K 7 iso 3200

jstevewhite · 07-15-2009, 07:08 PM

Originally posted by GaryM

Bingo. If the ultimate purpose of this stuff isn't to produce prints, then we would have never needed more than a few megapixels. IMO, the 100% pixel-view method of measuring image quality has seriously distorted the real measure of image quality - which may, perhaps, not be something that can ever be properly conveyed over the internet.

I think there's some balance to be struck here. I plainly admit to being a pixel-peeper, but I moderate that with the evaluation of "real prints"... As was said, the ISO 6400 pix from the K20D, printed at full frame on 8.5x11 photo paper with a pigment based 5-gray+black ink set produces beautiful black and white images reminiscent of everything that was good about Tri-X. The pixel-peeping is what I use to evaluate the 'enlargeability' of a system, if you will. If it's sharp at 100%, I can count on making some pretty large prints and retaining that stinging sharpness these systems are capable of. It's of academic interest to me to understand whether the noise of the K-7 or K20D looks more pleasing (I think it's a wash), but I constantly question the 'real life' view. Your point is VERY well made - the K20D @800 produces images of sufficient quality for a full cover bleed on a glossy magazine, and the 3200 good enough for a full single page.

kittykat46 · 07-15-2009, 07:59 PM

I've shot a bit with the K-7 at ISO 3200 and ISO6400 in somewhat dim conditions.
So far I'm quite pleased with the results.

Using a fast prime like the FA 50mm f/1.4, coupled with the K-7's much improved low-light Autofocus, the overall package really makes a VERY good low-light camera.

"Exposing to the Right" by applying a moderate +E/V, without blowing out any highlights, helps a lot with noise. Viewed full-screen or printed out on A4, its fine.

There's no point trying to pick on a 100% crop at ISO3200 - of course its noisy. But for practical photography - viewing the whole photo on-screen or printing - I'm quite happy with it.

Haakan · 07-19-2009, 11:30 AM

Even a "pushed backlit image" with a lot of noise as the example in this thread can become decent with proper use of noise reduction. I did 5 minutes of post processing of the posted clip and think it improved quite a lot. And if I had access to the original (and even better, in RAW) I think it could become quite OK (in the posted example there are some JPEG artifacts due to the compression of the image)

First the original, and then my processed.

Best regards,
Haakan

vinzer · 07-21-2009, 08:16 AM

Originally posted by Haakan

Even a "pushed backlit image" with a lot of noise as the example in this thread can become decent with proper use of noise reduction. I did 5 minutes of post processing of the posted clip and think it improved quite a lot. And if I had access to the original (and even better, in RAW) I think it could become quite OK (in the posted example there are some JPEG artifacts due to the compression of the image)

First the original, and then my processed.

Best regards,
Haakan

Looks good. Care to divulge your secrets for people like me?

Jonson PL · 07-21-2009, 08:40 AM

Originally posted by augustmoon

handheld, but I got some acceptable results by exposing to the right as far as I could get away with.

RAW from lightroom, no noise correction, but because I exposed to the right, I brought down the brightness a bit and boosted the black level to get a normal looking pic.

the 1600 looks even better. It's definitely "different" looking than K20D noise. More noticeable at first (coarser?) but maybe more acceptable due that it seems to have less chroma in it?

But I think that even ISO 400 (with the K-7) can be ruined if it gets even slightly underexposed.

Looking good, and thanks for the info

Haakan · 07-21-2009, 08:53 AM

Hi vinzer,
not much secrets. What I did was to use Neat Image noise reduction SW (even the free demo version works quite well). Instead of having the program select area itself I selected the trees as "noise area".

I also used Focus Magic (free demo is available, limited to 10 images), to sharpen the image.

I played with the curves somewhat in Photoshop to make it look less "flat" and make the grass in the foreground stand out more. I also used a trick that benjikan has shared in this forum, which is to use a very small radius, but rather high amount of sharpening in photoshop, but being careful to not have artifacts or emphasing noise again. This trick also makes the image look less flat.

philippe_e · 07-21-2009, 09:40 AM

i take one in 3200...its not so bad...no retouch.

**lawsonstone** · 07-21-2009, 10:34 AM

I think the issue in ETTR has to do with how digital camera sensors record data. Exposing to the right seems to cause the sensor to record more shadow data. Correcting back left in PP then brings out the shadow details better.

At least that was what I thought I heard when I thought I understood what I thought was probably an explanation that was not completely wrong....

...educate me folks if I've missed this

krypticide · 07-21-2009, 02:53 PM

ETTR deals with noise from posterization, not from the sensor. Because a sensor's response is linear, but light is exponential, it is advantageous to shoot TTR at a higher ISO than expose "properly" at a lower ISO. You simply collect more data that way, even though the resultant exposure is the same.

07-19-2009, 11:30 AM	#33
Haakan Senior Member Join Date: Jun 2008 Location: Stockholm Photos: Gallery Posts: 182	Even a "pushed backlit image" with a lot of noise as the example in this thread can become decent with proper use of noise reduction. I did 5 minutes of post processing of the posted clip and think it improved quite a lot. And if I had access to the original (and even better, in RAW) I think it could become quite OK (in the posted example there are some JPEG artifacts due to the compression of the image) First the original, and then my processed. Best regards, Haakan Attached Images

07-21-2009, 08:53 AM	#36
Haakan Senior Member Join Date: Jun 2008 Location: Stockholm Photos: Gallery Posts: 182	Hi vinzer, not much secrets. What I did was to use Neat Image noise reduction SW (even the free demo version works quite well). Instead of having the program select area itself I selected the trees as "noise area". I also used Focus Magic (free demo is available, limited to 10 images), to sharpen the image. I played with the curves somewhat in Photoshop to make it look less "flat" and make the grass in the foreground stand out more. I also used a trick that benjikan has shared in this forum, which is to use a very small radius, but rather high amount of sharpening in photoshop, but being careful to not have artifacts or emphasing noise again. This trick also makes the image look less flat. Last edited by Haakan; 07-21-2009 at 11:38 AM. Reason: Correcting spelling errors

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Black & White This is what ISO 3200 used to look like!	brkl	Post Your Photos!	5	01-10-2010 03:35 PM
iso 3200	jolee1990	Troubleshooting and Beginner Help	10	12-22-2009 08:26 AM
K-x vs K7 at iso 3200	artemapei	Pentax DSLR Discussion	21	11-18-2009 08:05 PM
ISO 3200! Only PP in lightroom	metroeloise	Post Your Photos!	3	12-12-2008 07:25 PM
Tests iso 200 to iso 3200 with k100d	Deni	Post Your Photos!	0	06-20-2007 05:17 AM

07-15-2009, 07:08 PM	#31
jstevewhite Veteran Member Join Date: Mar 2009 Location: Midwest Photos: Gallery Posts: 1,407	Originally posted by GaryM Bingo. If the ultimate purpose of this stuff isn't to produce prints, then we would have never needed more than a few megapixels. IMO, the 100% pixel-view method of measuring image quality has seriously distorted the real measure of image quality - which may, perhaps, not be something that can ever be properly conveyed over the internet. I think there's some balance to be struck here. I plainly admit to being a pixel-peeper, but I moderate that with the evaluation of "real prints"... As was said, the ISO 6400 pix from the K20D, printed at full frame on 8.5x11 photo paper with a pigment based 5-gray+black ink set produces beautiful black and white images reminiscent of everything that was good about Tri-X. The pixel-peeping is what I use to evaluate the 'enlargeability' of a system, if you will. If it's sharp at 100%, I can count on making some pretty large prints and retaining that stinging sharpness these systems are capable of. It's of academic interest to me to understand whether the noise of the K-7 or K20D looks more pleasing (I think it's a wash), but I constantly question the 'real life' view. Your point is VERY well made - the K20D @800 produces images of sufficient quality for a full cover bleed on a glossy magazine, and the 3200 good enough for a full single page.

07-15-2009, 07:59 PM	#32
kittykat46 Veteran Member Join Date: Dec 2007 Posts: 923	I've shot a bit with the K-7 at ISO 3200 and ISO6400 in somewhat dim conditions. So far I'm quite pleased with the results. Using a fast prime like the FA 50mm f/1.4, coupled with the K-7's much improved low-light Autofocus, the overall package really makes a VERY good low-light camera. "Exposing to the Right" by applying a moderate +E/V, without blowing out any highlights, helps a lot with noise. Viewed full-screen or printed out on A4, its fine. There's no point trying to pick on a 100% crop at ISO3200 - of course its noisy. But for practical photography - viewing the whole photo on-screen or printing - I'm quite happy with it.

07-21-2009, 08:16 AM	#34
vinzer Veteran Member Join Date: Mar 2007 Location: Philippines Posts: 1,399	Originally posted by Haakan Even a "pushed backlit image" with a lot of noise as the example in this thread can become decent with proper use of noise reduction. I did 5 minutes of post processing of the posted clip and think it improved quite a lot. And if I had access to the original (and even better, in RAW) I think it could become quite OK (in the posted example there are some JPEG artifacts due to the compression of the image) First the original, and then my processed. Best regards, Haakan Looks good. Care to divulge your secrets for people like me?

07-21-2009, 08:40 AM	#35
Jonson PL Veteran Member Join Date: Dec 2006 Location: Copenhagen Photos: Gallery Posts: 1,842	Originally posted by augustmoon handheld, but I got some acceptable results by exposing to the right as far as I could get away with. RAW from lightroom, no noise correction, but because I exposed to the right, I brought down the brightness a bit and boosted the black level to get a normal looking pic. the 1600 looks even better. It's definitely "different" looking than K20D noise. More noticeable at first (coarser?) but maybe more acceptable due that it seems to have less chroma in it? But I think that even ISO 400 (with the K-7) can be ruined if it gets even slightly underexposed. Looking good, and thanks for the info

07-21-2009, 09:40 AM	#37
philippe_e Inactive Account Join Date: Feb 2009 Location: grenoble france Photos: Gallery Posts: 4	i take one in 3200...its not so bad...no retouch.

07-21-2009, 10:34 AM	#38
lawsonstone Site Supporter Join Date: Oct 2008 Location: Wilmore, KY Posts: 376	I think the issue in ETTR has to do with how digital camera sensors record data. Exposing to the right seems to cause the sensor to record more shadow data. Correcting back left in PP then brings out the shadow details better. At least that was what I thought I heard when I thought I understood what I thought was probably an explanation that was not completely wrong.... ...educate me folks if I've missed this

07-21-2009, 02:53 PM	#39
krypticide Veteran Member Join Date: Dec 2007 Photos: Gallery Posts: 1,079	ETTR deals with noise from posterization, not from the sensor. Because a sensor's response is linear, but light is exponential, it is advantageous to shoot TTR at a higher ISO than expose "properly" at a lower ISO. You simply collect more data that way, even though the resultant exposure is the same.