Live Recording: Critique/Comments Needed

[ryanm]

Ok, let me preface by saying that we got some drum mics (thanks for the
offer Harvey, but we broke down and spent the money), we picked up some
Audix mics (D2s for toms & D6 for kick) and an SM57 for snare/hat. I like
the sound I can get with these, although I still would've liked to spend
more on the kick mic. The D6 seems to be good bang for the buck (~$150),
though, and sounds pretty good no matter where you put it. We also got the
new drummer, and this recording is of his first weekend with us (2nd night),
exactly 6 days after we gave him the song list, so the performance is
understandably a bit shaky (particularly the end of the Gypsy Queen part of
Black Magic Woman).

With that in mind, here are a couple tracks I captured that night. I'm
posting them in wav for if you have broadband, mp3 if you don't, and I'm
posting both some short clips as well as the full songs, in case anyone is
bored and wants to get the whole enchilada. I chose these tracks because
they encompass almost all of the instruments we use, everything but
harmonica. The mp3s are encoded at 256 kilobit with the LAME encoder.

Crossroads Short wav (11 megs):
http://www.horsefish.net/music/crossroads.wav

Crossroads Short mp3 (2 megs):
http://www.horsefish.net/music/crossroads.mp3

Crossroads Full mp3 (13 megs):
http://www.horsefish.net/music/Crossroads_full.mp3

Black Magic Woman Short wav (8.5 megs):
http://www.horsefish.net/music/bmw.wav

Black Magic Woman Short mp3 (1.5 megs):
http://www.horsefish.net/music/bmw.mp3

Black Magic Woman Full mp3 (7.75 megs):
http://www.horsefish.net/music/Black_Magic_Woman_full.mp3

And you can't play Black Magic Woman without playing...
Oye Como Va Full mp3 (6.88 megs):
http://www.horsefish.net/music/Oye_Como_Va_full.mp3

The room was pretty good (about as good as bars get for acoustics),
plenty of bodies, and we're playing quieter than ever, which is a good thing
as far as I'm concerned, so what you hear comes mostly from the mains rather
than the back line. The excerpt from crossroads is just a 1 minute clip that
has part of a guitar solo, one verse, and part of a drum solo in it, so you
can hear a bit of each of those instruments for reference. The drummer
didn't know he would be playing a solo, I kind of dropped it on him, but
it's a good clip to hear how the drums sound. The full song is a good
example of the kind of dynamics we like to use. The Black Magic Woman short
clip is just part of the intro so you can hear keys, guitar, and percussion.

To record this I used a DA-P1 and a pair of out of production AT mics
(3251 maybe?). I set the mics up together about 15-20 feet from the mains,
as close to center as possible, and pointed one a bit towards each side of
the stage.

Anyway, what I'm looking for are comments on the overall levels,
balance, quality, etc of the recording. If something sticks out at you, what
can I do to fix it (i.e. move mics around, change input levels, etc)? Is
there any post-processing I can do (eq, etc) to make these tracks sound
better. I'm not looking for perfection, just a good live recording.

Also, a technical question about the DA-P1. It seems to "skip"
sometimes. I'm not sure how else to describe it, it just skips now and then,
almost like a cd skips. I've spent countless hours re-dumping tracks to my
computer to fix what I thought were glitches caused by my computer running
out of ram while I was recording, but turned out to actually be on the tape.
Anyone have any idea why it might be doing this? Do I need to erase the
tapes between recordings, or is it a battery problem, or what?

Telling me we suck or that we sound like a bar band is preaching to the
choir, but any specific comments or suggestions are greatly appreciated.

ryanm

[Peter Larsen]

ryanm wrote:

> Also, a technical question about the DA-P1. It seems to "skip"
> sometimes.

To the repairshop it go, adjustment may fix it. I know several owners,
and one of them has had really really bad problems with this. Adjustment
seems to have solved it. Another owner claims that his never does it,
but it does - I have it on a CD with a listening copy, but it is only
very very short.

> I'm not sure how else to describe it, it just skips now and then,
> almost like a cd skips. I've spent countless hours re-dumping tracks to my
> computer to fix what I thought were glitches caused by my computer running
> out of ram while I was recording, but turned out to actually be on the tape.
> Anyone have any idea why it might be doing this? Do I need to erase the
> tapes between recordings, or is it a battery problem, or what?

It appears to be a matter of adjustment of the tape drive mechanism.

> Telling me we suck or that we sound like a bar band is preaching to the
> choir, but any specific comments or suggestions are greatly appreciated.

Later .... downloading right now.

> ryanm


Kind regards

Peter Larsen

--
************************************************** ***********
* My site is at: http://www.muyiovatki.dk *
************************************************** ***********

[ryanm]

> It appears to be a matter of adjustment of the tape drive mechanism.
>
Anyone know where, in Dallas, I could take it?

I'm reusing the tapes, but I've only recorded on them a half-dozen
times. They're 90 minute tapes, and I use them to record my band playing 60
minute sets at a few of the better sounding bars we play. They always run
from start to finish, because I start them ~10 minutes before we start
playing and then stop and fast forward them to the end before rewinding
after the set. I don't know if this is necessary, but I figured it can't
hurt to be sure the tape is wrapped evenly on the spools.

So this is a playback problem, rather than a recording problem? I've
noticed that the skips are always in precisely the same place, I've even
captured the part with the skips in it a few times to my computer and zoomed
in on them to see if the skip is in *exactly* the same place, and it seems
to be. I assumed, since they were always the same, that the skip was in the
recording rather than the playback.

thanks,
ryanm

[Peter Larsen]

ryanm wrote:

> > It appears to be a matter of adjustment of the tape drive mechanism.

> Anyone know where, in Dallas, I could take it?

> I'm reusing the tapes, but I've only recorded on them a half-dozen
> times. They're 90 minute tapes, and I use them to record my band playing 60
> minute sets at a few of the better sounding bars we play. They always run
> from start to finish, because I start them ~10 minutes before we start
> playing and then stop and fast forward them to the end before rewinding
> after the set. I don't know if this is necessary, but I figured it can't
> hurt to be sure the tape is wrapped evenly on the spools.

I'd skip rewinding, and just leave them at "the other end" so that they
get a rewind before playback, but that's just me and I may be old
fashioned. I find that I get lower error counts with DDS 90 meter
computer DAT tapes than with audio tapes with my SV3800, those are three
hour tapes.

At their price I use them once only. Using the tapes a maximum of 12
times is implicitly recommended in advice I have seen on when to use
cleaning tapes in a server room. Keeping the recorded tape is a simple
way to keep a backup copy of the recording after transfer to daw.

I too like to start the tape transport early, but that is because of a
known imperfect tape mechanism adjustment, the machine is prone to drop
outs 12 minutes into a cassette of 60 meter audio tape, my asumption is
that tape tension changes in a step thereabout.

> So this is a playback problem, rather than a recording problem? I've
> noticed that the skips are always in precisely the same place, I've even
> captured the part with the skips in it a few times to my computer and zoomed
> in on them to see if the skip is in *exactly* the same place, and it seems
> to be. I assumed, since they were always the same, that the skip was in the
> recording rather than the playback.

It is a _recording_ problem, playing back the tapes on a quality Sony
(77) tape transport makes no difference to the guy I know how had the
problem.

> thanks,
> ryanm


Kind regards

Peter Larsen

--
************************************************** ***********
* My site is at: http://www.muyiovatki.dk *
************************************************** ***********

[Peter Larsen]

ryanm wrote:

> With that in mind, here are a couple tracks I captured that night.

All things considered they are nice and airy.

> Telling me we suck or that we sound like a bar band is preaching to the
> choir, but any specific comments or suggestions are greatly appreciated.

For a sucking bar band you sound OK ... O;-)

Disclaimer: listened late at night.

> ryanm


Kind regards

Peter Larsen

--
************************************************** ***********
* My site is at: http://www.muyiovatki.dk *
************************************************** ***********

[Peter Larsen]

Sugarite wrote:

> Something is making the treble pretty brittle, probably one or all of three
> things:

> - the treble horns on the PA (definitely a contributor)

agree

> - the recording mics (be more specific, large/small diaphram)

agree

> - the DA-P1's preamps

nah, they're just slightly boring.

> Otherwise it's good material, nice job on the mic positioning, that's about
> 90% of the battle. I'd be interested to know how widely they were spaced,
> sound pretty much right on top of each other, which is what I always
> recommend when a PA is involved.
>
> I took the liberty of downloading the .wav file and tweaking it. I do that
> sort of thing a lot, and I can tell you a few post-processing things:
> - use a good analyzer to find any problem frequencies, in this case I
> thought 520Hz was a little hot
> - widen the **** out of it, a rear mic position gets the best tone but damn
> close to mono, widening makes it sound like a front row seat
> - almost always need a de-esser, compression horns expand when pushed, in
> this case at 9150Hz

Jez, zomethin's funny about the pa verzion of z and tz's but there is
also the A-T signature treble sound.

> - Multiband compression also helps add proximity, attack times start
> long-ish with the bass (~65ms) and get shorter as frequency rises, right
> down close to 1ms for 6kHz and up to help clean some of the brashness, but
> the release times should be shaped more like a smile, so ambient mud and
> sibilants are de-emphasized
> - peak limiter shaves the spikes, of which there are always many
>
> Here's what I came up with:
> http://home.cogeco.ca/~zigakly/crossroadstweaked.mp3

Perhaps I'll comment later, I may even try to do better ... O;-) ...
that of course depends on how good ya got it and at what sonic cost.


Kind regards

Peter Larsen

--
************************************************** ***********
* My site is at: http://www.muyiovatki.dk *
************************************************** ***********

[Peter Larsen]

Sugarite wrote:

> Here's what I came up with:
> http://home.cogeco.ca/~zigakly/crossroadstweaked.mp3

not a valid file.

--
************************************************** ***********
* My site is at: http://www.muyiovatki.dk *
************************************************** ***********

[ryanm]

> Something is making the treble pretty brittle, probably one or all of
three
> things:
> - the treble horns on the PA (definitely a contributor)
>
Our mains are Yamaha S215IV's, which have a 2" horn driver. They're not
what I would call great speakers, but for the money (about $750 per side)
they sound pretty good. I think most of the problem treble comes from the
cymbals, which aren't miced and you hear only from the room and mic bleed.
The snare mic does pick up a lot of hi-hat.

> - the recording mics (be more specific, large/small diaphram)
>
The mics are AT3528's, small-diaphram condensers. Specs here:
http://www.audiotechnica.com/prodpro/profiles/AT3528.html

> - the DA-P1's preamps
>
Not much I can do about that right now. At present my total investment
in this recording setup is $20 for the tapes. I was working for a game
company making racing games, using their DA-P1 at the track to capture track
sounds, pit noise, crowd noise, etc. I had the deck at home for several
months while collecting sounds, and then they went bankrupt (as game
companies do) without paying me that month. I hung on to the DAT figuring if
they came looking for it I'd tell them they could have it when I got my
paycheck, which was much larger than the cost of the deck. They never asked,
and that was over a year ago, so I started using it for the band. My
intention is to eventually pick up an HD24 to drop in our mixer rack, and
just take line input from each channel of our mixer, so that I can make
adjustments and add effects later if I want.

> Otherwise it's good material, nice job on the mic positioning, that's
about
> 90% of the battle. I'd be interested to know how widely they were spaced,
> sound pretty much right on top of each other, which is what I always
> recommend when a PA is involved.
>
Yeah, they were. They were actually on the same mic stand, at something
like a 20 degree spread, pointing approximately at each guitar position on
the stage, just inside the mains.

> I took the liberty of downloading the .wav file and tweaking it. I do
that
> sort of thing a lot, and I can tell you a few post-processing things:
>
I appreciate the effort. The difference is, uuuhh... less than subtle. I
have a couple questions.

> - use a good analyzer to find any problem frequencies, in this case I
> thought 520Hz was a little hot
>
Any suggestions of good software to use? I'm using Sound Forge now, but
I'm unable to glean any useful info from the built in spectrum analyzer.

> - widen the **** out of it, a rear mic position gets the best tone but
damn
> close to mono, widening makes it sound like a front row seat
>
Can you define "widen" for me? Do you mean move the mics farther apart,
or are you talking about something I can do in post to "widen" the sound?

> - Multiband compression also helps add proximity, attack times start
> long-ish with the bass (~65ms) and get shorter as frequency rises, right
> down close to 1ms for 6kHz and up to help clean some of the brashness, but
> the release times should be shaped more like a smile, so ambient mud and
> sibilants are de-emphasized
>
The best I can do for multiband compression, at present, is the Sonic
Foundry "multiband dynamics" plugin, which only offers 4 channels that are
assignable to frequency ranges. Did you use software or hardware? If
software, which software?

> Here's what I came up with:
> http://home.cogeco.ca/~zigakly/crossroadstweaked.mp3
>
Much brighter and fuller. The high end is a bit bright, but it's an mp3
and the compression artifacts are distracting, to say the least, and you
mentioned a problem with the treble in the source. I've spent some time
listening to it A->B with my original, and it's a definate improvement. Are
the things you mentioned the *only* things you did to it, or is there some
other basic EQ and stuff that you applied and just didn't mention? The
tracks I posted were straight from the tape, with no post processing
whatsoever.

> As for your DA-P1 troubles, sounds like an alignment issue like others
have
> said, not a costly repair. I've done alignments on Sony portables before,
> which I imagine are much less user-repair-friendly. However if I were you
> I'd fix it and sell it, DAT decks are dropping in value fast.
>
See my comment above about how much money I have in the current setup.
: )

Also, I should mention that the only people who will ever hear these
recordings are the band members (except for the samples I post here), and
are mostly for the purpose of improving our performances. However, I'm not
one to let the opportunity to get better at something pass, so I figured
these recordings are a perfect opportunity for me to learn. If we do happen
to capture a gem, it may find it's way onto the band website or a demo disc
as part of a promo pack or something, but we're not making discs to sell or
anything like that. We may eventually record a disc of originals, but we'll
go to a proper studio for that.

Thanks for the time and effort, it is appreciated.

ryanm

[Arny Krueger]

"ryanm" > wrote in message

> Ok, let me preface by saying that we got some drum mics (thanks
> for the offer Harvey, but we broke down and spent the money), we
> picked up some Audix mics (D2s for toms & D6 for kick) and an SM57
> for snare/hat. I like the sound I can get with these, although I
> still would've liked to spend more on the kick mic. The D6 seems to
> be good bang for the buck (~$150), though, and sounds pretty good no
> matter where you put it. We also got the new drummer, and this
> recording is of his first weekend with us (2nd night), exactly 6 days
> after we gave him the song list, so the performance is understandably
> a bit shaky (particularly the end of the Gypsy Queen part of Black
> Magic Woman).

First off, I like the playing.

Using the CEP/Audition parametric eq on Crossroads:

lo shelf f=20 Hz +13 dB
low notch f=30 Hz Q=2 -14 dB
high notch f=11 Khz Q=3 -12 dB
high shelf f=18 KHz +12 dB

In short the SM57 kinda screwed the cymbals, and there was probably some
kind of a room resonance affecting the low end.

Nice cover of the Cream. the Santana was OK, but...

[ryanm]

"Arny Krueger" > wrote in message
...
>
> First off, I like the playing.
>
> Using the CEP/Audition parametric eq on Crossroads:
>
> lo shelf f=20 Hz +13 dB
> low notch f=30 Hz Q=2 -14 dB
> high notch f=11 Khz Q=3 -12 dB
> high shelf f=18 KHz +12 dB
>
> In short the SM57 kinda screwed the cymbals, and there was probably some
> kind of a room resonance affecting the low end.
>
I don't use CEP/Audition, I have the Sonic Foundry tools, so I tried
several different parametric eq plugs that I have (4 of the built in SF
parametric EQs stacked and the Waves REQ 4 band), but they both gave
different results, so I don't know if I'm hearing it how you intended, but I
think I get the idea. The notches to remove the problem freqs (high for the
bright cymbals, low for the "boomy" bass) and the shelves to bring out the
outer freqs that weren't causing the problem. Also, you didn't mention the
rolloff on the shelves, but I tried everything between 2 and 10 octaves.

> Nice cover of the Cream. the Santana was OK, but...
>
I told you Gypsy Queen was a mess <g>. Like I said, 2nd performance with
the new drummer, and the arrangement on that one is pretty specific. We were
pretty much playing follow the leader. It's also better when the congas are
in tune <g>. We do normally do that one better, but if you have anything
specific to point out (I know some stuff was not in the right place, that
has been fixed since) I'm always open to critique. The crowd loves that song
because every other band in Dallas that does it does it with a 3 or 4 piece,
which isn't really even Santana (IMO). I'm doing triple duty
(vocals/congas/timbales) and one of the guitarists plays guiro on Oye Como
Va, but we're still short bongos, cowbell, and at least a shaker and clave
guy to do it justice. Like I said, any comments are appreciated, the crowd
is usually drunk enough not to notice anything less than a massive glaring
error, so the feedback I get from people who know how to listen and are
actually listening is limited. Anyway, I'm never offended by an actual
critique, it's just the "you guys suck" kind of thing I was trying to avoid,
because it isn't useful.

thanks for the feedback,
ryanm

[Arny Krueger]

"ryanm" > wrote in message


> "Arny Krueger" > wrote in message
> ...

>> First off, I like the playing.

>> Using the CEP/Audition parametric eq on Crossroads:

>> lo shelf f=20 Hz +13 dB
>> low notch f=30 Hz Q=2 -14 dB
>> high notch f=11 KHz Q=3 -12 dB
>> high shelf f=18 KHz +12 dB

>> In short the SM57 kinda screwed the cymbals, and there was probably
>> some kind of a room resonance affecting the low end.

> I don't use CEP/Audition, I have the Sonic Foundry tools, so I
> tried several different parametric eq plugs that I have (4 of the
> built in SF parametric EQs stacked and the Waves REQ 4 band), but
> they both gave different results, so I don't know if I'm hearing it
> how you intended, but I think I get the idea. The notches to remove
> the problem freqs (high for the bright cymbals, low for the "boomy"
> bass) and the shelves to bring out the outer freqs that weren't
> causing the problem.

I think you picked up on exactly what I was trying to say. Sounds better
doesn't it?

BTW, I went away and came back and did some more tuning and it sounded even
nicer, but in the end that's your job, not mine!

;-)

My hope was that parametric equalizers are similar enough across different
editing products so you could work with what you had. I had my hardware
parametric in mind as a target. It appears that SF & CEP were similar
enough so that with some insight and creativity on your part, the
information made its way through.

> Also, you didn't mention the rolloff on the
> shelves, but I tried everything between 2 and 10 octaves.

Shouldn't really matter since the areas I was trying to push back up are
near the edges of the normal audible range.

[Peter Larsen]

Sugarite wrote:
>
> > Sugarite wrote:
> >
> > > Here's what I came up with:
> > > http://home.cogeco.ca/~zigakly/crossroadstweaked.mp3
> >
> > not a valid file.
>
> Works here, try Quicktime

Did you try it off of the site, CE2k asks what the sampling rate is and
Winamp just gurgles. It is not a valid mp3 file. Quicktime says: "could
not open the file because it is not a file quicktime understands", i. e.
it too thinks that it is not a valid file.


Kind regards

Peter Larsen

--
************************************************** ***********
* My site is at: http://www.muyiovatki.dk *
************************************************** ***********

[Peter Larsen]

Arny Krueger wrote:

> >> lo shelf f=20 Hz +13 dB
> >> low notch f=30 Hz Q=2 -14 dB
> >> high notch f=11 KHz Q=3 -12 dB
> >> high shelf f=18 KHz +12 dB


I did something somewhat different. Listen at:
www.muyiovatki.dk/~mastering/rap/gt-tricycle.mp3

comments appreciated. The file will only stay up a couple of days &
there is no link to it.


Kind regards

Peter Larsen





--
************************************************** ***********
* My site is at: http://www.muyiovatki.dk *
************************************************** ***********

[Peter Larsen]

Peter Larsen wrote:
>
> Sugarite wrote:
>
> > Something is making the treble pretty brittle, probably one or all of three
> > things:
>
> > - the treble horns on the PA (definitely a contributor)
>
> agree

I no longer agree, the horns are a problem, but only when stressed with
vox and then at 5 kHz, I don't think on more listening that much - if
any - of the treble comes from them.


Kind regards

Peter Larsen


--
************************************************** ***********
* My site is at: http://www.muyiovatki.dk *
************************************************** ***********

[Arny Krueger]

"Peter Larsen" > wrote in message

> Arny Krueger wrote:
>
>>>> lo shelf f=20 Hz +13 dB
>>>> low notch f=30 Hz Q=2 -14 dB
>>>> high notch f=11 KHz Q=3 -12 dB
>>>> high shelf f=18 KHz +12 dB
>
>
> I did something somewhat different. Listen at:
> www.muyiovatki.dk/~mastering/rap/gt-tricycle.mp3
>
> comments appreciated. The file will only stay up a couple of days &
> there is no link to it.

Certainly a different direction. I'm wondering if the differences are
partially based on the reproducers we're monitoring with. If you told me
you're monitoring on some European phones I've heard I wouldn't be
surprised at all. I'll admit it, I'm committing the grave sin of using 7506s
for mastering. But this is for fun, right?

After listening to your version Peter, I went back and revisited Black Magic
Woman.

I ended up modifying the recipe above, adding a 10 dB peak at 440 Hz, Q=1
and a 15 dB notch at 60 Hz Q=2. At this point I'm consciously capitalizing
on the fact that the instruments and vocal that sound too distant to me,
have most of their energy in that range. The 60 Hz dip keeps the broad 440
Hz peak from adding tubbiness.

BTW for people who are falling off their chairs when they look at the
relatively large number of dBs of boost and cut, please understand that
there's a lot of cancellation going on and the resultant FR curve varies by
only a few dB (max = 8 dB @ 11 KHz) from flat. That's life with parametrics.

What I really want to do is remix the piece and maybe remic the tracking.

;-)

BTW I've put a DVD burner in my remote recording computer, so I now bring
home all 12 tracks on one little piece of optical media. I can then edit and
remix to my heart's content at home or with any computer with CEP, a DVD
drive, and a stereo sound card.

In my wildest dreams, we'd be able get the tracks of our fave songs
delivered this way so we could have things more "our way". It would be fun
to try some roll-my-own 6.1 surround mixing.

[Peter Larsen]

Arny Krueger wrote:

> > I did something somewhat different. Listen at:
> > www.muyiovatki.dk/~mastering/rap/gt-tricycle.mp3
> >
> > comments appreciated. The file will only stay up a couple of days &
> > there is no link to it.

> Certainly a different direction. I'm wondering if the differences are
> partially based on the reproducers we're monitoring with. If you told me
> you're monitoring on some European phones I've heard I wouldn't be
> surprised at all.

Kef Coda anno 1976 augmented by a pair of ATC 9" studio bass units in
4th. order reflex bins tucked away in my computer corner. It was almost
too bright afterwards, so I checked it on my L100's - if in doubt, they
decide and they decided OK and said "whopping great bass range". I
didn't test on my large speakers.

> I'll admit it, I'm committing the grave sin of using 7506s
> for mastering. But this is for fun, right?

It is indeed fun, but not just for fun ... I don't mind the advertizing
aspect of it .... O;-)

> After listening to your version Peter, I went back and revisited Black Magic
> Woman.

The most severe problem with this recording is that there is something
strange about the 220 Hz region, there is a narrow peak there that
suggests that the PA has - ahem - cooperated with a room resonance. Also
the guitar sounds strange, it could be a conflict between an open back
cabinet and reinforcement.

Headphone listening doesn't reveal this, the pressing for the ears when
listening to one of my loudspeaker sets did and analysis confirmed it.
Audition is busy so I haven't checked the phase-image, but something is
screwy in that range.

> I ended up modifying the recipe above, adding a 10 dB peak at 440 Hz, Q=1
> and a 15 dB notch at 60 Hz Q=2. At this point I'm consciously capitalizing
> on the fact that the instruments and vocal that sound too distant to me,
> have most of their energy in that range. The 60 Hz dip keeps the broad 440
> Hz peak from adding tubbiness.

Yes, the "mid" midrange is lacking a bit, possibly something with 15" +
horns (??), possibly just the instruments or the room or the way the
stage works or perhaps even simply audience absorption.


Kind regards

Peter Larsen




--
************************************************** ***********
* My site is at: http://www.muyiovatki.dk *
************************************************** ***********