I am currently working on a personal series of photographs using Portra 400 (35mm) and although I like the result of the images I find however that the grain is very apparent. What do you think about this?
I use a tungsten lighting, and I expose for the face (the background has according to the zones from 1 to almost 2 f-stops of difference at the bottom of the image)
Concerning the scan, I use the Reflecta RPS 10M with the Silverfast software which allows me to scan 48bit raw of my negatives that I then convert via NLP (version 2.3)
Thanks in advance to all those who can help me
It looks a little overcooked, was it maybe under/overexposed?
I don’t think so even if I’m not 100% sure. I used a spot meter to measure the highlights and depending on the image area there may be one or two f-stops difference, especially on the bottom of the image (full)
Could it be due to the scan? I scan at 5000ppp on the Reflecta rps 10m
To my eye that looks as if it’s either an underexposed negative, or that particular scanner is just noisy, particularly in shadow areas.
Have you examined the density of the negative on a light table to see if there are any exposure issues?
If there are no issues there, try getting a decent pro scan done and see if there’s a difference, in which case it would be the scanner.
You say you spot metered the highlights - did you apply compensation to that reading?
Hi ! I didn’t try checking the density of the negative, I didn’t know this technique but I’ll check some youtube video about this topic then verify negatives in a few days (I am currently abroad)
The scan lab comparison is also a good idea.
Thanks for the help
Hi ! I didn’t apply any compensation although I am not sure what it exactly means ?
I use a three point set up, a key light for the subject, one for the background and one for the hair (backlight)
I took a measure to correctly expose the face of the model and that’s it (the difference between key light and background light is 1 to 2 EV depending on the area of the background)
With respect, I suspect your knowledge of metering is shaky. The EV difference in the background is only relevant for modelling but is irrelevant to correct exposure for the subject.
If you took a spot reading from a true highlight, your film is 2.5 EV underexposed.
If you took a spot reading from the face, your film is about 1 EV underexposed.
Every meter is calibrated to middle grey and will indicate an exposure corresponding to middle grey no matter what you meter.
If you want a correct exposure you need to either meter off a mid-grey tone or meter off a white (highlight) area and add 2.5 EV extra exposure - compensate by adding 2.5 stops exposure.
Portra is a relatively fast (grainy) film stock so:
If you underexposed your film, the film will lose detail in the shadows and film grain will become more pronounced.
If you scan such a film and try to recover detail in the shadows (nearly impossible) by increasing exposure, film grain will be even more pronounced and you will further add digital noise.
Colour film like Portra is much more tolerant of overexposure than underexposure so if there is a doubt it is safer to overexpose than underexpose colour film.
Hope that helps.
Hi Belinda, thank you for your answer,
I realize that I made a big mistake in talking about spotmeters, in fact I used a lightmeter (Sekonic L308), sorry for that, I did not really make a difference between the two terms but they indeed refer to two types of devices that measure light differently.
So I took a measurement with a lightmeter, in incident light, from the face and I didn’t try to recover the shadows (bottom of the image) in post processing, the fact that they lose details didn’t bother me.
With the same configuration, I did a previous shoot where I overexposed by 1 f-stop, and I must admit that the images have less grain but in this case I had more hot spots on the model’s face and I had to compensate the exposure in post processing in a more important way to get a satisfactory exposure.
I think I’ll have to do a test session with different exposure values to determine what I like the most
Thank you for your help
Check this out for grain vs. exposure:
" …Overexposure doesn’t tend to cause heavy grain, but underexposure does. Whereas with the former, you mostly lose highlight details, with the latter you lose shadow detail, or even detail across the entire frame if underexposed severely.
Because the smaller light sensitive silver crystals are mostly left unexposed, with the bigger ones getting most of the exposure, the resulting image will be more grainy. But on top of that, your scanner will attempt to get more detail out of the shadows, resulting in increased digital noise on top of the grain, which is why the end result is poor."
Hi Digitizer, thank you for the link,
I am aware of the consequence of underexposure regarding the grain, but I don’t expect it to be the case in my situation since I used a light meter and applied the same settings to my camera. I didn’t try also to recover details from shadows.
I don’t currently have access to my negatives but I have this screenshot (maybe someone can tell if there is a density issue here ?)
Apologies for the lecture on spot metering if you already understand exposure principles.
It’s quite difficult to get a wildly incorrect exposure with incident metering so it must be something else. If you film was reasonably well exposed, well developed and well scanned, you shouldn’t be seeing excessive grain.
The suggestion to compare your scan with a lab scan is a good one as will get a scanning error out of the way.
There is just one more thing that occurs to me. Are you sure you had the correct ISO set on both the meter and/or the camera and that the camera wasn’t inadvertently set to Auto?
No worries, it allowed me to learn something about spot metering. I should have used the right words to express myself, it usually works better that way
I will indeed do the test with a scan in the lab. I also have thanks to a post on the NLP facebook account the suspicion that it could be a problem with the scan, several users of this scan (RPS 10M) seem dissatisfied, including one with a grain problem quite similar in my opinion.
For ISO I know I probably made this mistake once (see photo below, with a clear density problem on the negative) but that’s the only time. The grain problem I’m talking about appears all the time on 5 shoots where I use 2 rolls of film each time
I only suggested ISO because it’s the sort of simple mistake we have all made (in my case, more than once!) and which can seem baffling at the time.
Good luck with you scanner investigations.
In reply to the original post.
You heard the Irish joke about giving directions? If I were you I would not start from here.
Your picture falls into that scenario. You are asking a lot to expose correctly far more that your technology can cope with. Rephotograph the model and use a reflector or other source of lighting and reduce the shadow under her chin
Nice joke I don’t really mind the grain under the chin, I understand why it’s here and how to fix it, but what bothers me is the overall graininess of the image. But maybe you think that the grain is normal for the rest of the image ? (Portra 400)
My answer to your question would be “yes”. 400 iso col neg is fairly grainy, I’m not sure that the example you show is particularly unusual. Have you tried a similar shoot, same everything but with a 100 iso film?
And, in a studio setup, why use 400 iso at all?
Fair question, I started with portra 160 and shot some B&W fomapan 100 iso but I heard so much good thing about portra 400 that I wanted to try it. Both for skin tone rendition and the versatility of shooting a 400 iso. I felt limited at slower speed but I might reconsider that. I am currently traveling but I’ll try tomorrow to post photographs at 160 and 100 iso (the setup was similar) to compare. If I recall well even the 100 iso was grainy
100% Crop of Portra 160 (same setup as previous photos)
100 % Crop of Fomapan 100 (same setup)
To my eye some of that looks like digital noise and not film grain. Do you have the possibility to just pay for one drum scan by a known good operator of one neg to just rule out the scanner being the problem?