| |
| |
| |
|
Page: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
Comments:
<0> I didn't learn much of anything
<1> the scope will cause the smart pointers to destruct before CoUninitialize is called
<2> its just confusing me how they are both called the same, and seem to accomplish the same
<0> I don't remember if there was a book for it or not, I remember the teaching being mostly from the notes.
<2> yet one is pure stats the other is fourier analysis
<0> Fourier and Laplace transform did the same to me.
<3> what does the {} fix?
<1> your problem
<0> They're the same integral damn it!
<0> They solve the same problem
<0> I get the same transforms
<0> only the limits are different
<0> which never seemed to affect any problem we ever did.
<1> CoInitialize(...); { CComPtr<IFoo> spFoo; ... } CoUninitialize(...);
<3> hmm... true.. it works... but why?
<1> spFoo will destruct before CoUninitialize is called, which is required, as it uses COM
<3> aha
<3> peter, you from HU
<3> ?
<1> no, i estabolished that already, i'm american
<1> my last name starts with hu
<2> Noidea, you know I had a job offer at a bank for an IT job and I canceled it
<2> to do the phd
<2> just to get into those quant sections
<1> IT job
<1> pfft
<3> ah sorry
<2> peterhu :)
<0> yeah, it sounds really interesting.
<2> IT as in programming, not Win XP support
<2> ahem
<1> you know, if you get your doctorate, we're NOT calling you Dr.
<2> by then most people will be dead in here anyway
<1> you'll still be bitch to us
<2> uh huh
<4> is anyone familiar with posix threads
<1> partially
<1> haven't used them in some time
<1> and i'd probably favor boost::thread now
<4> is there anyway to attach a struct data type to them?
<1> thread local storage? or to p*** a structure to the thread's entry point?
<4> well, this is my first time working with threads, what needs to be done is that a subset of integers needs to be p***ed to x threads to determing if they are prime or not
<4> i have no idea on doing so
<4> any suggestions on how to '***ign' the subset to the threads ?
<1> the thread_function takes a void*, afair
<2> http://www.llnl.gov/computing/tutorials/pthreads/#P***ingArguments
<4> what exactly is a return type of void* ?
<2> what exactly is a C book?
<4> i mean a dereferenced no return ?
<4> a ptr with no return?
<2> void* was the way to describe "any type" - its not a function its a pointer to a block of memory
<1> a void* is a pointer to an unknown type
<2> its also archaic
<1> but required in API's like these
<1> i wonder how many schools teach boosdt
<1> -d
<1> boost
<4> ok, thx for the descrip. makes more sense now.
<2> most schools don't teach C/C++
<2> maybe C
<4> this is supposedly a cl*** in O.S.
<2> ah yes
<1> some teach C++--98
<2> mine started to teach generics in Java now
<2> leap of faith
<1> heh
<2> user, the idea is you have some memory structure, you cast it to void* to p*** it to a thread
<2> once inside the thread, you cast it back to the type you intended it to be
<4> is the casting done explicitly?
<2> yes
<2> else its a tad useless
<5> stiu ce este buna _Nes|mtiTA_
<5> moi tlt :P
<5> so prai
<2> continue running south
<6> can anyone help me with vocal removal from music? i got it as a high school project
<5> dosaga
<4> ok so it'd be something like, create the thread, p*** it to the function and the ptr to the subet ?
<2> go to the link I gave you
<4> *subset
<2> read it
<4> k
<2> then ask questions, it will help you
<2> im going back to my neurons
<7> hm another one..
<7> time_t is not a typedef but int -.-
<7> did i mention.. gnar
<7> :)
<4> oh nice....example 2 is all me...thx for the link
<2> no problems
<2> remember google is the entry to your work thread...
<8> neosisani you you care to elaborate?
<6> i have to make program which will remove voice from a song sent to it
<6> it has to leave all musical instruments
<8> and this is for what cl***?
<6> programming
<8> that's the name of the cl***? "programming" ?
<6> just give me a sec to translate it
<6> well it informatics
<4> is the purpose of a lock to lock global variables shared by all threads from being accessed simultaneously?
<8> user_xyz to "lock" whatever you want
<8> neosisani that program seems definitely non-trivial
<6> vawjrwrk, i was at researh station Petnica (for high school students) for one week and they gave me one month to make idea for program that will remove voice. If idea is good they will invite me there again to code the program.
<6> and i know it is not trivial, i hardly found anything about it
<8> well, have you any thoughts about how to do it?
<4> are vectors supported in C ?
<8> user_xyz what's a vector?
<4> array
<8> and why are you asking C questions in a C++ channel?
<4> simple question i thought u gurus might know
<6> the most common thing used is that it is ***umed that singer is in the middle. Then you change phase of left channel for pi/2 and add it to right channel. This way you get mono signal without voice.
<8> why pi/2 ? pi is "out of phase"
<6> Anyway i don't like that idea, so i thought to make analysis of music, and try to locate voice in it. Probably to find some common harmonics
<6> vawjrwrk, error in typing, sorry
<8> ok, thought maybe my memory was failing, lol
<6> :>
<8> I thnk that's a bad ***umption, btw (in the middle)
<8> I don't recall ever seeing any literature on the "harmonics" of the human voice
<6> i know, that is why balance is used to make sure singer is in the middle
<6> vawjrwrk, i haven't found it either
<8> will someone guarantee you of that
<6> no, that is why i will try to avoid that idea (singer in middle)
<8> and I suppose you don't have a sample of the singer alone
<6> i don't
<8> well, you do know the frequency range of the voice is rather narrow
<6> but, what about harmonics
<6> i know that most power iz below 1000 hz
<8> well, w/ a LOT of compute power and human intervention, you might be able to find the voice in the music
<8> but I suspect the harmonics change as the note changes also
<6> well my program doesn't need to run in real time, so computing power shouldn't be that big problem
<8> but I suspect your program should run unattended
<6> it should
<8> you're gonna have to do some research to see what "voices" look like, I suspect
<6> i know, do you have some book or link?
<2> this is a very complex topic btw
<2> its not some sort algo
<9> this is a rather new and expensive technology
<6> expensive?
<8> for what? doing fourier transforms to try to identify the "notes" being sung ?
<2> look at Neuron Networks, Hidden Markov Chains, Speech Analysis
<9> its really not something that can be found on line
<2> oh is that the research?
<6> yes
<8> no, remove the voice from a song
<9> and its too young and complex to write in a book
<6> it is supposed to be research project or sth like that
<6> this is like first phase
<2> its phd material in my opinion
<8> maybe past that
<8> on the order of proving Fermat's Last Theorem (which has apparently been done)
<2> having said that
Return to
#c++
or
Go to some related
logs:
#AllNiteCafe
#linux
ipod plaisio greece
sexmasculin
aqalla
voice reconition for sql coding
#AllNiteCafe
#AllNiteCafe
mysql semi-column sql
#AllNiteCafe
|
|