Feet of clay – My first impression of Harman Phoenix

The recent introduction of a color negative C41 film by Harman Technology (of Ilford fame) was both unusual and gutsy. Entering the market with a color product with a strong reputation specifically in black & white is a surprising move to begin with. Doing so in a way that resembles a 2020s startup more so than a firm rooted in the early 1900s is even more surprising. But what gives? Let’s look at some pictures! In particular, some optical RA4 enlargements.

No, I did not buy all of this film. I just happened to be in the store as it arrived. And since I had placed a few rolls on order, this was convenient, so I could drag them home on the spot to play with.

But I feel I should share a tiny bit of the backstory. I was skeptical, or actually incredulous, that this product was about to arrive in the first place. It started with rumors on the internet in early November, that Harman was up to something. Given their legacy in B&W silver halide technology, I figured it would be something like 1600-speed 35mm film – something in-between HP5+ and Delta 3200, and with maybe a Kentmere label on it. That seemed like the most sensible and realistic option to me.

Certainly NOT a color product of any kind.

Well, clearly, I was wrong. It was a color product alright, and within that category, a color negative film makes the most economic sense. This recent ‘film revival’ (not too happy with that moniker) consists largely of people shooting 35mm color negative. I guess Harman figured that if they wanted to tap into that apparent success, it was the only option they had.

Sure, the film revival also comprises people shooting B&W, and other formats besides 35mm, there’s a few people shooting color positives and some people probably even get some darkroom printing done – but it all pales by the dominating demand for 35mm C41 film. I guess Harman/Ilford realized this very well and decided to take a bet on it.

One of the reasons I totally misjudged the possibilities of this ever happening is that I was thinking along the lines of normal color negative film. By this, I mean a masked film with very little to no crossover, large latitude, a gamma of around 0.6 – in other words, something like Portra, Superia, Gold or any of the well-known color print films on the market for the past five decades, give or take. Although we (I) experience such films as commodities, they’re in fact highly complex products that require many emulsion layers (easily a dozen or more) and several emulsion technologies that a manufacturer of B&W products is unlikely to have – and certainly doesn’t actively use in their products.

My tacit assumption that if Harman would introduce a color negative film, it would be something along those lines (masked, normal gamma, multiple layers per color etc.) turned out to be totally wrong. It does seem like Harman might be on its way to do something like that – but they’re not quite there yet. They seem to be inching their way, starting with a sort of skeleton product that appears to (1) test the waters in terms of market response and (2) raise some money for future developments.

So for now, Phoenix is a maskless color negative film, with a seemingly rather randomly chosen very high contrast (for a color negative film) and a pretty weird color response. To illustrate, look at the gamma of Harman Phoenix 200 (in violet) vs. Kodak Gold 200 (cyan curves):

Note how much steeper the Phoenix curve is, and also how its toe and shoulder are far more pronounced. The Kodak Gold curve is pretty representative for what a normal color negative film looks like; if you plot this against Ektar, Superia and even Vision3 films, you’ll see the difference are rather subtle. The difference with Phoenix, however, is fundamental.

In terms of exposure latitude, a regular color negative film doesn’t struggle much recording a 14-stop brightness range with acceptable color fidelity. Scanning this is also not an issue with any decent scanner – printing such a negative optically on RA4 paper may be a bit more challenging, but is certainly feasible. Phoenix, on the other hand, seems to record something like 7 or 8 stops, but will start to do funny things beyond this: you either end up in a low-contrast toe or shoulder zone of the curve, with shadows dissolving in a murky mass or highlights compressing into a barely differentiated fog. Additionally, its color balance will change depending on how much exposure you give it, given how the blue channel doesn’t keep track with red and blue.

Spectral sensitivity data also shows an interesting picture. Again, comparing Phoenix 200 (in red) to Gold 200 (violet):

Note the distinct peaks in sensitivity in the green (around 550nm) and red (670nm) channels. There also seems to be a relatively deep but narrow gap around 570nm. Although this corresponds to yellow light, it doesn’t mean that yellow objects will therefore always render darkly and unsaturated on this film, since we experience yellow not only as a result of yellow light, but also any mix of red and green.

I’ve been told that the green and red peaks are signs of Harman working its way towards a masked color negative film. The rationale is apparently that the masking dyes themselves act as a filter in the unexposed film, and this would allegedly coincide with these peaks. As a result, if the familiar orange mask is added, the peaks should drop significantly, bringing the spectral response of the film more in line with regular color negative films. I’ve not yet been able to confirm this; I guess I still have some reading ahead of me about masking dyes.

So, before diving into the actual photos (sorry for taking so long), what can we expect based on this little theoretical foray? It seems we’re in for some very high-contrast negatives, with blue/yellow crossover, some compression of highlights and/or shadows on high-contrast scenes, and maybe a fierce response to specific colors. Vivid, punchy – and probably an interesting challenge to print in the darkroom.

I divided a single roll into two halves, and shot one half at its box speed of 200, and another half at E.I. 125 because rumor has it that this is closer to the real speed of the emulsion. I would take the latter with a grain of salt; with this kind of toe and high gamma in straight line section of the curve, I think establishing the ‘real’ ISO is a bit like trying to fix a screw in a wall made of jello. Anyway, I figured that given its pronounced toe, the film might do well with a tiny bit of additional exposure.

Tis the season for dreary weather, so I ended up shooting all of these test frames under similar conditions: overcast, low light, on the verge of raining or just emerging from a downpour. Since I already knew contrast would be…interesting, I figured that the very low inherent contrast of the scene would establish a useful starting point.

The first half of the roll, shot at 200, I developed for the normal 3m15s in C41 developer. The second half, shot at 125, I developed a little shorter; around 3m00s or so. To be frank, the net result isn’t all that different. Maybe the second set of negatives are slightly more manageable. It’s a close call, however.

Virtually all of the examples of this film I’ve seen so far are from scanned film. That’s perfectly fine, but I personally find it very hard to tell anything useful from scanned images. They mostly say something about the skill and preference of the person twisting the sliders in the scanning and post processing software, and very little about the film itself. Moreover, my interest in color photography is mostly in printing, with optical RA4 printing being my favorite at the moment. So where I’ll be heading with this film, anyway.

Enough talk; let’s look at some pictures. To get an impression, here’s that first half roll, all frames scanned on a flatbed at once in positive mode. I inverted and roughly color balanced these to more or less neutral.

Nothing special – looks perfectly neutral and normal, right? Dreary scenes of a dreary winter afternoon, rendered pretty much as you’d expect them. Here’s the other half roll, processed similarly. I did not bother to color-balance these ‘contact sheets’ identically, so they look a little different. I wouldn’t attribute too much meaning to the differences.

Things only get a little funny when we start looking at prints. For instance, take the bottom right frame of the first set. Here’s that frame, the digital version to the left, and an optical print onto Fuji DPII paper to the right:

The optical print is balanced differently; it’s far more red. This could be balanced out at will, by either making the print more cyan or the digital version more red (I’d prefer the latter). The main difference between the two images is the massive punch the print has. While the negative scan looks as drab and grey as the real scene looked, the print shows the massive contrast that Phoenix gives. And on a scene like this one, I actually quite like it. It just…gets out of hand pretty easily. Have a look at the bottom left frame of the other set:

Uhm, right. Interesting. There’s punch, and there’s…a wrecking ball!? The scan looks again pretty much like the real scene. The print looks like all dials were turned up to eleven!

Yes, I did try curtailing the contrast by preflashing, but frankly, it’s no use. There’s no amount of flashing that harnesses really helps, at least not without flashing the paper whites into a grey mess.

Burning in bright areas is more useful, and actually opens up interesting avenues. Drab skies can turn quite dramatic. Again, scan from negative on the left, DPII print on the right – this time with generous burning of the sky:

Burning really isn’t optional, either. If omitted, I ended up with the choice of either the shadows, or the highlights in the print – not both:

As you may note, I didn’t particularly carefully color balance the prints. In part, it’s because it took a *lot* of time to balance each shot individually. On normal color negative film, you balance out one frame and then the rest of the roll pretty much falls into place. But because the curves of Phoenix are so steep, each individual negative needs to be adjusted individually, with quite dramatic filter adjustments compared to normal color printing.

And in part, I have to admit, after two or three frames, I decided that there was simply no point in trying to color balance these prints ‘realistically’. It felt like taking a wild lion and trying to turn it into a meek lap cat. What’s the point? Instead, I decided to go with the flow, see what a test print or strip yielded, and then gently (or maybe not very gently) nudge the result into a direction that seemed pleasing. Let the colors do weird things and allow the contrast to go through the roof. Heck, why not.

Trying to coax this film into normality, at least in optical printing, just feels antithetic. This is one of those frames that drove this home with me:

The print on the right I pre-flashed to harness the contrast in the sky, and I burned the right side a little as well. The result has very little to do with what the scene looked like, nor with what the scan shows. It has become…something else. A straight print wasn’t all that more realistic, by the way. I just liked the heavily doctored version better. Note that most of the prints I show in this blog aren’t very heavily doctored; apart from some burning here and there and some adjustments to color balance, I didn’t use flashing to adjust color balance very often.

Well, I guess you get the point with the comparisons. The prints just look very different to the scans, which I processed as I would generally process color negative scans: adjust contrast so the scenes fit in the histogram, adjust color balance so everything looks pretty much neutral. Only when optically printing this film, I feel it starts to show its true nature – and lives up to the things we could initially glean from the curves.

Here’s some more – just the prints. Because those are what matter, to me, in the end.

Finally, here’s a fun little thing:

Notice the tiny little dot just to the right of the tree? Harman promises the possibility of coating defects (they’re quite clear about this being an experimental product) – and this is probably what they mean. They appear to be tiny globules of dye that somehow emerge in a fully colored state from processing. I guess it has something to do with the emulsification process where tiny dye globules should be dispersed among the gelatin matrix. It seems this process isn’t entirely under control, yet. Here’s a close-up of the little thing:

On the left, the globule as it appears on the negative scan – it’s a cyan globule, clearly. On the right, a slightly larger, inverted and approximately color-corrected view.

Finally, there’s the grain that needs to be remarked upon. Since I balanced my negative scans to a pretty muted, realistic looking result, the grain doesn’t stand out all too clearly there. However, printed on the natively high contrast of RA4 paper, Phoenix 200 proves to be quite grainy, indeed. The prints I made are fairly small ones; 12x18cm (roughly 5×7″) image size. Yet, the grain is quite apparent in them. Here’s a crop of a 240dpi scan of one of the print:

Yeah, that’s grainy alright. In fact, I don’t think I’ve printed anything this grainy on RA4 paper, ever.

All considered, Harman Phoenix 200 seems to offer something unique in the color negative film landscape. It’s a wayward, wild animal. I don’t doubt that Harman are on their way to tame the beast and work towards a more mundane C41 color negative film. I’m not quite sure yet if that would be an improvement. After all, isn’t it nice to be able to slide an Instagram filter into a 35mm camera once in a while?