The proof overshoot is a teaching example, not a validated recipe. No fixed resting rule follows from months in wood. No blinded preference test establishes improvement. Dated commercial availability and future opening plans are not current facts.
Text-only transcript review. Selected anchors locate important methods and disputed statements; the full supplied text was read.
Recalls early blending with retail bottles, directly connecting professional development to consumer practice. Uses 100 mL prototypes so milliliters correspond to percentage points. A four-component blend is adjusted, but deictic references and missing quantities prevent recovering the complete formula. Lower-strength whiskey can help lower blend proof; actual current ABVs are needed, not barrel-entry proofs. An approximately 118-proof sample is intended for 108 but measures around 103 after an imprecise water addition. Overdominant orange/marmalade prompts changing component balance. Different barrel yields require volume weighting, not barrel counting. Records include whiteboard photographs and saved notes, followed by rebarreling and reassessment.
Character 18838
Character 1
Entire supplied transcript; precise character-window anchors in body
Verified against supplied transcript only, not original audio or factual truth. Character indices are one-based Unicode positions in the extracted UTF-8 text.
No access
Source and review scope
No access · 
All 3,789 words / 18,838 extracted characters read, from first supplied text through the final supplied text. This does not verify the completeness or accuracy of the underlying video.
Located evidence windows
These are exact supplied-transcript strings for traceability, not audio-verified quotations. Character offsets are one-based and inclusive. Windows are selected anchors, not a claim that only these passages were read.
Characters 11734–12266
u take yeah so we proofed it down and it really brought out a lot more Oak to me and the proof that you brought it down to uh so it's a little lower than I wanted to so it's about 103 um but still that that orange marmalade notes a little strong so if I was to add a little bit more I'd use a little bit more of this uh Tennessee our Tennessee bourbon um just to give it a little more of that that bright strawberry uh raspberry and that chocolatey sweetness up front yeah and the other thing that is also as we kind of mentioned pro
Characters 14103–14642
igital right I because you know notebooks can get lost that can catch on fire like let's just make sure it's in the cloud somewhere yeah I do like writing stuff so I write it on a whiteboard when I'm mapping out so every blend we've done is on a whiteboard and I take a photo of it and upload it into ever note so got so that we can reference back uh you know to our Blends and whatnot I love it that's great and then once you land on it then you would put in the tote and you bottle it and you push it out well now we're we put them in the
Analytical digest
- Recalls early blending with retail bottles, directly connecting professional development to consumer practice. Uses 100 mL prototypes so milliliters correspond to percentage points.
- A four-component blend is adjusted, but deictic references and missing quantities prevent recovering the complete formula.
- Lower-strength whiskey can help lower blend proof; actual current ABVs are needed, not barrel-entry proofs. An approximately 118-proof sample is intended for 108 but measures around 103 after an imprecise water addition.
- Overdominant orange/marmalade prompts changing component balance. Different barrel yields require volume weighting, not barrel counting. Records include whiteboard photographs and saved notes, followed by rebarreling and reassessment.
Restrictions
- The proof overshoot is a teaching example, not a validated recipe. No fixed resting rule follows from months in wood.
- No blinded preference test establishes improvement. Dated commercial availability and future opening plans are not current facts.