Noise Reduction Walk-Through

[gmitchel]

I just wrote a twenty page walk-through for a photograph from noisy original to finished retouch. I used the Imagenomic suite of tools: Noiseware, Portraiture, and Real Grain.

This is a coaching session. So it's like looking over my shoulder and getting commentary as I work on the photo. (You can even see where I had to go back and start again to avoid a slight color smear.)

Comments will be sincerely appreciated.

I want to thank Kelly Chamberlain (aka Redcrown) for the photo used in this example. It's an ISO 6400 shot with a Canon 5D MkII. Come watch as it ends up looking almost like it was shot at ISO 100. (For the uninitiated, that's almost six stops of noise reduction).

There both an HTML and PDF version. The .PDF will be faster and gentler on the server. There are about two dozen screen captures.

http://www.thelightsright.com/CoachingSession3
http://www.thelightsright.com/files/coachi...ingsession3.pdf

Enjoy!
 
Cheers,

Mitch

[dalethorn]

The PDF I downloaded does give a really good look at noise reduction issues.  My main question would be - how many stops of actual noise reduction are we getting here as opposed to end-result (effective) noise reduction?  By effective, I mean given certain imtermediate steps that aren't noise reduction, such as enhancing an edge on the forehead to prevent smear.  It might seem like a purely academic question, but I wonder about all of those non-contiguous steps.

[Schewe]

I said in the Camera Raw tutorial that for ACR 4.x (and still applies to ACR 5.x) that Camera Raw (not just any software) was designed for about a 1 stop improvement from inside Camera Raw. I'm also a fan of NoiseWare and use it on really noisy high ISO images where needed. But the way I use it is so locally adjusted and blended that there is no way to do that sort of noise reduction inside Camera Raw (or any raw processor I'm aware of).

That's not to say the ACR/LR noise reduction can't improve (I happen to know for a fact that improved IQ is a primary goal of Thomas, Zalman and Eric for future versions).

So, if you are going to quote me, quote me accurately...(I'm pretty sure I know what I said–Michael and I also touted NoiseWare are our industrial strength noise reduction of choice inside Photoshop).

[gmitchel]

I said in the Camera Raw tutorial that for ACR 4.x (and still applies to ACR 5.x) that Camera Raw (not just any software) was designed for about a 1 stop improvement from inside Camera Raw. I'm also a fan of NoiseWare and use it on really noisy high ISO images where needed. But the way I use it is so locally adjusted and blended that there is no way to do that sort of noise reduction inside Camera Raw (or any raw processor I'm aware of).

That's not to say the ACR/LR noise reduction can't improve (I happen to know for a fact that improved IQ is a primary goal of Thomas, Zalman and Eric for future versions).

So, if you are going to quote me, quote me accurately...(I'm pretty sure I know what I said–Michael and I also touted NoiseWare are our industrial strength noise reduction of choice inside Photoshop).

Actually, your words were not qualified, Jeff. You said, "The aim is to get a one-stop reduction in noise."

But let's not quibble. I've deleted that comment from the walk-through and the associated .PDF.

Mitch

[gmitchel]

The PDF I downloaded does give a really good look at noise reduction issues.  My main question would be - how many stops of actual noise reduction are we getting here as opposed to end-result (effective) noise reduction?  By effective, I mean given certain imtermediate steps that aren't noise reduction, such as enhancing an edge on the forehead to prevent smear.  It might seem like a purely academic question, but I wonder about all of those non-contiguous steps.

I'm not an engineer. I'll let an engineer from somewhere like Imagenomic answer that.

ISO 6400 is certainly not equal across different cameras. A few years ago, ISO 400 or ISO 800 images would be a severe trial for noise reduction without significantly degrading the image. WIth the latest generation of in-camera noise reduction, the noise at higher ISO is more manageable.

I shoot at ISO 100 religiously because I prefer to avoid serious NR. Even ISO 100 images often have visible noise artifacts. But this ISO 6400 photo from a Canon 5D MkII is certainly very impressive. For a wedding photographer or event photographer shooting in available lighting . . . Well, it's been a brave new world over the last year or two.

Cheers,

Mitch

[Schewe]

Actually, your words were not qualified, Jeff. You said, "The aim is to get a one-stop reduction in noise."

Yes...while inside Camera Raw 4.x and 5.x. Did I not also point out that I use NoiseWare for heavy duty noise reduction?

As far as doing noise reduction inside a raw processing pipeline, aside for the obvious difficulties of trying to preserve "real" micro-detail while getting rid of the noise–and that's really hard in any event let alone in a global process-the fact is that most people will want to be able to localize and blend the noise reduced results rather than apply globally. And ACR/LR's local corrections are not designed for noise reduction (well, yet).

As far as ISO ratings goes, there's been a lot of discussion about noise on LuLa lately...I suggest doing a search to read up. As far as "real ISO", Canon is finally up to 3200 with the 5D MII but Nikon's D3 is at 6400. Anything other than those are false ISO that amounts to under exposing at the next lowest real ISO and over-processing.

[bjanes]

As far as ISO ratings goes, there's been a lot of discussion about noise on LuLa lately...I suggest doing a search to read up. As far as "real ISO", Canon is finally up to 3200 with the 5D MII but Nikon's D3 is at 6400. Anything other than those are false ISO that amounts to under exposing at the next lowest real ISO and over-processing.

"Real ISO" depends on the definition of ISO. With a saturation based ISO (see here), exposure of an 18% reflectance target with normal metering will yield a saturation of 12.5%, leaving 0.5 EV of headroom. At the lowest ISO, saturation is determined by the sensor itself and one could say that this is the true ISO of the sensor. When one sets the camera to a higher ISO, the sensor does not collect more electrons, but the gain of the camera amplifier is increased and saturation (overflow) occurs in the analog to digital converter (ADC). This is analogous to push processing of film. With the "fake" high ISOs, maximum amplifier gain has been reached, and the raw file is tagged with instructions to increase the "exposure" in post processing.

If cameras had better electronics, it would not be necessary to increase ISO with a camera setting, but one could achieve the same effect in post processing. This approach is already used with some medium format digital cameras. With current dSRLs, it is often advantageous to increase the ISO with a camera setting up to a certain limit. Beyond that limit, one can increase the ISO in post processing and gain extra headroom for the highlights. For example, with the Canon 1D MII, it makes little sense to increase the ISO beyond 1600 (see the excellent analysis by Emil Martinec (the section entitled "S/N and Exposure Decisions"). One could question if the 1D MII really has an ISO of 3200.

The images obtained at ISO 1600 with the 1D MII might not appear to be exposed to the right and one theoretically would lose levels in the upper f/stops of the dynamic range of the image. However, as Emil explains, the real rationale of exposing to the right is to obtain a better signal to noise ratio and not make better use of the number of raw levels available in the raw data. At high ISOs the effective number of levels is determined by noise even with a 12 bit ADC. With a 14 bit ACD, the extra levels are mainly used to quantize shot noise.

Bill

[Panopeeper]

With the "fake" high ISOs, maximum amplifier gain has been reached, and the raw file is tagged with instructions to increase the "exposure" in post processing

I don't know of any DSLR acting this way (though this is the logical way), only of MFDBs. For example your D3 does not "tag" the ISO 12800 and 25600 images to increase the intensity in post processing; instead, the pixel values get increased before creating the raw file, thereby reducing the dynamic range.