[nanok]

okay, so i had to get one, camera of the decade, etc. there's a problem: something is loose inside, sort of wobling when the camera is off, it's scary... i'm joking.

it is starting to be apparent to me that, with the dynamic range of this camera in _one_ shot, in the conditions of a typical shot i tend to take, my raw converter now needs to be able to compress that dr to make me happy. it seems the k-5 does something in-camera to handle that (didn't do tests though, yet) for it's own jpegs, but i want my raw converter to do some sort of tonemapping or exposure blending (preferably the later, i hate hdr-like tonemaping algorythms, i specifically want enfuse). as far as i know for now, the only application who had the insight (or is it foresight) to be close to that already is digikam, but i'm not sure yet it will do it from one raw file "automagically" (it does work with 3 separate ones), however this being opensource and with a very friendly and responsive developer team, i think it would be quick to include the tweak needed.

any other contenders? and do you guys think this is something you would require, or am i just plain mad? (i'm talking mostly to landscape photographers here, who need the dynamic range like air, i understand for some other types of photography it's just not relevant)

ps: of course i can make a script in 5 minutes to do just that, with ufraw/dcraw and enfuse, but i want controls over each image, so i can chose how far apart the three blending "source" images are, and so on; this is the point, i think the way of looking at raw processing is about to change, the dr difference to our screen/jpeg file/print is so high now, the way we used to do it seems increasingly inefficient)

---

Last edited by nanok; 11-16-2010 at 05:36 PM.

[Canada_Rockies]

okay, so i had to get one, camera of the decade, etc. there's a problem: something is loose inside, sort of wobling when the camera is off, it's scary... i'm joking.

it is starting to be apparent to me that, with the dynamic range of this camera in _one_ shot, in the conditions of a typical shot i tend to take, my raw converter now needs to be able to compress that dr to make me happy. it seems the k-5 does something in-camera to handle that (didn't do tests though, yet) for it's own jpegs, but i want my raw converter to do some sort of tonemapping or exposure blending (preferably the later, i hate hdr-like tonemaping algorythms, i specifically want enfuse). as far as i know for now, the only application who had the insight (or is it foresight) to be close to that already is digikam, but i'm not sure yet it will do it from one raw file "automagically" (it does work with 3 separate ones), however this being opensource and with a very friendly and responsive developer team, i think it would be quick to include the tweak needed.

any other contenders? and do you guys think this is something you would require, or am i just plain mad? (i'm talking mostly to landscape photographers here, who need the dynamic range like air, i understand for some other types of photography it's just not relevant)

ps: of course i can make a script in 5 minutes to do just that, with ufraw/dcraw and enfuse, but i want controls over each image, so i can chose how far apart the three blending "source" images are, and so on; this is the point, i think the way of looking at raw processing is about to change, the dr difference to our screen/jpeg file/print is so high now, the way we used to do it seems increasingly inefficient)

Check out the latest version of DxO Pro - they have a one shot HDR addition with v 6.5. I've just downloaded it and it is way too cold for me to go out and take some snow + shade shots to try it out. I suspect it is similar in effect to making a couple of copies of the original, one over exposed, the other underexposed and use these two with a Highlight and Shadow. I used Photomatix for it and it was far more natural to my eyes than any other HDR.

[nanok]

thanks albert for the hint.

wait. cold, snow? ugggh, now i'm envious. do take some pictures soon

[Canada_Rockies]

thanks albert for the hint.

wait. cold, snow? ugggh, now i'm envious. do take some pictures soon

It's supposed to warm up toward next week to just barely freezing. Right now we are expecting -20C overnight (-4 F) so it is a bit chilly out there. I'll see what I can do.

[nanok]

-20c sounds like heaven. make some tea, i'm coming your way

[Canada_Rockies]

I took an image inside my house, looking south on a cloudy day, today, to try out DxO 6.5 one image HDR. I was pleasantly surprised at the results. This is admittedly one of the worst things you can try to take a picture of - a darkish walled room and a glassed door looking at bright white snow and trees. Frankly, I am impressed with the results. Full size JPEGs are available in my Flickr site.

K10D, DA* 16-50, ISO 320, 1/80 @ f/5.6, -2 EV

DNG from camera using DxO RAW Conversion K10D preset without HDR, then imported into Lightroom to create JPEG.



The same image Virtual Copy with all the same except applied Strong HDR in DxO.

[nanok]

that's quite impressive. and the tonempaping doesn't look like shot on some exotic planet. such a shame they don't make it for linux too

[Canada_Rockies]

that's quite impressive. and the tonempaping doesn't look like shot on some exotic planet. such a shame they don't make it for linux too

Do you think it might run in a virtual Windows window? I'm running XP Pro, so you don't need the latest emulator.

[nanok]

yes, it probably would, but slowly. i think i'd rather write the damn script than use vmware or wine or whatever

doesn't matter, i'd probably never go for non-open software unless there's really nothing close available, but many people might find it useful (assuming anybody else is reading this..)

[Canada_Rockies]

yes, it probably would, but slowly. i think i'd rather write the damn script than use vmware or wine or whatever

doesn't matter, i'd probably never go for non-open software unless there's really nothing close available, but many people might find it useful (assuming anybody else is reading this..)

Maybe it's just down to you and me. Who knows? We're having fun, anyway!

[mithrandir]

DxO Optics Pro 6.5 (one shot HDR) has turned my K10D and K-7 into the camera I always wanted. It is almost magical. It is truly amazing what it does with the RAW files.

[falconeye]

i'd probably never go for non-open software unless there's really nothing close available, but many people might find it useful (assuming anybody else is reading this..)

It's a bit OT, but since you created the thread, I may ask anyway: Why?

I am asking as somebody living from people paying for software. Also, I observe a tendency for 99% of open source software being copycats and 1% results from publicly financed research. Don't misunderstand me, I love open source software. But why make it a dogma?

[nanok]

hi Falk,

no need to explain yourself, your question is perfectly reasonable.

as you seem to be familiar with open source, you probably know that price has nothing to do with it. when i say open source i don't mean "for free". i don't mind paying for software which is good, but i do mind using software which is closed, for many reasons. there are many businesses which make money based on open source software, so not making money is not something that has to do directly with oss.

the problems i have with non-oss: i am at the mercy of a company/developer/team of developers. this includes the possibility of the company just going out of business, and the software becoming "dead" (has happened before). it also includes more "day to day" annoyances, like updates not being done quickly because they are "low priority" (yeah, the users want them, but we can make money quicker from something else -- can't blame them really, can you?). open source gives me the security that the software will keep being developed by "somebody" if it's good and there's enough interest in it, it also gives me the certainty that, in most cases, features will be added because they are needed/wanted, not because they sell (big difference); it has an entire community to audit the code (for bugs, security issues, etc). in most ways, it is better, sometimes it is slow to "get there", sometimes (for some uses), might be too slow (see cad applications), but in most cases you can get by nicely. in most cases, oss software will be simply superior from a technical standpoint (there are exceptions).

i don't know where you got the 99/1 % quoted above (with all due respect, that's a bit malicious, i can feel a bit of displeisure towards opensource in that ), but copycats are rare imho, it usually goes the other way around (usually, not always).

last but not least, i have worked with coders due to the nature of my job, and i respect them and kind of understand what their work is like, so i find it very difficult to infringe on their copyrights. if a developer (or company paying that dev) decides that you need to pay this amount of money for using his app, my personal opinnion is that there are 2 options: you use it and pay up, or you don't use it, and don't pay. the third is unacceptable to me. i noticed this "dogma" is very common in the oss community btw. i do not think using pirated software is okay if it's just for personal use. in my experience, at least 90% of software used "at home" is not paid for, i dislike that, i would much rather use gpl software instead. if something is really good and worth paying for to use, and no suitable alternative exists, than i will just pay to use it. i tend to recommend linux and/or oss software (even oss software on windows) to people as a fair alternative to illegal usage of software they won't pay for.

there's a lot of money to be made in oss, it seems, the basic "pay for software" scheme has proven 1. not working (most people will not pay) 2. inefficient (the development is driven by the "wrong" motivation).

as a last note, speaking of raw decoding in particular: you are surely aware that a huge proportion of the third party raw applications these days (including commercial, closed ones) are based on the open source code developed and maintained by one guy. i'm wondering how well they would get by (and how much more money they would have to throw into reverse engineering and eventually the price of the final product) if dave would have released the code under gpl instead .

edit: to be clear, i created the thread being interested in all variants, and in the idea that some other people (landscape shooters?) might be interested. i personally need it to run natively on linux (it would have to be damn good to convince me to maintain a windows installation just for it), and ideally be open source. but this is just me, i don't expect everybody to have such requirements/preferences, that's why it's nto mentioned in the first post, and i am glad to discuss any software people reocommend.

[falconeye]

nanok,

interesting reply. If others in the thread think it's OT, please make a comment and I'll stop the topic from my side

Actually, you quote the best reason in favour of open source: it's fast, reliable and the source can be consulted to understand what it does and to build upon in own work. I'm not malicious against oss, I love it and I use it whereever possible.

However, I still think it is a bit overrated. The 99/1% quote is mine Just refer me to open source software projects which are not copycats (or a research delivery). I'm serious. From the back of my mind, I cannot find one.

BTW, we have released a piece of software which is free for non-commercial use. It's closed source and would be a perfect candidate for open source. It implements an algebra on XML documents which turns out to be about as powerful as XSL. It is our invention. It is a non-trivial component in one of our products too. Because it is our invention and we are sure the "free" community would copy us as soon as they understand how we did it, we closed (and obfuscated) the source. Nevertheless, people download and use this software just as if it were open source. We allow it's use in open source projects and people have no problems to adopt it. Because it has no peer. And nobody ever complained because in supporting it, we compete with any open source project out there. Some customers have the code in sealed enveloppes though ...

I only say this because as great as it is, IMHO open source is not driving the industry. So, I was surprised seeing it stated as a dogma.

Actually, I'm no friend of the consulting around oss business either. Oss is from developers for developers. JBoss comes to mind. Or Novell. IMHO, all this money should go into new technology. E.g., all XML parsers in Java or .NET are crap. But nobody can produce a better one because the market is dead after open source software covered that. If I should state things in a malicious way (to emphasize the point), then software innovations run until being halted by open source. It's not about pirated software. It's about copied intellectual property. IT need not fear China. China in IT is everywhere.

So, I'm in favour of oss for commodity software where innovation is not to be expected anymore. But not so much where ongoing innovation is made difficult business wise. Or the other way round, where no intellectual property is copied. And it cannot or at least shouldn't be protected. Because software patents are even more evil. But in the end, oss may just result in that: software patents. And Game Over.

[nanok]

hmm. you keep at this "dogma" label. it's not dogma. let me put it another way: life is just too short, so as long as i am able to without any major annoyance, i would rather use oss. in some cases, i will go out of my way to not be stuck with closed source (for instance, computer hardware: if there's no open source support for that hw, i'm not interested. closed "drivers"? commendable, but no thanks. thus, nvidia video cards, amongst others, are blacklisted. it's just too much hassle and i can't be bothered, that's all. the reasons are multiple, if you're curious).

it depends which industry you mean. i agree that in some sectors, open source is behind (i already mentioned cad, there are others). but the it industry for instance is led by opensource and has been for quite some time (and that's hardly arguable ). or, if you wish, the web is opensource-"owned" (what was it, 80% apache?)

oss projects: the linux kernel? kde? (and please don't say it's a copy of windows, i would be very disappointed ), digikam? come on, there are many. it's easy to get into a "chicken and egg" debate, but most of them are not copycats, they maybe are designed to fill a similar set of needs as some other (commercial) software, and some similarities are inevitable (like photo editing software working with the layers concept, and so on), but one should not fall into the trap of calling every car ever made a copycat of the ford model t

offtopic: probably is . i have a hunch the best place to move this debate might be private e-mail, i'm not sure there's many people here who really care, to be honest (i do find the debate interesting, though)