Jump to content

Blotchy blacks on scanned image


baci

Recommended Posts

Advertisement (gone after registration)

Well, I returned to shooting film at the start of the year and my progress up the experience curve has been fun and interesting. I'm shooting well, doing a good job on development and thought I was scanning well but have noticed more often than not where I have an expanse of darkness the blacks are quite splotchy. I figure this might be a by-product of me 'stretching' the histogram in Lightroom by moving the black and white points to extend the dynamic range. I need to do this because the histogram I get out of the Plustek 8100 scanner never extends close to either end of the x axis.

 

I'm wondering if I need to do something different somewhere, or if this is a limitation of the system? I'm using Vuescan and have all the settings where they should be for maximum quality.

 

Anyone else have the same experience?

 

Thanks!

 

Baci

Link to post
Share on other sites

 I'm using Vuescan and have all the settings where they should be for maximum quality.

 

Anyone else have the same experience?

 

Thanks!

 

Baci

 

Have you set your White and Black clipping points in Vuescan to as near to 0% as they will go?

 

The best scan to make is one that simply has all the information possible, so to increase the DR of the scan you need to minimise the Clipping points levels. This will result in a very low contrast scan, but that's what you need, the rest can then be easily adjusted in Lightroom where you can keep an eye on the histogram as you adjust contrast and brightness using Curves or Levels (does Lightroom have Curves and Levels, I don't know I only use Photoshop?)

 

Steve

Edited by 250swb
Link to post
Share on other sites

Thanks for the replies - yes I have black point and white point set to zero in Vuescan. I do two passes when scanning and, as the image is built up as the scans progress, it is usually contrasty and appealing. Then when the scan is finalised the image reverts to bring quite mid-tone heavy and flat on the screen. This is the image I export out of the scanner and histo curve has start and endpoints a long way from either end of the scale. Looks unappealing, but then I pull it into Lightroom and set the black and white clipping points, adjust the other levels until there's the appropriate tonal range, and save. The image looks fine from a distance but often the blacks really are lumpy. 

 

I'm using T-max 100, so black and white. I was wondering if the act of 'stretching' the histo in Lightroom causes interpolation errors where Lightroom is trying to invent data that doesn't exist in the image. And if this is so wouldn't it be better to get the fullest range of values from the scanner (ie not flat and mid range heavy) just to reduce the interpolation errors?

 

Again, I appreciate your input - there aren't that many people I can talk to about this stuff!

Link to post
Share on other sites

 I was wondering if the act of 'stretching' the histo in Lightroom causes interpolation errors where Lightroom is trying to invent data that doesn't exist in the image. And if this is so wouldn't it be better to get the fullest range of values from the scanner (ie not flat and mid range heavy) just to reduce the interpolation errors?

 

 

Lightroom isn't interpolating anything, as in inventing none existent tones or blacks. It isn't stretching anything. It isn't introducing errors. A low contrast scan sidesteps the relative crudity of the scanning software and allows you to make the tonal adjustments with a sophisticated software package. If it isn't you introducing lumpy blacks in your post processing technique then it may be you introducing lumpy blacks through your film exposure and processing. If there isn't enough detail in the shadow areas of the negative there will be nothing for the scanner to do other than create black instead of shadow detail.

 

Steve

  • Like 1
Link to post
Share on other sites

I can't be sure without seeing an actual example - but I suspect what you are seeing is "film-base-plus-fog" noise, exaggerated by your addition of contrast when you stretch the histogram, and the tendency of scanners to pull out every little bit of density variation in low-density (shadow, thin) areas of negatives.

 

Fog (more accurately, chemical fog - to distinguish it from accidentally opening the camera back and "fogging" the film) is a natural effect of developing - some tiny part of the silver halide in even the deepest (little exposed) shadows gets "exposed" by reaction with the developer chemicals, rather than by light, and then develops to silver grains.

 

In traditional darkroom printing, the diffraction of light passing through and around the grains in the negative, and the diffusion of the light in the paper gelatin, tended to "fill in" these microscopic speckles and give smooth blacks. It usually was no problem at all, and if it was, in a specific application, one could add anti-fogging agents to the developer

 

But scanners are hypercritical about revealing every little density variation unforgivingly. And if you scan for an original file that is "flat" (black and white come out dark gray and light gray) and then expand that histogram to reach pure black and white, you emphasize the contrast between the speckles and the purer black between them.

 

BTW - I'm not disagreeing with Steve and Chris. But too flat a scan can be too much of a good thing.

  • Like 4
Link to post
Share on other sites

Just to be sure, ICE is off, right?

 

An example would help, too. And seeing your Vuescan settings.

 

A very good point. ICE will treat the grain in a regular B&W negative as if it was dust and should only be used with colour or chromogenic B&W films. I imagine this would have a similar effect to aggressive digital noise reduction and could cause shadows to turn blotchy (along with a general image degradation) but I 've never done the experiment to find out.

 

 

Steve

  • Like 1
Link to post
Share on other sites

ICE definitely switched off - I'm using the scanner for predominantly B&W film so selected the 8100 scanner which doesn't have an IR channel.

 

I can only post small file sizes to the board so I've uploaded an image to Dropbox - you can get it here: 

 

https://www.dropbox.com/s/k882m64bo6z0pup/Chas.jpg?dl=0

 

Have a look at the roughness of the detail in the dark t-shirt compared with surrounding areas. If I can get smooth brighter tones why not smooth darker tones? 

 

I'll post the Vuescan settings separately. 

  • Like 1
Link to post
Share on other sites

Hi, Baci - thanks for the sample, and that is about what I thought you meant. I stand by my comments above.

 

As to why you get this in blacks and not in lighter tones - in the blacks there are random isolated silver grains (the base "fog" noise), that stand out like stars in a black sky. In the lighter tones, there are many overlapping silver grains (your film emulsion coating is about 50 grains "deep") that blend together so that the individual speckles don't stand out as much.

 

Remember that when you shoot a negative and "reverse" the tones to a positive, whether by scanning or traditional printing, the black(ish) silver grains in the negative become light(er) specks in the positive image. See attached, where I "inverted" a bit of your image to approximate the original negative.

 

(BTW - if you want to show small details to demonstrate this kind of effect in large scale, just crop a version of your image to an area showing what you want that meets the forum size limit, and attach that crop.)

 

I don't think there is much you can do with scanner settings, as such, to avoid this, outside of clipping the blacks to the point that the stand-out grains are also clipped to black. Which, as we've all said, causes other tonal problems. It is just the way high-res scanners image the grains, which is different from old-school photo printing.

 

What can be done with the image after scanning, or before scanning (camera exposure and development) I have some ideas about. But they are likely to be controversial, so I'll let Steve and others give their thoughts first.

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

  • Like 5
Link to post
Share on other sites

For completeness, here are the scanner settings if anyone would like to comment!

 

 

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

Link to post
Share on other sites

 

What can be done with the image after scanning, or before scanning (camera exposure and development) I have some ideas about. But they are likely to be controversial, so I'll let Steve and others give their thoughts first.

 

Just pile in there Andy, I've suggested earlier the situation may be improved with exposure and development so would love to hear your ideas. A better scanned tonal gradation is why for medium format and large format I've gone entirely over to compensating developers, but a general Zone System approach could be the answer as well, semi stand or stand development etc. all are possibilities to get some more detail/density in the shadows without blowing the highlights. I develop all my negatives now with scanning in mind and always question whether a development regime is darkroom or scanner friendly.

 

Steve

  • Like 1
Link to post
Share on other sites

For completeness, here are the scanner settings if anyone would like to comment!

 

Nothing wrong I can see, just one tip. For B&W some scanners give marginally sharper results if instead of taking the image from RGB you choose just one of the colour channels. So with my Plustek 120 I take the B&W image from the green channel not the default 'Auto' as per Vuescan. It is only marginal, but a test of each channel against the Auto result may reveal something advantageous.

 

Steve

  • Like 1
Link to post
Share on other sites

It's easy to check which is the sharpest channel. Tim Gray (who sometimes posts in this forum too) described it in this photo.net article. In my testing I haven't found that it makes much of a difference but I think this is very scanner dependent.

 

Question: should curve high be at 0.021?

 

And, if you're scanning T-Max perhaps use that film setting instead of Ilford XP2 (which is a C41 film, I believe)?

 

I sometimes get such blotchy darker zones in my scans and I've always thought it was, mainly, because of my developing skills (I use HC-110). Another reason for the graininess one sees in scans is due to how scanners "see" grain. An optically enlarged photo would not show the same type of grain. 

  • Like 1
Link to post
Share on other sites

Interesting thoughts, Philipus - thanks!

 

This raises a significant gap in my knowledge and that is I thought the curve settings affected the preview image but not the final scan. When the scanning is taking place and the image builds up on the screen it is contrasty and lovely as per the settings in the 'Color' tab. When the scanning has finished I get the round 'busy' symbol before the image snaps back to the flat, low contrast image that I then import into Lightroom and edit. I was of the (possible misinformed) opinion that none of the Vuescan settings affect the actual data in the file that gets exported. Basically the scanner records what the scanner sees irrespective of these settings.

 

I've been thinking about the role of the developer in all this too. This was the first roll I'd developed via stand processing and I was actually really pleased with the result - good density, great contrast. But if it's giving me the lumpy blacks then I may have to reconsider!

Link to post
Share on other sites

. I was of the (possible misinformed) opinion that none of the Vuescan settings affect the actual data in the file that gets exported. Basically the scanner records what the scanner sees irrespective of these settings.

 

 

Yes, misinformed, all those settings are there to affect the scan, not to perfect the preview image. Stop being obsessed with the preview image, other than cropping it and perhaps adjusting the brightness you don't need to 'see' anything, the settings for the best scan should be based on knowledge and attention to the histogram alone.

 

Steve

  • Like 1
Link to post
Share on other sites

Afaik the preview image has a gamma curve applied to it which makes it look different from what the actual scan.

 

One can scan in different ways but generally, and as Steve said at the beginning, it is best to leave all processing to another piece of software like Lightroom or Photoshop and only use the scanning software to pull out as much detail as possible from the film. This will result in flat, dull and boring looking scans. In other words, if that is what you get then you're doing the right thing.

 

I just checked my settings and I have curve low at 0,25 and curve high at 0,75. The result in the histogram is a straight diagonal line. This is good, I believe.

 

There's a lot on the internet about stand developing, particularly using Rodinal, I believe. I've seen nice results online but in my own tests (though with HC-110) have never managed to match those results. For that type of simplicity (ie more or less time-irrelevant developing), I'd use Diafine.

  • Like 1
Link to post
Share on other sites

This has been very thought provoking! One point that tripped me up was the fact that there are three distinct histograms relevant to any given image – the first is the raw histo showing the data straight from the scanning head arranged into a TIFF format. This histo is ‘pure’ showing no changes to the image as a result of the settings under the ‘color’ tab. The second histo shows where the black and white clipping levels sit in the image as currently selected, and the third shows the histo for the final image. There is a fourth curve used to show the inflection points of the Curve Low/Curve High settings – not relevant to this discussion for now! 

 

Here’s where I’m at: Under the Output tab in Vuescan I can save a raw TIFF file – it gives me a negative image but one straight from the scanner and without any of the changes to the image. As Lightroom has more refined control over adjustments to the image, best make all the modifications there instead of in Vuescan. To do that the negative has to be reversed. That’s a straightforward process and having done so I import into Lightroom what must be the most accurate representation of what was on the negative.

 

Blacks appear to be significantly better – I will do a scan of the original image to show the improvement when time permits. The images also appear to have more adjustment latitude (but that might just be confirmation bias!)

 

Now the big question I can’t quite answer is why the image histogram after the file has been imported directly into Lightroom looks different to the one depicting the data in the raw file straight from the scanner…

 

A learning process indeed – thanks for all your contributions to date.

Link to post
Share on other sites

I've never managed with my scanners (Coolscans) and regardless of whether I use Vuescan or NikonScan to get a graph that covers the full histogram. There are lots of posts about this over at RFF and elsewhere, btw. But I don't think it matters as long as the scan is flat and contains as much info as possible.

 

You say the image is negative - are you scanning your b&w as colour negatives? It says B/W Negative in the screenshot above. My b&w scans, with very similar settings (except the curve settings), are positive.

 

I'm leaning towards believing that the blotchy blacks are due to the stand developing.

 

 

Under the Output tab in Vuescan I can save a raw TIFF file – it gives me a negative image but one straight from the scanner and without any of the changes to the image. 

 

---

 

Now the big question I can’t quite answer is why the image histogram after the file has been imported directly into Lightroom looks different to the one depicting the data in the raw file straight from the scanner…

 

Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...