Multi Exposure a fake only?

[pike105]

Hi,

now the latest Version 6.5.0r4 supports 4k-resolution with Nikon Coolscan 5000.
However, only with 2x ME, not with 4x ME. You will find significant errors in the examples attached.

But what is the benefit of ME?

My test are showing (as expected) a zero-effect!

Who has made positive tests and can bring up examples?

Thanks

http://yogi2007.yo.funpic.de/Silverfast ... 0Exposure/

[anthony]

I don't think Multi-Exposure is fake, but I do think that it is an over-hyped feature. My experience with using it with an Epson 4990 is that only very dark areas of slide film benefit from the feature, and even then it's really only useful for scans of underexposed slides that will need heavy correction to brighten them. I have yet to try the feature with 6.5.0r4, though, because the new button is missing (see my post under Epson flatbed scanners).

From your examples, though, one has to wonder how they are doing multi-exposure with your scanner. The artifacts look like hot-pixels which could occur if they were increasing the CCD sensitivity to get the increased exposure. The blue-channel looks especially worse with multi-exposure. Along with the increased softness the feature is only degrading the image.

For negatives, I would never expect this feature to benefit them. Negatives just aren't dense enough to really suffer from CCD noise. Even then, the subsequent reversal and expansion of the original negative image would effectively hide the noise.

Anthony

[pike105]

Hi Anthony,

thanks for your remarks.
Can you supply us with examples supporting your findings?

The "noise" in my examples is film grain. Both films (negative and slide) are Kodak 100 ISO. CCD noise from a Nikon LS5000 is totaly negligible compared to grain.

The LS5000 (like all other scanners in this class) will exaggerate grain, since it has quite a simple, non-diffuse illumination system built in.

In case you want to eliminate CCD-noise (if any) you don't need ME, a normal multi-scan will do.

But what the people from Silverfast show here (the grey roof):

http://www.silverfast.com/highlights/mu ... re/en.html

is absolutely spectacular.

Probably the added white noise to a scanner before running ME, or the original is actually on the right side and for the picture on the left they added noise using e.g. photoshop.

If ME would be that good, the Swedish King would invite them for the Nobel Price in Physics!

[BOBRAPP]

I have two scanners - the Minolta 5400 mark I and a Microtex 120tf. M-E is available for both but presently only works with the Minolta.

I can offer that M-E works brilliantly with the Minolta and, when scanning dense transparencies, the noise reduction in the denser portions is phenomenal!

When scanning transparencies, I select M-E 2 (4 has been removed on versions 6.5.r5).

Cheers,

Bob Rapp

[pike105]

Good to hear, Bob,
can you show your phenomenal examples?
Thanks

[BOBRAPP]

Here is a scan as an example. It is from a Kodachrome 65 slide I shot in 95 and is of an alley way in Sydney. Here is the full image




[/img]

Here is a 5400 dpi of a wall section from the left. Note the noise in the dark areas. I adjusted the levels to highlight the noise.







Finally, here is the same portion with M-E 2. The levels were adjusted the same and before.




[/img]

The depth of the blacks and noise should be quite evident. Enough so, that all my slide and high-contrast negatives will be scanned using M-E.

Cheers,

Bob

[pike105]

Hi Bob,

nice examples. Did you also check with "normal" multiple scan? If you get the same effect, your scanner is probably noisy.

Here is again one of may examples

From left to right:

No ME, 2x ME, 4x ME, Nikon LS5000, Kodak Negative Film ISO 200

http://yogi2007.yo.funpic.de/Silverfast ... e_Film.jpg

[BOBRAPP]

The sample I posted was from an under-exposed Kodachrome slide and the example explored the ability or the scanner and M-E to extract the maximum detail and reduce the signal to noise ratio - which is does in a remarkable way.

Back it the days of slides, we normally underexposed our images to prevent washing out the high-lights knowing that projecting the image would take care of the denser areas.

Now that I am scanning some of these older slides, the ability of Silverfast M-E is now playing an important part.

In the case of Negatives, the improvement would be in cleaner highlights that will be cleaner and, thought curves, produce much more detail than a single scan.

M-E vs M-S is a no-brainier. M-E produces much better results in far less time!

As far as the scanner is concerned, the Minolta 5400 mark I is a good as they get. It has a cold-cathode light source and the ability to insert a diffuser to reduce grain. This works very well with all media - especially Black and White and Kodachrome.

The Nikon scanners and the Minolta 5400 mark II use led light source which is far more harsher in the examples notes above.

Kind Regards,

Bob Rapp

[pike105]

No problem, here is a slide film Ektachrome 100 showing the zero-ME effect:

<img src="/img/forum/left_no_me_right_4xme_kopie.jpg">

left no ME, right 4xME

Sorry, I'm still afraid, its a fake!

[degrub]

Perhaps here is your answer from the recent version log:

# REMOVED:

* Multi-Exposure: 4x Multi-Exposure no longer existent. Could have alignment issues in certain cases without improving the image above what 2x Multi-Eposure already does.

Try at 2x and see if that makes a difference ?

Frank

[pike105]

No, also with 2x ME and the most recent version, no effect at all!