Ad
Ad
Ad
Pages: [1] 2 »   Bottom of Page
Print
Author Topic: Film scanning histogram  (Read 9257 times)
Lol999
Newbie
*
Offline Offline

Posts: 7


« on: November 19, 2007, 08:18:44 AM »
ReplyReply

Hi, another brain picking question    I read somewhere, or could have even dreamed it, that when you do the preview scan the resultant histogram you should move the left and right pointers to 0 and 255 respectively ( or the other way round) to get as much information in the scan as possible. This will give a "flat" scan that can have it's curves adjusted in Photoshop. However, I'm sure the "standard" way to do it is to move the pointers until they just hit the bottom of the peaks at either end of the histogram and scan from there (if that makes sense?). Can anyone steer me in the right direction please?

Thanks, Lol
Logged
digitaldog
Sr. Member
****
Offline Offline

Posts: 9166



WWW
« Reply #1 on: November 19, 2007, 09:00:16 AM »
ReplyReply

If you move the Histogram such that you have data at either end (zero and 255), its not flat, its full tone. If you scan in 16-bit, you may want to scan a bit flat (that would be something along the lines of 5/8-245/250), then you can use the finer controls in Photoshop on full rez data to shape this tone and, have a bit of headroom should you increase say highlights from sharpening (although in good sharpening techniques, you want to keep from blowing out highlight data from the effects of sharpening just as you want to keep from sharpening noise in shadows).
Logged

Andrew Rodney
Author ďColor Management for PhotographersĒ
http://digitaldog.net/
Lol999
Newbie
*
Offline Offline

Posts: 7


« Reply #2 on: November 19, 2007, 10:26:48 AM »
ReplyReply

Thanks Andrew, got my terminology a bit mixed up  Makes sense now what you say.

Cheers, Lol
Logged
ashaughnessy
Newbie
*
Offline Offline

Posts: 21


« Reply #3 on: November 19, 2007, 11:13:22 AM »
ReplyReply

I don't understand how changing the histogram in the scanning software will change the amount of information you are capturing. Unless the histogram change affects the exposure setting of the hardware then isn't it just changing the data after the scan?
Why bother changing anything at all in the scanner, why not just do all changes in the photo editing program? (Assuming that you are carrying 16 bits over from the scanner software to the editing program and not losing any).
Anthony
Logged
Lol999
Newbie
*
Offline Offline

Posts: 7


« Reply #4 on: November 19, 2007, 11:45:41 AM »
ReplyReply

Well this is what I'm sure I read. the two arguments being:

1) Scan full tone so EVERYTHING is included that appears on the negative. More comprehensive tweaking is then required in PS.

2) Scan "flat" i.e. scan with sliders at initial "vertical points" for want of a better description and do less adjustments in PS. The argument being for this method is that you are putting the full scanning capabilities to work on the most important parts of the data from the histogram. Scanning full tone would include alot of "noise" etc that is unnecessary and not helpful!

Apologies for using the wrong terminology, new to this
Logged
digitaldog
Sr. Member
****
Offline Offline

Posts: 9166



WWW
« Reply #5 on: November 19, 2007, 11:49:54 AM »
ReplyReply

Quote
I don't understand how changing the histogram in the scanning software will change the amount of information you are capturing. Unless the histogram change affects the exposure setting of the hardware then isn't it just changing the data after the scan?

[a href=\"index.php?act=findpost&pid=154143\"][{POST_SNAPBACK}][/a]

Nope. Any decent scanning software is providing this data from the scanner settings, not post.
Logged

Andrew Rodney
Author ďColor Management for PhotographersĒ
http://digitaldog.net/
ashaughnessy
Newbie
*
Offline Offline

Posts: 21


« Reply #6 on: November 19, 2007, 03:29:23 PM »
ReplyReply

Unless the scanner software can actively modify the brightness of the scanning light source pixel by pixel then I don't understand how the histogram adjustment can extract any more information. And I don't see how the scanning software can control the exposure pixel by pixel when it needs to have already scanned the pixel in order to determine its brightness.
I just don't understand how this makes any difference. I can't see that it makes any difference to adjust the curve in the scanning software instead of leaving it till you've returned to your editing software. Perhaps there's some aspect of it that I'm missing.
Anthony
Logged
Eric Myrvaagnes
Sr. Member
****
Offline Offline

Posts: 7996



WWW
« Reply #7 on: November 19, 2007, 07:39:29 PM »
ReplyReply

Quote
Unless the scanner software can actively modify the brightness of the scanning light source pixel by pixel then I don't understand how the histogram adjustment can extract any more information. And I don't see how the scanning software can control the exposure pixel by pixel when it needs to have already scanned the pixel in order to determine its brightness.
I just don't understand how this makes any difference. I can't see that it makes any difference to adjust the curve in the scanning software instead of leaving it till you've returned to your editing software. Perhaps there's some aspect of it that I'm missing.
Anthony
[a href=\"index.php?act=findpost&pid=154204\"][{POST_SNAPBACK}][/a]
Every scanner that I've ever used makes (at least) two passes. The first pass uses the scanner's default settings (brightness, etc.) to produce the initial histogram and the preview image. When you make software adjustments to the histogram, those adjustments are then used to change the scanner settings for the second pass, which produces the final scan.

So yes, the software does indeed "actively modify the brightness of the scanning light source" in response to the adjustments you make.
Logged

-Eric Myrvaagnes

http://myrvaagnes.com  Visit my website. New images each season.
Kirk Gittings
Sr. Member
****
Offline Offline

Posts: 1548


WWW
« Reply #8 on: November 19, 2007, 09:00:42 PM »
ReplyReply

I agree in principal Eric. The point is to send a file with the fullest histogram to PS so that you have the most data to work with. So really large edits (color, clipping, strong curves etc.) are best done in the scanner software (at least ballpark close to what you want in the final image). SilverFast confirms this in their literature and in private communication.

You can test this easily with an 8 bit scan. Scan one flat and unclipped into PS. Then clip it and apply a strong curve. Do the same adjustments to one while in the scanning software. See how the histogram starts to break up when the same adjustments are done in PS. This is also true with 16 bits, but won't be obvious until you convert to 8 bit for say printing. Also look at the tone transition areas in the scans, tones that got enhanced or separated by the strong curve. The transition areas will show enhanced noise from the stretching of the tones in the one corrected in PS.

Scanning to raw formats like Imacon 3f only stop and save the file before the operators edits are applied to the raw file conversion.
« Last Edit: November 19, 2007, 10:53:08 PM by Kirk Gittings » Logged

Thanks,
Kirk

Kirk Gittings
Architecture and Landscape Photography
WWW.GITTINGSPHOTO.COM

LIGHT+SPACE+STRUCTURE (blog)
dmerger
Sr. Member
****
Offline Offline

Posts: 686


« Reply #9 on: November 19, 2007, 10:29:52 PM »
ReplyReply

Perhaps it depends on the scanner.  My Minolta 5400 doesn't work like Kirk, Eric and Andrew suggest.  I can adjust the exposure time, but there is no way to adjust the intensity of the lamp. I can move the histogram left or right, but I can't physically extend both the highlight and the shadow ends of the histogram, except post scan.  By post scan, I mean that all other exposure type adjustments are software only and just manipulate the data after the scan.  I prefer to make such adjustments in Photoshop instead of using my scanner software.

Perhaps other scanners work differently, but I agree with Anthony.  It seems impossible for any scanner physically to brighten one pixel and darken an adjacent pixel in order to expand the histogram.   Unless such a feat is possible, it seems to me that any such adjustment must be a manipulation of the data from the raw scan.

Maybe some scanners can manipulate the raw data like camera raw from a DSLR, but my understanding is that my Minolta canít do so and neither can the Nikon or Canon desktop scanners.  My post scan adjustments are pure software adjustments, i.e. similar to Photoshop.

Perhaps itís beside the point, but no digital camera Iím aware of can physically expand the histogram, either.  I mean, you canít expand the histogram (dynamic range) at the time you click the shutter.  You can only do so in raw conversion or post conversion with software like Photoshop.

Maybe I misunderstood what Kirk, Eric and Andrew are suggesting.  If so, Iím sure youíll set me straight.
Logged

Dean Erger
Eric Myrvaagnes
Sr. Member
****
Offline Offline

Posts: 7996



WWW
« Reply #10 on: November 19, 2007, 10:54:06 PM »
ReplyReply

I will certainly defer to Andrew and Kirk, who are the experts. But you raise an interesting question: How does a scanner change the contrast of a scan? I suppose it might multiply the values read at each pixel by a factor that varies with contrast, in which case it is somewhat analogous to in-camera processing of a digital image.

This is pure speculation on my part. Can Andrew or Kirk tell us how a scanner does its thing?
Logged

-Eric Myrvaagnes

http://myrvaagnes.com  Visit my website. New images each season.
Kirk Gittings
Sr. Member
****
Offline Offline

Posts: 1548


WWW
« Reply #11 on: November 19, 2007, 11:08:46 PM »
ReplyReply

Try the test. It is dramatic. The Sf one (adjustments in scanner) yields a full histogram while the PS one yields a pathetic histogram like a comb with broken teeth.

If I were not on my way out of town to visit my daughter in DC, I would post the samples. I'll try to get to it next week.
Logged

Thanks,
Kirk

Kirk Gittings
Architecture and Landscape Photography
WWW.GITTINGSPHOTO.COM

LIGHT+SPACE+STRUCTURE (blog)
TylerB
Sr. Member
****
Offline Offline

Posts: 361


WWW
« Reply #12 on: November 20, 2007, 12:14:33 AM »
ReplyReply

Quote
Try the test. It is dramatic. The Sf one (adjustments in scanner) yields a full histogram while the PS one yields a pathetic histogram like a comb with broken teeth.

If I were not on my way out of town to visit my daughter in DC, I would post the samples. I'll try to get to it next week.
[a href=\"index.php?act=findpost&pid=154300\"][{POST_SNAPBACK}][/a]

possibly because the adjustment is being made to 10, 12, or 14 bit data before saving out as 8 bit.
THe same adjustment in PS is being made on 8 bit information.
Very few scanners/software allow software analogue adjustment of the scan itself.
Tyler
Logged
dmerger
Sr. Member
****
Offline Offline

Posts: 686


« Reply #13 on: November 20, 2007, 12:50:15 AM »
ReplyReply

I did your test, Kirk.  The histograms are as you described them.  I couldnít see the effect in the image on screen, just on the histogram.  Of course, the image looked horrible after such extreme adjustments.

Iím not sure what conclusion to draw from this test.  Iím sure that the adjustments werenít adjusting the hardware.  The two scans took the same amount of time to scan.  As I mentioned before, my scanner canít vary the intensity of its light.  Therefore, the only way for my scanner software to make the adjustments was to manipulate the data from the scan.

Tylerís explanation makes sense to me.  My Minolta is a 16 bit scanner, and I usually scan at 16 bit.  So, I did the test again, but this time using 16 bit.  This time neither histogram had any gaps.  The histogram for the image with the adjustments made in Photoshop actually looked a little smoother than the scan using the scanner software to make the adjustments.
Logged

Dean Erger
TylerB
Sr. Member
****
Offline Offline

Posts: 361


WWW
« Reply #14 on: November 20, 2007, 01:20:20 AM »
ReplyReply

I did endless tests with Silverfast and a howtek 4500. Bringing the same stepwedge to the exact same tonalites various ways- all in Silverfast, all in Photoshop, and various combinations of both, I then determined how many remaining levels were in the file with Bruce Lindbloom's "levels" application. It's on his site, is quite old, and only works on a Mac in OS9.
The file remained hi bit in all cases.

The file with the most remaining levels was the one left completely unmanipulated in the scanner software, Including Silverfast's gamma setting, and the adjustments made in Photoshop.

Some software, Like DPL, can manipulate the log/linear amp, I believe in the A to D conversion (someone correct me if I have that wrong), and I believe some will adjust the lamp, maybe Nikon?
These seem like useful adjustments, but I've seen no tests to confirm.

I see value in doing some work in the scanner software regardless, as long as no needed info is tossed.
Some older software doesn't permit saving out in high bit, definitely do what you can in the scanner software in this case as the big moves will be made on the higher bit data off the scanner before saving out in 8 bit.

Tyler
Logged
Kirk Gittings
Sr. Member
****
Offline Offline

Posts: 1548


WWW
« Reply #15 on: November 20, 2007, 01:43:53 AM »
ReplyReply

Dmerger,

The 8 bit test scan gives you an inkling of how superior the workflow is doing the major adjustments in the scanning software regardless of whether you scan in 16 bit or 8 bit. Every hard edit you do destroys pixels even in 16 bit if you do adjustment layers and only flatten and convert to print. You can do the same test in 16 bit and do the conversion to 8 bit at the end and see the same issues. My printer, an Epson, converts to 8 bit to print and the issues will surface then in banding and noisy transitions. All printers convert to 8 bit except the new Canon. What are you making your files for? Almost all output is 8 bit. if you are seeing this in 8 bit, it is a warning about what will happen to your 16 bit files when they are converted.

Also, on my b&w images, I work those files allot. If I don't do the major adjustments in the scanning software my 16 bit files start to look like crap. The key is to bring to PS the fullest histogram you can, by letting the scanner software do the major adjustments where the most raw information is available. people at Silverfast shared this with me a few years ago and it is certainly true of all the sub "drum" scanners that I have tested.

Tyler,

I don't doubt your results. Which is why I worry about this allot less when i get drum scans. All of this is far less important with drum scans because of many things including minimal noise, but having said this, I'm not sure what level of scanning we are discussing here? For some reason I didn't think we were discussing drum scans, but rereading the posts makes me wonder. I am sharing my experience with the prosumer flatbeds, Epson 3200, 4870, 4990, 750, Microtek 1800f, canon 950f, Nikon 8000 and Imacons 800 and 900 series, scanners I have used extensively and  taught classes around. Everyone should test this on their equipment.
« Last Edit: November 20, 2007, 01:59:36 AM by Kirk Gittings » Logged

Thanks,
Kirk

Kirk Gittings
Architecture and Landscape Photography
WWW.GITTINGSPHOTO.COM

LIGHT+SPACE+STRUCTURE (blog)
ashaughnessy
Newbie
*
Offline Offline

Posts: 21


« Reply #16 on: November 20, 2007, 03:16:03 AM »
ReplyReply

Kirk, when you say there's more information available in the scanning software I don't understand this. What information is lost when the scan data is transferred to the photo editing software, assuming it is transferred as 16 bit data? I don't see that any data would be lost so how can the histogram adjustment be able to work with "more information"? If the information available is the same in both places then it wouldn't make any difference where you applied the histogram.

Can I just add I don't want to sound too aggressive here - I'm questioning as much as challenging. I'm very skeptical that histogram adjustments in scanning software make any difference but I'm not an expert in this field and I'm really willing to be wrong as it would allow me to get better scans but I really need to see a better explanation.
Anthony
« Last Edit: November 20, 2007, 03:29:34 AM by ashaughnessy » Logged
dmerger
Sr. Member
****
Offline Offline

Posts: 686


« Reply #17 on: November 20, 2007, 12:11:13 PM »
ReplyReply

Kirk, to paraphrase one of Americaís great philosophers, Sir Charles Barkley, you may be right, but I doubt it.

I donít believe that your test gives me any ďinkling of how superior the workflow is doing the major adjustments in the scanning software regardless of whether you scan in 16 bit or 8 bit.Ē  As Tyler explained, all your test shows me is the advantages of working in 16 bit versus 8 bit.  

I did the same test in 16 bit and then converted to 8 bit.  As before, neither histogram had any gaps, and the histogram for the image with the adjustments made in Photoshop actually looked a little smoother than the scan using the scanner software to make the adjustments.  Of course, I had to flatten my Photoshop version before converting to 8 bits.  Otherwise, my layered adjustments would have been done in 8 bit mode, with an ugly histogram as a result.

My results arenít surprising.  If I have two similar 16 bit images with similar histograms, it is simply not possible for there to be a significant difference between the two images after conversion to 8 bits.  To do otherwise, there would have to be some hidden information in the files.  No such hidden information exists.

Kirk, maybe you could do your 16 bit test again, and be sure to flatten your Photoshop version before conversion to 8 bits. Iíd be interested to hear about the results.  I canít understand how your results could possibly be different than my test results.  

Tyler provided a very rational, logical explanation of our test results, Kirk.  It seems to me that you have provided only a conclusion without an explanation.  Moreover, I canít duplicate your 16 bit test results. Nevertheless, I may be mistaken, and Iíd appreciate an explanation of your theory.  

Dean
Logged

Dean Erger
Diapositivo
Jr. Member
**
Offline Offline

Posts: 57


WWW
« Reply #18 on: November 22, 2007, 12:03:46 PM »
ReplyReply

Different scanner software behave differently.

After a preview you always have the "mountain graph" (the histogram graph). With certain scanner software (such as ScanView) you can have the basis of the mountain to coincide with the extremity of the graph field. So the darkest zone on the slide (let's imagine a black and white slide) will be rendered as 0 and the brightest will be rendered as 255.

That regardless of whether the darker zone is actually pure black and the brightest is actually pure white.

With some other software (NikonScan e.g.) you see the "mountain graph" but the mountain slopes do not necessarily reach the extremities of the field. This means that the darkest zone is not "0" and the brightest is not "255".

If you manually move the black point cursor rightbound, and the white point cursor leftbound, until they are at the beginning of the histogram, you are recreating a situation where the darkest point is 0 and the brightest point is 255.

If you do so in such a situation, you are obviously increasing the contrast of the scan.

If you do or don't do that is something which is left to your aesthetic appreciation, but you have a technical reason to do that anyway.

What you have read or dreamed is probably this: when the slopes of the histogram do not really arrive at the extremities of the graphs, if you adjust the white and black point cursors so as to put them just under the "mountain" extremities, you will as a result use the entire "numeric space width" of the scanner (be it 8-bit, 12-bit, 14-bit, 16-bit) on useful information from the slide.

If you don't do that, you are "wasting" part of the bit-depth of your scanner because the entire image info will be constraint into a "subset" of the numerical space that the scanner is able to represent.

To put it simply, let's say your scanner is 8-bit and your mountain is very narrow. Without any adjustment you would obtain from your slides let's say values from 15 to 225, so you will only have 210 shades of grey. If you adjust you obtain values from 0 to 255, higher contrast, and you will have 256 shades of grey. You will certainly have less problems in postprocessing (less posterization, less quantization error on resizing, better noise reduction) with 256 shades of grey than with 210.

As a final stage, you can always reduce again contrast in Photoshop if you want to obtain a less contrasted image.

So the logic is: use as much as you can the bit-depth of your scanner, by moving the black and white cursors (and increasing contrast) at scanning stage, so as to have a scan which suffers less during postprocessing, and if necessary reduce contrast again at the end of postprocessing.

If you like a comparison, that is a trick equivalent to the "expose to the right" with digital captures: you are adjusting your capture in such a way that you have more information to work with during postprocessing. You don't care about the appearance of the capture being "wrong" because you are doing postprocessing anyway and you will readjust things.

The entire problem is much less problematic with modern 14-bit or 16-bit scanners which do not really suffer that much of quantization errors (they do the same, but the errors are much less visible, if at all).

Cheers
Fabrizio
Logged
Diapositivo
Jr. Member
**
Offline Offline

Posts: 57


WWW
« Reply #19 on: November 22, 2007, 12:10:33 PM »
ReplyReply

Just to clarify a bit:

most of the times you will anyway adjust levels in your photoretouching program, at some stage of your workflow, so that the black point cursor and the white point cursor are just below the extremities of the "mountain".

Doing this prior to the scan process is better because you will have a scan with more tonal gradations, let's say 256 in the example of the previous post.

If you scan with 210 shades of grey and then adjust levels in Photoshop you will just "spread" those 210 levels into the 256 numeric space, with some quantization errors. So you'll better do that at the scanning stage and start with the 256 levels in the first place.

Again, if you have a high-bit scanner all those concerns are much less important than with a 8-bit scanner, nonetheless the same logic validly applies.

Cheers
Fabrizio
Logged
Pages: [1] 2 »   Top of Page
Print
Jump to:  

Ad
Ad
Ad