I’m thinking about recording my band next weekend using nTrack. I’m going to do 8 tracks 44.1/24. No-one will be running the computer (except me & I’ll of course be playing guitar).
I was thinking about have 3 project files, one for each set (we play 1 hour sets) & just recording the entire set. Of course, I’ll have to figure out some efficient way of breaking the large tracks up into individual songs.
Any thoughts about doing it this way or not?
Thanks in advance
I’d suggest not doing it this way. If anything happens before you click stop then the whole track can be lost. The header (wave data size) doesn’t get written to the file until that is done. I suggest breaking up the set into smaller chunks of no more than three songs or so - 10 to 15 minute chunks. It’s better to be safe than loose a whole set.
This is one of the big disadvantages of PC recording.
There are ways around it in software, but the app needs to be do a better job of dealing with it. The way to avoid it is to periodically write out smaller wave chunks in a way that can be reassembled into a large wave when recording is stopped (chop the saved wave into smaller waves automatically)- or keep writing the running wave file header that it always has the latest data size (almost the same thing but not quite). Of course that second one would GREATLY increase recording overhead.
I don’t know auto-safe will handle this.
I’ve done it a couple of times. Ntrack is certainly capable of doing it, but don’t stop and start recording for the heck of it - you’ll think ntrack has crashed the first time you press stop after a 30/40/50min set while you wait for it to draw the waveforms.
What I do is select and loop the first song. When I’m happy with that, I mixdown the selection,and save the .sng. Then I select the next song and use the previous settings as a starting point and mixing down the new selection. Save the .sng again if you want, or as a different name for each song so it’s easy enough to come back to. .sng files don’t take up any space anyway.
Phoo is right though, you want a completely rock solid system to do this with.
You could just hit stop, save, then start again-leaving the same file name and everything. that way each song is still it’s own set of wav’s but you don’t have to mess with settings again. It’ll all be one big song file, but each song will be its own section
ahh, my specialty,
Here’s what I do:
My band play 100% improv, so our practices are just as unique as ours shows, both of which I record regularly. We also tend to jam nonstop for extended periods, anywhere from 15 min. at practice to an hour+ at shows. What I do is setup my inputs (8-10 usually…on a MOTU 828), save the project in its own “session” folder with its own song name, and then begin recording. When we take a break, I stop recording, save the project, open a new project, and then immediately save it as a new song name but in the same “session” folder. When we’re ready to go again, all I have to do is hit record, and we’re off. Repeat ad infinitum.
Now you could adapt this to doing it between every song, or maybe every couple of songs (really it only takes a few seconds to save the project, open a new one, and save it again with a different name). You could have each of you’re three sets be a separate session with its own folder. You end up hopefully with separate project names for each song, with all the .sng files and all the correspondingly named .wav files in the same recording session folders. Nice and tidy.
Hope this helps.
If I understand you correctly, this is exactly what I'm proposing. So this will work I presume?
I like phoo's idea of smaller chunks.
Starting/stopping would create 8 tracks * number-of-songs in set which could make the total tracks very large. I'm not sure nTrack will handle that will it?
My machine is very powerful & stable.
I think you’ll have good success on your machine. I typically record 32 tracks in about 15 minute chunks. Glitches and errors are rare, occurring less than 1% of the time. And my PC is less capable than yours.
|Quote (MidnightToker @ Sep. 30 2004,16:28)|
|If I understand you correctly, this is exactly what I'm proposing. So this will work I presume?|
I like phoo's idea of smaller chunks.
Well, I'm not sure you understood 100%, cause that wasn't EXACTLY what you proposed. You asked about each hour-long set being a separate song. Like you said, phoo's idea of smaller chunks is probably the better way to go. I was trying to explain that it's easy to make a new project for every song or every couple of songs, so you won't have three hour-long projects to deal with or have mess up.
What you would have if you were to do what I do would be three folders, one for each set. Each folder would contain all the .sng files for each of the songs in the set (or groups of a few songs maybe) and the corresponding .wav files. Your sets would already be split up into their respective songs, your .wav files would all be easy to identify, and you wouldn't run into possible large losses if something went wrong, cause you'd only be recording for a few minutes at a time (assuming average song lengths.....?) between saves.
I hope this helps to clarify. Helll, I hope this helps. If there are questions, just shoot.
Thanks for the clarification.
The requirements I have are that no-one will be running nTrack so this is why I proposed letting it run for the entire set.
This is definitely NOT the way I would do it if I were able to engineer the project.
However, while I’m playing I don’t want to constantly be distracted with the recording.
I just want to know if anyone else has tried what I’m suggesting, i.e., let nTrack for an hour recording 8 tracks at 44.1/24 bit. It wasn’t clear that anyone had actually done it for this long.
Yeah, I was doing it with an ISIS/98lite/celeron600@900/abit bx6r2/ntrack2.3/16bit/44.1 for probably 45mins to an hour per set.
Thanks Willy! I’m gonna’ try it.
I know everyone’s worried about a crash killing long files, but it may not be that big a risk. As long as nTrack streams to disk under the assigned filenames, if a long recording crashes, I believe there are some soundfile utilities around that will open and properly complete a truncated WAV file so that it can be opened properly. You may lose a couple seconds of audio prior to the crash, but that’s it.
Can’t remember the name of the WAV repairer, though…
I always found ntrack 2.3 very stable for tracking. It was just when I started mixing and doing things like… drawing volumes or eqing tracks.
Yes Toker, I’ve also done hour long recordings of live shows with no problems (at least no problems related to N). I would think you should be fine.
Is it a laptop you’re recording to? I’ve had problems with drops and glitches on long recordings that I think are due to my laptop HD not being fast enough. If you are using a laptop, I’d suggest an external HD.
What kind of disk space am I looking at for 8-tracks of 44.1/24-bit? Is there an accurate way of calculating it. I’m going to run some tests tonight but I’m thinking that I may not have enough space (I have +80 GBs).
Thanks in advance.
Think for a second. There are 24 bits per sample and 44100 samples per second, so each second is 1058400 bits/8 = 132,300 bytes. Long story short, you will have 7.5648 MB per minute per channel (double that for stereo files) in 24/44.1 if I did my math right. So 1 hour x 8 tracks = 3631.104 MB ~ 3.5GB.
Well I ran a test for 10 seconds & each track was over 1 MB, so I started getting alittle concerned. Who is going to 2nd these calculations?
Um yeah, 7.5648MB per minute / 6 = 1.26 MB per 10 seconds. So um, what I said sure sounds about like what you found. This is basic math. Grab a calculator if you are unsure and crunch it out. There is going to be a few KB in header info etc. on each file, but the math I did should be pretty close. I’m not going to do all your home work for you.
You're starting to sound like Pete now