Henry outlines an in-camera test very similar to Adams. A few things puzzle me, but I will limit this to the x-axis of the curve plot. He retains the use of relative log H. Ok. But I don't get how you can start the scale at 0.0 where the two axes cross in a test like this. Henry begins with a metered exposure of the grey card and places it on Zone V (arbitrary since it is not a Zone System test per se). He then recommends stopping down (or reducing shutter speed) 6 or 7 stops to start the series of exposures so that you ensure you will start below a net D of 0.1, and of course also include a frame shot with the lens cap. This is all fairly straight forward. But if you call the lens cap exposure 0.0 on the x-axis, which exposure is 0.3? 0.3 relative to what? The first non-zero exposure could also be any other log number relative to the lens cap exposure. I don't get how you can do this unless you are somehow able to determine one of your first non-zero exposures happens to be precisely the threshold exposure for the film/processing combination. To me, in an in-camera test, once you base the series of exposures on a meter reading, the only reference point (ie 0.0 on the relative log H scale) is the middle grey exposure. You'd then label the x-axis with negative relative log H values below the metered exposure, and positive relative log H values above the metered exposure. I think I'm missing something here. Thoughts?

I agree, zero may not relate by 0.3 from lens cap zero to Zone I. Now it would relate if you created a Zone 0 exposure, but often the flare doesn't allow camera tests to go that low. I'd be more likely to try to estimate the exposure at the x-axis where you find a net D of 0.1, and start my labeling from there. I figure if you hit the ASA parameters with a standard developer this point is likely to correlate reasonably well to the manufacturer's rated speed.

Hi, this seems to be the only real issue. But when I looked through the section, I don't see that Henry actually said this. (If he DID, this would be improper; using the log axis does not allow for plotting an exposure point with zero light, which is what a lens cap should ideally produce.) So I don't see anything wrong. It looks like he simply plotted the actual exposure steps down to the point where they don't change anymore, which should be the same density as the lens-cap exposure. But it's not actually the lens-cap exposure which is plotted. As he said, the lens-cap exposure is a reference for base+fog, and I as I read things, that's all it is. I think this is the only problem you have, right? Otherwise, it's ok to use any value one likes as a starting point on the log exposure axis. (note for other readers: it's page 144 in second edition, Richard Henry's Controls in Black and White Photography)

Michael, if I understand your question correctly, you are just over thinking this. Relative log-H is an reference to measure the difference between exposure and not the actual exposure. The placement of 0.00 is arbitrary. For my graphs, I place the first step on the step tablet at 0.00 and the last step falls at 2.75.

0 is really a logarithm and it's arithmetic value is "1". It's not arithmetic "No Exposure". You can have -1 exposure on that scale, which would be a tenth of the reference. If the absolute measure were meter candle seconds, then where you sometimes arbitrarily assign 0 might be close to -3 log mcs. Again it's arbitrary. In your example 0 moves around. That helps to keep your graph simple. Stephen, I assume you place 0.00 on the right and 2.75 on the left of the x-axis, and it is a log measure of the "attenuation" from full exposure. 0 log mcs is 1 arithmetic mcs. 1 arithmetic mcs is a lot of light. A film which requires that much light to develop to 0.1 net density - if my graph is correct - is EI 0.7. -3.0 log mcs corresponds to the 0.1 speed point of EI 800 (Have I got my graph aligned correctly?). The arbitrary relative scale that reads 0 at the left and ascends 1, 2, 3 to the right, is setup that way just to be easy to visualize. A typical log mcs scale would read something like -3, -2, -1, 0 from left to right. But the arbitrary scale doesn't necessarily place its 0 where the log mcs scale has -3 it's just close to it sometimes.

The other way around. Even though my testing yields actual log-H, I don't graph it that way. I use relative log-H.

Exactly! However, making an "exposure" with a lens cap IS an arithmetic "No Exposure," and cannot be expressed as a log exposure value (try to find the log of zero). Therefore the lens-cap exposure cannot be graphed on this chart. I think this was Michael's only serious issue, apparently a misunderstanding as Henry didn't actually say to do this.

But since the scale is relative, you have to have a reference point. In Henry's test, the only reference point I can see is the metered exposure. Henry doesn't state that the lens cap exposure is his reference point, or that it is the exposure he placed at the intersection of the axes, but it must have either been that exposure or one of the other exposures which yielded zero net D. Suppose the exposure 6 stops below metered is the first one below metered that yields zero net D, and he decides to start the curve there. The only thing you can say is that this is 1.8 below metered on the relative log H scale, because the metered exposure is the reference point. You stopped down 6 stops below metered. It may be that 5 2/3 stops below metered also yields zero net D. Unless you maintain the metered exposure as a reference point, I don't see how you can plot your first non-zero net D exposure at 0.3 relative to a zero net D exposure. I'm probably not explaining this well. And I admit it is picky, but Henry is picky. That's one of the things I like about him. I get that it's an arbitrary scale, but it just seems like an odd way to label it if the test is one that bases everything on a metered "Zone V" exposure.

You don't even need to have the metered exposure point as a reference on the curve. Put 0.00 at the Δ1.80 log-H and the metered exposure will be at a relative log-H of 1.80. Now you might be thinking that the relative log-H should stand for an approximate actual log-H, but that's not what Henry is doing. How's it different than the resulting densities from a step tablet exposure (other than way less accurate in the exposure increments)? I still think you're over thinking this. Maybe if you walk through the details of Henry's procedure and point out the part you are having a problem with. And BTW, Henry's concept of speed and metered exposure isn't exactly on solid ground, but that isn't related to the problem you're having.

Well, there are a few things I find strange about his speed/metered exposure concept, and that section of the book (which we've discussed before). But I guess I was over-thinking this particular thing. In general I find the x-axis can be sometimes a confusing thing with H&D curves. Thanks to everyone who responded. Much appreciated.