Less noise with HDR and how to get more density

[newindustar]

1.First I was happy to find I am getting much less scan noise in my HDR scans than 48 to 24 scans when both were openes in PhotoShop. Why is this? If not the scanner itself, is the noise generated by the software?

2.I am scanning 66 negs as negs in HDR and Negafix is running Auto and CCR checked and I am getting good color balanced scans. After scanning HDR I am inverting in PS CS3. My issue is the density is too low, too light after inversion requiring big shits with levels. First I move sliders to the endpoints in each chennel seperately then I go back to rgb and drag the middle slider way over to get the density I saw in the preview window. The only way I found to get lighter at scan time = darker after inversion is to lie about the gamma in options, at the moment I have it on three to get a lighter preview, thus darker inversion. It seems to be the same thing as the brightness slider in the negafix window, is it? What is the best way to get more density after inversion?

3.I found Epson Scan to introduce more artifacts and micro windopane pixelation where SE 6.6 is not. Can you verify if this has been observed by others?

Thanks

[LSI_Heidorn]

Dear newindustar,
there is much more to Ngafix then just inversing and finding the RGB edges so you will have a hard time getting a similar result in PS !

If you experience more noise in the final scan with SilverFast, maybe this is because you had enabled the autosharpen filter which sharpens details but of course also will slightly amplify noise.
OR the image has a strong grain and that is what you are seeing.

best regards,

Nils Heidorn

[newindustar]

Thanks for the reply but I think my points were missed. Let me rephrase and simplify.

1. I had no concerns or questions nega fix in my question. I was comparing HDR vs 48 to 24 bit scans, also else unchanged. Why is HDR less noise than 48 to 24 bit scans?

2.Question two was about how to get more density with HDR, the only way I found was increasing the gamma option in options.

3. I found Epson Scan to introduce more artifacts and micro windopane (possibly stepping motor)pixelation where SE 6.6 is not. Can you verify if this has been observed by others? This is a plus for you software. Could you discuss it?

Thanks

[LSI_Heidorn]

Thanks for the reply but I think my points were missed. Let me rephrase and simplify.

1. I had no concerns or questions nega fix in my question. I was comparing HDR vs 48 to 24 bit scans, also else unchanged. Why is HDR less noise than 48 to 24 bit scans?

2.Question two was about how to get more density with HDR, the only way I found was increasing the gamma option in options.

3. I found Epson Scan to introduce more artifacts and micro windopane (possibly stepping motor)pixelation where SE 6.6 is not. Can you verify if this has been observed by others? This is a plus for you software. Could you discuss it?

Thanks

Dear newindustar,

-1- if you really mean HDR and not 48 Color mode, then the answer is simple:
The transform from negative to positive (amongst others) also uses very steep Gradation curves which simply amplify the noise/grain to seem much stronger.

-2- Maybe you should tell me what exactly you mean by density. The way you use it here makes me think that you mean the distribution of brightnesses is not to your liking. I would guess that you want to suppress the medium tones more to the dark tones, am i right ? Please explain...

-3- This is good to hear, but we do not reverse engineer the original Sotware and then change it, but we simply do OUR approach and try to be best.
So while this is a positive finding, i certainly do not know why EPSON Scan is failing here. We often had positive comments about the base level Quality of SilverFast.
I can only guess that we do a better choice of native Resolutions and a better way of interpolating in-between them...

best regards,

Nils Heidorn

[newindustar]

Thanks very much-

Re noise I was referencing HDR specifically.

When ones scans negatives in HDR ones needs to invert in PS then adjust levels. To get a darker (more density) after inversion one needs to get the preview lighter before the scan. The only way I found to do this is by upping the gamma settion in HDR options because all the normal adjustment option are not available. The problem with my approach one has to guess on the gamma setting by trial and error to get a dark enough image after inversion in PS.

I was asking if this approach is in error or could be improved somehow, (with negatives).

Thanks

[LSI_Heidorn]

Dear newindustar,
hmm i guess i did not miss your first point by far then

So i can just try to rephrase:
If you scan an image in HDR, no correction whatsoever is done to the image.
If you want to create a similar image in Photoshop you will have to simulate Negafix which is not an easy thing to do.
Depending on the film a Gamma Curve can bring you nearer but as i wrote before there is much more to NegaFix and you will not be able to simulate this in Photoshop.

Due to the nature of Negafix the grain will stand out a bit more in a "negafixed" image, but this simply reflects reality, it is not a flaw of Negafix.

However you can try to create a satisfying image in Photoshop with using Curves and Color Filters but it will be a major pain and the result will never match the original scene as much as it does if you use Negafix.
But if you just want an image that looks "okay" and not neccessarily very close to the original you can go that way and possibly create a Photoshop action for that later.

best regards,

Nils Heidorn

[newindustar]

I think I failed in the communication now, it is difficult to make the right point.

What I failed to say perhaps is that I was using negafix at the same time I was doing my HDR tests. There was the negafix brightness control but it had no effect on the brightness of the scan. This is why I went inot the gamma change in options. I found 3.0 enough on the frame I was working with to lighten it enough in preview, thus dark enough after inversion in PS to begin post re-leveling.

I can say even with HDR and gamma value changes and post re-leveling negafix helped with balance of colors such as neutrals but there is still much guessing. It is too bad that the brightness slider in negafix does not reflect in the preview what the image will look like after inversion.

If you can tell from my post my ideas about the whole process may be wrong so I am open to correction.

[LSI_Heidorn]

Dear newindustar,
yes, i guess we are still not understanding each other 100% ( maybe 66% ).



Did you watch our Negafix movie on the website ?
Please let me know if that clears up the problem.

If not, i will try harder, okay ?
( I am still confused that you choose HDR mode AND Negafix, combining these will result in the "brownish" original image, again: did you try to use 48 Bit Color instead of HDR ?? )

Greetings, and have a nice weekend.

Nils Heidorn.

[newindustar]

Ok - good that makes sense, I did not intentionally use it, it just pops up when you choose neg and does not go away when you choose HDR so it is not intutive. How do you make sure it is off?

This does not affect the question of using the 1-3 gamma option when for HDR output is selected. Negafix should not have been mixed in the discussion here.

We need to get back to my intial question if using increasing the gamma, 1-3 in options to increase inverted darkness is the right way to proceed if when using HDR the results (final after inversion) are too light, it does work for this but is it the best way

[LSI_Heidorn]

Dear newindustar,
i will try to write about the different purposes of HDR / 48 Bit Color / 24 Bit Color:

-In HDR Mode the Negafix Dialog is not used for processing ( even though it is shown ).
This mode is meant to produce archive-level quality --> The image data should not be changed at all, no matter what you do in Negafix or other filter
settings.
It seems that you checked the Option "for HDR Output" next to the Gamma value in the Options, so that is why you see a change at all when using a different Gamma value.
So: Only use HDR if you want to produce untouched archive data or if you want to do the negative to positive Conversion *completely* on your own.

-In 48 or 24 Bit mode, the Negafix settings (and other filters) will be processed on the scan data, so when you scan an image it should look very close to the color you see on the preview window.
This is the mode you should use if you want good looking images at an instant.
You should choose 48 Bit if you may want to do further processing / retouching ( to have more dynamic range as a reserve ) or 24 Bit if your work on the image is basically done and you want to have small storage consumtion and good compatibility with the most image-viewer software.

Many Users choose to actually do 2 scans, they do one HDR scan and archive it, then they do a 24 Bit scan to a JPG for instant viewing purposes.
When they choose later to try a different negative to positive match they can use our "SilverFast HDR" Software to load the HDR image and to do the negative to positive matching as if they still had the original film in their scanner.
This is an add on Software we sell.

If I understand you correctly the following is what you are doing right now ( I do not know if by mistake or on purpose to further optimize quality):

-1- you scan the image in HDR, using the "for HDR Output" switch on with different gamma values.
-2- you open that image in PS and invert it.
-3- you try to optimize colors.

Like I explained above, you can do that if you want to create your own custom negative to positive workflow but it will be hard work...
AND it would not make a difference if you choose to apply the gamma curve in SilverFast ( 1-3 ) or if you load the linear HDR in PS ( linear = Gamma 1.0 ) and apply a Gamma Curve in PS before inverting.

So, to summarize:
-If you want to scan good looking images, use 48 or 24 Bit Color with Negafix.
-if you want to do neg to pos conversion yourself OR want to archive the images use HDR.

best regards,

Nils Heidorn

[newindustar]

We are totally clear on all the above, great!

I am using HDR with negs for three reasons:

1. The HDR finals after I vert in PS have much less noise than do 48-24.

2. 48-24 with negafix is taking the histogram right to the endpoints before I get a chance to bring the endpoints in to my taste in Levels in PS. Even if you reset the auto setting it is still contrasted up beyond my control i.e. the histo is already out to the the very ends.. In EpsconScan if you reset it the histo values have a lot of margin on both sides to work with. I usually do all my endpoint on each channel seperately and by eye.

3. I prefer a 48 bit to work with in PS. I don't see anyway to get the 48 bit all the way into PS without using HDR. I am getting histo breakup after I adjust levels in PS so want all the 48 bits.

These are the reasons I am using HDR and why I have to use gamma in HDR output options to get enough density(darkness) after I invert in PS.

What I want regarding negatives and in some points positives:



1. If I could have a truly flat unleveled histo in 48-24 that would be good.

2. If negafix could just fix the mask color and not adjust the levels it might help with the noise.

3. If I could turn off Negafix with 48-24 and do my own levels it would be good.

4.If I could get 48 bit all the way into PS using negafix without using HDR it would be good.

Make sense? If you could tell me how to get some of what I want it would be good

Any suggestions would be helpful. The HDR scan are much cleaner even after inversion in PS so it is easier to post process them. If I have a way to maintain the cleanliness and 48 bit all the way into PS without using HDR and Gamma I would be happy because as you say it is much easier to get good color without HDR.

Thanks

[GetchHard]

with the information I got in this discussion my problem will be finish at last.,


_________________
Humidifier Filters

[LSI_Heidorn]

We are totally clear on all the above, great!
1. If I could have a truly flat unleveled histo in 48-24 that would be good.

2. If negafix could just fix the mask color and not adjust the levels it might help with the noise.

3. If I could turn off Negafix with 48-24 and do my own levels it would be good.

4.If I could get 48 bit all the way into PS using negafix without using HDR it would be good.

Dear newindustar,
1.
You always get the Histogramm of your current image processed by the current settings.
If you reset all, you get a flat unlevelled Histo EXCEPT the Gamma Curve.
If you don't want that, adjust Gamma to 1.0 in the options but do not be surprised to have a dark & dull image in the preview as well !

2. To fix the mask color we need to expand the levels in RGB seperately, so there is no fixing the mask color without touching the levels.
Maybe you do not want us to apply the negative curves within those levels. You can switch this off by choosing "linear" from the negafix film type (last dropdown menu, first entry).

3. You can do that by just switching to positive mode and setting the Gamma value to 1.0, this will get you untouched scans (except the 48 Bit to 24 Bit scaling). Do not be surprised by the dark & dull images that also lack dynamic range because of only 8 Bit per channel, as we discussed many posts above i do not recommend any of this

4. On the contrary you will not be able to use Negafix with HDR as HDR switches OFF Negafix. So the only way to combine 48 Bit with Negafix is to use the "48 Bit Color" mode.


best regards,

Nils Heidorn