Formerly https://lemm.ee/u/romkslqusz
I don’t know that PhotoRec was the ideal place to start for this case, as it is file signature based and will scrape for an insane number of results.
I recommend that you give DMDE a shot, run a full scan of your disk and then have a look through the FS Reconstruction with deleted files included.
Yes
Full quote:
One of the overlay’s most interesting quirks is the developer’s recommendation to run two GPUs. Dedicating one GPU to just running the CRT emulation GPU shader allegedly eliminates a lot of desync issues in most games compared to running everything on one dedicated graphics card (even if that graphics card is very quick, apparently). Luckily, the GPU shader portion is extremely lightweight, and even an old Intel HD 770 can allegedly run the shader at well over 800 FPS at 1080p. This will be an annoying quirk gamers with no integrated graphics chips will have to deal with (particularly AM4-based Ryzen users)
The recommendation is not necessarily to run two dedicated GPUs, it is to have an additional GPU (including iGPU) to dedicate to running the shader.
Use of the word “dedicate” made that a bit confusing since that is common terminology for discrete GPUs
OP has a 60Hz monitor
Friend has a 240Hz monitor and can’t enjoy a game unless it says 240 in the corner
/s
That’s about 3 hours 15 minutes per day or 25 hours a week.
I’m sure this will eventually reach a point of actually enshittification. For the time being, it seems like they’re making the highest bandwidth users - who have a part time job’s amount of time to play video games - pay more for their share.
It is an NPN, It isn’t flipping when base is pulled to ground / is behaving as expected.
When using NPN, the problems were related to the resistor. I managed to find a “sweet spot” at 330ohms where the module is able to turn on and play without cycling and the resistor sits at a comfortable enough temperature.
I’m going to try timing plugging in the battery so that the resistor doesn’t shed all of its power before the gift is opened :)
Probably, I used the term “Tripwire” as an example with relation to the circuit, in this application the intention is that a wire will be cut with snipe.
I don’t think it’s possible place the full module between 4V and Drain (or collector for NPN), so I shifted my focus to just the speaker without all the ICs.
The good news is that it works with both an NPN or a MOSFET - the bad news is that it totally degrades the quality of the audio to a point where the thing I am trying to play is unintelligible.
With how I’ve been going about this, technically, the load of U2 / Speaker Module would be on the Drain side of Q3 PMOS rather than the source side - and that’s the key difference between what I’m doing and the Piezo speaker tripwire circuits.
Could this be the reason I’m not seeing success?
Appreciate your response / effort! The first diagram that comes up in my post is one I drew of what I’ve been implementing, you can sub the NPN Transistor for a MOSFET to cover attempts with that component type. Base to Gate, Drain to Collector, Source to emitter.
For all three of your diagrams, SW2 exists on the module PCB but it is being bypassed by my circuit. I have the resistor / drain / collector wired to “1” on SW2, “2” is the load of the rest of the Module (in my diagram, “2W Speaker and some ICs”)
The blue wire is my “Tripwire” / pulldown to GND. In the last picture, I haven’t even bothered attaching it to pull down the signal from Gate because I had been experimenting with different MOSFETs and finding they weren’t turning on all the way / letting any more than 2V through, which is not enough to power the speaker module.
Nope! Mostly do computer repair / SMD rework, I’ve also tweaked some existing circuits to do what I want so this is probably my first foray into really tinkering and building something.
That’s where I think I’m having the biggest issue. I’ve been experimenting with different resistors, 20Ω is the only value so far that doesn’t result in the module power cycling the resistor starts to cook, I think it is rated for 1/8W and I could go with a higher wattage rating but I don’t want to drain the 300mAh battery too fast
Seems like you really just don’t like Persona.
This isn’t Persona.
As far as I understand it, the comparisons pretty much end at the combat system / gameplay mechanics.
Without giving away too much, you have a world where, every year, everybody of a certain age is erased. Every year, those with one year left to live set out to try and stop that from happening - and for ~77 years, none of them have returned.
This sets the stage for exploration of grief, loss, and associated trauma. In most games, there’s death everywhere but the emotional side is relegated to a 3 minute scene with sad piano music before the characters get back to the action. In this game, they drill a lot deeper and it really makes the characters come alive.
They’ve nailed the blend of sadness, joy, and even comedy.
This is all then set in a backdrop of some of the most visually interesting environments ever presented in a videogame with a completely insane musical score that brings all of those moments to life, the game is effectively a frisson machine.
Yup, that’s the full scan. Your quick scan results were chill enough that it wasn’t necessarily a bad idea to scan further, I usually stop a scan after 5-10 dead sectors because I know I’ll be replacing the disk.
You’ve got 47 dead / completely unresponsive sectors and quite a few that are slower to respond. Overall, it’s good that you don’t have a ton beyond the 250ms range.
I should have asked, is this disk making any clicking / grinding / scraping noises?
Before proceeding, I would prioritize manual copy of any critical / irreplaceable data. You mentioned using this disk for Program Files - Program installers can usually be sourced and rerun, so I would consider that low priority.
When it comes time to clone the disk, those dead sectors might pose a problem with cloning software. I know that, by default, Macrium Reflect and Clonezilla stop the clone on unresponsive sectors. Both have options to override that, leaving behind bad data.
If the disk is not clicky / scrapey / grindy, you could run chkdsk /r X: to have your filesystem adjusted and reallocate data away from problem areas before running a clone operation.
Please note that continued operation of a failing disk yields continued degradation. In cases of active / rapid failure, running a long filesystem repair operation like chkdsk can actually work against you.
I use OpenSuperClone, a Linux based advanced cloning tool, for most failing disk cases that come through my shop. It’s probably overkill for the current condition of your disk and there’s going to be a learning curve if you choose to go that route, but the main advantage is that it grabs all the “easy” sectors before working on problem areas in order of increasing difficulty. I would have been recommending this route if you were to have had a high count of sectors taking longer than 500ms to respond (green / orange / red). That recommendation would have been accompanied by the disclaimer “if you can’t afford to lose this data, hire a professional”
Everything I described relates mostly to motherboards for custom builds (ASUS, MSI, Gigabyte, etc)
Dell doesn’t have an “Advanced Mode” and profiles are limited to one “save as custom user settings” option.
Unless something has changed with their absolute latest models, their UEFI usually follows a “tree” layout that is super navigable with a keyboard. Arrow keys, spacebar, and the TAB key ought to get you everywhere you need to go - the latter being necessary to jump from the tree on the left over to managing the different settings on the right.
The SMART table is not necessarily an accurate representation of the disk’s condition - I get a lot of disks that come through with a “GOOD” condition yet they’re on the brink of data loss.
Disks in a caution state should have replacement and data recovery prioritized.
The appropriate approach to data recovery will vary based on the condition of the disk.
A good place to start would be to download the free Victoria disk diagnostic software, run the “quick” surface scan, and post the result here. Note that any disk interaction / usage during the test will negatively bias the result.
As far as replacement is concerned, provision for whatever falls within your budget plus a second disk to use for periodic backups.
In Windows, I use two disks in a mirrored storage space to tolerate one disk failure, plus an offsite backup.
In the OEM / Laptop / Prebuilt space I’m definitely seeing move of a pivot towards graphical UEFI, that ones with forced mouse navigation are currently the exception but when I come across it I hate it.
I own a repair shop so I’m sure my perspective comes from spending an unusual amount of time in UEFI/BIOS over the course of the years, but I find that they remain perfectly usable with a keyboard.
Most modern UEFI have an “advanced mode” (usually F7) that is a lot closer to traditional BIOS layout and navigability. I actually get unreasonably bothered when trainees insist on fumbling around with the mouse in “Easy Mode”
There’s so many options in modern UEFI that there are two features I’m actually incredibly grateful for:
The “Favorites” systems, so that you can have all of the settings you’ll actually change all in one place
The “Profiles” systems, so that you can easily hop between configurations
Without those, the onion meta-game of finding options that are arranged differently on AMD vs Intel or between different motherboard vendors is a rough time.
Some users with AMD graphics cards get a hard shutdown during the cutscenes immediately after boss fights.
I played through side by side with my partner, they had to repeat a difficult boss fight several times before they eventually just started skipping post-boss cutscenes on their system and watching them on mine.
The Lantern Boss
Bummed that this is a Switch exclusive, my partner and I live the Hazelight games
I can agree with some of the parts about how some humans can be really annoying but this mostly reads like AI propaganda from someone who has deluded themselves into believing an LLM is actually any good at critical thought and context awareness.