Taperssection.com

Gear / Technical Help => Recording Gear => Topic started by: Carrera2 on February 24, 2007, 06:26:05 PM

Title: Setting levels when wide dynamics expected
Post by: Carrera2 on February 24, 2007, 06:26:05 PM

Last night I recorded a strong orchestra from a good spot just left of center (my season tickets), in about the fifteenth row.  To be discrete, and never having taken these things out in the field, I tried running SP-TFB-2 in-ear mics with low gain on an R-09.  I clearly underestimated the gain setting that I should have used, and ended up boosting gain a bit more as each of the four pieces were performed.  However, even with the sometimes loud Brahams Symphony No. 4, I underestimated the gain.  In this case, the mics are seriously unmatched, I have now discovered also.  I did manage to normalize the channels and boost the levels so that it isn't a half bad recording.

All that said, how do you go about setting levels when you know you may have a performance with substantial dynamic range?  In this instance, experience will be something of a guide since I will return.  If you decide gain is too low (or too high) during a recording, will you adjust it while rolling?  Any other thoughts?

Thanks.


 
Title: Re: Setting levels when wide dynamics expected
Post by: Church-Audio on February 24, 2007, 07:43:08 PM

Last night I recorded a strong orchestra from a good spot just left of center (my season tickets), in about the fifteenth row.  To be discrete, and never having taken these things out in the field, I tried running SP-TFB-2 in-ear mics with low gain on an R-09.  I clearly underestimated the gain setting that I should have used, and ended up boosting gain a bit more as each of the four pieces were performed.  However, even with the sometimes loud Brahams Symphony No. 4, I underestimated the gain.  In this case, the mics are seriously unmatched, I have now discovered also.  I did manage to normalize the channels and boost the levels so that it isn't a half bad recording.

All that said, how do you go about setting levels when you know you may have a performance with substantial dynamic range?  In this instance, experience will be something of a guide since I will return.  If you decide gain is too low (or too high) during a recording, will you adjust it while rolling?  Any other thoughts?

Thanks.


 

That is a very hard question to answer. Let me try.

Setting gain is a product of distortion as we all know. If its too high you over load if its too low your noise floor go's bonkers. I think when your in this type of situation you need to know how quiet you can go before noise is a real issue. One of the ways I would do this is to simply sit in a room. A quiet room and set your levels with some soft music in the background. See how low you can set your recorder before noise is a real issue. You will then have to dump the files onto your computer and boost them. Then you will get a feel for the minimum level you can safely run before noise is becomes an issue.

You can by contrast do the same kind of test for maximum volume that your gear can handle before distortion. So that you have a "ball park" idea about the limits of your gear. As stupid as it sounds make notes as to what settings you set your levels too. So you can reset these levels at show time.

That's the first part of the problem

The second is knowing how much headroom you have at any one setting. That's almost impossible to do with out knowing your gear. Its very hard to record shows why the dynamic range is not known and do it with a small recording device. I would use your external preamp so that you can "stage the gain" have the preamp you own do half of the gain and your recording device do the other half. That way your splitting the headroom up and less likely to overload any one "stage" You still might have to adjust levels but at least you should have more to work with and less noise if you do need to boost later on. Your working with some of the hardest material to tape Classical music. Its not like rock and roll where the dynamics are pretty much fixed. With classical you can have 60db levels all the way up to 105db levels in one song! OUCH. That's a lot for any preamp to handle. and still get good levels to tape, try the idea I suggested about sitting around in a quiet room with some music on to get a feel of your noise threshold. Then at least you will know what the lowest level is you can set before you noise floor gets out of hand.

There are lots of guys around here Like Moke that do a lot of the type of recording your doing. They only difference is they are using very high end preamps and size is not an issue, regardless Moke really knows this type of recording I am sure he will chime in with a few ideas of his own.

Title: Re: Setting levels when wide dynamics expected
Post by: Church-Audio on February 24, 2007, 09:15:55 PM
clap it out at home.

Set the mics up, and clap a single sharp clap about 2' from the mic. Then repeat it until you get that sharp peak to just nudge the OVER light.
The audience is going to be your loudest influence, unless you've got some dream seating. So setting to the single clap can dial you in, or, at least give you a very solid starting point.
I calibrated my 4060->mma6000->r09 this way at 22.5dB gain at the pre, level 13 on the r09 line-in, and i don't even have to look at the levels, and haven't since at least october.
This puts my peaks at -8db / -6dB musically, and also, perhpas to detriment, goes over in the audience ovations, but not audibly. You can tame the clapping in post.

Great idea.. It makes perfect sense.

Title: Re: Setting levels when wide dynamics expected
Post by: SClassical on February 24, 2007, 09:26:21 PM

 If you decide gain is too low (or too high) during a recording, will you adjust it while rolling?  Any other thoughts?

Thanks.


 

You should not adjust anything while rolling - that will mess up your recording....If you think it is too low while rolling you have to go home and use an audio program to adjust it. You can adjust during intermission or between pieces.

What Moke said about clapping is good. Also quality of recording depends on where you're sitting. For best results try getting seats front row center that way the audience is behind you and the Orch is in front of you. Also you need to consider composers and the size of the hall. Your gain for a Mozart/Haydn concert is totally different from the gain for a late 19th (or early 20th) centry composer.

If I'm using the R-09 and MMA6000/DPA4060, I usually set my R-09 @8 and adjust my preamp depending on the concert. The more concerts you record the more you know your equipment and the better the result. The first few times it's always a try and erorr situation.

Good luck!
Title: Re: Setting levels when wide dynamics expected
Post by: WiFiJeff on February 25, 2007, 12:23:41 AM
Chiming in on this, I have found that trial and error (minimizing the error part as much as possible) works pretty well.  As Simon said, sitting up front maximizes your S/N ratio, but for chamber music the applause is going to be the loudest sound you get and I think it a mistake to try to keep this from overing, a distorted clipped clap is - just a clap, and the asshole whooping into your ear can be cut out from the final tape in most cases (except at the opera where they go nuts while the music is still playing, because after all no one is singing).  For DPA4060s > MMA6000 into either a D1 at 8 or a Sonosax MiniR82 at 0dB = 0 dBu, the range of settings I have found (YMMV, but the range should be comparable) is average +20dB for a piano recital up close, +22.5 to 25 dB further back in the hall, +17.5 to 20 for opera front row, +15 or 12.5 for the loudest orchestral music (Bruckner, Mahler) right up front.  With those settings I sometimes have to boost a bit in Wavelab (at 24 bits not a major issue with this gear), and sometimes I come close to running too hot, but no tapes wrecked by clipping lately.  Definitely set and forget, since playing around with equipment is the leading cause of getting bounced from halls ("for life," well not really) especially if you follow the advice to sit up front!

Jeff
Title: Re: Setting levels when wide dynamics expected
Post by: ShawnF on February 26, 2007, 10:44:17 AM
While the sitting up front idea makes sense from a reducing the audience noise point of view, I'm pretty sure that's not going to give you ideal sound from the orchestra in a large hall (assuming they are on a raised stage directly in front of you).  If you're in the front row and can't see anything but the first row of strings, that's all your mics can "see" too, and you're not going to have any direct sound from any of the other instruments.  It could still work, I suppose, but that doesn't agree with what I've been taught about the best place to hear an orchestral concert.
Title: Re: Setting levels when wide dynamics expected
Post by: WiFiJeff on February 26, 2007, 11:05:06 AM
While the sitting up front idea makes sense from a reducing the audience noise point of view, I'm pretty sure that's not going to give you ideal sound from the orchestra in a large hall (assuming they are on a raised stage directly in front of you).  If you're in the front row and can't see anything but the first row of strings, that's all your mics can "see" too, and you're not going to have any direct sound from any of the other instruments.  It could still work, I suppose, but that doesn't agree with what I've been taught about the best place to hear an orchestral concert.

Been doing it for years.  No question that it's the best place in the house for a tape.  There is a balance issue for soloists, piano for example (but piano guys sometimes love the over-the-top piano emphasis) but others (violin) need to be taped from there.  Nobody "taught" me this, I just had a Zen awakening the first time I pulled a tape from there.  It would be better to be directly over the conductor's head, but I haven't figured out how to do this yet.  Other advantages: even a hall with lousy acoustics sounds good from the front row, most of the time.  Maybe sit third row for vocal pieces where the soloist is upfront.

Try it, you'll like it.

Jeff
Title: Re: Setting levels when wide dynamics expected
Post by: powertenor on February 27, 2007, 10:02:59 AM
Hey guys,

I'm new to this stuff.  These questions should probebly be a new topic but since there seems
to be a lot of classical recordists responding here I thought I'de give it a try.
I'm an operatic tenor and have a R-09 and Core binaurals. 
My question is, which are the absolute best stealth mics for recording live opera?
I know my old friend Andrew Richards (another tenor) is using Core's high end binaurals.
(I think I should have bought the high-ends too). Oh well.
If I want these recordings to sound best from speakers which method of recording is best?
Do I really need a pre?  Without a pre should the R-09 be set to low or high gain?

Any help would be great.

Thanks,  Ian DeNolfo
Title: Re: Setting levels when wide dynamics expected
Post by: WiFiJeff on February 27, 2007, 11:02:33 AM
Hey guys,

I'm new to this stuff.  These questions should probebly be a new topic but since there seems
to be a lot of classical recordists responding here I thought I'de give it a try.
I'm an operatic tenor and have a R-09 and Core binaurals. 
My question is, which are the absolute best stealth mics for recording live opera?
I know my old friend Andrew Richards (another tenor) is using Core's high end binaurals.
(I think I should have bought the high-ends too). Oh well.
If I want these recordings to sound best from speakers which method of recording is best?
Do I really need a pre?  Without a pre should the R-09 be set to low or high gain?

Any help would be great.

Thanks,  Ian DeNolfo

Andrew has been on the Board (saw he cancelled a performance run recently, hope he is okay).  The HEBs are DPA4060s and are fantastic, especially when used (with the original DPA microdot connectors) with the DPA MMA6000 preamp.  People who are nuts about voices sometimes find they prefer the results with cardioid mics, like the Schoeps CCM4, with active cables (to a Sonosax "Lemo" preamp that powers the caps and cable); these are way bigger and harder to wear than the DPAs!  But smaller cardioids are less satisfactory, and the orchestral sound is better with omnis.

Jeff
Title: Re: Setting levels when wide dynamics expected
Post by: Carrera2 on February 27, 2007, 01:10:43 PM

Great discussion.  Thanks to everyone! 

With season tickets in the fifteenth row of a large concert hall, I may have to look and see if I could find a front row seat from which to experiment. 

I heard something I have never before heard the other night.  After a pause between the first and second movements of Brahams Fourth, a guy yells out "right on brother" referring to the conductor. 

Check it out.

Alan
Title: Re: Setting levels when wide dynamics expected
Post by: SClassical on February 27, 2007, 02:54:53 PM
Thanks...what orchestra is that? Where did you do the recording?
Title: Re: Setting levels when wide dynamics expected
Post by: Carrera2 on February 27, 2007, 03:08:28 PM
Thanks...what orchestra is that? Where did you do the recording?

Oakland/East Bay Symphony 
http://www.oebs.org/page/feb.htm

Paramount Theatre, Oakland, California, 15th row, just left of center

Title: Re: Setting levels when wide dynamics expected
Post by: SClassical on February 27, 2007, 03:16:09 PM

I know my old friend Andrew Richards (another tenor) is using Core's high end binaurals.


I don't understand why people buy the Core's HEB without the original DPA microdot connectors - it's not as if they are cheaper (are they?). Getting the 4060s with the microdot connectors allow you to use the MMA6000 preamp.
Title: Re: Setting levels when wide dynamics expected
Post by: powertenor on February 28, 2007, 09:57:11 AM
Thanks for all the replys.
How much of a differnece will using a MMA6000 with my R09 as opposed to the R09's pre's?
If I am on a budget, shouldn't I get better mics first?  How are my Core Binaurals (HighEnd Binaurals)
comparative to other mics?

Ian
Title: Re: Setting levels when wide dynamics expected
Post by: Gutbucket on February 28, 2007, 06:39:48 PM
You'll need at least a battery box to power the 4060's (doesn't matter if they're Core HEB's or straight from DPA with the microdots), the R-09 won't power them.  Batt box comes with the Cores I believe, or a DPA one is available for the microdot ends.  If you go the standard DPA microdot termination route you can add the DPA MMA6000 pre later if you care to and will benefit from a better, cleaner preamp with more available quiet gain (if needed for something else very quiet).  I don't have a batt box so I can't compare the two, but I really like the sound of the 4060>MMA6000>R09 combination.

PM coming in a few minutes on technique.
Title: Re: Setting levels when wide dynamics expected
Post by: powertenor on March 01, 2007, 08:11:10 AM
Gutbucket,  thanks for the PM.  I tried to reply but it won't accept the code.  What's the real life advantage of the pre?
Is it for more headroom? Does it let you set the levels lower and boost later?  Why can't you do that with the R-09 mic in?

Where's the best place to pick up 4060s?

Thanks again.

Ian
Title: Re: Setting levels when wide dynamics expected
Post by: Gutbucket on March 01, 2007, 11:45:33 AM
You can adjust levels either unit and you will want to set both correctly for lowest noise and enough headroom.  The MMA6k has detented gain pots, one for each channel, 2.5 dbs each click - where the R-09 has a single control for both channels with 1 db incremets.  Like the others here, I set the MMA to where I think the level going into the R-09 is in a good middle zone (level 12 or so) and then do the fine tuning on the R-09.  I use the same 'clap it out' method Moke mentions.  Sometimes I've had to adjust the levels lower to avoid clipping and I'll do that on the R-09 because it's a single stereo adjustment, doesn't make a click noise, and is easy to do while holding the unit in you hand looking at the meters. Like the others here I rarely need to adjust now if I know what to expect from a performance.  Once you record a few things you get to know the approximate settings you'll need.

Like I said, I haven't run the 4060's without the DPA pre so I'm not certain if the sound quality would be lacking without it.  I'd like to hear a good comparison of 4060>batt box>R-09 vs. 4060>MMA6k>R-09.  Eliminating the MMA6k would half the size of the rig which would be nice, but isn't a deal breaker for me.  The R-09 pre's aren't bad and you are running through them using either the line-in or the mic-in on the R-09 (only the available gain changes).  By using the DPA pre to make most of your gain you are using a better preamp that was specifically designed to work with those mics.  Consider that the preamp costs about the same as the entire R-09 (actually a little more). If you wanted to record really quiet stuff like ambient soundscapes, I'd definitely suggest the pre since you'd need all the quiet gain you can get.  You asked 'what is the best'.  Best is hard to say as taste differs, some value smallness differently than others, some need cardioids, etc.. but this is a tried and tested combination that gets high approval by many here.  Wipes the floor with the SP-TFB-2 in-ear mics strait into the R-09 (done it) but that's not a fair comparison.  Sorry I can't give you actual specifics.  Hope that helps.
Title: Re: Setting levels when wide dynamics expected
Post by: Carrera2 on March 25, 2007, 03:07:30 PM
Well, Friday was the next monthly installment of the recording of the orchestra.  As suggested early in the thread, I used the clap test a number of times and went in with my levels set.  Refrained from riding them, in spite of the temptation.  At the end of the first piece, it was rather clear that I was still set pretty low, so I nudged it up for the second piece.  Still low, nudged again for the Rimsky-Korsakov Sheherazade, the main event.

Nailed it.  I think I have some good settings now, maybe I will back off a notch or two.  Not once did the music peak, and of course, the clapping pegged the meter, but that's not a problem that the software hard limiter or some other technique cannot tame.

Thanks to all for the suggestions.  Next month I will go in with the same settings and believe that all will be well.  It's hard to be patient with trial and error.  But it's paying off!

Thanks again.


Follow-up edit:  This time I ran AT853-omni > ST-9100 > R-09 at 24/48.
Title: Re: Setting levels when wide dynamics expected
Post by: Shawn on March 25, 2007, 07:08:43 PM
Well, Friday was the next monthly installment of the recording of the orchestra.  As suggested early in the thread, I used the clap test a number of times and went in with my levels set.  Refrained from riding them, in spite of the temptation.  At the end of the first piece, it was rather clear that I was still set pretty low, so I nudged it up for the second piece.  Still low, nudged again for the Rimsky-Korsakov Sheherazade, the main event.

Nailed it.  I think I have some good settings now, maybe I will back off a notch or two.  Not once did the music peak, and of course, the clapping pegged the meter, but that's not a problem that the the software hard limiter or some other technique cannot tame.

Thanks to all for the suggestions.  Next month I will go in with the same settings and believe that all will be well.  It's hard to be patient with trial and error.  But it's paying off!

Thanks again.


Follow-up edit:  This time I ran AT853-omni > ST-9100 > R-09 at 24/48.

It's all about knowing your gear, and it seems like you are getting there.
Title: Re: Setting levels when wide dynamics expected
Post by: libfab on May 10, 2007, 06:54:26 AM
I've been through much of the stuff on recording classical music with an R-09 and so far I found little practical info as to optimal settings for stealth recording WITHOUT an external mic or preamp.
I did numerous tests and found the following, empirical results, which confirm the more scientific approach of this site's techies:

1. Very little interest in getting an external mic. The sony ECM-907 gives poorer results than the built-ins, whether at low or high mic-in settings. Other mics I've tried did little better noise-wise. Arguably the strategy only works with preamp or costly, cumbersome high-output mics.

2. Do record at 24-bit to boost post-editing efficiency and preserve a good wave shape for pianissimo playing. I know about the 16-bit v. 24-bit heated debate, yet to my ears, recording a solo violin at 24 DOES make a difference in pianissimos, even if you convert down to 16-bit for CD-building. No AGC, no low cut.

3. Noise-wise, avoid recording "high" as much as possible. For classical music, the following guidelines seem to work OK, provided the auditorium is average size, at around row 15-20 (I personally prefer first balcony, first row to get maximum reverb):

- solo acoustic guitar : you might have to get as high as "high" 25;
- string-only chamber music : "high" about 20 ;
- chamber music mith piano : "high" 15 to 20;
- all concertos : "low" 30 ; most jazz without saxo ;
- choral works, Beeth. 9th-style, or concertos with strong percussions ; jazz with strong saxo : as low as "low" 25 is OK.

The lower the setting, the lower the hiss after postediting. Empirical spectrum analysis shows "high" settings yield 10-20 000 Hz hiss noise at around -85/95 dB post editing, depending on your (conservative) hiss reduction techniques (see below). If the orchestra input is high enough to get down to "low", the same techniques will allow -110/120 dB hiss noise in the same spectral range. This said, recording too low gives poor violin pianissimos. So the settings above are a compromise.

4. Post-editing gives surprisingly good results, owing to the fairly regular hiss pattern of the R-09.

I got best results with a little bit of fine-tuning on Audition 2. Another simpler option is DC6.
For the newcomers in the hiss reduction business, the overall principle for fine-tuning a hiss filter is : take a pianissimo reference file of 10 sec, with slow violin (or flute) decay. Test the highest floor with strong hiss reduction level (at least -25dB) to rub off the noise below. Allow for some transition bandwith, to avoid abrupt cuts in the pianissimo tails.

For "low" recording at 30 and maximal orchestra input (Prokofiev's Nievski !) I did the following :

Normalize -0,5dB > Scientific filter (Remove under -28dB) > Noise reduction (automatic profile weighed by S curve between 20Hz (100%) and 200 Hz (0%), -30 dB setting) > Hiss reduction (automatic profile flattened for lows and lower mids, floor increased by +7-8dB w.r.t. to profile, -35dB reduction level yet with 10dB transition band, precision 20, special decay rate 80, high FFT settings )> Notch filters (7165 Hz and 14330 Hz need to be reduced by 12dB, super narrow ultra quiet)

5. Stereo separation: fairly disappointing with raw sound. I use a stereo expander filter (Audition 2. setting 200) in post. Great results, yet do this after hiss reduction, as this filter boosts hiss (and dynamics). You might have to compress (or normalize down) a little before doing it, to avoid clipping as a result of the dynamics boost.

Hope this will help all R-09 fans like me !
Title: Re: Setting levels when wide dynamics expected
Post by: Humbug on May 10, 2007, 08:11:36 AM
Something that's not been asked - is it possible to open tape a classical concert? If not, what are the reasons behind this?

I have stealthed a trio (plus pianist for some parts) once, where a couple of family members were playing - levels were too low, so it's slightly hissy, but listenable. The organiser (and cellist) was thrilled to receive a copy.

Interesting thread.

Humbug
Title: Re: Setting levels when wide dynamics expected
Post by: WiFiJeff on May 10, 2007, 09:11:20 AM
Something that's not been asked - is it possible to open tape a classical concert? If not, what are the reasons behind this?

I have stealthed a trio (plus pianist for some parts) once, where a couple of family members were playing - levels were too low, so it's slightly hissy, but listenable. The organiser (and cellist) was thrilled to receive a copy.

Interesting thread.

Humbug

I open tape a weekly chamber music series, the artists always give permission and love to get tapes, the hall has no union issues as it is not primarily a concert venue.  I love this, as I get to use real gear and do sound checks to set levels properly.  That said, I am officially the archivist, I'm not sure they would welcome the distraction of more than one setup.

Most major concert halls have unions that make it expensive to tape even when the artist wants a copy.  Hall rental costs can go up 50% just to allow the artist to have someone else tape.  And some artists don't want to be recorded. I can't imagine a classical venue tolerating a tapers' section, but it's fun to dream about it.  Also, for non-amplified classical music you need to be up close, the tapers' section would have to block out everyone else, which is not cool.

Jeff
Title: Re: Setting levels when wide dynamics expected
Post by: libfab on May 11, 2007, 05:15:40 AM
WiFiJeff, you're right but I don't agree that for classical music you should be close. My experience is that the added value of live recording this type of music is natural room reverb. Electronic reverb applied to classical may be fine, but nothing beats a good auditorium. My point is: when you record too close, you may get lower hiss, but the sound is too crude (shrill highs, excess of bass) or badly balanced (too much strings, not enough winds etc.). Standing back gives you a richer sound, but comes at a cost: quite a few coughs and some post editing too.
Title: Re: Setting levels when wide dynamics expected
Post by: WiFiJeff on May 11, 2007, 10:00:11 AM
This is a disagreement among non-tapers as well.  Some prefer to sit back further, some prefer right up front.  I've migrated to the front from experience, and I notice that people who try it like it, and usually don't go back to the "hall ambiance" areas (aka the cheap seats).  The migration is mostly one-way.  Anyway, I've taped from all areas, and the tapes I like best and that get the most positive comments on the sonics are pretty much without exception from the front.

Jeff
Title: Re: Setting levels when wide dynamics expected
Post by: SClassical on May 11, 2007, 11:20:58 AM
This is a disagreement among non-tapers as well.  Some prefer to sit back further, some prefer right up front.  I've migrated to the front from experience, and I notice that people who try it like it, and usually don't go back to the "hall ambiance" areas (aka the cheap seats).  The migration is mostly one-way.  Anyway, I've taped from all areas, and the tapes I like best and that get the most positive comments on the sonics are pretty much without exception from the front.

Jeff

I used to sit at the back for another stealthing activity (that's the only place where you are safe to do it). But for recording audio I like to sit as close to the artist as possible especially recording solo or chamber. But for large orchestral stuff it is best to sit not too close because the balance of the instruments becomes too favored towards the instrument you are closest to. I recorded a few piano/violin concertos sitting right in front of the solo artist and the end results sounded a bit strange...I think you can guess how that sounded.
Title: Re: Setting levels when wide dynamics expected
Post by: libfab on May 12, 2007, 05:28:28 AM
Yes Scyue, I agree with you about violins. Sitting back is often a good option for a violin concerto. Yet it depends on the quality of the auditorium a lot. In certain areas, natural room reverb somewhat makes up for the lesser sound pressure with the distance.
The mp3pro file attached is converted from a master taped in the best venue in Paris (salle Pleyel), first balcony. I took a risk but the result is stunning, to put it like Moke. Clear to me I would not get as much reverb sitting up front. (I simply used my R-09 without any external gear. Wide dynamics settings and post processing indicated above.)
It's Mendelssohn's, I hope you'll like it.