Monday, August 31st, 2009 | by matthew mcglynn
Jon from Audio Geek Zine pointed out a thread on the DUC that gives readers a chance to compare a stock Digi 003 to a Black Lion Audio-modded Digi 003. I’m a big fan of aftermarket mods for audio hardware, so this is a question that I have a lot of interest in — in part because I spent about $1200 getting my Digi 002-Rack modded by BLA earlier this year.
Is the mod worth the money?
The Black Lion Signature Series mod includes a complete rebuild of the mic pre’s, line inputs, headphone amp, and line outputs, using upgraded opamps from Burr-Brown (TI). According to BLA, “these modifications result in a cleaner signal, with much more high end definition.”
Next, the master clock is replaced with a new, low-jitter design. Clock improvements are said to deliver “improved signal clarity, fuller harmonic extension of the instrument you are tracking, an apparent increase in volume over the stock design (because of lower phase cancellations), and pinpoint image placement during mix-down.”
The ADC and DAC are upgraded by swapping “low-grade ceramic capacitors” for premium units. This reduces distortion and phase cancellation during conversion. Additional noise-reduction circuitry is installed as well.
Finally, the power supply is upgraded. I believe a fully separate supply is installed for the analog circuitry. This brings an increase in headroom that BLA claims gives audio more “heft” and “presence.”
Last month I had an opportunity to record a quick A/B comparison. I’ve just now circled back to do some comparative listening.
It was a very limited test. But I can’t hear any improvement at all.
My test was small: I recorded one track of a single acoustic instrument. I can imagine that different instruments or voices might reveal sonic improvements that this test does not. But still, I’m not sure I’d have spent the money if I knew I would only hear a difference sometimes.
For this test, I took a matched pair of 3 Zigma CHI microphones with small-diaphragm cardioid capsules. These are transformerless FET mics with low-noise amplifier circuits. We rubberbanded the two mics together and hung them from a single shockmount.
We positioned them so the capsules were about 18'' from the 12th fret of Michael Capella‘s acoustic guitar.
We plugged one mic into my BLA-modded 002 Rack. We plugged the other mic into Michael’s stock 002 Rack. We matched the audio levels visually as best we could. No, this was not a scientific test; I left both my tone generator and my oscilloscope in some other reality in which I’m an even bigger audio nerd.
We recorded a single performance simultaneously on both DAWs at 24-bit, 44.1kHz.
In my experience, performance differences tend to produce greater changes in the audio signal than would a swap of preamp or converter, or even microphone in many cases. So while it’s true that these two mics would have recorded subtly different signals by virtue of being in slightly different locations, the difference is far smaller than it would have been if we’d recorded two different performances with a single mic.
I copied the raw audio file from Michael’s DAW to my own, imported it to a new track in my Pro Tools session, and gain-matched it to the track I’d recorded. I first played them back through my BLA 002R into good headphones. On that first listen, the track recorded by the BLA unit sounded warmer in the mids. The last chord of the piece had more midrange presence than on the track from the stock 002R. It was a subtle difference, but favorable.
Later, in a blind playback test through a different DAC into the same headphones, I preferred the stock 002R track. Oops.
Playing the tracks back through the modded 002R and a pair of Mackie HR824 monitors, in a blind test I again picked the stock 002 track for its articulation — for example, in that final chord, I could hear the attack of the pick on every string. In the track recorded by my modded 002R, the chord seemed to have a softer attack, less articulation. Is the warmth I perceived in the BLA track really just a lack of high-frequency extension?
These differences are minor, and will evade notice by casual listeners. That’s a problem. I’m damn disappointed to not be hearing dramatic improvements in the modded unit.