why scanned negatives look so grainy?

[Warton]

Advertisement (gone after registration)

I hope someone can help me out with this issue.

I shot a roll of Ilford HP5 ISO400 BW, developed it myself using ilfosol3 according to spec (6.5 minutes developing, etc.)

And then I scanned the negatives using a Nikon D810 and 60mm Macro lens with good lightings.

Finally I used capture one to revert the negatives.

And it turns out most of the pictures are over exposed by 1 to 2 stops. Not a problem, so I adjust it to my taste.

However I found they are mostly very noisy with high grains.

I am puzzled. Which steps caused this grainy issue? Is it because I overexposed the shots? or because I developed for too long (such as when pushing the film, I should developed accordingly, i.e. developing time to be for ISO200?)

Thanks for your input.

[Advertisement]

The Leica Q3 is available for pre-order at B&H Photo Video and Adorama

[250swb]

Overexposure starts the ball rolling.

[adan]

4 hours ago, Warton said:

I am puzzled. Which steps caused this grainy issue? Is it because I overexposed the shots? or because I developed for too long (such as when pushing the film, I should developed accordingly, i.e. developing time to be for ISO200?)

Yes and yes. Film grain (unlike digital noise) tends to strongest in dense areas of the film (highlights, or overall overexposure). Overdevelopment "plates out" dissolved silver back onto the silver crystals, making them larger yet.

However another factor in reproducing film images onto digital pixels in grids or lines is "grain aliasing," which tends to be more common with color neg films, but can turn up anywhere.

Even this posting from 23 years ago discusses it.

http://www.photoscientia.co.uk/Grain.htm

Yet another factor is the character of the light back-lighting the film for copying. A smooth frosted panel for diffuse light helps hide grain and other artifacts. A point light source (e.g. undiffused LEDs) will make all of those stand out more strongly.

When I was shooting film in Leicas with digital reproduction in mind, I generally 1) used the lowest speed film possible (Ilford Pan F ISO 50), 2) uprated the metered ISO setting to 64 or 80, for a bit of underexposure - AND 3) reduced development by at least 10% from "normal box instructions."

It was the only way I got scans comparable to darkroom printing onto gelatin paper.

A 2003 image handled that way. 28 Elmarit, Pan F.

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

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

 

Edited August 14, 2023 by adan

[250swb]

Perhaps another thing to consider is what you are comparing scanned grain against? If you previously used a flatbed scanner or even a dedicated film scanner with a real resolution lower than your Nikon D810 you may now be seeing grain as it really is and not as diffused texture. Add to this over-exposure and it's a culture shock which can highlight the need to be careful at each stage.

[Warton]

Thanks for all the valuable inputs.

I am aware the high resolution scanning using D810 can contribute to excessive grains comparing to low resolution flatbed scanner. I don't have a scientific way to quantify the grains. They just look too coarse to my layman's eyes.

I read it somewhere else that underexposure can also contribute to the noise. Besides, I also think my lighting source is not well diffused. It's just a very bright working LED lamp. I am going to invest a good, universal, diffused light source.

A couple things I get out from this conversation is: 1, using low ISO T-grain films, like T-Max 100, 2, don't over expose the shots too much, maybe even under expose them by 1/3 stop etc. 3, developing the rolls in a shortened time, like 5-10% shorter? 4, using a universal well diffused lighting source.

I know as much as I can do the grain will always be there and I just have to adjust my expectation to that not to comparing the digital pictures.

[hansvons]

This is HP5 at box speed, developed in Xtol at box speed. The negative was scanned with the super-sharp Sigma 70mm at 1:1 on an SL2-S and converted in Capture One. However, for what it is, this 400 ISO film does the job, shows plenty of acutance, and is not too grainy. I prefer Delta 400 in the ISO 400 class, but I'm happy with my results. 

Ilford HP5 at box speed in Xtol, 35mm Nokton V2 SC @ f 8.  Click to enlarge.

 

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

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

 

Below is Delta 100 pushed to ISO 200, also cooked in Xtol, with the same scanning. This time the 35mm Summicron ASPH @ f 8. 

The grain is less pronounced than HP5, and the resolution is higher. Developed at box speed and exposed at IE 80 makes Delta 100 less contrasty and raises the resolution even further. Please note that this forum doesn't allow 4K large files. So, the examples are downsampled to 2K. Click to enlarge.

 

 

On 8/14/2023 at 10:13 PM, Warton said:

A couple things I get out from this conversation is: 1, using low ISO T-grain films, like T-Max 100, 2, don't over expose the shots too much, maybe even under expose them by 1/3 stop etc. 3, developing the rolls in a shortened time, like 5-10% shorter? 4, using a universal well diffused lighting source.

• Use any film you like. Don't worry about T-grain etc... 
• DO overexpose a bit, or at least don't underexpose. Film loves light.
• Develop the film with a proper developer accordingly to the film's manual. Because you don't like grain much, avoid Rodinal. Kodak D76, Xtol and Ilford's recommended developers do all a good job. Don't pull and push until you know precisely what you are doing. Boxspeed all the way for best results. 

 

 

 

Edited August 16, 2023 by hansvons

[charlesphoto99]

Advertisement (gone after registration)

Yes, over development and overexposure will increase the grain, sometimes dramatically. Be sure to expose for the highlights when dslr scanning (that is if the highlights are important to you). And check the white point clipping if using NLP in LR to convert, or however you convert. You can crush shadows to hide grain there. I find that the dynamic range is so great with my D850 scanning that I have to do a lot of dodging and burning in PS in post (b&w neg). It does expose everything, sometimes not such a good thing, at least compared to my inoperable Imacon. 

[Warton]

4 minutes ago, charlesphoto99 said:

Yes, over development and overexposure will increase the grain, sometimes dramatically. Be sure to expose for the highlights when dslr scanning (that is if the highlights are important to you). And check the white point clipping if using NLP in LR to convert, or however you convert. You can crush shadows to hide grain there. I find that the dynamic range is so great with my D850 scanning that I have to do a lot of dodging and burning in PS in post (b&w neg). It does expose everything, sometimes not such a good thing, at least compared to my inoperable Imacon. 

That’s another question coming to my mind, how do you exposure when scanning with SLR.

I have sort of good lighting source so I tried with small aperture like f8 or f11, and reasonably fast shutter like 250.

but how do I know if it’s properly exposed? I turned on liveview on Nikon d810 but it’s very difficult to judge the exposure by just looking at the negative or even histogram 

[hansvons]

15 minutes ago, Warton said:

That’s another question coming to my mind, how do you exposure when scanning with SLR.

In the neg, whites are the positive's blacks and vice versa. The best is to do a test scan, crop out everything that is not the image and evaluate the exposure. Without moving the exposure slider in C1, the positive should look ok from the beginning. If the image is too bright, raise the exposure by dialling down the exposure time. If it's too dark do the opposite. This is somewhat counterintuitive, but you are working inverted. 

[hansvons]

22 minutes ago, Warton said:

I have sort of good lighting source so I tried with small aperture like f8 or f11, and reasonably fast shutter like 250.

I have no idea what light source you are using, but when using an iPad or something similar (iPads show an even light, which is crucial), I get something of 1/3 at f 11 and ISO 100. Perhaps you are underexposing your scan and thus overexposing your positive.

[Warton]

14 minutes ago, hansvons said:

I have no idea what light source you are using, but when using an iPad or something similar (iPads show an even light, which is crucial), I get something of 1/3 at f 11 and ISO 100. Perhaps you are underexposing your scan and thus overexposing your positive.

I have a very bright workplace LED light with Nikon ES2 kit which has a diffuser. I’m kind of sure the light is sufficient and decent. But you maybe right I may underexpose the negative therefore overexposure. I may try again with slower shutter tonight with a tripod.

Thanks for the inputs, very helpful 

[250swb]

9 hours ago, Warton said:

 

but how do I know if it’s properly exposed? I turned on liveview on Nikon d810 but it’s very difficult to judge the exposure by just looking at the negative or even histogram 

It's properly exposed if there is no clipping in the shadows or highlights. Your DSLR only copies what information is in the negative, it can't add anything (like increase the dynamic range of the negative), it can only subtract information by clipping. Having too much (or any) border showing in the copy you are making can skew the exposure so make sure you are using centre-weighted metering, equally having too much border showing can skew the post processing result whether in ACR or NLP or whatever you use. But with that in mind I set my camera on auto and use aperture priority, it never fails, and then I crop the negative image in ACR before doing any small adjustments and before inverting it.

[charlesphoto99]

FWIW, I'm using a Skier Light Box, which gives me an average exposure of f18 @1/60th of a second. If the neg is blocked up, I might go to 1/30 or 1/15th, if thin then 1/100th. That's with a Nikon 850 and 100mm Tokina macro on a Kaiser copy stand with Negative Solution carriers and using Negative Lab Pro in Lightroom to convert. 

[Warton]

8 hours ago, 250swb said:

It's properly exposed if there is no clipping in the shadows or highlights. Your DSLR only copies what information is in the negative, it can't add anything (like increase the dynamic range of the negative), it can only subtract information by clipping. Having too much (or any) border showing in the copy you are making can skew the exposure so make sure you are using centre-weighted metering, equally having too much border showing can skew the post processing result whether in ACR or NLP or whatever you use. But with that in mind I set my camera on auto and use aperture priority, it never fails, and then I crop the negative image in ACR before doing any small adjustments and before inverting it.

Does this also assume the film was exposed and developed properly in the first place?

After I scan and invert the negative, when I see it's too bright. How do I know if it's because of the negative being scanned underexposed or film being overexposed?

 

[Warton]

2 hours ago, charlesphoto99 said:

FWIW, I'm using a Skier Light Box, which gives me an average exposure of f18 @1/60th of a second. If the neg is blocked up, I might go to 1/30 or 1/15th, if thin then 1/100th. That's with a Nikon 850 and 100mm Tokina macro on a Kaiser copy stand with Negative Solution carriers and using Negative Lab Pro in Lightroom to convert. 

I think you are right. I did try again last night. Still with Nikon D810+Micro 60mm, and the ES2 holder. The proper exposure seems to be around f/11, 1/3 and ISO100.

I think the diffuser coming with the ES2 blocks lots of light.

[charlesphoto99]

10 minutes ago, Warton said:

I think you are right. I did try again last night. Still with Nikon D810+Micro 60mm, and the ES2 holder. The proper exposure seems to be around f/11, 1/3 and ISO100.

I think the diffuser coming with the ES2 blocks lots of light.

I would look at Cinestill in the US for various light box options. Not only is the strength important, but also the accuracy (CRI) esp if doing color. 

[charlesphoto99]

14 minutes ago, Warton said:

Does this also assume the film was exposed and developed properly in the first place?

After I scan and invert the negative, when I see it's too bright. How do I know if it's because of the negative being scanned underexposed or film being overexposed?

 

Just look at the un-inverted photo in camera after you take it. Are the highlights blocked up? It won't be any different than shooting a properly exposed photo. Thing is, if you over or underexposed the original, then something (i.e. shadow or highlight) has to give up in the end, esp with color where post dodging and burning will introduce color shifts. B&w is a bit easier to save a poorly exposed image. 

[250swb]

39 minutes ago, Warton said:

How do I know if it's because of the negative being scanned underexposed or film being overexposed?

 

From the histogram, if not in-camera then the histogram in ACR. If the highlights on the negative are completely blocked through over exposure you can often guess by looking at the negative, but most modern DSLR's can bracket shots automatically make three exposures and choose the one where there is no clipping, and if there is still clipping you've over exposed the neg, over developed it, or both.

But I think you are coming at this from the wrong direction, that of having made a small mistake with the film, so don't try to over-compensate or over-think your workflow based on a mistake, base it on simple principles and let your DSLR deal with exposure most of the time. 

[Advertisement]

Hallo,

Look here: Interesting Leica gear

[Warton]

41 minutes ago, charlesphoto99 said:

I would look at Cinestill in the US for various light box options. Not only is the strength important, but also the accuracy (CRI) esp if doing color. 

I have been looking at the cinestill different lighting sources for a while. There is a cheap version like this https://cinestillfilm.com/collections/scanning-finishing/products/cs-litecamera-scanning-light-source-1

or something fancier like this https://cinestillfilm.com/collections/scanning-finishing/products/sunray-copybox-ii-scanning-film-workflow-set?variant=32145978720330

Other than the boxes/holders, what's the difference? One has CRI 95, the other is CRI+. Any brightness or color warmness difference?

I don't really need a fancy bamboo box or negative holder, since I have ES2 holding the negatives in front of the Micro 60mm lens nicely and snuggly.

 

[Warton]

29 minutes ago, 250swb said:

From the histogram, if not in-camera then the histogram in ACR. If the highlights on the negative are completely blocked through over exposure you can often guess by looking at the negative, but most modern DSLR's can bracket shots automatically make three exposures and choose the one where there is no clipping, and if there is still clipping you've over exposed the neg, over developed it, or both.

But I think you are coming at this from the wrong direction, that of having made a small mistake with the film, so don't try to over-compensate or over-think your workflow based on a mistake, base it on simple principles and let your DSLR deal with exposure most of the time. 

I just tried again this time let camera decide shutter so I set aperture priority with ISO 100 and f/16. The shutter goes to 1/60 but I found histogram is a bit right skewed meaning the dark is a bit clipped so I manually adjust the shutter to 1/50 and histogram looks perfectly balanced.

This is the picture. Without any adjustment other than inverted with C1.

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

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