I’m comparing a scan using a Canon R6ii, Canon 100mm macro lens, ISO 100, and a Negative Supply “Light Source Basic 4x5” for the backlight vs a scan using an OpticFilm 8200i and Silverfast. The image from the camera has more noise.
Noise can be reduced in LR, but I’d like to get even cleaner images. The exposure time for this was 1/5th second.The only improvement I can think of is a brighter light source for a shorter exposure, wondering if that would help?
The OpticFilm scan is on the left, R6 on the right, at 200% view in LR. The noise is most easily seen in the white panels of the building.
What you’re seeing here looks like a solid, typical scan. With the R6ii/100mm combo you’re pulling out more detail from the film, right down to the dye clouds / grain, while the OpticFilm is resolving less and blending some of that together.
To help further
- Could you share the full image along with this crop?
- What were all the R6ii settings?
- What was your camera, lens, and film emulsion type when you made the photo?
- I suspect this section of cropped image might be background rather than the main focus, but seeing both the whole frame and the crop will make that clearer.
- Posting the unedited file can also help with troubleshooting.
Hope that helps!
Here’s the full image (cropped from the display) to show where that cropped section is located, and its size. The section is indeed a very small area but I want to make sure the scan is set up well before doing a lot of work!
The camera was set to f/8, ISO 100, autofocus using the full frame and the camera picked the aircraft to focus on. The antenna tower is across the field in the background but it seemed a better object to use for comparing detail, the jet is too dark.
The original camera was a Minolta Maxxum 7000 with a mid-priced Tamron zoom. The film is Kodak and on the neg it’s marked “CA 100” and was most likely Kodak Gold.
I think if the scan is considered to be good then I would just use always noise reduction in LR. I am pleased the R6 version has better color detail in the field and the trees.
IMHO that looks pretty solid to me. The “noise” you’re seeing is just the film’s dye cloud/grain, which some scanners blur but your R6ii + 100mm is resolving clearly. Honestly, you’ve probably already pulled all the detail the film has to give. If you like the results, I see no reason not to get scanning.
Focusing: At 1:1 macro, AF will often say you’re in focus, but if you zoom in 5-10x and check manually you can often get it sharper. That’s been true for me on both the R5 and S1R. At f/8 you’re probably fine, but it’s worth double-checking for critical focus.
You may well find you can get more out of some shots vs the camera’s AF.
Noise: Try adjusting both the color and luminance noise sliders on your R6ii scans. You’ll see how much of what you’re calling noise is really just the film.
Ultimately, it depends on output: on screens it won’t be visible, and in print you’d need to go very large before it shows.
Vignette: This likely comes from the film-taking lens. You can test by pulling back 10-15% to avoid the corners of the lens at all and cropping in afterward. If that removes the vignette without losing meaningful detail, then the vignette’s from the 100mm setup, not the film itself.
You can also test if it is the NS 4x5 light by taking widely bracketed exposures of just the light and then playing with the exposure slider to see if the light is even across the whole light.
Lens/aperture/brightness/sharpness: Are you exposing to the right (of the histogram) without overexposing anything?
To balance the shutter speed and corner sharpness, try a range from f/4-f/8 while adjusting shutter. At 1:1, many macro lenses use internal extension, so f/4 behaves more like f/8 at infinity, and f/8 behaves closer to f/16. Exact results vary: some lenses are sharp in the corners closer to wide open, others don’t really clean up until f/5.6 or higher. Rule of thumb at infinity is often f8 so people use that for this too, but its worth testing to be certain.
Which 100mm are you using—RF, EF, or EF L?
My EF 100mm L was fine around f/5.6, though I rarely shot it at 1:1. I usually backed up a bit, captured more of the rebate, and cropped later. This avoided some/most of the vignette and softer corners on the macro lens produced while giving me more breathing room on shutter speed. Turning 44MP into 33MP was not a big deal for my use cases. So 24 to 20 is similar. 6000 vs 5450 on the long edge.
1 Like
Thanks for all those notes. It will take awhile to work out each of them, either as part of the basic setup or on each scan. I do expose to the right, and use FastRawViewer to check the histogram.
The lens is the EF-L. I had just expected f/8 would be the best tradeoff between exposure speed and depth of field for curved negatives but I’ll try f/5.6 and f/4.
This is a low res JPEG impossible to ascertain what the image quality is, and the sample of the crops above it do not strike me as particularly satisfactory if these are views at 100% or less. Both sides are fuzzy. There can be only two conditions worth talking about: (1) the scanning left a lot on the table, which can be discerned from the poor sharpness of the dye clouds (a.k.a “grain”) and/or (2) fuzzy photographs (which cannot be fully rescued). There is clearly evidence of (1) here, therefore can’t be sure about (2), but there could be some of that as well.
The first side by side comparison is clipped from the display, a 200% view in Lightroom. The tower is towards the far end of depth of field and a bit out of focus since it was the aircraft, at half the distance that was what I focused on. For that comparison though, it was the higher “noise” visible on the red/white building I was wondering about. The complete image is also a screen clip to show the size of the tower vs the total image.
To compare detail, I’ll find the negatives from a trip to Arizona. Cactus and their spines should be good for that purpose.
No wonder. To begin with, a 200% screen view includes pixelation. For these purposes one should never magnify greater than 100%. Now if part of the photo is out of focus it becomes impossible to distinguish quality problems between the photo itself and the scanning.
If you want to test whether your scanning set-up is making sharp captures it is best to choose a negative with very little detail in the photo itself, for example a photo with lots of sky. Scan it at the highest resolution your equipment allows, focused as well as you can, then magnify it to 100% and see whether the grain in the sky is sharp. Then turn the negative around 180 degrees and rescan it to get the sky in the other part of the frame and do likewise. That way you can tell approximately whether your scanning set-up is parallel between the media and the sensor. Alternatively, use a resolution target.
Now, reverting to your original post, noise and grain are completely different things. Noise is the result of electronic imperfections relative to sensors in digital cameras, whereas the film grain you see from colour negatives are the dye clouds remaining after development - those dye clouds are the image. You cannot compare them - they have totally different physical characteristics. You can make subjective decisions about which kind of interference bothers you more, but that’s a different matter. It has no implications for objective system performance quality comparison. Also be aware that a Plustek 8200 scanner is not the be all and end all of sharpness in scanners. It’s not bad, but a Minolta Scan Elite 5400, for example, beats it by country miles. For your camera, the kind of macro lens you are using also makes a difference. A flat-field macro copy lens made for photographing flat media at close range is the ideal. I don’t know whether what you are using is that kind of lens. Then there is the size of your sensor - is it full-frame, APSC, how many pixels high and wide? That and the lens determine the maximum resolution you can pull from the camera. In any event, either system you compared should give better results than what you showed, but I come back to the main point - let’s set up the testing so you can legitimately be able to distinguish the sharpness of the film “grain” from that of the photo itself.
@Andy1024, Mark really knows what he is talking about. Pay attention there!
If the negatives aren’t sharp it would be hard to judge anything fairly until they are FLAT (edited “flat” in later, typo!). Also, on the topic of:
These can be helped with more DoF but that is NOT the solve, it is more of a temporary bandaid. Keeping the negatives as flat as possible with a good negative holder will yield much better results.
Keep working at it and good luck. Always ask more questions!
SSelvidge - thanks.
Correct, and I would only add that Museum glass or Schott Anti-Newton glass could also do the job, but I agree with you - I’d first try a good negative holder. For example, I’m getting very good results using the 35mm strip carrier from my old Minolta Scan-Elite 5400 - but my negs don’t have much curl at all. For larger format negs, if they don’t lie completely flat, it’s the glass.
The problem with checking focus, is needing a negative that has excellent detail in the corners. This would mean a very good lens would have been needed 20 to 30 years ago or a modern test strip has to be bought but those are expensive.
It does seem that just finding a negative with relatively clear sky and checking the dye clouds is the simplest. The dye clouds are in a sense, perfectly focused across the entire negative because they were formed in the dye layer itself. I haven’t had time to search for some good negatives yet.
I use Negative Supply ‘s Basic Film Carrier 35 Mk2 which is promoted as being very good at holding negatives flat. If I find both a reasonably flat and a curved neg strip of good sky, that would be interesting to compare.
The sky over Kitt Peak, AZ was good for the dye clouds test at 100%. They’re also similarly defined near the corners, so focus, neg flatness, and platform levelling appear good.
For flattening curved negative strips, I can use a micrometer down into the well of the NS film holder and measure which does a better job, that or the film strip holder for the Plustek with its inter-frame crossbars, and see what to do about that.
Thanks for all the information on focussing, dye clouds, and depth of field.
