Studer V-Eight Review

ã 1999 by Eddie Ciletti

Click Here for More Options and Prices
 
 

I’m a lucky person, because tape machines just appear at my door! Take the new Studer V-Eight (List Price: $ X,xxx.xx), for example. It’s a 20-bit digital eight track (S-VHS) tape recorder that’s adat compatible, all the way back to the ancient 16-bit era. In terms of features, the V-Eight is seriously "deep," capable of full synchronization without the (extremely desirable) "Cockpit" remote control. The V-Eight is capable of providing a degree of reliability that the adat format has long needed yet this is not quite your father’s Studer. Translation: The power, the features and the interface alone could fill this review with enough details to make your head spin. I gave the transport hyper-serious inspection in the Sidebar: About the Guts. TEST DRIVE After confirming basic operation, the first "test drive" was to make the V-Eight "Master" to a pair of "vintage" slaves: a 16-bit XT as well as a Fostex RD-8, a "black-face" equivalent with built-in time code. The Cockpit can control any combination of Studer V-Eight or Alesis M-20, up to eight machines. Each can be toggled on- and off-line via dedicated button. My "ancient" machines chased perfectly well even though both the V-Eight and the Cockpit claimed the second slave was "unresponsive." This would have been a drag if I wanted to overdub on the RD-8. I swapped cables with no improvement. Conversation with Studer (so far) confirmed with my hunch, that this is a software thang. IT’S ALWAYS THE QUIET ONE! In all noise tests, the V-Eight is clearly a winner. I don’t have test equipment capable of measuring 102dB / 105 dB (weighted) dynamic range. (A-to-D and D-to-A, respectively. See Table One, for a detailed comparison of adats through history.) Instead, I chose to compare three generations of machines by connecting their analog outputs to a 24-input mixer. Of the two "tone" tests, the first routed a 40Hz sine wave (digitized by an Aardvark converter) to all eight channels of each machine via workstation. Then, to take full advantage of the 24-bit converters in each direction, the oscillator was connected to a single analog input channel on the Studer. (See the Sidebar: Noisy Neighbors.) MANHATTAN TRANSFER With the Studer as master, I used the three-machine configuration to load multitrack source material into my Soundscape workstation. The V-Eight has a dedicated Time Code track as well as an Aux track. Both are linear analog. Setting the TC generator to Internal and the TC track to "Input" converts the adat ABS (absolute) time code to SMPTE. In repro, the linear track time code is output, if recorded. Using the Input / Repro button as a TC source selector is a bit confusing. I would have preferred this to be a "Menu" item. Ultimately, whatever choice you make can be confirmed if selected via the many display pages, but I do not believe toggling Input / Repro automatically flips the display to the correct "page." Changing some parameters while the machine is in Play has no effect on the V-Eight. The unit must be stopped before changes are implemented. At the very least, the display should flash the message, "Please stop the machine for changes to take effect." I hate when machines flash the ILLEGAL OPERATION message, which are about as useless as a Microsoft HELP menu. After tangling with clocks, playing "who’s the daddy" (master or slave) and tasting SMPTE flavors by ear (to confirm whether legitimate code was being output), I could have gathered my notes and started this review. Instead, the holiday season provided an opportunity to take full advantage of the V-Eight. NERVE ANNA I received a call from Joe Hannigan of Weston Sound in Philadelphia. A last minute gig popped up — recording a "Holiday Concert" in a church with full orchestra, male and female chorus and pipe organ. While he was only hired to do a stereo recording, synchronicity linked his inspiration to call with my "holding" the goods and the perfect source material to capture on multitrack for re-mix in 5.1 surround. I believe one of the tricks to getting realistic 5.1 surround is to have real acoustic stereo pairs. The improved resolution of 20-bit preserved the spatial detail captured by a pair of B&K 6000 series mics (omni-directional, closely spaced). Add the chorus while stirring in stereo ambience tracks, spot mics (on tympani and soloists) to taste then warm with a low sub-woof flame and you can’t help but get goose bumps. It was very real!

 

PAT ON THE BACK Consuming seven inches of rack height, the V-Eight has more front- and back-panel real estate for stuff that really matters, balanced XLR analog I/O, space for AES I/O, etc. The Front Panel has a bumper crop of dedicated buttons — eighty-one — PLUS a Jog/Shuttle knob. With the exception of the high-current Power Switch, all buttons are illuminated and provide cushy-soft access to Inputs, Outputs, Presets and Preferences of the User kind. (Lotsa milk, no sugar, please!) There are two large fluorescent displays for metering and interface. The Fluorescent Level display represents 72 dB of dynamic range — enough resolution to "see" noise from a mic pre or room ambience. The Interface display has dedicated status "windows" as well as a text window that is capable of the full range of alphanumeric characters. Finally, a machine that can display status and error messages in plain, unabbreviated and unencrypted English! There is a "Quality" light (note the psychology here) that is a bit hypersensitive, certain tapes make it blink too often even though the error rate does not exceed two digits. If only the Quality Light "doubled" as button that directly led to the Error Rate display... Note: Studer has issued a service bulletin for a mod that desensitizes the Quality indicator. The Rear Panel is equally dense. Two rows of nine XLRs access the eight (primary) analog I/Os, plus a dedicated Linear AUX track. (Each includes a recessed level trim pot.) Like previous adats, the ELCO/EDAC connector makes for a quick get away. Another row includes adat optical, Word, Video, Timecode MIDI, RS-422 (edit controller), DB-9 (adat sync) and RJ-45 (meter bridge). The AUX and TC tracks do not appear on the multi-pin because there are no spare pins. OPTIONS The Cockpit Remote Control and Remote Metering Panel are pure sex. That and the fact that this review was written between two major holidaze saved me the hassle of decorating a tree. What a light show! It should be noted that the Cockpit includes the same wonderful interface screen, illuminated buttons and Jog/Shuttle knob as the V-Eight mother ship. There is more real estate, allowing the buttons to be more logically laid out than those on the transport — grouped according to function. The Cockpit is surprisingly user-friendly even without a manual (not available until after I had started the review). As mentioned, older machines may not be able to take full advantage of the Cockpit's features, but perhaps more recent versions of adat, such as the XT-20 might have the necessary hard and software? There can only be one Master, but the eight buttons labeled "Online" are used to select any machine for individual control or for enabling any or all of the Slaves. Above each are button-like LOCKED indicators that are illuminated when the slaves are behaved. The meter-bridge can display up to 32 tracks at once and can toggle between two groups of four machines. (Up to 64 tracks can be accessed.) Unlike the metering on the tape machine, no level-legend is provided; there isn’t even a reference mark. User parameters include Peak Hold, Fine (for calibrating levels), Headroom, Decay Time and Peak Hold. SUMMARY With the exception of the time limit imposed by S-VHS tape (60 minutes, max), the V-Twenty-Four "package" is the most affordable, finest sounding digital multitrack tape recorder ever. I expected the package price to be better than just totaling three machines and all of the accessories. (I didn't get the psychology of the "Quality" light, either...) No matter. The V-Eight alone is a steep jump up in quality from the land of the XT and, while reflected in the price, it is still more affordable than a two-inch analog machine. One sure reason to buy the V-Eight is for Studer’s digital converters, which are the cleanest I’ve tested. There is nothing about them that "sounds digital." The only better converters that I know of are made by Prism. Finally, I wish Studer luck with this machine. No adat clone has successfully been marketed or sold by a third party until now. The advantage to buying the Studer V-eight is having access to their support team. Studer has over thirty years of experience interfacing with world-class studios — their clients and technicians — around the world. FINAL NOTE:  (as of 6 jan'99) There is a new software version on the verge of being released. This includes some new features, some feature enhancements and some bug fixes.  Unfortunately it entails more than just a software download.  Hardware changes need to be made as well so we will be either recalling units or sending technicians out to perform a complete update including the Quality light fix. The list of new features will be provided in the near future.

 

SIDEBAR ONE: About The Guts The Studer V-Eight is a variation of the Alesis M20. The main parts are supplied by Alesis, assembled by Studer and augmented with front panel/electronics and A/D/A converters. The tape mechanism is by Panasonic. It is built from a rigid, cast-metal deck plate that is machined to tolerances hitherto unseen in previous versions of adat. The Alesis portion of the V-Eight consists of three boards: Main System, Motor Driver and I/O. The latter includes MIDI, WORD, OPTICAL, VIDEO, TIME CODE and RS-422 ports. Studer has added a nine-by-one mixer with headphone jack (for local monitoring), a parallel interface and a blank panel for an optional AES interface. The power supply is housed in a ventilated steel case with its own fan. FAST, PRECISE and SMOOTH With separate direct-drive reel motors (no tires to wear out), the Panasonic transport is super fast. It can "Locate Zero" from 44 minutes in 43 seconds with no overshoot. (An XT takes 75 seconds to complete the same task). A front panel knob has two positions: in / jog and out / shuttle. There are corresponding lights to indicate precise transport status — direction and mode — at all times. Shuttle works like you’d expect, from the center "détente" position, the machine moves in either forward (cw) or reverse (ccw) mode under capstan control. The further away from "détente" the faster it goes. Shuttle is limited to about 135 degrees rotation on either side of center. Jog allows a remarkable degree of control for precise "scrubbing." In Jog mode, the knob moves continuously throughout 360 degrees and behaves as if it were the capstan motor — like an Ampex ATR-100 or an Otari MTR-90 — in either direction.

 
 

COVER POPPIN’ DADDY Of course I popped the cover on the V-Eight straight away. I carefully observed tape handling under all conditions and immediately noticed three fixable bugs: 1.) Just before the tape meets the first precision guide there is a heavy, brass Impedance Roller to smooth out tape flutter as it exits the cassette shell. A support brace to the left of the impedance roller once served a useful function for Panasonic when this transport was used in their medical grade video decks. In the V-Eight, the brace can (and did) bend enough to impede the impedance roller. (I removed the offending device.) There is also full-track erase head that is not connected. 2.) I knew the perfect tape path test would be to use the Shuttle and Jog modes to move the tape in both directions under capstan control. Indeed, it did reveal a problem. On the opposite side of the head is a fixed guide (just before the Capstan Motor) that supports the lower edge of the tape. In the forward direction (Play, Shuttle or Jog), the tape bottomed out and began to curl. In reverse, the tape skewed up toward the top flange revealing that the motor's ever-so-slight deviance from "square" was the culprit. PLEASE NOTE THE BUG BELOW 3.) On more than one occasion, the machine boots up complaining that the PG timing needs to be re-calibrated.  I reset the value but it does not stick.  Could this be a grounding issue?  Is it possible that making connections from the outside world might "shock" the innards enough to make the machine lose this critical adjustment?  I did not bother to check whether other NOVRAM settings (like head hours) were lost, but I think this issue must be addressed as soon as possible. Studer reported two other instances of the PG delay value needing to be restored.  Once reset, the value stayed in memory and neither machine lost any other info like head hours, etc. At this time, there is no explanation of what could cause this to happen besides a battery problem.  (Even though the problem has existed since the XT.)  Studer is still looking into this matter.

 
 

Please note: Prior to publishing this review, I made my extensive modification notes (below) available to Studer and they now incorporate a visual inspection during quality control. In the month since Studer was alerted to the tape path problem, only one other machine was found to skew the tape when jogging back and forth.  This doesn't seem to cause any any short term performance problems but it was held it back from shipping until an expedient method of adjusting the capstan angle can be devised.   FIGURE ONE: Arrow indicates guide where curl occurs FLYING TOOLS There was no way I was going to put my master tapes in the V-Eight until the tape path problem was resolved. The audio signal is not recorded to the extreme edges of the tape, but a transport as good as this one should be capable of kindness plus take an adjustment tweak and stay there.  I found that exerting pressure on the top of the capstan housing either exaggerated or resolved the problem. So, unlike most reviewers, I started removing components until the capstan motor mounting screws became accessible. (I won’t mention that I took a Dremel motor tool to these parts to make them more service-friendly.) Figure Two: Colored Arrows indicate where head PCB and HEAD connections impede screwdriver access to capstan mounting screw. Black Arrows indicate places where metalwork makes this part difficult to remove without first removing the solenoid assembly. As a test, I created a shim that would hopefully move the capstan shaft into a position exactly perpendicular with tape travel. The first shim was 3 mils (.003 inches) thick, yielding nearly perfect results. I then proceeded to polish the opposite flange until 2.5 mils had been removed. Voila. The tape path is perfect and consistent with no shim! Now I can safely proceed with a long-awaited 24-track remix. FIGURE THREE: Black arrow improves serviceability.  Blue arrows indicate where trial shim was used. Yellow arrow points to hidden flange, which was polished to improve perpendicularity. WISH LIST I wish that Studer had been involved at the very beginning to bring their expertise to the design of the MD-20 / V-Eight. While it should be very reliable, I found that several steps could have been taken to make the machine more serviceable. The most obvious is the lack of access to the bottom of the transport. Currently, the steps involve either removing the front panel or the motor driver board and then the transport. Both steps make the unit inoperable and all steps are unnecessary, "if only..." The "Studer" method would have been to engineer a hole into the chassis and put the motor board on a hinge — with enough wiring slack — so that it could be safely positioned out of the way. The transport should also be mounted on a tray — or in some way provide improved access to the mounting screws, four of which are underneath the motor driver board — so that the transport can be removed without disassembling the rest of the unit.  Then, for example, if one needs to check the tach head clearance (of the capstan motor), there is infinitely less labor required. Instead, every place you look, wiring harnesses are exactly to length, on both sides of the front panel and motor board, a practice that restricts movement in any direction. These are fixable problems that in the long run encourage technicians to take the time to do the best job possible. The front panel could also be redesigned on hinges so that it can be folded down — again with enough wiring slack and connectors to permit service. The optional glide rails mean that the front panel does not need to support the unit.  In a rack, they only lock the unit into place. Return to where you left off: "Test Drive"

 

SIDEBAR TWO: Noisy Neighbors Using a low frequency sine wave can reveal a remarkable number of anomalies in both the analog (scratchy pots and switches) and the digital domains (converter non-linearities). A 40 Hz tone was recorded on a Soundscape workstation then played back and routed to all eight channels of the TDIF (Tascam Compatible) port. Soundscape’s SS8I/O translated TDIF to adat optical feeding first the V-Eight, then the two slaves via optical daisy chain. I made sure that each track of each machine produced the same level on the mix buss, then connected the mixer output to a Crane Song STC-8 Compressor to add enough gain (and protection) to make the noise floor recordable. (I will post Real Audio samples in the near future.) The V-Eight is so quiet and its noise spectrum so "sweet" (in an analog way) and devoid of artifacts that I had to raise four faders to make it obvious. The other machines were compared in the same way even though one fader made their presence known! The level of the 40 Hz tone was lowered until it created obvious artifacts on the Fostex RD-8. This lit the last segment of the V-Eight’s meters, -72 dBFS. (No segments were lit on the XT or the RD-8. The meters can’t display that low!) The V-Eight’s digital output can be toggled from 20-bit to 16-bit dithered (noise shaping) to 16-bit truncated. Changing the bit depth and dithering smoothed the XT’s low level chirping into "snow." The RD-8’s converters were quite inconsistent from track to track, making it hard to choose between the aggressive snowfall of noise (with chirps) or the resulting hash-n-fuzz that dither normally cleans up. Relative to the Studer, the adat-xt was about 10 dB noisier while the RD-8 was quite the chatterbox with nearly 20 dB more noise! That’s generational progress for ya! Return to where you left off: "Manhattan Transfer"

Specifications Throughout the Product's History
Compiled from Alesis and Studer on-line documentation
 

product	adat	XT	LX-20	XT-20	M-20	Studer V-Eight	Prism convertor only
record &  play	16-bit	16-bit	16 / 20 bit	16 / 20 bit	16 / 20 bit	16 / 20 bit	24-bit
Specifications  ( below )							For Reference only
Ref Level	-15dBFS	-15dBFS	-15dBFS	-15dBFS	-15dBFS	-15dBFS
Max level pro/con	+19dBu / +5 dBV	+19dBu / +5 dBV	+19dBu / +5 dBV	+19dBu / +5 dBV		+26 dBu / na
Record (A/D) Delta-Sigma	16 bit, 64x	18-bit, 128x	20 bit, 64x	20 bit linear, 128x	24-bit, 64x	24-bit, 128x	24-bit
Play (D/A) Delta-Sigma	18 bit  ?x not specified	20 bit 8x	20 bit, 128x	20 bit linear, 128	20-bit, 128x	24-bit, 128x	24-bit
Dynamic Range "A" weighted  20 Hz-20 kHz	92 dB	92dB	97 dB	102 dB	A/D: 115 dB,  D/A: 105 dB,  see note #1	A/D: 105 dB,  D/A: 102 dB,  see note #2	130 dB un-weighted
Distortion @1kHz  @ -0.1 dBFS	.009% THD + Noise  A-weighted	.009% THD + Noise  A-weighted	0.009% THD + Noise  A-weighted	0.003% THD + Noise  A-weighted	<.009% THD + Noise  A-weighted	0.002% THD + Noise  A-weighted
Crosstalk	-90dB@1kHz	-90dB@1kHz	-90dB@1kHz	-90dB@1kHz	-90dB@1kHz	100dB@1KHz	For Ref Only
Note #3	16 bit only	16 bit only	in 20 bit mode	in 20 bit mode	in 20 bit mode	in 20 bit mode	in 24-bit mode

Note #1: Alesis specs are not verified.  I invite Alesis to send an M-20 so a direct comparison can be made.

Note #2: Studer specs are measured and therefore traceable.

Note #3: single converter per channel

Return to where you left off: IT’S ALWAYS THE QUIET ONE!

Return to TOP of document

e-mail: edaudio@tangible-technology.com

Return to the Home Page?

Or, Visit the Directory?