Discussion of targeting technologies - Page 13 - Overclock.net - An Overclocking Community

Forum Jump: 

Discussion of targeting technologies

Reply
 
Thread Tools
post #121 of 130 (permalink) Old 07-21-2015, 11:18 AM
New to Overclock.net
 
wareya's Avatar
 
Join Date: Jun 2015
Posts: 1,699
Rep: 74 (Unique: 43)
The problem is, he doesn't agree with anything. I make arguments, he tells me I'm wrong. That's it. I give explanations, he says I explain nothing. It's not as simple as mediable disagreement.
wareya is offline  
Sponsored Links
Advertisement
 
post #122 of 130 (permalink) Old 07-21-2015, 12:39 PM
lololol
 
qsxcv's Avatar
 
Join Date: Feb 2014
Posts: 4,243
Rep: 365 (Unique: 152)
Quote:
Originally Posted by wareya View Post

The problem is, he doesn't agree with anything. I make arguments, he tells me I'm wrong. That's it.
/thread
apparently everything we say is groundless and everything he says is completely logical rolleyes.gif

too busy to check forums as regularly
pm me if i forget to respond
qsxcv is offline  
post #123 of 130 (permalink) Old 07-21-2015, 01:12 PM
New to Overclock.net
 
Tivan's Avatar
 
Join Date: May 2014
Posts: 2,628
Rep: 133 (Unique: 101)
Quote:
Originally Posted by Above8 View Post

angle snapping

That's a feature against human error if you ask me.

edit:
Quote:
Originally Posted by wareya View Post

The only reason "gaming mouse" is a thing now is because people are seriously stupid enough to make mice for office users or travellers that have tons of smoothing, angle snapping, and path correction. I learned it the hard way when I bought my Pyra.

Agreed, my 5euro basic mouse from some super market is pretty usable (at 500hz), aside from the low accuracy on short distance movements. (and it's just 1000dpi)
If it had any sort of tech in it to make the sensor less obviously weak, it'd be a lot worse as gaming mouse, but it's just fine the way it is. (sadly overclocking mice isn't as easy as it should be and used to be, also.)

~ Enjoy life, have a great day ~
Yes, I really am that tivvv3 guy from youtube who makes LoL videos by the same philosophy. Whoa!
Also Unconditional Basic Income advocate.
Tivan is offline  
Sponsored Links
Advertisement
 
post #124 of 130 (permalink) Old 07-23-2015, 10:13 AM - Thread Starter
New to Overclock.net
 
Above8's Avatar
 
Join Date: Dec 2013
Location: Russia, Moscow
Posts: 424
Rep: 40 (Unique: 34)
Well, I’ll try to explain what I mean.
Quote:
Originally Posted by wareya View Post

Did you completely miss the part where I laid everything out?

>I’m talking about amount of measurement errors. I hope, no one will claim, that this approach gives perfectly accurate result.

Yeah. It has the same fundamental limitations of any analog -> digital data conversion: quantization. Just like every other sensor that could ever exist.
1. What do you mean by “any analog”?
2. You want to say, that ball mice have same limitations as ICS?
3. By quantization you mean CPI?
Quote:
Originally Posted by wareya View Post

Specifically, the image sensor resolution and bit depth are the technical limitations to ICS. The real world limitation is noise.
4. What noise are you talking about?
5. Electromagnetic interference?
6. Photodiode noise?
7. Signal with frequency higher, than Nyquist frequency?
8. Improper tracking surface?
9. Math errors?
10. Quantization?
Quote:
Originally Posted by wareya View Post

All the "flaws" you claim are easy to compensate for
11. Which flaws exactly do you mean?
12. How can you compensate them?
13. What do you mean by “easy”?
14. With which errors it will be compensated?
15. Will it lead to different flaws?
16. Why do you think it’s possible?
Quote:
Originally Posted by wareya View Post

(DIC in particular is perfectly fine with a lot of things that seem like they should massacre it).
17. What things exactly?
18. What do you mean by “perfectly fine”?
19. Why do you think that these things don’t massacre it?
Quote:
Originally Posted by wareya View Post

However, 36x36 at 7bpp gives us 131072 possible levels of correlation (similarity) on a single correlation cell between black and white images when you merely use 32x32 of the 36x36 (which is both what they do and why the 3366 picked 36x36 instead of a mere power of two).
20. Any proof on 7 bpp?
21. What is this based on?
22. What do you want to say by this?
Quote:
Originally Posted by wareya View Post

This is just the insane theoretical nonsense, though. In the real world, it seems like you're typically going to get around 16384 levels of similarity per cell (taken from one example of real world data of a 1x1 downwards-right movement) on a typical cloth surface.
23. How did you get this number?
24. What do you want to say by this?
Quote:
Originally Posted by wareya View Post

This is more than enough.
25. More than enough for what?
26. What is enough?
27. Is it grounded on anything?
Quote:
Originally Posted by wareya View Post

It's way too much, actually.
28. If accuracy can’t be more than 100%, how can it be too much?
Quote:
Originally Posted by wareya View Post

I guess it's obvious how the 3366 manages to have high internal dpi by default
29. Do you mean hardware resolution or interpolation?
30. What do you mean by high?
31. How DPI is related with tracking accuracy, which we are talking about?
Quote:
Originally Posted by wareya View Post

with negligible jitter.
32. On what CPI?
33. What do you mean by negligible?
34. Do you know it’s exact influence?
35. How do you know, that sensor doesn’t hide jitter by smoothing and/or any other kind of (post)processing/path correction?
Quote:
Originally Posted by wareya View Post

>I hope, it’s obvious, that this technology has a lot of flaws and limitation.

It's not. It doesn't.
36. How can technology have no flaws and limitations?
37. Do you want to say, that ICS has no limitations?
38. What is this statement grounded on?
39. Why did you except Niquist frequency from ICS limitations?
40. Why did you except pixel array size and framerate from ICS limitations?
Quote:
Originally Posted by wareya View Post

Crank up the sensor framerate, resolution, and sensor smallness,
41. How do you know, that it’s possible?
42. How could it lower influence of limitations?
Quote:
Originally Posted by wareya View Post

add in correlation center prediction in order to raise the malfunction speed
43. Why don’t you think it’s possible?
44. How exactly will you do it?
45. Which measurement errors will have this?
Quote:
Originally Posted by wareya View Post

(no, this doesn't add extra artefacts or drift)
46. What is this statement based on?
47. How can prediction add no artifacts?
48. Any proof?
Quote:
Originally Posted by wareya View Post

and suddenly tons of theoretical problems disappear.
49. What do you mean by theoretical problems?
50. How can theoretical problems disappear?
51. Which theoretical problems exactly do you mean?
Quote:
Originally Posted by wareya View Post

Sensor quality is important, though: it limits framerate.
52. How is FPS related with sensor quality?
Quote:
Originally Posted by wareya View Post

A brighter LED can help.
53. How you think brightness of LED can eliminate fundamental flaws?
54. Which limitations can it possibly eliminate?
Quote:
Originally Posted by wareya View Post

I think that's part of why IR was/is popular.
55. Aren’t they popular because of invisibility?
Quote:
Originally Posted by wareya View Post

>Speaking about ICS we aren’t speaking about perfect accuracy. Otherwise there were jitter at all, no need in smoothing, angle snapping, path correction and any other variants of “post processing”.

Jitter happens when you run a sensor above its practical limit of DPI
56. What is practical limit of DPI?
57. Why do you think, that jitter happens only on some CPI settings?
58. Doesn’t jitter happen on level of sensor measurement?
Quote:
Originally Posted by wareya View Post

(as is defined by the image resolution and bit depth... But image grain and power consumption limit how far those can be pushed).
59. What do you want to say by this?
Quote:
Originally Posted by wareya View Post

If you get jitter at low DPIs, the sensor is doing something terribly wrong. Probably implementing DIC incorrectly in order to avoid patent lawsuits or royalties.
60. What do you mean by “low DPI”?
Quote:
Originally Posted by wareya View Post

Smoothing, angle snapping, and path correction are bait for casual PC users, nothing more.
61. What is this statement based on?
62. Do you know any ICS without such algorithms?
63. Is there any proof, that there are ICS without such algorithms?
Quote:
Originally Posted by wareya View Post

The real world disagrees with you, basically.
64. How can real world disagree with its working principle?
Quote:
Originally Posted by wareya View Post

There's no fundamental limitation here;
65. Why do you deny existence of fundamental limitations?
66. What is this statement based on?
Quote:
Originally Posted by wareya View Post

things keep getting pushed further and further with no obvious end in sight. We're way past what we need in terms of ICS quality. As for MCUs, no idea.
67. What is this statement based on?
Quote:
Originally Posted by wareya View Post

>There were no difference between “office” and “gaming” mice. There were no different sensors with completely different behavior.

Sure there were. The only reason "gaming mouse" is a thing now is because people are seriously stupid enough to make mice for office users or travellers that have tons of smoothing, angle snapping, and path correction. I learned it the hard way when I bought my Pyra.
68. You want to say, that ICS engineers are stupid enough to make mice for “office users”, but at the same time same engineers make sensors for “gamers”?
69. Why do you except, that more expensive mice don’t have tons of smoothing, angle snapping and path correction?
70. Do you have any proof?
Above8 is offline  
post #125 of 130 (permalink) Old 07-23-2015, 10:48 AM
New to Overclock.net
 
wareya's Avatar
 
Join Date: Jun 2015
Posts: 1,699
Rep: 74 (Unique: 43)
Unfortunately for you, your argument via verbosity doesn't scare me.

>1. What do you mean by “any analog”?
>2. You want to say, that ball mice have same limitations as ICS?
Any "analog -> digital conversion". Basic english.
>3. By quantization you mean CPI?
https://en.wikipedia.org/wiki/Quantization_%28signal_processing%29
You *CANNOT* convert analog data (motion) into digital format without introducting quantization, and quantization is *THE* limitation of *ALL* movement response systems. If you don't understand basic signal processing, you have no right to teach *ANYBODY* about it.

>4. What noise are you talking about?
Photographic grain.

>11. Which flaws exactly do you mean?
I did say all of them.
>12. How can you compensate them?
I already explained this. Your argument is "How?" after I've alread explained myself several times.
>13. What do you mean by “easy”?
Even an intern could do it.
>14. With which errors it will be compensated?
This is ungrammatical.
>15. Will it lead to different flaws?
No.
>16. Why do you think it’s possible?
Because it's actively done in the real world.

>17. What things exactly?
The things you claim keep it from working optimally.
>18. What do you mean by “perfectly fine”?
Are you seriously asking this question?
>19. Why do you think that these things don’t massacre it?
Because we know that they don't from a combination of "how DIC mathematically works" and "real world mice work".

>20. Any proof on 7 bpp?
>21. What is this based on?
qsxcv's debug images from the sensor are 36x36 with 7bpp. That is the amount of information that they contain. You can easily verify this yourself with any image editor capable of displaying color hex codes.
>22. What do you want to say by this?
This is not a question.

>23. How did you get this number?
>24. What do you want to say by this?
I ran an unsigned difference filter on shifted 32x32 matrixes from qsxcv's debug images and accumulated the result of the difference filter on each pixel. This resulted in a difference of 16384 raster levels. This is exactly how DIC operates and how it detects movement. A very small movement results in this large of a number. This is the level of precision that optical mice work under.

>25. More than enough for what?
100k DPI, for one thing.
>26. What is enough?
I think that 64 levels per sensor count would be enough in an ideal world, but multiplying that number by a couple hundred seems like a good idea Just In Case (tm).
>27. Is it grounded on anything?
The competent experience of literally every single person in the world that's using a good optical mouse.

>28. If accuracy can’t be more than 100%, how can it be too much?
Accuracy cannot ever be 100%, and we are far, far past the point of diminishing returns. If you want an LED that's so bright that it burns your eye, I'm sure that we could get 1% better than existing DIC implementations, but it's not worth it.

>29. Do you mean hardware resolution or interpolation?
The hardware resolution is 36x36, but this has no relation to the
>30. What do you mean by high?
More than 800
>31. How DPI is related with tracking accuracy, which we are talking about?
High DPI is more accurate, or do you want to claim that we shouldn't push the limits like you just said we should?

>32. On what CPI?
Any CPI.
>33. What do you mean by negligible?
https://en.wiktionary.org/wiki/negligible
>34. Do you know it’s exact influence?
It's low enough that we can't detect it, ever.
>35. How do you know, that sensor doesn’t hide jitter by smoothing and/or any other kind of (post)processing/path correction?
Because it doesn't need to. Actually, the FinalMouse does, but it's only necessary at 5000 DPI, and it doesn't reduce drift, because the smoothing implementation that the FinalMouse has is a simple infinite response filter that doesn't have any properties of reducing random walk.

>36. How can technology have no flaws and limitations?
You said "a lot", not "any".
>37. Do you want to say, that ICS has no limitations?
No; it has the limitations of quantization and sensor noise, just like literally every single other possible motion sensing method.
>38. What is this statement grounded on?
Math.
>39. Why did you except Niquist frequency from ICS limitations?
ICS operate on raster images. Raster images are mathematically equivalent to PCM waveforms. PCM waveforms have nyquist frequency.
>40. Why did you except pixel array size and framerate from ICS limitations?
Pixel array size and framerate are obvious forms of quantization. Having two doesn't suddenly make it worse.

>41. How do you know, that it’s possible?
Every new "next gen" sensor increases these things over the previous new sensor.
>42. How could it lower influence of limitations?
Increasing these reduce quantization.

>43. Why don’t you think it’s possible?
>44. How exactly will you do it?
All optical mice do this already. It's a basic element of DIC. It's mentioned in that interpolation patent that qsxcv posted in the other thread, if you want to verify it for yourself.
>45. Which measurement errors will have this?
None. It mathematically has zero drawbacks. You only need to increase the margins of the sensor picture array without increasing the size of the correlation array, then you get it for free.

>46. What is this statement based on?
Math.
>47. How can prediction add no artifacts?
It's not "prediction" in the sense you use it.
>48. Any proof?
1+1=2, pigs don't have wings, and moving over the frame of reference in a picture doesn't make it look different. It's essentially an optimization so that they can perform fewer computations and get the same result.

>49. What do you mean by theoretical problems?
Problems that don't exist on the real world.
>50. How can theoretical problems disappear?
>51. Which theoretical problems exactly do you mean?
You just asked a question using a term that you asked for a definition for and before receiving the definition. Invalid.

>52. How is FPS related with sensor quality?
Higher framerates reduce quantization error.

>53. How you think brightness of LED can eliminate fundamental flaws?
>54. Which limitations can it possibly eliminate?
Brighter LEDs allow for higher framerates and lower exposures without increasing sensor optical noise.

>55. Aren’t they popular because of invisibility?
My DA's IR LED emitted visible light in addition to its IR light.

>56. What is practical limit of DPI?
Right now? Looks like 10000.
>57. Why do you think, that jitter happens only on some CPI settings?
Jitter has many causes, but the cause of jitter that I'm referring to here is cheap interpolation of the correlation accumulation array causing jitter.
>58. Doesn’t jitter happen on level of sensor measurement?
No. Quantization happens on the level of sensor measurement. Quantization actually prevents jitter.

>59. What do you want to say by this?
It means what it says. If your grasp of English is too poor to understand it, you shouldn't be arguing with me.

>60. What do you mean by “low DPI”?
400

>61. What is this statement based on?
History of mouse marketing in America between 2005 and 2010.
>62. Do you know any ICS without such algorithms?
MLT04
>63. Is there any proof, that there are ICS without such algorithms?
Feel free to dump and reverse engineer the MLT04's firmware to prove the assertion that the MLT04 implements smoothing, angle sanpping, or path correction.

>64. How can real world disagree with its working principle?
It's not a fundamental principle, that should be clear given that the real world disagrees with it.

>65. Why do you deny existence of fundamental limitations?
>66. What is this statement based on?
You are chopping off a sentence.

>67. What is this statement based on?
3366 has higher stats than the 3310, which has higher stats than the MLT04. These stats are being pushed forward. MLT04 was good enough 15 years ago, and yet we're still improving these statistics.

>68. You want to say, that ICS engineers are stupid enough to make mice for “office users”, but at the same time same engineers make sensors for “gamers”?
Engineers don't make management decisions.
>69. Why do you except, that more expensive mice don’t have tons of smoothing, angle snapping and path correction?
I never said that there aren't mice with these things. I said that there are mice without these things. Your understanding of english negatives needs work.
>70. Do you have any proof?
Do you? No. You don't. You're the one enforcing spectacular claims here.
wareya is offline  
post #126 of 130 (permalink) Old 07-23-2015, 12:24 PM
lololol
 
qsxcv's Avatar
 
Join Date: Feb 2014
Posts: 4,243
Rep: 365 (Unique: 152)
this has got to be the most ridiculous thread ever on this forum

http://linustechtips.com/main/topic/272889-a-guide-to-buying-the-right-mouse/?p=3716426
Quote:
When it comes to Variance Twin Eye is the undisputed King with a Variance of under 0.1%
~0.1% variance is a very reasonable value given the speed resolution i estimated earlier. (speed resolution and minimum track speed go together for te sensors)

francois morier's comments to ino:
Quote:
Regarding PLN sensors, they have serious theoretical advantages:
- high speed accuracy, since their principle of work is based on delta speed to detect motion
- high DPI, it's purely arithmetic and they can split down to 1 DPI - less noise sensitive
- smalls system size and full integrated opto-ME - easy to integrate in product - cool ID

On the other end, they suffer from strong drawbacks:
- low speed accuracy is poor - sometimes miss departure if slow start !
- lift is unpredictable on soft surface - also mouse case deformation is reported...
- angle error is disappointing - especially 3 sigma deviation makes it hard to predict
- very surface dependent - hard, smooth to glossy, is prerequisite to grant PLN performance

So saying that PLN is better than image correlation sensor is a bit of a shortcut - especially accounting user and usage facts.
Recent image correlation sensor are factor better in resolution vs speed - zero acceleration - this makes them highly predictable.
i'm fairly sure these are only the case when considering low speed performance and/or pte sensors on cloth pads.

to me, it's still unclear how well te can perform on a cloth pad, and how noticeable it's low-speed issues are compared to those of ics sensors.

meanwhile, you guys can continue arguing over nothing rolleyes.gif

too busy to check forums as regularly
pm me if i forget to respond
qsxcv is offline  
post #127 of 130 (permalink) Old 08-22-2015, 06:40 AM - Thread Starter
New to Overclock.net
 
Above8's Avatar
 
Join Date: Dec 2013
Location: Russia, Moscow
Posts: 424
Rep: 40 (Unique: 34)
Quote:
Originally Posted by wareya View Post

Unfortunately for you, your argument via verbosity doesn't scare me.
You either don’t have sources of information/logics or/and hide it for some reason.

Parts of your words are milk and water:
17 > working optimally
19 > we know that they don't
27 > The competent experience of literally every single person in the world that's using a good optical mouse.
33 > negligible
34 > It's low enough that we can't detect it, ever.
35 > Because it doesn't need to.
59 > It means what it says.
67 > MLT04 was good enough
69 > there are mice without these things

Parts of your words are empty groundless statements:
15 > No.
(In the same way I could reply all your answers with one word. It makes no sense.)
16 > Because it's actively done in the real world.
32 > Any CPI.
38 > Math.
40 > Pixel array size and framerate are obvious forms of quantization. Having two doesn't suddenly make it worse.
45 > None. It mathematically has zero drawbacks.
46 > Math.
47 > It's not "prediction" in the sense you use it.
48 > It's essentially an optimization so that they can perform fewer computations and get the same result.
49 > Problems that don't exist on the real world.
52 > Higher framerates reduce quantization error.
54 > allow for higher framerates
56 > Right now? Looks like 10000.
61 > History of mouse marketing in America between 2005 and 2010.
68 > Engineers don't make management decisions.

Parts of your words are lies, errors, misinformation:
11 > I did say all of them.
12 > I already explained this. Your argument is "How?" after I've alread explained myself several times.
13 > Accuracy cannot ever be 100%,
14 > High DPI is more accurate
41 > Every new "next gen" sensor increases these things over the previous new sensor.
58 > Quantization actually prevents jitter.

Parts of your words show, that your understanding of English negatives needs work.
> >39. Why did you except Niquist frequency from ICS limitations?
> ICS operate on raster images. Raster images are mathematically equivalent to PCM waveforms. PCM waveforms have nyquist frequency.

The main idea behind all of your speeches is that your DA is enough for you. That modern IC sensors are good enough, errors are negligible, that people don’t need anything better. That search for better is a waste of time, because no one will ever notice difference.

But it’s the same thing as claiming than human eye can’t see more than 720p at 30 FPS, and we don’t need 4k at 144 FPS. It’s same as claiming that 1 000 Hz is not better than 500 Hz and no one will ever notice difference. But people notice difference in comparison, and this difference may be a selling point for some people.

Statements like “720p is good enough” and “difference between 30 and 60 FPS is low enough that we can’t detect it, ever” don’t make sense to me.

I’m trying to say, that there is difference between ICS and PTE, and that this difference is significant. And even it if it’s not that obvious at first sight, it doesn’t mean it doesn’t exist at all.

Quote:
Originally Posted by qsxcv View Post

i'm fairly sure these are only the case when considering low speed performance and/or pte sensors on cloth pads.

to me, it's still unclear how well te can perform on a cloth pad, and how noticeable it's low-speed issues are compared to those of ics sensors.
I have no idea what they tested and how (if they tested at all). All these statements sound very strange from him. I don’t get what they call “low speed accuracy” if their own sensors are much harder to control low speed movements, because of insane speed vs time errors and other errors.

From my experience, difference between sensor models and implementations are bigger, than difference in tracking between hard and cloth surfaces. Main problem with PTE on cloth is cursor jumps caused by clicks. This problem may be significantly reduced by increasing muse feet surface and reducing click resistance.

Recently I’ve found out that raw wood cutting board may be worthy alternative to cloth mouse pads for PTE. It’s hand friendly and has similar kind of friction. At the same time it has good characteristics for PTE tracking.



ICS has huge amount of their own problems related with small movements. First of all, it’s significant amount of noise, especially with small and/or fast movements.

In spite of the fact that slow speed movements are PTE’s weakness, from my experience it’s still more predictable than any ICS I’ve used. In practice, PTE’s slow movement flaw is predictable “negative acceleration” on extremely slow movements, which may be compensated with habit. In fact, like any kind of predictable acceleration, it’s not a direct flaw, it’s more like a tradeoff.

High quality ball mice leave far behind both ICS and PTE in terms of small slow movements. I’m only talking about high quality ball mice because recently I’ve used a couple of different ball mice. After this I have an idea about how quality of ball mouse influences performance. smile.gif


Quote:
Originally Posted by qsxcv View Post

meanwhile, you guys can continue arguing over nothing
It’s disappointing me as well.
Above8 is offline  
post #128 of 130 (permalink) Old 08-22-2015, 06:44 PM
New to Overclock.net
 
wareya's Avatar
 
Join Date: Jun 2015
Posts: 1,699
Rep: 74 (Unique: 43)
17 > working optimally

Your smoke and mirrors, not mine.

19 > we know that they don't

Yes, a fact that disproves you is smoke and mirrors just because you don't like it, definitely.

27 > The competent experience of literally every single person in the world that's using a good optical mouse.

Not smoke and mirrors. Empirical.

33 > negligible

See: statistics 101

34 > It's low enough that we can't detect it, ever.

http://sousuch.web.fc2.com/DIY/mouse_senor/index.html
Guess what has zero error? An optical mouse (DA).

35 > Because it doesn't need to.

Your smoke and mirrors, not mine.

59 > It means what it says.

Learn English.

67 > MLT04 was good enough

It was, though. It has perfect response below a broundary.

69 > there are mice without these things

Are you seriously trying to deny this?

----

15 > No.
(In the same way I could reply all your answers with one word. It makes no sense.)

Good for you! How about you actually break down why the things I'm saying are wrong, instead of saying I'm just wrong?

16 > Because it's actively done in the real world.

See: WMO

32 > Any CPI.

Real jitter from DIC scales linearly with CPI. This means that jitter is only removed at low CPIs because quantization error is stronger than it. See below. This is basic discrete mathematics. If you don't understand this, you need to go back to school before you continue trying to argue with me.

38 > Math.

It's not my fault if you literally don't understand the math behind what you're talking about.

40 > Pixel array size and framerate are obvious forms of quantization. Having two doesn't suddenly make it worse.

How is something like this possibly an "empty groundless statement"? Are you just picking things you don't like?

45 > None. It mathematically has zero drawbacks.

See above.

46 > Math.

See above.

47 > It's not "prediction" in the sense you use it.

See above.

48 > It's essentially an optimization so that they can perform fewer computations and get the same result.

See above.

49 > Problems that don't exist on the real world.

See 16

52 > Higher framerates reduce quantization error.

See 40

54 > allow for higher framerates

See 40

56 > Right now? Looks like 10000.

Find a mouse with over 10000 DPI that doesn't jitter or accelerate like ****.

61 > History of mouse marketing in America between 2005 and 2010.

See 40

68 > Engineers don't make management decisions.

See 40

----

11 > I did say all of them.

Learn English.

12 > I already explained this. Your argument is "How?" after I've alread explained myself several times.

My explanations are right there. Go ahead and read them again if you're actually openminded.

13 > Accuracy cannot ever be 100%,

Accuracy CANNOT ever be 100%. If you seriously think it can, you need to stop posting.

14 > High DPI is more accurate

It is, though, as long as the MCU isn't getting overloaded or doing something ******** because of it.

41 > Every new "next gen" sensor increases these things over the previous new sensor.

That's not misinformation. 3366 is better than 3310 is better than etc etc etc.

58 > Quantization actually prevents jitter.

Not misinformation. Quantization puts responses only at boundaries. Jitter won't occur away from these boundaries. Jitter slowly replaced by quantization as quantization increases.

---

>First of all, it’s significant amount of noise, especially with small and/or fast movements.

Brighten the LED, reduce the exposure. Voila! Less noise!
wareya is offline  
post #129 of 130 (permalink) Old 08-23-2015, 12:51 AM - Thread Starter
New to Overclock.net
 
Above8's Avatar
 
Join Date: Dec 2013
Location: Russia, Moscow
Posts: 424
Rep: 40 (Unique: 34)
Again, empty words and insults, that prove nothing. You either don’t have sources of information/logics or/and hide it for some reason.
Above8 is offline  
post #130 of 130 (permalink) Old 08-23-2015, 01:30 AM
New to Overclock.net
 
wareya's Avatar
 
Join Date: Jun 2015
Posts: 1,699
Rep: 74 (Unique: 43)
so much irony lol
wareya is offline  
Reply

Quick Reply
Message:
Options

Register Now

In order to be able to post messages on the Overclock.net - An Overclocking Community forums, you must first register.
Please enter your desired user name, your email address and other required details in the form below.
User Name:
If you do not want to register, fill this field only and the name will be used as user name for your post.
Password
Please enter a password for your user account. Note that passwords are case-sensitive.
Password:
Confirm Password:
Email Address
Please enter a valid email address for yourself.
Email Address:

Log-in



Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)
 
Thread Tools
Show Printable Version Show Printable Version
Email this Page Email this Page


Forum Jump: 

Posting Rules  
You may post new threads
You may post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off