Attending a Mahler 3rd Recording Session: An Audiophile’s Perspective

Audio: Listen to this article.

Attending a Mahler 3rd Recording Session:
An Audiophile’s Perspective

Rajiv Arora

Serendipity. That word was on my mind as Chris @The Computer Audiophile and I trudged to the stage entrance of Orchestra Hall in Minneapolis, that wintry Tuesday, to observe a recording session of Mahler’s 3rd Symphony by Osmo Vänskä and the Minnesota Orchestra. 

Serendipity had led us to that moment. When Chris and I decided on a date for my visit to his amazing immersive audio setup, which I described in this article, I did what I usually do when I travel – I checked to see if there were any concerts in the area. To my delight, I discovered that Osmo Vänskä and the Minnesota Orchestra were performing Mahler’s 3rd symphony the weekend of my visit. I’ve been a huge fan of Vänskä’s Mahler and Sibelius recordings on the BIS label, but had never attended one of his performances in person. We quickly made arrangements to attend. But what’s this? The concert web page mentioned that the orchestra would be recording the symphony with BIS the following week. Hmm! 

 I had always wanted to observe a classical recording in person, and get a much closer look at the entire technical process from recording to release. I had been wanting to write an article describing this experience from the perspective of an audiophile. Here was a golden opportunity, assuming we could get permission from the various stakeholders to observe a recording session. Luckily, once we set the wheels in motion, we were kindly granted access.

Our thanks go to the many people who helped us arrange our recording session visit, starting with "BIS boss" Robert von Bahr in Sweden, who put us in touch with the producer, Rob Suff. Rob could not have been more gracious, and he worked behind the scenes to get approval from Osmo Vänskä, as well as the Minnesota Orchestra management. We thank them all for allowing us access to what turned out to be an absolutely fascinating and memorable experience.

The Experience

Upon arrival at the stage entrance, we were buzzed in and asked to wait until Rob could come get us. While we waited, members of the orchestra streamed in wearing their street clothes, chatting and joking like any coworkers do. I must say, it was a bit surreal for me. I only ever see these accomplished performers dressed up and on stage, so to see them as regular people was fun. Several even smiled and nodded at us, which was thrilling. I was a bit starstruck!

Soon, Rob came by, and amiably chatted with us for a few minutes. He gave us a quick rundown of what we would experience, before walking us through what seemed like a maze backstage, to arrive at the recording “studio.” I say that in quotes because the studio was the repurposed ready room, where performers typically assemble before going onstage. While it looked makeshift, it was meticulously set up, as we learned.

For this occasion, it was arranged with two rows of tables. The first row was where Rob sat, alongside the recording engineer, Marion Schwebel, with her various screens and devices. At his station, Rob had a microphone with which he could communicate with the performers onstage. Directly in front was a large TV monitor with different views of the performers. A pair of speakers were available for playing back snippets for review, but were usually not engaged. Finally, for his inspiration, Rob told us, there was a large portrait of Gustav Mahler himself, conducting the Vienna Philharmonic.

Another row of tables behind Rob and Marion was set up with listening stations, connected to a collection of Sennheiser HD600 and HD650 headphones. We were invited to sit here and listen in on the headphones (more on this later).

These listening stations were actually there for the conductor and performers to come in during breaks to hear snippets and confer with Rob. During these times, we would vacate these chairs for their use.

The Recording Process

As we learned over the course of the afternoon, these recording sessions are on a very tight schedule. There was an overall deadline to record the entire symphony over the course of the week. Each day was divided into several recording sessions, and each was tightly managed. For example, union rules mandated that musicians get 20 minutes of break time every hour. Session start and end times were also rigidly enforced. As a result, efficiency and speed were of the essence, but without compromising artistic quality. This is where the experience and expertise of the producer comes in.

The Producer

Rob Suff, the recording producer for this project, has been a producer for over 30 years with the Swedish label BIS Records, and has collaborated on over 60 recordings with Osmo Vänskä. He is also the Artist and Repertoire Director for BIS Records.

Before this session, I really had no idea what a producer on a classical recording does. It turns out, they do a hell of a lot. They are effectively running the whole recording session. First, they manage the time within the rigid constraints required. This alone can be challenging, and yet this is the simplest part. 

The producer also has significant artistic control. While the conductor is inarguably the artistic director of the orchestra, and creates the performance they want, the producer is working to extract the best from the musicians to create the best recorded product. As you can imagine, there is a lot of overlap between these imperatives, and ample opportunity for disagreement. Yet, I saw none of that, which is a result of the level of trust these two men have built, collaborating for over 3 decades on a multitude of recordings, dating back to Vänskä’s early BIS recordings with the Lahti Symphony Orchestra.

To illustrate, let’s look at how the recording session progressed. Rob chose to divide each hour into a full 40 minutes of recording activity, followed by a 20 minute break. The recording activity would comprise a sequence of discussion/take/feedback cycles. These takes could be from a few minutes to a few seconds long, depending on the purpose. Typically a take would be stopped either by the conductor or producer. This is where the artistic collaboration would come in. In some instances, Vänskä would want to fine tune a section with his musicians. In others, it would be Rob who called a halt. His feedback could be recording related: “2 before 39, the trumpet could be a little louder, the 16th note is not coming through clearly,” while on other occasions, purely artistic. For example, at the start of the afternoon, there were some patches being completed on the 4th movement. At one point, Rob wanted to redo a section where he wanted mezzo-soprano Jennifer Johnston to sing the tief in the line “Tief ist ihr Weh” more expressively. This was accepted without fuss.

During the breaks, Rob and Vänskä would confer and map out changes. Occasionally, they and several of the performers would listen to takes at the listening stations.

This interplay was fascinating to watch. Rob is clearly a deep – or should I say tief! – musical scholar. His main tool was the score of the symphony, on which he was making notations in real time during the takes, and then using those to provide feedback after the take. The efficiency and willingness with which this feedback was accepted highlighted the experience and trust Rob has built with Vänskä and the members of the orchestra. 

Finally, to hear the immediate result in the next take was really quite thrilling.

The Recording Engineer

Sitting alongside Rob was recording engineer and tonmeister Marion Schwebel. Marion is a co-founder of Take 5 Music Productions, a recording and music production company that works with many record labels on classical music recording and crossover productions.

Marion presided over her screens to  ensure flawless recording of all the microphone feeds — as many as 62 of them! Her interactions with Rob had a practiced ease about them, in the manner of professionals who have worked together many times. In addition to managing the recording, she could quickly cue up a take for review by Rob, the conductor or a group of the musicians.

Of course, a lot of her work was done ahead of time. She spent many hours on the Sunday before the start of recording week in the hall to set up the microphones, and run the wiring for recording and monitoring. I had so many technical questions for Marion, which were impossible to ask while Marion was so busy with her actual job of recording! Fortunately, in a series of followup emails, Marion was kind enough to give me a ton more details about the technical setup, which I’ll describe in the appendix.

The Listening Mix 

As I was expecting, recording a work like Mahler 3rd, with a massive orchestra, a soloist, a women’s chorale, and a boys chorus, takes a lot of microphones! Each microphone feed goes through A/D conversion, and the resulting digital stream gets stored for later post-production work — editing, mixing, and mastering. So what 2-channel mix were we listening to during the session? This was the listening mix. As Marion explains:

The mix we listened to on headphones was a listening mix with the intention to get rather close to the final mix; I created it in the first 10-15 min of the first session on Monday while the orchestra was rehearsing. From that moment on, I optimise the mix constantly, adapting a little between the different movements. I store the mixing snapshots from the session - each movement has a different snapshot. This snapshot will be used later for the final mixdown as starting point.

Listening Sound Quality

Since Chris and I were not in the hall itself, all our listening was done on listening stations that Marion had set up for the purpose.

The headphones that were set up were a collection of Sennheiser HD600 and HD650’s. 
 

I used to own an HD650 for several years, so I thought I knew what I was going to hear. Boy, was I wrong!

As the session got underway, and the orchestra started playing, Chris and I both looked at each other in amazement. The sound quality of this live listening mix was just extraordinary! Chris’s expression says it all.

I know the Sennheiser HD650 well. Even on an excellent amp, with an excellent source, they can sound a bit closed in and wooly. Yet here, with this live mix, even on a very basic DAC and headphone amp, the level of clarity and palpable ambience was almost unbelievable. Garbage in/garbage out is an audiophone truism; it’s the principle that the quality of the recording plays a decisive role in the sound quality. Well, here it was being borne out! If I could have walked out of this session with a finished recording of the quality of this listening mix, I would have been a very happy man.

Summary

Before we knew it, our afternoon recording session had come to an end, and it was time to say our goodbyes. Despite the intensity and concentration that Rob and Marion had brought to bear during this session, they somehow found the energy and grace to chat with us a bit longer, and patiently answer our questions.

We had an interesting discussion about whether BIS was considering immersive audio formats like Dolby Atmos or Auro-3D, beyond the “traditional” 5.1 surround format. At this time, Rob said there are no plans, but it will be interesting to see if this shifts as the popularity of Atmos music grows.

I think I speak for Chris as well when I say this recording session was a real eye-opener. As audiophiles, we only hear and judge the end product: the released mastering of a recording project. We view these as immutable because by the time they reach us, they are. But in fact, what we are hearing is the product of a series of artistic and technical choices, accompanied by an incredible amount of hard work of many, many dedicated people.

Going into this, I had no idea exactly how a classical recording was made. This session taught me so much about the creative and logistical process that brings the concept to reality. I am extremely grateful to Rob, Marion, Osmo Vänskä, and the management at the Minnesota Orchestra and BIS for giving us this opportunity to learn.

Of course, this project is not complete. The interim output of these recording sessions is a set of raw data files, notes, and annotations that will form the input to the post-production phase. This is where editing, mixing, and mastering will happen. I am hoping that we can come back with a followup article detailing this phase.

Appendix: Technical Details

In a series of followup emails, Marion sent me technical details of the recording setup, which I think most of our readers will find very informative. Her responses are almost entirely her words, with only very light editing from me for clarity. I would like to thank Marion for providing us such detailed technical insights into the process. We are very grateful!

Microphones and Placement

Please describe the microphone setup for the recording.

• In total we had 62 microphones for the maximum set-up
• For the orchestra I used 50 microphones 
• for the vocal soloist for movement 4+5 I used 2 microphones (stereo pair) 
• 10 microphones were used for only movement 5, which involved a children’s choir (4 mics) and a women’s choir (6 mics)

Please describe the types of microphones, and where they were used.

• Omnidirectional
  • Two pairs in AB setting:
    • one pair of Neumann TLM50 width 4,5 m
    • one pair of Neumann KM130 width 2,5m
• Wide cardioid
  • I use them on violins and violas (Neumann KM143), for the 2 choirs (KM143), the vocal soloist (Sennheiser MKH800), the harps (Schoeps MKH21) and some brass instruments (Schoeps MKH21)
• Narrow cardioid
  • On celli, double bass, all woodwinds, horns, timpani, percussion, and some brass instruments (Neumann KM184, DPA 4011)

Please explain the purpose of each of these microphone types.

• Omnis: placed with some distance to the orchestra; they capture the hall, have a big depth and are open in sound. As a pair they capture the actual balance and the sound color of the hall
• Wide cardioids: I use the wider angle to capture several players with one microphone (for example a group of 4 violins playing on two desks or several choir singers); they give a wider, more natural sound than cardioids. Works great as a stereo pair for soloists also
• Cardioids: give good focus and less leakage than wide cardioids; you get some grip on the instrument. Cardioids sound less closed and more true to the instrument than e.g. hypercardiods or figure eight

Please elaborate on the exact microphone brand/model used, and why you like them.

• Neumann KM series: sound rather brilliant with a boost at around 8500Hz. I like them particularly on strings and percussion
• Neumann TLM50 are warm, detailed and rich. They direct a bit more than other omnis towards the high register
• DPA: neutral, clear with a very even frequency response; good choice as both spot mic and main mics. They work particularly well for woodwinds, piano, percussion. I use the omni for AB setting when knowing that the hall is good enough
• Schoeps: a bit warmer sound, silky, melt nicely together 
• Sennheiser MKH800 - switchable characteristics; great sound quality with good depth

ADC & Mic Preamps

What ADC (analog to digital converter) brand/model is used to convert each microphone's input to digital? Why did you select these particular ADC’s?

• For many years, we have used Preamplifiers and ADC by RME - ‘Micstasy’ and ‘Octamic XTC’. The sound of the units is clear, detailed and rich. They are reliable, light and easy to link together via MADI or Ethernet. We use them in a modular system defined by the amount of needed channels

Cables and Cabling

Do you use any particular brand or type of cabling, and why?

• We use professional high quality cables from different brands (Mogami, Neumann, Cordial; Neutrik plugs)
• The main line between stage (where the PreAmps are located) and control room is a Neutrik Quad Fiber-optical cable for all 64 channels

Recording Format and Software

What sample rate and bit depth does the ADC output?

• 96kHz and 24-bit integer; editing is done on 32-bit floating point

What software do you use to record the 62 streams?

• We use ‘Samplitude Pro’ by Magix for recording and do the editing on the ‘sister program’ Sequoia’ by Magix
• Sequoia is a great program especially for post-production of classical music. The editing crossfade window allows very advanced crossfade handling. In recording of acoustical classical music, there is a considerable amount of leakage between the microphones. Therefore the tracks cannot be edited and shifted individually but need to be handled at the same time (with exceptions of course). Sequoias edit crossfades work for most complex assembling issues
• ‘Samplitude/Sequoia’ is used by Take5/BIS Records on more than 20 systems. It is installed on recording equipment, our individual editing suites and the mixing studios. We take advantage of using only one edit platform when transferring projects at different project states between the colleagues

Data Storage and Resilience

What storage medium does all the data from the recording session get stored in?

• SSD for main computer, HDD for backup computer

Do you use redundancy like disk mirroring or RAID?

• Not during the session since we record two independent copies on two computers (main and backup). After each 3-hour session, a clone of the project data is done. On top, Rob makes a clone copy for his own listening purposes and multiple backups are done ‘back home'

Please describe your technique for recording 2 parallel streams to achieve additional redundancy.

• We use one Quad MADI line from the stage to the control room which feeds all channels into the AudioInterface of the main computer (MADIFace XT by RME). Then, I patch through all channels to a similar backup interface which connects to the backup computer. Both machines run independently, so in the worst case - if one computer refuses to record or hangs up itself - the other computer would not be affected; this ensures a safety net for all recording data

Notations and Organization

Since patches or snippets are recorded multiple times until Rob and/or Osmo were happy, how do you tag which patches were the "approved" ones? Is there a way to put bookmarks or some similar metadata during recording, to indicate this?

• All relevant notes are written directly into the producer's score. Each take gets  a marker in the recording software which corresponds to the producer's notes. We work with different ’project sessions’ for each movement so the take numbers don’t get too big
• Rob would annotate notes in his score according to his individual system. For example ‘+’ for good or ‘-‘ for mistake or bad, ‘i’ for intonation issues, special shortcuts for balance and ensemble remarks. Each producer has a slightly different system of annotation. The notes help to remember the issues during the session and to choose the best takes later during the mark-up stage
• During the session, the producer addresses some spots based on the annotations before playing the next take

Post-Production

After the recording session, what happens next and when?

• Within a few months Rob will listen to the material, choose the best takes  and write an edit plan directly into the score. Then the assembling will be done by a colleague following the mark-up plan 
• In the next step, Rob checks the edit and makes a few corrections, then sends out a ‘first edit’  to Osmo. In the next round, Osmo is commenting on general observations, balance issues and musical questions. There can be several ‘rounds’ of correction edits, depending on the complexity of the project and the artists
• After the producer and conductor have agreed on the final state of the edit and after having done all corrections, the last refinements of both stereo and surround Mixdown is done by the responsible Tonmeister/sound engineer (in case of Mahler 3 by me)
• Mastering is often handed over to an in-house mastering engineer

About the Author

Rajiv Arora — a.k.a. @austinpop — is both a computer geek and a lifelong audiophile. He doesn’t work much, but when he does, it’s as a consultant in the computer industry. Having retired from a corporate career as a researcher, technologist and executive, he now combines his passion for music and audio gear with his computer skills and his love of writing to author reviews and articles about high-end audio.

He  has "a special set of skills" that help him bring technical perspective to the audio hobby. No, they do not involve kicking criminal ass in exotic foreign locales! Starting with his Ph.D. research on computer networks, and extending over his professional career, his area of expertise is the performance and scalability of distributed computing systems. Tuning and optimization are in his blood. He is guided by the scientific method and robust experimental design. That said, he trusts his ears, and how a system or component sounds is always the final determinant in his findings. He does not need every audio effect to be measurable, as long as it is consistently audible.
 
Finally, he believes in integrity, honesty, civility and community, and this is what he strives to bring to every interaction, both as an author and as a forum contributor.