this press release. Manages to combine butt-ugly design with poorly thought through ergonomics in just one camera. Mark Newson may design products that some find attractive where aesthetics matter as much as functionality -- his web site boasts some interesting hair dryers -- but Pentax's making him the headline here on an under-specified bloated mirrorless camera is really dumb.

Isn't the real news here not the particulars of this lens but the fact that a third party has produced any lens that autofocuses on the e-mount, not to mention one that provides stabilization from someone other than the original manufacturer. If Tamron can do it, then surely Sigma and maybe another should have more e-mount AF lenses in the offing. This competition and choice makes the NEX platform even more attractive and should make Sony feel good about its decision to open the NEX platform.

As for those who complain about the size of this particular lens, isn't that mostly a constraint created by use of APS-C sensor, which has significant benefits too? Sure, I'd like an 16-300 pancake (f/1.4 of course) but in the interim I'm happy to see the beginning of what I suspect is more variety and somewhat reduced pricing in the e-mount field.