i'm finally convinced that i have this fixed, but it's bed time, now. i'm going to want to re-render pretty much everything for this project, but i actually think the mixing details will be pretty slight. even the stuff i was iffy on actually seems to sound pretty good.
my (hopefully) final analysis on this is as follows:
1) the initial problem was probably related to the codec install, although i can't state exactly what it did and will likely never know. it was a slightly different problem than showed up later on in the sense that it was at least consistent.
2) on the 21st, i ran the registry wipe. that would have broken the clock and put everything out of sync. i should have actually realized that, because i should have known to manually update the system drivers. i may have fixed the initial problem, but i wouldn't have been able to tell if i did because i broke the clock. since the 21st, this has been the primary problem, but i didn't realize it.
3) the reinstall ran the same registry wipe, which broke the clock again.
4) fixing the clock seems to have resolved the issue altogether.
i fully intend to get some serious work on this done tomorrow.
Monday, July 27, 2015
as i went into device manager to disable one of the cards, i noticed that the microsoft streaming proxy device had an exclamation mark and a code 19, telling me that the registry was wiped. and, that jogged my memory.
one of the reasons that i'm sticking with xp is that it's winlited. rather dramatically. the iso is...well, it's pretty big because it has install scripts and added programs. but, if it was just the winlited iso it would be around 50 mb. it's the size of one of those mini-linux distros. and it's cut down just as much. on top of that dramatic haircut, i run an install script that deletes hundreds of windows files and wipes out half the registry. it's pretty comprehensive. mine is the only non-network computer i've ever seen where trying to hit the internet from windows explorer actually seriously fails.
i can't know what caused the problem initially because i've wiped it down. it may have very well been the codec install - or perhaps the uninstall. but, i do know that i've seen this problem with the streaming device before, and i know that the reason it's telling me it's not in the registry is because it's not in the registry - the install script wiped it. the script is almost ten years old. i was mixing in cool edit via cut and paste at the time. my sleuthing told me it was a windows media service. i've just never updated the script..
...because it's an easy fix - i take it out of my vanilla sp3 virtual machine that i keep around for these reasons and install via device manager. i've actually done this before, albeit not for this reason. i forgot to do it on the recent reinstall.
time will tell if that's actually what the problem was. but, i think it makes sense.
thanks.
one of the reasons that i'm sticking with xp is that it's winlited. rather dramatically. the iso is...well, it's pretty big because it has install scripts and added programs. but, if it was just the winlited iso it would be around 50 mb. it's the size of one of those mini-linux distros. and it's cut down just as much. on top of that dramatic haircut, i run an install script that deletes hundreds of windows files and wipes out half the registry. it's pretty comprehensive. mine is the only non-network computer i've ever seen where trying to hit the internet from windows explorer actually seriously fails.
i can't know what caused the problem initially because i've wiped it down. it may have very well been the codec install - or perhaps the uninstall. but, i do know that i've seen this problem with the streaming device before, and i know that the reason it's telling me it's not in the registry is because it's not in the registry - the install script wiped it. the script is almost ten years old. i was mixing in cool edit via cut and paste at the time. my sleuthing told me it was a windows media service. i've just never updated the script..
...because it's an easy fix - i take it out of my vanilla sp3 virtual machine that i keep around for these reasons and install via device manager. i've actually done this before, albeit not for this reason. i forgot to do it on the recent reinstall.
time will tell if that's actually what the problem was. but, i think it makes sense.
thanks.
This sounds like a phasing problem where the audio is following two paths through the system and the clock on one path is running slightly by about 1/120Hz faster than the other.
well....i'm at no point sending it out through two directions. older versions of cubase like that require you use one card at a time. but, a conflict is something i can test for by disabling the cards. it's worth a shot. i think anything is at this point. but, i'm not putting too much faith in it.
i think the idea of the clock being off some other way is maybe a possible explanation, though. thanks for that as a path to explore.
well....i'm at no point sending it out through two directions. older versions of cubase like that require you use one card at a time. but, a conflict is something i can test for by disabling the cards. it's worth a shot. i think anything is at this point. but, i'm not putting too much faith in it.
i think the idea of the clock being off some other way is maybe a possible explanation, though. thanks for that as a path to explore.
completely stumped, resorting to "magnetic interference" as a possible explanation
hi.
i've got a very strange issue, just wanting to run it through some people to see if anybody's seen this before, before i start lining my walls with tinfoil. i'm not joking.
i'm running xp. of the soundcards in the machine, two are of importance: an alesis 16 channel firewire mixer and an m-audio delta 3296 that connects over good old pci. my main daw is cubase sx 3. yes, it's all very old. it works just fine. i'm going to need to move up to 64 bit one day to access more ram, but that day does not appear to be on the horizon.
a few weeks ago, i installed a windows codec pack in an attempt to get firefox to natively read flac over ogg. i misread the documentation; that's impossible. it shouldn't be impossible, but so be it. i had previously identified this as the root of the problem, but i've recently been backing away from that.
basically, i'm getting random tone issues out of both of these cards over all interfaces: asio, wdm and kmixer. from time to time, it actually sounds "correct". but, then it randomly modulates between being floored by the subbass and being cut around 2000 Hz. it happens in cubase over asio. it happens in foobar over wdm and through the kmixer.
this afternoon, i was finally able to isolate something specific, but this is such a fleeting problem that it's possible that it could change in a few hours, or have already changed.
imagine this..
you have a snare. it's a closed snare. it's just one hit. you put that snare hit on repeat in cubase. after about two minutes, that closed snare sounds like an open snare. so, you reset your drivers - and you get the closed snare again. two minutes later, it's an open snare again. and, this is repeatable through all of the following signal paths:
1) cubase-->asio-->alesis
2) cubase--->asio--->m-audio
3) foobar--->wdm--->alesis
4) foobar--->kmixer--->m-audio
i've tried driver reinstalls. i've tried an os reinstall. i've swapped every cord you can imagine. i've swapped out amps. consistency seems unobtainable.
is it time to line the walls with foil? does anybody have a better idea?
i've got a very strange issue, just wanting to run it through some people to see if anybody's seen this before, before i start lining my walls with tinfoil. i'm not joking.
i'm running xp. of the soundcards in the machine, two are of importance: an alesis 16 channel firewire mixer and an m-audio delta 3296 that connects over good old pci. my main daw is cubase sx 3. yes, it's all very old. it works just fine. i'm going to need to move up to 64 bit one day to access more ram, but that day does not appear to be on the horizon.
a few weeks ago, i installed a windows codec pack in an attempt to get firefox to natively read flac over ogg. i misread the documentation; that's impossible. it shouldn't be impossible, but so be it. i had previously identified this as the root of the problem, but i've recently been backing away from that.
basically, i'm getting random tone issues out of both of these cards over all interfaces: asio, wdm and kmixer. from time to time, it actually sounds "correct". but, then it randomly modulates between being floored by the subbass and being cut around 2000 Hz. it happens in cubase over asio. it happens in foobar over wdm and through the kmixer.
this afternoon, i was finally able to isolate something specific, but this is such a fleeting problem that it's possible that it could change in a few hours, or have already changed.
imagine this..
you have a snare. it's a closed snare. it's just one hit. you put that snare hit on repeat in cubase. after about two minutes, that closed snare sounds like an open snare. so, you reset your drivers - and you get the closed snare again. two minutes later, it's an open snare again. and, this is repeatable through all of the following signal paths:
1) cubase-->asio-->alesis
2) cubase--->asio--->m-audio
3) foobar--->wdm--->alesis
4) foobar--->kmixer--->m-audio
i've tried driver reinstalls. i've tried an os reinstall. i've swapped every cord you can imagine. i've swapped out amps. consistency seems unobtainable.
is it time to line the walls with foil? does anybody have a better idea?
i'm actually finally making some progress in figuring this out.
it was perfect. for a few minutes. the subbass cleared. the highs were crystalline. just as i remembered it. exciting.
however, by the time i got to the end of the sequence, the highs were starting to fade. the subbass stayed gone, and hasn't come back, but because the music is guitar oriented it can't cut out at 2000 hz. it's where all the tone is. it's keenly important...
i started thinking i mixed it a little dull. that's fine. i just have to establish that consistency, then i can adjust. but, when i restarted the sequence, it was dull again.
so, it seemed as though i was losing the high end over time. could i test this?
i switched cards for a minute to get a base signal, and then i put a single section on repeat.
as the section repeated, it faded from the perfect, crystalline sound i want to your typical am radio hip-hop garbage. but, it was a long section and hard to really hear directly.
i zoomed in to just a snare drum and repeated.
with the snare drum on repeat, it became audibly obvious. it began to fade from a nice tight snare to a loose, open mess. and, i was able to repeat this. directly through the mixer, with nothing in between.
this goes back to a driver problem, i believe. i suppose it could be a resistance issue, in theory - it could be overloading and turning down. but the fact that i'm able to get around it by eqing the mix makes that seem unlikely, as i'd still be overloading it and it would still turn down. i can test this by sending something directly through the device, bypassing the drivers.
the fact that it's predictable seems to rule out magnetic interference - for now. i would expect something more random. if i get more random behaviour, it will reassert interference.
but, finding a way to create the problem and test it and have repeatable results is a big breakthrough.
if it's drivers...it doesn't make sense, really. it's never done this before. i just reinstalled. i suppose i can try an updated driver set, but i picked this driver set for a reason. ultimately, i can't really make sense of it. i don't really care right now. i just want to fix it.
the thing is, though, that it often doesn't turn on to the clear sound i'm getting right now. i'm isolating what i'm hearing as a decay around 10,000 Hz. but it's only an aspect of the problem.
it was perfect. for a few minutes. the subbass cleared. the highs were crystalline. just as i remembered it. exciting.
however, by the time i got to the end of the sequence, the highs were starting to fade. the subbass stayed gone, and hasn't come back, but because the music is guitar oriented it can't cut out at 2000 hz. it's where all the tone is. it's keenly important...
i started thinking i mixed it a little dull. that's fine. i just have to establish that consistency, then i can adjust. but, when i restarted the sequence, it was dull again.
so, it seemed as though i was losing the high end over time. could i test this?
i switched cards for a minute to get a base signal, and then i put a single section on repeat.
as the section repeated, it faded from the perfect, crystalline sound i want to your typical am radio hip-hop garbage. but, it was a long section and hard to really hear directly.
i zoomed in to just a snare drum and repeated.
with the snare drum on repeat, it became audibly obvious. it began to fade from a nice tight snare to a loose, open mess. and, i was able to repeat this. directly through the mixer, with nothing in between.
this goes back to a driver problem, i believe. i suppose it could be a resistance issue, in theory - it could be overloading and turning down. but the fact that i'm able to get around it by eqing the mix makes that seem unlikely, as i'd still be overloading it and it would still turn down. i can test this by sending something directly through the device, bypassing the drivers.
the fact that it's predictable seems to rule out magnetic interference - for now. i would expect something more random. if i get more random behaviour, it will reassert interference.
but, finding a way to create the problem and test it and have repeatable results is a big breakthrough.
if it's drivers...it doesn't make sense, really. it's never done this before. i just reinstalled. i suppose i can try an updated driver set, but i picked this driver set for a reason. ultimately, i can't really make sense of it. i don't really care right now. i just want to fix it.
the thing is, though, that it often doesn't turn on to the clear sound i'm getting right now. i'm isolating what i'm hearing as a decay around 10,000 Hz. but it's only an aspect of the problem.
i've tried everything i can and cannot get a flat signal. i cannot mix on a device without a flat signal. i know it's not a flat signal for the following reasons:
1) the tone will not stay put.
2) i can sometimes get the same tone that i get on the neutral card in the laptop.
3) i generally do not get the same tone that i get on the neutral card in the laptop.
i cannot continue mixing until i consistently get the same tone that i get on the neutral card in the laptop. the reason is that i am unable to determine what the mix actually sounds like, and mixing is consequently entirely pointless.
so, i'm going to put this aside and jump ahead to creating the front end for aleph-1. that's a no mixing project. hopefully, by the time that's done, whatever's interfering with the signal will have stopped interfering with the signal. if not, i'll have to keep waiting.
it really seems like i'm dealing with an external mixer. i know that's insane. the only other option is that i'm dealing with some kind of electro-magnetic interference. i've ruled out a failure in the device. i've ruled out any patch cord problems. i've ruled out a driver issue, or at least escalated to the point that i don't know what else to do if it's a driver issue. the amp may be a little variant, but it's not the cause of the problem (the device has a headphone out, but i prefer not to use it because it gets very hot, and i'm not getting consistent tone out of it). i have no further leads. but i cannot do anything until the problem is fixed.
building the front-end for aleph-1 should take about a week. there are a few more front-ends to build should the problem continue beyond that point.
again: if i am dealing with an external entity, i cannot accept this. i give up, for now. but, giving up means giving up altogether rather than accepting a degradation in mixing quality.
if i was running a live sound system or something, i wouldn't care. somebody on the dance floor isn't going to care. somebody watching a rock show isn't going to care. but, a permanent document is another issue. there is only document. it is eternal. it will be as perfect as i can make it, or it will not exist at all.
even if the sound were to stabilize in this flat, gross mess that i've been stuck with for the last few hours, and even if i can mix the gross mess out of it, it's not useful - because it's not the actual sound of the actual file. if i can't get that, i'm not doing anything at all.
1) the tone will not stay put.
2) i can sometimes get the same tone that i get on the neutral card in the laptop.
3) i generally do not get the same tone that i get on the neutral card in the laptop.
i cannot continue mixing until i consistently get the same tone that i get on the neutral card in the laptop. the reason is that i am unable to determine what the mix actually sounds like, and mixing is consequently entirely pointless.
so, i'm going to put this aside and jump ahead to creating the front end for aleph-1. that's a no mixing project. hopefully, by the time that's done, whatever's interfering with the signal will have stopped interfering with the signal. if not, i'll have to keep waiting.
it really seems like i'm dealing with an external mixer. i know that's insane. the only other option is that i'm dealing with some kind of electro-magnetic interference. i've ruled out a failure in the device. i've ruled out any patch cord problems. i've ruled out a driver issue, or at least escalated to the point that i don't know what else to do if it's a driver issue. the amp may be a little variant, but it's not the cause of the problem (the device has a headphone out, but i prefer not to use it because it gets very hot, and i'm not getting consistent tone out of it). i have no further leads. but i cannot do anything until the problem is fixed.
building the front-end for aleph-1 should take about a week. there are a few more front-ends to build should the problem continue beyond that point.
again: if i am dealing with an external entity, i cannot accept this. i give up, for now. but, giving up means giving up altogether rather than accepting a degradation in mixing quality.
if i was running a live sound system or something, i wouldn't care. somebody on the dance floor isn't going to care. somebody watching a rock show isn't going to care. but, a permanent document is another issue. there is only document. it is eternal. it will be as perfect as i can make it, or it will not exist at all.
even if the sound were to stabilize in this flat, gross mess that i've been stuck with for the last few hours, and even if i can mix the gross mess out of it, it's not useful - because it's not the actual sound of the actual file. if i can't get that, i'm not doing anything at all.
ok. so, it most definitely *can't* be a blown capacitor, because i can get it pretty good by running it through an eq that is undoing whatever is fucking it up. if i blew a transistor, that wouldn't work, it would just plateau somewhere. the fact that i'm not getting any kind of push back indicates the hardware is just fine.
but that's putting me up against a wall in figuring out "da fuck".
if it stabilizes as a flat signal, i can eventually conclude that i was mixing relative to an unclean signal and remix all of it. but i should be able to at least recreate what exists through the neutral card.
i'm going to try playing with the firewire cord. that's about all i have left to grasp to. there's 8 possibilities to try.
but that's putting me up against a wall in figuring out "da fuck".
if it stabilizes as a flat signal, i can eventually conclude that i was mixing relative to an unclean signal and remix all of it. but i should be able to at least recreate what exists through the neutral card.
i'm going to try playing with the firewire cord. that's about all i have left to grasp to. there's 8 possibilities to try.
something is still wrong, but it doesn't make sense to think that there's anything blown in the device. right now, i'm getting a really dead signal, but it's not the same as before - so i'm still not getting consistency. if there was a blown capacitor or something, i'd expect it to consistently sound broken the same way. so, i can't shake this feeling that there's an external cause that is changing the output - whether it's interference or something else.
i've been playing with eqs to try and isolate it, and i've determined two things.
the first is that it seems like the sub-bass is floored, which has the effect of reducing the audible bass. if i cut the sub-bass and boost the audible bass, i get closer to the sound of a crisp, dynamic, fingered bass guitar, rather than this muddled synth bass sound that is coming out.
the second is that there seems to be a big cut over about 5000 hz. increasing the response from 5000 up to 10000 gives me back the tone on the guitar.
but, this is not the same problem i was dealing with a few days ago. a few days ago, the treble was just floored, making it sound like a tin can. i've also periodically been able to get it to sound the way i expect it to sound - and the way it sounds through the laptop, controlled for the decrease in quality through the laptop speakers.
again: if i blew a capacitor i would expect it to sound the same way every time. but, it's modulating between being too bassy and too trebly. so, it can't be a blown component.
the way it is right now - and it might be different in a few hours - is almost what i'd expect out of a system like beats, which is designed for hip-hop. it's basically a hip-hop frequency response. hip-hop doesn't tend to have much over 5000 hz, so it makes sense to cut it - you're cutting noise. and, it tends to use machines for low melodic parts that don't have a "live" frequency response, so you don't have to really worry about losing the harmonics - you can go ahead and floor the sub-bass.
i've never bought a hip-hop record in my life. this is certainly *not* hip-hop. it's white synth pop - somewhere between early nin and tears for fears, with a heavy psychedelic and grunge influence in the guitar parts. my guitar style is heavily influenced by jimi hendrix. but that's really the only thing close to black music as an influence in my writing. i'm a white person from canada that's spent my whole life listening to white music and has no desire but to create anything other than white music. sticking me with this hip-hop equalizer is just going to break the mix.
on the one hand, i know how to adjust to the effects as they currently exist. on the other hand, i know i'm adjusting to a false reality - because i know the frequency response i'm hearing is artificial. if i were to run the files through an eq that is designed to react to what i'm hearing, the files are going to lack any kind of definition on the bottom end.
so, i'm stuck. i don't think the device is broken - the testing is not consistent with this. i can't think of anything else i could do to resolve driver issues, and as i stated from the start: it doesn't seem like a driver issue. something has to be interfering with the output.
i've been playing with eqs to try and isolate it, and i've determined two things.
the first is that it seems like the sub-bass is floored, which has the effect of reducing the audible bass. if i cut the sub-bass and boost the audible bass, i get closer to the sound of a crisp, dynamic, fingered bass guitar, rather than this muddled synth bass sound that is coming out.
the second is that there seems to be a big cut over about 5000 hz. increasing the response from 5000 up to 10000 gives me back the tone on the guitar.
but, this is not the same problem i was dealing with a few days ago. a few days ago, the treble was just floored, making it sound like a tin can. i've also periodically been able to get it to sound the way i expect it to sound - and the way it sounds through the laptop, controlled for the decrease in quality through the laptop speakers.
again: if i blew a capacitor i would expect it to sound the same way every time. but, it's modulating between being too bassy and too trebly. so, it can't be a blown component.
the way it is right now - and it might be different in a few hours - is almost what i'd expect out of a system like beats, which is designed for hip-hop. it's basically a hip-hop frequency response. hip-hop doesn't tend to have much over 5000 hz, so it makes sense to cut it - you're cutting noise. and, it tends to use machines for low melodic parts that don't have a "live" frequency response, so you don't have to really worry about losing the harmonics - you can go ahead and floor the sub-bass.
i've never bought a hip-hop record in my life. this is certainly *not* hip-hop. it's white synth pop - somewhere between early nin and tears for fears, with a heavy psychedelic and grunge influence in the guitar parts. my guitar style is heavily influenced by jimi hendrix. but that's really the only thing close to black music as an influence in my writing. i'm a white person from canada that's spent my whole life listening to white music and has no desire but to create anything other than white music. sticking me with this hip-hop equalizer is just going to break the mix.
on the one hand, i know how to adjust to the effects as they currently exist. on the other hand, i know i'm adjusting to a false reality - because i know the frequency response i'm hearing is artificial. if i were to run the files through an eq that is designed to react to what i'm hearing, the files are going to lack any kind of definition on the bottom end.
so, i'm stuck. i don't think the device is broken - the testing is not consistent with this. i can't think of anything else i could do to resolve driver issues, and as i stated from the start: it doesn't seem like a driver issue. something has to be interfering with the output.
caulking the window
jessica
hi
i don't mind doing it, if you get me the caulk and let me use your gun.
it's good news and bad news. good news is that the pesticide is working. bad news is that there's at least a hundred dead ants in there. i'm going to leave them there for now, but i think it needs to be caulked relatively soon.
the landlord
I do have the caulking in my car. Sorry I have been busy and have not had the chance to come by. Pleae be patient I will come soon.
hi
i don't mind doing it, if you get me the caulk and let me use your gun.
it's good news and bad news. good news is that the pesticide is working. bad news is that there's at least a hundred dead ants in there. i'm going to leave them there for now, but i think it needs to be caulked relatively soon.
the landlord
I do have the caulking in my car. Sorry I have been busy and have not had the chance to come by. Pleae be patient I will come soon.
Subscribe to:
Posts (Atom)