Forgot Password
Pentax Camera Forums Home
 

Reply
Show Printable Version Search this Thread
03-14-2016, 01:36 PM   #196
Pentaxian




Join Date: Jan 2012
Photos: Gallery
Posts: 587


The K5 is a fantastic camera. It's better than the 5DS and compares favourably against the D750 and even matching it at high ISOs where noise performance matters.

My own theory is that we have reached the maximum efficiency of sensors about 4-5 years ago and what matters now is how much sensels manufacturers cram into such tiny spaces. It can be seen in the performance of the ancient E-M5 matching the 5DS.



03-14-2016, 02:44 PM - 2 Likes   #197
Pentaxian
jsherman999's Avatar

Join Date: Dec 2007
Photos: Gallery
Posts: 8,228
Screen Tab Follies! (actually, perfect timing)

QuoteOriginally posted by dtmateojr Quote

The K5 is a fantastic camera. It's better than the 5DS and compares favourably against the D750 and even matching it at high ISOs where noise performance matters.
<...presents capture of 'screen' tab on DxOMARK...>

An absolutely perfect example of using the 'screen' tab to draw the exact wrong conclusion - that the K5 (or any same-gen aps-c camera, not to pick on the K5) 'beats' or matches the FF 5DS or D750


Here's the 'print' tab, showing actual normalized output from each camera:




(Let's have a sanity-check show of hands - how many folks think that the K5 beats the D750 or 5DS FF (or Pentax K-1) cameras in noise performance? Clackers? Screen tab up there! He even has a micro 4/3 camera matching FF 5DS! What say you?

.

---------- Post added 03-14-16 at 04:41 PM ----------

QuoteOriginally posted by dtmateojr Quote
Here we go again. I can show you that the performance of my ancient Olympus E-M5 is the same as Canon's latest and greatest 5DS. This equivalence thingy has been debunked a million times. .
It has, huh?

Are you under the impression you 'debunked' it when you told Dr Bob Newman (sensorgen owner) he was full of crap on dpreview and got yourself banned there because the admins couldn't believe anyone posting such idiotic, baseless statements was doing anything but trolling?

And we see how you're using the screen tab to 'prove' your m/43 OMD matches the FF 5DS. That's very illuminating on a number of levels.

BTW, here's the 5DIII vs. your OMD-EM5, both at ISO 6400, without even size-normalization (ie the 5DII image is magnified more which will show more noise in comparison - displayed at the same size, the difference would be even greater.)

I think the 5DS should perform even better than the 5DIII.



.

Last edited by jsherman999; 03-14-2016 at 04:02 PM.
03-14-2016, 04:25 PM   #198
Pentaxian




Join Date: Jan 2012
Photos: Gallery
Posts: 587
QuoteOriginally posted by jsherman999 Quote
An absolutely perfect example of using the 'screen' tab to draw the exact wrong conclusion - that the K5 (or any same-gen aps-c camera, not to pick on the K5) 'beats' or matches the FF 5DS or D750


Here's the 'print' tab, showing actual normalized output from each camera:




(Let's have a sanity-check show of hands - how many folks think that the K5 beats the D750 or 5DS FF (or Pentax K-1) cameras in noise performance? Clackers? Screen tab up there! He even has a micro 4/3 camera matching FF 5DS! What say you?

.

---------- Post added 03-14-16 at 04:41 PM ----------



It has, huh?

Are you under the impression you 'debunked' it when you told Dr Bob Newman (sensorgen owner) he was full of crap on dpreview and got yourself banned there because the admins couldn't believe anyone posting such idiotic, baseless statements was doing anything but trolling?

And we see how you're using the screen tab to 'prove' your m/43 OMD matches the FF 5DS. That's very illuminating on a number of levels.

BTW, here's the 5DIII vs. your OMD-EM5, both at ISO 6400, without even size-normalization (ie the 5DII image is magnified more which will show more noise in comparison - displayed at the same size, the difference would be even greater.)

I think the 5DS should perform even better than the 5DIII.



.

Print SNR is normalised SNR i.e. FAKE SNR. That same normalised graphs can be replicated by moving forward and backward when viewing a print. Moving closer to the print increases PERCEIVED noise and moving farther away decreases PERCEIVED noise. Did I improve the sensor performance by moving forward backward and sideways? Of course not but you equivalence fan boys keep pushing this as truth. It's stupid.
03-14-2016, 06:53 PM - 1 Like   #199
Pentaxian




Join Date: May 2009
Location: Ponoka Alberta Canada
Photos: Gallery
Posts: 450
QuoteOriginally posted by dtmateojr Quote
Here we go again. I can show you that the performance of my ancient Olympus E-M5 is the same as Canon's latest and greatest 5DS. This equivalence thingy has been debunked a million times. Please stop spreading nonsense.
Eric Fossum on DPR in the Photographic Science and Technology forum spent a considerable amount of trying to explain how wrong you are and even after putting up with your insults he then finally got feed up you your antics and logged a formal complaint that setup the path in getting you banned. Even after his and other individuals who tried to educate you on the subject they put up with you for sometime and anytime one of these individuals showed you the error of your train of thought with hard science you resorted to insults and now you are trying to prove the same thing here. If you want to disprove any bodies work maybe you should take a look at your own.

Now who I am going to believe people like Eric ( a person that has likely forgot more about sensor design than most people in the industry know, written more published papers on sensor design, is leading the way to the next big thing in digital imaging "Joints photo sites", is a long time member( might even be the founder) of International Image Sensor Society. Or you and your miss guided ramblings
Welcome to Dr. Eric R. Fossum's Web Address

03-14-2016, 06:59 PM - 2 Likes   #200
Emperor and Senpai
VoiceOfReason's Avatar

Join Date: Jun 2010
Location: Nashville, IN
Posts: 5,123
QuoteOriginally posted by dtmateojr Quote
Print SNR is normalised SNR i.e. FAKE SNR. That same normalised graphs can be replicated by moving forward and backward when viewing a print. Moving closer to the print increases PERCEIVED noise and moving farther away decreases PERCEIVED noise. Did I improve the sensor performance by moving forward backward and sideways? Of course not but you equivalence fan boys keep pushing this as truth. It's stupid.
Aren't photographs always how we perceive them? Isn't this art about how people perceive your images? If the tools do the job who cares? If the images bring forth happiness or evoke what you wanted them to who cares?
03-14-2016, 07:09 PM - 1 Like   #201
Site Supporter
cali92rs's Avatar

Join Date: May 2011
Location: Long Beach, CA
Posts: 3,197
QuoteOriginally posted by dtmateojr Quote
Print SNR is normalised SNR i.e. FAKE SNR. That same normalised graphs can be replicated by moving forward and backward when viewing a print. Moving closer to the print increases PERCEIVED noise and moving farther away decreases PERCEIVED noise. Did I improve the sensor performance by moving forward backward and sideways? Of course not but you equivalence fan boys keep pushing this as truth. It's stupid.
And the screen capture of the coins which obviously shows more noise with the micro 4/3 camera?
03-14-2016, 07:30 PM - 1 Like   #202
Pentaxian




Join Date: Jan 2012
Photos: Gallery
Posts: 587
QuoteOriginally posted by Ian Stuart Forsyth Quote
Eric Fossum on DPR in the Photographic Science and Technology forum spent a considerable amount of trying to explain how wrong you are and even after putting up with your insults he then finally got feed up you your antics and logged a formal complaint that setup the path in getting you banned. Even after his and other individuals who tried to educate you on the subject they put up with you for sometime and anytime one of these individuals showed you the error of your train of thought with hard science you resorted to insults and now you are trying to prove the same thing here. If you want to disprove any bodies work maybe you should take a look at your own.

Now who I am going to believe people like Eric ( a person that has likely forgot more about sensor design than most people in the industry know, written more published papers on sensor design, is leading the way to the next big thing in digital imaging "Joints photo sites", is a long time member( might even be the founder) of International Image Sensor Society. Or you and your miss guided ramblings
Welcome to Dr. Eric R. Fossum's Web Address

Name dropping will get you nowhere. Photography existed way before digital sensors existed. The basic concepts of photographic exposure hasn't changed since the days of film. Equivalence fails miserably when applied to film photography because it is very wrong. Your Eric Fossum does not understand basic photography.
03-14-2016, 07:32 PM   #203
Lens Buying Addict
monochrome's Avatar

Join Date: Mar 2008
Location: Kirkwood (St. Louis) MO
Photos: Gallery | Albums
Posts: 19,472
This is why I never go to DPR. Take it somewhere else.

03-14-2016, 07:33 PM   #204
Pentaxian




Join Date: Jan 2012
Photos: Gallery
Posts: 587
QuoteOriginally posted by cali92rs Quote
And the screen capture of the coins which obviously shows more noise with the micro 4/3 camera?

Because a 5D3 and E-M5 do not have the same pixel density. Try comparing a 5DS and E-M5 though and you'll understand. The DXOMark graphs I posted will show you the real data.
03-14-2016, 09:00 PM   #205
Pentaxian




Join Date: May 2009
Location: Ponoka Alberta Canada
Photos: Gallery
Posts: 450
QuoteOriginally posted by dtmateojr Quote
Equivalence fails miserably when applied to film photography because it is very wrong
How so
If I was to use porta 160nc 135mm film in my 35mm camera and I take an image using my 6x6 with porta 400nc shoot the same landscape at the same DOF,SS and FOV (shot at equivalency )we would find that they would appear to contain the same level of noise when printed at the same 8x10 print. That sound more like proof of equivalency than not

IF you don't believe my then why not have a look at what Kodak says http://www.kodak.com/global/en/professional/support/techPubs/e58/e58.pdf
with regarded to grain pattern you will see that a 4x6 print from a portra 135 160nc has print grain of 30 and a 4x6 print for portra 120/220 400nc has almost the same print grain at 29
who would have ever thought this was possible.

Now I already know what you are going to say well they are using different emulsion between those 2 films, well you are right they have to each film was exposed to different densities of light so you use different emulsions, same as having to shoot at iso 225 for FF and iso 100 for cropped because of the lower density of light hitting the larger sensor. Now I know you will respond bla bla this and that or can we just jump to the insults as witnessed before.

Its funny that you should sidetrack this discussion about the E word from digital to film as Eric has already provided evidence and the underlining physics to the E word. Something you have yet to understand.
03-14-2016, 09:43 PM   #206
Loyal Site Supporter
clackers's Avatar

Join Date: Jul 2013
Location: Melbourne
Photos: Albums
Posts: 7,702
QuoteOriginally posted by Ian Stuart Forsyth Quote
If we use your own interpretation of the DXO data you will find that the k3 is not to far behind the k10 but none here would say that data represents what we see in the field or in the digital darkroom ...
Also you don't see many people using the K10 for its better performance up to iso1600

Well, as one Ian to another, these are DXOMark's actual measurements, whether we like it or not.


The K10's RAW files are more noise-free than the K-3's up to ISO 1600. We have people on the forum who still use the K-10 and in addition like its CCD rendering.


It's true, you would use the K-10D if the cleanest possible RAW file with the highest dynamic range was all you wanted.


After that, it's now post processing - nothing to do with the camera. That RAW file is what you're bringing into Photoshop.


If the 'Real World' is important to you, DXO Mark's Print Tab doesn't represent anything in the field or digital dark room, by definition - it's just all the Screen numbers with a mathematical adjustment - theoretical, not real. Their subject matter will not be your subject matter, their noise reduction assumption will not follow your noise reduction method, either - I can see that from the equation they use.


Now, when you make your JPEG, you can have wider objectives and use the K-3 for its higher resolution.


If you accept data loss because of noise reduction, it has many more pixels for an algorithm to clean up.


The APS-C K-3 downsamples equally well in the DXO 'normalizing' process as the full frame Nikon D750 - it gets about the same 5dB increase. This is of course troubling to the Full Frame Fetishists who worship the 'Print' tab.


Disclosure: I don't know what you use, but I own and shoot both full frame and APS-C.


Guess I didn't have a religious experience when I bought FF that makes some drop in from Canikony land and say "Your crop cameras are junk, everyone needs what I've had to pay for".

Last edited by clackers; 03-14-2016 at 10:21 PM.
03-14-2016, 09:44 PM   #207
Loyal Site Supporter
clackers's Avatar

Join Date: Jul 2013
Location: Melbourne
Photos: Albums
Posts: 7,702
QuoteOriginally posted by Gimbal Quote
Well, (holding up hand) I did.

You do realize that down sampling cleans up a picture? Noise (and details of course) are averaged out and ultimately lost.

And you know that the K50 has more Mp than the A7S, so it has to be down sampled more if one are to compare them printed at the same size. Therefore the K50 cleans up more (and loses more details, details the A7s never captured in the first place due to its lower Mp count.)

So compared pixel by pixel the A7S is way ahead, it has huge pixels compared to the K50. When printed at the same size the gap closes slightly, although the A7S still wins by a great margin due to its larger total sensor area. As expected.


I'm giving you nine out of ten for your answer, Gimbal, it's excellent but for the last bit.


'Total sensor area' could be interpreted by the Full Frame Fetishists as meaning that it's on a 35mm frame.


Of course, as you know, the only relevant areas are the individual surfaces of the pixels themselves (basically, the pitch), not the wafer they're on. And the A7S has some specific technology differences to its siblings in the A7 lineup.

Last edited by clackers; 03-14-2016 at 10:17 PM.
03-14-2016, 09:45 PM   #208
Loyal Site Supporter
clackers's Avatar

Join Date: Jul 2013
Location: Melbourne
Photos: Albums
Posts: 7,702
QuoteOriginally posted by Rondec Quote
Well, I think you need to have some way of normalizing the output of your camera. DXO Mark has picked a certain megapixel mark (8). But you could pick a certain screen size or print size. In the end, the question is, if you print from a K10, K3 or a K-1 at a given size, how do the photos look with regard to noise, colors and dynamic range?

Yes, there is a way of normalizing, Rondec, it's called ratios, and it's already done when DXO Mark do their Screen tab. All of the data for noise, dynamic and tonal ranges, et al, is in ratios.


You know that if the SNR is 0dB, you know that there is a pixel of noise for every pixel of the subject. Every 3dB gives you a 100% improvement.


Those ratios are true regardless of how many pixels are involved or how big the sensor is.


I can bore you with what their 'normalization' step involves, and how ridiculous it actually is to real shooters instead of Poindexters, but you'd better get drunk first.

Last edited by clackers; 03-14-2016 at 10:19 PM.
03-14-2016, 09:51 PM   #209
Loyal Site Supporter
clackers's Avatar

Join Date: Jul 2013
Location: Melbourne
Photos: Albums
Posts: 7,702
QuoteOriginally posted by jsherman999 Quote
There's 'appeal to authority' and then there's citing sources or aggregation of sources. One great tactic folks like you use is to call citations and references an 'appeal to authority' once you realize the citation makes you lose the argument.

You can do literally dozens of things to clear up your misconceptions - check Falk's site, or go root around on dpreview until you encounter Eric Fossum for example who provided information to JJ for some of the read noise information in his article and is one of the clear voices on Equivalence

I've taught some of this stuff, so it's rich hearing you trying to "put one over", Jay!


I'm afraid it's both the quality of the citations and when you quote them (like you made out Clark to agree with you last year- or should I not have brought this up?).


I know Falconeye is a forum member, but as I said last year, his 'essay' is a pointless polemic.


It first of all trots out standard formulas to hook you in, then does a non sequitir with its conclusion that "It is a myth that APS-C cameras "crop the sweet spot" of the image field of lenses which cover a larger image circle or are made for a full registration distance."


I'll forgive him his claim (made twice) that his procedure was empirical, because English is not his first language. It was a desk exercise at best, cherry picking a website. He fails to control variables (and admits it) in this convoluted thought experiment. Two cameras, two lenses, and somebody else's data, manipulated.


You're easily impressed.

Last edited by clackers; 03-14-2016 at 10:08 PM.
03-14-2016, 09:54 PM   #210
Pentaxian




Join Date: May 2009
Location: Ponoka Alberta Canada
Photos: Gallery
Posts: 450
QuoteOriginally posted by dtmateojr Quote
Print SNR is normalised SNR i.e. FAKE SNR. That same normalised graphs can be replicated by moving forward and backward when viewing a print. Moving closer to the print increases PERCEIVED noise and moving farther away decreases PERCEIVED noise. Did I improve the sensor performance by moving forward backward and sideways? Of course not but you equivalence fan boys keep pushing this as truth. It's stupid.
How about I change your words around some

That same normalised graphs prints can be replicated by moving forward and backward when viewing a print negative . Moving closer to the print negative increases PERCEIVED noise and moving farther away decreases PERCEIVED noise. Did I improve the sensor negative performance by moving forward backward and sideways? well no but it defiantly changes how we see noise,

How many here think that the larger negative is made up of the same amount of total light as a smaller one.? More light would have to be converted to the photographic image because of the area of the negative is larger

No one ever said that increasing the size sensor increases the performance per each unit of that sensor, just as the case with the same film on 2 different formats. What has changed is how we view that unit in a image ( Simply put that noise the sensor captured represents a smaller section in the final image) A small format has to be enlarged more to be view at the same size.

If you are to debunk equivalency you are to have to prove that a larger negative does not capture more light, its going to be tuff as most would agree that more photons struck that larger negative
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!
24x36mm, 300mm, advantages, app, aps-c, body, camera, cost, crop, dslr, f/2.8, ff, field, focus, frame, full-frame, k-1, k-3, k3, lens, ovf, pentax, pentax k-3, quality, sensor, sigma, view
Thread Tools Search this Thread
Search this Thread:

Advanced Search


Similar Threads
Thread Thread Starter Forum Replies Last Post
Crop Sensors vs Full Frame :: Crop Or Crap? i83N Photographic Industry and Professionals 44 07-30-2014 06:00 AM
No Pentax Full Frame DSLR, What to Do Now Then? RiceHigh Pentax DSLR Discussion 95 07-19-2014 10:57 AM
The Full Frame NEX is going to be announced in October ! jogiba Non-Pentax Cameras: Canon, Nikon, etc. 30 09-15-2013 03:23 PM
So.... Pentax Full Frame is a sure thing? LFLee Photographic Industry and Professionals 25 11-02-2012 12:55 PM
So It's True....a Full-Frame 'Professional' Samsung Dslr Coming !!!! Confused Pentax DSLR Discussion 9 05-02-2008 04:12 PM



All times are GMT -7. The time now is 04:09 PM. | See also: NikonForums.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