I've returned to black and white film photography after a break of about twenty years. To start with, I'm just using three of my old cameras, one each in 35, 120 and 4x5 format, and one film, HP5+. The cameras were CLA'd earlier this year. My intention is to spend some time with one developer, D23 stock, until I know it backwards (long ago I supplemented my income by taking commercial portraits, but outsourced the processing to a lab which used D23 for the black and white). I'm using a Minolta Spotmeter F, which I purchased new in the 1980s and which has been well cared for.

My first experiments are with the medium format camera, a Rolleicord Vb. The first little challenge is working out the exposure index and development times for my combination of camera and light meter. Most of the advice on the internet seems to be along the lines of "start with the recommended development times for D76/ID11 and work from there". Ilford's fact sheet recommends 7.5 minutes in these developers at the box speed of 400. The same recommendation is provided by the Massive Dev Chart, although the Digital Truth page presenting the D23 formula recommends the starting development time of 12 minutes.

I found a quote from the Kodak Data Sheet for D23 here:

which states that the "Average development time is about 12 minutes in a tank or 10 minutes in a tray at 20C".

So I ran a test at 10 minutes and, metering for Zone I, determined that, for my Rolleicord and spotmeter, the EI was indeed 400. The highlights, however seemed a little underexposed to me.

I developed my second roll for 12:23 at 20.5C (equivalent to 13:00 at 20C by my calculations) and this seems to be pretty close to correct to me. The attached image is from this roll. It's been scanned from the negative on an Epson V700, levels adjusted post-scan to a point that just avoids clipping of highlights and shadows, and then resized. No other manipulation. I metered on the darkest part of the water under the nearer pelican, Zone II1/2 according to my meter. The brightest highlights, on the backs of the pelicans, seem to have at least some hint of detail.

The original size image can be accessed here:

The original size is the full 20 megabyte scan - the file is a 3 megabyte JPEG.

I have two questions. First, to those more expert than I in judging negatives and development, do you think I am close to the correct developing time, to the extent you can tell from the image?

Second, and perhaps one for the chemistry experts, Ilford has a single time/temperature chart for all films and developers. From the little I understand about metol and superadditivity, the chart for a metol only developer might be quite different in linearity and slope to one for, say an MQ combination like ID11. I found some information, in Jacobson's book, about temperature coefficients for developers. This suggests that metol only developers do, indeed, have a different time/temperature function to MQ. On the basis for this, I'm intending to experiment with

Time = 35 - 1.1 * Temperature (Celsius)

for HP5+ in D23 stock.

This is for the temperature range in which I'm likely to be developing, 17C to 24C. Does this sound reasonable, or am I better to stick to the Ilford chart?