[er1kksen]

That's a pretty darn impressive NR job! I wish other software makers priced their products like topaz does... (cough adobe cough).

[LeeRunge]

That's a pretty darn impressive NR job! I wish other software makers priced their products like topaz does... (cough adobe cough).

Yes I am very impressed with the Topaz bundle. You can get good results with LR but the presets work pretty well with Topaz. It's a shame they aren't all combined so you could knock them all out in one pass.

[LeeRunge]

I should note that it seems like turning off NR helps Denoise out. I tried it with my Canon SX200IS which has it on by default and the results were a lot worse despite the picture being decent by default at it's max 1600 iso.

[JohnBee]

Decided to take a few late night test pictures at ISO 6400 with the K7.

Well I like it!
What base file were you working with(JPG, RAW etc)
Also, did you try to see how much of the wood grain you could preserve with TDN?

On a side note, I've found a number of objects to be easier to work with than others.
Glass or reflections and most of the uniform surfaces(for4 example) seem to be the most forgiving. Though when textures come into the scene, difficulty levels seem to escalate quite rapidly.

Either way, you image looks very good(much better than average) imo.

[LeeRunge]

I started with a bunch of shots around the hotel room set at 2mp with NR turned off. I tried with/without shadow/highlight correction.
Using Denoise it seems to respond best to no NR so fine grain is good.
I used the other tools as well "degrunge" in Clean. I used DeJPEG and one of the high compression settings as well.
The "degrunge" knocked out some of the woodgrain. Out of them all I think Denoise is the best as you can fix most color/exposure quickly with Lightroom. I should note that you can fix a lot of noise issues with Lightroom alone.
This is like another special tool in your box to fix noise and it seems to produce good results. I'll be buying Denoise when the trial is up and I'll decide if I want the bundle as I play with the settings through the month.
I took some pictures of palms outside here in Dubai but they are a mixed bag as far as the leaves being smudged out. You can apply a light/medium pass though and get a little noise but it's pretty clean at 6400, certainly usable for internet/small prints.
The K20/K-7 seem to do alright at max iso if you take the time to learn a few techniques in PP to wring the noise out of your images.
Johnbee your portraits came out excellent as well and those inspired me to take some time to give the software a fair trial.
It does take a good amount of time to become intimate with all the quirks in settings that a particular camera prefers. The K-7 metering is getting more familiar each day as my experience base with the camera grows. Really the only thing that has disappointed me significantly is the long exposures.
I with Denoise was a standalone product for more people to have access. If it could be worked into Google's Picasa (if it handled RAW) then it would probably sell a bunch more as well as meeting most casual users needs.
Lee
F
Added another, this one has a light pass (JPEG) with Denoise as you can see the detail in the Brush. ISO 6400 2 mp JPEG with no NR. This one is 4 Seconds at f 8 DA* 16-50 Firmware 1.10

[bayareahdv]

Were the above fairly long exposures (many seconds)? Or perhaps underexposed and then pushed in PP? Or shot RAW with no NR? Or maybe you shot JPEG and then tried to adjust color/WB a lot? I ask because I'm always curious about what variables affect the appearance of noise in different cameras. I know I get wildly different results on my K200D at the same ISO depending on factors like that. Hopefully you won't be offended if I observe that a lot of posted ISO 6400 shots from the K-x look noticeably better than the above, but I've definitely others that look quite similar. It's no reflection on the photographer - it seems to be these other factors that make the difference.

As a random point of comparison, the K200D *in general* won't do as well as the K-x, all else equal. But sometimes for whatever reason things come together and a very high ISO shot comes out noticeably better than I'd expect. Here's an example from the other night. There was basically no direct light on this fellow, which I find is usually a recipe for terrible results. Since my camera maxes out at ISO 1600, I shot there, but almost two stops underexposed and pushed in PP. So it's within half a stop of the equivalent of ISO 6400. Yet for some reason, it doesn't actually look that bad:

[IMGTALL]http://marcsabatella.zenfolio.com/img/s8/v10/p894954004-5.jpg[/IMGTALL]

It was handheld with I believe a 1/2 sec exposure time with 16-45 at f4 ... they where taken handheld at 11pm ... I also only take raw and did not apply any NR post ... second photo has a slight crop ... The ones I took earlier at iso 200 obviously look much better this was more an experiment as I didn't have a tripod on me.

[simico]

As a random point of comparison, the K200D *in general* won't do as well as the K-x, all else equal. But sometimes for whatever reason things come together and a very high ISO shot comes out noticeably better than I'd expect. Here's an example from the other night. There was basically no direct light on this fellow, which I find is usually a recipe for terrible results. Since my camera maxes out at ISO 1600, I shot there, but almost two stops underexposed and pushed in PP. So it's within half a stop of the equivalent of ISO 6400. Yet for some reason, it doesn't actually look that bad:

Given the conditions you mentioned, it looks surprisingly good to me.

[Marc Sabatella]

What's weird is that I've got a bunch of other shots from the same setting in the same conditions with more or less the same exposure and they don't look nearly as good. But I'm right at the point - between 1.5EV and 2EV push from ISO 1600 - where the results seem to drop off a cliff for me most of the time. Difference in his skin tones or the clothes he is wearing versus those of the subjects that were not treated as kindly, or differences in the color of the reflected light landing on him - who knows.

[EdMiller]

*ist DL
ISO-3200
f/5.6

No PP but was re-sized for the forum

[EdMiller]

Something a little different. Using the High ISO abilities of the K-x to allow me to capture 300mph cars at night with high shutter speeds and the 55-300.

Bristol NHRA night Racing. DFine Noise reduction and a bit of PP in aperture.

ISO 1600


ISO 3200

Both are very well done. I really like the dragsters...

[neoyoda]

Some great photos in this thread! I took this picture for the "Single in September" challenge using my K-x with a Takumar 50mm 1.4 lens. 6400 ISO, handheld with only light from shop windows.



I know this photo is maybe not technically that great but I totally love it. I caught this after the 2nd Annual Oshkosh Zombie Walk with almost no daylight left, handheld. You can see and hear this zombie playing in the following video, my first attempt at video on the K-x.
Oshkosh Zombie Walk Music Video IMGP1383 | Flickr - Photo Sharing!

You can also catch other shots of the event here (but mostly taken with my kit zoom lens, not the Takumar.)
http://www.flickr.com/photos/jyoda/s...7624757357837/

[Eric Seavey]

After seeing some very nice NR results using Topaz Denoise from LeeRunge, I downloaded the trial version to compare it to Noise Ninja. I have been using Noise Ninja for years, and am proficient at tweaking the settings. But I thought I'd give Topaz Denoise a fair trial, since I am always on the lookout for better products. In the K7 ISO 6400 photos below, I tried tweaking each plugin to obtain the best noise reduction, color balance and sharpness with both programs, and for comparison reasons, I also included the shots without noise reduction.

I still like the Noise Ninja performance best though the black in Topaz Denoise look much darker.

No Noise Reduction 1


Noise Ninja 1


Topaz Denoise 1


No Noise Reduction 2


Noise Ninja 2


Topaz Denoise 2

[mattdm]

I still like the Noise Ninja performance best though the black in Topaz Denoise look much darker.

In those images, the Noise Ninja results are hands-down the best. You can always adjust the curves a bit to darken blacks if you want.

[jgredline]

I too have both noise ninja and Topaz D5. What I have found is that for certain types of images like flowers, landscapes and buildings, noise ninja works better.For ''people'' images. Topaz D5 works better. But coming from me, take is as if it where a grain of salt.

[JohnBee]

I still like the Noise Ninja performance best though the black in Topaz Denoise look much darker.

I will admit, it took a while to understand and get TDN to work properly coming from wavelet based noise reduction software like NNinja and Noiseware Pro. However... I can assure you that TDN is far and away better at noise reduction than both those programs.

Looking at your samples I'd say you didn't optimize your the red and blue channels properly. Additionally, there is highlight level noise where none should be had, so you might want to try and adjust that slider as well.

One quick and dirty way to get a better feel for what's going on with TDN is to make use of the channel viewing options. This way you can better identify the type
of noise you are working with and how to best address it.

PS. I'm posting this entirely on assumption. Without seeing the original files, it's really not possible to know exactly how each system responds to the images.