recording with sony pcm d50, then direct to cd

[dep]

Hi folks. i have a bunch of questions.

the goal is to record a live concert -- solo singer-acoustic guitarist, live audience, medium-sized room i won't get to scope out much ahead of time -- using a sony pcm d50 stereo recorder, then take it to CD with as little processing as possible. to make it even more interesting, i won't have had much time to experiment with the recorder before the gig. and i'll be using its built-in microphones, which ought to be more than sufficient.

the little dab of post will be, i hope, limited to fading in and out the tracks.

my initial thought is to use the little dab of dither sony allows, making a 44.1/16 recording effectively 44.1/20 for a little extra dab of dynamic range. that, plus judicious gain setting, plus the supposedly great limiter, ought to give me a pretty clean recording.

so, now, my questions:

first, anyone here done that kind of recording with this machine? i'm interested especially in placement. my thought is to place it pretty close, as in on the mic stand itself. make sense? what would be better?

second, i had figured on using audacity to cut the thing up before burning it to red book CD. but my sense is that by merely bringing it into audacity i'm already altering the recording to some miniscule extent. the file will be a bog standard .wav. if this is so, might i be better off just recording at 24/96 and letting audacity do all the dithering? again, the goal here is to do as little as possible to the recording in post-production.

third, while the sony "super bit mapping" is supposed to be just great, it also supplies its own dithering -- just 20 to 16, but still -- and i'm wondering if this could somewhere cause more trouble than it's worth in audacity, if i record at 44.1 using it.

finally, in that all i want to do is fade the cuts in and out, is there something that would be better than audacity for doing it? maybe a very simple .wav editor of some sort (i'm running kubuntu linux).

thanks for suggestions and advice.

dep

[kozikowski]

IMHO, the recording conditions in the room and the proximity to the performers and other similar damage is going to be far worse than anything the digital services is going to do to it.

Are you recording this for the benefit of the performers? Audacity always works at 32-floating inside. It's a similar trick to Photoshop. Both use an internal format far better than any other format likely to arrive with a show.

<<<on the mic stand itself. >>>

Whose mic stand? The singer? Can you get a house feed? Is there even such a thing? Do you know that the performance is amplified or are you guessing?

If it's amplified, you are at the complete mercy of the house sound system. If it's not, you're at the mercy of the room. Have you been in the room? If you clap, does it come back for fifteen minutes?

I'm interested in what the others have to say. I have made perfectly delightful recordings at 48000/16 bit, but I was under tightly controlled conditions where I personally put the soundproofing in for the shoot.

Koz

[dep]

that's precisely the issue in the room: never been there and won't be there until sound check. do not want to take a feed off the board, and yes, i'm thinking the performer's mic stand -- the mic isn't worked closely and will be on a boom, so i have about a meter distance, unless i do something free-standing, which i may well be able to do if that's too close. i know the significance of positioning; was hoping someone here may have used this recorder in this way -- seems the standard recorder test is acoustic guitar, but no one ever specifies where the put the silly recorder relative to the instrument. (and yeah, i know that the room has a lot to do with it, but i'm hoping to minimize room effects.)

so audacity will deal with it in 32 bits even if going in and coming out everything is set to 16? even if default preferences are set to 16? i literally want to do nothing but cut it up and add a quick fade in and a slightly slower fade out.

goal is to see just how good a quality can be gained from doing as little as possible to the recording.

[billw58]

Dep:
It sounds like you have a pretty good handle on the recording set-up end of things. The picture on page 9 of the manual coincidentally shows the device being used to record a singer-guitarist!
Some suggestions:
1. Bring your own stand. A mic stand or tripod. So you can position the recorder independently of the stage mics. Bring a good set of headphones and make some recordings during soundcheck from several positions and choose the one you like best.
2. Forget the built-in limiter. Record at 24 bit 44.1 kHz and set your recording levels low enough that you will never get overload. Ask the performer to do their loudest bit and set you levels to this, at least 6 dB below max, possibly more. Even if the maximum level you record ends up being -12 db you'll still have 22 bits working for you.
3. Set Audacity's defaults to 24 bit 44.1 kHz and import the wav files without conversion. Work in 24-bit.
4. Now is the time to apply some limiting, if needed. At least with Audacity you get to try different settings to find one you like (or decide you don't like any of them), whereas if you use the Sony's built-in limiter you're stuck with it.
5. When the post production is finished, export to 16 bit 44.1 kHz and let Audacity do the dithering.

Hope this helps. I'd be interested to know how the project works out.

-- Bill

[dep]

ah. many thanks.

funny -- i used to do a lot of this kind of thing, but it was when sound editing involved a splice block and a razorblade! indeed, people could have made money by not investing in the things i mastered -- darkroom work, tape editing.

i'll certainly let you know how it works out.

[billw58]

funny -- i used to do a lot of this kind of thing, but it was when sound editing involved a splice block and a razorblade!

Oh dear, sounds like we're from the same era

I used to occasionally do this sort of thing, recording classical ensembles with coincident mics into a ReVox with dbx.

-- Bill

[dep]

best stuff i ever got of a group -- chamber orchestra in this case -- was done through the sennheiser binaural mic (on that silly plastic head with the good ears) into a half-track ampex. would be amusing to consider the s/n of those days, even at high tape speed, vs. even really cheapo gear now. though i wish i still had that mic. wasn't the flattest thing around, but it sure did put things where they belonged spatially. kinda like the original bose speakers.

[kozikowski]

Me and my Ampex 350 have no idea what any of you people are talking about.

48000,16-bit is the television broadcast standard. The tone level (overall average sound level) is -20dB in the US. I believe it's -18dB in Europe. That's 20dB headroom and 60dB noise floor. Any singer I ever met can blow 6dB right out of the water on an expressive peak. Even Consumer DV, widely reputed to be inadequate, uses -12dB.

We did have some Audacity installs, on a Mac, I believe, that had troubles with 24-bit sound. I've always been really leery about that bit depth.

I would tell you to test your brains out, except you can't.

Koz

[dep]

the -20 is, i think, pre-emphasis and came to us by law in connection with fm stations. it's a real pain, too, because it truly screws up s/n. which is why we got compressors, the goal being to make the quiet parts as loud as possible. which then became the standard for how most music sounds (producers thinking that if it's gonna get smashed, might as well do it in the studio and retain control). for a time i had a dbx compander, the idea being that i could record vinyl to tape with it all smashed flat, then play back with it expanded again and lose all the tape hiss. problem is, there was *a lot* of processing noise.

dynamic range isn't nearly as much an issue with digital stuff unless you're going to put it on commercial fm, in which case it will be smashed against the top no matter what you do.

as to the 16-vs.-24 thing -- wonder if that falls victim to the widely supposed exact multiples thing?

a good discussion of compression hell is here: http://www.rollingstone.com/news/story/ ... h_fidelity

[kozikowski]

<<<the -20 is, i think, pre-emphasis and came to us by law in connection with fm stations.>>>

Don't believe that too much. It's the digital television standard and it doesn't have pre-emphasis. And it doesn't have any trouble making it to my living room in digital form...

http://www.kozco.com/mytv/mytv.html

I've been doing this for almost a year and I went back to the analog TV in the corner (before we turned ananlog off) in order to record one show while I watched another. It was pretty dreadful and I have terrific television reception. I guess I used to watch this all the time and it was normal.

I think rental CDs and television reception should look the same now. Perfect.

That's not to say it all actually comes out the same. There are two shows on PBS with distorted audio. Badly distorted. I wrote to the producers wondering if anybody on the production staff actually listed to the air show.

http://kozco.com/bmjournal/bmjournal.html

Koz