OK, the saga continues. Bought a Presonus Inspire 1394 A-D → Firewire unit for digitizing my vinyl. Got it hooked up to my stereo and my MacBook Pro fine. Directed Audio MIDI setup to make this the default input and output device. Set the sample rate to 48kHz, 24 bit. Unit plays just fine, VU meters in the Inspire control panel look good, output through my speakers sounds good.
Launch Audacity 1.3.5. Tell it to use “Core Audio: Inspire 1394” for both recording and playback. Set quality to 48kHz, 24-bit sample rate. Back to project, confirm quality is set to 48kHz, 24-bit. Hit input monitor and there’s zero signal. Confirm that by recording a patch. Silence. Inspire control panel says there’s sound, and I can hear the sound in my speakers. Audacity says I got dead air.
OK, I figured out what has to be set wrong, by messing with Garageband and CuBase (the audio app included with the Inspire)-- the system defaults to audio input channels 1 and 2 (which are the mike inputs) instead of 3 and 4 (which are the phono/line inputs). Problem is that I can’t figure out how to configure that in Audacity-- how do I tell it which audio inputs I want to use from a particular device?
Garageband and CuBase have control settings for that, so I was able to make them work. Far as I could find, Audacity doesn’t, and this doesn’t seem to be settable in the MacOS Audio MIDI Setup or System Preference Sound control panels.
I’m not an Apple person - but I seem to remember from reading other postings that the problem may be with the 24-bit setting - IIRC Apple Macs seem to have a problem with this in Audacity (not sure whether this is a bug or a “feature”).
From looking at the specs and the manual fgor your box on the Presonus site I can’t see if you can reset the bitrate - it looks as though it only operates at 24-bit, so this may give you a problem.
Hopefully one of the Mac users will be able to comment on this in more detail - but meanwhile try Searching the MAC sections of the forum for “24 bit”, “24-bit” or “Presonus”
No it’s not got to do with the bit depth. If I plug an input into one of the mike jacks (audio inputs 1 and 2) on the Presonus, then Audacity sees the signal just fine. When I run audio into the Presonus via inputs 3 and 4 (the line/phono inputs I need to use to trascribe the vinyl), Audacity sees nothing.The two stereo channels in Audacity are looking at Presonus inputs 1 and 2, and I can’t figure out how to switch Audacity to look at 3 and 4.
Anybody know how to change Audacity’s settings for this?
pax / Ctein
PS-- yes, I know the Presonus is working correctly; I can use it with other apps.
If there isn’t any way of switching the inputs using the Presonus Inspire 1394 control panel, then try setting Audacity to record 4 channels - it may not work, or Audacity may record the line inputs on channels 3 and 4 (it depends on the drivers).
OK, setting Audacity Audio I/O preferences to 4 channel recording worked-- when I click the record button, it automatically creates four tracks and the phono audio comes thru on channels 3 and 4. I know how to delete channels 1 and 2 after I record, and mix 3 and 4 into a stereo track
Now, how do I get VU meters set up to monitor what’s happening on 3 and 4? I only see two VU recording meters and they seem fixed on channels 1 and 2. Is there some control for establishing more VU’s or associate the existing ones with specific channels?
One of the reasons I want to use Audacity for transcribing my vinyl is that the VU meters’ peaking bars make it really easy to get the input gain set right to insure no clipping.
Oh dear. Sorry I can’t help you with that one - I’ve not got a multi-channel sound card to even look at the issue. I’ll see if I can find out.
While I’m doing that, you could test if you can record in 32 bit (Edit ? Preferences > Quality). If you can record in 32 bit, that gives you a HUGE amount of headroom, so even if you record way below 0dB (and then Normalize), you will loose nothing in terms of SNR. (the dynamic range of 32 bit float will be considerably greater than the dynamic range of your audio equipment, allowing you leave yourself lots of headroom).
Thanks for the help so far. Seems terribly unlikely that Audacity would support multichannel recording with no way to monitor the signal strength.
Yeah, I could (and can) record 32-bit, but that just puts off the gain fiddlin’ nonsense to later on. I’d rather get it right from the get-go; it’s no more work to do trial recording than to fiddle late- actually, probably less. It’d just be better if I didn’t have to even bother. Feels kludgey to me.
Hi - Steve asked me, and I don’t know because I don’t have a multi-channel device, though it would not surprise me if this was actually true given our patchy multi-channel support. I’ll see if I can find out any more, and Steve could also PM Andy Latham as he records multi-channel I think.
Ctein, Is there any way to send the phono inputs as some kind of “master” or “mix” output, which might force them into channels 1 and 2? Have a look at the manual?
Another possibility is a standalone software VU meter. There are a number of these for Windows, but I can’t find any for Mac on a quick search.
No, unfortunately, the Inspire is not a mixer-- it’s designed solely to digitize four channels of audio and send them to software. It happens that Channels 1 and 2 are the Mike/Instrument inputs and 3 and 4 are the Line/Phono inputs. You can’t swap them around nor mix the digital signals.
A standalone VU meter would be usable, so long as it showed persistent peak level like Audacity’s. If it doesn’t, not much value.
The temporary directory should NOT be default-named “.audacity_temp” ; Audacity to fail with a directory-not-found error (on my MacIntel running 10.4.11). Starting a file name with dot on a Mac is considered bad form. Usually it doesn’t create problems, but in this case it does. Telling Audacity to name the directory “audacity_temp” instead fixes the problem.
Audacity is as I thought pretty much hard coded to meter two channels (meaning 1 and 2 if there are more than two). It would be something we’d want to improve later on, at least letting it meter two assigned channels out of a number of them.
Leland (our Mac developer) will reply to you shortly about the temp files directory on Mac.
Thanks for settling this. I’ll either grin and bear it or switch back to using CuBase (depends on which bothers me more, not having VU meters or not being able to save directly in flac format).
Look forward to hearing from Leland; glad to help out in any way I can. By the way, in case you couldn’t figure it out (unlikely), there were a couple of words accidentally deleted from my previous post. The sentence should have read “… this causes Audacity to fail with a directory-not-found error…” What can I say; I’m a writer, not a copy editor [ wry smile ].
~ pax Ctein
[ please excuse any word salad. MacSpeech in training! ]
The temporary directory should NOT be default-named “.audacity_temp” ; Audacity to fail with a directory-not-found error (on my MacIntel running 10.4.11). Starting a file name with dot on a Mac is considered bad form. Usually it doesn’t create problems, but in this case it does. Telling Audacity to name the directory “audacity_temp” instead fixes the problem.
I’ve just removed the dot and committed the change. Although I did not receive the errors you mentioned and don’t consider it bad form myself, there are reasons a user may want to use Finder and the like to access the Audacity temp folder. Having it hidden can most certainly be a pain.
In case this proves to be useful, let me describe my environment and the nature of the bug in more detail. This occurs on a Macbook Pro (Core Duo 2.0 GHz) running MacOS X 10.4.11 with 4 GB of RAM. My Audacity scratch directory is not on the system volume.
If the name of the scratch directory begins with a dot, then after 15-20 minutes of recording, Audacity starts to throw up a directory-not-found error. If I remove the dot, Audacity runs fine (well, at least out to an hour recording; I haven’t tried anything longer than that).
So, you may not have seen the error because it’s peculiar to my system, it doesn’t happen if you write the scratch directory to the system volume, or you didn’t run the recording long enough. Or door number four. Who knows? Not me!
It’s not uncommon for third-party software to name files and directories starting with a dot, but hard disk checking/repair utilities for the Mac will flag such things as not conforming to correct Mac practice. This may be an example of the reason why.
Anyway, kudos for the instant bug fix!
(By the way, FYI, “pax” is not a name but a sign-off, like “sincerely” or “yours truly”. It’s Latin.)
~ pax Ctein
[ please excuse any word salad. MacSpeech in training! ]
I recall now we did have a couple of instances last month of “unable to open temp files - no such file or directory” errors after recording for a while. They were both on MacBook Pros (Intel). So with hindsight that may have been the same problem.
This is probably born of ignorance of Mac, but when you say you need to write the Audacity temporary directory to a system volume, do you mean simply avoid writing to a hidden file with a “.”, or are you talking about having to write to a local drive rather than an external or network one (the latter being a bad idea for speed-of-access reasons anyway)?
No, that’s a misread-- it is definitely NOT necessary to write the temp directory to the system volume. In fact, I never do that. I was only noting as that since I hadn’t tested that case, it’s possible the dot-name problem doesn’t appear if one does so.
This is the first time I’ve run across a dot-name-related error. In fact, lots of Microsoft software creates dot stuff and it doesn’t break (no, I’m not holding them up as examples of good programming practices [s]). But in this case, it’s easily and quickly provable that the dot-name created a problem.
I agree that network storage might present performance problems. For similar reasons, I’d recommend against USB-2, because it’s CPU-moderated. If one’s not doing anything on the machine but transcribing audio, there’s plenty of horsepower to spare, but if audio’s being done in parallel with other tasks, it’s possible one could run into access-speed problems. I’m using external FireWire or eSATA drives.