Arkangyl suggested different soundcards depending on your OS, so I am wondering if a different sound card would be optimal if you are using Windows 7 (which hasn't officially been released yet, only the RC). If anyone knows the answer that would be helpful
![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
Arkangyls Great Big Audio Post
- Thread starter Arkangyl
- Start date
Arkangyl suggested different soundcards depending on your OS, so I am wondering if a different sound card would be optimal if you are using Windows 7 (which hasn't officially been released yet, only the RC). If anyone knows the answer that would be helpful
A couple things I need to correct:
1) Your info on THD specs isn't right. Speakers especially but amps as well do not have a consistent THD across their entire range. The question is not what THD something has, but what THD at a given level. What Logitech is doing is listing the power level at which the amplifier hits 10% THD. It is a somewhat misleading way of giving a higher power spec for the amp. That does not mean that it operates that high normally, however.
For example the Samson Servo 120a amp I use to power my rear speakers is spec'd to 50 watts at 0.05% THD, 55 watts at 1% THD. What this means is the transistors are operating outside their linear range past 50 watts, but are still capable of generating more power. It could, in fact, go past that though not much before its protection circuitry kicks in and shuts the amp down to protect the speakers.
Also speakers rarely have very useful THD specs. To the extent you find them, they list the level at which they produce 1% of the 2nd and 3rd harmonics. As you go above that level, it will increase. The more cone movement you have, the more harmonic distortion.
2) With regards to soundstage, you are just not correct. While a good 2 channel setup will give you a more convincing front image, as in a solid wall of sound rather than sounding like discreet speakers, they aren't going to give you the effect of a speaker behind you. Even if you have a recording set up for the precise speaker positioning, it doesn't work well. To get surround from two channels, you have to go with something close to the ears like headphones or earbuds and do a binaural recording.
I've got some high end speakers on the front of my computer (SVS MTS-01s) and while they offer imaging that blows away any computer speakers for two channel music, even a cheap speaker behind me reproduces something that sounds like it comes from the back better.
So while 2.0/2.1 is great for music and some kinds of gaming, if you desire positional surround sound, you can't get away from the need for rear speakers. For that matter it beats out headphones too since while headphones can do a pretty accurate job simulating a full sound field with good input, the sound field moves with your head. Your brain notices that as incorrect. Small head movements are part of the way humans localize sound.
There isn't a right answer in terms of more expensive 2.0 vs cheaper 5.1, it really depends on what you want. Personally if I had to choose I'd probably go the 2.0 route, but I'd really have trouble giving up the rear speakers. Of course personally I like the very expensive 5.1 option
.
3) With regards to SNR you are right in general, but wrong that 40dB is plenty. The problem is that when you are talking about SNR in an electrical system, it is also usually synonymous with dynamic range, and is referenced to peak output. What's that means? Well if you have an amp that has 80dB SNR that means that the smallest details it can reproduce are 80dB below it's rated output, and they'll probably be a bit mangled at that level.
So why's that matter? Well suppose your amp is 100 watts, however you are listening at a normal level with speakers close to you and thus running the amp at only 1 watt. Well the amp now only has 60dB SNR since it's power is 20dBW and you are only using 1dBW. Now that's still fine you think. After all if you are listening at 80dBSPL or so in a quiet room that has 30dBSPL of background noise. It's still under the background noise.
The problem is that humans can hear well below a noise threshold. How much depends on the signal and the kind of noise, but you are perfectly capable of hearing signals under the noise threshold. 10dB is easy, 20dB isn't uncommon.
So in this case, what you've got is an amp messing up the low signal. This isn't so much that it would sound awful, but if you switched over to a good amp, you'd notice the increase in clarity of the music. Maintaining a high SNR throughout the electrical signal is advantageous, when it can be done.
3) With regards to wire you are correct that better shielded wire can be useful, but only for line level interconnects. Speaker wire is not shielded, and does not need to be. The signal to speakers is low impedance, balanced, and high voltage. As such atmospheric noise is insufficient to induce any sort of response. For that, thick line cord is the best idea. Low cable impedance is the most important factor.
You pretty much have to run your speaker next to high current mains cables to induce an audible response.
Now line level interconnects are high impedance, unbalanced, low voltage connections and thus shielding is important. Good coaxial cable is the way to go. However you still needn't spend much on it. If you have an extremely noisy environment or are just worried, professional Belden cable can be had from people like Bluejeans cable and really, there isn't much point past that.
1) Your info on THD specs isn't right. Speakers especially but amps as well do not have a consistent THD across their entire range. The question is not what THD something has, but what THD at a given level. What Logitech is doing is listing the power level at which the amplifier hits 10% THD. It is a somewhat misleading way of giving a higher power spec for the amp. That does not mean that it operates that high normally, however.
For example the Samson Servo 120a amp I use to power my rear speakers is spec'd to 50 watts at 0.05% THD, 55 watts at 1% THD. What this means is the transistors are operating outside their linear range past 50 watts, but are still capable of generating more power. It could, in fact, go past that though not much before its protection circuitry kicks in and shuts the amp down to protect the speakers.
Also speakers rarely have very useful THD specs. To the extent you find them, they list the level at which they produce 1% of the 2nd and 3rd harmonics. As you go above that level, it will increase. The more cone movement you have, the more harmonic distortion.
2) With regards to soundstage, you are just not correct. While a good 2 channel setup will give you a more convincing front image, as in a solid wall of sound rather than sounding like discreet speakers, they aren't going to give you the effect of a speaker behind you. Even if you have a recording set up for the precise speaker positioning, it doesn't work well. To get surround from two channels, you have to go with something close to the ears like headphones or earbuds and do a binaural recording.
I've got some high end speakers on the front of my computer (SVS MTS-01s) and while they offer imaging that blows away any computer speakers for two channel music, even a cheap speaker behind me reproduces something that sounds like it comes from the back better.
So while 2.0/2.1 is great for music and some kinds of gaming, if you desire positional surround sound, you can't get away from the need for rear speakers. For that matter it beats out headphones too since while headphones can do a pretty accurate job simulating a full sound field with good input, the sound field moves with your head. Your brain notices that as incorrect. Small head movements are part of the way humans localize sound.
There isn't a right answer in terms of more expensive 2.0 vs cheaper 5.1, it really depends on what you want. Personally if I had to choose I'd probably go the 2.0 route, but I'd really have trouble giving up the rear speakers. Of course personally I like the very expensive 5.1 option
.3) With regards to SNR you are right in general, but wrong that 40dB is plenty. The problem is that when you are talking about SNR in an electrical system, it is also usually synonymous with dynamic range, and is referenced to peak output. What's that means? Well if you have an amp that has 80dB SNR that means that the smallest details it can reproduce are 80dB below it's rated output, and they'll probably be a bit mangled at that level.
So why's that matter? Well suppose your amp is 100 watts, however you are listening at a normal level with speakers close to you and thus running the amp at only 1 watt. Well the amp now only has 60dB SNR since it's power is 20dBW and you are only using 1dBW. Now that's still fine you think. After all if you are listening at 80dBSPL or so in a quiet room that has 30dBSPL of background noise. It's still under the background noise.
The problem is that humans can hear well below a noise threshold. How much depends on the signal and the kind of noise, but you are perfectly capable of hearing signals under the noise threshold. 10dB is easy, 20dB isn't uncommon.
So in this case, what you've got is an amp messing up the low signal. This isn't so much that it would sound awful, but if you switched over to a good amp, you'd notice the increase in clarity of the music. Maintaining a high SNR throughout the electrical signal is advantageous, when it can be done.
3) With regards to wire you are correct that better shielded wire can be useful, but only for line level interconnects. Speaker wire is not shielded, and does not need to be. The signal to speakers is low impedance, balanced, and high voltage. As such atmospheric noise is insufficient to induce any sort of response. For that, thick line cord is the best idea. Low cable impedance is the most important factor.
You pretty much have to run your speaker next to high current mains cables to induce an audible response.
Now line level interconnects are high impedance, unbalanced, low voltage connections and thus shielding is important. Good coaxial cable is the way to go. However you still needn't spend much on it. If you have an extremely noisy environment or are just worried, professional Belden cable can be had from people like Bluejeans cable and really, there isn't much point past that.
Sycraft said:
True, the main point I was trying to get across is to consider THD in general though, and IMO the Logitech's are worse than the competition and the THD number is a good way to illustrate that point. Obviously a 'true' THD number isn't a set number so much as a parabola that needs to be seen visually as it's a function of frequency and wattage where the limits have much more distortion than the sweet spots.
Sycraft said:2) With regards to soundstage, you are just not correct. While a good 2 channel setup will give you a more convincing front image, as in a solid wall of sound rather than sounding like discreet speakers, they aren't going to give you the effect of a speaker behind you. Even if you have a recording set up for the precise speaker positioning, it doesn't work well. To get surround from two channels, you have to go with something close to the ears like headphones or earbuds and do a binaural recording.
*cut to save space*
There isn't a right answer in terms of more expensive 2.0 vs cheaper 5.1, it really depends on what you want. Personally if I had to choose I'd probably go the 2.0 route, but I'd really have trouble giving up the rear speakers. Of course personally I like the very expensive 5.1 option
Again, you're absolutely right, however my point was more that for a $100-200 budget, a 2.0 / 2.1 system will create a much better overall sound image than a 5.1 system in that price range.
Personally I run Swan D1080's on my desktop (replacing Z-5500's) and it sounds much, much more like I'm being enveloped in the sound rather than just having the sound coming at me from different angles. With music and movies (which, in .mp3 and .avi formats, are largely 2.0/.1 based) the audio is far superior for half or a third the cost ($125 vs $200-300). I also run 6.1 for my HT with a Onkyo 606 receiver and Elemental Design speakers. The 2.0 by no means is the same experience as 6.1 HT audio, however the HT setup cost much, much more. It's basically a quality > quantity argument, personally I feel that the Swan D1080 give me far superior audio in all realms than the Logitech X-540 (or w/e their < $100 5.1 is called).
As for everything else, thanks for the input! Glad to see this thread is still alive and growing. I have a Computer Engineering background and I've done a decent bit of reading and learning about how audio systems work but my knowledge is certainly not complete. Always great to get these long replies with corrections & updates!
Excellent OP; coming from a fellow ECE.
Can someone clarify on this?
This doesn't make sense to me. 128/192/320 are the bit rates for the files; a 5 minute mp3 at 128 bps should produce exactly the same file size as a 5 minute aac at 128 bps.
Can someone clarify on this?
Depends on the mode of compression used. MP3s are often compressed as CBR, constant bitrate. That means that they use precisely a certain amount of data per second. AAC is often compressed VBR, variable bitrate. That means you give it a target bitrate and it shoots to get near that, but uses more or less as needed in the particular parts of the song.
So the bit comparisons aren't really meaningful. In CBR mode, 128kbps files will always be the same size. There might be a few bytes difference because of headers and such, but that's it. In VBR mode, they will vary based on the type of music. Some songs will be larger, some smaller, there isn't a defined size.
As an average point you can say 5 minutes * 60 seconds/minute * 128kbits/second / 8 bits/byte / 1024 kb/mb = 4.69mb. On average a 5 minute 128k file will be that size. Note the MP3 is, which implies it was CBR.
The easy rule of thumb is 1mb per minute for 128k. In reality it is a little less than that and as I said, VBR can vary, but it gives you an easy mental estimate. For 256k, simply double it, 2mb per minute.
So the bit comparisons aren't really meaningful. In CBR mode, 128kbps files will always be the same size. There might be a few bytes difference because of headers and such, but that's it. In VBR mode, they will vary based on the type of music. Some songs will be larger, some smaller, there isn't a defined size.
As an average point you can say 5 minutes * 60 seconds/minute * 128kbits/second / 8 bits/byte / 1024 kb/mb = 4.69mb. On average a 5 minute 128k file will be that size. Note the MP3 is, which implies it was CBR.
The easy rule of thumb is 1mb per minute for 128k. In reality it is a little less than that and as I said, VBR can vary, but it gives you an easy mental estimate. For 256k, simply double it, 2mb per minute.
QUESTION!
there seem to be different "HD audio codecs" ? called : "Core3D PCIe DSP", "ALC898", "X-Fi CA20K2 PCIe", "Game blaster" maybe more?
are "HD audio codecs" only something that is used on motherboards and sound cards use something else? I found here a comparative of some of these audio codecs from soundmax, realtek, VIA, C-media, and creative. Auzentech, Asus (and maybe some others) aren't in the list.
I'd like to be able to understand what "HD audio codecs" are and if they are present in sound cards or only motherboards and know which one is currently the best.
there seem to be different "HD audio codecs" ? called : "Core3D PCIe DSP", "ALC898", "X-Fi CA20K2 PCIe", "Game blaster" maybe more?
are "HD audio codecs" only something that is used on motherboards and sound cards use something else? I found here a comparative of some of these audio codecs from soundmax, realtek, VIA, C-media, and creative. Auzentech, Asus (and maybe some others) aren't in the list.
I'd like to be able to understand what "HD audio codecs" are and if they are present in sound cards or only motherboards and know which one is currently the best.
All the things you named are sound cards. You also necro'ed a thread that died in 2009. But at least it was in he name of learning things. You haven't bought any expensive power cables recently have you?
B00nie
[H]F Junkie
- Joined
- Nov 1, 2012
- Messages
- 9,327
Arkangyl for an audio engineer your knowledge is very entry level and simplistic - or perhaps you're just trying to keep the discussion at layman level?
The main problems with speaker design usually occur at directivity and power response level if you eliminate the obvious i.e. bad quality drivers or enclosure. It is relatively simple to make a speaker with a flat frequency response but it's exponentially difficult to create a speaker that does so while having a perfect power response and off-axis radiation control.
The frequency range starting from 100hz to 25Khz can easily be produced and with excellent results using full range ESL:s for example. In that sense you're right that from a 3" cone it's pretty much impossible to obtain reasonable full range audio. The sub range does not start from 30hz to 50 but from 5hz to 40hz on practical solutions.
Also your input on crossovers was awfully simplistic. The common problems a crossover designer has to haggle with have little to do with speaker element 'frequency range' per se. They're simple issues compared to design choices forced by the analog operation (impedance curve, phase behaviour, power response and offset radiation profiles etc). Also a crossover does not need to be (or even should be) at the speaker level. Active crossovers using multiple amplifiers yield the best results and are simpliest to design as they're immune to analog problems of the speaker.
In my experience you can invest at least tenfold amount to speakers before you even begin to approach the technical limits of a basic DAC or amplifier. Your analog to 128kbit MP3 is faulty as bitrate has nothing to do with the hardware it's played on.
ABX testing quickly and mercilessly reveals differences with speakers but 100dollar and 2000 dollar DACs fail to yield any statistically strong results. This reveals the magnitude of problems.
The main problems with speaker design usually occur at directivity and power response level if you eliminate the obvious i.e. bad quality drivers or enclosure. It is relatively simple to make a speaker with a flat frequency response but it's exponentially difficult to create a speaker that does so while having a perfect power response and off-axis radiation control.
The frequency range starting from 100hz to 25Khz can easily be produced and with excellent results using full range ESL:s for example. In that sense you're right that from a 3" cone it's pretty much impossible to obtain reasonable full range audio. The sub range does not start from 30hz to 50 but from 5hz to 40hz on practical solutions.
Also your input on crossovers was awfully simplistic. The common problems a crossover designer has to haggle with have little to do with speaker element 'frequency range' per se. They're simple issues compared to design choices forced by the analog operation (impedance curve, phase behaviour, power response and offset radiation profiles etc). Also a crossover does not need to be (or even should be) at the speaker level. Active crossovers using multiple amplifiers yield the best results and are simpliest to design as they're immune to analog problems of the speaker.
In my experience you can invest at least tenfold amount to speakers before you even begin to approach the technical limits of a basic DAC or amplifier. Your analog to 128kbit MP3 is faulty as bitrate has nothing to do with the hardware it's played on.
ABX testing quickly and mercilessly reveals differences with speakers but 100dollar and 2000 dollar DACs fail to yield any statistically strong results. This reveals the magnitude of problems.
[Q
In sort:
- If you have Vista, get an Asus Xonar card
- If you have XP, get an Auzentech X-Fi card
- Creative Labs cards are sort of pariahs due to Creative's treatment of its customers (intentionally disabled features in Vista on older cards to encourage upgrading, prosecuted a 3rd party driver maker when he made vastly superior drivers). CL cards have pretty good DACs but really shitty amps, there are mods out there to make them better, but Auzentech & Asus do a better job.
acording all forums read ,am getting the picture that asus xonar is best card ; and is best for audio; for gaming there are some other better ;now' why you suggest the asus just for vista ?
In sort:
- If you have Vista, get an Asus Xonar card
- If you have XP, get an Auzentech X-Fi card
- Creative Labs cards are sort of pariahs due to Creative's treatment of its customers (intentionally disabled features in Vista on older cards to encourage upgrading, prosecuted a 3rd party driver maker when he made vastly superior drivers). CL cards have pretty good DACs but really shitty amps, there are mods out there to make them better, but Auzentech & Asus do a better job.
acording all forums read ,am getting the picture that asus xonar is best card ; and is best for audio; for gaming there are some other better ;now' why you suggest the asus just for vista ?