OregonSun
Online
Upload & Sell: Off
|
 Re: Post your recent film shots! | |
Cloud75 wrote:
fjablo wrote:
Cloud75 wrote:
Desmolicious wrote:
Cloud75 wrote:
fjablo wrote:
Cloud75 wrote:
mamiya 6 + 50mm f/4 + ektar 100, pointed away from the big things in Yosemite everyone knows
 top of sentinel dome by Chris, on Flickr
Beautiful picture with a lot more potential if converted differently. It currently has a cyan cast and overall missing contrast. Did you convert manually? If so I can highly recommend this guide: https://www.alexburkephoto.com/blog/2019/10/16/manual-inversion-of-color-negative-film
If you don’t want to convert manually and also don’t want to pay for NLP or Smartconvert, try NegPy. It’s hosted on GitHub and is an open source film conversion software built with Python.
This is the kind of feedback I want. While I feel like my stuff is decent in black and white, I do not know what I'm doing with color. I scan using a cinestill cs-lite set to "cool", and a sony mirrorless with daylight white balance with manual exposure set just before zebras/highlight clipping shows up. From there, my process has shifted from darktable+negadoctor to lightroom classic + NLP. In lightroom, I set white balance with the border, invert using NLP and the Noritsu setting, and then... start flapping around with NLP white balance settings until it looks ok to my eye. Sometimes it comes out nice as-is, sometimes it looks weird and I need to randomly adjust more. I don't look at the histogram or curves. Since I really don't naturally "get" color, I think my color photos come out fine, not amazing. I'll walk through that blog post. But really, maybe step one for me is getting a color theory book, or taking a watercolor class or something. I'd love to hear what's worked for others.
Also, I work in software, and find it painful how many times I've manually done this while STILL not being able to articulate how I'd automate to a ~90% good starting point (whatever 90% even means).
I scan using a cinestill cs-lite set to "cool"
Why? By doing that you already have made sure that your images will look blue/cyan.
You need a neutral/daylight light source setting.
Button says “cool (color negative)” and I not so smart so I push the button.
Not sure if NLP can be fooled by this - easiest is try by taking two pics of the same negative 
It should *not* have an impact as long as none of your color channels is clipping. NLP works on the raw file and shifts white balance a lot anyway. The idea behind the cool light source is that you get a bit more signal to noise ratio in your green and blue color channels, which would otherwise be underexposed a lot given the orange mask.
Learned a bunch tonight. I tried all permutations of a photo with a cool backlight, a photo with a white backlight, and inversions with NLP using frontier or noritsu color models, and whether I set NLP white balance to "auto-neutral" or off (testing @Desmolicious@ comments). I skipped the extra dropper white balancing too.
What I found was that the crummy place I had originally edited from was the result of me unthinkingly inverting BEFORE I cropped to just the square negative. I'm scanning square images with a 3:2 aspect ratio camera. So, there are big areas of the negative that are black, which has to trip up NLP. This process doesn't negatively impact black and white.
Landed on this: for color negative, the cool backlight setting -> crop to the actual negative (a square here) -> invert with NLP defaulting to auto-neutral wb. This is using the "LAB - Standard" tone profile in NLP. No edits, all sliders untouched. Better?
 cool-backlight-cropped-nodropper-whitebalance-noritsu-nlp-auto-neutral by Chris, on Flickr
Looks good to me!
My philosophy with NLP is "less is more".
I've found that every choice bakes in a specific look that is hard to change or adjust later on. Most of my pics get additional minor adjustments in LR.
Another thing that really helps is to only convert pics from a single photo shoot/lighting scenario in the same batch. It helps make the conversion better, and it makes it easier to just adjust one pic and sync the settings to the others.
Here are the Setting I use:
Convert
- Basic Color Model 99% of the time, sometimes None will fix bad color shifts
- Pre-Saturation - 2
- Roll Analysis on
- Roll Process - Pro Lab
- Adjust Border Buffer so that only exposed negative is visible (be sure to exclude sprocket holes and DX edge coding)
Edit
- Settings - NLP Standard
- Tone Profile - This is the main decision I make, I pretty much always cycle through all the options for every batch I convert. Sometimes they all get the same one, other times it's different for specific photos. The ones I end up using the most are LAB - Standard, LINEAR, LINEAR - Gamma (especially for B&W), and LINEAR - Deep. The other LAB profiles are good for specific scenarios, e.g., very thin or dense negs or ones that need more/less contrast.
- WB - Leave it on Auto-Neutral unless there's an obvious issue, then adjust as needed to get it close enough. Final WB adjustment in LR if needed.
- HSL - Mostly just leave this on Lab, but sometimes use one of the others for a specific look
- Sometimes I'll use the general exposure/contrast sliders, but not always
- I never touch any of the other more fine grained options
|
