24 bit question

[SouthSIDE Glen]

I agree with the quality distinction between the two but what I'm having trouble with is the idea that the more tracks you add to the soundblaster the more noticeable the noise. Have you heard noise difference between one track and 10 on a blaster?

I think what he's saying is the effect that we all know from all sources of noise; that when the noise is recorded, the noise form all recorded tracks creates a summed noise floor when the tracks are summed together. It doesn't matter at that point just what it's played through, nor whether it's a playback of all th tracks at once or of a summed mixdown; the noise is already there on the recordings.

G.

 

[Ethan Winer]

I heard a vocal track recorded through a soundcard and a consumer grade interface (like an M-Audio box) for comparison sake and could hear a difference. The soundcard track was more grainy or raspy. That was just one track.

Soundblaster isn't meant for anything serious.

I question this advice. First, unless you recorded the same vocal performance through two sound cards at once, you're comparing the performance differences rather than the converters. This is Rule #1 with subjective listening (not measuring) audio tests. You can't compare different performances! And you can't use two microphones either because tiny position differences make large changes in the frequency response. You must use one microphone and one preamp, then split that signal to the different converters.

As for SoundBlasters not being capable of pro results, I question that too. I'm not saying an SB card is as good as a high-end converter, but it's rarely the limiting factor in a recording's quality. Here are two files (1.5 MB each) of the same acoustic guitar performance, with the signal split to different converters after the microphone preamp. One was recorded through a $6,000 Apogee 8000, and the other through a $25 SoundBlaster X-Fi sound card:

Clip 1
Clip 2

Can you tell which is which?

--Ethan

 

[bouldersoundguy]

Here are two files (1.5 MB each) of the same acoustic guitar performance, with the signal split to different converters after the microphone preamp. One was recorded through a $6,000 Apogee 8000, and the other through a $25 SoundBlaster X-Fi sound card:

Clip 1
Clip 2

Can you tell which is which?

The first track sounds a bit more "present". The string squeaks seem more clear and real, while they sound slightly "plastic" or rolled off on the second track. There is a slight level difference between the two which can skew things since louder almost universally sounds better. The two tracks also have slightly different lengths, probably from the varying clocks of the two converters. If nothing else it demonstrates why you'd want to avoid using multiple converters simultaneously unless they're clocked together.

 

[noisewreck]

As for SoundBlasters not being capable of pro results, I question that too. I'm not saying an SB card is as good as a high-end converter, but it's rarely the limiting factor in a recording's quality. Here are two files (1.5 MB each) of the same acoustic guitar performance, with the signal split to different converters after the microphone preamp. One was recorded through a $6,000 Apogee 8000, and the other through a $25 SoundBlaster X-Fi sound card:

Clip 1
Clip 2

Can you tell which is which?

--Ethan

Ethan, admittedly I haven't listened to the examples, as the only thing I can listen to is the crappy laptop speakers I am typing this on at the moment. However, going from a personal experience, when I started in this noise, my first "real" soundcard was a Terratec 2 channel "mastering" thing. After I upgraded to my current Aardvark Pro Q-10, I heard a marked difference in sound quality. I am not talking about material that was recorded externally, but completely ITB stuff as well (i.e. sounds generated by VSTis).

Another concern is the quality of drivers themselves. Try having a project with a bunch of tracks with some serious processing going on, and you'll see a major difference in latency, CPU utilization, etc.

Soundblasters are great for gaming and watching to occasional DVD, but I stand by my comment: Soundblasters shouldn't be used for anything serious.

 

[Ethan Winer]

The first track sounds a bit more "present". The string squeaks seem more clear and real, while they sound slightly "plastic" or rolled off on the second track. There is a slight level difference between the two which can skew things since louder almost universally sounds better. The two tracks also have slightly different lengths, probably from the varying clocks of the two converters. If nothing else it demonstrates why you'd want to avoid using multiple converters simultaneously unless they're clocked together.

Right, I'd never use two non-sync'd sound cards for real. This was just an experiment to show that consumer grade sound cards are not as horrible as so many people claim. I don't normally use my SB card for recording or playing music. I have it only for editing SoundFonts which I do rarely these days.

--Ethan

 

[Ethan Winer]

Ethan, admittedly I haven't listened to the examples, as the only thing I can listen to is the crappy laptop speakers I am typing this on at the moment.

You really should find the time to listen, and tell me which file you think was recorded through which sound card. I can't speak for an old Terratec card, and cards from 10 to 15 years ago were noticeably inferior to what's sold to consumers today.

As for drivers and multiple tracks, the driver has nothing to do with how many tracks are in your DAW. Your DAW sums all the tracks, and adds effects etc, then sends a single stereo stream to the sound card. A driver is not hardware with resistors and capacitors and sampling circuits, so all it has to do is pass a data stream accurately, which is pretty simple for software.

When you're at a computer with a decent sound card, try this more elaborate sound card test:

Converter Comparison

This is a small multi-track project that compares three sound cards from very expensive through very cheap. Then email me your guesses from the link on my site. I'm still collecting replies, and when I have enough to be statistically significant I'll do an analysis.

--Ethan

 

[Chili]

Can you tell which is which?

--Ethan

No, they both sound exactly the same, can't tell them apart. But, it is only one track, maybe two. I still contend that after 24 tracks or so, the noise from conversion errors and poor components will accrue and be noticeable. Granted, I'm not going to go through the pain of trying to find out.

Ha, I even have a Phonic interface and Lavry Blacks, extreme opposites, and I'm still not going to perform any experiments to see if there is a difference. I'm sure it's very little; especially in my room, with my mics and my monitors.

 

[noisewreck]

As for drivers and multiple tracks, the driver has nothing to do with how many tracks are in your DAW.

The driver has a ton to do with the performance of the card/DAW/system.

It is a matter of crappy Windoze drivers VS ASIO.

 

[Ethan Winer]

I still contend that after 24 tracks or so, the noise from conversion errors and poor components will accrue and be noticeable. Granted, I'm not going to go through the pain of trying to find out.

I busted the "stacking myth" with a good explanation (IMO of course) in my AES Audio Myths video. That section starts at 28:28 into the video. The 5-track project I linked to above also disproves stacking to my satisfaction. If anything, stacking gives less distortion compared to summing then going through a single conversion. This is from a magazine article that's about to come out:

"There is some truth to the stacking myth, but it’s the opposite of what people think. Summing before versus after passing through a preamp or converter can be different if the device adds an audible amount of distortion. The key is if the distortion is actually audible ... Mixing tracks after they pass through a device that adds distortion avoids intermodulation (IM) products caused when the instruments combine. So if anything, recording many tracks through separate converters (or preamps) is potentially cleaner than when the tracks are mixed first and sent through one device."

There's much more detail in the article, and I can post the rest once the issue is out.

--Ethan

 

[Ethan Winer]

The driver has a ton to do with the performance of the card/DAW/system.

Please explain with more detail. Not "performance" as in latency, but how a driver affects sound quality at normal operation levels. (Not how a driver handles digital overs.)

--Ethan

 

[Farview]

Please explain with more detail. Not "performance" as in latency, but how a driver affects sound quality at normal operation levels. (Not how a driver handles digital overs.)

--Ethan

The driver doesn't affect sound quality, I don't think that is what he was saying. The driver does affect the stability of the system, latency, and the manner in which the interface can be used.

I'm sure we can all agree that the biggest difference between cheap recording interfaces and on board or gaming soundcards are the features that they offer.

 

[Ethan Winer]

The driver doesn't affect sound quality, I don't think that is what he was saying.

Okay, maybe I read too much into his post. But I still don't see why modern budget sound cards are necessarily incapable of professional results. One could even argue the opposite, that consumer sound cards are sold in larger quantities, by companies with much more at stake and with budgets large enough to hire expert hardware and software designers. I'm not saying that budget gear is better than high end stuff! But I think the common prejudice against budget gear is misguided. That's all I'm saying.

--Ethan

 

[Dogbreath]

I still contend that after 24 tracks or so, the noise from conversion errors and poor components will accrue and be noticeable.

Agreed.

Crappy components on a couple of tracks won't make as much difference as crappy components on 25 or 30 tracks.
It's accumulative.
Just like crappy mic pres on one or two tracks vs 25 or 30 tracks.
Accumulative.

 

[SouthSIDE Glen]

One could even argue the opposite, that consumer sound cards are sold in larger quantities, by companies with much more at stake and with budgets large enough to hire expert hardware and software designers.

That's an erroneous argument. First of all, as someone who spent many years working for and working with software and hardware development companies, I can tell you you'd probably be surprised to find out just how small these companies design and development operations usually are. We're not talking Google and Microsoft here.

Secondly, Soundblaster-class cards are not being designed and built for home recording use, regardless of how many people or what sized budgets may be behind them. Their creators have never meant them to compete with something like a small stand-alone Tascam or Presonus, so the size of the company or budget is irrelevant. That would be like saying that a Ford Fiesta could compete with a Formula 1 car built by a team of 20 people simply because Ford is such a large company with such a large budget. Even Ford will tell you that the Fiesta was never meant to run with the F1 crowd, even if it is technically possible for one to compete with them on the streets of the Toronto Gran Prix.

I'll grant you that some good things *could* be done on occasion with a well-disciplined Soundblaster. I've done it myself in a pinch, as I'm sure many folks here have done, including many who are recommending against them. In that you're right.

But occasionally being able to get away with using them if you need to is far different from *recommending* doing things that way. Even the folks at Creative Labs don't tout their products for use that way. And providing some Halton Arp-style half-science to argue against those that recommend against choosing them just to make a point isn't all that helpful.

G.

 

[NL5]

Clip 1
Clip 2

Can you tell which is which?

--Ethan

G1 sounds a lot more detailed.

 

[Ethan Winer]

That would be like saying that a Ford Fiesta could compete with a Formula 1 car built by a team of 20 people simply because Ford is such a large company with such a large budget. Even Ford will tell you that the Fiesta was never meant to run with the F1 crowd

Good point Glen, though I stand by my comment that modern consumer grade sound cards are never the reason a newbie's recordings suck.

--Ethan

 

[PDP]

with budgets large enough to hire expert hardware and software designers. --Ethan

LOL Your a dreamer like me, in reality its all about cutting expenses, I do agree that for a couple hundred bucks you can get some very good sound, the thing is, as the budget gear improves so does the high end gear, I'll always be chasing the best sound. ...........

 

[mshilarious]

I had a peek at a SB I have laying around, I/O is 4558s. There are a few discrete bits around them, could be anything so I won't venture a guess. Looks like all A/D is handled by a Creative CODEC.

I don't see a reason that the line input should not function adequately if used intelligently with an external preamp, but the problems are practical: latency, monitoring, limited I/O, etc.

 

[SouthSIDE Glen]

Good point Glen, though I stand by my comment that modern consumer grade sound cards are never the reason a newbie's recordings suck.

That I'll agree with 100%. But that's a whole 'nother thing from being so adamant about shooting down the idea that they're better off getting something other than a SB card. There are a dozen reasons, physical, logistical and sonic not to recommend a card meant for games and Skype playback, *including* the fact that they *will* sound different after stacking twenty tracks together in a standard mix.

Why should anybody ever recommend the worst potential solution? It's not like a decent EMU card or something similar - which will be MUCH better than a SB in most ways, including sonically - is in a whole 'nother price range or something. One can get an interface card to replace a SB for $50 - $90 that'll eat the SB for lunch. I've experienced that first-hand a few times.

G.