COE 104 Annealing Schedule: What Glass Makers Publish

COE 104 annealing temperatures from Effetre and Double Helix, a CiM user report, why they disagree, soak and cool-down, and why Bullseye's 900 F is not for 104.

By Joe Blanchard, sixteen years on the distribution side of the glass industry · Updated 2026-09-27

Short answer: There is no single COE 104 annealing schedule, because the glass makers do not publish one number. Effetre lists an annealing temperature of 470 °C (about 878 °F) for its rods. Double Helix describes a broad range of roughly 900 to 1000 °F, says it typically anneals at 950 °F, and notes that most small objects are relieved within about 30 minutes at around that temperature. A user comment on Creation is Messy’s own Messy Color Clear page reports that the Clear generally cracks when annealed below about 950 °F; that is a reader’s report, not a CiM figure. Community practice among bead makers clusters around 940 to 968 °F. Pick the figure published for the glass you actually used, hold long enough for the thickest part to equalize, and cool slowly afterward. Bullseye’s widely quoted 900 °F soak is for its COE 90 glass, not for 104.

This page narrows our general glass annealing schedules guide to soft glass in the 104 family and sticks to what the makers themselves publish. For turning a schedule into controller segments, see kiln controller programming.

What the 104 makers publish

Source	What it publishes	Notes
Effetre (Murano)	Annealing temperature 470 °C (about 878 °F); expansion 104 ± 1.5 x 10^-7 per °C	Effetre’s own rod page; no soak time given there
Double Helix	Annealing range about 900 to 1000 °F; typically anneals at 950 °F; most small objects relieved within about 30 minutes; kiln shut off after an hour or two	Its annealing FAQ; also warns about restriking silver and burning off luster at higher temperatures
Creation is Messy (CiM)	No annealing figure of its own on the Clear’s page; a user comment there reports Messy Color Clear “generally cracks” below about 950 °F	A user report, not a CiM figure; the same comment warns against using the Clear with silver-laden glass
Community practice	Soaks clustered around 940 to 968 °F	Lampwork Etc. poll; shop experience, not a maker figure

Sources: Effetre Murano, glass rods; Double Helix, annealing FAQ; CiM, Messy Color Clear (user comment); Lampwork Etc. poll.

Editor’s note: these sources span more than 100 °F at the extremes, from Effetre’s 878 °F to the upper end of Double Helix’s range, and none of them is wrong for its own glass. We have not found an Effetre statement of how its 470 °C figure should translate into a kiln soak, so treat it as the maker’s published property, and use the maker’s figure for the specific glass in your piece.

Why the numbers disagree

Three reasons come out of the makers’ own notes and the reports around them.

Annealing is time and temperature, not a single point. Double Helix says so directly: the lower the temperature, the longer the soak typically needs to be. Two workers can anneal the same bead successfully at different temperatures if the one running cooler holds longer. That is why its FAQ gives a range and a typical setting rather than a mandate.

Specific colors push the answer up or down. CiM’s Clear is the reported example on the high side: the user comment on its page says it tends to crack when annealed below about 950 °F, which would put a piece built on it at or above that figure. Silver and striking colors push the other way. Double Helix warns that higher annealing temperatures can oxidize off (burn off) the surface luster on reduction colors and shift striking colors further along their sequence. A common community approach when silver glass keeps restriking (see how to strike silver glass without burning it) is to fire spacer beads and bring the annealing temperature down in small steps between sessions until it stops; that is shop practice, not a published schedule.

“104” is a family, not one recipe. CiM describes developing its colors to sit within the median range of the existing 104 clear bases, testing every melt against Effetre, Lauscha, and Vetrofond clear with a trident seal test viewed through a polarimeter (CiM, testing practices). Compatible is not the same as identical, and the annealing behavior of each glass follows its own formula. Our soft glass brands guide and the Effetre vs Double Helix compatibility page cover mixing brands.

Building a schedule from what is published

Every 104 annealing cycle has the same four parts. The makers supply the temperatures; your kiln’s controller runs the steps.

Garage. During the session, finished pieces go straight into a kiln held near the annealing temperature so they never cool into the stress range while you keep working. Garaging is a holding pattern, not annealing.
Soak. When the last piece is in, hold at the annealing temperature for your glass. Double Helix’s own figure for small objects is about 30 minutes around 950 °F; thicker or more massive pieces need longer, and so does a cooler soak.
Controlled cool. Bring the kiln down slowly through the range below the soak before letting it coast. Double Helix describes simply shutting the kiln off after an hour or two for small work, letting the closed kiln cool on its own; thicker work is where a programmed down ramp earns its place.
Off. Once the glass is well below the soak and cooling slowly, the kiln can cool to room temperature closed.

Which soak temperature to choose is the one real decision, and the makers point the same way: use the figure for the glass in the piece. If a piece mixes glasses with different needs, you are choosing a compromise, so test it. A bead built on CiM Clear and decorated with a Double Helix silver color is the pairing the user comment on CiM’s page warns against: the Clear reportedly wants 950 °F or more, and Double Helix warns that higher temperatures push striking colors further along. Choose a different clear for silver work, or run a spacer bead through your actual cycle before you commit a piece.

Bullseye’s 900 °F is a different glass

One figure travels around bead forums that does not belong here. Bullseye lowered its recommended anneal soak from 960 °F to 900 °F in 2009 and uses it for everything from simple fused pieces to large castings (Bullseye, recommended annealing cycle). Bullseye glass is its own tested-compatible fusing system, commonly called COE 90, not a 104 lampworking glass. Its schedule is a good example of how carefully a maker can document annealing, including its point that uniform temperature through the glass matters more than the soak number itself, but the 900 °F figure should not be copied onto Effetre, CiM, or Double Helix work.

What the kiln maker’s controller does, and does not do

Kiln makers build the tool; the glass makers supply the numbers. Paragon’s BlueBird bead annealer, for example, is described by Paragon as using a touchscreen controller that maintains the annealing temperature for as long as you program it and lets you adjust the cooling rate through the annealing range (Paragon BlueBird). That is exactly the job: hold the maker’s temperature and control the descent. Paragon also publishes an instruction sheet on annealing glass beads; we have not been able to read its current version and do not reproduce figures from it here. Your controller’s manual governs how to enter the segments, and our lampworking kiln guide covers choosing the kiln.

One caution applies to every brand: the display shows the thermocouple temperature, not the glass temperature. Small bead kilns can read differently from the beads inside them, so a schedule that works in one kiln may need testing in another.

Checking your schedule

Clear and transparent pieces can be checked for leftover stress between crossed polarizing filters: well-annealed glass stays dark, and stress shows as light. It is the fastest way to confirm a temperature and soak before you commit a batch of finished work. Opaque colors cannot be read this way, which is one more reason to follow the published figure rather than trusting that a piece did not crack.