Comparing Imacon X5, Creo iQ3 and Sony A7R IV scans

[Steven Lee]

This thread is derived from this one, where I was comparing the ease of inverting color using camera and Imacon scans. Some of you have asked to see the full-sized crops to evaluate the available resolution.

WARNING: I am not a resolution aficionado. Back in my digital days I never felt the need for more than 24 bayer megapixels. I am only doing this because some of you asked me to. I did not have a controlled environment, I simply sharing my vacation images here.

Photo equipment:

• 6x6 Fuji 400H exposed at f/8 with Hasselblad 60mm Distagon mounted on a tripod.
• 35mm Portra 160 exposed at f/5.6 with Voigtlander 28mm Ultron II handheld.

Scanning equipment:

• Sony A7R IVa with Sigma 105mm macro lens operated by me.
• Imacon X5 operated by a lab.
• Creo iQ3 operated by https://www.scansolutionsonline.com/ - Michael is extremely knowledgeable and a pleasure to work with.

First let me show you how the full images look like, followed by the links to full-sized crops.

35mm Portra 160

6x6 Fuji 400H Pro

Full-sized crops to follow...

WARNING: once again, this is not a controlled test. For example, since 3 different people did the scanning, the sharpening settings are not the same. This is particularly noticeable with the Imacon scans. My camera scans used 10% sharpening in Lightroom with 1.0 radius (vs 40% Lightroom default).

 

[Steven Lee]

These are JPEGs exported from Photoshop using its "Great" quality setting, so the JPEG compression artifacts should be negligible/absent.

135:

1. Sony 100% crop.
2. Creo 100% crop.
3. Imacon 100% crop.

6x6:

1. Sony 100% crop.
2. Creo 100% crop.
3. Imacon 100% crop.

 

[NortheastPhotographic]

Ah "red", the sworn enemy of camera scanning.

 

[4season]

It looks like the software is applying heavy chroma noise reduction to the Sony scans, and it's not doing you any favors.

 

[Steven Lee]

It looks like the software is applying heavy chroma noise reduction to the Sony scans, and it's not doing you any favors.

It's not software. It's me. I use software to get what I want. I'm doing myself all the favors. Software is not weather, it's a tool.

Ah "red", the sworn enemy of camera scanning.

You may want to look at the thread I started earlier, where I evaluated colors instead of resolution. These scans are not color balanced to look the same. And, in case you're wondering, if I had the time to adjust colors, I'd be upgrading the scanner output to my Sony file.

 

[koraks]

Especially the scanned 6x6 frames seem to suffer from clipped highlights; the Sony/camera scan doesn't. This seems to be due to operator error more than anything else.
I'm surprised at the relatively mushy Creo scans overall, although the X5 6x6 scan seems excessively sharpened, which influences the overall appearance. T
There's a couple more things we could nitpick on, but personally, I take the following from your tests:
1: The Sony 'scans' are perfectly usable.
2: From 6x6 upwards, proper film scanners may have the edge, but this depends on...
3:...operator choices which influence the end result greatly.

Thanks for posting!

 

[GLS]

Are the Sony files from a single capture, or did you use pixel-shift?

 

[Helge]

This thread is derived from this one, where I was comparing the ease of inverting color using camera and Imacon scans. Some of you have asked to see the full-sized crops to evaluate the available resolution.

WARNING: I am not a resolution aficionado. Back in my digital days I never felt the need for more than 24 bayer megapixels. I am only doing this because some of you asked me to. I did not have a controlled environment, I simply sharing my vacation images here.

Photo equipment:

• 6x6 Fuji 400H exposed at f/8 with Hasselblad 60mm Distagon mounted on a tripod.
• 35mm Portra 160 exposed at f/5.6 with Voigtlander 28mm Ultron II handheld.

Scanning equipment:

• Sony A7R IVa with Sigma 105mm macro lens operated by me.
• Imacon X5 operated by a lab.
• Creo iQ3 operated by https://www.scansolutionsonline.com/ - Michael is extremely knowledgeable and a pleasure to work with.

First let me show you how the full images look like, followed by the links to full-sized crops.

35mm Portra 160

View attachment 346065

6x6 Fuji 400H Pro

View attachment 346066

Full-sized crops to follow...

WARNING: once again, this is not a controlled test. For example, since 3 different people did the scanning, the sharpening settings are not the same. This is particularly noticeable with the Imacon scans. My camera scans used 10% sharpening in Lightroom with 1.0 radius (vs 40% Lightroom default).

This is great! Thanks a lot!
Whether you are a “resolution aficionado” or not, once you reproduce at sizes over the common phone screen, ultimate resolution and the MTF curve has a big impact on psycho-optical impact of the photo, in ways that is not quickly put into words.
In much the same way that highs can affect the lower registers in audio reproduction.

It’s noticeable that the Creo has what appears to be blooming or at least general micro contrast reduction.
Is that due to the mylar and scanner liquid, or the aperture technique of scanning in general?
I appears to pick up more dirt or scratches the a the other two scans too.
General resolution appears slightly lower than the cam scan. But colour resolution is a smidgen better.
Both pretty predictable from what we theorize about the two techniques.

The Imacon appears the worst of the three, though still respectable.
I don’t know about koraks suspension of sharpening. To me it looks like plain sensor noise.

Overall the Sony scan appears overall best by a bit.
It also has better dynamic range both in the shade and the highlights.

 

[Helge]

Are the Sony files from a single capture, or did you use pixel-shift?

I’d like to know too, now that you have been to good as to supply the rest of the technical data.

 

[Steven Lee]

This is great! Thanks a lot!
Whether you are a “resolution aficionado” or not, once you reproduce at sizes over the common phone screen, ultimate resolution and the MTF curve has a big impact on psycho-optical impact of the photo, in ways that is not quickly put into words.

Heh... I am mentioning this as an excuse for a poor test. If I were to test for resolution, I would have exposed film specifically for such test, with a better test target, more careful approach to exposure, and finer-grained film.

It also has better dynamic range both in the shade and the highlights.

You may be right, but I would be careful about evaluating dynamic range using just these scans. First of all, the images do not have the same curve and clipping applied to them, this was outside of my control. And AFAIK the Creo software, and possibly Imacon too, has several scanning settings that may affect the captured range. I did not have access to "RAW equivalent" output for both images from these scanners.

Are the Sony files from a single capture, or did you use pixel-shift?

Single-shot.

@koraks I agree with your conclusions. Again, my primary motivation was color. I was always slightly worried by some folks claiming that you can't get "true colors" out of a Bayer-filtered area sensor. When I couldn't get those flowers to show up in my Portra scans, it was the final trigger to go and get some 16-bit TIFFs from the two highly recommended scanners.

This made me like my Sony-based rig even more. For 35mm it's already awesome, and I can easily keep up with both scanners on resolution (and beat them at speed) by stitching two exposures of medium format negatives. But the Creo looks really appealing for 4x5 though...

 

[GLS]

Single-shot.

@koraks I agree with your conclusions. Again, my primary motivation was color. I was always slightly worried by some folks claiming that you can't get "true colors" out of a Bayer-filtered area sensor

Thanks for confirming. As to the second point, in addition to resolution benefits pixel-shift captures avoid the problem of false colour from de-Bayering (at least in theory).

 

[Steven Lee]

@GLS I should play with pixel-shifting some more, but my early experiments with it were discouraging. I was getting high % of pixel-shifted captures affected by camera shake, even on a copy stand tethered to a PC. This requires stopping, checking the combined image, and re-shooting if necessary. Slows you down, especially on windy days when the house is apparently micro-shaking. Additionally, the giant DNG files were kinda painful to work with. I found that multi-shot stitching to be a better approach, that's what I do for 4x5.

 

[GLS]

@GLS I should play with pixel-shifting some more, but my early experiments with it were discouraging. I was getting high % of pixel-shifted captures affected by camera shake, even on a copy stand tethered to a PC. This requires stopping, checking the combined image, and re-shooting if necessary. Slows you down, especially on windy days when the house is apparently micro-shaking. Additionally, the giant DNG files were kinda painful to work with. I found that multi-shot stitching to be a better approach, that's what I do for 4x5.

Yeah, edge artifacts from micro-vibrations can be an issue on pixel-shifted images. Cars driving by the house, fridge pumps in the house coming on etc can cause it. A really sturdy copy stand helps keep them to a minimum.

Multi-shot stitching has its own potential problems too, including uneven focus across segments (field curvature etc), overall distortion in the stitched image, seam artifacts etc.

 

[Lachlan Young]

For what it matters, Creo and the Imacon/ Hasselblad both used Kodak sensors. I think the Eversmart and the Imacon used the same 8000px sensor, and the iQ3 and the Durst Sigmas used the 10,200px one.

 

[Lachlan Young]

It’s noticeable that the Creo has what appears to be blooming or at least general micro contrast reduction.

If you consider the complexity of the optical system compared to camera scanning or the Imacon, there's a lot more glass in the way.

The Imacon appears the worst of the three,

The 35mm scans tell a different story - the iQSmart should be much closer to the others than it appears to be in terms of fine detail handling at higher frequncies - the differences should be very small indeed, especially vis-a-vis the X5.

The big problem with Imacon/ Hasselblad is the software (but that's another story) - if you get the file out without Flexcolor doing weird things to the sharpness/ noise, then there's not a lot (pretty low single digit percentages) between a Flextight & an Eversmart Supreme II at equivalent resolutions, if everything is in spec, properly operated etc. Unquestionably, there are issues with the 120 X5 scan in this comparison - quite possibly Flexcolor related.

And ultimately, if the software issues with getting camera scanning to be really good can be solved, this discussion would be entirely moot.

 

[Helge]

The 35mm scans tell a different story - the iQSmart should be much closer to the others than it appears to be in terms of fine detail handling at higher frequncies - the differences should be very small indeed, especially vis-a-vis the X5.

The big problem with Imacon/ Hasselblad is the software (but that's another story) - if you get the file out without Flexcolor doing weird things to the sharpness/ noise, then there's not a lot (pretty low single digit percentages) between a Flextight & an Eversmart Supreme II at equivalent resolutions, if everything is in spec, properly operated etc. Unquestionably, there are issues with the 120 X5 scan in this comparison - quite possibly Flexcolor related.

And ultimately, if the software issues with getting camera scanning to be really good can be solved, this discussion would be entirely moot.

I’m mainly looking at the 135 shots, since they put a point on the differences.
Camera top, Imacon bottom.
Pretty much looks like readout noise to me.

 

[Steven Lee]

@Helge Some of this can be explained by color grain. The Sony images were processed with 10% chroma noise reduction in Lightroom, which is the default setting and I actually like using it. @4season noticed it above. But with Portra 160 the difference is negligible and noticeable only in smooth areas such as the sky.

 

[albireo]

Those Bayer sensor scans pretty much look like what you would expect images from an interpolating non dedicated line sensor to look.

No offense at all but I've seen better colour from v600 scans.

Usable images I guess if one already has a mirrorless camera kicking around, and needs to quickly scan entire rolls (eg people who inherit entire rolls from their grandpa etc).

 

[Steven Lee]

No offense at all but I've seen better colour from v600 scans.

None taken because you haven't. I have the 3F files from the Imacon right here, next to my Sony scans. Color-wise, I can make them look 100% identical, conforming to my taste. I posted some quick samples earlier in a separate thread. What you're probably seeing is that your taste for color doesn't match mine. This has nothing to do with equipment.

I spent hundreds of dollars to get several rolls scanned on these top of the line machines with a goal in mind. To finally throw the dedicated scanner color superiority myth out the window once and for all.

Besides, this thread is about resolution. The images aren't color-balanced to match, they're Negmaster defaults. I posted several disclaimers above.

 

[GLS]

they're Negmaster defaults

Slightly off topic here, but have you compared Negmaster to NLP and/or ColorPerfect? How do you find it overall?

I'm curious about it but they don't offer a trial, which is a rather odd decision. Also to read or participate in their discussion fora you need to have a Facebook account; a markedly terrible decision.

 

[brbo]

Thanks for confirming. As to the second point, in addition to resolution benefits pixel-shift captures avoid the problem of false colour from de-Bayering (at least in theory).

I tested that theory when I was bored during covid. I bought the cheapest used camera with pixel-shift that I could find (Olympus E-M5 II). Pixel-shift is just as good with getting proper colour resolution (at pixel level) as regular single shot. Or should I say, just as bad.

Further more, I scanned a frame with different scanners, optically printed it in darkroom and camera scanned it (with pixel shift). When you get "readout noise" (at exactly the same locations) on all scanners and optical print, but not on a camera scan, I would say that is a proof that "noise" on scanner scans is not sensor noise. It's detail on film. And even if that detail is just a dye cloud with no pictorial information, I prefer to have that in my scans.

Olympus E-M5 II (processed with Adobe Camera Raw), 200%* (pixel-shift first, then regular):



Noritsu LS-600:



Minolta 5400:




(click for full resolution)

Maybe pixel-shift cameras with higher resolution or from other manufacturers can get better results, but all those are out of my "playing around" budget. As reasonable as camera scanning is for people that don't have a decent scanner at home, it's equally pointless for people that don't have a digital camera or already have a decent scanner unless the speed of individual capture is the most important thing to you.

* I downsized all scans to the lowest common native resolution of the scanner (4000dpi of LS-600) and then made 200% enlargement to better see the details

 

[koraks]

@brbo to what extent would you consider the differences you noted as practically relevant? That's a subjective question, I know, but in my mind, it's still (perhaps even more so) relevant.

 

[brbo]

@koraks, I refuse to answer that!

It wasn't me who started pixel peeping here, but I certainly won't miss a chance to participate in such...

 

[koraks]

That's fair enough! I know what I think, so let's leave it that anyone can decide for themselves.

 

[Steven Lee]

Slightly off topic here, but have you compared Negmaster to NLP and/or ColorPerfect? How do you find it overall?

Never tried ColorPerfect, but I use both NLP and Negmaster. Their authors have very different philosophies towards color inversion, if you will.

1. NLP takes the aggressive guessing approach: give me an image, no matter what, and I will try to guess what the most pleasing colors for it should be. It is optimized for one-click simplicity and speed. In the latest version you feed it an entire roll, click a single button, and you'll get mostly decent results by default. I am not surprised it's so popular.
2. Negmaster takes "I am RA4 paper" approach, for the lack of a better description. Its algorithm feels far simpler, but more consistent. It is extremely sensitive to digitization exposure. So much so that I recommend exposure bracketing. Its workflow is not as quick. First step is to apply a custom DCP profile which strips Adobe color science [1], then you trasnfer the file into Photoshop and run the plugin. Batch processing is so primitive that it's fair to say it's absent.

Both of them give you a 16-bit image to apply further tweaks to, but Negmaster output is more malleable and colors don't jump all over from scene to scene like they do with NLP.

Usually I approach them this way:

• Apply the Negmaster DCP profile in Lightroom, open in Photoshop and invert manually. That's the golden standard. No automatic tool can beat that. That's how I easily matched color from all of these scanners. For high quality "portfolio" images this is where it stops. Everything below is to speed up batch processing for vacation rolls or everyday snapshots.
• Select a sequence of shots of the same scene / light conditions and run them through NLP. If the output is close to the reference image above, I am done.
• If some images aren't quite there, I run them through Negmaster conversion and then tweak them into shape in PS, usually just a couple of slight curves.

There's also Negmaster BR, which is a completely different product. I have not tried it because it's meant to be used with scanners only. Knowing the author, I'm sure it applies a similar color philosophy.

If I were to choose one, it would be Negmaster.

P.S. I am also fond of grain2pixel. It produces really nice flat & desaturated TIFFs that are easy to edit. It is also free. My only criticism of it is that it applies additional tweaks to the image, completely removes color noise, for example. It also makes all film emulsions look more or less the same. Its batch mode is pretty good too.

[1] I believe that Adobe color fuckery is the reason for people complaining about cameras not capable of capturing true film colors. If you strip it away, you get the same uninverted image as what X5 or Creo would give you.