You’ve probably heard about OpenClaw, the AI framework that blew up on GitHub a few weeks ago and was just acquired by OpenAI.
I tested connecting OpenClaw to Gladys to try out the possibilities, and it’s really impressive
I tell you more in the video:
Note: I don’t recommend installing OpenClaw on your Gladys server; it’s still early-stage software that touches a bit of everything and has been widely criticized for security flaws. For this test, I deployed OpenClaw on a VM in the Cloud to stay in an isolated environment
I think it’s a good video / a good title to attract people to Gladys, nice job
Personally I wouldn’t go near OpenClaw ^^ too intrusive for me!
One small remark, especially if you keep making videos: I find your mic’s sound very muffled. I watched the video at home on my TV (which has 5.1 sound) — there was A LOT of midrange and almost no highs, so the audio wasn’t very clear; I’d already noticed that on other videos.
Either you need to adjust your mic… or plan on getting a higher-quality mic in my opinion
That’s the goal! I’m going to continue with lots of topics in the same style, but always within different algorithmic bubbles from the connected home, to reach lots of people
You’re not the first to complain about it, but I admit I’m a bit stuck on what to do, I have a Rode SmartLav+ mic :
Then in editing, I boost the sound via this option in my ScreenStudio software, which normalizes the volume and I find it boosts the sound well :
Personally, I don’t think it comes from the room — what matters above all is the microphone.
And in the end I suspect your software, when normalizing the level, doesn’t just do that but also over-compresses the sound. This causes all the frequencies to be buried in the midrange.
I can’t tell you how to fix that since I don’t know this software or its settings, but in my opinion that’s where the problem lies.
For the previous video, that was fine for me. But for this one, I don’t know if you changed your settings, but the sound was really muffled; I had to strain to hear.
I had to restart the beginning of the video three times before understanding that they were clips, because I couldn’t understand what you were saying.
I spoke with Claude about the problem, and according to him, it doesn’t come from the microphone itself, but from the « complete audio chain », he says:
Your SmartLav+ plugged directly into the Mac goes through the Mac’s jack, which is a basic audio input with a mediocre preamp and low gain. Result: weak signal, you have to boost it in post and you bring up the noise with it.
With the Wireless Go, the lavalier’s signal first goes through the Rode transmitter (which has its own quality preamps), then through the USB receiver which is itself a real audio interface. You get a clean, strong, well-processed signal from the start — without needing to boost it in ScreenStudio.
So it’s less the microphone that changes, it’s the quality of the conversion and amplification of the signal. That’s often where it all comes down to.
On his advice, I therefore ordered a Rode Wireless Go III:
This will allow me to have real audio recorded through a proper, clean audio interface.
For the microphone, either I’ll keep the SmartLav+ (connected to the Wireless Go), or I’ll use the microphone built into the Go transmitter — I’ll run tests!
I receive it tomorrow, just in time to be ready for my next video recording on Monday
I’m sure it’ll make all the difference! Whether it’s with your current mic or via the module.
That way, in your editing software you won’t have to touch that setting anymore (or even if you had to normalize, there wouldn’t be much to change).
