Scanning Your Own Negatives Demo/Guidelines
We're gonna take a look at a piece of software called SilverFast. So, SilverFast is just one of the two kind of big ones. Like I said, the other one is VueScan. They both do very similar things. They drive the scanner. They both have access to basically the same options, just kind of in different places. But they do create slightly different scans. VueScan, one of the nice things about it, is it works with every scanner. SilverFast is actually attached to the individual scanners. So this is what's called a pre-scan. So the last time I ran this piece of software, we ended up with these images. We'll be looking at those images in Photoshop in a little bit. These are some of the images we're actually going to process as negatives. But when I'm in here into the work flow, I've got a number of different options. I can choose reflective, transparency, or wide transparency. So reflective is kind of like I was doing a piece of paper and it's being reflected back up. Transparency is something I...
can see through. So in this case... And then wide transparency is a long transparency, so if you're doing a panorama negative or something long. So we're gonna choose a transparency in this case, because I can see through the negative. Then I have the option of positive, Kodachrome, or negative. Kodachrome was there because Kodachrome was such a magical, magical film that it requires everything special to be treated. It was the iconic... Our way of seeing was Kodachrome in photographs, in color photographs, for decades. National Geographic... I mean, Paul Simon wrote a song about Kodachrome. So, it's no longer available, but there are so many slides out there, and it just has a certain saturation and grain structure, so most of the scanning software treats Kodachrome a little different, but that's the only reason. When you launch something, you'll see it. Kodachrome was a very special film. If I do a positive, basically what that says is create a positive of whatever's there. So in this case, a positive of the negative is actually the negative. If I choose negative, it says, "Oh, take my negative, "and I actually want that to be the photograph." Okay, so it's a little bit backwards. You're like, "Oh, if I choose negative, "I get a positive on the screen. "If I choose positive I get negative." And that's because I'm scanning the negative and telling it it is the positive. If I was doing a slide film, that's what that's originally there for. A slide film would be a positive, like a Velvia or Provia. And the negative is anything else. Then here under this HDR button in this piece of software is how do I sample the data for the image? So I've got 48 bit, down-sample the 24 bit. 16 to eight, 16 to one, and then I have this HDR stuff. The HDR stuff is this company has a separate piece of software for processing HDR images. But what I want is the HDR information. I don't like 24 bit data in my photography. I like 16 bit data in my photography. The reason for that is eight bit gives me 256 available total transitions. 16 bit gives me 65,000. Now there are people that will argue that in the black and white world, the 256 is enough, and that's true, if good was enough. But in photography, the difference between a great print and an exceptionally amazing print is in small percentages. So I want that last 1%, I want that last 1/2% I wanna kick it as far as I can, I wanna suck as much data out of it as I can. So in that case, I don't want the step down to eight bit. So what that does is scan at 16 bit, and then it down-samples it to the eight bit. So I don't want that. What I want is the 16 bit. Now you see there's an HDRI. That's their version of a digital ice technology. That's the infrared. So for black and white, I gotta stay off anything with an I. That's bad. So I got a choice there of 48 bit HDR raw, or I got a choice of 16 bit HDR raw. 48 bit HDR says capture the red channel, the green channel, and the blue channel, just like a digital file would. Capture those three channels, assemble them together, and give me the negative. 16 bit raw file says capture a 16 bit gray scale image. So what it does is it's gonna either do all of the channels and blend them, select one of the channels, which is one of my preferences, take the red channel, the blue channel, the green channel, and scan that, or take a little bit of all three channels and give me that. If I choose the 84 bit, I just made up a whole new technology right here on CreativeLive. If I do 48 bit, I end up with a file that's three times the size of the 16 bit, because there's three channels of information. If I choose the 16 bit, I end up with less. Now, I've gone back and forth on this, and I've scanned both ways, and there are days I'm convinced I get more out of 48 bit, and there's days I figure it's just as easy to use the 16 bit. The benefit of the 48 bit, in some cases with some scanners, is one of the channels might actually produce a more sharp image. So if I can come in, and then after the fact and Photoshop, look at the red channel, the green channel, and the blue channel, and say that the blue channel gives me a little bit more sharpness, or the red channel gave me a little more contrast. They're gonna look almost identical, because we're capturing basically just these gray scale values across the image, and unlike a digital capture of the world through a digital camera, the film has already taken care of all of that. The film has already made the associations of zones and tones and exposure. We're just trying to pull that data off. So it's not as varied as it would be with a digital capture. If you're starting out, you're gonna scan on your own, I would experiment with both of those. If you're in VueScan, you're gonna have the option of creating a 16 bit channel or a 48 bit file. Scan it out of both. You're gonna want that option just to see. Mostly what we're gonna be looking at today are 16 bit scans. As I went through and looked, there wasn't enough there that I wanted to the hit on the file size. And like I say, sometimes, though, I look, and I definitely can see a difference, and so it's a time behind the scanner, so if I scan it and I'm like, "Could I get a little more?" I can just run the scan again. Scans are gonna take somewhere between two minutes and five to eight minutes, ten minutes at the most, for the kind of resolution that we're gonna get. So in that regard running it more than once is not gonna really damage anything other than the amount of time I have. So, in this case I'll just go ahead and choose a 16 bit HDR. And I wanna go back and scan a positive. So this is one of the things that I think is important in the scanning. And when I work with a lab, this is what I ask for. I want the original image. I want the original negative, and I wanna make the inversion decision. So when I make the inversion of the negative to a positive in Photoshop, I wanna control that, and I wanna show you why in a couple of minutes, but that control to me is really important. So what I want the lab to do is give me the absolute most shadow detail and the absolute most highlight detail possible. That's a linear scan. I don't want any correction, I don't want any adjustments. I literally want you to give me as much as you can out of the scan and let me determine what to do with it after the fact. So we went over to the lab, I'm just like, "I want a linearized positive of the image." So that's what I'm gonna scan here. The other piece, when it comes to scanning is how big should I scan? That is such a complicated question sometimes, and yet so easy. Like I said, it's some basic math. If you print at 200, 240, 300, depending on the size print, 30 by 40, 200, other than a photographer who thinks photographs should be looked at right here. Looks a little fuzzy. Standard viewing distance looks fine. So then photographers, you can print down to without any significant interpellation if you're getting bigger in size. So what you wanna do is think about what's the maximum potential size I would ever want to print? And then double that. And that's where to start. Everybody tells me, "I will never in my life, "ever print bigger than 16 by 20." Do you know how much it costs to print bigger than 16 by 20? And then they get this picture of the international space station going across the solar eclipse, and there's a mastodon flying by, and it's this thing that's never happened, they're like, "Aw, I gotta get that huge." I'm like, "Well, that's too bad. "Cause you said you were never "gonna go bigger than 16 by 20." So now we gotta go back into the scanner, we gotta re-scan everything, we gotta clean all the dust again. We gotta do everything again. So just take the size and double it. A little bit more in the hard drive. That's all it's gonna cost you. So in here, in this particular software, the preset says, "What are you printing?" I'm printing Photo Quality at 300 ppi. And then here's my scan resolution. This software has green for this scanner that tells me up to green, things are gonna be good. Get into yellow, scanner might be doing some magic math. Red, scanner's probably really doing some magic math. So we move that little slider up. Now this is the file size, so at 800, I can basically double the size. So I can get 2418. If I bring that up to, say, 2400, now my file size is 200 megabytes. Now remember, when I go into Photoshop, I open that image in Photoshop, that's a 200 megabyte file. Then I'm gonna double that layer, 'cause I gotta apply some sharpening. That's now a 400 megabyte file. I gotta double it again, because I need to add in a smoothing layer. It's now a 600 megabyte file. Okay, so, you're working in these large files, things are gonna be bigger. So that's the other thing to take into account, is if you're working with big negatives, bigger sizes, you're gonna have bigger things to deal with, and it exponentially gets bigger in Photoshop. In the case of what we're doing, I'm gonna drop this just back down to about 800. And it's a speed thing. The higher the resolution, the slower the scanner. It goes from (groans painfully) to (grunts rapidly). and then you're like, "What are you doing?" And it's (grunts rapidly), and you're like, (huffs). And then you leave, and you watch Game of Thrones, like a marathon, and you come back and it's like (grunts rapidly). So, high resolution runs slower. Some of the information else you have down here is navigator, so if you're zooming over here, there's a little bit of information about a densitometer. You can kinda see the percentages of blacks change as I mouse over different parts of the image. Little bit of information there. The one other piece that is important in this piece of software, is this thing called the multi-exposure. You have the same thing in VueScan. You have a multi-exposure there. This is the other thing to turn on that I think is really important, particularly with black and white images. What the multi-exposure does is it's gonna basically make an exposure for the entire image. So just floats by and makes the exposure. And then, it makes another exposure where it's magically and intensely over-exposes the scan in an attempt to suck out as much of the shadow detail as possible. Now remember, I spent a lot of time with my meter, getting that exposure right for the shadows. The last thing I want is my scan to truncate that. When I look at scans and when people call me, and they're like, "Oh, I've scanned my black and white stuff, "and my prints look terrible, and nothing's working," Bad scan job. If you got the exposure right on the film, and you have that latitude in that part of the film, it's a scanning problem. So whether the lab did it or you did it yourself, and usually it's what happens is they muddy up, so without sufficient detail, we try to bump the exposure up in Photoshop, and just this weird kind of muddy gray shows up, and that's 'cause there's no data there for it to work with. So that multi-exposure is an important part of the process. Everything else gets shut off, no matter what software I'm using. If I'm in VueScan, auto-sharpening gets turned off, auto color correction gets turned off. Everything gets turned off. If I wanted to make this a positive, so I would be scanning the negative, down here is the negative fix, there's an auto button, there's a CCR button. Those get turned off. Everything gets turned off. I do not want the software doing anything but giving me the scan data. Now there's a lot of people who go in, and they tweak with that, and they get it all set, and they're fine with the level of sharpening, but as you'll see in my work flow, sharpening is a critical part of how we judge and experience the photograph, and particularly with the black and white images, that grain is gonna make a wicked difference on some of the sharpening decisions we're gonna make, on how we accentuate that or minimize that, so I don't want the software here doing anything, because whatever the scanning software gives me, I'm done. The only way to go back and fix it would be to go back to the scanner. So I would go on ahead and make that decision to stop. Okay, so I kinda get my basic settings set up, and then up here I have the option of what I'm gonna save. TIFF, DNG, JP2. I always save as a TIFF file. It's a universal file format, opens up in anything. Everybody's happy. Okay, I'm gonna hit a pre-scan here, so what's gonna happen now is the scanner's gotta warm up the lamps, so the other thing that's really nice about these scanners is there's a lot of brain power behind them. It warms up the lamp so that it can make sure that whatever it's doing its pre-exposure with is similar to what I'm actually going to see when it does the actual image. So as you can see here now, these are the negatives that were on that roll of film that was brought out of the magic cooker this morning after it dried. These bounding boxes are the bounding boxes for the four by five So one of the things I can do with all the software, different locations, is I can say, "Oh, can you come up here "and find frames for me?" And they are six by seven. And now it's found a frame, found a frame, found a frame, found a frame, and this one, it's like, "Oh, I couldn't figure out what to do there," so I can move that frame and resize that frame. Now I have one, two, three, four, five. If I wanna create a sixth, I can come in and draw a bounty box on six. So now I have the six options there. Now this box is selected, because it's got the red boarder. This box over here... Wow, it's really hard to watch up there. I should watch over here. I'm like, that's just really weird. The white box here tells me that frame's not selected. If I select that, I'm gonna scan frame number four now. Here I was scanning number two, number one. If I come up here to the scan button now and click scan, I'm gonna get that one frame. But let's say like I wanted all six. If I click and hold that down, there's a batch scan option. So on the batch scan option, it's gonna scan all six images and then drop them down. In this case it'll just take the name Image One, and then there'll be an Image Two, an Image Three, an Image Four, an Image Five, and it'll just auto increment those up. And it automatically does it. So if you've got your images, just load 'em up, scan, and you'll get a little individualized scan off of there, and you're getting the linearized data off of this. So this is kind of the cool, cool piece. So then all I have to do is I wanna scan, I'll pick this one, I just click scan, The scanner warms back up, and what we're gonna see is it's then gonna give me a weird preview here as it flies across, and then it'll go on ahead and pick up that second exposure on the multi-exposure. And it goes slower. Now you can here how slow it is (grunts rapidly) versus the pre-scan, and it's having to work itself across now. So it's gonna scan the image. It's 40% done scanning the original image. You get a nice little status bar down here that's telling you, actually, what's happening. So, because the process is slow, sometimes you might freak out, and think, "Oh my, nothing's happening!" As long as this is making a noise and that bar is moving, something's happening, and it may just take a little bit of time. Now just scan the image, hit 100%, but that was only 50% of the scan. And it moves across. Now one of the other options you'll see, particularly in VueScan, is the samples. Right below is the multi-exposure. Right above that is a number of samples. That's how many times it should read each individual pixel to calculate the exposure. In black and white photography, I've never seen all my negatives where more than one has made a huge difference. With my color work, sometimes I do a four, five, to try to really make sure I get the right information right there, but I'm basically minimizing that. So this is basically now making the small black image. And you can see down here, it's starting to make a little bit of the image draw, as it's going through and trying to get those exposures done. So it basically does that, and then it does the merging at this point, so I don't end up with too files. That's the part I really wanted to demonstrate. I don't end up with part of the image, and then the black part on, and I gotta somehow auto-stack and merge and do all that. The scanning software figures all that out. The one other thing to watch for in scanning software is they will set up certain areas for highlight clipping and shadow clipping. They call it protection. What they mean is clipping. So they're like, "Oh, to make sure we get "the dynamic range of your image, "we're gonna go ahead and just lop off the ends, "and make sure everything fits." Everything turned off, so if you see anything in the highlight warning clipping, and it'll be like at .01, .2, 1%, take them all to zero. You want everything to come back in as much as possible. Once it's done with the scan, then it's gonna drop the file in, and we're gonna be ready to go into Photoshop. So we're at 95%. We'll just let this finish. I've already scanned a bunch of other images, so we're gonna jump into Photoshop here, and talk about, actually, how we do the process. So it says I'm finished now, and if we jump out and take a quick look in the pictures folder, there's that negative scanned. Looks terrible. Good! We love terrible-looking coming out of the scanner. Really, just to reemphasize, the biggest thing with scanning is getting the focus set right for whatever is your holder. I spent six months pulling my hair out because nothing was sharp. And I'm like in Photoshop trying things, I called the software people, and finally one of the engineers was like, "Well, try a different holder," so I ordered a different holder, and I could have more fine-eyed adjustment. And instantly also I got great scans, and then moving to the wet mount scanning, huge, huge difference. That is selling out the dust, big thing. So the settings in here are pretty easy, because you're linearizing the scan, you're not having to worry about all the color correction and sharpening in here, you're just wanting the raw data. That's really, I think, one of the big keys. And for my color negatives, by the way, I do the same thing. I want just as much raw data as possible, because I'm not a commercial photographer moving high in the volume. I'll come in, look at the negative, cherry pick it, scan a few, make the conversion. If I was moving at a higher volume, I would definitely just have a scanning operator do it, and then just have them give me as much of a linearized response back for the final correction, or do what Slater did and have a color pack made, and just know that everything's gonna show up fine.