Troubleshooting Grain/Clarity

Hi There!

Just scanned 120mm film with NLP for the first time using a Sony A7III, Nikkor 55mm 2.8, Lobster Holder and Cinestill backlight.

I’m finding a pretty significant amount of grain considering I shot at 100iso on my Sony. Wondering if anyone has any ideas? The result feel exciting and decent for my first time, but not yet up to the level that I expect for the quality of sensor.

Also, I shot 1:2 instead of 1:1 because I found a 28mm tube was too zoomed in for me to be able to fit a frame into view… should I be taking 2 photos and stitching them together..? Feels daunting,

Any advice is much appreciated! Pictures here:

Hi Ezra,

Welcome to the NLP community! You have some great photos. Thank you for sharing with us and asking for feedback. My honest first impression is that your images actually look quite nice with an acceptable/expected amount of grain appearance for film photography. ISO 100 on a Sony A7 III with good exposure is going to be super clean, so the question is what filmstock are you using (Tri-X, Delta 100, HP5, etc.)? It could just be film grain depending on your stock. Something else to consider is how much extra space is around your negatives when scanning at 1:2? Cropping an image from a digital sensor magnifies the appearance of digital noise. Thanks for the post, Ezra!

Thank you for the kind note! I shot Tri-X! I think you’re right.. I may be noticing more of a focus quality issue than a grain issue.

There was definitely some extra space around my negs, however I found that a 28mm tube was too large so maybe I need to use 12mm?

1 Like

You might find info about imaging ratios in your lens’s technical data or user manual. Search the Internet, if you don’t have the original booklet.

The 55mm Micro-Nikkor focuses to 1:2 without an extension tube, so that means that it will encompass a 72mm x 48mm piece of film or larger on to Full Frame without an extension tube. Since ‘6x6’ 120 film is 56mm x 56mm you will be fine capturing it with no margins top & bottom, just wind the focus on the lens back a bit from full extension.

Your A7III is 24MP so that means the resulting ‘scan’ from 6x6 will be a maximum of 4000 x 4000 pixels. To get the same pixel size using an actual medium format scanner you’d need to set it to scan at 1800 ppi.

That might be fine for your requirements, say a 13” square print, but if you want to make bigger prints or get more detail from your negs then you will need to stitch as you say. That’s a bit of a performance so maybe just save that for particular negatives. The easiest way would be to match the long side of your sensor with the 6x6 frame and so get 6000 x 6000 pixels from 2 frames (just), a 50% increase in print size if you like or equivalent to 2700 ppi from a film scanner.

Hi Ezra,

I second piphlenitricity’s assessment. I really like your photos, and I also think it’s just film grain. To me, that’s the whole point: I want to see the grain when it’s an analog photo. Otherwise, I’d just shoot digital, where the result is technically much “better” / “cleaner” but also more “sterile”.
What was the developer you used? Thanks!

2 Likes

These are good, clean scans and considering it’s Tri-X, the grain is well controlled and normal.

The amount of resolution you need depends on the end-use you intend for these scans. Print would probably be the most demanding use, and large prints more so. If you would be using an Epson printer for example, the printer normally wants an input resolution of 360 PPI. Without resampling, as your photos look to be square, as mentioned here your maximum resolution is 4000 PPI on each dimension, which allows you linear dimensions of 4000/360 = a print of 11.1 x 11.1 inches.

But that said, these days resampling algorithms are pretty darn good at inventing information such that retained quality would be fine. You can easily expand the linear dimensions so that input PPI gets diluted to say 240 PPI and let the printing system upgrade it under the hood. In this case you can print to 4000/240= 16.7 x 16.7 inches. That would be neatly compatible with a 17” printer.

If you want to print larger than this, and you don’t want to stretch the resampling beyond what I’m discussing here, then indeed you would increase the magnification at the scan time such that you photograph the negative in sections, which produces more sensor pixels over the same image area and stitch the sections in post-capture processing. As long as you provide for overlap when making the captures, the stitching is easily accomplished in Lightroom using PhotoMerge>Panorama. I do it often and have never had issues with it. The quality of the result as usual depends mostly on the capability of your lens. The camera sensors in these models are fine.