| Author |
 Topic  |
|
Mark E
Lokahi
USA
186 Posts |
 Posted - 07/27/2007 : 05:33:13 AM
|
Hi and Aloha!
Question: Is there some way to use a Zoom H4 and/or Audacity to be able to record one track while listening to another? The purpose here is to add tracks with harmonies, backup, etc. to my original recording without recording them separately and then trying to line them up, mess with the tempo and so on.
Thanks for any advice.
Mark E
|
|
|
cpatch
Ahonui
USA
2187 Posts |
Posted - 07/27/2007 : 06:43:25 AM
|
| You can do it with the H4 using 4-track mode...see pp. 21-26 in the manual for a quick guide. |
Craig
My goal is to be able to play as well as people think I can. |
Edited by - cpatch on 07/27/2007 06:48:29 AM |
 |
|
|
hapakid
Luna Ho`omalu
USA
1533 Posts |
Posted - 07/27/2007 : 11:12:15 AM
|
I've tried using Audacity (open source, free sound editor) for multitracking. It is supposed to work but I've always had latency problems.
Jesse Tinsley |
|
|
|
Mark E
Lokahi
USA
186 Posts |
Posted - 07/29/2007 : 6:13:27 PM
|
Thanks, Craig -
I've tried reading that but I don't see how to get it to play something while recording something else. Doubtless, I'm being dumb but if you don't mind explaining more, it would be a big help.
Jesse - If "latency" means trouble lining things up, I've sure run into that. I even tried playing one part on Audacity through a headset while recording the accompaniment on the H4 and then adding the accompaniment as another track in the Audacity project. Perhaps it was my wobbly timing, but the tracks matched in some places and not in others. (Granted, I was playing Keola's part of Tiare Tahiti against Mark O's recording and Keola used all kinds of syncopation.)
I remember Little Brother's tutorial in which he added many tracks separately - using some different equipment, I suppose?
Again, Mahalo for all help,
Mark E |
|
|
|
Absolute
Lokahi
275 Posts |
Posted - 08/02/2007 : 2:26:03 PM
|
It's probably not Audacity that is causing the problem with multi-track alignment. I've run into it myself. You need to go into control panel in Windows. Go to Multimedia, then to the Audio tab, then Recording and Advanced Properties. Push the Hardware Acceleration slider under Advanced Properties to "Full" (all the way to the right), and the Sample Rate Conversion slider to "Good" (all the way to the left).
You can experiment to see if this helps. Just start an Audacity recording, and after a few seconds hum some tone for a second or two, then stop the recording. Don't erase it; just mute it. Hit the record button again to start a second track, and when you get to the start of the blip for your first recording, hum something again just at the leading edge of the blip, then stop the second recording. You should now have two recordings on your screen. They should start pretty close to the same point (almost exactly if you timed the second tone right and you're computer has been optimized for multi-track recording and is fast enough to do it). If you don't watch the conversion quality slider in Control Panel under Multi-media Audo Recording Advanced Properties and make the error of pushing it all the way to the right for best quality conversion, you'll get nice quality samples at a very slow conversion rate, which makes synch-ing impossible on a computer that isn't fast enough to perform fast A/D conversions (are there any? - I don't know), particulary if your computer uses the CPU for everything with a motherboard that incorporates the sound and video. This is one reason why people go for faster computers and separate sound and video cards that have their own memory and CPU. It also lets them upgrade as audio and video progresses, particularly if they don't want to spend a lot on a high end video or audio card at the start. Of course, even all-in-one boards typically have jumpers that will disable the on-board audio or video and let you plug an upgrade card into an expansion slot. CONSULT AN EXPERT BEFORE PLAYING AROUND WITH COMPUTERS, UNLESS YOU'RE A FORMER MEMBER OF THE GEEK SQUAD!
To listen to a track while recording another, go to Edit, Preferences, and Audio I/O tab in Audacity. Check the "Play Other Track While Recording New Ones" box. Plug headphones or ear buds into your speakers or into your sound card, whichever mutes the speaker output, if you don't want to record the playback along with the second instrument track.
If you use a high quality conversion (many bit A/D conversion) on a slow computer, the samples come so slowly due to the time required for each conversion that you get samples at a much slower rate than is required for normal playback, so when you play it back at the normal rate for samples, it can make you sound like a member of the "lolly pop guild", just as though you took a normal rate recording and played it much faster than intended. |
Thank you. |
Edited by - Absolute on 08/03/2007 03:25:23 AM |
|
|
|
hwnmusiclives
`Olu`olu
USA
580 Posts |
Posted - 08/03/2007 : 03:32:13 AM
|
quote:
Originally posted by Absolute
It's probably not Audacity that is causing the problem with multi-track alignment. I've run into it myself. You need to go into control panel in Windows. Go to Multimedia, then to the Audio tab, then Recording and Advanced Properties. Push the Hardware Acceleration slider under Advanced Properties to "Full" (all the way to the right), and the Sample Rate Conversion slider to "Good" (all the way to the left).
Absolute, you're hitting on territory very near and dear to my heart here because I haven't made a multitrack recording since last November when I first experienced the latency issue with a newly-purchased state of the art PC. (Oh, why didn't I go Mac this time?)
I upgraded from a Pentium III machine to an AMD Athlon dual-core to solve some of my other problems (like running mixing and mastering software at the same time). And in the process, I discovered latency - an issue I NEVER had with the old machine. I have never seen any solutions to latency issues outside of tinkering with sampling rates on the multitrack software itself (in this case, "Guitar Tracks Pro III"). I had never heard of the "Hardware Acceleration" slider. I am going to go home and play around with this and let you know what happens...
Thanks for an invaluable tip! |
Join me for the history of Hawaiian music and its musicians at Ho`olohe Hou at www.hoolohehou.org. |
|
|
|
cpatch
Ahonui
USA
2187 Posts |
Posted - 08/03/2007 : 06:56:00 AM
|
Hey Mark, I haven't had a chance to sit down with the H4 recently to be able to give you specific instructions. The pages in the manual that I referenced are a quick overview; there is a section later on that goes into more detail.
As for latency, it's not just a Windows issue. It really depends on the combination of software and hardware interface you're using and is often introduced at the interface level. |
Craig
My goal is to be able to play as well as people think I can. |
|
|
|
Lawrence
Ha`aha`a
USA
1597 Posts |
Posted - 08/03/2007 : 08:08:39 AM
|
quote:
I've tried using Audacity (open source, free sound editor) for multitracking. It is supposed to work but I've always had latency problems.
Jesse Tinsley
I have not tried to use Audacity for layered tracking, so I do not know if it adds much latency with the default settings.
There is always some latency added in computer recording systems. (both Mac and PC). Unfortunately Computers are designed to handle information in "chunks" and latency is an inevitable result of this design. The "chunks" come in various differing sizes depending upon the device and application. Buffers are established to hold these chunks while the computer is doing other things (and it is always doing many other things). The digitizing hardware (your sound card or sound interface ) also has buffers of its own and passes the these chunks into the computer at regular intervals.
The latency is directly related to the SIZE of these buffers (often expressed in number of samples for audio interfaces), the bigger the buffer is the more stable the software is BUT the longer the latency.
Most audio Apps have a setup dialog panel somewhere where you can adjust the buffer size. Audacity probably has one somewhere. Part of the setup you should do for any Multi track Audio App is to adjust this buffer size to the SMALLEST value that you can get away with and still get glitch-free sound. This number will vary greatly depending upon the actual hardware that you are using, but can usually be adjusted down to just a few milliseconds. When the software is installed this buffer size number is set very large on purpose in order to guarantee that the software will work on every system, but needs to be adjusted downward by each user.
In addition, most multi-track apps have latency compensation. There will still be latency when you record new tracks (relative to existing tracks). The application will move the new track(s) you record forward by the amount of this latency setting, thereby aligning the tracks correctly (if the compensation number is correct). This number also has to be set by each user because of the hardware dependencies.
You can measure and adjust this yourself: 1) turn off the latency compensation 2) record a track of clicks (impulses). 3) Play that track back and connect it's OUTPUT into the next track INPUT. 4) Playback the first track while recording the second. 5) Use the cursor to measure the time offset from the first track to the second. This is your multi-track latency. Turn the latency compensation back on and set it to this number. If you have a mixer in the loop (like I do) this number will also reflect any delays going thru the mixer (which are usually very small).
P.S. I recommend that you download Reaper instead of Audacity. The demo is fully functional (and will remain fully functional with a "Nag Screen"). It beats the pants off Audacity and if you like it you can register for $40 to remove the nag screen.
http://reaper.fm/download.php
|
Mahope Kākou...
...El Lorenzo de Ondas Sonoras |
Edited by - Lawrence on 08/03/2007 09:23:01 AM |
|
|
|
Absolute
Lokahi
275 Posts |
Posted - 08/03/2007 : 1:39:30 PM
|
Dear Mr. Wynne:
I'd been experiencing the song of the "lolly pop guild" when I tried to record my voice for some time. My relatively slow tin whistle playing was transformed into hot jigs by the settings that I was using due to problems with the A/D conversion setting I was using with Windows ("sample rate conversion" factor set too high on the audio tab). I note that it was not just increasing the "hardware acceleration", but also reducing the "sample rate conversion" setting to the lowest value ("good") that has restored my capability to multi-track with Audacity (a capacity that was probably eliminated when I decided to "optimize" my recording settings without actually experimenting to make certain my hardware could handle many, lengthy and progressive A/D conversions as the price for potentially pointless levels of accuracy in representating a signal with many sampled points and a high bit resolution).
I don't know about your A/D converter chip, but if it operates using successive approximations, and you assert a need for many samples and many bits per sample to represent a sampled analog value, the A/D converter chip would go through a series of steps associated with progressively comparing the digital value of the signal with the analog value for each sample by converting the digital approximation to an analog signal and sending it a comparator circuit that could take quite some time. If the digital representation (value in Hex) is low relative to the actual, analog value at any point in this A/D cycle (for a single audio sample) as determined by the comparator circuit, it will continue to step the digital value up until the comparator indicates that the latest shift in the LSB (least significant bit) has pushed the digital value higher than the analog value. How that is handled is up to the party who developed the conversion algorithm for the chip (it either uses the last value or the one before it, thus either slighly over- or under-estimating the digital representation of the analog signal).
At any rate, I'm sure you see that using many samples each represented by many bits to represent an analog sample's value causes a more lengthy time delay than cruder representations with fewer bits. If the sample rate is higher than the A/D algorithm can correctly support for the specified sample rate and bit accuracy level per sample (on a given computer system), the system may start skipping samples due to delays in conversion. (IF the A/D conversion takes too long, the computer can't do anything else other than interpolate between points to try to restore the desired number of samples, and then only if the software using the sampled data is set up to detect dropped samples and hold a space for interpolated data points.) If you then tell the computer to play back expecting the number of samples that you said you wanted it to provide per second when it recorded the signal, but which it couldn't acquire due to a "slow" A/D conversion process, you effectively speed up the rate of playback. (Two data points that are supposed to be separated by an intermediate point are now played sequentially.)
The need for high bit depth representations is questionable for home recording (and other forms of recording as well under most circumstances). You can set Audacity for 16 bit audio resolution, and still not have to worry about slow conversion affecting multi-tracking, so long as you set the conversion to "good" (all the way to the left) on the audio tab (at least with my system - others may differ) and the "hardware acceleration" to maximum (all the way to the right).
I don't argue that there is a time delay to transmit data from memory to the CPU, and from the CPU back to memory and to temporary files on the hard drive, so a faster system is useful for home recording, and particularly for multi-tracking, where you need to have a system, including the software set-up, that will permit alignment of a recorded and a live track in real time. (Think of this: You play back one track from the hard drive's data, and record another from your soundcard when multi-tracking. If your data pathways are slow, it can prove difficult to align the two tracks in real time. If you force too much data through the usual pathways due to high sample rates and bit resolutions, that may also slow things down, causing misalignment. Sample rates and bit resolutions have an effect on the amount of data being transmitted per unit time from two different locations.) I'm using Windows 98SE, by the way, and an older motherboard and CPU, so you don't have to have the latest and greatest to make this work.
I'm still using 44 kHz sample rates in Audacity with 16 bit resolution. Note that Nyquist's sampling theorem asserts that one should sample at twice the highest frequency of interest to avoid aliasing problems. If the human ear is capable of up to 20,000 Hz, 44,000 Hz should satisfy any reasonable sampling need for audio, particularly if you do an FFT (Fast Fourier Transform) of the audio in Audacity to produce a spectrum plot, and discover that the most energy in the spectrum of the sound occurs in a frequency range that is well under 10,000 Hz.
Choose settings that work for the recording options under Control Panel's multimedia icon that work for your system, not options that satisfy your personal computer vanity, by CONFIRMING that the settings you have selected work with the software you are using.
(Note: This discussion of data pathways in the CPU does not explain the "munchkin music" effects I experienced on my computer before I altered the Control Panel Audio settings. Those effects resulted from pseudo-acceleration of the data representing the music even with only one track that was already recorded, under the assumption that slow cache doesn't drop data points during recording, it merely slows down their transmission to the hard drive. I didn't have this problem with Windows Voice Recorder with the same multimedia settings, so there must be some capacity for misalignment between Audacity's settings and those presumed to exist by Windows.)
I'm using an AMD Duron processor running just over a MegaHertz and an "all-in-one" PCI motherboard. You should experiment with care with Audacity to establish how the "advanced properties" menu (accessed via a button) under audio recording in Windows's multimedia option under control panel will affect your system. It was only due to my sense of wishing to push my system to its limits (that PC "vanity" issue) that caused me to push my computer too far and not immediately confirm with Audacity that what I had done under Control Panel was viable for my needs. (Up until I learned of these two recording settings in Windows and made the proper adjustments I was suspicious of some mystery virus that my virus program simply could not detect that was designed to target audio.)
P.S. If you don't have a virus scanning program, you should probably try to get one just to be sure your computer memory isn't being allocated to something meant to slow your computer down. I'd suggest downloading the free trial version of Panda Software's anti-virus program and running a full system scan just for peace of mind. (I don't work for them, I just found their last version was one of the few that would still run under Windows 98SE.) |
Thank you. |
Edited by - Absolute on 08/04/2007 03:50:50 AM |
|
|
|
Lawrence
Ha`aha`a
USA
1597 Posts |
Posted - 08/03/2007 : 2:01:59 PM
|
quote:
You can set Audacity for 16 bit audio resolution, and still not have to worry about slow conversion affecting multi-tracking, so long as you set the conversion to "good" (all the way to the left) on the audio tab.
If the ADC design supports a given bit depth and sample rate, that means the successive conversion process is complete before each sample transferred out of the ADC. Most modern converters will easily support 24 bits at 44.1KHz or even 96Khz. If you exeed the sample rate that the converter supports you just get bad data (garbage noise which is quite obvious and loud - full scale) since the sample rate is FIXED. Additionally almost all modern converters use an oversampling method where they sample at a very high rate (like 300Khz to 1.5MHz) with a low-bit depth flash converter (4 to 8 bits typical) and then use decimation to construct a higher-resolution lower rate data stream. In any case the "sound card" hardware designers do not allow the user to change bit depth at the converter stage in the circuit.
Changing the Audacity bit depth will not change the converter behavior at all (it will still do 24bits or whatever the hardware is designed for). Setting Audacity to 16 bits WILL save time in the writing to disc since there will be less data written per unit time. Changing the sample rate is another matter. Most designs actually change the sample rate at the converter, however, some designs (such as the infamous SoundBlasters) run the converters at a fixed rate and use rate-conversion hardware to convert to other rates like 44.1 or 96. I recommend avoiding all SoundBlaster products for this reason!
|
Mahope Kākou...
...El Lorenzo de Ondas Sonoras |
Edited by - Lawrence on 08/03/2007 2:32:27 PM |
|
|
|
Absolute
Lokahi
275 Posts |
Posted - 08/04/2007 : 04:05:12 AM
|
| Thanks. My background is power, not electronics. The details of how the boards are designed in reality (not theory) isn't my specialty. If you are right, and it sounds like you know this architecture well, then Window's "Sample Rate Conversion" setting sounds like a key factor. If Audacity is assuming that the card is operating at some other speed, that may be the basis for the problem I encountered with acceleration of audio recordings due to some mismatch causing either dropped data points or playback at a rate based upon assumption of a higher data acquisition speed. The hardware acceleration option seems also likely to impact how memory is being accessed to provide for greater speed. |
Thank you. |
Edited by - Absolute on 08/04/2007 04:14:38 AM |
|
|
|
Lawrence
Ha`aha`a
USA
1597 Posts |
Posted - 08/04/2007 : 07:28:50 AM
|
Sounds like a sample rate mismatch to me (the chimpmunk effect that is). Somewhere in the system one function is "thinking" that it is at a different sample rate than another. Could be the converter driver(s) or Windows or an interaction between Audacity and Windows.
I fired up my old (six months or more) copy of Audacity to check the settings on my work computer (with onboard "SoundMax" chip set). Looks like Audacity does not give many "low level" adjustments of much use to the user (like buffer size). Better tools, such as Audition and Reaper allow much more fine-tuning in this respect (which of course could cause some trouble if set incorrectly, usually won't crash the system, just make awful noises).
I reccommend switching to a better tool such as Reaper. Reaper is small (2.5Mb install file) so it downloads quickly. There is a free 6.8Mb HTML users guide you can download too. In the preferences panel you will see much better (more extensive) hardware setup options, including the buffer size option. And Reaper is FAST, it runs my copy of the Ozone mastering tool (a DX plugin) about SIX times faster than Audition does). Some apps (like maybe Audacity) might run very large buffers, perhaps on the order of 22000 samples (and they don't seem to report this in the preferences settings), which would result in half a second guaranteed latency at 44.1Khz! Some folks usng Reaper have reported good results with buffer sizes as small as 64 samples which is only 64/44100 or 1.5 milliseconds.
The really unfortunate thing is that Windows was never designed as (and still isn't) a true real time operating system. It goes and steals signifcant time just to paint the cursor (highest priority in the system) and do other silly things (like spying for Uncle Bill), so these buffers are absolutely needed to hold the streaming audio data while the computer is off twiddling it's thumbs (or spyng on you)!! Yes - Uncle Bill does not want any other viruses or trojans on your computer except than genuine Microsoft ones!! - 
|
Mahope Kākou...
...El Lorenzo de Ondas Sonoras |
Edited by - Lawrence on 08/04/2007 07:36:43 AM |
|
|
|
Absolute
Lokahi
275 Posts |
Posted - 08/04/2007 : 08:56:09 AM
|
I go along with the mismatch theory in the Windows/Audacity settings. Changing the Windows settings fixed the accelerated playback problem. I've got half a gigabyte of RAM, which can permit some pretty significant buffering.
Once you get the Windows settings I reference above under multi-media in Control Panel set to the recommended levels (from Microsoft) that I describe, Audacity seems to perform fairly well. I see your point regarding the Windows overhead theory. I'd be more inclined to look forward to switching to Linux someday than abandoning Audacity. (Enough to think it might be worth having a second workstation purchased cheap - even used - on-line at an auction site for use with one or two memory intensive applications.) I like Audacity quite a bit when everything is set up properly and there is no issue with multi-tracking, though Audacity could use a better noise clean-up algorithm, more like SoundSoap, for home users. (Some might add some need for a front end mixer option.) If you're not multi-tracking, and there is no acceleration issue with playback due to bad settings, Audacity is nice, fairly easy to use software, that permits direct output of WAV and MP3 files (with the proper plug-in for the MP3 capability - also free on-line). |
Thank you. |
|
|
|
Absolute
Lokahi
275 Posts |
Posted - 08/14/2007 : 5:06:10 PM
|
Changing Windows Control panel settings got rid of the accelerated playback problem in Audacity. Another fix got rid of the problem synchronizing a live track to a recorded one.
I did some more experimenting tonight, because I'd been able to synchronize to a recorded track in the past. THE ONE FACTOR that enabled me to play along to another track I'd already created while it played back in real time was disabling the "noise cancellation" option on the AC97 control on my computer. (If my mouse lingers over the related icon on the lower bar tray, Windows identifies it as "Sound Effect".) The AC97 driver is the driver for the AC97 audio chipset used on my motherboard. (Other computers probably have something like it comprising a proprietary driver for their audio board or on-board chipset.) If I check noise cancellation, there is a delay between the playback and recording. This delay enables the processing to occur that identifies and cancels background noise. If you eliminate noise cancellation, you eliminate the need for the time delay required to provide for noise cancellation. This worked consistently in terms of rechecking "noise cancellation" causing the time delay and synchronization problem to redevelop in Audacity. |
Thank you. |
Edited by - Absolute on 08/15/2007 04:07:26 AM |
|
|
|
Absolute
Lokahi
275 Posts |
Posted - 08/15/2007 : 04:43:21 AM
|
POPS AND CRACKLES IN RECORDING?
If you experience odd pops and sudden bursts of crackles in recordings, look at the wire connected to your microphone. Is it tangled among various cords in the back of your computer? This problem with noise can be fixed or substantially reduced by eliminating the potential for inductive interaction between those wires as follows:
Step 1: Unplug microphone from back of computer and de-tangle the chord until the microphone and its chord are entirely free.
Step 2: Re-route microphone cord away from other cords in back of computer (or any unshielded cords used by other equipment), away from the monitor, and away from the computers back plate where all the cords come out. (Some transparent tape can be useful here, but be careful sticking it to walls or expensive furniture, as it can remove the paint. The tacky white putty some use for hanging posters can be helpul to permanently reroute the cord, but will probably require a trip to a local retailer unless you use it a lot.)
Note: Make sure you don't route the cord behind the monitor or the computer, and leave enough cord to permit you to place the microphone some distance from the computer and monitor. (They're electrical and audio noise sources.)
Step 4: When recording, place the microphone some reasonable distance form the computer to avoid recording the sound of the computer's fan (to the maximum extent possible). You may have to sit a few feet away when recording to reduce fan noise.
(Hint: Try to get your recording, mistakes and corrections, in one recording at one time, even if you have to let Audacity run a while, so the background noise level and microphone level are constant.) |
Thank you. |
Edited by - Absolute on 08/15/2007 04:55:40 AM |
|
|
|
wcerto
Ahonui
USA
5052 Posts |
Posted - 08/15/2007 : 05:42:33 AM
|
| Does it make a difference if you use shielded cable? EMI is everywhere. |
Me ke aloha
Malama pono,
Wanda |
|
|
|
Topic |
|
|
|
Topic Locked
