Why does the Marantz PMD661MKIII appear to limit amplitude?

Question

I have a Sennheiser K6/ME66 microphone attached with an XLR cable to a Marantz PMD661 MKIII recorder.

I set my presets to: 16-bit/44.1khz sampling rate. Low cut off/high cut off/mic attenuation to 0db/manual recording, etc. I also have an internal battery in the sennheiser mic and phantom power is turned to off.

When I do this, I get a decent signal in my recording, but there appears to some amplitude cutoff. Almost like a limit/ceiling for the signal. Indeed, when I examine the file you can see a clear cutoff, where the audio above a certain threshhold is clipped completely. I attached a photo to show what I mean.

It’s like someone took scissors and cut the rest of my signal. I want my signal to be as “natural as possible”, so even if it goes over 0db in recording I want that to be reflected in the waveform. Very weirdly the problem gets sort of fixed when I turn low-cut on. I have no idea why this is, and it’s not a great solution since I don’t want to manipulate my signal prior to analysis.

Does anyone have any insight as to how to set up the marantz recorder so that it won’t limit my signal? If it helps, when I change the recording level, the “limit” just shifts a little but the same thing occurs.

Answer 1

i have the 661mk3. same scrissor-like limited signal effects and distortion as described(661 internal gain at -18db: brickwalling occurs at -7db with a 12mV/Pa (‑38 dBV) sensitivity condenser mic)..problem seems to be the preamps overloading(analog stage) with hot signals, with a -20db pad on the mic output signal, shouting in the mic > the marantz 661mk3 doesn't distorts and clips at 0db (digital scale)...me not like it either but donno how the other recorders(in this price range) performs -- pros: no need for expensive, high sensitivity mics :)

Answer 2

The problem is most likely to do with the input attenuation. The waveform in the OP looks like the digital converter is clipping. I recommend you attenuate the microphone input to fix this. Start at the highest level of attenuation (-18dB) and then decrease until clipping starts, then reduce back one level.

Answer 3

In digital 16bit or 24 bit recording 0dB is the maximum. There simply is no way to encode any stronger signal.

I recommend doing what I have done the last 20 years. Turn down input gain to have a healthy headroom. I aim for -6dB. I check every part of the signal chain from mic to the analog parts of the recorder to the digital parts. The ”lack of gain” is Rhen compensated in my Daw where I have tools to remove unnecessary frequencys and can decrease the dynamic range with a compressor.

Answer 4

I wanted to add something that may or may not help. I use PMD661 recorders in classes I teach. I still have a few MKI that work, some MKII, and MKIII. The MKIII work in mono mode with EV RE50 mics and with Audio Technica AT897 short shotgun mics using the same presets as the O/P uses: 16-bit/44.1khz sampling rate. Low cut off/high cut off/mic attenuation to 0db/manual recording and with Mono input.

But there is loud distortion with the AT875R when the mic is in the ideal range for field recording 4-6 inches. If I change to -18db then the distortion drops but the quality is nowhere near the AT897.

The AT875R works fine with the MKI and MKII recorders.

This suggests to me that there are certain mics that have problems with the MKIII recorders. I don't possess the same tech knowledge as some other folks who have shared in this thread but the theory about the amp in these recorders makes sense. Thus, the trick is to find a mic that works well. Again, the RE50 and the AT897 have worked for field recording interviews.