Forgot Password
Pentax Camera Forums Home
 

Reply
Show Printable Version Search this Thread
07-21-2016, 10:30 PM   #181
Loyal Site Supporter
clackers's Avatar

Join Date: Jul 2013
Location: Melbourne
Photos: Albums
Posts: 8,508
QuoteOriginally posted by Nicolas06 Quote
I must have issues with my eyes because I don't see the 2 image having the same size....

Then resize them!


I have provided everyone links to the two untouched RAW files I guessed you used ... do as you like with them, zoom in, zoom out, pan to different areas.


Except don't try the postprocessing you did to prove things with.


In your Post 157, the top picture does not match what we can see of the K-1 file you began with (screenshot attached).


Kh123456789 thought they might be demosaicing artifacts - you were silent on the subject. But your weird smears to me look like you've attempted noise reduction and sharpening. What has happened to all the pink chroma dots in the K-1 image?


PS My apologies for calling you 'Nicholas' back there. No disrespect intended!

Attached Images
 

Last edited by clackers; 07-21-2016 at 11:04 PM.
07-21-2016, 10:34 PM   #182
Loyal Site Supporter
clackers's Avatar

Join Date: Jul 2013
Location: Melbourne
Photos: Albums
Posts: 8,508
QuoteOriginally posted by Gimbal Quote
There is definitely a lot more information in the K1 picture.

Do you think they are equal?

We aren't talking about resolution - that's Nicolas' distraction, we're talking about the noise.


Yes, I do think they're the same within the variation of two photos taken four years apart with only roughly the same setup.


And that's what DxOMark found in their much more precise, controlled measurements of a uniform subject, Gimbal - 21.4dB and 21.5dB 18% SNR when they did both cameras around ISO 6400.


Pentax K-5 II vs Nikon D800

Last edited by clackers; 07-21-2016 at 10:56 PM.
07-21-2016, 10:39 PM   #183
Loyal Site Supporter
clackers's Avatar

Join Date: Jul 2013
Location: Melbourne
Photos: Albums
Posts: 8,508
QuoteOriginally posted by Gimbal Quote
Really?

Signal being information and noise being spurious and random, I would say that there is a lot more information (signal) in the K1 picture (looking at the whole picture of course, or a downsized picture, whatever you like). .

Noise is the *ratio* of random data to information, Gimbal - this you seem to have redefined to suit your 'logic'.


An honest post from you would admit to us that the K-1 also in fact offers worse read noise than the K-5. See attachment.


You like astrophotography, right? Then you would remember writing these words about the topic:

It's not easy to understand the figures, I mean some of the samples with a low mean value looks much worse then samples with a higher value.

Anyhow, does this mean that if I where to shoot some deep space object like the Andromeda galaxy with a
200mm lens, (which means cropping to 100%) that I would be better off using the K5 than the K1?

The number suggests this, and if so it's kind of disappointing, I where expecting at least the same performance on a pixel level.





Attached Images
 

Last edited by BigMackCam; 07-24-2016 at 12:27 AM. Reason: Edited mild vulgarity
07-21-2016, 10:48 PM   #184
Site Supporter




Join Date: Mar 2008
Location: Prince George, BC
Photos: Gallery
Posts: 2,279
This is all angels on the head of a pin bickering. For single-shot astrophotos, the read noise difference *may* be seen in the result... but I doubt it. For any other type of astrophotography where stacking is used, it makes no difference at all. In fact, the paltry chroma noise of the K-1 gives it an advantage. But they are both stellar astro machines.

07-21-2016, 11:17 PM   #185
Loyal Site Supporter
clackers's Avatar

Join Date: Jul 2013
Location: Melbourne
Photos: Albums
Posts: 8,508
QuoteOriginally posted by jbinpg Quote
In fact, the paltry chroma noise of the K-1 gives it an advantage.

???


Look at the attachment above, Jbinpg ... the coloured specs on the K-1 shots with the lens cap on are chroma noise!


As for stacking and dark frame subtraction, neither are methods for obtaining clean, original data, they are postprocessing you do on the noise that's there in the first place.
07-21-2016, 11:51 PM   #186
Site Supporter




Join Date: Mar 2008
Location: Prince George, BC
Photos: Gallery
Posts: 2,279
I shoot with a K-1 and I know firsthand how little chroma noise it has. The difference between my shots and the ones in your post is that I do not shoot with the lens cap on.
07-22-2016, 12:04 AM   #187
Pentaxian




Join Date: Nov 2013
Posts: 4,621
QuoteOriginally posted by clackers Quote
Then resize them!


I have provided everyone links to the two untouched RAW files I guessed you used ... do as you like with them, zoom in, zoom out, pan to different areas.


Except don't try the postprocessing you did to prove things with.


In your Post 157, the top picture does not match what we can see of the K-1 file you began with (screenshot attached).
I did raw => JPEG and denoising in Raw Therapee. The noise level for both camera was the same exactly to be fair and quite light to not wash details. RawTherapee (that I find too complex) is able to find details Lightroom and DxO can't see. I seen it on my own pictures, this at all iso but you can see chromatic aberation even at iso 100. My understanding is that their algorithm apply a very light touch by default.

The resizing was done in DxO; No image even took sharpening.

In post 157 the K1 is viewed at 100% and the K5-IIs enlarged to fit. This is representative of large print looking at it from near distance. The K5-IIs look quite moire noisy because the noise is magnified.

On other posts, on the contrary the K1 image are downsized to K5-IIs size, averaging the noise of K1. And of course there less details as there only 16MP instead of 36MP.
07-22-2016, 01:26 AM   #188
Pentaxian




Join Date: Apr 2007
Location: Sweden
Posts: 1,916
QuoteOriginally posted by clackers Quote
Noise is the *ratio* of random crap to information
Really? I would call that SNR.


QuoteOriginally posted by clackers Quote
An honest post from you would admit to us that the K-1 also in fact offers worse read noise than the K-5. See attachment.


You like astrophotography, right? Then you would remember writing these words about the topic:

It's not easy to understand the figures, I mean some of the samples with a low mean value looks much worse then samples with a higher value.

Anyhow, does this mean that if I where to shoot some deep space object like the Andromeda galaxy with a
200mm lens, (which means cropping to 100%) that I would be better off using the K5 than the K1?

The number suggests this, and if so it's kind of disappointing, I where expecting at least the same performance on a pixel level.

We have been through this so many times, but here we go again.

At pixel level, the K-5 and the K-1 were expected to be more or less equal. ( no measurements of the K-1 are available at DXO as of yet ).

The measurements from "Sensor noise DB" does seem to indicate that, at pixel level, the K-1 is worse then the K-5.
And yes, that is a disappointment and will be visible (well, in theory) when viewing pictures at 100%. (For instance the Andromeda shoot with a 200mm lens)

But in those cases were we will be using the entire sensor (hopefully most of the time) we should be looking at the SNR of the whole picture, and the fact that the K-1 has the double amount of (roughly the same quality) pixels of the K-5 will of course result in a cleaner image. (with all the normal disclaimers of the same viewing size and so on and so forth.)

07-24-2016, 05:25 AM - 1 Like   #189
Veteran Member
benjikan's Avatar

Join Date: Jan 2007
Location: Paris, France
Photos: Gallery
Posts: 4,312
QuoteOriginally posted by normhead Quote
Oh some folks have fudged some pretty convincing images.

But my APS_c images from this morning....








as posted here.
https://www.pentaxforums.com/forums/10-pentax-slr-lens-discussion/132843-boke...ml#post3686194

This is definitely a place where you can challenge yourself, by looking at the images and seeing if you can pick out the FF images. According to the constant roar of Full Frame nonsense, it should be a snap.
That's because you are shooting relatively close up to a small subject. It becomes a lot more apparent when shooting full shots of people. I own a 2.8 300mm lens and the bokeh differences are quite evident.
07-24-2016, 07:17 AM   #190
Pentaxian




Join Date: Nov 2013
Posts: 4,621
QuoteOriginally posted by benjikan Quote
That's because you are shooting relatively close up to a small subject. It becomes a lot more apparent when shooting full shots of people. I own a 2.8 300mm lens and the bokeh differences are quite evident.
Yep using the same lens APSC has less dof than FF.

300mm f/2.8 on APSC focussed at 10meters => 2.3meters of dof.
300mm f/2.8 on FF focussed at 10 meters => 3.5meters of dof.

More, if you keep the framing, same apperture give the same dof, whaterever the focal length. The background blur would be different, but the dof the same. Sure the sensor size would count. But if you take an f/1.4 lens on APSC, say a 50 or 85mm you'll achieve less dof than you can with that 300mm f/2.8 as long as the framing kept the same. The perspective compression would be also much different.

The following link would give you an idea on how the background would like from different lenses at different apperture and different sensors: http://howmuchblur.com/#compare-1.5x-300mm-f2.8-and-1x-300mm-f2.8-and-1.5x-8...m-wide-subject

In particular it is only until 5 meters distance between subject and background that the 300mm f/2.8 on FF would provide more subject separation than a 85mm f/1.2 on APSC. If the background more near than that, the 85mm f/1.2 on APSC give more subject separation... It would need 10 meters distance between subject and background to beat 85mm f/1.2 on FF. And honestly 85mm f/1.2 still cheaper than 300mm f/2.8... and more practical

Last edited by Nicolas06; 07-24-2016 at 07:34 AM.
07-26-2016, 09:52 PM   #191
Loyal Site Supporter
clackers's Avatar

Join Date: Jul 2013
Location: Melbourne
Photos: Albums
Posts: 8,508
QuoteOriginally posted by Nicolas06 Quote
I did raw => JPEG and denoising in Raw Therapee. The noise level for both camera was the same exactly to be fair and quite light to not wash details.

Well, yes, that's the problem with postprocessing, Nicolas. We all do it in different ways.


We can see in your Post 157 it hasn't worked. That wall is a plain grey one that's not supposed to have any texture.

Last edited by clackers; 07-26-2016 at 10:16 PM.
07-26-2016, 10:07 PM   #192
Loyal Site Supporter
clackers's Avatar

Join Date: Jul 2013
Location: Melbourne
Photos: Albums
Posts: 8,508
QuoteOriginally posted by Gimbal Quote
At pixel level, the K-5 and the K-1 were expected to be more or less equal.

Well, that's hypocritical of you.


Why did you imply otherwise in your Post 172?


Remember the topic is noise, not resolution. A K-3 has more resolution than the FF Canon 5D MkIII.




QuoteOriginally posted by Gimbal Quote

But in those cases were we will be using the entire sensor (hopefully most of the time) we should be looking at the SNR of the whole picture.

@Gimbal, you still have this lack of understanding. The SNR of the whole pictures using the full sensor areas of the K-5 IIS and D810 is the same - that's what DxO found. That's the screen tab. They measured the RAW files, and got 21.4dB and 21.5dB.


You can try and average out the noise though by downsampling in post processing - the more pixels the better, but this is software and your algorithm may have dreadful consequences as Nicolas found.


I speak as someone who shoots and post-processes both 16Mp APS-C and 36Mp FF.


Again, noise is all about the number of pixels, their technology and their well capacity, not the wafer underneath them. 'Total light' and 'sensor area' is bollocks.

Last edited by clackers; 07-26-2016 at 10:26 PM.
07-26-2016, 10:37 PM   #193
Pentaxian




Join Date: Nov 2013
Posts: 4,621
@Clackers

maybe after a few more years of post processing and printing 16MP and 36MP pictures, on day you'll finally notice with your own eyes that the K1 pictures look better, in particular when both are at quite high iso. Maybe even one day you'll understand the concept of normalization.

Maybe one day I'll understand that sensor size is not a factor, equivalence and all the fuss about FF is lie and that only individual pixel size count... I'll buy a *ist DL with 6MP or maybe a 5D that would beat everything due to their huge pixels.

In the meantime, it isn't really worth to discuss futher the subject. Neither you or us would change position.

Last edited by Nicolas06; 07-26-2016 at 11:19 PM.
07-27-2016, 01:08 AM   #194
Pentaxian




Join Date: Apr 2007
Location: Sweden
Posts: 1,916
QuoteOriginally posted by clackers Quote
Well, that's hypocritical of you.


Why did you imply otherwise in your Post 172?
That was not my intention.
The K-5 and the K-1 has been expected to perform similar (at pixel level) ever since it became known that they would have a similar pixel size. Nothing new there. The performance increase per square mm of sensor has become smaller and smaller for every generation. 10 years ago a new sensor could mean vastly improved performance, not so much anymore, slight increases is all we can hope for.

If you want noticeably better iso perfomance now, you will have to jump to bigger sensors, bigger as in bigger total sensor area.



QuoteOriginally posted by clackers Quote
You can try and average out the noise though by downsampling in post processing - the more pixels the better, but this is software and your algorithm may have dreadful consequences as Nicolas found.
Software is unavoidable when it comes to digital photo. You will have to use it and it will affect end result. You will use up or downsampling to match print/view size, you can either use your preferred software or some other software will do it for you (as in printer driver or screen viewer), and it will affect noise. Why pretend otherwise?
07-27-2016, 01:51 AM   #195
Site Supporter
kenspo's Avatar

Join Date: Mar 2014
Location: Oslo
Posts: 1,931
K-1 is way over any other camera from Pentax in High ISO, at least when it comes to real life use of a camera..Which is the way i prefer to use and test my cameras.

So others can pixelpeep, use online testings etc. I couldn't care less
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!
aperture, aps-c, background, blur, camera, dof, dslr, feet, ff, ff over aps-c, field, focus, frame, full frame, full-frame, k-1, k1, lens, paper, pentax k-1, people, picture, posters, stitching
Thread Tools Search this Thread
Search this Thread:

Advanced Search


Similar Threads
Thread Thread Starter Forum Replies Last Post
FF vs APS-C Field of View revisited Ole Pentax K-1 2 05-07-2016 02:13 PM
Confused about Angle of View of Lenses on FF vs APS-C? Kath Pentax SLR Lens Discussion 3 10-01-2015 09:55 AM
Does FF vs APS-C affect amount of light? windhorse General Photography 46 03-02-2015 07:07 PM
Quick question regarding field of view - FF vs APS-C glass? Julie Pentax SLR Lens Discussion 5 12-23-2012 05:33 PM
APS-C does not increase focal length over FF, it decreases field of view. TomTextura Photographic Technique 135 06-09-2012 04:58 PM



All times are GMT -7. The time now is 09:50 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