Rockbox Technical Forums
Support and General Use => Recording => Topic started by: xmixahlx on May 05, 2009, 06:00:54 PM
-
hi all,
i was hoping y'all would be able to give me advice for recording from the soundboard. there are AWESOME guides on doing this for audience recording, but i didn't find much for soundboard recording.
i'm using a rockbox'ed h320 with ~week old CVS (updated before the shows, on 4/30)
i actually ALREADY DID record about 15 hours of sets (over two days), but was definitely not sure of the settings i was using. i did my best with what little i knew (it was a totally impromptu thing). i had my h320 hooked up to power at the soundboard and was using a mini-jack to stereo RCA cable connected to RCA to XLR converters (female RCA to female XLR, which i had to run to the guitar center to pick up...) and that was patched into the soundboard's stero XLR outputs. i also taped (masking tape) the inputs so that nothing would get loose or accidentally removed while i was away.
anyways, of course i probably did a ton of stupid recording mistakes and i'm hoping to avoid those this weekend when i do this again (and in the future!).
here are the settings that were used:
format: wavpack
frequency: 44.1k
source: line-in
channels: stereo
split measure: time
time measure: 4 hours (to avoid >4GB WAV?)
pre-record time: off
clipping light: main unit only
trigger: off
AGC: Live
AGC time: 200ms
peak meter settings (ALL DEFAULT)
i was essentially recording ~8 hours of sets per day (which was why i chose wavpack. i deleted tons of stuff off of the player, but filespace was still an issue) and would restart the recording every 3 bands or so (~1:30 to 2:00 hours) just because the file splitting was kinda sketching me out. i was monitoring levels at the soundboard and levels seemed to be a range (during play, not break down and introductions with just mic input) at roughly -4dB to -9dB and very rarely up to -10dB. also, i noticed that the right channel was often moved up 0.5dB higher than the left channel.
the "AGC Live" setting that i had selected would step back the gain during breakdown and hit about ~0dB until the next band played and then would quickly go back to the -6dB to -9dB range.
i REALLY didn't want this to clip so was babying the recording at times.
should i have used these settings? i realize now reading threads here that folks are only using the AGC Safety setting, but perhaps that is strictly for audience recordings. (?) looking back it seems like just setting the gain at -10dB should have sufficed, but i was just lacking knowledge of how to do this.
listening to the sets, quality is amazing. there are a few moments where it seems that the gain boosts (perhaps detecting where the AGC kicks in?) but i'm very pleased with how it turned out for such low expectations going in.
sooo... my questions are these:
1 what (if any) gain control should be used for soundboard recordings?
2 any of my steps wrong here? how can i do this better in the future?
3 how do you handle the difference in volume (i.e. mic-only input versus band playing)? is it even an issue?
4 are there any problems with using wavpack versus wav?
5 is file splitting working correctly? (i.e. no lost audio samples between splits)?
6 what would be the best "set it and forget it" type of settings for soundboard recording?
7 should i setup a mic for audience participation (i.e. crowd noise, etc.)?
i guess i just don't understand the gain and peakmeter stuff that well. i've read a bunch of threads, the manual and some other sources and it just isn't solid for me yet.
thanks for any responses,
-michael
-
I use my iriver H300s extensively to record concerts off the soundboard at the guitar summer camp I go to every year. Your settings look fine to me.
I do some things a little differently, but they are not of major consequence.
I get my line out from the mixing board using 1/4" unbalanced outputs rather than XLR. I use a 1/4" to RCA adapter, then a RCA to 1/8" mini-jack cable. The sound man graciously sets up a separate bus on the board from which I get my feed. Another output setup we used to use in the past was just from the tape out jacks, RCA to 1/8" mini-jack cable.
I like to use the AGC Safety setting because it automatically lowers the gain, then leaves it lower. That way you don't get volume levels drifting up and down and up and down. If you are recording at 16-bit resolution, you probably have enough room above the noise level to readjust volume levels in post-processing. But if I know someone is going to be really quiet (e.g., dulcimer player), what I do is raise the gain before they start playing, so that I have a good clear signal. If the signal is really small, you can get quantization effects that you can't fix in post-processing. (That is, a REALLY small signal not only has lots of system noise -- hiss -- but also the range between no signal and its highest signal doesn't have many volume steps, because it's only using a small part of the entire 32,000 volume steps available in 16 bits).
You are just going to have to learn from trial and error what the best gain settings are, and be prepared to change them some from one night to another. One problem I have sometimes run into is getting a signal from the board that is too hot (too high in volume), giving me clipping and distortion. If a signal is too strong, it overloads the amplifier circuits, and the sine wave pattern gets a haircut (clipped). Sounds awful. If it happens to you, you may have to work with the sound man on adjusting his output settings, since you can't fix that in your recorder, only on his board.
Like you, I also use Rockbox's file split settings, but since I tend to the recording during the whole concert (they range from about 3 hours to about 7 hours, with breaks), I split files manually. I really love the way Rockbox lets you just push the record button to start a new file. There is absolutely no delay, and absolutely no loss of audio samples (unlike the old days when I had to turn over a tape...). I also like to use the pre-record, set at 30 seconds. Those MCs and bands sometimes fool you and start up before you are ready. Pre-record means never missing something at the start.
With a 60GB and an 80GB iriver, I do not have your capacity limitations, so I record at a full 16-bit 44.1K wav format.
I find it is a whole lot easier to go back and re-adjust volume levels in the recording (musician or MC speaking, vs band playing) using an audio file editing program after the event, rather than trying to do it on the fly, on the spot. I like to use Audacity, an excellent open-source program for doing this, and it's free.
I like to have a separate mic for the audience. It's especially valuable when people sing along. At a music summer camp, they can actually sing competently, in rich harmonies. I have one channel for the audience, one for the stage, with some panning for overlap. The stage is panned close to the center of the audio field, and the audience is further to the side. This also has to be done on the sound board. In the past, I had a separate mixer taking feeds of the main board to set up this simple two-channel mix. Now my friend the sound man sets that up on the main board on a separate bus from the one used to send signal to the room speakers.
Good luck, and enjoy!
-
hi pete, thanks a lot for the reply.
it sounds like we are doing similar configurations and your input is a big help.
regarding a "hot signal" > how would you tell this is happening? would the peakmeter be overloaded even at low gain settings?
i will definitely use AGC Safety instead. everyone is recommending it and after i detected some gain boosting in my recordings i'll now have to process those... :(
how do you usually "setup" the recorder? do you just monitor the peak meter and slowly step down the gain before hitting record? or do you always start at a certain setting? and do you have any tips on processing the audio? any patterns that you usually run into? - i too use audacity for simple processing (fade in/out, cutting, etc.)
i was actually stopping and starting the recording after listening back to the last minute or so to make sure that the levels were good and the recording sounded OK. but once i'm confident in these recording settings i can tell that will be a great feature.
pre-record sounds like a good idea, too. i was just beginning the recording during band setup and tear down (so like... 10min of PA background music is recorded before & after each band) and i was frequently not around the soundboard (i.e. in the crowd, etc.) so i didn't want to wait to press "record".
oooh so you swapped out the hard drives of your h3x0's? how hard was that? anything i should keep in mind - i DEFINITELY want to do this. the hard drive on my h320 seems to be on the fritz.
what AGC peak time setting do you use?
do you find that the gain varies widely from start to finish? (i.e. 0dB to -10dB)
thanks again,
michael
-
+1 for using AGC-safety, not AGC live or other options
If you get your input from the soundboard, it should be pretty line-level, and the guy at the soundboard would probably be able to adjust. I'd even go for setting it to 0dB and no AGC.
-
hi petur,
so do you think there was a problem with the output from the mixer if the AGC was stepping it down to about -9dB during each set? if i had set it at 0dB (and no AGC) that would be clipping all the time, right?
-michael
-
for some odd reason the irivers can record without clipping at -xxdB, but not too much. I think -9dB is too much or on the limit. The best test is to check your recording. If there's no clipping, you're fine ;)
-
I agree with Petur, the best strategy is to check your recording. Finding the best initial setting is trial-and-error, and after a while you will have a sense for suitable starting levels.
Clipping is a signal that is too strong for the electronics to handle without distortion, and it appears in the waveform (that you view on Audacity) like a signal with a flat-top hairdo. You might want to experiment by running a line-level signal into your iriver from your stereo (e.g., from tape out) and push the iriver record levels up until it overloads and distorts. You'll get a clear idea of what clipping sounds like, and then when you are monitoring the recording at the concert, you'll know what to look out for.
There are actually several possible sources of clipping in the recording chain. The recording chain has several discrete stages at which some amplification may take place, one after the other, each of which could be overloaded. Sometimes a singer will get right up on the microphone and holler into it, and overload the microphone circuit, which might have a limit of 100 dB, for example. Or, the pre-amp (trim pot) on the mixing board could be set too high, overloading that circuit. Then there are one or more other amp circuits that follow in the mixing board. As an amateur, I have learned from wretched experience how to screw up the gain sequence in a mixing board and cause that clipping. Pros are unlikely to do that to you, but if this is a beginner sound man, it can happen.
When the signal is passed to the iriver, its input circuits can be overloaded on the entry (i.e., a signal that is "too hot"), and even though you lower the internal amp gain of the iriver through Rockbox, you still get a distorted signal recorded, despite not having the clipping signals going off on the iriver display. The solution is to lower the signal output from the sound board; the sound man will know which knob controls the level of the feed to you.
I especially like the AGC-Safety setting because it gently lowers the gain in 1/2 dB increments so that it is not too obvious in the final recording.
Pre-record is handy for keeping the band from giving you a rude surprise by starting up when you didn't expect it. But take note: you can only set it to a maximum of 30 seconds (not several minutes). It records continuously, maintaining a 30-second buffer of audio data, dropping off the oldest data to make room for the newest. So if you are out in the audience and the band starts up, you have 30 seconds to dash back to the recorder to push the record button and catch the whole act. Pretty cool.
My H360 and H380 started life as H340s. It is very simple to swap the 40 GB hard drive for a 60 or 80 GB hard drive (Toshiba MK6006GAH or MK8007GAH respectively). These are 8mm thick, and will not fit in an iriver H320. For that, you need a 5mm thick Toshiba MK3006GAL (30 GB) or MK4007GAL (40 GB). The 30 GB drive can be found on the net for $30 to $50 or so, the 40 GB one for $130 to $150. Another option is to replace the hard drive with a 32 GB compact flash card and adaptor for around $80-$100 depending on the speed of the card you choose. It is more complicated to install the card, but the results are very pleasing. My H332 and H308 started life as H320s. (Actually, I have now changed my H308 to another H332, using a different brand of CF card.) There is more info on either the hard drive upgrade or the CF upgrade elsewhere on the Rockbox site.
I think a 200ms AGC delay time should work fine.
When the AGC setting is Safety, the gain can vary a fair amount from start of recording to finish, but don't worry. Use Audacity to level out the volume levels in the recording for the final product.
-
thanks for the replies (again) guys,
the recording reached -9.5dB, but i didn't amplify the signal more than 7dB. it DID clip on a few occasions (3 times in ~10 hours of sets), but only because i had been using the AGC Live setting, which moved the gain too high for the signal.
to fix the recordings, i used audacity and a combination of amplify, leveller (light, -70dB floor) and repair for some short (non-clipped) sounds in the mix (as well as fade in/out). hopefully i won't have to do that again! although i am now pretty confident using audacity :) - are there any other tools or methods that i should check out?
i wasn't able to record again (they already had someone recording the soundboard) to use the AGC Safety settings that y'all recommended, but i sure will do that next time.
if you want to check out how my recordings turned out, here are some mp3 preview encodings (lame V6 ~128kbps):
(btw, this is hardcore music, burning fight book release shows in chicago...)
http://rapidshare.com/files/232598393/Burning_Fight_Chicago_20090502-Phase1-106-Ringworm.mp3.rar
http://rapidshare.com/files/232607685/Burning_Fight_Chicago_20090502-Phase1-107-Guilt.mp3.rar
http://rapidshare.com/files/232618778/Burning_Fight_Chicago_20090502-Phase1-108-Underdog.mp3.rar
http://rapidshare.com/files/232627219/Burning_Fight_Chicago_20090502-Phase1-110-Killing_Time.mp3.rar
http://rapidshare.com/files/232639765/Burning_Fight_Chicago_20090502-Phase1-111-Trial.mp3.rar
http://rapidshare.com/files/232645443/Burning_Fight_Chicago_20090503-Phase1-201-Harms_Way.mp3.rar
http://rapidshare.com/files/232651940/Burning_Fight_Chicago_20090503-Phase1-202-The_Killer.mp3.rar
http://rapidshare.com/files/232658543/Burning_Fight_Chicago_20090503-Phase1-203-Blacklisted.mp3.rar
http://rapidshare.com/files/232663679/Burning_Fight_Chicago_20090503-Phase1-204-Have_Heart.mp3.rar
http://rapidshare.com/files/232670994/Burning_Fight_Chicago_20090503-Phase1-205-Threadbare.mp3.rar
http://rapidshare.com/files/232677389/Burning_Fight_Chicago_20090503-Phase1-206-Split_Lip.mp3.rar
http://rapidshare.com/files/232683018/Burning_Fight_Chicago_20090503-Phase1-207-Reach_the_Sky.mp3.rar
http://rapidshare.com/files/232689390/Burning_Fight_Chicago_20090503-Phase1-208-Bane.mp3.rar
http://rapidshare.com/files/232698599/Burning_Fight_Chicago_20090503-Phase1-209-Converge.mp3.rar
http://rapidshare.com/files/232708923/Burning_Fight_Chicago_20090503-Phase1-210-Unbroken.mp3.rar