Nicholas Guttenberg
@ngutten.bsky.social
Same original soil (purchased compost/mulch mix), but one sample was stored in my garage for two years, one was outside in a raised bed for two years, and one was from a polytunnel where I'm currently growing tomatoes.
I'll take a second set of samples tomorrow to see if it replicates.
I'll take a second set of samples tomorrow to see if it replicates.
June 28, 2025 at 10:28 PM
Same original soil (purchased compost/mulch mix), but one sample was stored in my garage for two years, one was outside in a raised bed for two years, and one was from a polytunnel where I'm currently growing tomatoes.
I'll take a second set of samples tomorrow to see if it replicates.
I'll take a second set of samples tomorrow to see if it replicates.
Incidentally, you do see different features with different electrodes. The previous plot was a pair of copper electrodes made of copper tape. This is what you get with zinc-coated steel nails (yeah, I should get proper electrodes one of these days...)
June 28, 2025 at 9:09 PM
Incidentally, you do see different features with different electrodes. The previous plot was a pair of copper electrodes made of copper tape. This is what you get with zinc-coated steel nails (yeah, I should get proper electrodes one of these days...)
Promising! I take the sample, apply voltage Vh, let go, then measure the voltage Vc and take dVc/dVh. The negative portion is an interesting feature - I think its because of alterations to the electrodes.
Now if I only had a thousand or so soil samples with known nutrient profiles to train on...
Now if I only had a thousand or so soil samples with known nutrient profiles to train on...
June 28, 2025 at 9:09 PM
Promising! I take the sample, apply voltage Vh, let go, then measure the voltage Vc and take dVc/dVh. The negative portion is an interesting feature - I think its because of alterations to the electrodes.
Now if I only had a thousand or so soil samples with known nutrient profiles to train on...
Now if I only had a thousand or so soil samples with known nutrient profiles to train on...
This is the normalized voltage just after switching the DAC to ground.
June 22, 2025 at 11:26 PM
This is the normalized voltage just after switching the DAC to ground.
Doing some experimentation with charging/discharging curves of flooded soils. There's a weird voltage-dependent effect I don't understand that seems sensitive to the soil type. Or of course there could be a bug. Specifically, this immediate drop.
June 22, 2025 at 11:23 PM
Doing some experimentation with charging/discharging curves of flooded soils. There's a weird voltage-dependent effect I don't understand that seems sensitive to the soil type. Or of course there could be a bug. Specifically, this immediate drop.
A bit of light (running average over 1100 seconds) vs temperature data. It gets really hot in there, for an outside air temp during the day of roughly 26C! This is with windows and door flaps open.
June 1, 2025 at 6:40 PM
A bit of light (running average over 1100 seconds) vs temperature data. It gets really hot in there, for an outside air temp during the day of roughly 26C! This is with windows and door flaps open.
Messed around with making a digital theremin from a capacitative sensing board. Drifting calibration makes this really painful and MIDI seems bad for smooth pitch shifts and volume envelopes. I still think the potential is there, but I need to do it another way.
May 29, 2025 at 9:48 PM
Messed around with making a digital theremin from a capacitative sensing board. Drifting calibration makes this really painful and MIDI seems bad for smooth pitch shifts and volume envelopes. I still think the potential is there, but I need to do it another way.
Capacitative soil moisture measurement. There's more going on here than I would have expected! I don't understand this pattern at all. Why the sudden increase? Why does it increase after noon?
May 28, 2025 at 3:37 AM
Capacitative soil moisture measurement. There's more going on here than I would have expected! I don't understand this pattern at all. Why the sudden increase? Why does it increase after noon?
Now, this one is a bit painful. Yesterday I got temperatures around 45C in the greenhouse (mostly closed) so I left it totally open today. Peak temp is still 40C, and its still spring.
May 28, 2025 at 3:29 AM
Now, this one is a bit painful. Yesterday I got temperatures around 45C in the greenhouse (mostly closed) so I left it totally open today. Peak temp is still 40C, and its still spring.
Spectral profile, using an improved normalization. Basically, first divide by sum over time, then divide by sum over channels.
May 28, 2025 at 3:24 AM
Spectral profile, using an improved normalization. Basically, first divide by sum over time, then divide by sum over channels.
Today's light curve in the greenhouse. What the heck is that jump near the end there?!
May 28, 2025 at 3:20 AM
Today's light curve in the greenhouse. What the heck is that jump near the end there?!
Well, I got a bunch of sensors and took data once a minute in my greenhouse. It gets up to 110F in there during the day! It's mostly driven by the light - when a cloud passes, the temperature drops within a minute or two.
The color of the light does some interesting things too...
The color of the light does some interesting things too...
May 27, 2025 at 3:58 AM
Well, I got a bunch of sensors and took data once a minute in my greenhouse. It gets up to 110F in there during the day! It's mostly driven by the light - when a cloud passes, the temperature drops within a minute or two.
The color of the light does some interesting things too...
The color of the light does some interesting things too...
Crude impedance spectroscopy with a Raspberry Pi Pico. Really crude. Drive the sample with square waves at different frequencies via PWM, Fourier transform, divide, and hope. Not supposed to put a capacitative load on GPIO pins? Uh, well nothing happened, its fine right?
May 13, 2025 at 4:15 AM
Crude impedance spectroscopy with a Raspberry Pi Pico. Really crude. Drive the sample with square waves at different frequencies via PWM, Fourier transform, divide, and hope. Not supposed to put a capacitative load on GPIO pins? Uh, well nothing happened, its fine right?
Got it up to 360kHz. The first thing I measured is evidently the effect of the DMA retrigger/wifi packet sending/??? on the voltage of an unconnected pin. This is averaged over 1000 cycles.
May 4, 2025 at 1:27 PM
Got it up to 360kHz. The first thing I measured is evidently the effect of the DMA retrigger/wifi packet sending/??? on the voltage of an unconnected pin. This is averaged over 1000 cycles.
Added a bit of vital wheat gluten and the dough was strong enough I could basically stretch it very thin by hand (and then deepfry it). Its more chewy than crispy in the end, but good.
March 3, 2025 at 2:02 AM
Added a bit of vital wheat gluten and the dough was strong enough I could basically stretch it very thin by hand (and then deepfry it). Its more chewy than crispy in the end, but good.
Ok went back and found this stuff, calling it 'clusters' was way too generous. I'm pretty sure this is arbitrarily thresholded so I could do reverse Cuthill-McKee to diagonalize though.
February 19, 2025 at 6:38 PM
Ok went back and found this stuff, calling it 'clusters' was way too generous. I'm pretty sure this is arbitrarily thresholded so I could do reverse Cuthill-McKee to diagonalize though.
Okay, I did the thing. Sort of works, but its slow, and there's a diagonal attachment that isn't supposed to happen here. Periodic boundaries for ease of programming.
January 23, 2025 at 9:20 PM
Okay, I did the thing. Sort of works, but its slow, and there's a diagonal attachment that isn't supposed to happen here. Periodic boundaries for ease of programming.
A roasted beet thing because why not, it was sunny out.
January 22, 2025 at 3:00 AM
A roasted beet thing because why not, it was sunny out.
The chain of thought if you provide information and then ask about it retroactively explains the action as 'because it's a sensitive topic'. This is from a local instance of the 7B distilled model.
January 20, 2025 at 10:15 PM
The chain of thought if you provide information and then ask about it retroactively explains the action as 'because it's a sensitive topic'. This is from a local instance of the 7B distilled model.
Well the output here is actually pretty good now so I wouldn't say unreadable. But like, this feels like a glimpse into the mind of someone with social anxiety and a need to please (or, I suppose, a PR committee).
January 20, 2025 at 4:58 PM
Well the output here is actually pretty good now so I wouldn't say unreadable. But like, this feels like a glimpse into the mind of someone with social anxiety and a need to please (or, I suppose, a PR committee).
I think I got it right this time.
January 5, 2025 at 1:10 AM
I think I got it right this time.
Borscht soup pierogi, made by freezing the soup, wrapping, boiling, and pan frying.
That said, I've yet to get the trick of heating it enough to melt the soup without also bursting the pierogi.
That said, I've yet to get the trick of heating it enough to melt the soup without also bursting the pierogi.
December 18, 2024 at 3:28 AM
Borscht soup pierogi, made by freezing the soup, wrapping, boiling, and pan frying.
That said, I've yet to get the trick of heating it enough to melt the soup without also bursting the pierogi.
That said, I've yet to get the trick of heating it enough to melt the soup without also bursting the pierogi.
I got some liquid nitrogen to experiment with in the kitchen. This is a pandan whipcream 'meringue', like an oversized dippin' dot. Base still needs adjustment (it went to the bitter side, and I'd like to keep the aeration but include more pandan flavor)
December 14, 2024 at 8:31 PM
I got some liquid nitrogen to experiment with in the kitchen. This is a pandan whipcream 'meringue', like an oversized dippin' dot. Base still needs adjustment (it went to the bitter side, and I'd like to keep the aeration but include more pandan flavor)
I then tried to modify the old textual inversion and the new one with obvious adjectives like 'red'. The image on the left is 'red lissajous' with the original embedding (disconnected), while the image on the right is 'red lissajous' with the new embedding.
November 22, 2024 at 7:00 PM
I then tried to modify the old textual inversion and the new one with obvious adjectives like 'red'. The image on the left is 'red lissajous' with the original embedding (disconnected), while the image on the right is 'red lissajous' with the new embedding.
However, img2img and ControlNets still work, so I could generate my own 'Lissajous cat' images - I made seven of these, added them to the training set with captions, and retrained.
November 22, 2024 at 7:00 PM
However, img2img and ControlNets still work, so I could generate my own 'Lissajous cat' images - I made seven of these, added them to the training set with captions, and retrained.