External audio interface - scam, or necessity? (intel mac)

[bouldersoundguy]

The file is 32 bit float, but I'm pretty sure that doesn't prove anything about your interface. I've recorded 24 bit files from a 16 bit interface.

 

[pazu]

I see your point bouldersoundguy. I am looking for ways to analyze this myself, but can't find anything for free for mac. Thanks for having taken a look. If I could look at this waveform in individual slices, it would probably be evident whether a 44,100 digital signal were fed to the 48K recording. Not sure what that might look at in analysis software, but curious... 65.5K of amplitude levels verses 16 million plus, there ought to also be big chunka chunkas of amplitude levels spanning the much higher resolution of 24 bit, revealing a lower res signal's having been recorded. Don't know!

Edit - In Adobe Audition, I was able to take the horizontal down to the sample level and scroll it - no jaggies which suggests to me that I am recording at 48K samples as expected.

In terms of amplitude, I think that would be much more obvious but it is also difficult to look at; I can't easily follow the waveform up and down along its path to see if there are shelves that might indicate 16 bit.

 

[bouldersoundguy]

I think you may be looking for the wrong symptoms.

I would think it's possible to determine the original bit depth by detecting the dither or by seeing if the LSBs filled with zeros, but that's a bit beyond my knowledge. Maybe this will help: Bitter - Stillwell Audio Stillwell Audio

 

[ecc83]

I include two shots of RMA of my 2496 and your mac. (I KNEW my card was better than before! Needed a shorting plug in it, suggest you do the same)

You can see that there are a lot of spikes along the noise floor (take no notice of the impossibly low N floor! Feature of the software. The rms noise is about -100dB and pk about -83dBFS) but that the mac card is much worse in this respect. It is the integrated sum of all this crap that it the noise as measured on a dBFS meter.

There is a bit of a learning curve with RM Anny! Once setup my MO is..
Put the signal on the desktop (must be a 16bit .wav file, fortunately Samplitude exports 16/44.1 unless I tell it nay)
Open RMA and click Analyser, find Desktop, find file. "Open". Varder and click the disc icon to save. N.B! If you want a printout flip the picture to "colour on white" else you get thru a shedload of toner PDQ!

The results with the app' must only be used comparatively and not to condemn in any absolute way various equipments. It is however very useful for tracking down hums and other "noises off" .

Re 24 bits or not. IF you think a file is 24bit .wav it will NOT, as I said, open in RMA.

More soonest.
Dave.

 

[ecc83]

The KA6 results show I think that 3 decades or so of converter development have not been in vain!

I do not know how much store to put in the results but the very low (-90dBFS pk) and very clean noise floor have to count for something?

Again, I don't know how much the 1kHz self generated tone is telling us but the absence of sidebands is very encouraging!

I also have a Levell TG200DM audio oscillator which is way "too good" for lining up tape machines but is of course crap compared to the distortion levels digital achieves!

As I said before, these excellent specifications are largely academic but it is nice to know I think that our hardware is SO much better than the signal we put in?

Dave.

 

[ecc83]

Ooops!

Missed one!

Dave.

 

[pazu]

ecc83, I am grateful to you for having generated these results. I don't however understand them. I will come back to this info tonight and try to make better sense of it. I am not sure if you are saying that my card is junk, or that it is an okay card relative to other stock internal cards, or if it is an okay card relative to external interfaces... Still working on comprehending the info! Thanks again Dave ! - jonny d

 

[ecc83]

ecc83, I am grateful to you for having generated these results. I don't however understand them. I will come back to this info tonight and try to make better sense of it. I am not sure if you are saying that my card is junk, or that it is an okay card relative to other stock internal cards, or if it is an okay card relative to external interfaces... Still working on comprehending the info! Thanks again Dave ! - jonny d

Don't sweat the understanding too much! The Samplitude shot shows that the KA6 is quieter than even my very acceptable M-Audio 2496 and WAY quieter than the clip of the mac you sent me.

The analyser noise shots show the NATURE of the noise. Note that in both internal cards, mine and yours, there is a definite hum component at 50, 100, 150 Hz and some harmonics. I would guess this comes from less than perfect power supply rejection in the cards (tho the 50Hz must be direct mains coupling) .

The KA6 on the other hand, even tho fed from a mains powered PC, obviously has another stage of power filtering because there is no hum to speak of at all.

All very nice of course but it bears repeating, hum at -80dBFS ain't really a worry!

The KA6 shot of 1kHz at neg1 shows that the interface generates almost no harmonic distortion else there would be spikes at 2, 3, 4, 5 .....etc kHz. If it would make things clearer for you I could inject 1kHz (say) from my oscillator and you could see the result. All you will get however is the vastly greater distortion of a mid priced (and very old!) bit of test kit!

As for your soundcard being rubbish? No, I don't think so, as others have said, if it works for you?? All I wish to show is that AT THE LIMIT built in sound cards are technically inferior to even a sub £200 USB interface.

Dave.

 

[Bobbsy]

Late to the party, but...

One aspect I haven't seen mentioned yet is that the performance of internal sound cards on Macs can be variable. I've seen some that seem to do a good job like the one above but, on the other hand, many's the time I've used the noise reduction in Audition to fix a -55ish noise floor on some other Macs. I haven't bothered investigating whether it the sound card itself of some kind of mismatch on the input but results weren't very nice on some machines I've received material from.

Beyond that, I have to say that external interfaces are NOT a scam. I have both USB and Firewire units and both work reliably and well, They also give me facilities I need (for example the FW one allows 32 channel recording and playback) while the simple 2 channel USB one allows the use of professional (i.e. XLR) mics and phantom power without an adaptor. The use of a mixer mitigates some of these issues but, as mjbphotos says, by purchasing a mixer and plugging it into your computer you've effectively built an interface anyway.

Only the OP can decide if the built in sound card does everything he needs...and if he's happy, then fine. But, for most people getting serious about recording, the built in sound card has too many drawbacks. You may have decided you don't need an external interface but that doesn't make them a scam. It just means your needs are different (and, in many cases, less) than most hobbyists.

 

[pazu]

Late to the party, but...

One aspect I haven't seen mentioned yet is that the performance of internal sound cards on Macs can be variable. I've seen some that seem to do a good job like the one above but, on the other hand, many's the time I've used the noise reduction in Audition to fix a -55ish noise floor on some other Macs. I haven't bothered investigating whether it the sound card itself of some kind of mismatch on the input but results weren't very nice on some machines I've received material from.

Beyond that, I have to say that external interfaces are NOT a scam. I have both USB and Firewire units and both work reliably and well, They also give me facilities I need (for example the FW one allows 32 channel recording and playback) while the simple 2 channel USB one allows the use of professional (i.e. XLR) mics and phantom power without an adaptor. The use of a mixer mitigates some of these issues but, as mjbphotos says, by purchasing a mixer and plugging it into your computer you've effectively built an interface anyway.

Only the OP can decide if the built in sound card does everything he needs...and if he's happy, then fine. But, for most people getting serious about recording, the built in sound card has too many drawbacks. You may have decided you don't need an external interface but that doesn't make them a scam. It just means your needs are different (and, in many cases, less) than most hobbyists.

Thanks. I have read that some MACs lose dynamic range when the volume is lowered, so they should be run at full volume to the amp but not sure of that, or of whether that situation would affect sound card performance. I run mine at 100% in the case of my DAW mac, and that goes into the same Berry PMP4000 that I record with. I did turn down the input level recently, however.

I have not decided that I do not need an external interface; I merely asked for evidence to suggest that my mac was notably under-performing its recording task, without one. This is an evaluative step in my re-thinking my recording process. It does not speak to my needs as a singer songwriter; those needs are relative to what I can learn about potential solutions and techniques. For instance, I am thinking of paying for studio time to see if I can get a better acoustic guitar sound out of my Ibanez, and if so how. I've tried many things but remain wholly unsatisfied.

In asking the question, I have learned of many factors worth considering, all of which point toward the solution of external interfaces. At this point actually, I am sold and conclude that I do need one, to help to take my audio recordings to the next level.

The issue of 16 bit internal sound cards in many computers as compared to 24 bit external interfaces, is huge, even though not relevant in my particular case (could be with my XP boxes). The phantom power, XLR/1/4" combo in's and preamps are not necessary either in my case. I can always use more computer processing power to run Adobe Audition CS6, and the external interface would grant me that. I would love to have MIDI built into my record interface and it seems that many external interfaces can do this as well. This could save me a USB port.

In the end I conclude as many have stated, that USB external interfaces are not a scam, now I know why, and that it is only happenstance that has resulted in my luckily not having been very adversely affected by my lack of one. Since I have been adversely affected in many other aspects of home recording, I can't feel great about that - my recordings still suck. Congrat's to those of you here whose recordings, do not!

A lot of people suggest that, 'if it works for you then you are all set.'

It doesn't work for me. My recordings are garbage thus far. I know now that in my unique circumstances, acquiring an external audio interface is probably not going to make the difference. I'll be saving my pennies for one, now that I understand their advantages, and in the meantime continue browsing, looking for a path to decent acoustic guitar recordings.

Thanks to all for the help with understanding this! Good job! Some was over my head, but now I basically get it.

Jonny D

 

[ecc83]

Morning Jonny, you might find this interesting?
Recording Acoustic Guitar
There is another very good article on recording electric guitar in the Aug 07 issue.

With regard to internal levels, my 2496 controls are all set at max. There are two reasons for this..
1) The card defaults to max upon driver install and who am I to argue with M-A?
2) If left at some intermediate value I would be forever checking them!

However, at the risk of "telling Granny....." the levels you RECORD at in the DAW software should be nowhere near "max". For a truly 24bit system, the received wisdom is that average levels should be at -18dBFS. I prefer to run at -20dBFS for no other reason that 2dB don't matter a hoot and neg 20 is marked on most, certainly Samplitude's, meters and -18 ain't!

For a 16bit setup you would in theory come up a few dB on those levels (and risk clipping) but in practice, for a GOOD 16bit system you could stay at -18/-16 or so. In any case, as mentioned before, any source you have is going to be noisier (room if nothing else) than even a poor 16bit digital recorder.

Dave.

 

[pazu]

Morning Jonny, you might find this interesting?
Recording Acoustic Guitar
There is another very good article on recording electric guitar in the Aug 07 issue.

With regard to internal levels, my 2496 controls are all set at max. There are two reasons for this..
1) The card defaults to max upon driver install and who am I to argue with M-A?
2) If left at some intermediate value I would be forever checking them!

However, at the risk of "telling Granny....." the levels you RECORD at in the DAW software should be nowhere near "max". For a truly 24bit system, the received wisdom is that average levels should be at -18dBFS. I prefer to run at -20dBFS for no other reason that 2dB don't matter a hoot and neg 20 is marked on most, certainly Samplitude's, meters and -18 ain't!

For a 16bit setup you would in theory come up a few dB on those levels (and risk clipping) but in practice, for a GOOD 16bit system you could stay at -18/-16 or so. In any case, as mentioned before, any source you have is going to be noisier (room if nothing else) than even a poor 16bit digital recorder.

Dave.

Good Morning Dave, and thank you for including the article. I've begun reading it and will continue prior to my recording this afternoon. I am familiar with a lot of the basic advise, the 12th fret placement and guitar selection in the first place being so important, and properly playing it.

I may have shot myself in the foot, by going with elixers but, due to the texture of my fingers (I guess) I get a lot of string slide noise if I use regular strings. It's a crutch, theory being, enough crutches & I'll be walking.

What I've been doing over the past few days of record attempts, is monitoring the output with headphones as I play and move around the guitar near the MXL 990, to try to find the better sound. It doesn't seem to be there, however. Voice sounds reproduce great, but I just can't capture the guitar sound. I tried another room, actually it was my pentagon 'bobhouse' or ice shanty. It was no good. I can try putting foam on the MXL 990 this afternoon, and also try opening all of the windows in the small room that I use for recording. Anyhow, I look forward to finishing the article and maybe pulling more info from there. I like sound on sound and am on that board as pazu as well, but the focus may be too advanced on that board for me. I emailed some files to their mix clinic a month ago but received no reply. Others have experienced that as well. Doesn't do much for the ego... and perhaps there is a point to that.

I believe that I am recording in 24 bit not 16 at this point, though I can't prove that my input signals are actually 24 bit as they are recorded. It seems fairly obvious that they are as they can't really 'not be.' The DR-05 recordings are definitely 24bit 96K.

My board, the way that I have it configured right now, may be outputting a little hot. Not distorted, but a hotter signal so that's why I decided to turn down the record in of the mac. I haven't recorded much with that setting yet. I have an M audio card but don't use it as it won't fit in my modern macs.

Interesting I think, that the guitar sounds just as muddy when recorded with the DR-05 at max res, as it does when recording it via the mac. Perhaps that is just the sound that the guitar makes, though that is not the sound that I think that I am hearing when playing it live. I have raging tinitis, maybe that has something to do with it.

The single best thing that I have been able to do with guitar tone thus far, is attach two corks to the back of my instrument which stand it off from my body. It resonates more then, and achieves a better tone! But, that is sort of crazy to do.

I have a fender 6 string and a fender 12 string as well as my primary instrument, which is the crappy sounding ibanez. So I'll try mike-ing them as well this afternoon.

Since I am left handed, it is difficult for me to even find other instruments to play. It is very limiting, guitar shops don't carry lefty guitars, not around here, I do better in pawn shops. If they do have a lefty it's a $250 lefty.

My record levels are generally very low to stay well away from clipping, I'll up them a little to stay in the -20 range. Looking forward to finishing that article, thanks again! - Jonny

 

[DM60]

On acoustic, the general guideline is a two mic setup. One around the bridge and the other on the 12th fret. The bridge mic not pointing to the sound hole. For the bridge setting, I have used, straight towards the bridge, from the back of the guitar and over the should pointing to the bridge. I have tried with dynamics and condensers.

My favorite would be, condenser on the 12th fret, dynamic on the bridge.

If you try a couple of different approaches, you will find the sound you are looking for.

 

[pazu]

On acoustic, the general guideline is a two mic setup. One around the bridge and the other on the 12th fret. The bridge mic not pointing to the sound hole. For the bridge setting, I have used, straight towards the bridge, from the back of the guitar and over the should pointing to the bridge. I have tried with dynamics and condensers.

My favorite would be, condenser on the 12th fret, dynamic on the bridge.

If you try a couple of different approaches, you will find the sound you are looking for.

In the myriad of approaches that I have tried, I have not tried this approach. I have an akg-d320b and an MXL 990 and will try them in this configuration. Reading down through the detailed S O S article, I find that this is sort of referenced, at least the mic on the bridge idea. Good idea and thanks particularly for the thought that I will find the sound eventually! - Jonny D

edit - thinking about it, this is pretty interesting your thoughts as to bridge placement behind or over the shoulder... I'm hearing the sound from above the instrument, but recording it from a place that I can never hear this guitar from. I know no other lefty guitarists so nobody else plays my instrument. Open mic guys say that it sounds great, but maybe they are just trying to make me feel comfortable (or themselves comfortable). Definitely trying all of the new ideas this afternoon!

 

[DM60]

For the most part, most of us close mic it because our rooms aren't that good. If you have a good room, then you can use something further out. Using 12th and bridge provides close micing advantages, but helps provide a more balanced final. I suggest two tracks, not stereo so you can get your sound balanced with the faders before you even EQ.

I usually use faders for the two tracks, then group or send to a single EQ to go to the master and treat it as a single instrument.

 

[pazu]

For the most part, most of us close mic it because our rooms aren't that good. If you have a good room, then you can use something further out. Using 12th and bridge provides close micing advantages, but helps provide a more balanced final. I suggest two tracks, not stereo so you can get your sound balanced with the faders before you even EQ.

I usually use faders for the two tracks, then group or send to a single EQ to go to the master and treat it as a single instrument.

I think that I follow you. I use a 16 channel board for input with 8 XLR in's so I can easily bring the two mikes in and try to get them right before I go to the DAW. It is actually the only way I can use this board because I use monitor sends to send the mix from all of the inputs to either one or the other, and I can only record two tracks at a time, and one has to be vocal because if I don't sing and play, I tend to get lost in time. Plus singing and playing, trying to do both well, is a lot of the fun of hobby audio for me. Great clarification, thanks DM60! - Jonny D

 

[DM60]

I think that I follow you. I use a 16 channel board for input with 8 XLR in's so I can easily bring the two mikes in and try to get them right before I go to the DAW. It is actually the only way I can use this board because I use monitor sends to send the mix from all of the inputs to either one or the other, and I can only record two tracks at a time, and one has to be vocal because if I don't sing and play, I tend to get lost in time. Plus singing and playing, trying to do both well, is a lot of the fun of hobby audio for me. Great clarification, thanks DM60! - Jonny D

OK, food for thought. I use a mixer as well, but mainly for the connection. What I do is on the mixer, pan one mic left, the other mic right and arm two tracks in the DAW. I set each track to record mono. You should see a channel 1 (left) and channel 2 (right) for your inputs. If not, that is a good case of a simple two channel interface. You get that type of flexibility.

 

[jonny deep]

You might want to be recording a stereo mix of your two mics, although that is of course not necessary. If you're having trouble knowing where you are in the song, you can lay down a scratch rhythm & vocal to guide you or use markers on the DAW to help.

 

[mjbphotos]

What don't you like about the acoustic tone you are recording? Are you using Elixer phosphor bronze (8/92) or 20/80's? I didn't like the Elixer sound when I tried them on my Taylor (factory strings, and one replacement set), and I've switched to Martin Lifespan 20/80's - also coated but I don't find them as 'dead when new'.
The problem may indeed be your guitars themselves, though I could get a decent recorded sound from my Fender CD140sce by using one mic 9-12" from the 12th fret, slightly pointed towards the soundhole, and a second mic on the lower bout at the same distance. Before I had any acoustic treatment, I would set up in my largest room and point towards the farthest walls, at a slight angle to try to avoid any reflection pickup, but now I just face a corner/wall of bass traps.

 

[gecko zzed]

Here is a video of a typical two mike on guitar set up. Rode NT1 pointing at bridge, and a very hard to see AT853 pointing to 12th fret. Audio is recorded via Firepod into Reaper and replaces the video audio.


Doodling on the guitar