Anyone using Capture One for converting film negatives?

I've recently started shooting film again and scanning and converting my negatives at home. For scanning I am using my Sony a7iii with the Valoi Easy35 which seems to be a decent solution (apart from the pain of keeping it clean of dust).

For converting the RAW files I've compared Capture One, which is my default RAW converter, to Filmlab and Filmomat SmartConvert. I found that you can achieve great results with Capture One, sometimes even better than with the dedicated apps. But with most files it took me way longer to get a good result with Capture One than with the other two apps.

For those of you who are using C1 for converting negatives: What is your exact workflow?

I basically proceed like this:

1. Pick white balance from film base
2. Crop
3. Set curve to "Linear response"
4. Invert colors using the Levels tool: Set black point to 255, white point to 0
5. Auto-adjust levels (each color individually)
6. Adjust mid-point in Levels
7. Fine-tune RGB curve with Curve tool

What I find particulary difficult is correcting colors manually if auto-adjusting levels doesn't result in good colors. Because if you shift white balance after doing auto levels your auto levels aren't in the right place anymore. My best solution often was to use the Color Balance tool in this case, but that was always time-consuming.

Here's an example of an image where I got the best result with Capture One. Usually, negatives processed with Filmlab and SmartConvert have a more "film-like" look than the ones processed with C1, which is also visible here to some degree. However the skin looks a bit too pale for my taste in the Filmlab and Filmomat images.


Left: Capture One; center: Filmlab; right: SmartConvert

I haven't decided yet if I should purchase either Filmlab or Filmomat SmartConvert. Overall, I slightly prefer Filmlab, but 200 $ is a steep price for such a simple app IMO.

What I like about FilmLab is the ability to take a set of edits, copy them from one exposure and paste them to the whole roll of 36 if you want to.

Usually a paste like that doesn't work unless all 36 shots were of the exact same subject in the same light with the same shutter and f-stop but it was a selling point when I made the purchase.

I agree, your FilmLab example the skin could be rendered better.

grsnovi said:

What I like about FilmLab is the ability to take a set of edits, copy them from one exposure and paste them to the whole roll of 36 if you want to.

Usually a paste like that doesn't work unless all 36 shots were of the exact same subject in the same light with the same shutter and f-stop but it was a selling point when I made the purchase.

Click to expand...

Yes, that sounds good - in theory at least. I am not yet experienced enough but it does seem to me that, like you say, every shot needs its own adjustments - unless taken under the same conditions.

I read one of your older posts about negative conversion. I remember you saying you bought Filmomat Smartconvert? So did you switch to Filmlab?

For me, the one crucial point where Filmlab has an advantage over Filmomat ist sharpness. I've noticed that the images produced by Capture One are the sharpest, followed by Filmlab, then Filmomat. In the case of Filmlab I think that the reduced sharpness compared to C1 is actually pleasing in most cases because the film grain then looks more organic. Filmomat on the other hand I find not to be sharp enough in many cases - especially in those cases where I hadn't exactly nailed focus with my manual focus Nikon FE2.

Easy Lee said:

I read one of your older posts about negative conversion. I remember you saying you bought Filmomat Smartconvert? So did you switch to Filmlab?

Click to expand...

Wasn't one of my posts - I have never used or owned Filmomat.

SmartConvert does an admirable job for me but the thing I dislike about it is the lack of control for highlight retention. I have tried doing conversions in CaptureOne and I just can't get good results. I was hopeful it would be simply in C1 but really its not, it takes a good amount of work to get it to look even ok. Below are some SmartConvert outputs for me.

Easy Lee said:

For me, the one crucial point where Filmlab has an advantage over Filmomat ist sharpness. I've noticed that the images produced by Capture One are the sharpest, followed by Filmlab, then Filmomat. In the case of Filmlab I think that the reduced sharpness compared to C1 is actually pleasing in most cases because the film grain then looks more organic. Filmomat on the other hand I find not to be sharp enough in many cases - especially in those cases where I hadn't exactly nailed focus with my manual focus Nikon FE2.

Click to expand...

I find it strange that there should be a difference in sharpness between different inversion software - surely the sharpness is defined by the initial image capture ? Is Capture One not just applying some software sharpness filter to result in a sharper image ?

HaroldC3 said:

SmartConvert does an admirable job for me but the thing I dislike about it is the lack of control for highlight retention. I have tried doing conversions in CaptureOne and I just can't get good results. I was hopeful it would be simply in C1 but really its not, it takes a good amount of work to get it to look even ok. Below are some SmartConvert outputs for me.

Click to expand...

Great colors! Is this medium format? Did you never think SmartConvert was lacking a bit in sharpness? (These images seem quite sharp for film negatives scanned at 40+ MP, but I wonder what results you'd get with C1 or Filmlab.)

Yes, Filmlab does have a highlight control but honestly I don't find it extremely useful. I've found that many images from SmartConvert and Filmlab need some additional tweaking anyway, especially some more contrast by setting black and white points. So I just make sure the highlights look good and then maybe boost the midtones a bit in Affinity Photo.

What problems did you encounter with Capture One? There is definitely a steeper learning curve than with Filmlab and SmartConvert but by now I am pretty happy with most outputs. However, there are some "problematic" negatives which I just can't get to look good with C1.

In the last few days I've found a little trick that makes most negative conversions in C1 look better. Between steps 6 and 7 in my workflow (see above) I use the Color Balance tool to shift the colors a bit towards orange - sometimes all tones, sometimes only midtones and highlights. With this method, colors turn out more like the actual prints from the lab.

Overrank said:

Easy Lee said:

For me, the one crucial point where Filmlab has an advantage over Filmomat ist sharpness. I've noticed that the images produced by Capture One are the sharpest, followed by Filmlab, then Filmomat. In the case of Filmlab I think that the reduced sharpness compared to C1 is actually pleasing in most cases because the film grain then looks more organic. Filmomat on the other hand I find not to be sharp enough in many cases - especially in those cases where I hadn't exactly nailed focus with my manual focus Nikon FE2.

Click to expand...

I find it strange that there should be a difference in sharpness between different inversion software - surely the sharpness is defined by the initial image capture ? Is Capture One not just applying some software sharpness filter to result in a sharper image ?

Click to expand...

I don't think it's that strange. There is some sharpening involved in every RAW conversion. In C1 you have complete control over the sharpening process, but obviously there are limits as to how sharp you can make a 35mm film scan. I found between 100 and 120 to be a good value to make the images look as sharp as possible without making it look unnatural. (Between 120 and 180 would be "normal" values for RAW files from digital cameras.)

In Filmlab you can also control sharpness, but the difference between the lowest value 0 and the highest value 10 is not that big. I usually set it to 6. In SmartConvert unfortunately there is no way to control sharpness and the outputs are about as sharp as if I set sharpness in Filmlab to zero.

You can see the difference well in these 100% crops from one of my test shots. The C1 and Filmlab outputs are about equally sharp but the SmartConvert output is definitely less sharp. (The watermark in SmartConvert is because it's a demo version.)


Capture One crop 100%


Filmlab crop 100%


Filmomat SmartConvert crop 100%

Easy Lee said:

Overrank said:

Easy Lee said:

For me, the one crucial point where Filmlab has an advantage over Filmomat ist sharpness. I've noticed that the images produced by Capture One are the sharpest, followed by Filmlab, then Filmomat. In the case of Filmlab I think that the reduced sharpness compared to C1 is actually pleasing in most cases because the film grain then looks more organic. Filmomat on the other hand I find not to be sharp enough in many cases - especially in those cases where I hadn't exactly nailed focus with my manual focus Nikon FE2.

Click to expand...

I find it strange that there should be a difference in sharpness between different inversion software - surely the sharpness is defined by the initial image capture ? Is Capture One not just applying some software sharpness filter to result in a sharper image ?

Click to expand...

I don't think it's that strange. There is some sharpening involved in every RAW conversion. In C1 you have complete control over the sharpening process, but obviously there are limits as to how sharp you can make a 35mm film scan. I found between 100 and 120 to be a good value to make the images look as sharp as possible without making it look unnatural. (Between 120 and 180 would be "normal" values for RAW files from digital cameras.)

In Filmlab you can also control sharpness, but the difference between the lowest value 0 and the highest value 10 is not that big. I usually set it to 6. In SmartConvert unfortunately there is no way to control sharpness and the outputs are about as sharp as if I set sharpness in Filmlab to zero.

Click to expand...

I always scan (Nikon CoolScan) with the sharpening control set as low as it can be, and then store the unsharpened version. I will then apply sharpening to any version that intend to print / publish etc since quantity of sharpening needed will depend on the end use and image size. For example the CoolScan images are around 20MP - if I was to share one via social media etc (so an end size of a couple of MP at best) I wouldn’t apply any sharpening other than reducing the size of the image, whereas if I was going to print the image full size I would apply some slight sharpening.

I have a licence to FilmLab which I use occasionally for images that SilverFast inverts poorly. I’ve found it particularly good for maskless films like Harman Phoenix.

[ATTACH]3713604[/ATTACH] [ATTACH]3713605[/ATTACH] Easy Lee said:

HaroldC3 said:

SmartConvert does an admirable job for me but the thing I dislike about it is the lack of control for highlight retention. I have tried doing conversions in CaptureOne and I just can't get good results. I was hopeful it would be simply in C1 but really its not, it takes a good amount of work to get it to look even ok. Below are some SmartConvert outputs for me.

Click to expand...

Great colors! Is this medium format? Did you never think SmartConvert was lacking a bit in sharpness? (These images seem quite sharp for film negatives scanned at 40+ MP, but I wonder what results you'd get with C1 or Filmlab.)

Click to expand...

Yes, I have shot some 6x6 and the above is 6x9. My original examples were me trying to figure out how to scan them so the sharpness isn't great. The above is where the lab scanned them using a Z7 II and sent me the RAWs. The top one is from SmartConvert and the bottom one is from CaptureOne. If I'm being honest, I would normally then take it into Photoshop CS6 and do some final tweaks but I wanted to see how close I could get with C1. Also, I have some filmstyles that I like to use as well in C1.

Looking at the two at 100% I do see the differences in sharpness. I did apply just a little sharpening in C1.

Easy Lee said:

Yes, Filmlab does have a highlight control but honestly I don't find it extremely useful. I've found that many images from SmartConvert and Filmlab need some additional tweaking anyway, especially some more contrast by setting black and white points. So I just make sure the highlights look good and then maybe boost the midtones a bit in Affinity Photo.

What problems did you encounter with Capture One? There is definitely a steeper learning curve than with Filmlab and SmartConvert but by now I am pretty happy with most outputs. However, there are some "problematic" negatives which I just can't get to look good with C1.

In the last few days I've found a little trick that makes most negative conversions in C1 look better. Between steps 6 and 7 in my workflow (see above) I use the Color Balance tool to shift the colors a bit towards orange - sometimes all tones, sometimes only midtones and highlights. With this method, colors turn out more like the actual prints from the lab.

Click to expand...

I think my problem is simply getting the color to look good. IMO it just takes a lot of effort and that's why I thought SmartConvert would work well but I'm not thrilled with it.

Overrank said:

I always scan (Nikon CoolScan) with the sharpening control set as low as it can be, and then store the unsharpened version. I will then apply sharpening to any version that intend to print / publish etc since quantity of sharpening needed will depend on the end use and image size. For example the CoolScan images are around 20MP - if I was to share one via social media etc (so an end size of a couple of MP at best) I wouldn’t apply any sharpening other than reducing the size of the image, whereas if I was going to print the image full size I would apply some slight sharpening.

I have a licence to FilmLab which I use occasionally for images that SilverFast inverts poorly. I’ve found it particularly good for maskless films like Harman Phoenix.

Click to expand...

I think you can't quite compare the sharpening process of a negative scanner and a RAW converter. With a RAW converter (C1, but in this case also Filmlab) there needs to be some sharpening to make up for the loss of sharpness that happens inside the camera. Nobody would store their RAW conversions at zero sharpness with Capture One.

I did try to apply some additional sharpening to the outputs from SmartConvert, to see if I could get the same sharpness as with C1 and Filmlab if needed. But for some reason it didn't work well.

HaroldC3 said:

Yes, I have shot some 6x6 and the above is 6x9. My original examples were me trying to figure out how to scan them so the sharpness isn't great. The above is where the lab scanned them using a Z7 II and sent me the RAWs. The top one is from SmartConvert and the bottom one is from CaptureOne. If I'm being honest, I would normally then take it into Photoshop CS6 and do some final tweaks but I wanted to see how close I could get with C1. Also, I have some filmstyles that I like to use as well in C1.

Looking at the two at 100% I do see the differences in sharpness. I did apply just a little sharpening in C1.

Click to expand...

Yes, the lower image is slightly sharper, but the top one is also plenty sharp. I think probably sharpness is far less of an issue with medium format negatives. In this case I would be prefectly happy with the output from SmartConvert. Colors / WB look a bit more pleasing to me in the upper image. It is just lacking a bit in the shadows.

HaroldC3 said:

I think my problem is simply getting the color to look good. IMO it just takes a lot of effort and that's why I thought SmartConvert would work well but I'm not thrilled with it.

Click to expand...

I agree, color balance is the difficulty in C1. How are you adjusting colors if they seem off? By shifting the mid points in the levels tool? Or with the color balance tool?

Are you using the "linear response" curve? Are you adjusting the tone curve with the HDR tool or with the curve tool?

Easy Lee said:

Overrank said:

I always scan (Nikon CoolScan) with the sharpening control set as low as it can be, and then store the unsharpened version. I will then apply sharpening to any version that intend to print / publish etc since quantity of sharpening needed will depend on the end use and image size. For example the CoolScan images are around 20MP - if I was to share one via social media etc (so an end size of a couple of MP at best) I wouldn’t apply any sharpening other than reducing the size of the image, whereas if I was going to print the image full size I would apply some slight sharpening.

I have a licence to FilmLab which I use occasionally for images that SilverFast inverts poorly. I’ve found it particularly good for maskless films like Harman Phoenix.

Click to expand...

I think you can't quite compare the sharpening process of a negative scanner and a RAW converter. With a RAW converter (C1, but in this case also Filmlab) there needs to be some sharpening to make up for the loss of sharpness that happens inside the camera. Nobody would store their RAW conversions at zero sharpness with Capture One.

I did try to apply some additional sharpening to the outputs from SmartConvert, to see if I could get the same sharpness as with C1 and Filmlab if needed. But for some reason it didn't work well.

Click to expand...

We may be typing at cross purposes - of course I’m always interested in the scan itself being as optically “sharp” as possible ( by which I’m meaning the maximum measurable resolution - which may not be what you think it is - see https://www.dpreview.com/forums/thread/4547755 ), however I don’t then apply any additional sharpening through software until I am actually going to use the image. I think what you’re talking about is this software sharpening, which is likely some variation of applying a 3x3 convolution kernel matrix to the image, I.e. a bit of maths. Given that it’s just a software element I can apply that anywhere in the toolchain, which I choose to do at the end.

What I’ve found interesting is to subtract one image from another - e.g. in your case one converted by Filmlab and one by SmartConvert and see what the residual image is - is it just a chroma / contrast difference or is there additional sharpening / softening of the image. I’ve used this in the past to visualise the supposed scanner noise (practically non-existent).