Perceived Sharpness.

[Tony Evans]

As a continuing learner, I am somewhat disappointed with the apparent sharpness of my 4x5 (B&W) scans on my monitor. However, I have not got a keeper worth printing yet. Is there a known difference in sharpness (perceived) between monitor & print?

[Lenny Eiger]

Uh, what scanner?
That's probably it...

Lenny

[John Rodriguez]

As a continuing learner, I am somewhat disappointed with the apparent sharpness of my 4x5 (B&W) scans on my monitor. However, I have not got a keeper worth printing yet. Is there a known difference in sharpness (perceived) between monitor & print?

Yes, your printer can print more detail then your monitor can display.

Some things to keep in mind -

- With high spi scans, viewing at 100% is never going to look super crisp. You have to keep in mind the magnification you're viewing the scans at.
- If you're using a flat bed you must do some input-sharpening

[Tony Evans]

John,
Thanks for that. It's about what I thought. My 18 x 20 inch prints from 135 look much sharper than anything I have seen on the monitor with 4x5. Must start to print! Must start to print!......Must....
(Scans are at 2400 on a V700 + BSH set at optimum 3.5mm).

[domaz]

If you zoom in on your 135 scans and 4x5 scans it should be pretty clear there's a lot more detail in the 4x5 scans. Yes you will need to run the Unsharp Mask filter to get a sharp looking monitor image.

[Bruce Watson]

Is there a known difference in sharpness (perceived) between monitor & print?

Of course. Monitors bigger than an iPhone typically have 72-100 or so ppi available. Most photographic printers print at least 300-360 ppi.

You can not accurately judge print sharpness on a monitor.

And I won't even get into the difference between a light source (your monitor) and a reflective source (your print).

The only way to accurately judge what a print will look like, is to make a print. That's just the laws of physics talkin' to ya.

[ashlee52]

Germain to this discussion, I have a similar issue... but I do print. My Epson R2400 13x19B&W prints from nice 5x7 Negatives look no better than prints from my Nikon D300. And they look noticeably less detailed than darkroom prints of 4x5 back when I had a functioning dark room. I've been using 5x7 because the larger scan area... and frustratingly I can't just go print it in a dark room because I have only a 4x5 enlarger. In color, even using a loupe on FUJI NPS negatives I get no more detail than with my digital camera. Which of the following do people think is right:

1) Quit using f32 so much because diffraction limits are reducing the advantages of a big negative. Stick to f22 max.
2) Use your scanner better... scanning for 360 dpi output isn't giving enough detail
3) Learn to "sharpen" better. Your standard 100% sharpening with a 1.4 radius isn't enough. Try settings of _____???.
4) Give up on a digital work flow and set your darkroom back up... and therefore go back to 4x5 for practicality. It's cheaper anyway.
5) Pay $100 to have every half decent shot drum scanned.
6) Use only sharper film... T-max and Ektar.
7) Find better paper than Hahnemuhle Fine Art Pearl.
8) Your problem is the R2400... you will get better prints off a Costco Light Jet... plus you can get a 20x30 inch print for less than what you pay to print 13x19 at home
9) Use your digital camera, it works as well as a 5x7.
10) Something else...

[David Higgs]

There are more experienced people here than me, but I haven't been blown away by LF 'sharpness', partly due to a hybrid workflow - I'm getting good at scanning on a flat bed now, although some nice drum scans has shown me potential left I'm not accessing.

In truth those pesky digital cameras are very very good - I love film so much, but for sharpness even my Micro 4/3rds camera is hard to beat - up to a certain size. If you don't print large, and sharpness is the only quality you are measuring ( I know it won't be) , digital is going to be hard to beat.

I've found drum scans for colour are pretty much essential, I can get away with using the Epson for B+W for most jobs.

You don't mention lenses, I have some LF lenses that really cannot compete against 35mm resolution, and some that are really sharp.

Get a print done elsewhere, but at 13x19 I wouldn't necessarily expect to see a striking difference with a D300 - tonality and feel are a different matter though.

[John Rodriguez]

Germain to this discussion, I have a similar issue... but I do print. My Epson R2400 13x19B&W prints from nice 5x7 Negatives look no better than prints from my Nikon D300. And they look noticeably less detailed than darkroom prints of 4x5 back when I had a functioning dark room. I've been using 5x7 because the larger scan area... and frustratingly I can't just go print it in a dark room because I have only a 4x5 enlarger. In color, even using a loupe on FUJI NPS negatives I get no more detail than with my digital camera. Which of the following do people think is right:

1) Quit using f32 so much because diffraction limits are reducing the advantages of a big negative. Stick to f22 max.
2) Use your scanner better... scanning for 360 dpi output isn't giving enough detail
3) Learn to "sharpen" better. Your standard 100% sharpening with a 1.4 radius isn't enough. Try settings of _____???.
4) Give up on a digital work flow and set your darkroom back up... and therefore go back to 4x5 for practicality. It's cheaper anyway.
5) Pay $100 to have every half decent shot drum scanned.
6) Use only sharper film... T-max and Ektar.
7) Find better paper than Hahnemuhle Fine Art Pearl.
8) Your problem is the R2400... you will get better prints off a Costco Light Jet... plus you can get a 20x30 inch print for less than what you pay to print 13x19 at home
9) Use your digital camera, it works as well as a 5x7.
10) Something else...

10) Something else - print bigger.

My 12x18 prints from D200 5D prints are not too far off in sharpness from prints from my 4x5=>4870 prints of the same size. In terms of tonality and color depth the 4x5 prints are much better, and they still have a resolution advantage, albeit small. However, bump up the size any and there's a huge difference. I just don't think 12x18 is large enough to utilize the resolution advantage.

I also have some prints from 4x5s scanned on an Aztec Premier. The files are shaper, but they don't show up in a print for me until I get into larger sizes.

Try this - shoot the same scene with your D300 and your 4x5. Scan the negative. Pixel peep both. Do you get more info with one or the other?

[David Higgs]

I agree with John!

[ashlee52]

Try this - shoot the same scene with your D300 and your 4x5. Scan the negative. Pixel peep both. Do you get more info with one or the other?

Oh how I hate to type this. I looked at 100% files of a 4x5 scan (FUJI NPS) and my D300 (not even the Leica M9 that for whatever reason I can't bond with)... and the resolution is fundamentally equal. Then I put a loupe to the 4x5 negative, and the scan is getting everything that is there. I do suspect that other films would have more detail than the FUJI NPS... but I am using up what I have. Eventually I expect to try Portra (I love that palette) and Ektar (said to be the sharpest).

With HP4 on 5x7 however I know that there is much more detail in the negative than is making it to the print. Maybe I need to print bigger... and no way do I want to spend what a larger printer costs... so that means having someone print it. Does anybody have experience with light jet RA-4 prints from B&W? I've heard they have color casts.

[Lenny Eiger]

In color, even using a loupe on FUJI NPS negatives I get no more detail than with my digital camera. Which of the following do people think is right:

1) Quit using f32 so much because diffraction limits are reducing the advantages of a big negative. Stick to f22 max.

This is lens dependent. And some other things. I have just done a test with my 150 Sironar S and f45 looks just as good as f22.


2) Use your scanner better... scanning for 360 dpi output isn't giving enough detail

There is always more to learn about scanning.


3) Learn to "sharpen" better. Your standard 100% sharpening with a 1.4 radius isn't enough. Try settings of _____???.

If you can, use a better scanner, or actually focus the camera. I would never consider 1.4. My scans are sharpened at a radius of .2. For other scanners and digital cameras, I might recommend as high as .8. But 1.4 is definitely too far.


4) Give up on a digital work flow and set your darkroom back up... and therefore go back to 4x5 for practicality. It's cheaper anyway.

Not a good question. The medium's are different. Inkjet prints are very different, or they can be. Go look at some good ones and decide which fits your aesthetic.


5) Pay $100 to have every half decent shot drum scanned.

No, pay for drum scans when either the shot needs it, your aesthetic needs it, it's one of your top images, or you want to print large.


6) Use only sharper film... T-max and Ektar.

Not a bad idea. It is good to choose one and make it work over time. Don't forget Delta in there...


7) Find better paper than Hahnemuhle Fine Art Pearl.

There are papers that are better, but only marginally. Hahnemuhle's Baryta is very good. However, I would also consider looking at what a really good printer can do on PhotoRag. It's not shiny, and that means it has some very interesting qualities. Personally, I like it much better than the shiny stuff.


8) Your problem is the R2400... you will get better prints off a Costco Light Jet... plus you can get a 20x30 inch print for less than what you pay to print 13x19 at home

No, learn to print. Get better at it, study other good printers, learn more, get better, etc.


9) Use your digital camera, it works as well as a 5x7.

Patently ridiculous..


10) Something else...

There's always something else. How about my favorite, get a copy of the History of Photography, do some reading, learn about the different genres in photography, figure out what speaks most to you and get clear about what you are trying to accomplish. It will be much easier to print when you have a style that is all your own that you are after...


Lenny

[ashlee52]

However, I would also consider looking at what a really good printer can do on PhotoRag. It's not shiny, and that means it has some very interesting qualities. Personally, I like it much better than the shiny stuff.

I have tried to print on both Hahnemuhle Photo Rag and Moab Entrada, using their profiles for the 2400 and get muddy prints with blotchy shadows. I'd love to be able to print on those richer papers. Plus I'd love to be able to do something with the pile of very expensive paper sitting here.

[Kimberly Anderson]

Learn sound Color Management practices.

Your problem with your prints (Ashlee52...) probably lie there. It would behoove us all to make sure we understand color management execution as well as philosophy before we start making excuses for our prints looking not as expected.

[Ken Lee]

I have tried to print on both Hahnemuhle Photo Rag and Moab Entrada, using their profiles for the 2400 and get muddy prints with blotchy shadows. I'd love to be able to print on those richer papers. Plus I'd love to be able to do something with the pile of very expensive paper sitting here.

Is your monitor calbrated?

[Lenny Eiger]

I have tried to print on both Hahnemuhle Photo Rag and Moab Entrada, using their profiles for the 2400 and get muddy prints with blotchy shadows. I'd love to be able to print on those richer papers. Plus I'd love to be able to do something with the pile of very expensive paper sitting here.

As others have said, you can't blame this on paper. I've been working with PhotoRag for a very long time and I believe it is fair to say that I understand it well. It does not deliver muddy prints unless you send it a muddy file to print.

I don't quite agree with the comments about color management. Maybe I'm not being fair tho', I am color-managed to the hilt. I see it as the beginning, however. I spend most of my time looking at the print vs the screen. I don't imagine the screen will look like the print. I print just like I used to in the darkroom, I make a print, look at it and see if I can make a "better" one. If I can I make an adjustment and send it off again until I get what I am looking for. Color management gets me a predictable response.

One can certainly over-expose and get some muddiness, or scan improperly and get same. However, I think the most important aspect is to understand the curves dialog, and how to mask so that you affect only the areas you want to change. If you have blocked up shadows, change the curve in that area, lighten, add some contrast, etc.

Everyone's trying to promote something these days (not to mention some of the incorrect answers people can find over at photo.net), and they talk about some cool shortcut or another and the best thing people can do, IMO, is use the curves until they understand them. Then learn to mask....


Lenny

[John Rodriguez]

Oh how I hate to type this. I looked at 100% files of a 4x5 scan (FUJI NPS) and my D300 (not even the Leica M9 that for whatever reason I can't bond with)... and the resolution is fundamentally equal. Then I put a loupe to the 4x5 negative, and the scan is getting everything that is there. I do suspect that other films would have more detail than the FUJI NPS... but I am using up what I have. Eventually I expect to try Portra (I love that palette) and Ektar (said to be the sharpest).


Wow, I don't know what to tell you. I haven't tried NPS, but it can't be that bad. So far I've shot Portra 160, Ektar, Provia and Acros. Viewing a 4870 scan at 2400 dpi on screen shows way more image info then my 5D or D200 files ever did, it's not even remotely close. Once again, I have to zoom in to see it (just like printing big). It's not even possible for me to view a DSLR file at the same scale as 100% zoomed in on a 4x5.

[ashlee52]

Is your monitor calbrated?

Yep, calibrated monitor.

The muddying up I had was with Velvia source scans, and perhaps would be less severe with color neg which has less dense shadows.

But no way are my Epson prints as sharp say as the printed page in high quality photography books... on any paper. Nor are they as sharp as 10x10 prints from medium format on traditional darkroom papers.

I guess I need a workshop. I live a few miles from Andersen Ranch... I'll have to find something there.

[Lenny Eiger]

Yep, calibrated monitor.

The muddying up I had was with Velvia source scans, and perhaps would be less severe with color neg which has less dense shadows.

But no way are my Epson prints as sharp say as the printed page in high quality photography books... on any paper. Nor are they as sharp as 10x10 prints from medium format on traditional darkroom papers.

If you want sharpness, try a drum scanner.... at least for comparison and for something to shoot for with a lesser scanner and some sharpening technique. I have great luck with all kinds of b&w and color film.... it's not the film...

Lenny

[Peter York]

Then I put a loupe to the 4x5 negative, and the scan is getting everything that is there.

Can you give an example of a scene you shot and how you focussed the camera? It may be an issue with your technique. With LF, there are so many ways to screw up a shot.

[Ken Lee]

You might find it easier to work with b&w film that has been exposed and developed to fit the dynamic range of the scanner. Getting the scanner to fit the dynamic range of a contrasty slide is... an art.

[Kirk Gittings]

You might find it easier to work with b&w film that has been exposed and developed to fit the dynamic range of the scanner. Getting the scanner to fit the dynamic range of a contrasty slide is... an art.

ditto.

Easier to expand DR in the scan than compress a contrasty negative. I develop my negs to print normally on a grade 3 paper-meaning they are a bit low in contrast. They then scan well for ink prints and or print well on silver papers.

[Brent Long]

I would never consider 1.4. My scans are sharpened at a radius of .2. For other scanners and digital cameras, I might recommend as high as .8. But 1.4 is definitely too far.


Lenny


Hmmm . . . I think I'd like to take issue with this. I've recently found that using a radius of up to 2, even, can be useful. In scanning a piece of 35mm neg film on both a Canon FS4000US and then on my Epson V700, I found that sharpening at 2, and a fairly strong amount, gave me practically the identical image as the dedicated film scanner, which is quite sharp (with the added "bonus" of emphasizing the noise, but that was able to be dealt with as well).

Since then, I've been experimenting with higher radius sharpening and am finding that there aren't as many rules as I once thought.

I am intrigued by your comments, though, and will continue to experiment.

[Ken Lee]

One of the most important factors in sharpening, is to sharpen the different regions of the brightness scale, separately (http://www.kenleegallery.com/html/tech/index.php#Sharpening).

Another, when working in color, is to sharpen only a duplicate layer whose Blend Mode is Luminosity. This amounts to sharpening only the grayscale portion of the image. (This can be combined with the method described above.)

[Brent Long]

Ken, yes, I had noticed your tutorial before and haven't yet gone back to play with it. It looks quite good. I have/had been using PhotoKit Sharpener and it is useful, but I think there are some things I can get better by myself.

[ashlee52]

Thanks to the pushing here I finally made some good prints on my EPSON R2400 and Entrada Rag... the secret was reading this:

"The Front feed is for extremely thick media and the heads are farther away from the paper than the sheet and manual roll feeder. As a result if the paper is not thick enough for the front feeder you get overspray, which looks like smearing as the colors overlap.

When you say sheet feeder are you talking about the manual roll feeder in the back or the regular sheet feeder on top. Heavy Fine Art Paper must be fed through the Manual Roll feeder in the back and "Manual Roll" selected as the paper source in the printer driver."

So basically with the thick media I had been front loading the printer which resulted in an overspray problem.

Getting 300 GSM papers to feed into this printer is a chore any way you do it, but I now know that I need to use rear feed.

[Bob McCarthy]

One of the most important factors in sharpening, is to sharpen the different regions of the brightness scale, separately (http://www.kenleegallery.com/html/tech/index.php#Sharpening).

Another, when working in color, is to sharpen only a duplicate layer whose Blend Mode is Luminosity. This amounts to sharpening only the grayscale portion of the image. (This can be combined with the method described above.)


Ken, have you tried the new Photokit sharpener 2.0. Works very well in latest iteration.

Also, to improve scanner microcontrast try unsharp mask

at 15-20 % radius 50-70 yes I know odd numbers, but it makes up for low mtf of most scanners.

Glad your liking the 200M, I do have sellers remorse. Use it in good health.

bob

[rdenney]

Yep, calibrated monitor.

The muddying up I had was with Velvia source scans, and perhaps would be less severe with color neg which has less dense shadows.

But no way are my Epson prints as sharp say as the printed page in high quality photography books... on any paper. Nor are they as sharp as 10x10 prints from medium format on traditional darkroom papers.

I guess I need a workshop. I live a few miles from Andersen Ranch... I'll have to find something there.

Here are a couple of data points from another amateur: I use an Epson V750 with 4x5 and a Nikon 8000 with roll film. I have an Epson 3800 printer and make 16x20 prints. My 4x5 prints at that size, scanned in the Epson flatbed, are very sharp. When I compare the scan to the negative using a 50x microscope, I see that the scanner is slightly aliasing the grain (Ilford FP4 film, which does have some grain). Thus, images from that scanner will show a graininess on enlargement before a conventional print would show it. But I see NO grain at 16x20 print size, so whatever size is required is bigger than that.

I scan at 2400 and end up with scans that are about 9100x11400 pixels from 4x5. I'm therefore printing at over 550 pixels/inch or a bit more. I have tried downsampling it to 360 and sharpening at that resolution (with small sharpening numbers, as Lenny suggests), but I can't see any difference between that and just printing it at 566 and letting the Epson driver do the resampling. The prints look critically sharp to me.

A 6x7 frame scanned in the Nikon produces a file with about 8800x10000 pixels. It prints to 16x20 at the same number of pixels/inch as the 4x5. At 16x20, I get largely similar results, and the photos are very sharp. Color photos never have the crispness of black and white, and much of what I've printed from rollfilm has been color while most of my 4x5 prints have been black and white, so comparisons are difficult. But I think roll film in the Nikon matches 4x5 in the Epson pretty closely.

I'm not sure either would hold up to bigger prints.

I have made 12x15 prints in an ancient Epson 1270, scanned at 1200 spi from 4x5 black and white negatives in a scanner that was unlikely to do better than about 800, that were stunningly sharp. Those scans lose the "stunning" when printed larger, but if you can't get "stunning" from your setup, then something is wrong.

On the muddiness: Make sure you are using a color space with a large gamut, and scan at 48 bits/pixel. Don't use JPG formats for intermediate images. I need a little air on each end of the histogram to really have something to work with in Photoshop. And I often really pump up the contrast (to make up for my scanner). Those results may not be as smooth as a conventional print, but in return for that they more closely match my visualization. Trying to do that with a 24-bits/pixel file will lose smoothness much more quickly, and those shadows will be pounded into the limit of the histogram where you have no room to stretch them. This is especially true with scans of Velvia.

And don't oversharpen. In my eyes, that actually makes the image less sharp. That's the reason for Lenny's suggestion of small numbers for sharpening radius. I have actually salvaged photos that were not perfectly focused by giving them a soft-focus effect with a transparent duplicate layer of the image with gaussian blur applied. This actually makes the underlying image look sharper than the original, because it reduces the contrast at those fat, blurry edges. The fatness of those blurry edges is made more apparent by sharpening at too large a radius.

Finally, on screen resolution: Prints have something like 3 or 4 times the visual resolution that monitors do. So, you can roughly simulate the sharpness you might perceive from a monitor by backing up to 3 or 4 times the viewing distance. That way, the size of the pixels on the monitor shrink to about the same apparent size as the print. Display the image to the same apparent size from these viewing distances, and you'll get a slightly better impression. But the print will always look sharper, because our brain perceives more sharpness when the only limit to apparent sharpness is our ability to focus our eyes closely.

Rick "my own home digital darkroom findings" Denney

[Tony Evans]

Rick,

Thank you!

[cps]

Ashlee52,

I know your paper path issue seems to have been a culprit, but I will echo what others say here about color management too. A few years ago I switched to making my own color profiles for the paper I use (with a Spyder) instead of using the canned ones from the manufacturer and never looked back. My profiles aren't always perfect, but work quite a bit better for me than off-the-shelf profiles. A lot of hours and expensive paper and ink were wasted before this.

[Brent Long]

And don't oversharpen. In my eyes, that actually makes the image less sharp. That's the reason for Lenny's suggestion of small numbers for sharpening radius. I have actually salvaged photos that were not perfectly focused by giving them a soft-focus effect with a transparent duplicate layer of the image with gaussian blur applied. This actually makes the underlying image look sharper than the original, because it reduces the contrast at those fat, blurry edges. The fatness of those blurry edges is made more apparent by sharpening at too large a radius.

Rick "my own home digital darkroom findings" Denney

Rick, I'm sure I'm just being very dense, but what, exactly, do you mean by "transparent duplicate layer? For some reason this isn't clicking for me. I think you mean to set the duplicate layer to "Luminosity"??? From a quick test I guess that must be it, but would like to be sure.

[Brent Long]

One of the most important factors in sharpening, is to sharpen the different regions of the brightness scale, separately (http://www.kenleegallery.com/html/tech/index.php#Sharpening).

Another, when working in color, is to sharpen only a duplicate layer whose Blend Mode is Luminosity. This amounts to sharpening only the grayscale portion of the image. (This can be combined with the method described above.)

Ken, I've taken a few minutes to give your advice a try. I must say, it's very, very good. I didn't imagine it would be that useful, to be honest. Thank you very much.

[Ken Lee]

"I must say, it's very, very good."

Excellent :)

One of the most annoying things about sharpened images is the artifacts. Sharpening the different regions of the scale by differing amounts - and with different settings - helps avoid them in a big way.

The big culprit here is trying to use only one setting. For Fine Art images, using only one setting is rarely appropriate across the entire image.

In the same vein, sometimes it's best to sharpen only certain regions of the scale - like the mid-tones - and leave other areas alone, not sharpened at all. It all depends on the individual image.

When we sharpen on duplicate layers, we can also use the Erase Tool to remove areas that we don't want sharpened - like clouds or moving water. We can set the Erase Tool's percentage, so that we can partially erase certain areas. That lets us "paint-in" - or paint out - the sharpening, so to speak, wherever we want it.

[Brent Long]

Thanks again, Ken. The thing that was putting me off of PhotoKit Sharpener was that I was getting some artifacts around edges adjacent to smooth graduated areas such as the horizon and sky. I started sharpening the whole image so that at least the artifacts would be, more or less, even across the image, which appeared better than only along certain edges. I was quickly able to control this using your method.

As for PK sharpener, I tried to hide the artifacts by adjusting opacities, but just couldn't get it right on many images.

[rdenney]

Rick, I'm sure I'm just being very dense, but what, exactly, do you mean by "transparent duplicate layer? For some reason this isn't clicking for me. I think you mean to set the duplicate layer to "Luminosity"??? From a quick test I guess that must be it, but would like to be sure.

It's easy. I make a duplicate of the background layer, apply a Gaussian blur with a very large value for the number of pixels, and the adjust the transparency of that layer down to about 50% or thereabouts. Whatever sharpness is in the background layer shows through, with finer apparent detail than it originally had. It's an optical illusion, of course, but it demonstrates how too much sharpening at too large a radius can make an image seem less sharp. I just did it the other day for an image we wanted to put on the cover of a Blurb wedding book. My wife had made the picture, and missed the focus slightly.

Rick "surprised by how well this works for small prints" Denney

[Greg Miller]

[COLOR="Blue"]When we sharpen on duplicate layers, we can also use the Erase Tool to remove areas that we don't want sharpened - like clouds or moving water. We can set the Erase Tool's percentage, so that we can [I]partially erase certain areas. That lets us "paint-in" - or paint out - the sharpening, so to speak, wherever we want it.

I really hate to see people using the eraser tool, since it is destructive (once you close the file, those pixels are gone forever so future fine tuning means redoing it all). I much prefer that people use layer mask and paint out an effect using a black paint brush. Opacity of the brush can be set just like the Eraser's percentage. Using this non-destructive method, the sharpening that has been hidden can fine tuned forever, and yields the same result as the eraser.

[Ken Lee]

- Excellent - Thanks !

[Greg Blank]

Is there a reason you can't save two versions of the file or is storage space that cost prohibitive?

[Brian Ellis]

Absolutely. A typical monitor is about 110 ppi (so I've read). You'll likely print at something like 300 or 360 ppi.

[Brent Long]

It's easy. I make a duplicate of the background layer, apply a Gaussian blur with a very large value for the number of pixels, and the adjust the transparency of that layer down to about 50% or thereabouts. Whatever sharpness is in the background layer shows through, with finer apparent detail than it originally had. It's an optical illusion, of course, but it demonstrates how too much sharpening at too large a radius can make an image seem less sharp. I just did it the other day for an image we wanted to put on the cover of a Blurb wedding book. My wife had made the picture, and missed the focus slightly.

Rick "surprised by how well this works for small prints" Denney

Sure . . . sure, she did. That's what they all say, Rick :D .

Thanks.

[Brent Long]

I really hate to see people using the eraser tool, since it is destructive (once you close the file, those pixels are gone forever so future fine tuning means redoing it all). I much prefer that people use layer mask and paint out an effect using a black paint brush. Opacity of the brush can be set just like the Eraser's percentage. Using this non-destructive method, the sharpening that has been hidden can fine tuned forever, and yields the same result as the eraser.

Glad you chimed in, Greg. I usually use masks for these types of things and had never played with the eraser tool. I thought I may be missing something, but if the eraser tool us not undoable as a mask is, I'll not bother.

The history brush can be used similarly, but I guess it, too, must throw away future possible adjustments when closed. I guess what I heard when I first got into PS was true, learn to mask, learn to mask, learn to mask.

[Ken Lee]

I really hate to see people using the eraser tool, since it is destructive (once you close the file, those pixels are gone forever so future fine tuning means redoing it all). I much prefer that people use layer mask and paint out an effect using a black paint brush. Opacity of the brush can be set just like the Eraser's percentage. Using this non-destructive method, the sharpening that has been hidden can fine tuned forever, and yields the same result as the eraser.

When we paint with black (or gray) inside the layer mask, it looks as though we have deleted the pixels, but they can be restored by painting them white. It's a bit surprising to see, but it works great.

The same effect can be had by deleting inside the layer mask. Painting white over the area, restores the pixels.

Thanks again for pointing this out !

[Ken Lee]

...to improve scanner microcontrast try unsharp mask at 15-20 % radius 50-70 yes I know odd numbers, but it makes up for low mtf of most scanners.


Brilliant.

As Brent said: I must say, it's very, very good. :)

[mdm]

Thanks for that. The answer I was looking for because this one printed grey and mushy.

[rdenney]

Sure . . . sure, she did. That's what they all say, Rick :D

It's that crappy Nikon D300 she uses. I never have that problem with my Canon 5D.

Rick "happily married" Denney

[SeanEsopenko]

Sorry for esurrecting a thread that hasn't been touched in 5 days.

From what I've seen digital is susceptible to edge aliasing and I think it lacks a general tonality that even small format films have. One Software's Perfect Resize uses some good algorithms but it can't fix the heavy aliasing that can be coaxed from a DSLR with a very sharp lens and the right subject. It prints great but only up to a certain size. I find my 5D, even with Zeiss glass, can only print up to about 13x19 before it starts to require huge amounts of laborious tweaking to go any larger. A 4x5 scan on my V700 with some dust spotting beats it at 20x24 in 1/4 the time spent. For those interested I scan at 6400 dpi then downsample to 2200 or 2400 (depending on my mood). With my new Mamiya 7II I notice more "perceived" sharpness when I downsample to 3000 dpi instead of 2400 like I usually do.

I print "sharpening test strips" for certain texture regions of a print to see what various sharpening settings do with my R3000. I wish I spent the extra $500 or so and got a printer one size up (hindsight's always better than foresight) but from the inkjet crops I print there's a lot to be done with the tweaking of sharpening because what's seen on the monitor does not translate to what's seen on the print, like others have said.

As for "perceived sharpness" because this is what this thread is all about, I've been trying to figure out ways to increase the perceived sharpness of optical RA4 prints. There's tons of cool stuff to with a 10x loupe when enlarging a 4x5 to 16x20 but step back and I'm not getting the perceived sharpness I get with an inkjet. I'm just getting into masking but I'd like to get to the point where I'm churning RA4 prints out that are perceptibly as sharp as an inkjet but with the viewing loupe sharpness inherent in the optical process.

With the inkjet and a 10x loupe all I see are dots but honestly viewing prints that close is for the artist's benefit; nobody else is going to stick their nose up that close and they can't if the print's behind framing glass :)