samfan

Joined on Sep 20, 2012

Comments

Total: 696, showing: 1 – 20
« First‹ Previous12345Next ›Last »
On article Pentax KP Review (493 comments in total)
In reply to:

samfan: It's just gorgeous. Pentax nailed what Nikon was trying to do with the Df and screwed up.

Well, almost. I with they'd finally upgrade their 4-way controller to something more modern and that fat border around the LCD is a bit odd.

Either way K1 and KP seem like a killer combo as far as camera bodies are concerned, at least if you like the retro-ish SLR looks like I do.

Heh, this has to be the first time I really like a camera for the looks and barely think about the insides. Okay, second time as I had the just as gorgeous Pentax i-10 compact at one point too. But this one is certainly also a decent camera on its own.

@vscd
Well this IS a DSLR, so it won't suck batteries when not using it.

Anyway I've never quite understood the problem with low battery life in cameras. Sure, spare batteries aren't free but there are good cheap aftermarket ones available. No need to buy overpriced originals.

Carrying them shouldn't be a hassle (unless you need a dozen or so), I mean if you carry a ILC you need a bag and accessories anyway don't you? And even if travelling very light, just pop a spare in the pocket or such.

I also don't quite agree with the point that swapping batts can mean lost shots. Technically it's true but that's just as true for doing *anything* with the camera or not having the camera fully ready.

I used to shot weddings with Nikon D200 which could do what, 200 shots on average? I survived. Better battery life is better but I don't thin it's much of a big deal.

In fact, compare 390 shots to a smartphone which dies within a few hours of shooting at most.

Link | Posted on Mar 29, 2017 at 21:23 UTC
In reply to:

samfan: Foveon ftw.

I really don't understand what's all the excitement about bayer filtering. Sure it was a nice idea but it should've been replaced with something like Foveon ages ago. It hasn't only because it's the cheapest and simplest hardware solution.

Software solutions are cheap, so all those inherent problems such as low color resolution, aliasing and all that can be done outside of the sensor.

No, bayer filtering is not good. You appreciate it only if you've not seen anything better. Foveon provides better resolution and no aliasing. Native b&w sensors such as in Leica M mono blow bayer out of the water with monochromatic resolution and sensitivity.

Let's call it what it is. Bayer filtering is nothing but a compromise between affordability and acceptable quality. It made it possible to produce color cameras in mass quantity and I guess the idea itself is kind of genius but it's outdated as hell.

(And yes I know Foveon has its own problems. I still prefer that kind of solution.)

And, I guess it is true that with enough resolution, bayer filtering does begin to lose the disadvantages. Indeed maybe 100 Mpx is enough. Even 24 Mpx is good enough for most people, even maybe for me (given regular print and display sizes) to get those results I'm talking about. (Not talking about pure raw spatial resolution, but the feel of the image.)

Then again I don't want to work with 100 Mpx images. I'd rather have a 10-20 Mpx Foveon (talking flat resolution here) and had the nice results 'naturally' not by scaling down ridiculous resolutions. Though I guess that's just my own preference.

Link | Posted on Mar 29, 2017 at 20:43 UTC
In reply to:

samfan: Foveon ftw.

I really don't understand what's all the excitement about bayer filtering. Sure it was a nice idea but it should've been replaced with something like Foveon ages ago. It hasn't only because it's the cheapest and simplest hardware solution.

Software solutions are cheap, so all those inherent problems such as low color resolution, aliasing and all that can be done outside of the sensor.

No, bayer filtering is not good. You appreciate it only if you've not seen anything better. Foveon provides better resolution and no aliasing. Native b&w sensors such as in Leica M mono blow bayer out of the water with monochromatic resolution and sensitivity.

Let's call it what it is. Bayer filtering is nothing but a compromise between affordability and acceptable quality. It made it possible to produce color cameras in mass quantity and I guess the idea itself is kind of genius but it's outdated as hell.

(And yes I know Foveon has its own problems. I still prefer that kind of solution.)

Okay, so a high-end, latest 100 MPx MF sensor by the behemoth that is Sony may beat a APS-H Foveon sensor designed by a tiny company that's already a few years old.

What would happen in an alternate reality where Sony instead of developing bayer designs and traditional sensors for the last few decades, instead had made the same developments with Foveon-like sensors?

These days it's kinda difficult to make apple to apple comparisons because the existing sensors simply aren't comparable by size and development resources.

I for one still fondly remember my Sigma DP1. While technically, in resolution and other areas the 4 Mpx sensor was comparable to 10-16 Mpx APS-C bayer sensors with the same size, the natural feel of the images was really unbeatable. Of course, sadly, Sigma doesn't have the resources to keep up with Sony. Still, yes i think their Quattro cameras beat comparable (price, sensor size) cameras to the ground in certain aspects.

I just wish for a new generation.

Link | Posted on Mar 29, 2017 at 20:40 UTC
In reply to:

samfan: Foveon ftw.

I really don't understand what's all the excitement about bayer filtering. Sure it was a nice idea but it should've been replaced with something like Foveon ages ago. It hasn't only because it's the cheapest and simplest hardware solution.

Software solutions are cheap, so all those inherent problems such as low color resolution, aliasing and all that can be done outside of the sensor.

No, bayer filtering is not good. You appreciate it only if you've not seen anything better. Foveon provides better resolution and no aliasing. Native b&w sensors such as in Leica M mono blow bayer out of the water with monochromatic resolution and sensitivity.

Let's call it what it is. Bayer filtering is nothing but a compromise between affordability and acceptable quality. It made it possible to produce color cameras in mass quantity and I guess the idea itself is kind of genius but it's outdated as hell.

(And yes I know Foveon has its own problems. I still prefer that kind of solution.)

Ian: First, for use in MF there would still be need to develop the sensor in the first place. which would be costly. Also I believe Sigma said their Foveon design wasn't suitable for FF so that would need to be developed further, but that would mean more costs. Even with a 20k camera you need to stay competitive.

But good point though, stacked sensors would be wonderful for MF if they were done well.

Spectro: but understand that you're talking about cameras from a tiny manufacturer for whom cameras aren't even a primary concern and who develop their own sensor technology. Those aren't inherent flaws that couldn't be overcome if given enough attention and resources. Such as they are, Sigma Foveon cameras are specialty tools but given the results they are able to provide, I'd highly welcome similar technology being developed further by bigger players and used in mainstream products.

Link | Posted on Mar 29, 2017 at 18:28 UTC
In reply to:

samfan: Foveon ftw.

I really don't understand what's all the excitement about bayer filtering. Sure it was a nice idea but it should've been replaced with something like Foveon ages ago. It hasn't only because it's the cheapest and simplest hardware solution.

Software solutions are cheap, so all those inherent problems such as low color resolution, aliasing and all that can be done outside of the sensor.

No, bayer filtering is not good. You appreciate it only if you've not seen anything better. Foveon provides better resolution and no aliasing. Native b&w sensors such as in Leica M mono blow bayer out of the water with monochromatic resolution and sensitivity.

Let's call it what it is. Bayer filtering is nothing but a compromise between affordability and acceptable quality. It made it possible to produce color cameras in mass quantity and I guess the idea itself is kind of genius but it's outdated as hell.

(And yes I know Foveon has its own problems. I still prefer that kind of solution.)

Bayer survives because it's cheap and "good enough". Anything else proved to be expensive to develop and produce, difficult to implement and at the end went unappreciated anyway.

Despite all that, Foveon manages to provide results that beat bayer designs to the ground for some purposes. Sigma by no chance has the resources and will to turn it into a multi-purpose mainstream solution. It would take someone like Sony or Canon (or Apple) to really dig into it and produce something that's all-around better and with comparable price.

It's like with CCD/CMOS. Nobody wanted to use CMOS sensors in higher-end because the technology was unproven and results mediocre for those purposes. It was only Canon that made it work after the invested a lot of resources into developing the technology.

And yes obviously bayer is not super-simple either, the algorithms these days are really complex but they evolved gradually from a pretty simple idea.

Link | Posted on Mar 29, 2017 at 17:56 UTC
In reply to:

samfan: Foveon ftw.

I really don't understand what's all the excitement about bayer filtering. Sure it was a nice idea but it should've been replaced with something like Foveon ages ago. It hasn't only because it's the cheapest and simplest hardware solution.

Software solutions are cheap, so all those inherent problems such as low color resolution, aliasing and all that can be done outside of the sensor.

No, bayer filtering is not good. You appreciate it only if you've not seen anything better. Foveon provides better resolution and no aliasing. Native b&w sensors such as in Leica M mono blow bayer out of the water with monochromatic resolution and sensitivity.

Let's call it what it is. Bayer filtering is nothing but a compromise between affordability and acceptable quality. It made it possible to produce color cameras in mass quantity and I guess the idea itself is kind of genius but it's outdated as hell.

(And yes I know Foveon has its own problems. I still prefer that kind of solution.)

Also it's worth pointing out that there have been different color arrays used.

Canon I think used CMY color filter back in the day, Sony had an RGB and Cyan (I think) filter after they acquired Konica Minolta's camera division, and there was RGBW by Kodak I believe? With 1/4 of the pixels being monochromatic.

One major issue with bayer fltering is how much light it throws away. At some point, late 2000's I believe, when CMOS sensors became more prevalent in DSLRs (which are/were inherently noisier at the same sensitivity than CCD), camera/sensor makers tried to make up to it by using weaker color filters.

It increased the sensors' sensitivity but also made the color suffer. Which is one reason why when the switch to CMOS happened, people noticed the new sensors don't provide such bright, vivid colors (combined with the necessary on-chip noise reduction of CMOS sensors).

Link | Posted on Mar 29, 2017 at 17:24 UTC

As a side note, DPR. Could you please not use those asterisk'd notes below the article? One has 2 choices, either scroll down to read the explanation and then scroll up, having to find the point where the article continues, or finish reading and then get to the notes, forgetting the context in the meantime.

Just put them somewhere in separate boxes in the article, or make them appear on hover/click or such.

Link | Posted on Mar 29, 2017 at 17:18 UTC as 18th comment | 2 replies

Foveon ftw.

I really don't understand what's all the excitement about bayer filtering. Sure it was a nice idea but it should've been replaced with something like Foveon ages ago. It hasn't only because it's the cheapest and simplest hardware solution.

Software solutions are cheap, so all those inherent problems such as low color resolution, aliasing and all that can be done outside of the sensor.

No, bayer filtering is not good. You appreciate it only if you've not seen anything better. Foveon provides better resolution and no aliasing. Native b&w sensors such as in Leica M mono blow bayer out of the water with monochromatic resolution and sensitivity.

Let's call it what it is. Bayer filtering is nothing but a compromise between affordability and acceptable quality. It made it possible to produce color cameras in mass quantity and I guess the idea itself is kind of genius but it's outdated as hell.

(And yes I know Foveon has its own problems. I still prefer that kind of solution.)

Link | Posted on Mar 29, 2017 at 17:14 UTC as 19th comment | 11 replies
On article Ming Thein joins Hasselblad as Chief of Strategy (241 comments in total)

Huh. A camera company hires a photographer. Why does the idea sound so novel even if it's the most logical thing in the world?

Hey, Nikon, you hear that?

Link | Posted on Mar 29, 2017 at 15:28 UTC as 14th comment | 3 replies
On article Pentax KP Review (493 comments in total)
In reply to:

samfan: It's just gorgeous. Pentax nailed what Nikon was trying to do with the Df and screwed up.

Well, almost. I with they'd finally upgrade their 4-way controller to something more modern and that fat border around the LCD is a bit odd.

Either way K1 and KP seem like a killer combo as far as camera bodies are concerned, at least if you like the retro-ish SLR looks like I do.

Heh, this has to be the first time I really like a camera for the looks and barely think about the insides. Okay, second time as I had the just as gorgeous Pentax i-10 compact at one point too. But this one is certainly also a decent camera on its own.

@vscd "CIPA 390? Nailing nothing in the field..."
Um how many attempts do you need to nail one shot?

@Leandros S "if you put a flash with GN6 on the camera (as in the KP) rather than GN10 (as in the K-3), you'll get a better CIPA score."
Dang! Good point. I've been wondering why camera makers use such pathetic, useless flashes. That's probably one reason. I really don't use the on-board flash at all, BUT it's useful to have as a last-resort backup. But for that it has to have at least some use, which the weakest units actually don't. Indeed I'd prefer a no-flash battery rating (even though I don't care about battery rating either, but at least it wouldn't be skewed so badly).

@nicolaiecostel
The Df looks silly because the front and back are like from 2 completely different cameras, especially on the silver version. KP looks much more consistent overall (even though the LCD is odd). The basic exposure controls are also poorly thought out on the Df, which is something Pentax excels at.

Link | Posted on Mar 28, 2017 at 09:01 UTC
On article Pentax KP Review (493 comments in total)

It's just gorgeous. Pentax nailed what Nikon was trying to do with the Df and screwed up.

Well, almost. I with they'd finally upgrade their 4-way controller to something more modern and that fat border around the LCD is a bit odd.

Either way K1 and KP seem like a killer combo as far as camera bodies are concerned, at least if you like the retro-ish SLR looks like I do.

Heh, this has to be the first time I really like a camera for the looks and barely think about the insides. Okay, second time as I had the just as gorgeous Pentax i-10 compact at one point too. But this one is certainly also a decent camera on its own.

Link | Posted on Mar 27, 2017 at 21:03 UTC as 56th comment | 9 replies
On article Re-make/Re-model: Leica Summaron 28mm F5.6 Samples (201 comments in total)
In reply to:

samfan: I like it. Yes, it certainly wouldn't score high in synthetic tests and in fact isn't impressive even to the eye from a technical perspective. The flaws are apparent. But the lens has character, just like many of the primes of old designs.

I wish more camera/lens companies would go back to their older designs and just update them with modern tech such as better AF (or any AF in fact) or coating. While I respect modern lens designs, the images usually don't talk to me in a way like many of these older lenses do.

And I don't mean just Leica primes or other overly expensive or antique pieces. Some of the older Sigma lenses for example I really loved before they went all high-end on us. Older Nikon lenses also, even zooms.

These days it's all about sharpness and other technical merits but the images provided by the lenses are sterile and lifeless. I used to enjoy picking one lens over another just based on certain characteristics.

This photo for example
https://2.img-dpreview.com/files/p/TS5976x3984~sample_galleries/8329529667/8142963012.jpg
really suits the lens. Taking a photo of such a subject matter with a modern, ultra-sharp fully-automatic overly complicated lens is kind of... Weird I think?

And yea sure you can add all the lens faults in post or might as well just add grain and all the effects and filters to make it look vintage, but there is certain specific fun to walk around desolate buildings and areas and use a camera/lens combo that is as much low-tech as possible (i.e. even if using digital is unavoidable, a small manual lens with old design and an optical viewfinder still fit right).

Just my 2 cents anyway!

Link | Posted on Mar 24, 2017 at 14:40 UTC
On article Re-make/Re-model: Leica Summaron 28mm F5.6 Samples (201 comments in total)

I like it. Yes, it certainly wouldn't score high in synthetic tests and in fact isn't impressive even to the eye from a technical perspective. The flaws are apparent. But the lens has character, just like many of the primes of old designs.

I wish more camera/lens companies would go back to their older designs and just update them with modern tech such as better AF (or any AF in fact) or coating. While I respect modern lens designs, the images usually don't talk to me in a way like many of these older lenses do.

And I don't mean just Leica primes or other overly expensive or antique pieces. Some of the older Sigma lenses for example I really loved before they went all high-end on us. Older Nikon lenses also, even zooms.

These days it's all about sharpness and other technical merits but the images provided by the lenses are sterile and lifeless. I used to enjoy picking one lens over another just based on certain characteristics.

Link | Posted on Mar 24, 2017 at 14:21 UTC as 58th comment | 3 replies
On article OPPO launches F3 Plus with dual selfie cameras (13 comments in total)
In reply to:

RedFox88: Super wide group selfies? Mega perspective distortion try!

I'm sure there's an app for that.

Link | Posted on Mar 24, 2017 at 10:29 UTC
On article OPPO launches F3 Plus with dual selfie cameras (13 comments in total)

I always wonder with these selfie phones, why don't they stick the exact same lens/sensor unit to the back and to the front?

I mean if selfies are so important to some people that they might pick a certain phone over another based on selfie capabilities, wouldn't it make sense to have the same quality (or "quality" more likely anyway) on both sides?

Link | Posted on Mar 24, 2017 at 10:29 UTC as 3rd comment | 1 reply
In reply to:

james s. kennedy: After 22 years in the Air Force, I spent 24 years as a Boeing software engineer. The last half of my career was on the F-22 stealth fight, which carried the equivalence of 2 Cray supercomputers, original design had 3. I was asked to study the feasibility of automated target recognition. I had all the necessary security clearances (my USAF career was largely spy satellites that used film), and I had long discussions with every involved key player I could find, Lincoln Labs, Sandra, Livermore Labs, USAF aeronautical systems. This was 1995. My conclusion was not in my lifetime, and maybe not in God's lifetime. Very frustrating since we humans can easily do it, tell our children apart, but no one knew how we did it. Now even very cheap digital cameras have face detection. How? Was the answer neural nets, which held the most promise back in 1995. I have been retired since 2005, and no security clearances.

Basic face detection doesn't sound too complicated in theory. One needs to look for a blob of pixels with a fairly narrow range of colors with 3 slightly darker blobs of specific color difference. At least that's how I imagine it.

Now recognizing any object, or even a range of objects, that's a different thing. Honestly I'm still amazed that even radar target recognition works as well as it does, but that works again because the theory is fairly simple and radar resolution is so low so the computer can be programmed to identify a blob with range of A to B to be object X according to the database.

Optical recognition with nuancing like a human eye/brain can do is a different matter.

Link | Posted on Mar 23, 2017 at 22:53 UTC
In reply to:

samfan: I've always hated how Panasonic cameras operate - the controls, handling, how they sit in the hand, I've always found those aspects appaling.

But yea, admittedly they always knew how to fit some very nice technology and some cute ideas into their cameras. They had a few compacts with RAW capability when everyone else was abandoning it (even Canon in their G series).

The digital age has been kind of odd in this regard. Companies that have never made film cameras such as Casio, Panasonic and Sony were able to cram some incredible technology especially into compacts, but didn't know, and in my opinion still don't know, how to make cameras with proper ergonomics.

While the traditional camera companies like Canon, Nikon, Pentax, Leica were often struggling with latest tech and often messed up when trying to look hip.

Olympus and Samsung were kind of exceptions, where the former often had a lot of new ideas for digital and the latter learned a lot from their collaboration with Pentax.

I know Panasonic cameras have the reputation of being photographer-friendly, but I've always found them to be the opposite.

Back in 2005 when I was picking up a compact, I was choosing between Canon A610 and some Panasonic equivalent. I tried both without having any experience with either brand and within a few minutes I hated the Pana. It was supposedly in the same class as A610 if not higher, but it wouldn't focus in the dark, image was noisy even on the screen and the controls were clumsy as hell. Later I've tried a few other Pana cameras and always came to the same conclusion. I had one for a week and I hated it so much I didn't even realize it could shoot RAW, I just didn't want to use it.

On the other hand I always loved Pentax and Ricoh. Pentax' green button that would reset aperture and shutter speed was a stroke of genius. Cameras that were a joy to use. Samsung inherited some of Pentax' ideas and philosophy and while they always messed up some detail, I usually liked them.

Link | Posted on Mar 23, 2017 at 22:37 UTC
In reply to:

samfan: I know these social networks are all the rage these days, but I often wonder what are all those fancy cameras and lenses for if the images are meant for instagram?

Yea well I do the same except I don't even post to social networks at all :P

It's kind of a shame though, to have all the equipment and not all that much outlet to post images in good quality. (Yea I know there's Flickr and others, but everyone just wants the 1024*1024 IG/FB/Tublr images and the likes with filters and captions.) Of course, due to mind-blowing amount of photos being made and posted everywhere, we can't really do much about it.

Still, Instagram, ugh.

Link | Posted on Mar 23, 2017 at 15:27 UTC
In reply to:

marc petzold: I want this Sensor Size (1/1.63", 1/1.7") into a smartphone, as i said more than twice.
1" is too huge, only Panasonic did this, with mixed results. Current 1/2.5" to 1/2.3" sensors are virtually a bad joke in terms of IQ - and rely therefore on heavy asic-processing (JPEGs) inside the smartphone. DR, Tonality is also nothing great onto these very small sensors.

The LX3 was way good into it's heyday (the D90 came out into 2008, too) and the only real successor into my eyes since then is the LX100, but this model could use an upgrade (LX200) anytime soon....

I have the EX1 myself ;) The camera is awesome. Pretty much blew Panasonic, Canon and Nikon competition out of the water, combining the advantages of all of them.

But again, while the 10 MPx CCD sensor was truly great, the 12 later MPx CMOS variant really wasn't anything special and I don't think it's just because of the extra 2 MPx; it just wasn't as good. So even if I had that exact sensor in my phone I wouldn't be very happy anyway despite the size. And I doubt we'll ever get that CCD sensor back... Or any CCD in any mainstream camera for that matter.

And yea, I also enjoy my CCD DSLRs too ;) I never liked the D300 sensor for example, but at least with larger sensors eventually we got good CMOS too; the 14 MPx one in Nikon D3100 was really good (really felt like a CCD really) and even the 10 MPx 1" in Nikon V1/J1/S1 etc. was pretty cool. And the later 24 Mpx APS-C models are good too.

But it's not just the tech or size... With phone sensors you'll always get the cheapest ones.

Link | Posted on Mar 23, 2017 at 15:21 UTC
In reply to:

marc petzold: I want this Sensor Size (1/1.63", 1/1.7") into a smartphone, as i said more than twice.
1" is too huge, only Panasonic did this, with mixed results. Current 1/2.5" to 1/2.3" sensors are virtually a bad joke in terms of IQ - and rely therefore on heavy asic-processing (JPEGs) inside the smartphone. DR, Tonality is also nothing great onto these very small sensors.

The LX3 was way good into it's heyday (the D90 came out into 2008, too) and the only real successor into my eyes since then is the LX100, but this model could use an upgrade (LX200) anytime soon....

It's not just the sensor size, it's how good the sensor is. The Sony 1/1.7" 10 MPx CCD was amazing, in fact even Canon and Nikon used it in their later models (G11, P7000) despite the fact that the previous ones had higher pixel count.

However, the Sony 12 MPx CMOS followup with the same size was garbage in comparison.

If they'd make a 1/1.7 sensor but crammed 30 MPx or whatever they surely couldn't resist into a mobile phone use and combined it with some pencil-tip-sized lens, you wouldn't get any advantage. There's a reason while most of those 1/1.7" compact cameras were pretty large (and the smaller ones weren't that impressive).

Link | Posted on Mar 23, 2017 at 11:59 UTC
Total: 696, showing: 1 – 20
« First‹ Previous12345Next ›Last »