Author |
Topic Search Topic Options
|
Lindsay Lohan
Forum Senior Member
Joined: May 25 2005
Location: Norway
Status: Offline
Points: 3254
|
Posted: September 13 2005 at 08:03 |
MikeEnRegalia wrote:
maidenrulez wrote:
MikeEnRegalia wrote:
maidenrulez wrote:
Wich basically means that if we record speech wich got a bandwith of about 3khz then you would need a samplingfrequency of 6khz |
Exactly. The reason why modern soundcards (and SACD) use sampling frequencies up to 192khz is that when you process digital sound (add reverb, eq ... anything) the rasterisation used during A/D conversion can cause artefacts ... nasty enharmonic overtones. So you need a large safety margin, although twice the frequency is enough if the signal is just played back.
|
Yup NOMATTER how good your sound cards there is really a huge limitation in using the PCI-bus because it has great limitiations when it comes to sound.
|
What does the bus have to do with it?
|
Infact CREATIVE wich is a huge manufacturer of sound card admitted that there where no point in making their sound card's better until they have gotten a better BUS to fit it... Actually they are working for a special bus only for sound-cards that should be optimal
|
|
|
MikeEnRegalia
Special Collaborator
Honorary Collaborator
Joined: April 22 2005
Location: Sweden
Status: Online
Points: 21149
|
Posted: September 13 2005 at 08:04 |
maidenrulez wrote:
MikeEnRegalia wrote:
maidenrulez wrote:
MikeEnRegalia wrote:
maidenrulez wrote:
Wich basically means that if we record speech wich got a bandwith of about 3khz then you would need a samplingfrequency of 6khz |
Exactly. The reason why modern soundcards (and SACD) use sampling frequencies up to 192khz is that when you process digital sound (add reverb, eq ... anything) the rasterisation used during A/D conversion can cause artefacts ... nasty enharmonic overtones. So you need a large safety margin, although twice the frequency is enough if the signal is just played back.
|
Yup NOMATTER how good your sound cards there is really a huge limitation in using the PCI-bus because it has great limitiations when it comes to sound.
|
What does the bus have to do with it?
|
Infact CREATIVE wich is a huge manufacturer of sound card admitted that there where no point in making their sound card's better until they have gotten a better BUS to fit it... Actually they are working for a special bus only for sound-cards that should be optimal
|
The bus doesn't affect digital playback in any way. Of course if they want to transmit 7.1 signals in 24bit/192khz you'd need PCIe ...
|
|
|
Lindsay Lohan
Forum Senior Member
Joined: May 25 2005
Location: Norway
Status: Offline
Points: 3254
|
Posted: September 13 2005 at 08:05 |
MikeEnRegalia wrote:
maidenrulez wrote:
MikeEnRegalia wrote:
maidenrulez wrote:
MikeEnRegalia wrote:
maidenrulez wrote:
Wich basically means that if we record speech wich got a bandwith of about 3khz then you would need a samplingfrequency of 6khz |
Exactly. The reason why modern soundcards (and SACD) use sampling frequencies up to 192khz is that when you process digital sound (add reverb, eq ... anything) the rasterisation used during A/D conversion can cause artefacts ... nasty enharmonic overtones. So you need a large safety margin, although twice the frequency is enough if the signal is just played back.
|
Yup NOMATTER how good your sound cards there is really a huge limitation in using the PCI-bus because it has great limitiations when it comes to sound.
|
What does the bus have to do with it?
|
Infact CREATIVE wich is a huge manufacturer of sound card admitted that there where no point in making their sound card's better until they have gotten a better BUS to fit it... Actually they are working for a special bus only for sound-cards that should be optimal
|
The bus doesn't affect digital playback in any way. Of course if they want to transmit 7.1 signals in 24bit/192khz you'd need PCIe ...
|
Nope actually PCI-E is even worse...i can explain later but for now i gotta go...so bye then...
|
|
|
goose
Forum Senior Member
Joined: June 20 2004
Location: United Kingdom
Status: Offline
Points: 4097
|
Posted: September 13 2005 at 08:06 |
If anyone's interested, I have a collection of sine waves with much more than one bit difference in amplitude. I can't hear any difference (I am listening through an absolutely dire system, and yes, it's a computer), but if anyone wants to see if they can tell the difference (by burning them onto CD, if needs be (they don't)) I'll upload somewhere.
|
|
goose
Forum Senior Member
Joined: June 20 2004
Location: United Kingdom
Status: Offline
Points: 4097
|
Posted: September 13 2005 at 08:09 |
maidenrulez wrote:
MikeEnRegalia wrote:
maidenrulez wrote:
MikeEnRegalia wrote:
maidenrulez wrote:
Wich basically means that if we record speech wich got a bandwith of about 3khz then you would need a samplingfrequency of 6khz |
Exactly. The reason why modern soundcards (and SACD) use sampling frequencies up to 192khz is that when you process digital sound (add reverb, eq ... anything) the rasterisation used during A/D conversion can cause artefacts ... nasty enharmonic overtones. So you need a large safety margin, although twice the frequency is enough if the signal is just played back.
|
Yup NOMATTER how good your sound cards there is really a huge limitation in using the PCI-bus because it has great limitiations when it comes to sound.
|
What does the bus have to do with it?
|
Infact CREATIVE wich is a huge manufacturer of sound card admitted that there where no point in making their sound card's better until they have gotten a better BUS to fit it... Actually they are working for a special bus only for sound-cards that should be optimal
|
Creative is a huge manufacturer of sound cards intended mainly for playing games with and usually not sounding very good for music (I'm going to reserve judgement on the new one which I forget the name of, since I haven't heard it).
|
|
MikeEnRegalia
Special Collaborator
Honorary Collaborator
Joined: April 22 2005
Location: Sweden
Status: Online
Points: 21149
|
Posted: September 13 2005 at 08:16 |
^ That's the X-Fi.
BTW: A lot of manufacturers of professional audio equipment offer sound interfaces for the PC which arebased on PCI. I'm curious how the bus affects the audio quality ... there aren't any bit errors, I certain of that.
|
|
|
oliverstoned
Special Collaborator
Honorary Collaborator
Joined: March 26 2004
Location: France
Status: Offline
Points: 6308
|
Posted: September 13 2005 at 08:17 |
maidenrulez wrote:
goose wrote:
oliverstoned wrote:
You can't prove it "scientifically". It's just like proving that your favourite prog band is better than the last Britney spear record. But when you hear a good tube amp versus a transistor, there's no need to explain. There's only music and emotion. |
You've said before it doesn't matter if people prefer digital to analogue because analogue is better. But all you can prove by listening is that you prefer the sound of analogue, not that it's any more accurate than digital.
Re the "cutting off" of sounds. There are only two losses that occur during processing from analogue to CD-audio (PCM)
Firstly the cutting off of any frequencies above 22.05kHz. This is basically irrelevant from a musical point of view - although these frequencies can affect people, they're inaudible.
Secondly the rounding up or down of every sample to one of 65,536 values. I can't prove that that isn't audible (quite possibly it is, I'll work on test files within the next couple of days) but consider that DVD-A is capable of 16,777,216 values and up to 48kHz. If any other medium sounds particularly different to this, then it must be less accurate. Maybe that sounds nicer to some peoples' tastes, but science really is real.
|
Now if you look at schematics over the amplitude of transistors when they distort you will see that the amplitude turns alot sharper than the amplitude over the tubes. Just like in digital signals the signals can only be either 1 or 0 there is not a smooth transition between these two so you can say really about the transistors they are either not distorted or distorted it is no smooth transiton between these two |
Actually, everybody agrees that tubes sound is better.
And for analog, some can't stand a few cracks on vynil or tape noise. But they agree on the fact that it's bettern excepting these details.
|
|
goose
Forum Senior Member
Joined: June 20 2004
Location: United Kingdom
Status: Offline
Points: 4097
|
Posted: September 13 2005 at 08:19 |
...but do they sound more accurate?
|
|
oliverstoned
Special Collaborator
Honorary Collaborator
Joined: March 26 2004
Location: France
Status: Offline
Points: 6308
|
Posted: September 13 2005 at 08:23 |
Yes, it goes much further than any numeric source, i will repeat endlessly. Low, high, dynamic, image, everthing (with a GOOD analog source).
|
|
cobb
Forum Senior Member
Joined: July 10 2005
Location: Australia
Status: Offline
Points: 1149
|
Posted: September 13 2005 at 08:33 |
Hey guys, I've just done some research on the MPC cable from cdrom to
soundcard. I did this because I know for a fact that I have installed
computers without a cable attached and they have not produced sound
from the cdrom. Now, it appears that the cable is a direct connection
to stream audio to a soundcard and if the MPC cable is present this is
how the audio will be fed to the card. The CPU has nothing to do with
sound other than to start the player. If the cable is not present then
WDM drivers must be present on the system so the standard bus can be
used. The WDM drivers will encourage the problems that have been
discussed. Am I correct in this?
|
|
MikeEnRegalia
Special Collaborator
Honorary Collaborator
Joined: April 22 2005
Location: Sweden
Status: Online
Points: 21149
|
Posted: September 13 2005 at 08:51 |
cobb wrote:
Hey guys, I've just done some research on the MPC cable from cdrom to soundcard. I did this because I know for a fact that I have installed computers without a cable attached and they have not produced sound from the cdrom. Now, it appears that the cable is a direct connection to stream audio to a soundcard and if the MPC cable is present this is how the audio will be fed to the card. The CPU has nothing to do with sound other than to start the player. If the cable is not present then WDM drivers must be present on the system so the standard bus can be used. The WDM drivers will encourage the problems that have been discussed. Am I correct in this? |
Which problems?
|
|
|
cobb
Forum Senior Member
Joined: July 10 2005
Location: Australia
Status: Offline
Points: 1149
|
Posted: September 13 2005 at 09:00 |
maidenrulez wrote:
There has even been sound card with TUBES
on...regardless of how much money you spend on your soundcard it will
not sound right as the PCI-INTERFACE in wich the soundcards are put is
not designed for optimal sound performance at all |
The one that is alluded to above. I should have said may
in my final sentence, as I don't know whether or not WDM drivers have
any inherent problems in sound reproduction. But even if they do, the
point is moot if there is an MPC cable connecting the cdrom to
soundcard, as the PCI bus will not be used for sound transference.
[edit] But the whole point of this is moot as well, because I know that
you are using the bus to transfer sound to your card, as you are
generating the original signal from harddrive, which can only travel on
the bus, via the wdm drivers? So perhaps you might need to look into
this?
Edited by cobb
|
|
MikeEnRegalia
Special Collaborator
Honorary Collaborator
Joined: April 22 2005
Location: Sweden
Status: Online
Points: 21149
|
Posted: September 13 2005 at 09:02 |
cobb wrote:
maidenrulez wrote:
There has even been sound card with TUBES on...regardless of how much money you spend on your soundcard it will not sound right as the PCI-INTERFACE in wich the soundcards are put is not designed for optimal sound performance at all |
The one that is alluded to above. I should have said may in my final sentence, as I don't know whether or not WDM drivers have any inherent problems in sound reproduction. But even if they do, the point is moot if there is an MPC cable connecting the cdrom to soundcard, as the PCI bus will not be used for sound transference.
|
- There's nothing wrong with the PCI bus, it does not change the audio data in any way
- Using the analog connection between drive and soundcard is about 10 times worse.
|
|
|
cobb
Forum Senior Member
Joined: July 10 2005
Location: Australia
Status: Offline
Points: 1149
|
Posted: September 13 2005 at 09:14 |
MikeEnRegalia wrote:
- Using the analog connection between drive and soundcard is about 10 times worse.
|
Where does this information come from? I'd be interested to read up on
it. Why do we connect the cd and card together if this is worse?
[edit] or did I miss it in a previous post?
Edited by cobb
|
|
MikeEnRegalia
Special Collaborator
Honorary Collaborator
Joined: April 22 2005
Location: Sweden
Status: Online
Points: 21149
|
Posted: September 13 2005 at 09:21 |
cobb wrote:
MikeEnRegalia wrote:
- Using the analog connection between drive and soundcard is about 10 times worse.
|
Where does this information come from? I'd be interested to read up on it. Why do we connect the cd and card together if this is worse?
|
This was done more than five years ago, when sound cards were limited to the ISA bus.
What happens is this:
- The drive reads the digital audio and converts it to analog. This is not the primary function of a computer CD drive, so not the best D/A converters are used, and the result is a reasonably good signal at best, a bad jittery one at worst.
- Then the signal is routed via a cheap analog cable without proper shielding to the sound card.
- The sound card converts the signal to digital again. It's like recording analog audio through the line in - it has to be digitized, and this ALWAYS results in signal degradation. The signal is also sometimes distorted, or sometimes too weak, because some CDs are "louder" than others ...
- From here on the signal path is the same as for digitally extracted audio. But in the following D/A conversion (either in the amp or the soundcard) the errors introduced by the previous A/D conversions multiply.
Edited by MikeEnRegalia
|
|
|
cobb
Forum Senior Member
Joined: July 10 2005
Location: Australia
Status: Offline
Points: 1149
|
Posted: September 13 2005 at 09:28 |
Okay- thanks Mike. So I should pull out the cord then! Not too sure
that WindowsDM driver would not be just as buggy as any other Windows
product, though
|
|
MikeEnRegalia
Special Collaborator
Honorary Collaborator
Joined: April 22 2005
Location: Sweden
Status: Online
Points: 21149
|
Posted: September 13 2005 at 09:37 |
cobb wrote:
Okay- thanks Mike. So I should pull out the cord then! Not too sure that WindowsDM driver would not be just as buggy as any other Windows product, though |
You can remove the connection and test if the audio still works ... it should, unless it's a really old soundcard (like from the 90's).
|
|
|
cobb
Forum Senior Member
Joined: July 10 2005
Location: Australia
Status: Offline
Points: 1149
|
Posted: September 13 2005 at 09:45 |
Forget all that- The DSP 24 is using its own WDM drivers, no matter
that I have the MPC cable connected. Sh*t it's been so long since I
looked at the settings on the card, I had completely forgotten.
Edited by cobb
|
|
goose
Forum Senior Member
Joined: June 20 2004
Location: United Kingdom
Status: Offline
Points: 4097
|
Posted: September 13 2005 at 11:34 |
oliverstoned wrote:
Yes, it goes much further than any numeric source, i will repeat endlessly. Low, high, dynamic, image, everthing (with a GOOD analog source). |
I'm not saying the sound isn't nicer to listen to!
If it has better lows, how do you know that that's because the recording hasn't been made with enough lows for your tastes and the turntable isn't compensating for this?
If it has better highs, how do you know that that's because the recording hasn't been made with enough highs for your tastes and the turntable isn't compensating for this?
Dynamic range of most CDs is lower than most vinyls and that's a fact. That's to do with the loudness race rather than limitations of the format.
Stereo image, again there are 65,536 different "positions" at half of max volume possible on CD (I think, maybe my reasonings a bit askew). I can produce tests for that too.
|
|
oliverstoned
Special Collaborator
Honorary Collaborator
Joined: March 26 2004
Location: France
Status: Offline
Points: 6308
|
Posted: September 13 2005 at 11:38 |
Theory...
A good listening is better than nay speech!
You would understand in 10 seconds!
|
|
Donate monthly and keep PA fast-loading and ad-free forever.