Androole: Does anyone have any actual examples that shows that the PRIME noise reduction is worthwhile?
I've tried the "best" NR packages in the past (i.e. Topaz DeNoise) and find that they all still obliterate too much detail. I'm happy removing chroma noise which is dead easy in any package, but I haven't yet found a noise reduction solution for luminance noise that looks better than leaving it untouched. Your mileage and aesthetic preferences may vary.
@Mark9473 This is exactly the sort of image that can give a misleading impression of Prime NR. It's mostly bokeh with a touch of high complexity detail in the center that's been shot in wide spectrum daylight. Of course Prime looks good... the bokeh is infinitely plasticize-able. An image in poor quality light with fine textures is another story completely.
Jacques Cornell: DxO OP 10's PRIME NR is by far the best I've seen anywhere, and I've worked extensively with Dfine and DeNoise. And, it's one-click simple. A game-changer for this high-ISO low-light event photog shooting with Micro Four Thirds.
I have to backup what curio77 said above after working with DxO 11 for the past day on several images of mine from various cameras. I was interested to see if DxO merited purchase specifically as a backup RAW convertor for high ISO images.
Overall I find that the color accuracy, tonal rendition, and even high ISO noise reduction still look better in C1 Pro. Also, the time to get to those results was faster in C1 Pro if not immediate with default settings. With DxO the default starting points often looked off to me and then it took a lot of clicking to get the color and tonal balance right or more natural.
The good news is that Prime NR indeed appears to be faster and didn't produce the posterization in shadow areas that I saw it render in earlier editions. The bad news is that it still plasticizes textured surfaces far too much for my liking. I don't shoot anything above ISO 6400 on my D700 but at least up to this ISO range C1 Pro produces a better balance of noise reduction.
fPrime: Very in tune with the trend in lens design these days... big, heavy, sharp, and well corrected but with relatively flat, characterless rendering.
I could tolerate a little CA and less than perfect bokeh in trade for more 3D pop. Over-correcting the lens for aberrations also seems to make it flatter.
Very in tune with the trend in lens design these days... big, heavy, sharp, and well corrected but with relatively flat, characterless rendering.
Don Sata: Nice review but is it still relevant two years late?
No, but it made a vacation in Iceland a tax write-off or an all expense paid extravaganza for somebody.
Lovely, a Leica camera with only the "essentials." I guess building it with a wide latitude CMOS sensor was also a must to allow M-D shooters to fix their pictures with the "essential" RAW post processing required when shooting without an LCD?
Note to Leica, read the user forums here on occasion. Your M9 customers weren't asking for a CMOS camera without LCD because they needed more anticipation in their photography. They wanted an updated, narrow latitude CCD based camera for a film-like, ultimate picture quality shooting experience.
Instead of creating something unique with a real differentiator worth spending real money on like that, you broke something that didn't need fixing. Good luck with that!
fPrime: DPR: We understand that the color filter array on the D5 sensor has been changed to improve low light performance. Can you comment on this?
NIKON: It’s very difficult to explain exactly how we achieved this, but the basic concept is that we improved light-gathering ability. And by doing that we reduced noise and increased sensitivity.
Thanks, Nikon... with every iteration you give us optically thinner CFA's to support your drive for higher ISO's. But then color fidelity suffers from weak CFA's that are less able to strictly separate light into distinct color channels. The RAW color transforms used to compensate for this on Macbeth color charts in turn twist neighboring subtle hues in unintended directions. No wonder why color enthusiasts users don't upgrade their old cameras. Better to buy another legacy color champ body on Craigslist.
IMHO the 35mm camera industry jumped the shark for high ISO performance somewhere after 2007. The MF crowd followed suit much later with the introduction of CMOS sensors for CCD in 2014. Before 2008 all of the major 35mm manufacturers held to strict CFA designs as originally laid out by Kodak. After 2008 they began tinkering with the CFA to let more light get through to the sensor.
My personal choices for pre-2008 bodies with strong legacy CFA's has been the Nikon D700 (huge pixel CMOS), Canon 5D Classic (huge pixel CMOS), Nikon D200 (big pixel CCD), and Nikon D1x (big pixel CCD). Note the importance of pixel size... the more sensitive (bigger) the native pixel well is, the better it can support a strong CFA. In the era before micro lenses and back side illumination, CMOS was significantly less sensitive than CCD so it is best to stick to low density designs (12MP) for FX. CCD can support a strong CFA with resolutions of up to 10MP in crop sensors.
fPrime: Automated AF adjustment for high resolution sensors has been way overdue for years now. This firmware option should have premiered in 2012 with the launch of the D800 with it's woeful left AF focus issues. The fact that this is launching in 2016 with only support for calibrating the center AF point under one type of lighting should be an embarrassment for Nikon.
When this can be automated for all AF points at all zoom positions under multiple light sources, it'll be a worthwhile upgrade. Until then let's call this very minor improvement to DSLR focusing tech what it really is... a half-baked step in the right direction. Underwhelmed.
+1. Think about it, primeshooter... AF adjustment is purely an offset value stored in a data table within camera memory. With flash memory as big and inexpensive as it has become today it's a crime that Nikon created this algorithm to only handle one offset value per lens.
There isn't a technical limitation to overcome here. Nikon is simply too lazy to write the program code for storing more values. That's why I'm tough on them regards AF adjustment and wish DPR would follow suit.
Automated AF adjustment for high resolution sensors has been way overdue for years now. This firmware option should have premiered in 2012 with the launch of the D800 with it's woeful left AF focus issues. The fact that this is launching in 2016 with only support for calibrating the center AF point under one type of lighting should be an embarrassment for Nikon.
DPR: We understand that the color filter array on the D5 sensor has been changed to improve low light performance. Can you comment on this?
Dr_Jon: That's a lot of elements for an 85mm (13, the Canon is 9), it will be interesting to see how well the "SP" name still holds up these days.
Maybe not the only one... have a look at this blog:
@Thematic... YMMV but I find Sigma's ART lenses render clinical, sterile, and perhaps the better phrase to use than lifeless is "less life-like" images in that they don't have the micro-contrast and color depth that much simpler lens designs have to communicate the three dimensionality of the scene. With an ART lens you basically have to shoot wide open to separate subject from background using bokeh. A simpler lens design can separate at any aperture.
In astrophotography flat and accurate imaging is desirable, but for everything else I'll prefer less-perfect, simpler lenses that can yield better micro-contrast bite and richer color.
Agreed, seems Tamron is headed down the Sigma track... more glass elements for greater lens correction. Probably will render a flat, lifeless image like the ART series do.
exapixel: Want to make this article more helpful? Run through a list of credible "best cameras for somebody", and for each, legitimately try to describe the photographers who would indeed be best served by that product.
You do realize that Dpreview are in the media biz and a thousand replies or more are better than none, right? The innuendo in the argument against their current approach is that their reviews are too milk toast... every camera is good says nothing of interest.
The right way to apply use-cases is to make tough judgements based on real test data. So for your example of pro sports things like AF accuracy, precision, tracking and ISO performance can all be tested and measured. I'd love to hear if Canon can trounce Nikon on these metrics regardless of my affiliation. The problem is that not enough of this type of testing is done, not that there's too much and fanboys are offended.
I think the reason we don't see Dpreview doing use cases is that they require in depth field & operational testing that they either don't have resources or expertise to conduct. Consequently we are largely left with static studio shot comparisons and subjective commentary on features and use.
@ Richard Butler
No, he's not talking about "every" camera... he's basically saying there are only "best" cameras in regards to certain uses.
Setup common use-cases (like shooting low light, shooting landscape, shooting action, shooting video) and review cameras against their prescribed uses... regardless of if they are in the compact, mirrorless ICL, or DSLR camera category.
The category review approach dpreview currently uses to cross compare brands (as in the mid-range interchangeable lens camera category referenced above) is relatively meaningless to buyers.
Another message the D500 sends is that Nikon finally learned that their 2007 era AF systems were no longer sufficiently accurate or precise any longer in today's era of ART lenses and high megapixel sensors. Even the D500 got a top of the line AF system.
I feel sorry for anyone who bought a high megapixel camera with either inconsistent focus (looking at you, D800) or now wantonly outdated focus (that's you, D810). The D900 with new AF will substantially devalue both of you, I'll bet.
Generally terrible greenish skin tones and lots of harsh, unflattering contrast. If I were a Sony exec I'd be calling Dpreview to deep six these images lest they impact last minute Christmas sales.
AshMills: "Featured story" - Lets keep this non-news as hot as we can, stir up more haters!
And good to see the Nikon apologists are still alive as well. Yes, let's censor and diminish the facts and truth, shall we, just so we don't stir up those "haters".
AngryCorgi: As usual, Canon puts out another camera that get's spanked by a 4-year-old Nikon in the IQ department. When will they think about either (a) purchasing Sony's sensors or (b) partnering with Sony in some way to improve their sensors to be competitive with what Sony has been producing for more than half a decade??
@Rishi: Let's agree to diasagree... you can say "updated" and I'll stick with "warmed over" Nikon AF. Bottom line, Nikon still call the D810 ad D4s AF module CAM 3500FX.
As far as things that need improvement in the AF module, certainly more cross type AF points and wider AF cross type spread are at the top of the list. But beyond that, Nikon has had long term issues with reproducible AF using fast glass wide open with high density sensors that beg for tighter tolerances. As far a forward looking designs, ideas include allowing users to calibrate individual AF points separately, allowing users to save different AF offsets for various zoom lens positions, allowing users to save different AF offsets for focus shift, allowing users to save different AF offsets for lighting. Ultimately Nikon needs pro-level PDAF on sensor to move to mirrorless.
PDAF is hardly a "finished" technology.