Recording session 2

Phonograph recordings – temperature study


Cylinder type: 2 minute

Piano: Inja Stanović

Instrument: Broadwood (1910)

Room temperature: 21.5°C

Temperatures tested: 30°C, 20°C and 10°C

Mechanical equipment: Edison Fireside – model B; Edison New recorder (1901) – glass diagram; brass flared horn (762mm by 355mm)

Digital equipment: pair of AKG414 microphones; pair of Neuman 184 microphones; Steinberg UR22 soundcard; DAW; wave-editor

Other equipment: thermochromic strips; USB microscope; portable incubator; fridge


As described in Recording session 1, this project started with two studies using two different pianos and various mechanical equipment. The conditions of both recording sessions were quite different and produced two contrasting results, as one room in which we recorded was slightly cold (around 15°C) and the other one was warmer (around 22°C). The experience of recording in slightly warmer room was far more rewarding, as wax was not as hard and I didn’t need to play overly loudly. Not only that the process was more straight forward, and I did not need to play beyond my strength; but also the cylinders produced in the case-study were better quality and much louder than the ones from the pilot-study. Even though it is a common sense that a warm wax will cut better than the cold, I wanted to measure the process as much as possible, and understand this through numbers and not just descriptions.

This recording session was very straightforward. I played Chopin’s Mazurka Op. 67 No.2 numerous times in a row, but registering my playing on wax cylinders kept at the different temperature. The playing needed to be the same, or as close to the same, on all the cylinders so this was the major challenge of this recording session – to play the piece so many times in a row, being careful that I am actually playing it with the same strength. The piano used was upright Broadwood from 1910, which was kindly loaned to me by my colleague from Huddersfield University Dr. David Milsom. I am extremely grateful to have an upright piano whose age fits this project, and in such a fine playing condition, so a huge thanks to David! This piano is later used in all the recording sessions, as it was (unsurprisingly) proven that the recording results were much more satisfying when we used this instrument, rather the grand pianos from the previous studies.

The room temperature was measured 21.5°C, while the reading for outside temperature refers to hallway in the University of Huddersfield. We recorded this on the 28th of December, with a significant difference in temperature outside. It would be very surprising that outside temperature was 21.1°C in the North of England (and a dream come true for me).

The process

I will not go into details of recording process here, because it was very similar to Recording session 1, described here. We used the knowledge and experience gained from the previous sessions, and decided that we will use Edison Fireside (Model B) with Edison New recorder (1901) with a glass diagram. After some initial testing, we chose the brass flared horn (762mm by 355mm) because it gave the clearest results.

As with previous sessions, all transfers which you could hear on this page, were made in situ straight after the recording session ended. Like this, we could monitor the conditions so they match with each transfer made.

Here are some photos from the process, and you can see we really had a lot of fun!

The first batch of wax cylinders was put in the fridge, keeping them cool on 10°C. Luckily there was no staff in the Music Department on the day, as they would be very surprised to see a batch of cylinders in the community fridge, instead of food. Duncan measured each cylinder’s temperature using thermochromic strips throughout the session. The second batch was put in the room, so they adapt to the room temperature (21.5°C). The third batch was put in the portable incubator which kept them all on 30°C.

Room temperature cylinders

The first cylinder was used to run some small checks. We wanted to decide which recording horn, out of two we brought, we would like to use – long zinc horn (first) or a brass flared horn (second), or a brass flared horn with a different cutter (third). Also, we never recorded with the aforementioned Broadwood, so we needed to check how it resonates. The three test takes you can hear here:

Cylinder 1. Tests.

After the tests, we recorded a few cylinders on the room temperature. Both cylinders measured 20.6°C.

Cylinder 2. Room temperature wax. 20.6°C
Cylinder 3. Room temperature wax. 20.6°C

Using the upright Broadwood really made a difference, even though it wasn’t as loud as the grand pianos we used previously. It was much easier to set both mechanical and digital equipment up. From the pianist’s perspective, I was very happy not being able to see any of the equipment around me – as it was all facing the back of the piano. The only small problem was to peep over the piano, waiting for Duncan to give me a sign to start playing – as in mechanical recording sessions it is that way around, which I had to get used to. The piano recorded perfectly, and its slightly brassy sound registered wonderfully to cylinders, with an occasional high pitched tone poking through.

The cool cylinders

While we were busy recording the room temperature cylinders, our little batch in the fridge was reaching a cool temperature of 10°C. Once the cylinders were taken out of the fridge, we had to be extremely fast as they warm up very easily and quickly. The first cylinder of the cool batch measured a high 11.7°C, just after a few minutes which we needed to take it one floor up and put it on already prepared machine. I also needed a moment to concentrate before I started playing again. With the next cylinder we were quicker, and even so once it got out of the fridge it showed 10.6°C, and only a few minutes later it was already on 11.2°C.

The results were surprisingly good, as I assumed that most of things will not be cut deeply enough to make a quality recording. On both of the cylinder transfers below you can hear that the takes were quite stable, with not many interruptions .

Cylinder 4. Cool. 11.7°C
Cylinder 5. Cool. 11.2°C

The warm cylinders

The portable incubator was keeping the cylinders warm for all this time, and we were very excited to see how they will cut. The same problem appeared soon: the moment we would take the cylinders out of the incubator, they would start rapidly cooling down and once again, we had to be very fast. The first cylinder measured 29°C, the second dropped significantly to 27.8°C, and the third came to 29.2°C.

Cylinder 6. Warm. 29°C
Cylinder 7. Warm. 27.8°C
Cylinder 8. Warm. 29.2°C

One can hear how ‘wobbly’ the sound is on all three cylinders here. The sound quality was particularly bad, as the large quantities of suave showings were in the way of cutting the cylinder properly. That is the reason for fading in and out throughout all three cylinders. The unstable cylinder 6 showcases this clearly. It is interesting that the difference in temperature is quite small between the three takes, but the end result is different.

We also took photos of each cylinder with an USB microscope. Here how they look, for cylinders nos. 2, 5 and 8.


Even though I assumed that the warm cylinders will be the most successful ones – I was wrong, and they proved to be the least good ones. The cold cylinders were surprisingly good, with an occasional wobble and a fade out. The best were the room temperature ones and we will strive to leave the cylinders for a significant amount of time in the room before we start recording in the future – this session proved that this is necessary. This is a unique study in all of the sessions I made, whose purpose was to better the rest of them, and to help anybody who is attempting to make the cylinder of their own. Even though very straightforward, putting values on descriptions helps for this study to be recreated, challenged and ultimately improved.