How do you tweak your scans?

[madsolitaire]

Photoshop is a very time-consuming process, and time is always a scarce commodity. I can never have enough time and even if i do, i would very much prefer to be out there somewhere taking pictures. I guess i just don't enjoy digital manipulation that much, and sometimes, i like the pictures 'raw' (in every sense of the word).

cheers,

 

[Keith]

For my, working a scanned image can take anything from around 20~40 minutes for 35mm up to 1.5 ~ 3hrs for a 6x7. My first comment is that I only work on a flat image and never use layers.

Initial work is global - black and white points, some global contrast adjustments through curves usually in conjunction with luminosity masks. After the global adjustments I start on the local areas. Pick an area and make curves adjustments for the particular area (with no consideration for the rest of the image), mark this as a history state, go back to the previous state and paint the areas in to effect. I keep repeating this across different local areas until satisfied. I sometimes work further with more luminosity masks and gaussian blur before sharpening but the reasons for when and why I do this are not part of typical adjustments so I'll leave it here. This is all 99% on B&W; if and when I ever use colour, I convert to L*a*b, switch of the "a" and "b" channels then just work it as a B&W before converting back to an RGB space.

This image is nothing special but to get the overall tonality to where it is took about 3 hours - admittedly its a 185MB single channel file from a 6x7 scan, which slows things down. Off the scan it took a lot of time to bring the sky and distant hills into detail relative to the foreground, even before the foregrounds own processing. This was not something a more "correct exposure" could have fixed.

From my perspective, direct scanned film, without post-processing is a long way from a presentable image.

That is a nice shot ... I'm curious though what drives you to spend so much time on an image. Is it a need for perfection personally or are you preparing an image for exhibiting or selling?

 

[craygc]

That is a nice image ... I'm curious though what drives you to spend so much time on an image. Is it a need for perfection personally or are you preparing an image for exhibiting or selling?

Hi Keith,

I guess the amount of time does seem a little inordinate. I certainly don't sell or typically exhibit; and perfection is a slightly loaded word. For me these days, the process is about producing a high-quality image. I certainly enjoy the shooting experience itself, but for what I tend to shoot it involves regional travel, and usually planned travel at that. At home, I don't shoot anything and have time to process images (although never enough and getting less).

For me its all about production of a product (even if its only for myself or web posting) and this production doesn't stop at capture nor does it happen in 10 minutes of post-processing tweaking (especially on very large files 😀). Everyone has their reasons why they photography with differing expectations about what they want from it. I personally get satisfaction in being able to produce a final image that not most can achieve (...and that isn't intended to be a conceited comment). Furthermore, if anyone every feels I have something to offer in my approach to post-production I'm always willing to discuss and share.

I just enjoy doing it 😀

 

[Keith]

Hi Keith,

I guess the amount of time does seem a little inordinate. I certainly don't sell or typically exhibit; and perfection is a slightly loaded word. For me these days, the process is about producing a high-quality image. I certainly enjoy the shooting experience itself, but for what I tend to shoot it involves regional travel, and usually planned travel at that. At home, I don't shoot anything and have time to process images (although never enough and getting less).

For me its all about production of a product (even if its only for myself or web posting) and this production doesn't stop at capture nor does it happen in 10 minutes of post-processing tweaking (especially on very large files 😀). Everyone has their reasons why they photography with differing expectations about what they want from it. I personally get satisfaction in being able to produce a final image that not most can achieve (...and that isn't intended to be a conceited comment). Furthermore, if anyone every feels I have something to offer in my approach to post-production I'm always willing to discuss and share.

I just enjoy doing it 😀

Thanks for the response ... it was very interesting and an insight into how we are all different here but all enjoy our photography in whatever way fulfills us. 🙂

 

[remegius]

www.radiantvista.com

Watch the free tutorials and you will learn more than enough about Photoshop to get you started. The new generation of tutorials is geared toward CS3, but 90% is applicable to earlier versions.

Cheers...

Rem

 

[amateriat]

Oh, my stars and garters, let's start at the beginning...

- The Monitor. You won't get anywhere without getting it in order. This can be as simple as using the basic calibration tools within the OS, or as intricate/anal as using a third-party hardware/software solution. (For the record, I use a Spyder 2 Pro system for calibrating both my monitors.) To keep things in reasonable order, you'll need to re-calibrate on the order of every other month or so. A bit of a PITA, but it'll save you a good deal in aggravation and wasted time/materials. (Note: I just re-cal'd my monitors tonight.)

- Scanner Profiling. Yep, you might need to tweak things there, too. If thoughts of profiles make your teeth itch, one simplistic rule of thumb is this: more often than not, Adobe RGB is your friend.

- Get to know a few film types well, then stick to 'em. My fave color films are, descending order: Kodak Portra, Fuji Pro, Kodak E200. In b/w, the order is: Ilford XP2 Super, Kodak BW400CN, Ilford HP5. With my film scanner, I find that I need do only minor tweaking in Photoshop with all but one of the above films (HP5, which requires somewhat more tweaking, not to mention manual spotting since I can't use Digital ICE on conventional b/w film). Which gets us to the subject of...

- Know you scanner. Maybe you couldn't get the scanner you wanted. Hopefully, you have something that does at least a halfway decent job. You might need to do a bit more work with that scanner as a result, and maybe you need to use something other than the driver that came with that scanner. (I recommend trying VueScan, which is designed to be used with a mind-bogglingly huge range of scanners, and won't cost two weeks' groceries to get hold of. You can download it to try out at http://www.hamrick.com.)

- Practice. Practice. Practice.

I fell down this rabbit hole just about ten years ago, and it's been a worthwhile trip. Somehow, my first prints from my first scans turned out somewhat better than the otherwise first-rate prints I got from a Really Good Lab. Perhaps I got lucky, but it's not impossible, or necessarily silly-expensive, to get good results from your own handiwork.


- Barrett

 

[craygc]

Watch the free tutorials and you will learn more than enough about Photoshop to get you started. The new generation of tutorials is geared toward CS3, but 90% is applicable to earlier versions.

Not completely sure how it ever came about now; and not being against any of these approaches and tutorials but for some reason no one ever seems to post-process how I work 😱 Probably just me ....or maybe I was dropped on my head as a child :bang:

In seriousness, my observations on this over time are that many/most people begin post-processing by the numbers - this is about the only way post-processing tutorials can work. Do this, then do that, set adjustment values to some number to achieve an effect, mask off areas that fall within a particular range of tonal levels, and so on... As this methodology tends to work, many/most just continue with the same way, maybe adding new techniques but rarely taking an entirely different approach.

For me, my mother was an artist (painting). Not in the amateur sense. She painted, ran a gallery and exhibited. This was how she derived her sole income. I learnt a lot over the years from talking and watching her work and I ended up seeing any adjustments and changes as an artistic vision that was transferred from my mother to the painting through the brush - these days, I live post-processing almost entirely through a Wacom tablet and pen 😀. I think my more artistic approach to working an image by feel and look on a non-layered environment, rather than by formula, has been strongly influenced from what I adopted from years of observing my mother's painting.

 

[amateriat]

Craig: Yours isn't a bad methodology, although I would hope you have some method of record-keeping for the sake of repeatability of technique when desired (something I've had to work hard at doing for myself). Of course, my opinion might have something to do with my mother having been an artist as well. 🙂


- Barrett

 

[boy_lah]

I looked into scanning myself but I decided given where I am with my free time, it's not for me so I cheat. 😛 I get my local minilab to dev + scan for me.

In my mind, ultimately scanning and and post processing takes time, and quiet a lot of it. That's a key reason why you would do it, to have full control and max quality of your output. For me, my minilab's quality is "good enough" with zero scanning and minimal post processing effort from me.

Their default scans are quiet good most times but can be a bit contrastly at times. But I can just post/print them straight off and spend zero time in front of the PC (a benefit of shooting film over digital in my mind). I have however recently asked them to lower the contrast so i now get a 'flatter' pix - which means I have a better 'starting point' to tweak (in PS or LR) but it also means I need to tweak. Undecided as to which i prefer atm. Long term I guess I probably would want control, so the flatter scan is probably best.

All the pix on my flickr atm are 'default' from my minilab - 90% are pretty good off the bat. Some are a tad blown out esp bright days (e.g. 3rd pix below). Note, 90% of them has had no post processing, maybe some crop here and there.

 

[wizofz2k]

Coolscan V for 35mm and a 9000 ED for Medium Format.
Use both vuescan and nikonscan, depending how I feel or if I want more or less control over the scan or if I need to use ROC/GEM/DEE for old negatives and slides.

All colour adjustments are done first by changing the scan light colour balance until I have an "equi-colour" histogram, then I use the master control to centre the histogram. Very little curves if at all on the scanner, try to do it all by changing the light source intensity in the first place. Digital-Ice for colour and chromogenic. Also use the chroma control in nikonscan: very useful for faded colours.

Always scan at 4000 and 16-bit anyway. The post-processing consists of Neat Image at 16-bit to slightly clean-up the grain on older films. With the newer stuff I rarely bother to do it. It's a batch process anyway. Then I use GIMP for crops and/or slight colour balance adjustement and save as TIFF 8-bit compressed. That's my "master" image.

From there I use Irfanview and Focus Magic to end up with either large jpgs for printing or smaller ones for web use.

Most of the work is scripted and/or batched. Every once in a while an image or roll needs more attention. No biggie.

 

[Sjixxxy]

For me, the insanity is having some huge library of PSD files just to support the layers. With all images being being 4000 dpi scans and mostly B&W thats ~40MB for 35mm and ~185MB for 6x7 as flat TIFFs. Already, this is hard on storage, especially when Im using RAID 1 configurations for working files plus additional incrementally backups.

Here is what I do to ease the file size impact of multilateral TIFFS when working from medium or large format negatives. First, I take my massive untouched file from the scanner and do any cropping/straightening that needs to be done. Save it, and then save it again as a second "work" copy of the file. I scale down this second copy to something my system can handle better, like 2000 or 3000 pixels wide, and do all of my tonal adjustments on this small file via layers and layer masks. When it comes time to print, I upscale the work file, and down scale the full resolution file, then cut & paste the high res file into the work file below all the adjustments. Saves me a ton of time in file read/writes & processing time.

 

[craygc]

Here is what I do to ease the file size impact of multilateral TIFFS when working from medium or large format negatives. First, I take my massive untouched file from the scanner and do any cropping/straightening that needs to be done. Save it, and then save it again as a second "work" copy of the file. I scale down this second copy to something my system can handle better, like 2000 or 3000 pixels wide, and do all of my tonal adjustments on this small file via layers and layer masks. When it comes time to print, I upscale the work file, and down scale the full resolution file, then cut & paste the high res file into the work file below all the adjustments. Saves me a ton of time in file read/writes & processing time.

Ooooh.... that is a clever little approach 😀

My issue now is to work out how to do all my direct work in layers. For example; I'll do a global curves adjustment, mark it as history then go back to the state before and start painting those changes in only to areas I want - and those areas are all painted in by sight, not through masking

-------------------------------------------
a day or so later and...

Well I've now worked out how to deal with these aspects in layers. This has to be one of those very infrequent tips you get that just revolutionizes your workflow. Really powerful and extremely time saving on 4000dpi 6x7 scans. Thank you very much for this...

 

[januaryman]

I had been playing with sliders and curves and filters on my Coolscan V ED using first Nikon's SW, then Vuescan. The other day I just fed the damn thing a strip of film, set everything to "off" except ICE (which I could have set to off, also, I now see) and got a scan of the entire, uncropped color negative that had been save at 4000, to a TIF - the results looked gruesome.

I just sent it over to Elements, dropped on a level adjustment layer, took each color channel one by one and matched the arrows to the beginning or end of the curve and WOW the image just popped. I may go back to the many previous scans I did and just scan it and play later. I have no idea how this is working out, but the results were pretty amazing to my eyes. Perhaps because the original scan was so poor to begin with when it was done with a simple scan. No USM, no curves, no nothing except (afterwards, in PS Elements) adjusting the levels for each color channel. (Well, for a few dark shots I needed to add or reduce contrast and/or brightness, but that was all) - When I get home I'll post a before and after of the basic, quick PP work I did.

 

[luketrash]

My methods change with the seasons. I have a refurbished Epson 3170 I bought in 2005. The lights in the head are getting dimmer and wearing out. I've scanned over 5000 unique negatives with it to date. Most at 3200dpi. Those two light tubes have been on for countless hours at this point.

THIN negatives scan better than thick ones for me now, thanks to this. I still shoot thick, since I plan on having printable negs in the future.

Anyway, here's my really ignorant and simple-minded approach:

I use the Epson software in Windows XP. I open up the histogram window and slide the sliders to the end of the 'mountain' on both sides. This gives me the highest contrast without killing shadows and highlights. I scan.

In photoshop I open up the .tif file I created and I crop&rotate to get things nicely aligned and read for some work. I typically do auto levels and see if that looks good. If not, I open up the curve and pull it around at three points to adjust the shadows, highlights, and midtones.

Then I use a plugin called Fred Miranda's Web Presenter Pro to do a size down to 1600 pixels wide for my Flickr uploading. It sizes down without creating jagged marks on stuff.

So that's it. Each 35mm frame at 3200dpi takes about 3 minutes. Medium format stuff get scanned at 1800dpi takes about 5 minutes of my time. I am very much a 'good enough' sort of guy.

If I was working at exhibiting my photos, A) I'd probably wet print them in my darkroom, or B) pay someone else to do it right, who has proper training.

I literally stumble through the dark because I'm too stubborn to sit down and read a book, and find that there are 1000 ways to arrive at a very similar conclusion, so why waste time listening to ONE book?

I don't do layers, I don't dodge and burn, I DO clone out dust specks occasionally. I still think it's important to actually know how to expose your negatives correctly. Then you won't need to spend much time layering and darkening the sky or whatnot.

As for color work? I spend time with the color sliders to make the photos look JUST HOW I WANT THEM. I don't fool myself into thinking that Fuji Superia 400 has to look like Fuji Superia 400. I make it look how I want it, not how internet purists think it's supposed to look. Anyone who does PROFESSIONAL work at a photography shop will tell you that their pro grade work has been modified greatly to look its best.

Purists. They're only kidding themselves the moment they poke the 'SCAN' button.

I remember someone emailed me to bitch that this is not what Fuji Provia 100 looks like:

Oh well. That's what the shot looked like in my mind.

 

[januaryman]

Hi Luke, how you been? Still play with the Yashicas? - Anyway, your point about getting to look how you want is the key. In fact, I can usually (not always) pick out Fuji Superia400 photos out of a lineup and WISH I knew how to get my photos to NOT look like that. I have been successful about 10% of the time and may just have to jump the Fuji ship to Kodak.

 

[luketrash]

Doing good! This winter I plan on sitting down and repairing all of my Yashica E35s to 'like new' performance.

I hear you about film. What I don't like is when people leave their scans dead and lifeless, thinking they are elite purists. Edit the freakin things! It's not like every print I make in the darkroom is for the same exact amount of light exposure with the same contrast filter... For God's sake, edit the stuff! 😉

 

[Toby]

I've got a HP G4050 and this is how I scan (I only shoot black and white film I use digital for colour):

I use vuescan to produce a DNG file of the negative with no adjustments at all not even inversion ( I usually scan for a 16 bit RGB file). I do this because the scanner in essence does exactly the same thing whatever settings you put into vuecan. The settings in vuescan are on the whole software tweaks and Photoshop is a far more flexible program for altering files.

I then open this file in adobe camera raw. I use the curves tab, select the linear curve and place the two ends of the curve just outside the the histogram so I don't clip out any information. If you want to reduce some grain at this stage you can do it with the luminance smoothing toggle.

Once I have the file in photoshop I invert it and flip to its correct orientation. Then I convert the file to LAB colour and select the only the lightness channel. I then convert to greyscale discarding the other two channels. I fine this massively opens up the shadow detail in the negtive and smooths out the contrast. From this point I edit as I would any other image.

You can view images I've processed in this way on my new website http://www.tobiaskeycommercial.com

 

[Pherdinand]

Well, my mother is not an artist🙂 😉
but sometimes i manage to scan a well exposed 35mm bw frame on auto adjust, then only sliiightly tweak the levels in photoshop, do some basic USM and get a 40x60 cm print out of it that makes my jaws drop.
The moral is, if the film is exposed properly and the scene is not too difficult, it's all easy and straightforward.

But maybe my jaw is too lose.

 

[mh2000]

back in my darkroom days... I always thought that my good photos almost printed themselves and the photos that I spent hours on were somehow not the good ones... so why did I waste my time? Same goes for scanning... good photos scan and require very little work (if you know how to run your scanner).

Layers, don't use them much, if I mess up I can just hit undo a few times... don't want to save them all.

I only do b&w. From Vuescan I scan to 16-bit grayscale, from Nikonscan, I scan as color and convert in Photoshop.

Then mainly levels, curves, crop (if needed) and sharpen... and I'm there. Occasionally I dodge and burn... just like in a "real darkroom." Yes Curves give you more control than VC filters and paper did, but it's the same thing still... and Levels is really just a simple curve.

From my experience, film scans cannot be pushed around nearly as much as digital images can...

Unless I am scanning "real b&w," scan and processing takes less than 5 minutes which is much faster than I ever got in a darkroom, but with real b&w and no DigitalICE I can waste a lot of time cleaning up dust and scratches (My Coolscan V really brings out every defect... so I mostly shoot BW400CN... love that film!).

 

[luketrash]

My only issue with that fact is that the negatives that 'scan themselves' so to speak happen to be somewhat thin. About one stop under what I'd consider normal. I like to shoot my negs about .5 stops fat of normal so that I have all the information in the shadows. Highlights can be managed with a wet darkroom print.

So... There's both and art and a science to my madness 😉