Forgot Password
Pentax Camera Forums Home
 

Reply
Show Printable Version Search this Thread
02-13-2008, 01:59 PM   #61
Veteran Member
Gooshin's Avatar

Join Date: Sep 2007
Location: Toronto, the one in Canada.
Posts: 5,611
also remember that for the most part, unlike hockey, blowing out that lamp on the wall isnt going to ruin the picture as much as blowing out the ice that covers nearly half your shot.

02-13-2008, 02:08 PM   #62
Veteran Member
attack11's Avatar

Join Date: Apr 2007
Location: Ottawa, ON - Canada
Photos: Gallery
Posts: 659
QuoteOriginally posted by Gooshin Quote
all i know, from personal experience, is that unpleasant noise originating in high ISO pictures (1600, since i dont have options between that and 800) is largely determined by whether or not i properly or over expose a shot.

trying to "push" underexposed high ISO pictures will most definetly get you horrible results, if thats how you do things then i'm not at all surprised at your distaste for using high ISO.
you're assuming too much about how i shoot. this was situation dependant; and i brought it up because everyone relies well lit / high iso images for the basis of opinion. it's mostly misleading when you do that; consider this image and this image. the people are literally sitting beside each other, but one was closer to an over head light. if both had been shot at 1600 they'd both look much worse due to loss of detail from increased light sensitivity.

if your monitor is calibrated, you'll also see a grid in the blue of the hoody.

QuoteOriginally posted by Gooshin Quote
trying to "push" underexposed high ISO pictures will most definetly get you horrible results, if thats how you do things then i'm not at all surprised at your distaste for using high ISO.
this makes no sense. the noise is recorded in the raw; turning up the exposure is not creating more noise.

Last edited by attack11; 02-13-2008 at 02:14 PM.
02-13-2008, 02:22 PM   #63
Veteran Member
Gooshin's Avatar

Join Date: Sep 2007
Location: Toronto, the one in Canada.
Posts: 5,611
QuoteOriginally posted by attack11 Quote
this makes no sense. the noise is recorded in the raw; turning up the exposure is not creating more noise.
turning up exposure, digitaly, means you turn up the signal, when you turn up the signal, you amplify the digital information, if there is noise in that information, you amplify the noise.

this is how digital sensors work, this is how "iso" in digital sensors work, its all about electricity and signal amplification.
02-13-2008, 02:27 PM   #64
Veteran Member
Gooshin's Avatar

Join Date: Sep 2007
Location: Toronto, the one in Canada.
Posts: 5,611
this is why the 5D is going to slowly get left in the dust, its only advantage now is larger pixles, hence better iso control since the signal is cleaner

but if the smaller pixel is designed to capture light with the same clarity and information loss as the larger sensor in the 5D, you will not see a difference in IQ

the 5D is 3 years old, have faith in the future development of APS-C sensors.

02-13-2008, 04:03 PM   #65
Veteran Member
attack11's Avatar

Join Date: Apr 2007
Location: Ottawa, ON - Canada
Photos: Gallery
Posts: 659
QuoteOriginally posted by Gooshin Quote
turning up exposure, digitaly, means you turn up the signal, when you turn up the signal, you amplify the digital information, if there is noise in that information, you amplify the noise.

this is how digital sensors work, this is how "iso" in digital sensors work, its all about electricity and signal amplification.
who told you that? you under expose the noise and then get say iso800 noise with iso1600 exposure (without the degradation due to the higher sensitivity). you've never done this apparently, so why are you guessing at how it works? regardless, it won't stop you from getting absolutely horrible noise in the dark once you shoot above iso1000.


QuoteOriginally posted by Gooshin Quote
this is why the 5D is going to slowly get left in the dust, its only advantage now is larger pixles, hence better iso control since the signal is cleaner
that's subjectional and you've stated your opinion of ff vs aps-c regarding how the light is recorded (dof/bokeh). to some people ff dof matters.

Last edited by attack11; 02-13-2008 at 04:21 PM.
02-13-2008, 05:55 PM   #66
Veteran Member
Tom M's Avatar

Join Date: Aug 2007
Location: Lincoln Park, NJ
Photos: Gallery
Posts: 782
If you want to underexpose your digital shots and then whine like a school girl about how your images are all noisy, then by all means do so. But, can you do it in private? Thanks.

Ps. Not all film techniques translate to digital. And, I've had enough of this nonsense conversation.
02-13-2008, 06:27 PM   #67
Veteran Member
Gooshin's Avatar

Join Date: Sep 2007
Location: Toronto, the one in Canada.
Posts: 5,611
QuoteOriginally posted by attack11 Quote
who told you that?
thats how digital light sensors work, period.


QuoteOriginally posted by attack11 Quote
you've never done this apparently, so why are you guessing at how it works?
no i havent, because its a silly idea.

QuoteOriginally posted by attack11 Quote
regardless, it won't stop you from getting absolutely horrible noise in the dark once you shoot above iso1000
please refer to examples below:


since you're alot like me, and unless you either do it or have someone show you something, taking someones word for whatever is difficult.

so here are some very controlled tests for your pleasure.

all done off a tripod, with mirror lock up, Pentax FA43 at f4.0 manualy focused.



FIRST ROUND:

ISO 1600 VS ISO 800 PUSHED.

camera meters for 1/6 seconds, therefore both shots are done at 1/6 to underexpose the 800.

iso 1600


iso 800 + 1ev in lightroom




even at +1ev in lightroom the shadows dont fill as much as iso 1600, there is more black in the underexposed shot, and rightly so, noise in my opinion is identical.



ROUND TWO, more extreme.

ISO 3200 VS ISO 200 PUSHED 4 stops

metered at 1/10 for iso 3200


iso 3200



iso 200 exposed 4 stops in lightroom



as you can see, ISO 200 pushed still amplifies some noise in the areas where there was some noise hiding, but overall, nothing happened, why? because you cant increase the signal of something thats not there!


ROUND THREE, reverse fight

ISO 200 underexposed 4 stops in camera VS Properly Exposed ISO 3200, underexposed 4 stops in lightroom.

ISO 3200 underexposed (pulled back?)




ISO 200 underexposed in camera (not enough shutter)






Now that i have done these experiments, i revert back to my original belief, its better to over expose than to underexpose (unless ofcourse you are dealing with a scene with significant amount of differential in lighting, where its better to have more black than blown out whites)


remember that you cant "push" information that is not there

when you push ISO 800, if for any given pixel, you dont increase the noise, that means by definition you cant increase any of the light either, because there isnt any!
02-13-2008, 06:32 PM   #68
Veteran Member
Gooshin's Avatar

Join Date: Sep 2007
Location: Toronto, the one in Canada.
Posts: 5,611
go nuts with the originals:

Zenfolio | Serge Guschin | pushing film

02-13-2008, 06:36 PM   #69
Veteran Member
Gooshin's Avatar

Join Date: Sep 2007
Location: Toronto, the one in Canada.
Posts: 5,611
some 200% crops or so

this is the ISO200 underexposed vs ISO3200 underexposed in lightroom

absolutely zero difference in detail, they both suck. Except that the ISO3200, even underexposed, has significantly more light in the shadow areas.



02-13-2008, 08:36 PM   #70
Veteran Member
jeffkrol's Avatar

Join Date: Sep 2006
Location: Wisconsin USA
Photos: Gallery
Posts: 8,435
Interesting read

QuoteOriginally posted by Gooshin Quote
all i know, from personal experience, is that unpleasant noise originating in high ISO pictures (1600, since i dont have options between that and 800) is largely determined by whether or not i properly or over expose a shot.

trying to "push" underexposed high ISO pictures will most definetly get you horrible results, if thats how you do things then i'm not at all surprised at your distaste for using high ISO.
Think both of you may want to ponder Julia's threads and some responses:
Trying to "push" underexposed high iso shots is bad. Pushing low iso shots is not bad....and may be better than properly exposed high iso shots (at equal shutter/aperature) ...
Underexpose at ISO 800 can be better than proper exposure at HI-1: Nikon D3/D2/D1 Forum: Digital Photography Review
As a summary:
I used to think that it was better to properly expose at a higher ISO rather than runa risk of underexposing at a lower ISO, but I was convinced by some experiments done by another dpreview forum member (Julia Borg) that, if you shoot RAW and have a good RAW processor, it does not matter much whether you shoot at proper exposure at ISO HI-1 or underexpose by one stop at ISO 800. In fact, in some cases which I still don't fully understand, it's possible to get better results, not raising the ISO at all and doing all of the "pushing" in a good RAW processor. See these prior threads for the discussion of this including a bunch of tests: push and noise - a sample: Nikon D3/D2/D1 Forum: Digital Photography Review and oh no: Nikon D3/D2/D1 Forum: Digital Photography Review.

There is some logic to this because both exposures are giving you the exact same amount of photons onto the sensor. In the "underexposed" ISO 800 case, you're pushing the exposure in the RAW processor. In the "properly exposed" HI-1 case, you're pushing the exposure in the camera with amplification before recording the RAW values.

In both cases, you got the exact same amount of light photons to each pixel on the sensor and thus ended up with the same signal-to-noise ratio as the signal comes off the sensor. In one case, you amplify the signal more in the camera (the HI-1 case). In the other case, you amplify the signal some in the camera and some in the RAW processor (the ISO 800 case). There are some differences in the two methods of amplification, but according to the experiments I saw, you can get pretty similar results with either shot.

In fact, when doing the ISO amplification in the camera in an irreversible manner, you are vulnerable to permanently blowing highlights whereas if you push in "underexposed" case, you won't blow highlights in the RAW file and can freely push it only as much as you need.

The one gotcha here is that apparently ACR doesn't "push" underexposed shots well compared to other RAW processors so it is not advised to use it.
02-13-2008, 08:38 PM   #71
Veteran Member
Gooshin's Avatar

Join Date: Sep 2007
Location: Toronto, the one in Canada.
Posts: 5,611
QuoteOriginally posted by jeffkrol Quote
Think both of you may want to ponder Julia's threads and some responses:
Trying to "push" underexposed high iso shots is bad. Pushing low iso shots is not bad....and may be better than properly exposed high iso shots (at equal shutter/aperature) ...
Underexpose at ISO 800 can be better than proper exposure at HI-1: Nikon D3/D2/D1 Forum: Digital Photography Review
As a summary:
I used to think that it was better to properly expose at a higher ISO rather than runa risk of underexposing at a lower ISO, but I was convinced by some experiments done by another dpreview forum member (Julia Borg) that, if you shoot RAW and have a good RAW processor, it does not matter much whether you shoot at proper exposure at ISO HI-1 or underexpose by one stop at ISO 800. In fact, in some cases which I still don't fully understand, it's possible to get better results, not raising the ISO at all and doing all of the "pushing" in a good RAW processor. See these prior threads for the discussion of this including a bunch of tests: push and noise - a sample: Nikon D3/D2/D1 Forum: Digital Photography Review and oh no: Nikon D3/D2/D1 Forum: Digital Photography Review.

There is some logic to this because both exposures are giving you the exact same amount of photons onto the sensor. In the "underexposed" ISO 800 case, you're pushing the exposure in the RAW processor. In the "properly exposed" HI-1 case, you're pushing the exposure in the camera with amplification before recording the RAW values.

In both cases, you got the exact same amount of light photons to each pixel on the sensor and thus ended up with the same signal-to-noise ratio as the signal comes off the sensor. In one case, you amplify the signal more in the camera (the HI-1 case). In the other case, you amplify the signal some in the camera and some in the RAW processor (the ISO 800 case). There are some differences in the two methods of amplification, but according to the experiments I saw, you can get pretty similar results with either shot.

In fact, when doing the ISO amplification in the camera in an irreversible manner, you are vulnerable to permanently blowing highlights whereas if you push in "underexposed" case, you won't blow highlights in the RAW file and can freely push it only as much as you need.

The one gotcha here is that apparently ACR doesn't "push" underexposed shots well compared to other RAW processors so it is not advised to use it.

well accordign to my own little self made test, the results were not identical... so...

anyway the way i like to think about it (and assuming no blow-outs), its better to turn white into black than trying to turn black into white.

atleast if you over expose, or properly expose, you have information to work with, as in my case with ISO 200, if you severly under expose, yoou will be left with NO information at all, and no amount of photoshoping will be able to add light to that area completly void of digital information.

Last edited by Gooshin; 02-13-2008 at 09:11 PM.
02-13-2008, 09:11 PM   #72
Veteran Member
jeffkrol's Avatar

Join Date: Sep 2006
Location: Wisconsin USA
Photos: Gallery
Posts: 8,435
QuoteOriginally posted by Gooshin Quote
well accordign to my own little self made test, the results were not identical... so...

anyway the way i like to think about it (and assuming no blow-outs), its better to turn white into black than trying to turn black into white.

at least if you over expose, or properly expose, you have information to work with, as in my case with ISO 200, if you severly under expose, yoou will be left with NO information at all, and no amount of photoshoping will be able to add light to that area completly devoid of digital information.
Yes but you always have to remember (at least I do) if a proper exposure is f4 at 1/10 sec for a 1600 ISO, the EXACT same amount of light hits the sensor and the EXACT same amount of detail is there regardless of the fact that you set the camera to 200 and -3 EV IF the shutter speed and aperature are identical ( f4 at 1/10 sec). The brighter 1600 iso image is only created by multiplying the original data....... Digital cameras really have no iso. Now if f4 at 1/10 sec doesn't allow an efficient amount of photons in it makes no difference what the iso is as you really do state. You need a proper exposure but it is ISO independent. Only sensor dependent. The rest is just math. From what I read the k10 has no amplification. All isos are strictly "developed" w/ a multiplier....
And a little more complication:
the higher ISO setting is used in camera the more highlight clipping occurs in linear amplification/multiplication scheme.
Re: No free lunch?: Nikon D3/D2/D1 Forum: Digital Photography Review
Take it for what it's worth. Her logic and the examples through weeks of debate (wait till you hear about Uni-Whitebal) seem pretty convincing.
Oh and it is quite Raw developer dependent... and that includes the one built into the camera (to create jpgs).
Her father is the co-creator of Raw Magik Lite software. Interesting program. I downloaded a Pentax D file for him to incorpoate into his program. Uses floating point math for development (making it less prone to rounding errors but a bit pokey . Just a side note.

Last edited by jeffkrol; 02-13-2008 at 09:17 PM.
02-13-2008, 09:26 PM   #73
Veteran Member
Gooshin's Avatar

Join Date: Sep 2007
Location: Toronto, the one in Canada.
Posts: 5,611
thats nuts, heheh

guess i'll have somethign to read tommorow at work thanks for the links.
02-14-2008, 01:11 AM   #74
Senior Member




Join Date: Jan 2008
Location: Boston
Posts: 142
Can we please get back to discussing the 5d vs k20d?

And BTW, kicking the 5D is like abusing a warm and fuzzy dog. The 5D is/was a landmark camera that's still in a class of one (D3 doesn't count; it's over twice as expensive) *three* years after it was released. Let's wait until we see how the K20D fares come 2011. As this is a Pentax forum, I obviously hope the K20D does extremely well, if only because I want Pentax to keep on producing these great cameras.
02-14-2008, 09:22 AM   #75
Veteran Member
Gooshin's Avatar

Join Date: Sep 2007
Location: Toronto, the one in Canada.
Posts: 5,611
QuoteOriginally posted by twinda1 Quote
Can we please get back to discussing the 5d vs k20d?

And BTW, kicking the 5D is like abusing a warm and fuzzy dog. The 5D is/was a landmark camera that's still in a class of one (D3 doesn't count; it's over twice as expensive) *three* years after it was released. Let's wait until we see how the K20D fares come 2011. As this is a Pentax forum, I obviously hope the K20D does extremely well, if only because I want Pentax to keep on producing these great cameras.
we are discussing, an individual brought up doubts of the K20D's ability to produce stellar high ISO pictures and that even the 3 year old 5D will kick its ass.

i'm trying to show how even a flimsy little K100D can produce really good high ISO pictures, and that Pentax would not be stupid to downgrade their sensors, or allow their sensors to produce downgraded images.
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!
5d, camera, dslr, k20d, photography
Thread Tools Search this Thread
Search this Thread:

Advanced Search


Similar Threads
Thread Thread Starter Forum Replies Last Post
For Sale - Sold: K20D, Battery Grip, NEW K20D battery, cable remote (Worldwide) Albert Siegel Sold Items 6 09-23-2010 08:02 AM
In Canada: Summer Rebates for K20d or K20d plus lens Pentaxtic Pentax DSLR Discussion 1 07-17-2009 11:34 AM
Magic Lantern Guides: Pentax K20D and MasterWorks: Jumpstart Guide for the K20D. Reportage Troubleshooting and Beginner Help 10 02-12-2009 10:24 PM



All times are GMT -7. The time now is 03:07 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