« The Amazing Shrinking Panasonic | Main | Jazz Notes: Ben and Sweets »

Saturday, 06 November 2010

Comments

I want one. I can't afford it, so I am waiting for some reviews of the K-r, which uses a different sensor. At least as far as the sensor goes, it does not look like a slouch either, with perhaps only one or arguably a couple of cameras in its price range scoring higher.

You know, Mike, that posts like this don't do anything to help quell gear lust? Your irresponsibility may have to be submitted in evidence during my next in-depth examination of the economic situation on the home front... :)

It's been a little funny and sad reading some of the comments on the web about the DxO readings. There seems to be such an anti-Pentax bias out there that having finally put one over the wall no one wants to give Pentax credit. The reactions range from "DxO screwed up" to "THE D7000 WILL DO BETTER!!"

Hats off to Pentax for a job well done and for firmly planting a boot in the Biggies' tush.

[you are welcome ; )].
Thing is, what astonishes me is the ability to improve smaller sensors´ performance. Which will put smaller sensor cameras to shame as the mobile phone gets pervasive.

Is it my birthday yet?

Wow. Nice work, Pentax.

As good as the sensor new K5 might be:

1) DXO itself clearly disclaims direct correlations between sensor performance and overall performance of a camera: “Disclaimer:This dxomark review evaluates only the selected camera’s RAW sensor performance metrics (i.e., Color Depth, Dynamic Range, and Low-Light ISO), and should not be construed as a review of the camera’s overall performance, as it does not address such other important criteria as image signal processing, mechanical robustness, ease of use, flexibility, optics, value for money, etc. While RAW sensor performance is critically important, it is not the only factor that should be taken into consideration when choosing a digital camera.”

In other words, without testing the camera results DXO is merely testing one component in a parts-chain; the processing engine and supporting electronics and design elements are at least as (and almost certainly more) important.

(2) According to DXO, a difference of just "one f-stop offsets the Overall Sensor Score by approximately 15 points", so the differences between competitive sensors are actually less important or impressive than they seem at first to be.

(3) The mavens running dpreview have been dealing with explaining and deconstructing the meaning of the DXO results for many months now. One notable point is that DXO results may be skewed compared to real world results because they use a different definition of ISO than just about anyone else. A quick summary by dpreview's Any Westlake here: http://forums.dpreview.com/forums/read.asp?forum=1018&message=32965502&q=dxo+iso&qf=m

Just checked, but still no fast(ish) 35-40mm equivalent prime... I'll check back in another 3 years.

"Just checked, but still no fast(ish) 35-40mm equivalent prime... I'll check back in another 3 years."

Nick,
That's my big problem. Speaking not as a reviewer but just as me. I've been writing about that particular problem at least since August of 2008.

Mike

Since the megapixel race has shifted from affordable and marginally affordable point-n-shoots and DSLRs to prohibitively expensive medium format wonders, commenters and pundits have been starved for new material.

Let the DxOmark wars begin.

I was surprised by this, and just looking at the numbers, it's impressive. Ahead of D3 and D700! Surely not? Bit more to it than the end figure though.

Looking at the graphs, I noted two areas of interest. First: something happens above ISO 1600 which 'kicks' several of the graphs over to the right from their general downward curve. Second: We've become accustomed to the camera manufacturer's extending the performance in to the higher ISO ranges, but what Pentax appear to have done is give an extrapolation of the graphs in opposite direction by giving a lower ISO value of 80.

Maybe others will copy the later, and give us ISO 64, 50, and 25... remember those days?

I'd be surprised if Nikon's D7000 does as well. Though I note its base ISO has changed from the D90's value of 200, to 100.

I await the other reviews with interest.

It feels good to see, for once, the almost permanent underdog, taking the main spot in media, and putting to shame the big gorilla players.

It feels good also to realize that you don't need to be a huge company to make cameras that please photographers and provide truly top notch image quality.

I don't know what effect will have this kind of sudden, massive free promotion derived from objective tests; but certainly is welcome by those of use who always had a very soft spot for Pentax.

As a person who began photography with Pentax cameras, and who once loved his k20d with 21/40/70mm DA Limiteds and 16-50 DA*, I'm glad to see Pentax deliver such a knockout of a camera, sensor-wise. And I wish that I could see this turning into a big benefit for Pentax. I just can't see it happening. This result may cause Pentax owners to choose to upgrade, and may even offer cover to the small number of enthusiasts that were considering Pentax anyway; the public at large, however, "knows" that a good camera is a Canon or a Nikon, and that sort of bias is too much to overcome quickly, I fear.

Here's hoping they keep up the good work, continue to innovate with lenses, and use the 645D to increase their profile amongst professionals. Combining those last two points, I'd love to see more than one lens available new at retail for the 645D...

Dear Sandro,

You're trying to compare apples (in-camera JPEG quality) to oranges (in-camera RAW quality). Most of the 'deconstruction' going on in dpreview is the result of that confusion.

When you photograph JPEGs, DxO tests have much less relevance to you. When you make RAWs, dpreview's tests have much less relevance.

The whole ISO kerfuffle is nonsense. Ignore it. dpreview is right enough regarding JPEGs; they're pretty much wrong when it comes to RAW. If what you're making are RAW photos, the camera's default midtone placement point for JPEGs is irrelevant. I'm not going to get into the technical details, just saying skip it all.

'Sides, you have people taking up arms over whole half-stop differences. Really!

You misunderstand what DxO tests-- they test the sensor in-camera, so their tests do include the camera's support and processing electronics: There's no other way to get the RAW image out, 'cept through that signal processing chain.

pax / Ctein

Well, there is the Pentax 35/2.0 or 43/1.9, or depending on your definition of fast(ish), the wonderful 40/2.8 Ltd pancake. Or if you want faster, the Sigma 30/1.4 or 50/1.4 in PK mount.

"still no fast(ish) 35-40mm equivalent prime"

It's a dead horse, I know, but Pentax could instantly offer a nice prime-based system by producing just a full frame body--it still offers a bunch of fast FF lenses from 31/1.8 to 100/2.8.

Say, a limited run minimalist FF body in the vein of a 5D or A850. It's a somewhat tall order, I guess, but I would think not nearly as tall as a medium format DSLR.

Alex,
None of those are 35mm equivalent on APS-C.

I love the 35mm f/2.8 DA Macro; wonderful lens. It's just a bit too long for me on APS-C is all.

Mike

Ctein explained: You misunderstand what DxO tests-- they test the sensor in-camera, so their tests do include the camera's support and processing electronics: There's no other way to get the RAW image out, 'cept through that signal processing chain.

This is precisely correct, and the idea that DxO somehow tests sensors apart from the cameras in which they are embedded is a bizarre misconception that, IMO, DxO have made little attempt to dispell. Moreover, there are aspects of the DxO testing methodology that are "black box," that is, aspects that are proprietary and that therefore make the DxO results non-reproducible by independent workers (see the link in the next paragraph, especially the comments).

All that said, for those who have followed the results coming from the Pentax K-x, none of this comes as any surprise. The K-x was already nipping at the D700's heels, so it's no surprise that with another round of sensor development the K-5 is even better.

Note: I am not a Pentax partisan. I don't work for 'em. I don't have any Pentax film cameras, and the only Pentax digital that I ever bought (K20D), I returned to B&H for a refund.

"Just checked, but still no fast(ish) 35-40mm equivalent prime... I'll check back in another 3 years."

Pentax used to make a SUPERB FA* 24mm F/2.0

If you can find one in the used market, it's "THE" lens for that range!
(note: the FA* line of the lenses, used to be their TOP professional line)

Since I have a K-7, I can add that for the mechanical robustness, ease of use, flexibility, optics, there's absolutely no cause for concern.

As for the 35mm-e lens, there is the DA 21mm Limited, which isn't exactly fast at f/3.2, but it is very small, solidly built, and optically nice. (And like all the other Limited lenses, decent wide open.) With the great results the K-5 apparently can produce at medium and high ISOs, the fast part of the equation might not be such a show-stopper.

If we pay too much attention to DXO then my little EPL1 should pretty much suck. Damn thing takes wonderful photos.

Dear Semilog,

I don't have a problem with DxO having proprietary tests; they're not publishing scientific papers for peer review. It may be frustrating for some armchair scientists, but not me.

In those few cases where I could perform tests to compare with DxO's, I've gotten close to the same results as they do, so I'm satisfied.

Now, my big gripe with EVERYONE is their reporting three significant figures in test results. Nobody out there can measure stuff accurately to 1 part in 1000. Most of them are lucky if they can hit 5%.

A perpetual problem in science, to be sure. (Did you know that the putative body temperature of 98.6 F is false precision from converting C to F? Took almost a century for medical scientists to notice.) But none of the magazines and web sites are helping!

pax / Ctein

It's amazing to see some real competition out there, other than the two big ones.

You have to give them kudos for this one. I've been looking at the samples posted and thought that both DR and high iso performance looked quite remarkable. Wish I could squeeze it into my Olympus.

@ Andre: I'm not that hard up and the manual focus...let's just say I'd rather do that on a rangefinder :-)

I do want to like Pentax, my first camera was a ME Super and it was great, but the failure to produce that 35-40mm prime for their cameras...it disqualifies them as a cameramaker that takes photographers seriously. Same with Nikon and Canon; they're catering for a customer I can't identify with. The only modern 35-40 eq. primes they make are for Full Frame and are as big as their zooms.

Luckily other companies have stepped in to fill the void and ditch the mirror in the process.

Hooray for Pentax! Going all the way back to 1969, one of the best choices one could make for lenses and price/performance was Pentax. I could have saved a system change or two if I'd started there (or with Nikon), and another if I'd skipped the switch to Olympus (I really wanted spot metering, especially their multi-spot. In practice it was of less benefit than I had thought. And I didn't see auto-focus coming or Olympus missing that boat, so that was a 7-year excursion of little value.)

Isn't that sensor made by Sony?

I'm still waiting for the Sigma SD1.

Please note, I am not dogmatic. With film, I used Pentax, starting with a Spotmatic. With digital, I switched to Oly, since they had the only dust buster at the time.

All of my digicams are Fuji. I want the X100.

Before the DXO announcement I had the K5 on my list of next body to purchase. Just seeing the RAW results of early production models I'm very impressed. I'm a bit disappointed that there has been no roadmap announcement of a 35-40 fast prime, but then I plan on buying the X100 as my street camera anyway. Problem solved...

yes for 35mm FX equivalent there is only the old and now discontinued (so eBay only) FA*24mm f2 lens.

Pentax would do well to introduce the successor to that one, and make it f2 again, forgoe the size for the speed, as 21mm f3,2 has small size already.

I have noticed a significant ramp up in Pentax's relative competetiveness verses Canon and Nikon since the K10D and since the Hoya takeover an acceleration of this ramp up. The latest the K-5, is just following this trend, perhaps has reached the summit in APS-C land :)
I am itching to get hold of one of these, I have been delayed a little by an impulse buy of a marvellous K50f1.2, seems I am still gripped firmly by Pentax lba instead of Pentax cba.
All I can say is Bravo to the Pentax and Sony engineers who have delivered a great little camera. Is this perhaps a modern day LX?

Mike, regarding your never-ending quest for a 35mm-e: I never thought I'd pull this card, but...with the great, nay, fantastic high ISO performance the K-5 is demonstrating, a lens such as the DA 21mm f/3.2 might no longer be "slow", and it's a 32mm-e on the Pentax sensor. It's also cute as a button :-)

I just got mine a few weeks ago and my oh my it sure is an awesome camera. No complaints that it took top spot either ;)

Ahh I see...

The "how big is your sensor" has become the new "how many pixels"...

just imagine if those photosites were on a full frame sensor. it would be 36mp.

body integral image stabilization, 100% viewfinder, weather sealing, body not much bigger than the k-5, 35mm f2 limited...i know what i want!

"Isn't that sensor made by Sony?"
"All I can say is Bravo to the Pentax and Sony engineers"

I thought the image chip was made by Samsung.

Anyone?

"Ahh I see...

The "how big is your sensor" has become the new "how many pixels"..."

Mine go to eleven.


About DXO -- Michael Reichmann just gave a rave review to the Panasonic GH2 based on field testing and eyeball examination of prints made from GH2 RAW files. DXO promptly weighed in with a test that shows the GH2 somewhat *inferior* to the GH1 in precisely the areas in which Reichmann thought the GH2 excelled.

Reichmann said in a discussion-forum comment that he simply couldn't explain the differences in what he observed subjectively and the DXO claims. Reichmann is an experienced field tester and print-maker, and it's difficult to suggest that he simply somehow was comparing apples to oranges, or made some fundamental mistake in his printing procedures, etc. It also seemed weird that Panasonic would put out a camera with a sensor notably inferior to a previous one. Wouldn't their own testing show that?

Then (perhaps contra Ctein?) another guy commented that DXO's ISO testing influences scores in ways that it shouldn't, and that simply by adjusting their own proprietary scoring system to a more conventional system, suddenly the GH2 scores better than the GH1.

So - no reflection on the Pentax -- I'm becoming a little more skeptical not of DXOs tests, but of their scoring system. It always seemed a little weird to me to give a multi-dimensional problem a single "score," but I don't know enough either of math, physics or engineering to sustain my skepticism.

JC

I have to admit I don't pay much attention to their overall scores, only to the evaluations of specific properties--which they now (unfortunately, in my view) call "Portrait" (color depth), "Landscape" (dynamic range), and "Sports" (high ISO). Too dumbed down for my taste--those names just muddy and obscure what they're talking about. We're not chimpanzees--we can understand "color depth," "dynamic range," and "high ISO."

Mike

I recently bought a K-5, and it's a wonderful little camera. Not just because of its awesome sensor, but because of pretty much everything, really. It sort of has me wondering what they could possibly improve on for the next generation of cameras (I'm not just talking about Pentax here). Aren't things way beyond good enough by now? Perhaps we're going to be seeing a lot more cameras that follow the ethos of 'less is more' now that more is no longer more.

Dear Mike and JC,

I don't pay attention to the "scores" either. They're just dumbed-down derivatives of the real data. Not worthy of attention and certainly not of the time and energy to argue about them.

pax / Ctein

"I thought the image chip was made by Samsung."

The sensors of the K20D an the K-7 were. The K-5 has a Sony "Exmor" sensor and thank Heavens this miserable Korean intermezzo is over.

That Samsung sensor was so rotten that Pentax couldn't even afford to let us switch off the camera's noise reduction by darkframe subtraction (DFS) at exposure times beyond 6 sec (K20D) or 30 sec (K-7). This made both models practically unusable for lightning, fireworks and astro photography.

I've got my K-5 on Thursday and it's by far the greatest leap between any two camera generations I've ever seen. The K-5 is the camera the K-7 should have been right from the beginning.

Ralf

Okay.
Here it goes.
My point when I did address this stuff to Mike was that, regardless of the comparative stuff, there is something happening underneath, and does not take into account which measurement or system comes over.

Thing is, sensor technology and its implementation is jumping leaps and bounds regardless of the metrics used, and at some point, and subjective evaluation aside, larger sensors are becoming sort of irrelevant, and APS-C size sensors SEEM to be the sweet spot for photography AS A SYSTEM [pardon my uppercase, not shouting, just stressing the fact].

They do seem to be the better system, and that includes the fine balance between the sensor, imaging engine and the rest of parametrics related to that system have started to come together in quite a harmonious way.

I honestly do not care if DxO is wrong, or whatever analysis website is wrong. Thing is that they do compare cameras under the same circumstances. And under those circumstances some are ahead, some behind. That will happen with ANY comparative system, even for EuroNCAP for cars, or "howfastIcantype" typewritters´ comparative sheets of some decades ago.

The other thing I see compared is if some RAW date is massaged or not. Which again I do not care. The whole rest of the digital photo workflow has become so outrageously convoluted [who knew at the begining it will become so, rather than get simplified] that what happens between the front glass element of the lens and the Adobe Lightroom file has nothing to do with me as far as I am concerned. I could even have some skilled sketcher fastly painting the scene over the sensor that I wouldn´t care had he or she be fast and accurate enough.

Meaning that I do not care if the RAW data is massaged with noise reduction or not. That doesn´t make the camera better or worse.

I need a good output, full stop.

Printing has become so difficult that it is a science on its own, and has not changed for the last 30 years, something that I do not understand why has it happened, and why the printer does not read on advance my thoughts according to what I do on my computer usually.

So all in all, APS-C [please, PLEASE, that is the name of that format, no DX or FX or whatever, that is a commercial name solely for Nikon cameras] seems to be the very sweet spot for digital cameras regarding SIZE [mirrored or mirrorless cameras], speed and output quality and dark situations.

It seems to allow great compact professional grade cameras, such as the D300s, K5, 7D, Alpha 700, E5 [no, the D7000 is a partly professional camera in that sense, as the K20D was, as the "magnesium alloy" is just a cover and does not include the top or bottom plates of it, as far as I can remember], rugged and handsomely fast due to buffer size and file size balance.

It seems to allow great compact cameras, such as the Samsung and Sony efforts in that regard.

And they can actually have petite lenses [NOT pancake lenses, there are very few Tessar type lenses out there of modern construction].

Who would have ever thought that a half frame camera could go up to 51000 ISO, or have a rather usable 25000 ISO output some years ago?

As a matter of interest, how would something like Velvia or Portra stack up against these sensors? I appreciate that it's not an apples for apples comparison. But film too has DR limits, quality degradation with increasing ISO, etc.

I have never generated an image with my Nikon D200 that begins to approach a well-exposed Velvia 50 image in impact (I use a lightbox for looking at my old slides). I'm sure that the real reason is that I have never invested £500 in a high end piece of software, or a lot more than £500 of my time to learn it. The D200 sensor is, I am sure, capable of generating RAW files that could with a lot of effort replicate the look and feel of Velvia,

Ah, the perfect fastish 35-40mm equivalent on APS-C prime quest. I've been using the Nikon 24mm f/2.8D on a D90 for a while now. I'm pretty happy with it, but I have this nagging feeling that everything would be better with that dreamy little designed-for-digital Pentax DA 21mm pancake. These new K-5 test numbers aren't helping. (The new Nikon 24mm f/1.8 isn't in the running: too big, too heavy, too expensive, too ridiculously everything.)

Misha wrote:
I thought the image chip was made by Samsung.

Anyone?

Misha, the K-5 sensor is manufactured by Sony. The sensors in the K20D and K-7 were manufactured by Samsung. All of Pentax's previous cameras used sensors by Sony.

"I thought the image chip was made by Samsung."

The last Samsung sensor was in the K-7, Pentax seems to have switched to Sony since the K-x and stated that the K-5 was a Sony sensor at Photokina.

Anyways, it's pretty clear that Sony is investing a lot of research into their Sensors and it would be pretty interesting to see this one expanded up to FF. In the meanwhile, I'm extremely happy with my K-5.

Sometimes you see comments like some of the above that imply that it is the sensor manufacturer, and not the camera manufacturer who is responsible for the good (or bad) performance of a camera. The sensor manufacturer is clearly critical.

However, did other manufacturers' products perform exactly the same as the Pentax K-x when using the same sensor? From what I saw, there were tons of raves about the outstanding low light performance of this relatively inexpensive camera, and not a lot of similar reviews about other cameras.

Was the K-x the only camera using that sensor? Or did Pentax do something in between the sensor and the photograph that made a difference?

Certainly great news for those of us who use Pentax. Although I am always uneasy buying a digital camera with full knowledge that it will be replaced in a year or two by an improved version. I suppose I could consdier the cost of a camera every few years replacing the cost of film and processing. It would be a wash if it weren't for the necessary computer costs which accompany digital. Speaking of obsolescence on the horizon, Kodak announced a 29 MP 35mm sized chip:
http://www.kodak.com/global/en/business/ISS/News/pressReleases/archive/2010/pr2.jhtml
I wonder how long before it's incorporated into a newer, higher resolution camera. Too bad Pentax painted themselves into an APS-C corner with the DA line of lenses, or it might appear inthe K4.

One of the interesting DxO graphs is the all sensors comparison, setting image quality against price. In that one you get an average function of quality/$, and the K5 stands way out in terms of bang for the buck.
I'd love a lens equivalent to the panasonic 20/1.7 in the lineup. The K5 with such a lens in the Limited build quality and APS-C size with be amazing.
For those aching for a full frame model, It would be a shot in the foot for Pentax. Too many expensive lovely APS-C primes on which they spent most of their recent R&D would be useless with such a camera, actually pressing towards buying older lens designs, not a good strategy at all, in my opinion. The Canon or Nikon lineups always kept their top notch lens manufacturing for their full frame cameras, it's a very different approach.
I believe with proper lenses you can get a lot out of high quality APSC sensors, the deep DOF can be quite helpful, but still there's need for fast lenses for that specific kind of image.

"I thought the image chip was made by Samsung."

K20D and K-7 sensors were by Samsung. K-x, K-r and K-5 sensors are by Sony. I expect Nikon D7000 and Sony A580 housing the same or a very similar Sony sensor to do very well on DxOMark as well.

Adam Maas:

Thanks for that post. It makes sense to me.

Kudos to Pentax for the K-5! FF will certainly jump ahead in their next iteration, but APS-C is progressing quite nicely and will soon be considered "good enough" for all but the demanding situations.

I'm hoping that someday, all cameras will be sufficiently good to be essentially a wash. Maybe, just maybe, attention will shift back to the part of the equation with the most variance - the photographer.

I know...wishful thinking. But we can dream, can't we?

Hi, gents.
I'm not sure if anybody will read a message that goes too that low and delayed, but nevertheless.

1. K-5 scores #1 out of DxO marks:
well... For eras Pentax bodies used to be nothing special but just good bodies for Pentax lenses that are smth special comparing to mass market. Now Pentax enters the area occupied but #1 (C) and #2 (N) photo brands and being a pentaxian I'm a bit worried: damn, what's next? Should Pentax happen to follow #1 and #2 and release DA* (24)28-70/2.8 and 70-200/2.8 - I quit. Pentax must be [a bit] behind of #1 and #2 in terms of AF and high iso performance but get grand prix with it's glass - or it is not The Pentax.

2. Andre, Mike and others keened on fastish 35-40mm's: could I please bother you asking to show or describe a scene where you do need f1.4-2.0 aperture? A scene where you do need that fast glass and where f2.8-3.5 does not work. Once again, we're talking about 35-40mm effective (converted to 36x24mm frame) focal range, aren't we? I do understand 85-105/1.8 and 200-300/2.8 but I can't see any real need for fast wide glass. What am I missing?

Yours,
Zig

P.S.
Iñaki,
you can use bold and italics instead of CAPS.

with the K-5's new chart topping low-light ISO ability, the absolute need for large aperture lenses goes down quite a bit. I'm sure similar things were said when the D700 was first tested.

Suddenly the size and expense of the DA* line of lenses aren't quite as necessary, and the DA Limiteds start making even more sense.

"That's my big problem. Speaking not as a reviewer but just as me. I've been writing about that particular problem at least since August of 2008."

And mentioned that when Pentax announced their entry level 35mm a while ago. But what's exactly wrong with K28/2.8 - or the like? I find it hard to believe it's lack of AF that you can't bear...

If i was starting fresh, the K-5 would be my next camera, but as I have all the Nikon glass staring at me, it's got me quite excited about the D7000, as it's very likely sharing the K-5's base sensor. At least it seems that we're at a point where 35mm cameras got to with the Nikon F100(or for me, the EOS A2) - good enough for just about anything and enough staying power that it just might not be disposable after 2 years. It is very nice, tho, to see Pentax having a camera worthy of their limited primes.

In reply to Adam Mass, I quote Thom Hogan: "Nikon committed to using only Nikon-design sensors. That doesn't mean that the sensors don't have elements from other makers in them or that they stopped using Sony as a supplier, only that Nikon's sensors appear to now be 100% exclusive to Nikon."

So the K5 almost certainly doesn't use the same sensor as the D7000. FWIW.

I can heartily recommend the Sigma Super-Wide II AF 24mm/2.8 as a good "35mm equivalent" for Pentax. This lens is a stalwart member of my prime lineup, and I am always pleased with the image quality it delivers. It's also perfectly sized and balanced on the small Pentax camera bodies.

Having received my Pentax K-5 last week, I can also say that it is simply the best camera I have ever owned, film or digital, Pentax or Nikon. Solid. Functional. Fast. Great pictures. Takes all of my K-mount and M42 glass. I'm happy!

"But what's exactly wrong with K28/2.8 - or the like? I find it hard to believe it's lack of AF that you can't bear..."

Lukasz,
I actually would prefer manual focus, but I've had some problems with my right eye, and I'm not so good at focusing manually at the moment...good thing there's AF as an alternative.

Mike

DxO marks aside, check this out (make sure to scroll down to see the better processed and exported image, not just a screen shot):

http://www.pentaxeros.com/forum/index.php?topic=37916

This is the first example I have seen of extreme shadow, er... darkness recovery. Guys at PDReview and PentaxForums.com keep posting new stuff, such as:

http://www.pentaxforums.com/forums/pentax-dslr-discussion/121337-k5-low-iso-noise-not.html

Regardless of their approach and methodology it seems that DxO mark score for K-5 is fairly accurate.

It's about time someone finally knocked the D90 off it's APS-C pedestal-- for two years that camera ruled the rankings. Good for Pentax! Now if they would just lower that price a bit...

Zig,....

The lack of a large aperture moderate wide,..ie a 35mm e, is preventing me from going any further with Pentax. In over 10 years of making DSLR's Pentax has only managed ONE moderate wide for APS-C,..the 21mm f3.2 Leica considers teh moderate wide as a vital focal length and has had up to 5 lenses of 35mm focal length in it's list at the same time!

For me, such a lens really does need to have a max aperture of at least 2.8 and preferably larger, both for accurate focus when not using AF (most of the time for me) and for use in low light simce the max aperture is effectively the viewfinder brightness and essential for composition etc.

....for me, no 24mm f2 then no pentax...

Aren't numbers fun?

But I'm in agreement with MJFerron's comment. I realize many photography fans enjoy comparing gear as much or more than actually using it, but I'm too busy with a job and family to have much time for online number crunching. I really don't know how my two year old E-510 stacks up numbers-wise with the current crop of DSLRs (probably not too well) because I've never checked the DXO site. Despite what the stats may say, I'm well pleased with the results I get with my camera, and I'd rather spend my limited free time making photographs. At the pace new models are introduced, no camera stays at the top of the heap for more than a few months. I say go out and learn to make the best use of the camera you already own, and don't let online test envy rob you of enjoying what's already right there in your camera bag.

"I can heartily recommend the Sigma Super-Wide II AF 24mm/2.8 as a good "35mm equivalent" for Pentax."

+1 Great lens.

"2. Andre, Mike and others keened on fastish 35-40mm's: could I please bother you asking to show or describe a scene where you do need f1.4-2.0 aperture? A scene where you do need that fast glass and where f2.8-3.5 does not work. Once again, we're talking about 35-40mm effective (converted to 36x24mm frame) focal range, aren't we? I do understand 85-105/1.8 and 200-300/2.8 but I can't see any real need for fast wide glass. What am I missing?"

+2 Also curious about this.

"don't let online test envy rob you of enjoying what's already right there in your camera bag"

Good advice John. They're all still just cameras. I might go a little further than you, and note that people who chase an elusive "best" based on measurements are usually disappointed that it seldom imparts special qualities to their pictures.

(Of course, that isn't stopping me from being interested in the K-5....)

Mike

D7000 figures are up - 80 vs. K5's 82.

2. Andre, Mike and others keened on fastish 35-40mm's: could I please bother you asking to show or describe a scene where you do need f1.4-2.0 aperture? A scene where you do need that fast glass and where f2.8-3.5 does not work.

I just got back a roll of photos taken in available darkness at a bar and at a bowling alley. In the former, banging away with flash would've ruined the mood, in the latter it's disallowed because it distracts the bowlers. In either case I likely wouldn't have been able to get any usable shots at all if not for my f/2 prime. In both cases I still would've preferred a couple of stops more headroom over the ISO1600 color neg I was using--I wanted more DOF in the former and it would've been nice to have some chance of freezing motion in the latter.

This isn't everyone's style of photography, but it is what I like to do. I was using an RF this time, because that's what I had, but I do the same work with a smallish DSLR and a fast prime.

@Rob: The D3100 definitely seems to be a Nikon sensor, I'm unsure about the D7000 (which is why I note it may share the K-5's sensor), the specs are quite similar overall although the Nikon versions of recent Sony sensors are significantly different than the Sony or Pentax versions (totally different output chain, Nikon uses classic channelized output, Sony and Pentax use Sony's per-line on-sensor ADC's).

@Siegfried: As someone who shoots a lot in very low light f2.8 simply isn't enough to get usable shutter speeds. I regularly run out of shutter speed at f1.4 and ISO 12,800. But I shoot in extreme situations (dim bars shooting candids, so I need 1/60 or better for a 50/1.4). A fast 24 would make my day (I'm currently drooling over the Zeiss ZA 24/2 for my A33, having recently switched to Sony from Pentax due to the EVF of the SLT's which is a huge benefit for ultra-low-light work)

Siegfried raises a good point about the extent to which one truly needs a "fastish" 35-40mm-equivalent lens for Pentax APS-C. There are choices available. Pentax currently offers a 21mm f/3.2 (32mm-equivalent) and a 31mm f/1.9 (46mm-equivalent).

The 21mm is a bit wider than you have in mind, but with 16MP to work with one can afford to crop a bit if necessary. The maximum aperture of f/3.2 is only a third-stop slower than f/2.8, so that shouldn't be a major deal-breaker either, especially on a camera body that has built-in IS and can shoot comfortably at ISO 1600. If you really want that extra stop of light, it's available on the 31mm f/1.9. Just be prepared to pony up $1000 for it.

Or you could continue to wait. You could be waiting for quite a while though. Your original complaint was made back in 2008 and so far neither Pentax nor any other camera manufacturer seems in any big hurry to produce a 24mm or 28mm f/2.0 prime for their APS-C DSLRs.

Hi bob,
There are other alternatives, but there is a 24 f2 for pentax which can be had both online [via Ebay] or on many local stores:
the FA* 24 f2

On the other hand, and contrary to Cosina believing, you might want to order any CZ [CosinaZeiss] or CV [CosinaVoightlander] manual focussing lenses, or get the 31 Limited for that intend.

Can´t get the überstepfurther, as it is an exclusive:

http://www.cosina.co.jp/seihin/voigt/v-lens/micro-ft/index.html

[S***t, that is the first specialty dedicated lens I see for micro 4/3 mount, and with an aperture of 0.95 none the less!]

Inaki,..

Yes, I'm aware of the (old)pentax 24mm f2 FA*....however, since I live in the UK it's more or less unobtainable on Ebay UK and "local Stores" who actually remember Pentax are very thin on the ground. I have been trying for 3 years to get this lens, so far without luck.

The 31mm is effectively a 46mm e....

"Pentax currently offers a 21mm f/3.2 (32mm-equivalent) and a 31mm f/1.9 (46mm-equivalent)."

Gordon,
I think you have semi-inadvertently put your finger on the problem. If what we're looking for is something in the 35-40mm range, and the system has a gap between 32mm-e and 46mm-e...well, then that means there's a void smack-dab where you don't want there to be a void, right?

I'm looking into the Sigma, though. Never had much luck with Sigma stuff, but I'm willing to (struggling to?) remain open-minded.

Mike

John,
fully agreed! Absolutely, entirely, completely. From all of my shots ever taken I recall only a single one that was ruined only due to the limits of the gear used - but it was very very special shot: winter night, a lonely road and there was smth falling down from the sky that caused a very special affect to happen - every single source of light (high-beam from a car in two miles away and such) to create a beam like there was UFO looking for touch down. And there were almost black trees covered with white snow around me and the road had snowbanks around... I did my best PP'ing that shot taken with Oly SP570UZ but whatever I did it turned out to be just a straight white line (that very UFO light beam) in a black noisy rectangle (well... the surroundings).
I'm almost ready to move to K-5 from my oldish *ist DS mainly to gain some of ease of use. Built-in IS will help me to save on tripod that I too often forget to take with me, high DR will save the shot if camera's metering fails and highlight clipping occurs (and when B'n'W isn't the way to cause of the scene), and on, and on. Will I shoot better with K-5? No. Well, I mean to say that sooner or later I will, but it'll be my own achievement not K-5's. And most probably I'll gain more and in less time with current istDS rather then with K-5.

Peter (+Adam),
I think I got your point, but I'm using another approach. Maybe 'cause I'm 100% digital.
If you've got a scene and you're to take a photo of it then you've got the following controls:
- focus to... well, to focus
- focal length and distance-to-subject to control the perspective
- shutter speed to control the motion
- aperture to control the depth of field
- sensitivity to control the light
Since you appear to stuck at analogue (film) photography it may make sense to control the light also thru the aperture since you can hardly move higher then 1600ISO. But from my perspective it's not the right way to go. You need more light? twist the shutter dial, go BULB if needed. Though this isn't the right way either :)

As for that shallow DOF that a fast glass gives you - I just checked pentaxforums.com and found some shots a chap did on wedding shooting and those 16/2.8 shots he published do not look that impressive to me. I mean they're ok but I don't see any need to go to f2.8 there.
Perspective separation, bokeh? nah... Those shots would be at least the same if not better if shoot with f4, imo.

I'm referring the review by Urkeldaedalus here:
http://www.pentaxforums.com/forums/pentax-digital-slr-camera-reviews/114233-pentax-k-5-a.html


Lemme put it this way: 24/2 lens appears for me to be a pure math object just like 50/1.2 or a point particle. Or like 300hp FWD car.

Yours,
Zig

"... especially on a camera body that has built-in IS and can shoot comfortably at ISO 1600 ..."

I shot this at ISO 20000 on the K-5: http://grumpyjohns.com/files/gallery/180/07.jpg

Some noise reduction and high pass sharpening applied, I'm not a pro and hardly an expert at making an image top line, someone with much better skills would turn this into a razor sharp image. Net effect, I wouldn't even blink at shooting at ISO 6400, I know I'll get a usable image at much higher ISO than that.

@Siegfried:

I'm talking f1.4 and ISO 12,800 on digital, not ISO 1600 on film in those situations at 1/60 or even 1/30 (note I'm shooting people in dimly lit bars, so I need the shutter speed to freeze them). You can't shoot people reliably below 1/60 or so and below 1/30th I'm tossing 80-90% of the shots. And flash is an absolute no-no, you'll blind and annoy everyone. I've heard the same point you make many times, and every single time it's from people who've never shot in the same situations I do. These aren't composed scenes but very low-light candids and for this, every bit of ISO and aperture I can get matter. The K-5's improvement in High ISO performance would make f2 viable but F2.8 simply isn't there.

And to the natural objection of 'what did I do before the current high ISO monsters' the answer is before ISO 6400 became practical, I simply didn't get more than the occasional usable shot. Modern cameras made this sort of photography practical to everybody who wasn't doing obscene things to Tri-X in terms of pushing.

Note also that maximum aperture affects focusing in lower light as well. I can focus with an f1.4 lens in light where a f2.8 lens will be very difficult to focus. This is as big a deal as getting the shutter speed up to needed levels and one which cannot be offset by improved high ISO performance.

If I'm not shooting people I'm almost always somewhere where ISO 6400, f2.8 and a shutter speed of at least 1/30 is achievable (IE outdoors in the city). Body-IS makes this practical even handheld.

Even when shooting B&W film in these situations I'm pushing to 3200 or 6400 (I don't do much of that anymore, it's the one spot where film simply cannot touch digital).

DoF isn't an issue for me, the tiems I need that sort of shallow DoF I pull out my Mamiya 645 and 80/1.9 and shoot slow film wide open. That nets me DoF that's nearly impossible to shoot practically with 35mm.

Just fyi, KEH.com currently has two EX+ condition FA* 24 mm f/2 lenses available for purchase. I hope the people of this thread can buy them so I don't have to keep staring longingly at them all the time.

This said, I agree with Mike that a new fast wide between 20 mm and 24 mm would really help the Pentax lineup. If they came out with that I would be much less tempted by a full frame camera. (HINT... HINT... PENTAX)

What Adam said... you can't open up to bulb and hold the shutter open when you're trying to get a photo of someone across the table shooting the person next to them a quirky smile over a pint of something. (Working from across a standard-size table also makes you appreciate the whole 35-40mm eq. thing, but that's a whole other discussion.)

And just to reiterate what some others have already mentioned, the other thing about fast glass in this environment is the viewfinder brightness--you can't compose if you can't see. This is one huge advantage of an RF, film or no. That being said, if ever I find I can afford an M9, don't do anything to wake me up, because I am clearly dreaming... :-(

The comments to this entry are closed.