| Home |
| Search |
| Today's Posts |
|
#1
|
|||
|
|||
Help:bought Audiophile 2496;sounds just like my SB Live!
I just bought the M-Audio Audiophile 2496, hoping to get better sound
than my Sound Blaster Live. Audiophile is 24 bit, 96 khz. SB is 16 bit, 44.1 khz. In Audobe Audition, I try recording the exact same thing (a short piano piece) on each card for a back-to-back comparison, but both recordings sound the same. I have a basic understanding of bit depth and sample rate, but I don't know how the bit depth and sample rate of my sound card relates to the BD and SR that I set my recording session to, and what occurs if I then convert the BD and SR. So could it have something to do with those settings? Please let me know, in plain English, when I will be able to hear the difference between different BD's and SR's, and between these two sound cards. So much rave about the Audiphile cards, and negativity towards Sound Blaster, and I can't hear the difference! |
|
#2
|
|||
|
|||
|
"Bob" wrote in message om... I just bought the M-Audio Audiophile 2496, hoping to get better sound than my Sound Blaster Live. Audiophile is 24 bit, 96 khz. SB is 16 bit, 44.1 khz. In Audobe Audition, I try recording the exact same thing (a short piano piece) on each card for a back-to-back comparison, but both recordings sound the same. I have a basic understanding of bit depth and sample rate, but I don't know how the bit depth and sample rate of my sound card relates to the BD and SR that I set my recording session to, and what occurs if I then convert the BD and SR. So could it have something to do with those settings? Please let me know, in plain English, when I will be able to hear the difference between different BD's and SR's, and between these two sound cards. So much rave about the Audiphile cards, and negativity towards Sound Blaster, and I can't hear the difference! Nobody said you would necessarily be able to hear the difference. You will be able to measure it though. Get a copy of RMAA, which is freeware, and test both cards. You will see that the AP2496 gives you much greater margin for recording error. TonyP. |
|
#3
|
|||
|
|||
|
"Bob" wrote in message
om I just bought the M-Audio Audiophile 2496, hoping to get better sound than my Sound Blaster Live. Audiophile is 24 bit, 96 kHz. SB is 16 bit, 44.1 kHz. Which SB Live! are you talking about? There are at least three major permutations of this hardware. The first two had definite technical issues that caused audible coloration but a more recent version called the SB Live! 5.1 was improved to the point where the audible coloration can be difficult or impossible to hear under a wider range of real-world circumstances. In Adobe Audition, I try recording the exact same thing (a short piano piece) on each card for a back-to-back comparison, but both recordings sound the same. IME this is potentially a valid procedure depending on the associated gear used to record and play back the piano piece. Key issues are to use recording and monitoring equipment of sufficient quality, and to time-synch and level-match the samples being compared. Recordings being compared need to be played back in quick succession for maximum listener sensitivity. For maximum sensitivity, a number of different sound sources, and a number of different instrumental sounds should be used. There should also be some bias controls, as listener foreknowledge of which alternative is being played can be distracting and lead to spurious results. A number of programs, mostly freeware that assist with making close, sensitive, and bias-controlled listening tests can be found at www.pcabx.com . A set of recordings made with the second revision of the SB Live! can be found at http://www.pcabx.com/product/ct4830/index.htm . A number of these test musical samples are known to produce positive results for audible differences because of the technical flaws previously mentioned. If you work with them you will notice that different kinds of music make technical flaws more or less audible. One of the dirty little secrets about listening tests is that not all music reveals all audible differences due to technical flaws. A listening test with one piece of music can give positive results, and another listening test with a different piece of music can give negative results, even when associated recording and playback equipment are of the highest quality, the listener is properly trained and sensitized to the difference, etc. For example, consider the SBLive! 4830. The technical flaw that is probably most audible with this audio interface relates to upper midrange coloration. While upper midrange content is fairly common in music, various musical sounds differ in the degree to which they make upper midrange colorations audible. Musical sounds can have weak upper midrange content, or the content that is there can be chaotic, and therefore tend to conceal differences in this critical area. Or, stronger content in other musical ranges can distract the listener. I have a basic understanding of bit depth and sample rate, but I don't know how the bit depth and sample rate of my sound card relates to the BD and SR that I set my recording session to, and what occurs if I then convert the BD and SR. Audition has three basic BDs, 8, 16, and 32 bits. The 32 bit format is floating point and has immense dynamic range. In any case you can't make or playback recordings with effective BDs that exceed those that the particular card supports, and you can't exploit high BD recordings with an audio interfaces that don't support those BDs. Audition tends to make these differences less obvious from a UI standpoint, as it will automatically downsample and upsample to mitigate mismatches between the audio interface and the recorded format. So could it have something to do with those settings? Audition will allow you to record and playback a 8 bit recording with a 32 bit audio interface, or record and playback a 32 bit recording with a 16 bit interface for your convenience, but in the end the rule of the weakest link will still limit the quality of your recordings and how they sound when you play them back. Please let me know, in plain English, when I will be able to hear the difference between different BD's and SR's, and between these two sound cards. It all depends on the various issues that I've raised, and probably some other ones as well. Furthermore, some of the benefits of better audio interfaces relate to safety margins. A card with better dynamic range gives you more wiggle room for practical suboptimalities. For example one card can have 80 dB dynamic range and another one can have 120 dB dynamic range. Depending on other parameters, headroom, quality of associated equipment, the room you record and listen in, and the basic nature of the actual musical event being recorded, the dynamic range of the recording might top out at 70 dB. If you get all the levels optimized, both cards are capable of making a recording with about 70 dB worth of dynamic range. However, you only have 10 dB worth of margin with the 80 dB DR card, while you have a whopping 50 dB of margin with the 120 dB dynamic range card. You are much more likely to exploit the dynamic range of the basic musical event with the better card, but if you do everything right, and shave all the margins optimally, the lesser card just might deliver roughly comparable results. So much rave about the Audiophile cards, and negativity towards Sound Blaster, and I can't hear the difference! It all depends... |
|
#6
|
|||
|
|||
|
"Arny Krueger" wrote in message ...
"Bob" wrote in message om I just bought the M-Audio Audiophile 2496, hoping to get better sound than my Sound Blaster Live. Audiophile is 24 bit, 96 kHz. SB is 16 bit, 44.1 kHz. Thanks very much for your detailed reply. Many helpful suggestions. I have a couple more questions. Key issues are to use recording and monitoring equipment of sufficient quality, and to time-synch and level-match the samples being compared. Don't understand what time-sync and level match mean. Audition has three basic BDs, 8, 16, and 32 bits. The 32 bit format is floating point and has immense dynamic range. In any case you can't make or playback recordings with effective BDs that exceed those that the particular card supports, and you can't exploit high BD recordings with an audio interfaces that don't support those BDs. Audition tends to make these differences less obvious from a UI standpoint, as it will automatically downsample and upsample to mitigate mismatches between the audio interface and the recorded format. Does this mean that if my card is advertised as having 24-bit 96khz quality, I won't get any use out of the 32 bit setting in audition? Or if I use my SB, which is a 16 bit/44.1khz card it won't do me any good to set the sample rate to 96khz at the beggining of my session? Also, what is floating point? Furthermore, some of the benefits of better audio interfaces relate to safety margins. A card with better dynamic range gives you more wiggle room for practical suboptimalities. For example one card can have 80 dB dynamic range and another one can have 120 dB dynamic range. Depending on other parameters, headroom, quality of associated equipment, the room you record and listen in, and the basic nature of the actual musical event being recorded, the dynamic range of the recording might top out at 70 dB. If you get all the levels optimized, both cards are capable of making a recording with about 70 dB worth of dynamic range. However, you only have 10 dB worth of margin with the 80 dB DR card, while you have a whopping 50 dB of margin with the 120 dB dynamic range card. You are much more likely to exploit the dynamic range of the basic musical event with the better card, but if you do everything right, and shave all the margins optimally, the lesser card just might deliver roughly comparable results. Does this mean that if my card has a bigger dynamic range, my musician can afford to play louder without causing clipping or distortion in the recording? |
|
#7
|
|||
|
|||
|
"Arny Krueger" wrote in message ...
"Bob" wrote in message om I just bought the M-Audio Audiophile 2496, hoping to get better sound than my Sound Blaster Live. Audiophile is 24 bit, 96 kHz. SB is 16 bit, 44.1 kHz. Thanks very much for your detailed reply. Many helpful suggestions. I have a couple more questions. Key issues are to use recording and monitoring equipment of sufficient quality, and to time-synch and level-match the samples being compared. Don't understand what time-sync and level match mean. Audition has three basic BDs, 8, 16, and 32 bits. The 32 bit format is floating point and has immense dynamic range. In any case you can't make or playback recordings with effective BDs that exceed those that the particular card supports, and you can't exploit high BD recordings with an audio interfaces that don't support those BDs. Audition tends to make these differences less obvious from a UI standpoint, as it will automatically downsample and upsample to mitigate mismatches between the audio interface and the recorded format. Does this mean that if my card is advertised as having 24-bit 96khz quality, I won't get any use out of the 32 bit setting in audition? Or if I use my SB, which is a 16 bit/44.1khz card it won't do me any good to set the sample rate to 96khz at the beggining of my session? Also, what is floating point? Furthermore, some of the benefits of better audio interfaces relate to safety margins. A card with better dynamic range gives you more wiggle room for practical suboptimalities. For example one card can have 80 dB dynamic range and another one can have 120 dB dynamic range. Depending on other parameters, headroom, quality of associated equipment, the room you record and listen in, and the basic nature of the actual musical event being recorded, the dynamic range of the recording might top out at 70 dB. If you get all the levels optimized, both cards are capable of making a recording with about 70 dB worth of dynamic range. However, you only have 10 dB worth of margin with the 80 dB DR card, while you have a whopping 50 dB of margin with the 120 dB dynamic range card. You are much more likely to exploit the dynamic range of the basic musical event with the better card, but if you do everything right, and shave all the margins optimally, the lesser card just might deliver roughly comparable results. Does this mean that if my card has a bigger dynamic range, my musician can afford to play louder without causing clipping or distortion in the recording? |
|
#8
|
|||
|
|||
|
Also, could it be some settings that I don't have right on my ASIO
drivers, in my program under device properties and settings, or on my soundcard software, that might make the difference between these cards more obvious and optimize the sound of the Audiophile? |
|
#9
|
|||
|
|||
|
On 16 Aug 2004 19:45:01 -0700, (Bob) wrote:
Key issues are to use recording and monitoring equipment of sufficient quality, and to time-synch and level-match the samples being compared. Don't understand what time-sync and level match mean. He's talking about taking a rigorous approach to comparing the two soundcards. Don't worry about it. if the card sounds good to you - enjoy :-) Does this mean that if my card is advertised as having 24-bit 96khz quality, I won't get any use out of the 32 bit setting in audition? Or if I use my SB, which is a 16 bit/44.1khz card it won't do me any good to set the sample rate to 96khz at the beggining of my session? Also, what is floating point? I don't know Audition particularly. But many quality multi-track programs run internally at a high bit-rate, quite independently of the chosen input and output sample rate. This gives headroom, making it almost impossible to run into overload within the mixer. All you have to do is check input levels and the final output level. Floating point is a method of mapping audio levels to bytes of data. You can't do anything about it. Nod sagely at the buzz-word and get on with your music :-) Your SB is actually a 48KHz card, with on-the fly re-sampling to other rates on input and output. This is perhaps its main limitation. But, yes, I see little point in setting up a 96KHz project if you have no 96KHz source. Does this mean that if my card has a bigger dynamic range, my musician can afford to play louder without causing clipping or distortion in the recording? Sort of. Actually, it allows you to set levels LOWER, without falling into the noise floor. CubaseFAQ www.laurencepayne.co.uk/CubaseFAQ.htm "Possibly the world's least impressive web site": George Perfect |
|
#10
|
|||
|
|||
|
"Bob" wrote in message
om Thanks for the input. I just thought that better sample rate and bit depth equals higher fidelity, which I thought would be audible. Not true. If you do some listening to the files you can download from www.pcabx.com you will find out that a really good 16 bit card, operating at optimal levels, does little or nothing to reduce sound quality. What do you mean when you say "greater margin for recording error"? What is recording error? Level mismatch. Obviously, if your levels are too high, there will be clipping and audible distortion. However, the actual levels developed by a living human musician or vocalist during a performance are subject to considerable variations. Therefore, recording levels are usually set low and conservatively. Every 6 dB the levels go below the absolute max, you lose one bit's worth of resolution. In effect, the 16 bit card becomes a 15 bit card. Normal 10 dB or so worth of headroom costs you nearly 2 bits worth of resolution. If you don't get the levels just right, and have reasonable allowance for headroom, several bits of resolution are lost. Our nice little 16 bit audio interface becomes, practically speaking perhaps a 12 bit audio interface. The point where the loss of resolution becomes audible can be someplace between 13 and 15 bits. 12 bits might not cut it! But if you are really careful, and really lucky, you might not end up here. Contrast this with using a premium audio interface with 19 bits worth of resolution. If you have a 4 bit resolution loss due to level mismatch and your allowance for headroom, you are now down to 15 bits, which probably won't be a problem at all. |
|
#11
|
|||
|
|||
|
"Bob" wrote in message
m "Arny Krueger" wrote in message ... "Bob" wrote in message om I just bought the M-Audio Audiophile 2496, hoping to get better sound than my Sound Blaster Live. Audiophile is 24 bit, 96 kHz. SB is 16 bit, 44.1 kHz. Thanks very much for your detailed reply. Many helpful suggestions. I have a couple more questions. Key issues are to use recording and monitoring equipment of sufficient quality, and to time-synch and level-match the samples being compared. Don't understand what time-sync and level match mean. When you are comparing two recordings, the most sensitive comparison requires that the only difference be the effect you are listening for, in this case the difference between two sound cards. Of course the music you use for each test g must be absolutely identical - a very high quality recording can provide this kind of source. The basic loudness of the recordings must be absolutely identical, because you don't want to think that an audio interface sounds better just because it is a bit louder, which can easily happen. Also, the musical selections being compared have to start and end at the same time in the music. You want to be comparing apples and apples, the only audible difference being due to the inherent charateristics of the two audio interfaces. Audition has three basic BDs, 8, 16, and 32 bits. The 32 bit format is floating point and has immense dynamic range. In any case you can't make or playback recordings with effective BDs that exceed those that the particular card supports, and you can't exploit high BD recordings with an audio interfaces that don't support those BDs. Audition tends to make these differences less obvious from a UI standpoint, as it will automatically downsample and upsample to mitigate mismatches between the audio interface and the recorded format. Does this mean that if my card is advertised as having 24-bit 96khz quality, I won't get any use out of the 32 bit setting in audition? What it means is that when Audition records from a 24 bit interface, using Audioin's 32 bit mode, the 24 bit resolution limit of the interface rules. The law of the weakest link is in force, and your recording can't possibly have more than 24 bit resolution. The good news is that the 24 bits won't be further corrupted by the 32 bit resolution of Audition. Or if I use my SB, which is a 16 bit/44.1khz card it won't do me any good to set the sample rate to 96khz at the beggining of my session? Any recording made through a 16/44 audio interface is limited to the frequency response limits of 44 KHz sampling, no matter what what follows. Also, what is floating point? Floating point is a way for computers to store data with exceptional dynamic range of about 1000 dB. Just another way to keep the computer from limiting the resolution of the recording. Furthermore, some of the benefits of better audio interfaces relate to safety margins. A card with better dynamic range gives you more wiggle room for practical suboptimalities. For example one card can have 80 dB dynamic range and another one can have 120 dB dynamic range. Depending on other parameters, headroom, quality of associated equipment, the room you record and listen in, and the basic nature of the actual musical event being recorded, the dynamic range of the recording might top out at 70 dB. If you get all the levels optimized, both cards are capable of making a recording with about 70 dB worth of dynamic range. However, you only have 10 dB worth of margin with the 80 dB DR card, while you have a whopping 50 dB of margin with the 120 dB dynamic range card. You are much more likely to exploit the dynamic range of the basic musical event with the better card, but if you do everything right, and shave all the margins optimally, the lesser card just might deliver roughly comparable results. Does this mean that if my card has a bigger dynamic range, my musician can afford to play louder without causing clipping or distortion in the recording? If you set levels with enough headroom, yes. I just covered this in more detail in the response to you I just posted. |
|
#12
|
|||
|
|||
|
"Bob" wrote in message
om Also, could it be some settings that I don't have right on my ASIO drivers, in my program under device properties and settings, or on my soundcard software, that might make the difference between these cards more obvious and optimize the sound of the Audiophile? At this time Audition doesn't *know* that ASIO exists. |
|
#13
|
|||
|
|||
|
Just a thought;
I tended to find the difference between my old SB Live and my new M-Audio Delta 1010LT's became more apparent when I started mixing more and more recorded tracks together. It just seems to have more clarity. This is probably a combination of a bigger dynamic range, as a few others have mentioned, and less sonic distortion. I'm no expert, but I imagine distortion in a particular frequency range will increase when tracks are summed. Cheers, Mark. -- Bob wrote: I just bought the M-Audio Audiophile 2496, hoping to get better sound than my Sound Blaster Live. Audiophile is 24 bit, 96 khz. SB is 16 bit, 44.1 khz. In Audobe Audition, I try recording the exact same thing (a short piano piece) on each card for a back-to-back comparison, but both recordings sound the same. I have a basic understanding of bit depth and sample rate, but I don't know how the bit depth and sample rate of my sound card relates to the BD and SR that I set my recording session to, and what occurs if I then convert the BD and SR. So could it have something to do with those settings? Please let me know, in plain English, when I will be able to hear the difference between different BD's and SR's, and between these two sound cards. So much rave about the Audiphile cards, and negativity towards Sound Blaster, and I can't hear the difference! |
|
#14
|
|||
|
|||
|
"Arny Krueger" wrote in message ... "Bob" wrote in message om What do you mean when you say "greater margin for recording error"? What is recording error? Level mismatch. Obviously, if your levels are too high, there will be clipping and audible distortion. However, the actual levels developed by a living human musician or vocalist during a performance are subject to considerable variations. Therefore, recording levels are usually set low and conservatively. Every 6 dB the levels go below the absolute max, you lose one bit's worth of resolution. In effect, the 16 bit card becomes a 15 bit card. Normal 10 dB or so worth of headroom costs you nearly 2 bits worth of resolution. Yes, and since the Live is only good for a real 14 bits to start with, you can't waste too many more allowing for headroom. Get it all just right, and the SB live is OK, but there is no margin for error. The chances of clipping or excessive noise is high. TonyP. |
|
#15
|
|||
|
|||
|
"TonyP" wrote in message
u "Arny Krueger" wrote in message ... "Bob" wrote in message om What do you mean when you say "greater margin for recording error"? What is recording error? Level mismatch. Obviously, if your levels are too high, there will be clipping and audible distortion. However, the actual levels developed by a living human musician or vocalist during a performance are subject to considerable variations. Therefore, recording levels are usually set low and conservatively. Every 6 dB the levels go below the absolute max, you lose one bit's worth of resolution. In effect, the 16 bit card becomes a 15 bit card. Normal 10 dB or so worth of headroom costs you nearly 2 bits worth of resolution. Yes, and since the Live is only good for a real 14 bits to start with, you can't waste too many more allowing for headroom. This is a key point. Audio production is very much prone to slips and glitches, so safety margin is very important. To put things into perspective, Good quality line-level audio gear is capable of dynamic range on the order of 110-120 dB, sometimes more. If people use practice and procedures that work with that kind of massive headroom reserves, but this time with sound cards that have only 80 dB dynamic range, its a recipie for disaster. If you want a sound card with the same kind of dynamic range as good quality line-level audio gear, you need to move smartly towards the high end. I feel pretty comfortable with audio interfaces with performance in the 95-103 dB range. Get it all just right, and the SB live is OK, but there is no margin for error. The chances of clipping or excessive noise is high. Noise builds up as you add processing and tracks. |
|
#16
|
|||
|
|||
|
"Arny Krueger" wrote in message ... "TonyP" wrote in message Get it all just right, and the SB live is OK, but there is no margin for error. The chances of clipping or excessive noise is high. Noise builds up as you add processing and tracks. Not at all. The processing can be done at 32 bit FP (or even 64 bit!), and no need for track bouncing with digital. TonyP. |
|
#17
|
|||
|
|||
|
"TonyP" wrote in message
"Arny Krueger" wrote in message ... "TonyP" wrote in message Get it all just right, and the SB live is OK, but there is no margin for error. The chances of clipping or excessive noise is high. Noise builds up as you add processing and tracks. Not at all. The processing can be done at 32 bit FP (or even 64 bit!), That can be true, but the noise still builds up as you add processing. It's just that with 32 bit FP, it adds up at such a level that it is irrelevant. However, I was speaking in the context of the general case, which includes 16 bit processing. If you're making the point that you can track at 16 bits and then use 32 bit processingm, that point is well-taken. and no need for track bouncing with digital. I wasn't speaking of track bouncing. If you mix a lot of tracks, each one brings in its noise floor. Hopefully, each track's noise floor will be random and be attenuated during the mixing process. However, any discrete noise that is common to a number of tracks will add with simple arithmetic instead of adding geometrically like random signals. Things add up much faster with simple arithmetic. |
|
#18
|
|||
|
|||
|
a few things to help you out:
1)i thought soundblaster was 48k, not 44.1...you sure about that? 2) make sure your software is set to 24 bit recording. many software apps default to 16 bit. 3) if you're using cheap computer speakers, it can be hard to tell the differences, epecially if you have background computer fan noise clouding your listening 4) the audiophile doesn't sound that amazing. it's "ok". what it does have is excellent drivers. the thing just "works" with midi, ASIO, GSIF, etc. I think of it as kind of an ugly deep sea fish. It's not pretty, but it's tough and it gets you there. |
|
#19
|
|||
|
|||
|
"Arny Krueger" wrote in message ... That can be true, but the noise still builds up as you add processing. It's just that with 32 bit FP, it adds up at such a level that it is irrelevant. If you're going to pick nits, it should be easy to use a NR algorythm that will recover the fraction of a dB lost, without being too audible :-) However, I was speaking in the context of the general case, which includes 16 bit processing. If you're making the point that you can track at 16 bits and then use 32 bit processingm, that point is well-taken. Of course I was. You don't even have to choose with most programs. and no need for track bouncing with digital. I wasn't speaking of track bouncing. If you mix a lot of tracks, each one brings in its noise floor. Hopefully, each track's noise floor will be random and be attenuated during the mixing process. However, any discrete noise that is common to a number of tracks will add with simple arithmetic instead of adding geometrically like random signals. Things add up much faster with simple arithmetic. Not true, this occurs only if you don't gate unused tracks. Otherwise the real signals all add as well as the noise. In fact the signal will be more correlated than the noise in most cases. TonyP. |
| Reply |
| Thread Tools | |
| Display Modes | |
|
|
Similar Threads
|
||||
| Thread | Forum | |||
| Lyrics for We Live | Tech | |||
| M-Audio audiophile 2496 + Serial ATA + WinXP HELP!!!! | Pro Audio | |||
| new issue of AUDIOPHILE AUDITION | Marketplace | |||
| Dec. of brand new AUDIOPHILE AUDITION | Marketplace | |||
| Live Sound Vs Studio Sound | Pro Audio | |||
Linear Mode
