While trying to put some sort of order into my jumble of slides and negatives from the past 100 years or so, I noticed a small grey paper envelope tucked away in a corner somewhere. Inside this were three frames that I shot on the margins of Dronning Maud Land, Antarctica, way back in 1992.

These three shots were almost certainly the amongst first medium format photos I ever took. It was during a period of somewhat nerve-wracking waiting around, in gorgeous weather, but with rapidly decreasing temperature - the full story is documented here.

I seem to remember I had almost run out of film at that point. However, one of our little group, a technician attached to the Swedish oceanographic team who’s name I sadly forgotten, gave me a couple of rolls of 120 film (Kodak EPR 6017, which is apparently Kodak Ektachrome Professional 64), and lent me a camera to use them in. The camera was a vintage folding rangefinder, either an Agfa or a Voigtländer - its owner was clearly an early adopter in the FilmsNotDead scene, even before Film wasn’t Not Dead! It was also the first time I’d used a rangefinder, in all probability.

I’ve certainly got 6 or so frames somewhere around, but these three I think I’ve never scanned before. They’ve survived pretty well.

Whatever the photographic merits of these three may be, I think they tell an interesting story. Together with other film-era photos I have of Antarctica, largely I think I could say that there is a good chance I would not have taken them in this way today. Certainly it has something to do with the cameras, and something to do with film, and possibly quite a lot to do with experience, but the overwhelming factor is quite different.

Back in those days, there was no Flickr, no Facebook, no Twitter, no Instagram, etc. The world wide web had barely got started, and probably the Mosaic browser had just started to support GIF images. This meant that the audience for anybody’s photography, apart from a small select group of professionals, was their immediate family and friends. I was taking these photos to show my mum what the Antarctic was like, and hopefully to impress a few girls (well, I was in my mid-20s). Today, it is extremely difficult to ignore the ever-present need for “Likes”, “Faves” and whatever, as well as conforming to guru-set standards and peer approval. And there is also an almost intolerable (to me) omnipresent feeling of competition.

The middle photo of the three is really the key.  It’s a photo of, quite honestly, nothing. It ignores the rule of thirds. It isn’t going to get approved by anybody, and it would sink with trace on Flickr. Today I probably wouldn’t ever bother with it.

And this is also probably why I have very little interest in the whole Film revival movement, because for me the golden age of photography was that innocent time when all this pressure didn’t exist, when the only way to “share” was to invite a few friends around for a slide show, and when there was genuine interaction between photographers sharing a hobby, not constant competition and fighting for visibility and approval. The fact that the cameras were (arguably) more interesting is just a coincidence. And frankly, at least so far as 35mm colour is concerned, film has no advantage at all over well-informed use of digital. All the various film websites, feeds, communities seem to be doing is to take the whole squabbling mess of internet photography and switch the veneer of digital with that of analog. I’m not sure I see the attraction.

Or maybe I’m just a miserable old git. It has been suggested a few times…

 

I realised the other day that my last 12 posts have been almost exclusively about film photography. Since May, I’ve mentioned digital just once, and that was in the context of comparing with film.  This seems to have started in March, but it really wasn’t intended. I’ve also noticed that by and large, the quality of my photography has dropped significantly. Possibly I have fallen prey to the very same strain of gear obsession that in past posts I have charged the “film community” with. It may also be that I’m not finding much inspiration, and am just repeating myself.

I suppose really I’ve been dedicating quite a lot of time to getting my film photography back up to speed again, and ensuring that all the stuff I need to work works as well as it can. I think I’m almost there on that front. I’ve also been getting familiar with the Linhof 612, which is not that simple. Actually, the Linhof seems to have quite a serious fault which is causing uneven film winding, in some cases resulting frames overlapping. So it looks like its going off to the factory for servicing, which is going to be expensive. The previous owner told me he never had any issues, and I have no reason to disbelieve him. Caveat emptor, I suppose, especially when buying through eBay.

The above photograph is absolutely digital. It was taken two weeks ago, way off the beaten track in the Aspromonte region of Calabria, right in the south of Italy. Aspromonte, most of which is a national park, is absolutely stunning. It is harsh, arid, with precipitous abandoned villages connected by crumbling, vanishing roads (Europcar would have a fit…), and astonishingly beautiful.  There are few people around, but those few are welcoming, friendly and embarrassingly generous. We had only 2 days there, but I’m certain I’ll be going back.

I suppose Aspromonte would look even more stunning on Portra or Provia. But hauling medium format film cameras down there would be a real struggle. And would it even be worth it? I’m not going to try to pretend: in terms of real resolution, even a 5300dpi scan from medium format film doesn’t beat a 16MPix Olympus file, never mind a Sigma Quattro file. Resolution isn’t everything though, and there remains a clinical precision in digital which I sense rather than see. It don’t like it, but I can live with it.  Just as the lens I took the above shot with, the Olympus 14-150 zoom, is probably optically my worst. The bottom right corner is really soft at wide to medium focal lengths. But it is extremely light, very flexible, and great to travel with. So, like digital, I tolerate it.

So yes, I am quite conflicted about film versus digital, and I suppose I always will be. I wish I could just choose one, but I don’t suppose I ever will. But it does seem that the less I bother about gear in general, the more enjoyable I find photography. Maybe I should turn off the internet.

A few posts ago, I wrote a rather dismissive impression of the new Rollei Variochrom film. Unfortunately, I’d bought 4 rolls of the stuff, so I felt I should do something with it. Having discovered what it actually does, which is to transport one back to the Good Olde Days of wildly inaccurate colour and grain you could eat for breakfast, it occurred to me that the part of the world I’m constrained to wander during the working week might actually benefit from this treatment. Well, it would be hard to make it look more dull than it actually is - although Dog knows I’ve tried over the years.

I’m pretty much at odds with todays retro film community, which seems only interested in the flaws and weaknesses of film. There are certainly people doing fabulous work today with film, for example Bruce Percy, but the film camera hipsters don’t actually seem to be interested in photographing much else than their cameras. 

Oh dear, have I got off track again ? Where was I ? Oh, yes ... Variochrome.

When used forewarned and with intent, I have to admit it can be quite interesting.  I quite like the following sample, although its not really my thing.  In the right context Variochrome is interesting, but I still pretty much stand by my earlier comments.

The canister light leak I encountered on the first roll repeated itself, by the way, despite my taking special care in loading, unloading and handling the film.

Oh well, only another 2 rolls to go.

This is an addendum to my two previous posts comparing the Plustek OpticFilm 120 to the Hasselblad Flextight X5. The X5 is supposed to be the nirvana of desktop scanners, a dream machine with a nightmare price tag, which provides the benefits of a drum scanner with none of the downsides, such as fluid mounting, even huger cost, and elephant-level proportions.

So, what’s the X5 actually like to use ? Well, overall, it’s pretty nice. The hardware is on a completely different level. It is easy to use and very fast in operation. It scans a 6x12 frame in under a minute. But there are some drawbacks. First of all, the film holders. They’re a little weird, to be honest. The heavy rubber masks remind me of the protective shroud things you have to wear when being X-rayed! And actually they’re not that easy to use. The problem is, you lay the film on a base plate, then lower the flexible mask, which has an aperture corresponding to the frame size. But there are no guides to align the film on the baseplate, so getting it to fit correctly in the aperture is quite fiddly. Maybe with more than two hours’ experience I might realise that there is a better way to do it, but actually I find that the Plustek holders are much easier to load accurately. And them the fixed aperture size is a bit of an issue: although in theory Medium Format film has defined sizes, different cameras have slightly different film gate dimensions. So, my Bessa 667 seems to expose slightly more area in the cross-film direction, and the Linhof 612 is more like a 612.5. Actually the Plustek holders are a bit narrow in the cross-direction too, but they are fully adjustable in length. The X5 holders have no adjustment whatsoever, although I believe Hasselblad will be delighted to sell you a custom holder. Once you’re loaded, though, it’s really smooth: just slide the holder forward between the guides until the magnets latch on, and it disappears into the scanner and does its stuff. Then, there’s the software. Really, I have yet to meet a simple, straightforward and well-designed piece of scanning software, and Flexcolor is no exception. For example, if I select the 6x12 film holder, would it be unreasonable of me to expect the scan size to default to 6x12 ? Obviously it would: it took some time for me to realise that the bloody thing was producing two scans (overlapping, fortunately).  Why it does this for 6x12, but not 24x65, or 6x7, will remain a mystery. Another unattractive thing about Flexcolor is that it only responds to adjustments after you let go of the respective control (slider or whatever). This makes it exceedingly annoying to use. And I still haven’t figured out exactly what it is that makes it show a full resolution rather than horribly pixelated low resolution preview - but then again, Silverfast still hasn’t figured out zooming. Flexcolor includes a fairly limited set of negative profiles, a bit like Silverfast Negafix. However, they are much less extensive, far less adjustable, and too my eyes less accurate than Silverfast, although they do produce very flattering results.

X5/Flexcolor interpretation of Kodak Provia 400

Plustek/Silverfast interpretation of Kodak Provia 400

I don’t think either are fully accurate (actually, the Silverfast version suffers from using Multi-Exposure, which doesn’t play nice with Negafix - the marketing material doesn’t tell you that). So, I could do better “out of the box” from the Plustek. Even as it stands, to my subjective eyes the Plustek/Silverfast version looks more like how Provia is supposed to look, The X5/Flexcolor version is almost Velvia, and too neutral. Of course, Silverfast is regularly updated. Flexcolor isn’t.

On the plus side, Flexcolor coupled with the X5 can really pull out clean shadow detail to an extent I’ve never seen before from a film scanner. Of course that is limited by the exposure range of the film, so it is more useful for negatives. But still, the difference is clear.

Generally Flexcolor feels old and fairly clunky, but I suppose that is because it is old. The X1 & X5 scanners are just rebadged Imacon Precision 848 and 949 models, which Hasselblad inherited when they bought / merged with Imacon to get digital back technology. They’ve carried on selling them, but they certainly haven’t tried very hard. I reckon they could drop the price by 50% and quadruple their sales, and still make a decent profit. They’re just not interested. The only update to the hardware is case redesign and a Hasselblad logo. The backplate is still branded “Flextight Precsion”. And the backplate also features a Firewire 400 interface, the one and only port, fully obsolete. At least it isn’t SCSI. The Mac Pro running the X5 at Light & Byte is running Mac OS X Snow Leopard, which must be 8 years old or something.

So, is the X5 worth €25’000 ? No, unless you’ve got a business model which will subsidise it, or you work for a Swiss bank or something. Is the X1 worth €16’000, then ? I don’t think so - it loses the X5’s diffuse light source, it’s slower, and (if this matters to you) it can’t do batch scanning, or, indeed reflective scanning (which I really doubt is a major selling point of the X5, but whatever).  If Hasselblad were showing some sign of continuing development, even if just to add a USB3 interface, and updating Flexcolor, then just maybe it could be considered a long term investment, but I would even be concerned that it will continue to work with current operating systems for much longer.

After all, Hasselblad recently named their new pride and joy camera the “X1D”. Presumably they didn’t even remember they’ve already got an X1 in the catalogue.

Last week I published the first part of my Plutek OpticFilm 120 vs Hasselblad Flextight X5 scanner bake-off. The results from scanning XPan frames were surprisingly close, with the X5 subjectively winning by a whisker.

The key word in the above paragraph is “subjectively”. Because film scanning is nothing if not subjective. While preparing this second part, I once again fell down the rabbit hole of comparing output from different scanners, different software, different settings, reproducible bugs, irreproducible bugs, and indifferent customer support.  I learned, or-relearned, a couple of key points about scanning, which can be summed up by saying that magic features - like “multi-exposure” - generally don’t work. It’s better to keep to the basics.

That out of the way, this time around I’m going to look at two Medium Format scans, a Portra 400 negative taken with the Voigtländer Bessa III 667, and a Provia 100F positive taken with the Linhox 612.

First up, the Portra shot. This was taken in a secluded spot in Venice, far from the tourist gyres.  Note, you can click on any of these images to see them in larger size on Flickr.

Left: Plustek OpticFilm 120 (OF120), right Hasselblad Flextight X5 (X5)

In this case I decided to apply a simple grey balance using the same reference (the letter box in the background, on the right) in Silverfast HDR for the OF120, and in FlexColor for the X5. For the OF120 I used Silverfast’s Negafix Portra 400 profile, for the X5 I used FlexColor’s Portra 400NC. As you can see, they’re pretty close. Preferences are subjective. I could show you a similar frame from my Sigma DP0, and you would see that there are differences, but if I went down that path it would never end.

So, what about detail ? Well, I decided to zoom in on the notice in front of the staircase. Since the OF120 allows a scan a 5300dpi, and the X5 only at 3200dpi, the size at 1:1 is different.

1:1 zoom - left: Plustek OpticFilm 120 (OF120), right Hasselblad Flextight X5 (X5)

The extra resolution of the OF120 doesn’t really seem to add much, here. But it might be interesting to downsample to 3200dpi to get a better idea.

Matched resolution - left: Plustek OpticFilm 120 (OF120), right Hasselblad Flextight X5 (X5)

Well, that’s interesting, isn’t it ? As we saw in the last episode, the X5 is perhaps slightly less noisy, or possibly the diffuse light source is decreasing grain contrast, but in terms of resolution it’s pretty much a dead heat.  The OF120 scan is a little more contrasty, which may give the impression of more detail. On the other hand, it may really be delivering more detail.

So far so good, the Plustek has nothing to be ashamed about.  Let’s move on to the Provia frame. While colour negative film presents substantial challenges in colour representation, it is generally low contrast. Slide film , on the other hand, should present less problems for colour accuracy, but contrast is another matter altogether. Shadow areas can be extremely dense, and detail easily visible on a light table can be completely lost in a scan. Provia isn’t too bad in this respect, but Velvia is very tricky.  It’s just as well that I don’t much care for Velvia.

So, here’s our Provia frame:

Top: Plustek OpticFilm 120 (OF120), bottom: Hasselblad Flextight X5 (X5)

You can immediately see a colour difference. The X5 scan is direct from FlexColor with all settings zeroed, and sharpening off. The OF120 scan has had some magenta removed from the highlights and shadows, and contrast reduced slightly. The FlexColor scan is a remarkably accurate representation of the slide on the light table, to my eyes. The greens especially are more accurate. The OF120 scan seems to be lacking a certain amount of tonal separation in the higher midtones. Still, I’m not sure there’s $23’000’s worth of difference.

So what about detail? Slide film is generally sharper than negative film, so this also could be more challenging.  Note, however, I’m not really very familiar with the Linhof 612 yet, and I have some question if I was using an optimal aperture here.

Top: Plustek OpticFilm 120 (OF120), bottom: Hasselblad Flextight X5 (X5)

In the above illustration I have downsampled the Plustek scan to 3200dpi. Although contrast and micro-contrast might be playing a part here, I have to - just - give the prize to the X5. Looking at the tree branch, you can see a smidgeon more detail. But let’s face it, unless you make a print the size of a bus, it is totally insignificant.

Note though, if you allow FlexColor to do post-capture sharpening, the gap widens considerably. FlexColor appears to have very good sharpening algorithms, tuned to scan resolution and scan type.  Silverfast’s USM on “automatic” setting is also not so bad. But nowhere as good as the FlexColor / Flextight combination. Still, there are many options for sharpening.

However, there is one area where the X5 nails it. Just as we saw for the XPan slide scans, for shadow detail the X5 wins easily.  In the sample below, I haven’t even touched the shadow depth slider in FlexColor, which can widen the difference still further with significant downside.

Shadow detail - left: Hasselblad Flextight X5 (X5), right Plustek OpticFilm 120 (OF120)

So, from this point of view, it is difficult to avoid envious glances in the direction of the X5, or at least it’s X1 sibling, which is only astronomically priced rather than absurdly.

But wait.  There’s an elephant in the room, keeping very quiet over there in the corner. Well, a very small elephant.  Take a look at this:

Same slide scanned using the Canoscan 9000F

The above version was scanned on the Canoscan 9000F flatbed scanner using Silverfast Ai Studio, calibrated with the same IT8 slide as the OF120. This is straight from the scanner, with all Silverfast settings flat. The colour accuracy is quite noticeably better than the pre-adjusted OF120 version, and it appears to have more shadow depth.

This is a bit scary. So what about resolution ?

1:1 - top: Canoscan 9000F, middle OpticFilm 120, bottom: Flextight X5 (X5)

Well, the 9000F can’t quite keep up, but at less than 1/100th of the cost of the X5, I guess it does a reasonable job. I only really bought the 9000F some years ago because I needed a document scanner, and remarkably it came bundled with Silverfast HDR: this was easily the cheapest way to acquire that software. So really I haven’t paid that much attention to it. It just sits there and does what it’s told. I am vaguely aware that it has a reputation of being rather under-appreciated, especially compared to Epson flatbeds.  These days you can pick up a 9000F MkII, sadly without Silverfast, for just $200. The film holders are truly horrible, but otherwise, it’s pretty good.

So, what about the OF120 vs X5 ? Well, I think the OF120 delivers quite enough resolution. Colour accuracy is another matter. Of course, one could blame Silverfast, but I have used the latest versio of Vuescan as well and have found similar issues. And the same Silverfast delivers much better results with the Canoscan.  I have made a whole series of Portra 400 scans with various combinations, which I may present as addendum, but in that case getting good colour out of the OF120/Silverfast combination proved quite a challenge.

On balance I think the Of120 put up a pretty good fight against the Flextight X5, but at the same time it is not as superior to the 9000F as it should be.  For scanning 35mm film I’d still go with the higher resolution of the OF120, but for medium format, honestly I’d be tempted to recommend the cheap but excellent Canoscan, or perhaps the more expensive but theoretically superior Epson V850 (which I’m tempted to try). One advantage that a dedicated film scanner should have is delivering better shadow depth and tonal separation for slide film. I’m not convinced that the OpticFilm 120 achieves that.

As a final note, I should point out that I haven’t yet extended testing to black & white film. That brings a completely different set of issues, and may well result in very different conclusions.  I don’t generally have much to do with black & white, but since I have 5 rolls of Ferrania P50 to burn through, I may have something to say on this later.

Finally, if you have any questions, feel free to ask…