Forgot Password
Pentax Camera Forums Home
 

Reply
Show Printable Version Search this Thread
09-09-2014, 09:21 PM - 1 Like   #1
Senior Member




Join Date: Mar 2010
Location: Lewiston, NY
Posts: 108
Noise and other noise

Musings on a few topics ...

The K5 has an ISO 80 setting. What does that really mean? Just my opinion but I would think it goes something like this. ISO is an abbreviation for International Standards Organization. In olden days we used to talk about ASA which I think stood for American Standards Association. Now some cameras have sensitivity settings like L1, L2, H1, H2, but the K5 has ISO 80. Perhaps that is because Pentax is able to use the ISO trademark since the K5 does comply with the complete definition of ISO 80. That means, within some tolerance, it is equivalent to an ISO standard for film grain and sensitivity documented as standard number 80. Other manufacturers use things like L1, L2 simply because these sensitivities do not comply with an ISO standard, therefore they are unable to use the ISO trademark.

Noise: Our camera club had a discussion about "native ISO". To me this makes no sense. Now my electronics education was when the vacuum tube was king, but pretty well in any amplification system, the more gain, the more noise. With semi-conductor sensors I expect the least noise occurs when the amplification gain is lowest and in the K5 that is at ISO 80. If that is what is meant by native, OK.
This discussion went on to claim what I think is another misconception; that mid ISO settings are interpolated and it is better to shoot 1/3 stop lower than an integer multiplier of native than it is to shoot 1/3 stop higher. In other words. since the native ISO of the K5 is 80 and the next integer multiplier is 160, it less noisy to shoot at ISO 125 than ISO 100 etc. This rumor may have come from some video/Technicolor issue, but IMO I call it Kazunga when it comes to still shot cameras. So I did a test...

Completely dark room, lens and viewfinder caps on, and the camera covered by a black lightproof blanket. Let's say about -8ev (who knows, it was too dark to see). The process is shoot, increase sensitivity by 1/3 stop, shoot again. Then take all those files (in DNG format) and import them into photoshop as layers. Apply a threshold adjustment layer (set at level 1) at the top of the layer stack, and make 4 101x101 pixel samples around the middle of the frame. Then average the RGB levels from all 4 samples and plot the graph.

This is what I got. A couple of "shelves" in the curve, but nothing other than what I would call a sample anomaly. The results I got, are pretty much what I expected.

Thoughts? Comments? By the way, although I don't have the curve to show, a friend did the same thing with his Nikon D7100. The K5 was less noisy by far.

Attached Images
 
09-09-2014, 09:27 PM   #2
Veteran Member




Join Date: Nov 2013
Location: North Bohemia
Photos: Gallery
Posts: 1,251
Thanks for sharing. Now I understand a bit more why my K30 uses ISO 200 when highlight correction kicks in.
09-09-2014, 10:52 PM - 1 Like   #3
Senior Member




Join Date: Jan 2014
Location: Halifax, UK
Posts: 120
QuoteOriginally posted by AldaCZ Quote
Thanks for sharing. Now I understand a bit more why my K30 uses ISO 200 when highlight correction kicks in.
??

I thought the highlight protection shot at an ISO level ( 1 stop?) below the one selected for exposure to protect the highlights and then brightened the picture to the correct level in post production. As 100 is the lowest ISO level on k30, it can't use highlight protection at this ISO level.
09-10-2014, 05:12 AM - 1 Like   #4
Veteran Member
kh1234567890's Avatar

Join Date: Sep 2010
Location: Manchester, UK
Photos: Gallery
Posts: 2,653
The discontinuity at about ISO2000 is when the camera switches from analog to digital 'amplification'. Above this ISO any extra sensitivity is just a digital multiplication at the expense of dynamic range.

09-10-2014, 06:00 AM   #5
Senior Member




Join Date: Mar 2010
Location: Lewiston, NY
Posts: 108
Original Poster
QuoteOriginally posted by kh1234567890 Quote
The discontinuity at about ISO2000 is when the camera switches from analog to digital 'amplification'. Above this ISO any extra sensitivity is just a digital multiplication at the expense of dynamic range.
Ahha. That makes sense. I was wondering if it was truly an anomaly. It did that quite consistently from trial to trial. Thank you.
09-10-2014, 06:08 AM   #6
Pentaxian
normhead's Avatar

Join Date: Jun 2007
Location: Near Algonquin Park
Photos: Gallery | Albums
Posts: 40,442
Which explains why 1600 is the absolute highest I like to shoot with a K-5. And why I think of 100 and 200 as pretty much interchangeable, but I'll go to 400 under duress, and 800 under even more duress. Anyone know what the curve for a K-3 is?
09-11-2014, 08:21 PM   #7
Senior Member




Join Date: Mar 2010
Location: Lewiston, NY
Posts: 108
Original Poster
QuoteOriginally posted by normhead Quote
Which explains why 1600 is the absolute highest I like to shoot with a K-5. And why I think of 100 and 200 as pretty much interchangeable, but I'll go to 400 under duress, and 800 under even more duress. Anyone know what the curve for a K-3 is?

I think you are pretty safe up to ISO 800 for most lighting situations. The average amount of white in the samples is only 25. You could probably see an RGB value of 25, 25, 25 on a pure black background, but since this is just an average of 101x101 pixels the actual noise would be distributed and not clustered in any one place. If you are shooting a fairly bright scene you wouldn't see it at all. Based on what kh mentioned though I'd keep even bright shots to a max of 1600. The last thing I want to sacrifice is dynamic range.

09-11-2014, 11:38 PM   #8
Veteran Member




Join Date: Nov 2013
Posts: 4,854
QuoteOriginally posted by sandilands Quote
I think you are pretty safe up to ISO 800 for most lighting situations. The average amount of white in the samples is only 25. You could probably see an RGB value of 25, 25, 25 on a pure black background, but since this is just an average of 101x101 pixels the actual noise would be distributed and not clustered in any one place. If you are shooting a fairly bright scene you wouldn't see it at all. Based on what kh mentioned though I'd keep even bright shots to a max of 1600. The last thing I want to sacrifice is dynamic range.
Iso 800 on K5 or K3 I can see the difference. In good light this is really ok and there isn't much against it. On dim light, it already start to show visible noise, even at screen size and if you didn't denoise it, it would be disturbing.

iso 800 for me if the last good isos available. 1600 is still ok for a nice shoot while 3200+ is really starting to be disturbing and clearly visible in output.

It also a lot dependant of having the right exposure: noise is in shadows. So iso 1600 with lot of highlight will be really good, while iso 800 with lot of shadows is already bad. The subject itself also count. Some subject need lower isos than others. I take car shoot at 3200isos, and they look, very very good to me. While landscape shoot iso 3200 (by error) I have seen of some friend K30 in daylight looked crap with noise visible in the sky.
09-20-2014, 04:00 PM   #9
Veteran Member




Join Date: May 2008
Location: Virginia
Posts: 376
Only ISO 800 out of a K-5? I just ordered a K-5ii and thought I would get good results at 1000 or 1250. It sounds like these cameras, no matter the brand, all pretty much do the same thing--nice up to 400 and afterward the nasty stuff starts coming in. My "ancient" Nikon D200 is pretty much the same. I guess one has to spend $6,000 to get any real improvement.
09-20-2014, 04:07 PM   #10
Veteran Member
kh1234567890's Avatar

Join Date: Sep 2010
Location: Manchester, UK
Photos: Gallery
Posts: 2,653
https://www.pentaxforums.com/forums/6-pentax-dslr-discussion/64295-pentax-hig...post-here.html
09-20-2014, 04:20 PM   #11
Veteran Member




Join Date: May 2008
Location: Virginia
Posts: 376
Some nice stuff there. I am feeling a little better now.
09-20-2014, 04:29 PM   #12
Veteran Member




Join Date: Jun 2009
Posts: 11,913
QuoteOriginally posted by zx-m Quote
nice up to 400 and afterward the nasty stuff starts coming in.
Let's not go overboard.

I used to quite happily shoot birds with my K-x at ISO 1600, in order to get the DOF and shutter I needed sometimes, and nothing ugly would happen. Pic related:


Scarlet Honeyeater w 'Orange Marmalade' grevillea

With the K-5 and K-3 I would happily shoot gigs and sports at ISO 3200, with nothing horrific about the images at all.

It's not all gloom and doom above ISO 800 with the K-3 or K-5. Something must be wrong with people's cameras or technique if they believe so.
09-20-2014, 05:26 PM   #13
Veteran Member




Join Date: May 2008
Location: Virginia
Posts: 376
Beautiful image...
09-21-2014, 01:30 PM   #14
Veteran Member




Join Date: Nov 2013
Posts: 4,854
QuoteOriginally posted by zx-m Quote
Only ISO 800 out of a K-5? I just ordered a K-5ii and thought I would get good results at 1000 or 1250. It sounds like these cameras, no matter the brand, all pretty much do the same thing--nice up to 400 and afterward the nasty stuff starts coming in. My "ancient" Nikon D200 is pretty much the same. I guess one has to spend $6,000 to get any real improvement.
My understanding is that current sensors gain 1-2 EV compared to what you could expect 4-5 years ago fom say in Pentax land a K7. With good denoising software (DxO, Lightroom), you also make high iso much more pleasing to see than before. 1-2 EV more, depending on the scene.

400isos for examples, there no way for me to see it on normal viewing size.

And 800isos is still very very good, giving nice photos in most circumstances, except really in the shadows. There at least 1 EV difference in performance between what you get from high isos in good light, good exposure, fast shutter speed than with slow shutter shitty light.

1600 is like the limit. Under 1600, most of the time this is very good, more than 1600, and there more post processing involved, and there is visible quality deterioration. Let say that if you shoot the right exposure, correct well the white balance and find the good color tones, 3200 isos can be supprisely good, really. Many time you shoot something with bad lighting anyway, and the results are so-so.

---------- Post added 09-21-14 at 10:36 PM ----------

QuoteOriginally posted by rawr Quote
Let's not go overboard.

I used to quite happily shoot birds with my K-x at ISO 1600, in order to get the DOF and shutter I needed sometimes, and nothing ugly would happen. Pic related:


Scarlet Honeyeater w 'Orange Marmalade' grevillea

With the K-5 and K-3 I would happily shoot gigs and sports at ISO 3200, with nothing horrific about the images at all.

It's not all gloom and doom above ISO 800 with the K-3 or K-5. Something must be wrong with people's cameras or technique if they believe so.
This is because here the light is good and there nice contrasty colors. For this 1600iso will do a good job. You did need more isos for more speed or the focal lens. This typically work better than to bump the isos due to bad light quality.

Noise destroy color deph, dynamic range and subtle tones. This is far less visible on a good contrasty shoot than on a dim light shoot.
09-21-2014, 02:04 PM   #15
Veteran Member




Join Date: May 2008
Location: Virginia
Posts: 376
QuoteOriginally posted by Nicolas06 Quote
This is because here the light is good and there nice contrasty colors. For this 1600iso will do a good job. You did need more isos for more speed or the focal lens. This typically work better than to bump the isos due to bad light quality.

Noise destroy color deph, dynamic range and subtle tones. This is far less visible on a good contrasty shoot than on a dim light shoot.
This is so true. When the light is good, some higher ISOs look okay. When it is dim/poor, 800 is a stretch. Things have changed, but not all that much.
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!
camera, color, contrasty, curve, details, dslr, ev, experiments, gain, gear, iso, isos, k-5, k5, lace, layer, light, multiplier, noise, photography, quality, sensitivity, shutter, target, test, tones
Thread Tools Search this Thread
Search this Thread:

Advanced Search


Similar Threads
Thread Thread Starter Forum Replies Last Post
Bitrate and noise neostyles Canon, Nikon, Sony, and Other Camera Brands 2 07-26-2014 04:18 AM
K-5 noise even with ISO 200 (was: K-5 IIs noise) lunany Pentax K-5 & K-5 II 9 12-28-2013 06:54 PM
Lenses and noise Na Horuk Pentax SLR Lens Discussion 8 02-13-2013 06:45 PM
Noise...noise..do I need software? ejbpesca Photographic Technique 15 04-07-2011 03:39 PM
Noise on other brands fishy Pentax DSLR Discussion 17 07-22-2008 04:57 PM



All times are GMT -7. The time now is 07:50 PM. | 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