Sanken Css-5 : Normal Mode Ambience Recording Impressions

Introduction

In this blog post, we'll take a look at how the Sanken Css-5 performs as a tool for recording ambiences in its "Normal Mode" configuration. We will examine select recordings and their spectrograms to better gauge this microphones ability and/or inability to perform sufficiently in certain outlined criteria. Key points of interest in this analysis are the microphones Signal to Noise Ratio, Self Noise, Frequency Response, and the width of it's conveyed Stereo Image. The goal of this inquiry is to determine whether or not the Css-5, in its Normal Mode, is suitable for recording ambiences. This blog post is the companion piece to our recent video which can be found here https://www.youtube.com/watch?v=Mki9PR8om64

Setting the Stage

The Sanken Css-5 is a Mono-Stereo switchable shotgun microphone. Developed by Sanken Microphones, in conjunction with Japan's national broadcaster, NHK. With film and television use in mind, the Css-5 was designed to capture on-axis sounds clearly, "accentuating the frontal image while the side signals are positioned slightly behind." It's 5 capsule design, with 3 capsules being employed in its mono mode, aid in reducing reflections in reverberant spaces. Some even compare its ability to the likes of the popular Sanken Cs-3e. Though, we will further investigate its mono capabilities and use as a boom mic in a future blog post.

Aditional information on this mic's specs can be found here https://www.sankenmicrophones.com/production/shotgun/css-5/

Signal to Noise Ratio & Self Noise

A microphones SNR is often a determining factor for its eligibility as an ambience recording microphone. Sanken's Website states that the
Css-5 has a self noise level of 18dB-A, theoretically being equal to an SNR of 76dB-A. The popular Audio Technica BP4025 stereo microphone has an SNR of 80dB-A or a self noise level of 14dB-A, and for good measure, the Sennheiser MKH 416 has a self noise of 13dB-A. So the Css-5 is quite a bit louder than both those microphones, but does that difference even matter? In most cases there shouldn't be an audible difference.

Let's take a look and listen at our first recording.

With the gain on my Mixpre 6 II set to 40dB, recording at 24bit 48kHz, I captured this quiet suburban scene with distant chatter and a couple of passing cars. Below, the spectrogram from Izotope's Audio Editor displays the amplitude of all frequencies captured in the recording above.

You can click on the image to get a better view.

Noise created by a microphone often manifests itself in the higher frequency ranges of recordings, commonly described as a "hissing sound". As can be seen in the image and heard in the recording above, the "hiss" is nowhere to be found. Even with peaks normalized to -3dB, the Sanken's SNR is still silent enough for use in recording quiet suburban ambiences and possibly even quieter ambiences.

40dB - 43dB of gain seems to be the sweet spot for my particular microphone in Normal Mode, with 45dB of gain being the outer limit before recordings become too noisy for my tastes. That being said, If I had to score this Sanken Css-5's Normal Mode SNR, I would give it a 8/10 for what it accomplishes, given its age and form factor.

Radio Frequency Interference

Keen eyes may have noticed streaks present in both channels on spectrogram. I suspect this to be radio frequency interference from the nearby homes coming into contact with poorly shielded XLR cables. I don't believe the Css-5 to be the culprit in this instance, though, I would need to do more testing. Regardless, I find RFI with the Css-5 to be a rare occurrence and an overall non-issue when it comes to most recording scenarios.

Frequency Response

The Sanken's frequency response Is certainly not as wide as many of today's commonly used microphones, the new standard seeming to be 20Hz-20kHz, but that doesn't mean it captures a lesser recording. In fact, I believe the Css-5's frequency response and curve helps waive the need to EQ recordings in post. Often, I find that ambience recordings made on my Clippy EM272's are too hot in the lows and don't always sound pleasant or faithful to how I remember the scene, so I usually end up adding a high pass to roll off those sub 100Hz frequencies.

This isn't to say all ambiences sound better with attenuated low end, It's simply a preference that I've grown accustomed to while recording in my immediate environment. Sanken states that this microphone's Frequency response is 100Hz-15kHz. Not a particularly spectacular frequency response when looking only at the numbers but it really isn't as bad as it seems.

Neighbourhood ambience recording from the Css-5. Gain set to 40dB with no post processing.

The frequency response curve of the Css-5.

Or maybe it is. This kind of curve doesn't necessarily inspire confidence at first glance, but in the world of sound, it's not always about looks! Take a listen to this sample of creaking trees on the edge of a ravine.

These scenes usually carry lots of low frequency energy due to strong winds, but this microphone, in conjunction with its wind protection do a good job at preserving the integrity of the soundscape while lessening the rumble of the wind. Those with preceptive ears my have also heard the very slight sound of distant traffic. The Css-5 does a great job at attenuating that distant traffic rumble right off the bat! So I'm a fan of it for that reason as well.

Taking a look at the spectrogram for this recording, we'll see that the drop off only really starts to pick up around the 70Hz mark.

Ravine ambience recording from the Css-5. Gain set to 40dB. +3dB in post to accentuat low frequencies.

I would consider the Css-5's frequency response to be more than adequate for ambience recording, especially in urbanized environments where you might want to lessen the rumble of distant or nearby traffic.

Stereo Image

Normal Mode operates at a stereo angle of 120°, with a stereo image closely resembling an XY miking. The phantom center of the Normal Mode is quite pronounced, reflecting the broadcast focused design of the microphone. ENG applications seem to be where this mode accels. For ambience recordings however, I find the spacing to be a little narrow for my tastes. I strongly dislike XY arrays for there narrow stereo fields and the Css-5, when recording ambience in this mode, leaves lots to be desired on this front. Not to say that it is particularly bad in any way; my personal tastes simply lean more towards favoring ORTF and AB set ups.

Take a listen to the birds in this forest. Does the stereo image feel too narrow to you?

Just look at the polar pattern, the lobe is colossal and the angle just doesn't feel like it's enough to capture an ambience in a faithful manner. This is almost certainly due to its prioritization of maintaining a strong phantom center. The fact is: The Css-5's Normal Mode wasn't made to record ambience, it was made to record stereo.

All ambience is enhanced by stereo but not all stereo enhances ambience.

This is one of those cases. But it's not all doom and gloom for the Css-5. It's a switchable microphone with 3 modes after all! Sanken's stereo saving grace could lie only a-third of a turn of a dial away! Say that five times fast.

It's Wide Mode is advertised as the mode for "cinematic ambiance and stereo sound effects" with its gloriously wide 140° stereo angle! Surely this mode will provide the same quality of recording without any drawbacks, right? Well, we'll just have to test and see in a future blog post!

Final Thoughts

The Sanken Css-5, in its Normal Mode configuration is suitable for ambience recording but not suited for them. While the Signal to Noise Ratio is adequate at providing recordings free from audible noise and it's Frequency Response inherently attenuating sometimes overly present low end frequencies, It under-delivers in defining an immersive soundscape due to its narrow stereo width and overly present phantom center.
But alas, these are simply but my own sentiments derived from objective inquiry. Sound, like art, will forever be shrouded in subjectivity. And for that reason I say, like most others do: "if it sounds good, it is good."

If you like how the Sanken Css-5 sounds in the examples above, or in other examples you find online and you see value in its Normal Mode or switchability, perhaps this mic is for you.

Personally, I love my Css-5 because of its versatility. It's Normal Mode just wouldn't always be my first choice when recording ambience.