DNA 1  de-noiser de-clicker ambience processor

The Weiss DNA1 is an ex­tremely ver­sa­tile tool­box for restora­tion work. It han­dles broad­band noise re­duc­tion as well as de-click­ing and de-crack­ling. The K-Stereo Am­bi­ence Proces­sor brings back life to the mu­sic in case it lost some dur­ing de-nois­ing, or gen­er­ates a pseudo stereo sig­nal out of mono. Fi­nally an M/S ma­trix al­lows to de­fine the stereo width.

DNA1 – a mile­stone in the Gam­bit Se­ries equip­ment de­vel­op­ments at Weiss En­gi­neer­ing. Three state of the art al­go­rithms are packed into the DNA1. The de-nois­ing sec­tion has been de­vel­oped in co­op­er­a­tion with the Pure Notes Com­pany (UK), the de-crack­ling and de-click­ing al­go­rithms are the Weiss R&D lab’s de­sign and fi­nally the K-Stereo Am­bi­ence Proces­sor we li­censed from Mr. Bob Katz (USA). All func­tions can be used con­cur­rently. The DNA1 can be switched to M/S mode which can be use­ful e.g. for de-nois­ing the M chan­nel only while ap­ply­ing less de-nois­ing on the S chan­nel to leave the am­bi­ence in­tact.

Con­trols, display

There are a to­tal of 15 knobs in or­der to al­most have a “one knob per pa­ra­me­ter” op­er­a­tion. The knobs are touch sen­si­tive to switch the LCD to dis­play the as­so­ci­ated pa­ra­me­ters. The LCD shows an FFT of the in­put sig­nal and me­ters to in­di­cate the amount of de-nois­ing go­ing on. In ad­di­tion all pa­ra­me­ters are shown in nu­mer­i­cal format.

Snap­shots, MIDI, metering

Other ba­sic fea­tures are an A/B com­pare mem­ory, a 128 po­si­tion snap­shot bank with two ad­di­tional banks for back-up (can also be dumped/re­cov­ered via MIDI), over in­di­ca­tors, MIDI con­trol for each pa­ra­me­ter, an over­all by­pass switch and an over­all gain con­trol, a chan­nel gang­ing switch and an M/S mode switch.

Sig­nal pro­cess­ing, connectivity

The in­ter­nal pro­cess­ing is done in a 40 Bit float­ing-point for­mat. In­put sig­nals can be be­tween 44.1 and 96 kHz. The out­put can be POW-R dithered to 16, 20 or 24 Bits. The dig­i­tal in­put/out­put are in AES/EBU for­mat on XLR connectors.

Broad­band de-noiser

The broad­band de-noiser sec­tion works based on an FFT ap­proach with a se­lec­table num­ber of fre­quency bands. The en­ergy in each band is an­a­lyzed and thus de­ter­mines whether the band is of rel­e­vance for the mu­sic sig­nal. The de-noiser does not need to take a noise fin­ger­print, it con­tin­u­ally adapts to the noise sig­na­ture. Three op­er­at­ing modes can be se­lected, the au­to­matic mode for un­crit­i­cal de-nois­ing, i.e. for cases where the noise floor is fairly low al­ready and thus the de-nois­ing is not that in­tru­sive. The semi-auto mode where some of the de-noiser’s pa­ra­me­ters are es­ti­mated first based on the auto-mode and sub­se­quently are re­fined man­u­ally by the user. And fi­nally the fully man­ual mode where all pa­ra­me­ters are ac­ces­si­ble, for those cases which re­quire more sub­tle tun­ing to get the best re­sult. Like e.g. cases of old record­ings with high lev­els of noise which need a sub­tle bal­ance be­tween re­moved noise and noise re­quired to stay such that the “vin­tage” touch of the record­ing is kept alive.

De-clicker, de-crackler

The de-clicker sec­tion is made up of nine main pa­ra­me­ters. There are three groups of three pa­ra­me­ters each. Each group, named “qual­ity”, “sen­si­tiv­ity” and “click type”, has a few in­di­vid­u­ally se­lec­table snap­shots which al­low to zoom in on a par­tic­u­lar de-click­ing prob­lem at hand. Af­ter set­ting the three snap­shots for the three groups, the pa­ra­me­ters can be trimmed fur­ther in­di­vid­u­ally. With that scheme it is rel­a­tively sim­ple to get a re­sult quickly.

K-Stereo

The K-Stereo proces­sor uses one main pa­ra­me­ter, the K-level, which de­fines the level of the ef­fect. In ad­di­tion there is a three band equal­izer with low cut, high cut and a para­met­ric cen­ter band in peak­ing mode. The equal­izer shapes the au­dio go­ing through the K-Stereo process. The “wide” and “deep” on/off switches al­low for ad­di­tional shap­ing of the effect.

Out­put M/S Matrix

The out­put sec­tion con­sists of an ad­di­tional M/S en­cod­ing/de­cod­ing ma­trix for fi­nal stereo width shap­ing. In ad­di­tion left and right chan­nels can be trimmed in­di­vid­u­ally plus fi­nally the vol­ume of the stereo sig­nal can be changed with a sin­gle knob.

Power

 • Mains voltage: 115 V or 230 V with voltage selector

 • Fuse rating: 500 mA slow blow

 • Power consumption: 40 VA max.

 Size

 • Depth: 30 cm

 • Width: 43,3 cm (19")

 • Height: 8,8 cm (2HU)

 AES/EBU Input

 • Sampling frequencies: 44.1 kHz, 48 kHz, 88.2 kHz, 96 kHz

 • Maximum input word-length: 24 Bits

 • Channel Status Data: Input accepts professional or consumer format

 • Channel Status Data Bits forwarded to AES/EBU output: see table below

 • Connector: XLR female

AES/EBU Output

 • Sampling frequencies: 44.1 kHz, 48 kHz, 88.2 kHz, 96 kHz (always the same as the input)

 • Output word-length: 24 Bits

 • Connector: XLR male

 AES/EBU Channel Status Data

 The DNA1 allows to convert the incoming Channel Status Data as follows:

 • From Consumer to Professional

 • From Professional to Consumer

 • Transparent mode, i.e. Channel Status Data is fed forward to the output without any conversion