Logitech MX Master 3S

[tjc]

Hi all,

Thinking about getting this mouse.

https://www.amazon.co.uk/Logitech-Master-Performance-Ultra-fast-Scrolling/dp/B07W5JKHFZ/ref=mp_s_a_1_3_maf_1?keywords=logitech+mx+master+3s&qid=1708009974&sr=8-3

Just wondering if anyone has it or has used. Thoughts on it in general?

Cheers

tj

 

[Patrick Dontigny]

I have 2 of them, one for my home-gaming PC and I have now bought another one for work that I carry between home (teleworking) and work (office). For me, it is the best mouse there is. Movement is precise, buttons give a good feedback, scrolling is the best. I like also that it weights quite a bit. I like heavy mouse. The battery also last a long time between recharges.

I have not tried a tons of mouse, but i have tried pretty much all Logitech mouses, and for me, the MX Master 3s is the best out there.

To each its owns, I am sure some will disagree. There are also tons of reviews out there.

 

[tjc]

Many thanks for your reply Patrick. Much appreciated.

 

[RasmusP]

It's a brilliant mouse. The only cons will be that some report that it becomes faulty too quickly or that it doesn't fit your hand.
But apart from that it's a great mouse.

 

[roPe Gargamel]

Bit expensive but its really great, feeling of smoooothnes is awesome. Had it for few years, no issues.

 

[tjc]

Many thanks for the replies guys. Appreciate the info.

 

[Wreckoning]

I’ve got one, I would say great for ”work” stuff and general PC use but may not be the best shape if you’re wanting to play fast paced games.

 

[Deleted member 197115]

You don't want wireless for gaming.

 

[RasmusP]

You don't want wireless for gaming.

That's not exactly true. Wireless is awesome for gaming, but it needs to be at least 500 Hz polling rate.
I have Logitech G Pro X Superlight and don't game much with it, but it's awesome for CAD, office and anything else.
1000 Hz polling rate via dongle and so light, that it feels like just moving your hand around, without a mouse, but with some buttons.

Very expensive and not a lot of buttons, but hands down the best mouse I've ever used and worth every cent.
I'm using it since 3 years now.
But I had go swap out the mouse wheel. It would sometimes stop between steps and then "randomly" scroll in the wrong direction when starting to scroll.

I also have a cheap Bluetooth mouse and every time I use it, I want to destroy it..

 

[Deleted member 197115]

That's not exactly true. Wireless is awesome for gaming, but it needs to be at least 500 Hz polling rate.

With all due respect, don't be ridiculous.
G502 is cheaper, has better tracking and ergonomics with 1000Hz report rate. If you are half serious about shooters, corded is the only way to go.

 

[RasmusP]

If you are half serious about shooters, corded is the only way to go.

Maybe watch this:

G502 is cheaper

Wasn't part of my argument, but it's cheaper, yes. And weights 121 grams corded, 114 g wireless.
G pro x superlight weights 63 g.
Before testing it, I never thought it would make such a difference.

has better tracking

Source? I'm seeing hero 25k sensor for all three mice (g pro x sl, G502 lightspeed and G502 corded)

and ergonomics

Not for me, a buddy has the G502 and I didn't like it as much. Fingers got a bit cramped.

with 1000Hz report rate

Same for basically all serious gaming mice. But I tested to set it lower and 125 Hz and 250 Hz are a mess in comparison. 500 Hz is almost the same as 1000 Hz but not exactly.

 

[Neilski]

Maybe watch this:

Interesting video.
Am puzzled by this data though, where the latency is measured vs polling rate for a single mouse:

At face value, this data is telling us that the latency improvement with shorter polling intervals is actually better than the reduction in the intervals... (wtf?)
125 Hz = 8 ms, 1000 Hz = 1 ms and yet we save not 7ish ms by going from 125 to 1000, but 10.5 ms. And it's the same at every step in polling rate, so it's hard to believe it's just random wiggles within the statistical variation. We might expect to save 4 ms going from 125 to 250, but we save 4.9. From 250 to to 500 we save not 2 ms but 4 (!?), and finally from 500 to 1000 we save not 1 ms but 1.6.
This is kinda bonkers, and for me it raises questions about the whole dataset.

Spoiler: Statistics... suggests this is not just random fluctuations

Assuming a roughly normal distribution, taking 250 samples with the standard deviations indicated on that graph should mean a standard deviation on the calculated mean of 1/sqrt(250) (about one sixteenth) of the standard deviation of the individual measurements, or roughly 0.18 ms to 0.22 ms from the fastest to the slowest polling rate. Taking two standard deviations brings us to 0.36 - 0.44 ms, and we should expect around 95% of the true means to lie within 2 stdevs of the measured means. Then taking the differences between the means (for say 250 Hz latency minus 500 Hz latency), we just combine those stdevs in quadrature. Tl;dr: it doesn't stack up.

The maximum 2*stdev among the set of differences is for the 125 Hz - 250 Hz case, which I estimate at 0.61 ms, so in 95% of universes that measured latency improvement of 4.9 ms should lie within 0.61 ms of the true improvement, and it's hard for me to imagine why that true improvement wouldn't just be 4 ms. Not exactly a nail in the coffin on its own of course, because a 3-stdev gap will absolutely happen now and then (less than one time in 300 though).

But not a single one of the 6 differences that we can generate (from the 4 means) lies within 2 stdevs of my expected value...

Most significant of all: the 2*stdev of the measured difference between 1000 Hz and 125 Hz is around 0.56 ms, while the actual deviation from the mean difference you might expect based on the difference in polling interval (8 ms - 1 ms = 7 ms) is 3.5 ms. That's 12.5 standard deviations away from what I expected, so this experimental data is clearly totally incompatible with my assumption that the true difference should basically be the difference in the polling interval.
But why is it improving by so much more than "it should"?
Well, either my assumption is broken, or the measurements are broken...

 

[RasmusP]

Interesting video.
Am puzzled by this data though, where the latency is measured vs polling rate for a single mouse:
View attachment 730731
At face value, this data is telling us that the latency improvement with shorter polling intervals is actually better than the reduction in the intervals... (wtf?)
125 Hz = 8 ms, 1000 Hz = 1 ms and yet we save not 7ish ms by going from 125 to 1000, but 10.5 ms. And it's the same at every step in polling rate, so it's hard to believe it's just random wiggles within the statistical variation. We might expect to save 4 ms going from 125 to 250, but we save 4.9. From 250 to to 500 we save not 2 ms but 4 (!?), and finally from 500 to 1000 we save not 1 ms but 1.6.
This is kinda bonkers, and for me it raises questions about the whole dataset.

Spoiler: Statistics... suggests this is not just random fluctuations

Assuming a roughly normal distribution, taking 250 samples with the standard deviations indicated on that graph should mean a standard deviation on the calculated mean of 1/sqrt(250) (about one sixteenth) of the standard deviation of the individual measurements, or roughly 0.18 ms to 0.22 ms from the fastest to the slowest polling rate. Taking two standard deviations brings us to 0.36 - 0.44 ms, and we should expect around 95% of the true means to lie within 2 stdevs of the measured means. Then taking the differences between the means (for say 250 Hz latency minus 500 Hz latency), we just combine those stdevs in quadrature. Tl;dr: it doesn't stack up.

The maximum 2*stdev among the set of differences is for the 125 Hz - 250 Hz case, which I estimate at 0.61 ms, so in 95% of universes that measured latency improvement of 4.9 ms should lie within 0.61 ms of the true improvement, and it's hard for me to imagine why that true improvement wouldn't just be 4 ms. Not exactly a nail in the coffin on its own of course, because a 3-stdev gap will absolutely happen now and then (less than one time in 300 though).

But not a single one of the 6 differences that we can generate (from the 4 means) lies within 2 stdevs of my expected value...

Most significant of all: the 2*stdev of the measured difference between 1000 Hz and 125 Hz is around 0.56 ms, while the actual deviation from the mean difference you might expect based on the difference in polling interval (8 ms - 1 ms = 7 ms) is 3.5 ms. That's 12.5 standard deviations away from what I expected, so this experimental data is clearly totally incompatible with my assumption that the true difference should basically be the difference in the polling interval.
But why is it improving by so much more than "it should"?
Well, either my assumption is broken, or the measurements are broken...

I agree, yep. And I have no explanation either.
His hardware kit measures the latency between the mouse action and the screen showing movement.
It's consistent, repeatable, how nvidia suggests how to do it and what other people also use to measure this stuff.
I guess changing the polling rate isn't only changing the intervals of the PC "checking the USB connection to the mouse", but is changing a bunch of other things.
I have no clue about USB connections, but I can imagine something like a faster polling rate might also get other prioritisations from Windows, which improves the average latency even more.
1000 Hz polling rate = 1 ms for the mouse latency, in theory.
Mouse controller comes on top, no one knows how much it adds.

Then 360 Hz monitor = 2.78 ms in theory.
Add one buffered frame = 5.56 ms.

So we have 6.56 ms in minimum latency, but the real latency is 19.8 ms.

I'd think the part that causes the other ~13 ms gets affected by the different polling rate settings.


In any case though, I can clearly feel the different between 125/250/500 Hz. The difference between 500 Hz and 1000 Hz is very minimal though.
I can also very clearly feel the latency of all BT mice, that I've ever used.
However I absolutely can't notice any difference between my wired G302 Daedalus Apex and the G Pro X Superlight.
Apart from not being massively annoyed by the cable, lol.

 

[Neilski]

So we have 6.56 ms in minimum latency, but the real latency is 19.8 ms.

Yeah, that's a big gap to explain for sure, and I was wondering about that myself.

 

[Deleted member 197115]

Maybe watch this:

Wasn't part of my argument, but it's cheaper, yes. And weights 121 grams corded, 114 g wireless.
G pro x superlight weights 63 g.
Before testing it, I never thought it would make such a difference.

Source? I'm seeing hero 25k sensor for all three mice (g pro x sl, G502 lightspeed and G502 corded)

Not for me, a buddy has the G502 and I didn't like it as much. Fingers got a bit cramped.

Same for basically all serious gaming mice. But I tested to set it lower and 125 Hz and 250 Hz are a mess in comparison. 500 Hz is almost the same as 1000 Hz but not exactly.

Slightly confused, which mice you compare with G502.
OP MX Master 3S, your G Pro X Superlight, or most expensive G502 Lightspeed which is G502 with wireless capabilities?
They all are different profiles, different use cases, general use office mouse vs gaming, different spec, response time, sensor and features.
Even wireless technologies are different BT in MX Master 3S and proprietary Lightspeed in G502 and G Pro X Superlight.
I must admit, on paper new Lightspeed protocol looks impressive, my experience is based on BT and older Logitech wireless protocol with unified receiver, and corded was always beating them all single handedly.
If you compare MX MASTER 3S vs G502, price, features, performance, unless you really need wireless, G502 is a much better choice. For general use though nothing beats convenience of generic BT mouse and keyboard.

 

[RasmusP]

Slightly confused, which mice you compare with G502.
OP MX Master 3S, your G Pro X Superlight, or most expensive G502 Lightspeed which is G502 with wireless capabilities?
They all are different profiles, different use cases, general use office mouse vs gaming, different spec, response time, sensor and features.
Even wireless technologies are different BT in MX Master 3S and proprietary Lightspeed in G502 and G Pro X Superlight.
I must admit, on paper new Lightspeed protocol looks impressive, my experience is based on BT and older Logitech wireless protocol with unified receiver, and corded was always beating them all single handedly.
If you compare MX MASTER 3S vs G502, price, features, performance, unless you really need wireless, G502 is a much better choice. For general use though nothing beats convenience of generic BT mouse and keyboard.

Sorry, didn't get to reply earlier.
This post is a complete different one to the earlier post from you and I simply agree with everything.
I'm a bit confused too about the comparisons though.
My point was just about your general statement, that you'd never want wireless for gaming.
And I agree with that for everything wireless, apart from technologies for gaming, like Logitech's Lightspeed.

The rest of my post was just fanboying about my Superlight and that the G502 wasn't really a comparable mouse to the Superlight, because the super low weight is as impressive (for me), as going from 60 Hz to 120 Hz.
I don't move the mouse, I just move my hand.
But it might be an alternative to the MX Master.
Logitech seems to have quality issues with their scroll wheels though (mine isn't perfect either.. Gonna need to replace it soon).
Which is unacceptable at the price point.
Other brands sadly didn't match with my hand. I would've liked to not buy Logitech.

About BT mice:
There might be good ones, but from the ~20 BT mice that I got my hands on over the years from different people, mostly university, all of them were imprecise and laggy in a way that even navigating Windows was a pain
I'd be curious to try the MX Master via BT though, might be a lot better!

I'd also be very curious to see some LDAT latency comparison of the MX Master with the Unified Dongle against a Lightspeed mouse.
Might be acceptable or horrendous.

 

[Deleted member 197115]

The rest of my post was just fanboying about my Superlight and that the G502 wasn't really a comparable mouse to the Superlight, because the super low weight is as impressive (for me), as going from 60 Hz to 120 Hz.

Funny thing that G502 comes with extra weights you can install to get a steadier aim. So it's not necessarily a bad thing, call it body positive.

 

[RasmusP]

Funny thing that G502 comes with extra weights you can install to get a steadier aim. So it's not necessarily a bad thing, call it body positive.

It's like putting extra weight on your wheel, because you like the "chunky" feeling of the added inertia.
I guess that can feel cool for drifting or cruising, but I prefer my "Carbon fiber formula wheel".
It might make sense for corded mice that aren't meant to be used for fast paced games, since it makes you forget the cable a bit easier.

My superlight feels absolutely awful when plugged in.
And if you take your hands off, the cable will move it around a bit.

 

[Deleted member 197115]

It's like putting extra weight on your wheel, because you like the "chunky" feeling of the added inertia.
I guess that can feel cool for drifting or cruising, but I prefer my "Carbon fiber formula wheel".
It might make sense for corded mice that aren't meant to be used for fast paced games, since it makes you forget the cable a bit easier.

My superlight feels absolutely awful when plugged in.
And if you take your hands off, the cable will move it around a bit.

Depends on what you play and personal preferences, with high DPI weights helps with stability and precision, once you get used to it, there is no way back.
Typically you set up dpi and in game mouse sensitivity to turn 180 degrees with just move of your wrist, heavier mouse allows higher precision and mouse control in that narrow range without additional jitter and overshooting.

Top 7 Heaviest Gaming Mouses in 2023 (Best Picks)

When buying a mouse, one of the most important things to consider is whether it is light or heavy. Lightweight mice are more comfortable to hold, but t

leaguefeed.net

But in the end I guess you can get used and be good with either one.

 

[RasmusP]

That's up to debate and I don't think there's a real right or wrong. A heavy mouse will basically reduce your "hand jitter" by smoothing it out with inertia.
But I'd say a light mouse with a mouse pad that has a bit higher resistance, but no extra initial resistance, will work better since you just get more friction, without adding inertia to the movement.
I have very calm hands and I love that I'm able to move the mouse around with my finger tips. I'm using a mix of palm grip and fingertip grip: Wrist on the mouse pad, hand relaxed onto the mouse. Rear end of it is slightly, but not really touching my palm.
But I can lay my whole palm on the mouse to relax too, it's quite big.

On the Desktop, I can move the cursor from left edge to right edge of my 3440x1440 monitor without lifting my wrist. I guess 32:9 would be the limit with this sensitivity.
Pixel-accurate movements are done by letting my thumb and little finger touching the mouse pad and moving the mouse with just my fingertips.
Most movements aren't via the wrist and more just by moving my fingers.

Got a low to semi friction, cloth pad:
Razer Pro Glide - 36x27.5 cm

When carefully touching the mouse with a fingertip, there's no initial resistance to overcome, just a tiny bit of friction.

I can recommend, when in need for a new mouse or unhappy with the current one, to try this experience. For my friend, who has pretty bad hand jitter, it's awful. But he doesn't play any fast paced games at all. For another friend who mostly plays Apex Legends, it was a game changer. I gave him my Superlight for a week and after a few days he said he could never go back to a heavier mouse.
A third mate was like "You guys have problems... I just play and save money with my corded Razer Deathadder for sub 30€"

But anyway, to "troll" a bit :

The 10 Best Tips to Achieve Perfect Mouse Aim - Overclockers UK

They key to success in any FPS is perfect mouse aim, and with our list of ten tips you'll be sure to never miss your target!

www.overclockers.co.uk

While there’s no easy or short solution, there are plenty of things you can be doing to improve your aim. In fact, we’ve got ten useful tips to help you achieve perfect mouse aim.

Get a Lightweight Mouse​

First things first, let’s take a look at the tools you’re working with. Since you will be using your gaming mouse to aim you want to choose yours carefully. As well as ensuring the shape is comfortable, a lighter weight can be a key difference maker. A lightweight mouse is easier to accelerate and stop, giving you finer control if you adjust your sensitivity accordingly. It’s also less strenuous if you find yourself moving your gaming mouse frequently and extensively.