My Theory on the MLT04 sensor..

[tp4tissue]

There are many Users Claiming, that the MLT04 feels the Most-Raw among all mouse sensors..


I don't think this has anything to do with the sensor itself..

I think this has to do with the fact, that people who make this claim are USING the WMO which is lighter than almost all of the other --perfect--sensor-- mice.



The light weight makes the perception of direction change less, because you carry 20% less momentum than say a G400..

This lower momentum may be responsible for the Raw-Feel (but only RELATIVE to heavier mice), and such perception has nothing to do with the Sensor output itself.. 

[munch]

I think so too. I feel lighter mice feel snappier.
although it does not have nearly as much processing as more advanced sensors, so there might be something to that too. since IME3.0 isn't that light.

[tp4tissue]

I think so too. I feel lighter mice feel snappier.
although it does not have nearly as much processing as more advanced sensors, so there might be something to that too. since IME3.0 isn't that light.

It's hard to say whether their perceptions carried over to the IME or FROM the ime,..

[Bucake]

people claim it about the IMO and IME as well, don't they? those mice aren't that light.
and the raw feel is not completely a myth. i'm sure you know that anti-jitter smoothing is a real thing in many mice.
i also bet you've seen the test results, and the MLT04 has great scores overall, except for perfect tracking speed and malfunction speed.

i don't think it's just nostalgia or weight, it's probably just all the tiny things added up. low jitter, low delay, low angle snapping, low variance, etc.
and you can't underestimate stuff like SROM and firmware, maybe microsoft did it all just right in that regard as well.

i assume that the escalating "DPI war" might have contributed a lot to "worse feeling sensors", because it made the companies invest in funny tricks to bump up the CPI to crazy levels.

the MLT04 is certainly not the only 'snappy' feeling sensor, but most other sensors had stuff going on like undeniable jitter, prediction or variance.
the A2020 and S2020 might've been more liked if they had less prediction, the S3688 if it wasn't for the jitter, laser sensors if it wasn't for the variance, etc.
another good example would be PTE, i guess.


but you definitely have a point. shape, weight, coating, etc, it all matters. the S3095 would probably have been more hyped if it didn't only come in the heavy, ergonomic G400. some other sensors might've been a lot more popular if they had been implemented better. Logitech is known for doing that well, and maybe Microsoft was good at it too.

[tp4tissue]

I have not seen any scientific tests on part of the consumer besides G4m3rboyExtreme running mouse tester and mouse-rate tester.. the two things ever reported..

I don't think prediction is the issue here, because it's always guessing on SOME part of the incoming data..

[Bucake]

yeah, that's my understanding, and also why i said low angle snapping, and not no angle snapping.
if you try a A2020 / S2020 mouse, you feel and see your cursor snap to and stick to angles. not everyone finds that it feels very raw when this happens.

and this is also where srom / firmware is important (algorithms and such). with the MLT04, Microsoft or ST Microelectronics might've done it in such a way that feels 'just right' for a lot of people.

but what's wrong with those "non-scientific" tests?
they might not have been professional, but that does not mean they are invalid or useless. stuff like stable polling rate, motion delay, jitter, variance can all be seen.
i'm not claiming that these tests or their results are perfect, but they are certainly reliable to a degree.


i just don't believe it's just the weight.

edit: to be clear, i'm not claiming it's the best sensor or anything. i just think that this is more or less why most people claim it feels so raw.

[woll3]

Few things coming together, hoewever due to its fixed framerate and how it does Delty X and Y at once it is actually at a disadvantage at low speeds, it can actually even be tricked into going the wrong direction. among other things. Maybe i whip mine out later just to fumble around, despite the signed driver i have no use for it as i cant just be like "can i install this driver please", there has already been an outcry about the timer tool once so i rather avoid any accusations, not to mention that when it bugs out the USB Port can become useless until you get rid of the changes again.

ST Microelectronics might've done it in such a way that feels 'just right' for a lot of people.

It was ST, MLT04 is actually a SoC system that comes straight out of the factory, just solder it, done, and just to mention it, it has no impact on latency. Guy who did MLT04 actually went on to Avago, dont know the exact time frame though.


Also some other stuff i wanted to write but ive forgotten it.

[tp4tissue]

Few things coming together, hoewever due to its fixed framerate and how it does Delty X and Y at once it is actually at a disadvantage at low speeds, it can actually even be tricked into going the wrong direction. among other things. Maybe i whip mine out later just to fumble around, despite the signed driver i have no use for it as i cant just be like "can i install this driver please", there has already been an outcry about the timer tool once so i rather avoid any accusations, not to mention that when it bugs out the USB Port can become useless until you get rid of the changes again.

Hrrrm....

So you're saying the MLT04 is doing  -1x-1  while the new sensors are doing -1x0 n 0x-1

?

So that reduces the diagonal motion to 1 count ? vs 2 counts ?

[woll3]

Hrrrm....

So you're saying the MLT04 is doing  -1x-1  while the new sensors are doing -1x0 n 0x-1

So that reduces the diagonal motion to 1 count ? vs 2 counts ?

Basically yes, but going XY separately is actually the natural thing for a 4 sided matrix.

[tp4tissue]

Hrrrm....

So you're saying the MLT04 is doing  -1x-1  while the new sensors are doing -1x0 n 0x-1

So that reduces the diagonal motion to 1 count ? vs 2 counts ?

Basically yes, but going XY separately is actually the natural thing for a 4 sided matrix.

Hrrrmmmmmmmmm.....

[tp4tissue]

How can the MLT04 be tricked to go in the wrong direction? does it need to be rotated?

[munch]

Few things coming together, hoewever due to its fixed framerate and how it does Delty X and Y at once it is actually at a disadvantage at low speeds, it can actually even be tricked into going the wrong direction. among other things. Maybe i whip mine out later just to fumble around, despite the signed driver i have no use for it as i cant just be like "can i install this driver please", there has already been an outcry about the timer tool once so i rather avoid any accusations, not to mention that when it bugs out the USB Port can become useless until you get rid of the changes again.

ST Microelectronics might've done it in such a way that feels 'just right' for a lot of people.

It was ST, MLT04 is actually a SoC system that comes straight out of the factory, just solder it, done, and just to mention it, it has no impact on latency. Guy who did MLT04 actually went on to Avago, dont know the exact time frame though.


Also some other stuff i wanted to write but ive forgotten it.

I think I've had that happen a few times with my WMO... cursor would move the opposite direction when trying to do pixel precise shots.
not sure if that's what you mean but kinda sucked for railgunning in quake.

[woll3]

How can the MLT04 be tricked to go in the wrong direction? does it need to be rotated?

No, but it is somewhat surface dependent, i blame the fixed Framerate for it, which is also responsible for the slight deadzone.

not sure if that's what you mean

Indeed it is.

[Elrick]

Indeed it is.

TRUE the MLT04 is entirely dependent on the right mouse pad and that will determine how the sensor operates at any given moment.

Remembered one professional Gamer using his WMO for years but he has the BIGGEST stock pile of mouse pads that he carry's to every match.  It makes a world of difference between standing on the podium to receive your trophy or crying in the alleyway, alone and embarrassed to show his face again.

He actually still uses his WMO simply because he knows how it will perform with every mouse pad in his collection.

[tp4tissue]

Indeed it is.

TRUE the MLT04 is entirely dependent on the right mouse pad and that will determine how the sensor operates at any given moment.

Remembered one professional Gamer using his WMO for years but he has the BIGGEST stock pile of mouse pads that he carry's to every match.  It makes a world of difference between standing on the podium to receive your trophy or crying in the alleyway, alone and embarrassed to show his face again.

He actually still uses his WMO simply because he knows how it will perform with every mouse pad in his collection.

which pro gamer is this..  and why don't I remember him,   clearly he is not very good.. hell, even YOU don't actually remember him...

[Elrick]

which pro gamer is this..  and why don't I remember him,   clearly he is not very good.. hell, even YOU don't actually remember him...

Show Image

It was here in West Australia during the mid to late 1990s, hence full-time professional gaming didn't exist just yet but since he kept earning those prizes it would make him a semi-professional as opposed to an outright amateur?

There had been the rage for LAN gaming at that time and I was at WAIT when it eventually converted over to Curtin Uni and he was really good.  Mostly saw him carry his gear to and from the meetings and he only had a plain beige WMO, the prizes he won were anything from the standard type trophies, Cash, PC keyboards/mice, a couple of actual PCs and a motor scooter (which was cool when you were a kid).

Sorry for not knowing his name but this is WA and here you don't ask for names,  plus I don't have that feminine part of wanting to know everyone's business or concern's because you live longer that way.

[iLLucionist]

In one of the logitech videos, they explain that increasing DPI is like zooming in in a picture. It gives more detail, but also more noise.

Isn't it just the case that the MLT04 feels so perfect because the DPI is so low, relatively to newer mice, that no irregularities can occur?

[Bucake]

having only one (low) cpi step surely makes things easier. for one, as you suggested there will be less need to come up with ways to battle excessive noise.
that's why i mentioned the "dpi war" earlier: companies going for interpolation, getting more noise/jitter by pushing cpi, implementing smoothing to counter it..
i can imagine that with all that processing going on, it becomes harder/impossible to get cursor movement that's low-latency, low-jitter, and accurate.
the main issue is probably just that if all this stuff is needed at all to achieve high cpi, and if high cpi comes with such high a cost.. then the sensor probably just isn't suited. (3360 or go home!)

things were just simpler in the beginning, i suppose. why bother trying to come up with a way to get 800cpi on your mouse when everyone is running at resolutions like 800 and 1024, you know?
and of course the fact that those mice were optical, and not mechanical, was reason enough to get a mouse like it.

few of the companies could/can really handle (the pace of) the dpi war, in my opinion.
rather than trying to come up with ways to get decently high cpi while cursor movement remains "sharp", they prioritized maximum cpi and other "gimmicks" (lighting, more buttons, etc).
in my opinion the high cpi came with too high a cost. but luckily for these companies, the average consumer doesn't care about or even see that cost.
(i realize that cursor precision isn't the main concern for every kind of mouse, but of course i'm primarily talking about gaming mice.)

edit: also, the gaming peripheral market has become quite a silly one over the years.. for example the 3310 received a huge hype-bomb, which never really stopped exploding, and now there are so many people thinking and claiming that it's such an amazing super fantastic sensor, that every new mouse that uses this sensor automatically gets 4 out of 5 stars.
i can not un-forget that TekSyndicate themselves claimed it's the best sensor on the market, after the 3360 was already publicly available.
you could argue that them claiming that is just marketing, but i consider it laughable.
(but finalmouse is probably the funniest company.)


/rant
numbers sell


edit: again thanks woll3, the insight and detail is always appreciated :-)

[munch]

I don't think 3310 is a bad sensor though.
many of the 3310 mice that I've tried are more pleasant to use than many 3090 ones.
and just cause of the sensor feel, I'd also pick 3310 over MLT04. but smoothing and such was fairly noticeable on a few of them which did make it less desireable.

[iLLucionist]

having only one (low) cpi step surely makes things easier. for one, as you suggested there will be less need to come up with ways to battle excessive noise.
that's why i mentioned the "dpi war" earlier: companies going for interpolation, getting more noise/jitter by pushing cpi, implementing smoothing to counter it..
i can imagine that with all that processing going on, it becomes harder/impossible to get cursor movement that's low-latency, low-jitter, and accurate.
the main issue is probably just that if all this stuff is needed at all to achieve high cpi, and if high cpi comes with such high a cost.. then the sensor probably just isn't suited. (3360 or go home!)

things were just simpler in the beginning, i suppose. why bother trying to come up with a way to get 800cpi on your mouse when everyone is running at resolutions like 800 and 1024, you know?
and of course the fact that those mice were optical, and not mechanical, was reason enough to get a mouse like it.

few of the companies could/can really handle (the pace of) the dpi war, in my opinion.
rather than trying to come up with ways to get decently high cpi while cursor movement remains "sharp", they prioritized maximum cpi and other "gimmicks" (lighting, more buttons, etc).
in my opinion the high cpi came with too high a cost. but luckily for these companies, the average consumer doesn't care about or even see that cost.
(i realize that cursor precision isn't the main concern for every kind of mouse, but of course i'm primarily talking about gaming mice.)

edit: also, the gaming peripheral market has become quite a silly one over the years.. for example the 3310 received a huge hype-bomb, which never really stopped exploding, and now there are so many people thinking and claiming that it's such an amazing super fantastic sensor, that every new mouse that uses this sensor automatically gets 4 out of 5 stars.
i can not un-forget that TekSyndicate themselves claimed it's the best sensor on the market, after the 3360 was already publicly available.
you could argue that them claiming that is just marketing, but i consider it laughable.
(but finalmouse is probably the funniest company.)


/rant
numbers sell


edit: again thanks woll3, the insight and detail is always appreciated :-)

Good insight. But I do find the 3310 amongst the best I have ever tried. Except for perhaps the G502 sensor. But I won't touch the fugly G502. Waiting for the sensor to appear in other mice.

[Bucake]

i never considered the 3310 to be bad, it had its obvious strengths. but it never was the best sensor, and imo just didn't deserve that huge hype
the MLT04 also has its downsides, including low cpi, as well as the less apparent downsides woll3 mentioned

didn't mean to sound like a 3310-hater or MLT04-lover :-P

[iLLucionist]

i never considered the 3310 to be bad, it had its obvious strengths. but it never was the best sensor, and imo just didn't deserve that huge hype
the MLT04 also has its downsides, including low cpi, as well as the less apparent downsides woll3 mentioned

didn't mean to sound like a 3310-hater or MLT04-lover :-P

You crushed my soul! Nah.. joking.

What sensor do you consider better than 3310 or MLT04?

[Elrick]

What sensor do you consider better than 3310 or MLT04?

3366 sunshine  .

[iLLucionist]

What sensor do you consider better than 3310 or MLT04?

3366 sunshine  .

How only will it take before it appears in mice other than G502? I'm waaaiiitinnng...

[tp4tissue]

What sensor do you consider better than 3310 or MLT04?

3366 sunshine  .

How only will it take before it appears in mice other than G502? I'm waaaiiitinnng...

Just get g303.. hhahahahaha.. it is theoretically ideal assuming you don't get a rattler..

But even if you do get a rattler, just put some double sided tape on it.. there's really nothing seriously wrong with it..

[iLLucionist]

What sensor do you consider better than 3310 or MLT04?

3366 sunshine  .

How only will it take before it appears in mice other than G502? I'm waaaiiitinnng...

Just get g303.. hhahahahaha.. it is theoretically ideal assuming you don't get a rattler..

But even if you do get a rattler, just put some double sided tape on it.. there's really nothing seriously wrong with it..

If I'd get a rattler, I'll probably throw the mouse at my wall.

[Bucake]

the finalmouse scream keeps getting delayed. planned release was originally set to end februari - early march, but it has been pushed back to late may. i wouldn't be surprised if it gets delayed even further though.
anyway, the steelseries rival 700 was aimed for "spring 2016" iirc. and there is that corsair m65 mouse, but i haven't followed that one at all

other than that, nothing has been announced yet.. i bet more mice with the 3360 are in development, but the fact that nothing has been announced yet might just mean that it will take months before those would get released.

edit: i actually just purchased a used G303, bring on the rattling! it should be here within 2 days.
cost me just over €35 including shipping. seems to be in excellent shape judging by the pictures

[iLLucionist]

the finalmouse scream keeps getting delayed. planned release was originally set to end februari - early march, but it has been pushed back to late may. i wouldn't be surprised if it gets delayed even further though.
anyway, the steelseries rival 700 was aimed for "spring 2016" iirc. and there is that corsair m65 mouse, but i haven't followed that one at all

other than that, nothing has been announced yet.. i bet more mice with the 3360 are in development, but the fact that nothing has been announced yet might just mean that it will take months before those would get released.

All I hope for is an Avior 7000 with 3360 sensor. That would be my dream mouse... basically a "new" WMO / IM3.0

edit: i actually just purchased a used G303, bring on the rattling! it should be here within 2 days.
cost me just over €35 including shipping. seems to be in excellent shape judging by the pictures

Please tell me when you have it if you have teh rattle! So G303 has 3360 as well?

[Bucake]

haven't i already answered that 2 times? :-D could also google(!), but yes dude the G303 has a 3366, but the main concern is indeed chance for sensor rattle.
should be an easy fix, but you might end up needing a new set of feet, kind of a bummer.

[iLLucionist]

haven't i already answered that 2 times? :-D yes dude the G303 has a 3366, but the main concern is indeed chance for sensor rattle.
should be an easy fix, but you might end up needing a new set of feet, kind of a bummer.

Sorry... you did. I should've paid more attention.

Meh... feet are easily replaceable. But dat rattle. Not sure how to fix it myself. I'm not really handy with those things. Or soon I will... heheh

[Bucake]

handy? you unscrew like 5 or 6 screws, apply some tape, then put the screws back in.

this brother shows where the screws are located and everything. ezpz!:

edit: ugh. why are youtube videos centered?

[iLLucionist]

handy? you unscrew like 5 or 6 screws, apply some tape, then put the screws back in.

this brother shows where the screws are located and everything. ezpz!:

edit: ugh. why are youtube videos centered?

Thanks! Dat mouse looks so sweet.. gotta get me one of those.

[tp4tissue]

I've peeled the newer logitech mousefeet... They're a layer of teflon, then this foam layer, then the adhesive layer..

So, the Primary issue is , if you peel it, and peel into the foam layer, as that can't be stickied back together..

So........... all you need to do is use a boxcutter, and make sure you're lifting up the adhesive layer and not the middle which is the foam layer..

[iLLucionist]

I've peeled the newer logitech mousefeet... They're a layer of teflon, then this foam layer, then the adhesive layer..

So, the Primary issue is , if you peel it, and peel into the foam layer, as that can't be stickied back together..

So........... all you need to do is use a boxcutter, and make sure you're lifting up the adhesive layer and not the middle which is the foam layer..

How do you put those bastards back on? Same adhesive? Is it strong enough to use a second time once you peeled them off?

[tp4tissue]

I've peeled the newer logitech mousefeet... They're a layer of teflon, then this foam layer, then the adhesive layer..

So, the Primary issue is , if you peel it, and peel into the foam layer, as that can't be stickied back together..

So........... all you need to do is use a boxcutter, and make sure you're lifting up the adhesive layer and not the middle which is the foam layer..

How do you put those bastards back on? Same adhesive? Is it strong enough to use a second time once you peeled them off?

First , clean table of dust..

Then, wash your own hands.. "thoroughly dry them"..

Then, dedust the mice..

Then, alcohol swab the bottom of the mice

Then, alcohol swab the box cutter/utility knife

Then, insert tip of knife beneath mouse-feet

Then, pry lightly until you have 0.5 cubic cm of area accessible

Then, sticky that to the knife, and pull slowly..

Put  piece of tape against its sticky side for protection..

Set aside..

Do rest of mod.

[iLLucionist]

I've peeled the newer logitech mousefeet... They're a layer of teflon, then this foam layer, then the adhesive layer..

So, the Primary issue is , if you peel it, and peel into the foam layer, as that can't be stickied back together..

So........... all you need to do is use a boxcutter, and make sure you're lifting up the adhesive layer and not the middle which is the foam layer..

How do you put those bastards back on? Same adhesive? Is it strong enough to use a second time once you peeled them off?

First , clean table of dust..

Then, wash your own hands.. "thoroughly dry them"..

Then, dedust the mice..

Then, alcohol swab the bottom of the mice

Then, alcohol swab the box cutter/utility knife

Then, insert tip of knife beneath mouse-feet

Then, pry lightly until you have 0.5 cubic cm of area accessible

Then, sticky that to the knife, and pull slowly..

Put  piece of tape against its sticky side for protection..

Set aside..

Do rest of mod.

Is alcohol a good idea? Coz if shell is ABS... or not if you do it once / quickly?

[Bucake]

usually the adhesive is pretty good and definitely works multiple times. i haven't encountered feet yet that wouldn't go back on after taking them off.
my issue has always been that i rip the layers off of one another, or just bend the thing into uselessness.
why do they all make such terrible mousefeet? :E

also lol @ tissue the professional. i'm lazy and i just take them off without any "prep work".
but he's right though, you should take the extra minute to do it right.
and yeah afaik ABS is resistant to alcohol. some coatings are not though

[iLLucionist]

usually the adhesive is pretty good and definitely works multiple times. i haven't encountered feet yet that wouldn't go back on after taking them off.
my issue has always been that i rip the layers off of one another, or just bend the thing into uselessness.
why do they all make such terrible mousefeet? :E

also lol @ tissue the professional. i'm lazy and i just take them off without any "prep work".
but he's right though, you should take the extra minute to do it right.
and yeah afaik ABS is resistant to alcohol. some coatings are not though

Mionix has these sheets of glidez you can buy. You can always cut it in right shape.

[tp4tissue]

usually the adhesive is pretty good and definitely works multiple times. i haven't encountered feet yet that wouldn't go back on after taking them off.
my issue has always been that i rip the layers off of one another, or just bend the thing into uselessness.
why do they all make such terrible mousefeet? :E

also lol @ tissue the professional. i'm lazy and i just take them off without any "prep work".
but he's right though, you should take the extra minute to do it right.
and yeah afaik ABS is resistant to alcohol. some coatings are not though

Oh right..

Don't touch any rubber coating with alcohol  it will damage the finish..

[iLLucionist]

usually the adhesive is pretty good and definitely works multiple times. i haven't encountered feet yet that wouldn't go back on after taking them off.
my issue has always been that i rip the layers off of one another, or just bend the thing into uselessness.
why do they all make such terrible mousefeet? :E

also lol @ tissue the professional. i'm lazy and i just take them off without any "prep work".
but he's right though, you should take the extra minute to do it right.
and yeah afaik ABS is resistant to alcohol. some coatings are not though

Oh right..

Don't touch any rubber coating with alcohol  it will damage the finish..

tp... you modding mice all the time? seems like you know what you are doing