Forgot Password
Pentax Camera Forums Home
 

Reply
Show Printable Version Search this Thread
01-28-2013, 06:20 PM   #1
Pentaxian
normhead's Avatar

Join Date: Jun 2007
Location: Near Algonquin Park
Photos: Gallery | Albums
Posts: 40,442
They must have made a mistake.

I looked up the resolution numbers for the D600 D6 and K-30. To my surprise , the K30 came out at 3000lw/ph, while the D600 came out at 2800, and the 6D came out about the same but with false colour. I couldn't find that these tests on a K-5 or K-5 II. According to these charts the K-30 out resolves both the D600, D6. A d800 can top 4000 lw/ph. Previous tests done on a k-5 , which I can no longer locates indicated a top resolution of 2100 lw/ph.



So, what gives?

01-28-2013, 07:38 PM   #2
Veteran Member




Join Date: Apr 2010
Location: Tennessee
Posts: 6,617
Unless I am reading it wrong:

The K-30 JPEG = 2400lph
The D6 JPEG = 2600lph
The D600 JPEG = 2800lph

The K-30 RAW file resulted in 3000lph
The D6 RAW file resulted in 2800lph
They don't say what the RAW file for the D600 scored after sharpening.

Not sharper than the FF competition, but definitely a strong performance. What lens was used for each camera test? At one point I thought DPR was using a Sigma macro lens in each mount for the test, but I'm not sure what they do any more.

The D3200 (24MP APS-C) JPEGs hit 3000lph and RAW files scored 4000lph with a Nikon 50mm F/1.4 lens.
01-29-2013, 06:39 AM   #3
Pentaxian
normhead's Avatar

Join Date: Jun 2007
Location: Near Algonquin Park
Photos: Gallery | Albums
Posts: 40,442
Original Poster
winder, you're reading the jpeg resolutions, so of course when take into the consideration the degradation of the image caused by coding and decoding a jpeg, the K-30 clearly suffers more than the others. In the raw files, which is what you use if you are serious about resolution, the K-30 comes out on top. Although, for me it's a one piece of the puzzle type of thing. I'd really like to know this isn't a typo, before I take it seriously. As I said, previous research indicated the base resolution of a k-5 was about 2100 lw/ph and with the best lens I've seen it rated for it had a lw/ph of 2900, so that would mean the K-30 out performed even the K-5 at it's best.

Also interesting that the Canon 6D can lw/ph over 2800 but not without colour artifacting and moire issues. It's really starting to look like a bit of a dud photographically.

I believe DPR tests with a 70 mm, lens which puts APS-c at a disadvantage in that they are shooting from further away. And as everyone knows, "if you don't like your pictures, get closer." IN many cases, light diffraction in the atmosphere over distance degrades images more than different lens characteristics. The closer you are, the better in terms of reproducing resolution.

ANother shocking note, talking about the D800 (not the E)

QuoteQuote:
We should note, however, that we had to work quite hard to get this amount of resolution. We used flash to eliminate any risk of blurring due to vibration, we focus-bracketed in extremely fine increments, and we used an excellent lens (the Nikkor AF-S 50mm f/1.4 G) at an aperture optimal for central sharpness of F4.5. Even with the inevitable softening that results from using an anti-aliasing filter, this methodology allowed us to tease resolution out of the camera as high as 3200 LPH, which, for practical purposes is about the maximum we'd have expected to see.
( Nikon D800 Review: Digital Photography Review )

WIth the anti-aliasing filter on, again based on these highly suspect numbers, the D800 produces only marginally better resolution than the K-30.

Compare this to the 800e

QuoteQuote:
As you can see in the samples above, the resolution performance of the D800E is truly impressive. Individual lines can be distinguished nearer to our chart's 4000 LPH limit than with any other 35mm format DSLR we've yet tested in our studio. However this is accompanied by more-obvious false colour than we saw with the D800, especially in Raw - an inevitable side-effect of cancelling the anti-aliasing filter.
It would seem (again with the warning that I'm very suspicious of these numbers), that you can't get much better than a K-30 without false colour. It's disappointing IR doesn't have the tests for the K-5 II and K-5 IIs available, in fact i can't even find the test of the K-5 at this point.

If anyone else can find a K-5 review, I looked here.... I know I've seen it before. I tmust be on a part of the site I can't figure out how to access.

Here's where I'm looking.

Camera Reviews / Previews: Digital Photography Review
01-29-2013, 06:55 AM   #4
Veteran Member
causey's Avatar

Join Date: Feb 2009
Location: Arlington, VA
Photos: Albums
Posts: 3,757
QuoteOriginally posted by normhead Quote
winder, you're reading the jpeg resolutions, so of course when take into the consideration the degradation of the image caused by coding and decoding a jpeg, the K-30 clearly suffers more than the others.
No, I think winder is right. From DPR:

D600: "Its JPEG performance on our resolution chart is very good with output that retains fine detail up to roughly 2800LPH, which is about as good as we'd expect from a 24MP Bayer sensor."

K-30: "... on the converted Raw file detail is accurately described up to almost 3000LPH ..."


Very good resolution number from K-30. But look why I'm waiting for a 24mp APS-C Pentax camera:

Nex-7: "The NEX-7's 24MP pixel count pays off in high levels of detail, accurately resolving up to at least 3400 lines per picture height." (Not clear if they are talking raw or jpeg.)

Nikon D3200: "... the out-of-camera JPEGs, while clean of artifacts, are a little soft and and only resolve accurately up to approximately 3000 lp/ph which is less than we would expect from a camera with a 24MP sensor.

Converting your raw files and applying a customized unsharp mask (170%, 0.4 radius in our case) will get you a significant amount of additional detail. In our examples, here, there is some detail (albeit not 'true' almost up to the Nyquist limit of 4000lp/ph, which is the theoretical limit of the D3200's sensor resolution."

A lot of cropping room!


Last edited by causey; 01-29-2013 at 07:02 AM.
01-29-2013, 07:01 AM   #5
Pentaxian
normhead's Avatar

Join Date: Jun 2007
Location: Near Algonquin Park
Photos: Gallery | Albums
Posts: 40,442
Original Poster
The improvement to D600 files to RAW is described as "a bit more". A bit more could mean another 200 lw/ph meaning the same as the rated k-30. Not enough information to understand exactly what they are saying. It doesn't say what the RAW file resolution is on a D600. I'm still going with the 3000 lw/ph being a typo. If not the K-30 is the rough equivalent of the D600 in lw/ph.
01-29-2013, 07:06 AM   #6
Veteran Member
causey's Avatar

Join Date: Feb 2009
Location: Arlington, VA
Photos: Albums
Posts: 3,757
QuoteOriginally posted by normhead Quote
The improvement to D600 files to RAW is described as "a bit more". A bit more could mean another 200 lw/ph meaning the same as the rated k-30. Not enough information to understand exactly what they are saying. It doesn't say what the RAW file resolution is on a D600. I'm still going with the 3000 lw/ph being a typo. If not the K-30 is the rough equivalent of the D600 in lw/ph.

Any way we take it, the resolution of the new APS-C sensors--16mp and 24mp--is fantastic. Resolution is no longer a FF advantage. Now "APS-C vs FF" basically means "(lightweight) vs (DOF control & better ISO performance)."
01-29-2013, 07:54 AM   #7
Veteran Member




Join Date: Apr 2010
Location: Tennessee
Posts: 6,617
Measuring resolution with LPH equalizes for sensor size (height) but not for pixel density. The higher density sensor will yield competitive resolution at lower ISO setting and under good light. Since the D800 and the K-30 sensors have the same pixel density they should score very similarly on a LPH test. Since the APS-C sensor has to be enlarged more for real world output it would need to capture at a higher resolution to compensate and produce an equal quality output.

01-29-2013, 09:51 AM   #8
Pentaxian
normhead's Avatar

Join Date: Jun 2007
Location: Near Algonquin Park
Photos: Gallery | Albums
Posts: 40,442
Original Poster
QuoteOriginally posted by Winder Quote
Measuring resolution with LPH equalizes for sensor size (height) but not for pixel density. The higher density sensor will yield competitive resolution at lower ISO setting and under good light. Since the D800 and the K-30 sensors have the same pixel density they should score very similarly on a LPH test. Since the APS-C sensor has to be enlarged more for real world output it would need to capture at a higher resolution to compensate and produce an equal quality output.
LW/PH is Line Width divided into Picture Height. It is a ratio of the smallest resolvable line as a function of the Height of the frame. So even though the sensor may be smaller, it's still resolving 3000 lines, over the frame, same as the FF camera is resolving 3000 line over it's frame. A 2 dimenesional image blown up to the same size should look the same whether FF or APS-c. They will both have the equivalent of 3000 horizontal lines. I say equivalent because obviously a 3200 pixel image can't resolve 3000 lines, but, you can pad your images with photoshop if you want bigger detail, not more detail.

To be a superior in LW/PH the images has to resolve more, no matter what size the sensor. You can observe the same thing with the 645D and the D800. One would think that a larger sensor would mean less trouble with false colour, but even that doesn't hold up. Neither the D600 nor the K-30 exhibit the kind of false colour at high REs that the 6D does, so clearly there is something else in play besides sensor size, going by pixel size alone the 6D should be the best of the bunch. it's the worst.

Last edited by normhead; 01-29-2013 at 10:02 AM.
01-29-2013, 10:45 AM   #9
Pentaxian
normhead's Avatar

Join Date: Jun 2007
Location: Near Algonquin Park
Photos: Gallery | Albums
Posts: 40,442
Original Poster
Because this is digital, you can actually define exactly what the LW/PH max is for a given sensor. If you took an image from a k-30, you have 3264 horizontal lines, so, of you ignore the lens, take an image that size and divided it into one pixel wide black and white lines, you would have 3264 distinguishable lines. Blow the image up to wall size an you'll be able to count them.

Using a D600 doing the same thing, one pixel arrangement with a horizontal resolution of 4016, you should be able to blow up and observe 4016 horizontal lines.

Unless I'm thinking this through wrong... The D600 should be able to resolve 1000 more LW/PH. But neither the D600 nor the K-30 are capable of doing that. Apparently however the K-30 is capable of resolving the same or better LW/PH, in practical, even though it has much less theoretical resolving power. It's this gap between what's possible and what's observed that has me confused with the larger sensors.

A D800 has 4900 and resolves over 4000 lw/ph.

The rule or thumb would seem to be that a camera resolves at best 80% of it's theoretical max. That would mean, compared to the others a K-30 should be somewhere around 2500 lw/ph.

Which is why I still feel, that this must be a mistake. Either that or Pentax are really a bunch of wizards and our cameras all have a secret compartment to hold the fairy dust.

Using this example you can also see why, if you resolve 3000 distinct lines on APS-c a 3000 Distinct lines on an FF, it makes absolutely no difference how big you blow them up. They are in every sense digitally equivalent.

Where this falls down in practice is that in many cameras, like the D800 and 645D you can see that there is a pattern, in part of the image that is unclear, long after the distinct lines have disappeared. SO even though you aren't getting distinct lines, you are getting a representation of something beyond what LW/PH would sugggest without following each line pair to extinction.

Last edited by normhead; 01-29-2013 at 10:53 AM.
02-02-2013, 10:55 AM   #10
Pentaxian
normhead's Avatar

Join Date: Jun 2007
Location: Near Algonquin Park
Photos: Gallery | Albums
Posts: 40,442
Original Poster
The world is as it should be..... Pentax K-30 Camera Exposure - Review

QuoteQuote:
We were able to resolve a little more with an Adobe Camera Raw conversion, about 2,250 lines of resolution in the horizontal direction and about 2,200 in the vertical, and complete extinction of the pattern was extended beyond 3,600
QuoteQuote:
Impressive 24 x 36-inch prints at ISO 100; good 11 x 14s at ISO 3,200; ISO 25,600 makes a good 5 x 7.
So for a 645D (and by inference a D800) you should be able to get a good 40x60 print.

Imaging resource seem to be in flux right now. Hopefully at some time in the future they'll get a handle on this stuff and you'll get comparisons that use the same standards for every camera. As of today they don't have k-5, K-5II or K-5IIs data... but it appears they are working on producing consistent results for all camera bodies. If I was rich, I'd send them money.
02-15-2013, 01:21 AM   #11
Inactive Account




Join Date: Jan 2011
Posts: 12
If someone points to a dog in front of you, and says "this is a hippo", do you believe that it's a hippo? Or do you believe your own sensory organs? Nevermind what the dpreview writer writes. After all, "roughly" and "almost" are not exactly precise terms of measurement. From what I see of the results on the screen, the K-30 raw vertical resolution starts giving false detail at 28. The d600 is still giving accurate line count at 32.
02-15-2013, 06:55 AM   #12
Pentaxian
normhead's Avatar

Join Date: Jun 2007
Location: Near Algonquin Park
Photos: Gallery | Albums
Posts: 40,442
Original Poster
The universe is as it should be.
Reply

Bookmarks
  • Submit Thread to Facebook Facebook
  • Submit Thread to Twitter Twitter
  • Submit Thread to Digg Digg
Tags - Make this thread easier to find by adding keywords to it!
d600, k-30, k-5, lw/ph, resolution, tests
Thread Tools Search this Thread
Search this Thread:

Advanced Search


Similar Threads
Thread Thread Starter Forum Replies Last Post
DPR K-01 Preview - They've Made A Hash Of It mervis Pentax K-01 92 03-28-2012 03:50 PM
Must We Act As If They Mean What They Say? deadwolfbones General Talk 23 09-07-2011 10:49 AM
I made a mistake! How to clean my sensor?? camo Troubleshooting and Beginner Help 8 07-11-2011 01:30 AM
Must own to have a valid opinion BillO General Talk 23 03-19-2011 08:41 AM
istDS vs isD: how big a mistake have I made? WMBP Pentax DSLR Discussion 12 09-01-2007 09:26 AM



All times are GMT -7. The time now is 07:43 AM. | See also: NikonForums.com, CanonForums.com part of our network of photo forums!
  • Red (Default)
  • Green
  • Gray
  • Dark
  • Dark Yellow
  • Dark Blue
  • Old Red
  • Old Green
  • Old Gray
  • Dial-Up Style
Hello! It's great to see you back on the forum! Have you considered joining the community?
register
Creating a FREE ACCOUNT takes under a minute, removes ads, and lets you post! [Dismiss]
Top