How much of a difference does the audio interface make when recording?

  • Thread starter Thread starter notoriousben15
  • Start date Start date
N

notoriousben15

New member
I have an M-Audio Fast track that i've played around with here and there. I just moved in with a guy who has an "Audiobox USB" audio interface, made by PreSonus. The PreSonus is definitely more expensive than my interface, but does that make it better? I only need to use 1 mic at a time, so having multiple outlets and stuff like that does not make a difference to me. Should i learn how to use his interface or stick to what i know?
 
It's never a bad thing to learn to use another piece of gear. I think on the software side of things, most of us stick with what we know and are used to because it works for us, but I'd have a hard time turning down a chance to use a mixing console, and even an audio interface.

A lot of the things you learn early on carry on to the next phase. So even if your audio interface only has one input it's not bad to branch out. Maybe eventually you'll get another interface with more ins and outs and you can hit the ground running instead of going through the initial learning curve.

As far as being better, more channels is not necessarily better. I'm just picking out random price brackets in the following examples. If you look at the different price brackets, everything under $100 will sound different, but not necessarily better. Anything from $100-300, same thing. Anything from $300-$600 will be about the same; again, different, but not necessarily better. When you start to get to about $1,000+, you'd hopefully have the ear and gear to be able to differentiate between the nuances in converters. If you're spending that much on gear, you better be able to tell the difference. Some people like a certain interface for one reason and others will hate it for the same reason. So it's not that people disagree on the sound of a device, but preference plays a huge role in gear.

I stuck with some old MOTU 2408 MKII boxes because I like the flexibility. I've used them with the PCI324/424 card as an audio interface but I've also used them as stand-alone converters with some RME cards. They also worked well through TDIF with a yamah 02r, and through the years they've done everything I needed. They're not the greatest, but they're fine for me. I've had a pair of these since 2003 and I've gone through a bunch of soundcards but in the end, I stayed with MOTU because it worked for me.
 
It's never a bad thing to learn to use another piece of gear. I think on the software side of things, most of us stick with what we know and are used to because it works for us, but I'd have a hard time turning down a chance to use a mixing console, and even an audio interface.

A lot of the things you learn early on carry on to the next phase. So even if your audio interface only has one input it's not bad to branch out. Maybe eventually you'll get another interface with more ins and outs and you can hit the ground running instead of going through the initial learning curve.

As far as being better, more channels is not necessarily better. I'm just picking out random price brackets in the following examples. If you look at the different price brackets, everything under $100 will sound different, but not necessarily better. Anything from $100-300, same thing. Anything from $300-$600 will be about the same; again, different, but not necessarily better. When you start to get to about $1,000+, you'd hopefully have the ear and gear to be able to differentiate between the nuances in converters. If you're spending that much on gear, you better be able to tell the difference. Some people like a certain interface for one reason and others will hate it for the same reason. So it's not that people disagree on the sound of a device, but preference plays a huge role in gear.

I stuck with some old MOTU 2408 MKII boxes because I like the flexibility. I've used them with the PCI324/424 card as an audio interface but I've also used them as stand-alone converters with some RME cards. They also worked well through TDIF with a yamah 02r, and through the years they've done everything I needed. They're not the greatest, but they're fine for me. I've had a pair of these since 2003 and I've gone through a bunch of soundcards but in the end, I stayed with MOTU because it worked for me.

What about obvious differences in signal-to-noise ratio? Or overall differences in fidelity between D/A and A/D converters? Or different transfer protocols like usb 2.0 (3.0 hasn't caught on with audio interfaces), firewire, or even motherboard and/or processor speed for pci-e cards? Some pci-e cards offer the best signal-to-noise around, at around 121 decibels, but even though certain cards are stereo, there aren't many ins and outs available. This may be what you are looking for. You must go high-end if pci-e is your choice. The ones with only s/pdif inputs and outputs are not good enough for audio creation. Also, if you have more than one available 1x slot, ASIO4ALL drivers offer the capacity to run two sound cards at once as long as they are integers of the same sample rate.
 
The difference won't be obvious as far as SNR is concerned. Even built-in soundcards don't have much noise. Not that I'd recommend using a built-in soundcard, but the problem is usually at the mixer/gain stage where a lot of the noise is introduced.

Differences between AD/DA converters are always there. Even the same converters in different devices will be different. Better? Not necessarily. That was my point.

USB 3 is irrelevant as far as quality is concerned. Processor speed and PCI-e is irrelevant as far as sound quality is concerned. The processor or slot does not affect sound. I've used both PCI and PCIx motu cards and they make no difference since it's all about the 2408 and its converters. If a device offers great SNR, it's not due to it using PCIe or PCI, or USB, or firewire. It's due to great design, not communication interface. There aren't many cards out there that use SPDIF only and they can be great if used with great converters. The fact that something uses or doesn't use SPDIF is not a determining factor as far as quality is concerned.

ASIO4ALL is unnecesary when using even low-end audio interfaces. They come with their own ASIO drivers and they perform better with their dedicated drivers.
 
To sum it up: when the signal has been converted to digital, there will be no changes in quality, whether it's USB, FireWire, PCI or something else. Digital is 1s and 0s, it's all the same. The things that affect the quality are the converters and the preamps (ok, things like clocking come into play when you're using multiple pieces of digital hardware, routing things out of the box etc. but this stands for basic recording).
 
It's never a bad thing to learn to use another piece of gear. I think on the software side of things, most of us stick with what we know and are used to because it works for us, but I'd have a hard time turning down a chance to use a mixing console, and even an audio interface.

A lot of the things you learn early on carry on to the next phase. So even if your audio interface only has one input it's not bad to branch out. Maybe eventually you'll get another interface with more ins and outs and you can hit the ground running instead of going through the initial learning curve.

As far as being better, more channels is not necessarily better. I'm just picking out random price brackets in the following examples. If you look at the different price brackets, everything under $100 will sound different, but not necessarily better. Anything from $100-300, same thing. Anything from $300-$600 will be about the same; again, different, but not necessarily better. When you start to get to about $1,000+, you'd hopefully have the ear and gear to be able to differentiate between the nuances in converters. If you're spending that much on gear, you better be able to tell the difference. Some people like a certain interface for one reason and others will hate it for the same reason. So it's not that people disagree on the sound of a device, but preference plays a huge role in gear.

I stuck with some old MOTU 2408 MKII boxes because I like the flexibility. I've used them with the PCI324/424 card as an audio interface but I've also used them as stand-alone converters with some RME cards. They also worked well through TDIF with a yamah 02r, and through the years they've done everything I needed. They're not the greatest, but they're fine for me. I've had a pair of these since 2003 and I've gone through a bunch of soundcards but in the end, I stayed with MOTU because it worked for me.

The difference won't be obvious as far as SNR is concerned. Even built-in soundcards don't have much noise. Not that I'd recommend using a built-in soundcard, but the problem is usually at the mixer/gain stage where a lot of the noise is introduced.

Differences between AD/DA converters are always there. Even the same converters in different devices will be different. Better? Not necessarily. That was my point.

USB 3 is irrelevant as far as quality is concerned. Processor speed and PCI-e is irrelevant as far as sound quality is concerned. The processor or slot does not affect sound. I've used both PCI and PCIx motu cards and they make no difference since it's all about the 2408 and its converters. If a device offers great SNR, it's not due to it using PCIe or PCI, or USB, or firewire. It's due to great design, not communication interface. There aren't many cards out there that use SPDIF only and they can be great if used with great converters. The fact that something uses or doesn't use SPDIF is not a determining factor as far as quality is concerned.

ASIO4ALL is unnecesary when using even low-end audio interfaces. They come with their own ASIO drivers and they perform better with their dedicated drivers.

I started off this post with arrogance and then realized I actually am not sure what device handles what when it comes to pci-e cards, cpu's, and motherboards, and decided I will learn from you. CPU doesn't affect sound? What about 24-bit, 32-bit sound? Or 96,000 sample rate? Doesn't processor power and motherboard speed come into effect when using more than one pci-e card?... each running 24-bit, 32-bit sound, at 96,000? I am using two Sound Blaster x-fi Titanium's (very high-end, at least in the expansion slot world), using ASIO4ALL to run them both at the same time, as the final stage before mixdown on my dj mixer. I guess what I am wondering about, is how much workload does the internal workings of a pci-e card take on... and more to the point, where does cpu begin and pci-e card end in this workload?

I also admit I am still very unclear on what is considered s/pdif lol. What I meant was, for example, a Sound Blaster that is meant for entertainment systems, that have no discernible stereo in's and out's... just a single jack for each surround sound channel. I know there is optical, non-optical, rca, coaxial etc., but to save time, can s/pdif be used for music creation? or just with video? My favorite discipline in recording is routing. I love routing in general... whether it's MIDI mapping or signal flow, and obviously, I need to know about s/pdif.
 
SPDIF is one of the many digital protocols available these days. It's nothing more than a digital stereo outlet. It's def nothing new and has been around since I've known somewhere around the early 90's. As far as pci & pci-e system bus is concerned, there's no corolation with it determining better performance vs. usb or firewire.

Someone mentioned SNR, I think they were trying to say that dynamic range is one of the governing factors that depicts the audio clarity of a sound card. Me personally, I'm still working with two pci cards. But because they work around the 115 dB dynamic range spectrum and haven't given me any problems, I see no need to upgrade and replace them.

Both are duplicates and both have spdif coax and optical as well as adat ports.
 
Last edited:
Sooo thanks for the help everyone, but if i'm being honest, i have no clue what any of these posts meant. Like, no clue at all. From what i can gather, the more expensive interface might make a small difference, but there really isnt much difference unless i'm using a much much more expensive interface...right? If someone could give me the "idiots guide to an audio interface," that would be perfect.
 
CPU doesn't affect sound? What about 24-bit, 32-bit sound? Or 96,000 sample rate?

No, it doesn't. It affects the computer's ability to handle & process the data. If it affected the sound, it'd mean that a different processor would spit out a different set of 0s and 1s, and that doesn't make any sense.

(...) but to save time, can s/pdif be used for music creation? or just with video?

Doesn't have anything to do with video, even though it's often used to carry the audio signal in, say, a home theatre system. But S/PDIF is strictly an (digital) audio connection. And yes, it can be used "for music creation" in the sense that some digital hardware - modern synths or effect units - have S/PDIF i/o, which just means that you get to avoid unnecessary conversions from digital to analog and vice versa.
 
I started off this post with arrogance and then realized I actually am not sure what device handles what when it comes to pci-e cards, cpu's, and motherboards, and decided I will learn from you. CPU doesn't affect sound? What about 24-bit, 32-bit sound? Or 96,000 sample rate? Doesn't processor power and motherboard speed come into effect when using more than one pci-e card?... each running 24-bit, 32-bit sound, at 96,000? I am using two Sound Blaster x-fi Titanium's (very high-end, at least in the expansion slot world), using ASIO4ALL to run them both at the same time, as the final stage before mixdown on my dj mixer. I guess what I am wondering about, is how much workload does the internal workings of a pci-e card take on... and more to the point, where does cpu begin and pci-e card end in this workload?

I also admit I am still very unclear on what is considered s/pdif lol. What I meant was, for example, a Sound Blaster that is meant for entertainment systems, that have no discernible stereo in's and out's... just a single jack for each surround sound channel. I know there is optical, non-optical, rca, coaxial etc., but to save time, can s/pdif be used for music creation? or just with video? My favorite discipline in recording is routing. I love routing in general... whether it's MIDI mapping or signal flow, and obviously, I need to know about s/pdif.

The only reason you need to run both cards at the same time is if you are using all of the physical outputs of your soundcards. If you're only running one set of outputs from the soundcard to your speakers, then you're not using both cards at all.

Your Soundblaster cards do have a processing chip but it is not used in the DAW. So if you're running a DAW, like Cubase/Sonar, or any other, the card doesn't do anything special. The X-fi processors are likely used for the basic stuff when convertins from digital to analog, and analog to digital, as well as hardware buffer, and other basic functions of all soundcards.

The bit depth (24bit/32bit) and sample rate (96,000hz) is determined by your DAW. My device supports 24 bit and 48khz, but I run it at 44.1khz in most projects. You do need more processing power to run higher bit depths and bit rates. Krushing already touched on the difference.
Sooo thanks for the help everyone, but if i'm being honest, i have no clue what any of these posts meant. Like, no clue at all. From what i can gather, the more expensive interface might make a small difference, but there really isnt much difference unless i'm using a much much more expensive interface...right? If someone could give me the "idiots guide to an audio interface," that would be perfect.

Simple answer, yes, more expensive soundcards will generally be better. If you can get the most out of it, then the difference will be huge. Most people can't tell the difference.
 
I started using 96,000khz last year and have been using 24-bit sound for longer and I can certainly tell the difference. And 115 decebels is horrible! I can notice a huge difference between a card I used years ago that was 115 decibels and the ones I'm using now. I have a PA system, so that must be why. I use my headphones when programming, and they are hooked up to a virtual reference monitoring system. Also, not many devices offer 121, 122 decibels SNR. Show me a hardware device under $400 that does. Sleepy, are you also gonna tell me you can't tell the difference between compressed and uncompressed audio files, too?

I must not be asking my questions properly. I knew all that stuff you guys said. Sleepy, you must be reproducing Oldies or something... that stuff was crazy that you said. And you must not have great gear.
 
I started using 96,000khz last year and have been using 24-bit sound for longer and I can certainly tell the difference. And 115 decebels is horrible! I can notice a huge difference between a card I used years ago that was 115 decibels and the ones I'm using now. I have a PA system, so that must be why. I use my headphones when programming, and they are hooked up to a virtual reference monitoring system. Also, not many devices offer 121, 122 decibels SNR. Show me a hardware device under $400 that does. Sleepy, are you also gonna tell me you can't tell the difference between compressed and uncompressed audio files, too?

I must not be asking my questions properly. I knew all that stuff you guys said. Sleepy, you must be reproducing Oldies or something... that stuff was crazy that you said. And you must not have great gear.

There's no single figure that describes the workings of a soundcard, regarding SNR. The inputs have their SNR figures. The preamps have their own figures. The outputs have different figures. And 115dB of dynamic range is excellent - take a look at the specs of the Prism Orpheus, which is one of the best interfaces money can buy. The theoretical real-world maximum is around 124dB; the X-Fi's dynamic range is about 109dB, which is just fine for a prosumer card.
 
. And 115 decebels is horrible! I can notice a huge difference between a card I used years ago that was 115 decibels and the ones I'm using now

Much more likely that the two interfaces had other differences that you heard .
 
Much more likely that the two interfaces had other differences that you heard .

Yes! Beside the fact that consumer brands don't take their measurements to seriously (they only measure the parts of it that look nice), the SNR has nearly nothing to do with "audio quality". Even high end analogue devices worth a small car have serious issues to reach better SNR values than ~90dB in practice. SNR is pretty worthless for audio. And that's why manufacturer like to use it!

The true audio quality relates to the precision of the internal clock, the quality of the converters sample and hold circuit, the anti-aliasing strategy and shape of its imposed bandwidth restriction, the analogue part inside the DA is very important too (the amp's linearity, distortion, slew rate, etc). None of these parts are explained by manufacturers.

A very similar "let's sh!t inside the consumers head" approach can be seen in the loud-speaker business.


Reality is, the specs of a carefully crafted high end device and on board sound-cards will virtually look equal. But sound quality wise, you may find out that there are worlds between both.
 
Last edited:
I started using 96,000khz last year and have been using 24-bit sound for longer and I can certainly tell the difference. And 115 decebels is horrible! I can notice a huge difference between a card I used years ago that was 115 decibels and the ones I'm using now. I have a PA system, so that must be why. I use my headphones when programming, and they are hooked up to a virtual reference monitoring system. Also, not many devices offer 121, 122 decibels SNR. Show me a hardware device under $400 that does. Sleepy, are you also gonna tell me you can't tell the difference between compressed and uncompressed audio files, too?

I must not be asking my questions properly. I knew all that stuff you guys said. Sleepy, you must be reproducing Oldies or something... that stuff was crazy that you said. And you must not have great gear.

You're paying too much attention to the marketing. Soundblaster cards, X-fi or not, are not great, they're okay at best. I don't know why you're so suck on SNR. You got caught up so much on the marketing, you bought two of them. By choice, I wouldn't use anything that made me use 1/8" connections.
 
I'm not an audio engineer but as an upcoming electrical engineer here is my view.

I believe Soundblasters focus more on D/A conversion for your average consumer's computer speakers. Which means they will have less than optimal A/D conversion and amplification for the audio input. Why? It's simple, when the engineers are designing a soundcard with a limited budget and knows that the majority of your consumer does not need high quality preamp and A/D converter for the mic input (that's because they will likely to use usb mic instead for gaming, chat, and other casual activities) you will know where to trim out.

Now, back to the topic of audio interfaces. Here's the breakdown of how an audio interface works: you have an analog signal from mic or an instruments going into a preamp and then it goes into A/D converter so it becomes 1's and 0's aka digital signal. Once the signal becomes digital (1/0) it's quality won't change through firewire, usb, pci-e, faster processor, motherboard, driver, etc.

In my opinion any modern audio interfaces are good for recording. They are capable of producing clean output (these boxes have reached the SNR threshold so your average ear can't notice! If you are hearing something the most probable cause will be bad grounding or noises from other devices picked up by your mic wires) However, not all audio interfaces will sound the same because there are many factors in the preamp and A/D conversion (probably preamp will effect more) which boxes sound better will come as a matter of taste. (Some boxes even uses boutique amplifiers just to take more of your money lol)

Now even though I said all these, I still didn't buy a typical $100-200 audio interface. I've chosen RME Babyface as my audio interface not for its high quality preamp and A/D converter but by its ability to expand and stable driver (never had a crash yet for 9 months and still going strong).

Lastly, your sound will come out vastly different from your recording techniques and mic placements than switching out gears. So stop looking for the perfect gear because it doesn't exist and make more music instead!
 
Last edited:
SNR is important on the first conversion of gain staging... whatever the noise level, it is amplified through each gain stage all the way until it reaches the monitors. So the SNR from the first source (my PC) is important.
 
SNR is important on the first conversion of gain staging... whatever the noise level, it is amplified through each gain stage all the way until it reaches the monitors. So the SNR from the first source (my PC) is important.

Well of course, SNR is important up to certain point, like I said before most modern audio interfaces now has enough SNR threshold to keep the signal clean enough for noise to be unheard. Having a better SNR isn't gonna mean much once you reach certain threshold, it's just an overkill, and a number that marketers like to use to hypnotize people who are unfamiliar with electronics.

This is also true for people who buys into 96k, 192k sample rate. If you look up and study the Nyquist-Shannon sampling theorem, you'll find that those high sample rate in reality doesn't make much difference at all. Your brain tricks you to believe that it is different if you know it beforehand! (You can confirm this in a A/B blind-test)

This is what happens when a majority of companies that sell these box with similar functions pulls out irrelevant numbers to lure in consumers.

Anyway, you can believe whatever you want. We need more people like you to buy into marketing gimmicks so those greedy companies grow fatter.
 
Well of course, SNR is important up to certain point, like I said before most modern audio interfaces now has enough SNR threshold to keep the signal clean enough for noise to be unheard. Having a better SNR isn't gonna mean much once you reach certain threshold, it's just an overkill, and a number that marketers like to use to hypnotize people who are unfamiliar with electronics.

This is also true for people who buys into 96k, 192k sample rate. If you look up and study the Nyquist-Shannon sampling theorem, you'll find that those high sample rate in reality doesn't make much difference at all. Your brain tricks you to believe that it is different if you know it beforehand! (You can confirm this in a A/B blind-test)

This is what happens when a majority of companies that sell these box with similar functions pulls out irrelevant numbers to lure in consumers.

Anyway, you can believe whatever you want. We need more people like you to buy into marketing gimmicks so those greedy companies grow fatter.

What I said was correct. That's all you needed to say.
 
Back
Top