Dpreview REALLY scraping the barrel bottom now!
in [Digital Cameras]
|
From: Doug McDonald on 8 Jul 2010 08:43 On 7/7/2010 6:30 PM, Stuffed Crust wrote: > In rec.photo.digital.slr-systems Savageduck<savageduck1@{removespam}me.com> wrote: >> I also thought that odd. I think that is because the manufacturers know >> their target market is never going to use RAW, let alone know what RAW >> is. So why make any effort to add a feature they know will never be >> used. > > Or perhaps this is due to the fact that most compact cameras now correct > the images for lens flaws (eg pincushion/barrel distortion) when > generating JPGs, but if you pull up their RAW output, you'll have to > apply the corection manually.. and the corrections would likely change > at each combination of zoom stepping and focal distance. > > So unless the manufacturer provides an official RAW converter, that's a > hell of a lot of work to get a non-distorted image... > > - Solomon But what they could provide would be a converted 16 bit file that had either an un-gamma-corrected image or one with a fixed, true gamma of say 2.0 .. by that I mean no heel nor toe, and no fixing for blown highlights. It would have distortion correction and lateral CA correction only ... no noise reduction. Doug McDonald
From: J. Clarke on 8 Jul 2010 08:34 On 7/8/2010 4:27 AM, Chris Malcolm wrote: > In rec.photo.digital John Navas<spamfilter1(a)navasgroup.com> wrote: >> On Thu, 08 Jul 2010 12:16:38 +1200, in<i135dk$672$1(a)news.albasani.net>, >> Me<user(a)domain.invalid> wrote: >>> On 8/07/2010 11:56 a.m., John Navas wrote: >>>> On 07 Jul 2010 23:30:48 GMT, in >>>> <4c350e28$0$4840$9a6e19ea(a)unlimited.newshosting.com>, Stuffed Crust >>>> <pizza(a)spam.shaftnet.org> wrote: > >>>>> Or perhaps this is due to the fact that most compact cameras now correct >>>>> the images for lens flaws (eg pincushion/barrel distortion) when >>>>> generating JPGs, but if you pull up their RAW output, you'll have to >>>>> apply the corection manually.. and the corrections would likely change >>>>> at each combination of zoom stepping and focal distance. >>>>> >>>>> So unless the manufacturer provides an official RAW converter, that's a >>>>> hell of a lot of work to get a non-distorted image... >>>> >>>> Bingo! >>>> >>> How well does PTlens go in correcting this distortion? >>> Last time I used it, it would read metadata in jpeg, camera model, lens >>> including focal length the zoom was set at, and make a correction from >>> the database automatically (okay - sometimes you would need to intervene >>> to select the lens if it misidentified slr lenses of the same type/focal >>> length, but this is not very hard to do). >>> It also corrects complex ("moustache" pattern) commonly found on wide >>> zooms. The internal interpolation algorithm used by PTlens seems good, >>> even on jpegs, and options for recompression after editing are >>> available. Last time I used it, CA correction was manual - so that was >>> slower and less precise than automatic CA correction. >>> There seems to be quite a range of compact cameras supported: >>> http://epaperpress.com/ptlens/ > >> I used to use PTLens, but eventually gave up on it because: (a) I rarely >> need any distortion correction with current cameras; (b) PTLens was a >> pain to use, and (c) Photoshop got so good at distortion correction. > > Has Photoshop got around to doing complex distortion correction such > as moustache? Supposedely CS5 can. > Last time I looked people were complaining that it > didn't. Does it do specific lens correcyions from a database? CS5 does. > That's > essential if you're very fussy about geometry distortions. A bit of > barrel distortion doesn't matter much in landscape, and (especially at > the edges of a wide view) can be aesthetically advantageous in > portraits, but are a serious problem in architectural photography. > > For very precise geometry correction and perspective adjustments the > problem with PTLens is the lack of easy access to a high degrees of > pannable zoom for tiny fractional corrections. Lines which look > straight enough on a screen can be obviously squint on a A2 print. But > PTLens is so good and precise at the corrections it can do that when I > need such precision for a gallery sized print I'll jump between PTLens > and a fast zoomable viewer just to get things precisely right. > > If Photoshop is now as precise and accurate as PTLens, but a lot > easier to use, I might consider getting it.
From: John Navas on 8 Jul 2010 10:20 On 8 Jul 2010 08:27:11 GMT, in <89lgevFh11U2(a)mid.individual.net>, Chris Malcolm <cam(a)holyrood.ed.ac.uk> wrote: >In rec.photo.digital John Navas <spamfilter1(a)navasgroup.com> wrote: >> I used to use PTLens, but eventually gave up on it because: (a) I rarely >> need any distortion correction with current cameras; (b) PTLens was a >> pain to use, and (c) Photoshop got so good at distortion correction. > >Has Photoshop got around to doing complex distortion correction such >as moustache? Last time I looked people were complaining that it >didn't. Does it do specific lens correcyions from a database? That's >essential if you're very fussy about geometry distortions. A bit of >barrel distortion doesn't matter much in landscape, and (especially at >the edges of a wide view) can be aesthetically advantageous in >portraits, but are a serious problem in architectural photography. Depends on the camera and the lens, and as I wrote, my current cameras need little post-correction, and Photoshop is quite effective at giving me the correction I need (architectural photography included). >For very precise geometry correction and perspective adjustments the >problem with PTLens is the lack of easy access to a high degrees of >pannable zoom for tiny fractional corrections. Lines which look >straight enough on a screen can be obviously squint on a A2 print. But >PTLens is so good and precise at the corrections it can do that when I >need such precision for a gallery sized print I'll jump between PTLens >and a fast zoomable viewer just to get things precisely right. I'm personally not interested in that kind of precision. I'm just interested in good images. "A Ming vase can be well-designed and well-made and is beautiful for that reason alone. I don't think this can be true for photography. Unless there is something a little incomplete and a little strange, it will simply look like a copy of something pretty. We won't take an interest in it." ~John Loengard, "Pictures Under Discussion" >If Photoshop is now as precise and accurate as PTLens, but a lot >easier to use, I might consider getting it. It's certainly worth a trial. CS5 is amazingly capable, though I tend to use Elements for less demanding work (and Corel PHOTO-PAINT for certain types of manipulation). But as always, YMMV. -- John "Assumption is the mother of all screw ups." [Wethern�s Law of Suspended Judgement]
From: Bruce on 8 Jul 2010 10:21 On Wed, 7 Jul 2010 15:32:12 -0700, Savageduck <savageduck1@{REMOVESPAM}me.com> wrote: >On 2010-07-07 14:19:01 -0700, Bruce <docnews2011(a)gmail.com> said: >> What I found very surprising was that the Panasonic's RAW performance >> was actually marked significantly *lower* than its JPEG performance. >> It is usually the case that post processing RAW files outside the >> camera gives far superior image quality. >> >> The Panasonic was one of only a few models tested that offered RAW - >> it was notable that the Canon and Nikon models did not - so it was >> disturbing to see that shooting RAW would have been pointless. > >I also thought that odd. I think that is because the manufacturers know >their target market is never going to use RAW, let alone know what RAW >is. So why make any effort to add a feature they know will never be >used. > >I think you will find the marketing war cry will be, "If you want RAW, >buy one of our DSLR's!" > >Few of the buyers of those cameras will use the more complex features. >The great majority will use them as a "big snap shot camera." Perhaps >that is what we shoutd term them "BSSC"? ;-) That's all true, but Panasonic has made a point of including RAW in almost all its high-end small-sensor cameras (if that last phrase isn't an oxymoron, that is!). I have used the LX2 and LX3 professionally and found that RAW added 1.5 to 2.0 stops of extra dynamic range as well as the opportunity to avoid the camera's over-heavy noise reduction and the sometimes quite grotesque results of over-heavy in-camera sharpening. The anti-DSLR troll appears completely unaware of the value of this feature of RAW. One reason for changing to the Micro 4/3 GF-1 model was that much less post-processing is required and you can even use the JPEGs out of the camera is you are in a hurry. Paradoxically, RAW is not essential with the GF-1 whereas it is with the small-sensor LX2 and LX3! I couldn't work out from the section of the review which dealt with the Panasonic FZ-35/38 why the RAW feature got such poor marks. Ir was mentioned only in passing: "It is also one of four cameras in this group that capture RAW files. This allows you to remove the camera's default noise reduction and to process the images exactly as you wish. However, in our tests this did not get us much more resolution." Perhaps the reviewer missed the point? The greatest value of using RAW is to extract the maximum possible dynamic range for the final image. Except for avoiding the smudging caused by over-powerful in-camera noise reduction, RAW doesn't materially increase resolution. And in the Summary, the reviewer, having criticised and marked down the RAW feature, actually included it on the "We like" list! I think the conclusion is that this particular DPReviewer doesn't really understand what the feature is for. DPReview strikes again!
From: John Navas on 8 Jul 2010 10:33
On Thu, 08 Jul 2010 15:21:50 +0100, in <udmb365fs9bnja9uadreaag34mkj6nkuag(a)4ax.com>, Bruce <docnews2011(a)gmail.com> wrote: >I couldn't work out from the section of the review which dealt with >the Panasonic FZ-35/38 why the RAW feature got such poor marks. Ir >was mentioned only in passing: "It is also one of four cameras in this >group that capture RAW files. This allows you to remove the camera's >default noise reduction and to process the images exactly as you wish. >However, in our tests this did not get us much more resolution." > >Perhaps the reviewer missed the point? The greatest value of using >RAW is to extract the maximum possible dynamic range for the final >image. Except for avoiding the smudging caused by over-powerful >in-camera noise reduction, RAW doesn't materially increase resolution. > >And in the Summary, the reviewer, having criticised and marked down >the RAW feature, actually included it on the "We like" list! I think >the conclusion is that this particular DPReviewer doesn't really >understand what the feature is for. DPReview strikes again! Fair observation, except for the childish taunt at the end. The key is to learn how to use a given tool effectively. I personally find less and less need for RAW output from my FZ28 as I get better and better at using it. For example, when shooting sailboats (with large white sails), if I set the metering mode properly and exposure bias of -1/3, I get consistent good results right out of camera. -- John "It's a poor workman who blames his tools." [proverb] |