[Search for users] [Overall Top Noters] [List of all Conferences] [Download this site]

Conference napalm::commusic_v1

Title:* * Computer Music, MIDI, and Related Topics * *
Notice:Conference has been write-locked. Use new version.
Moderator:DYPSS1::SCHAFER
Created:Thu Feb 20 1986
Last Modified:Mon Aug 29 1994
Last Successful Update:Fri Jun 06 1997
Number of topics:2852
Total number of notes:33157

1911.0. "samplers -- generic comparison" by NORGE::CHAD () Tue Feb 14 1989 13:57

Mr. Moderator:  Please move this if you think there is already a generic
discussion of samplers (I couldn't find one that hadn't degenerated :-)

March 89 KEYBOARD was in my mialbox yesterday.  A "special issue" on samplers.

They lab-tested a bunch and list "specs" for a whole lot more than tested, as 
well as doing a blind listening test.

The blind test had the following catagories: (rated from 1 to 10)

Test #1. 

Unprocessed original pitch

     a. Thin/Fat
     b. Noisy/Clean
     c. Dull/Bright 
     d. Processed/Transparent

Samplers ranked from top to bottom:

           (overall)
Casio FZ-10M  8.56  a. 8.37  b. 9.00  c. 8.37  d. 8.50
E-MU Emul III 7.96  a. 7.75  b. 8.25  c. 8.00  d. 7.87
Akai S1000    7.56  a. 7.87  b. 7.12  c. 7.37  d. 7.87
Simmons SDX   7.25  a. 7.50  b. 8.75  c. 5.87  d. 6.87
Seq. Prpht
2002          5.90  a. 6.50  b. 5.50  c. 6.75  d. 4.87
Rol. S-550    5.81  a. 6.50  b. 4.50  c. 7.00  d. 5.25
Ens. EPS      5.71  a. 6.12  b. 5.87  c. 6.12  d. 4.75
KORG DSM-1    5.46  a. 6.00  b. 6.50  c. 4.62  d. 4.75
Rol. S-330    5.37  a. 6.00  b. 4.75  c. 5.87  d. 4.87
Akai S950     4.62  a. 5.87  b. 3.25  c. 5.75  d. 3.62
Yam. TX16W    4.50  a. 5.75  b. 3.75  c. 4.75  d. 3.75
Akai S900     4.31  a. 5.50  b. 2.12  c. 5.00  d. 4.62
E-mu Emax     4.18  a. 5.62  b. 3.12  c. 4.37  d. 3.62
K-250 w/
sampling opt. 3.25  a. 5.00  b. 1.50  c. 4.12  d. 2.37
Ens. Mirage   1.53  a. 2.50  b. 1.00  c. 1.62  d. 1.00

Test #2

Transposition

Akai S1000      7.42
E-mu Emul. III  7.28
Akai S950       6.71
Yam TX16W       6.57
Rol. S-550      6.14
E-mu Emax       5.57
Seq. Prpht 2002 5.28
Simmons SDX     5.14
K-250 w/smpl    5.14
Rol. S-330      5.14
Akai S900       4.28
Korg DSM-1      3.14
Ensoniq EPS     2.85
Casio FZ-10M    1.71
Ensoniq Mirage  1.14


Test #3

Chords (a test of playing same note more than once, to test DACs etc.)

Akai S950       7.00
Akai S1000      6.42
Yam TX16W       6.28
E-mu Emul. III  6.00
Rol. S-550      5.71
Prophet 2002    5.57
Simmons SDX     5.42
Roland S-330    5.28
E-mu Emax       5.00
K-250           4.85
Akai S900       4.71
Korg DSM-1      4.28
Ensoniq EPS     3.28
Casio FZ10-M    2.71
Ensoniq Mirage  2.00

Test #4

Processed sound (internal analog processing)

     a. Thin/Fat
     b. Noisy/Clean
     c. Dull/Bright 
     d. Processed/Transparent

Akai S1000     8.10  a. 7.42  b. 8.57  c. 8.42  d. 8.00
E-mu Emul III  7.64  a. 7.28  b. 8.28  c. 7.42  d. 7.57
Rol. S-550     6.89  a. 7.14  b. 8.00  c. 7.00  d. 5.42
K-250          6.78  a. 6.42  b. 7.14  c. 7.14  d. 6.42
Akai S900      6.75  a. 6.00  b. 8.28  c. 6.28  d. 6.42
Korg DSM-1     6.50  a. 7.00  b. 7.71  c. 5.85  d. 5.42
Akai S950      5.96  a. 5.28  b. 7.57  c. 5.85  d. 5.14
Simmons SDX    5.75  a. 5.42  b. 7.00  c. 6.00  d. 4.57
Rol. S-330     5.74  a. 6.28  b. 6.00  c. 6.28  d. 4.42
Prophet 2002   5.64  a. 6.14  b. 6.28  c. 5.42  d. 4.71
E-mu Emax      5.60  a. 5.57  b. 6.00  c. 5.42  d. 5.42
Yam TX16W      5.42  a. 5.71  b. 5.42  c. 5.42  d. 5.14
Casio FZ-10M   4.46  a. 5.14  b. 4.00  c. 4.57  d. 4.14
Ensoniq EPS    4.10  a. 4.85  b. 3.57  c. 4.42  d. 3.57
Ensoniq Mirage 2.32  a. 5.28  b. 1.14  c. 1.71  d. 1.14


Rating #5 CLEAN  (comb. unprocessed and processed scores)

E0mu Emula. III  8.26
Simmons SDX      7.87
Akai S1000       7.84
Korg DSM-1       7.10
Casio FZ-10M     6.78
Rol. S-550       6.25
Prophet 2002     5.89
Akai S950        5.41
Rol. S-330       5.37
Akai S900        5.20
Ensoniq EPS      4.72
Yam TX16W        4.58
E-mu Emax        4.56
K-250            4.32
Ensoniq Mirage   1.07


Overall rating (pure average of 4 tests)

Akai S1000       7.34
E-mu Emul III    7.22
Rol. S-550       6.14
Akai S950        6.07
Simmons SDX      5.89
Yam TX16W        5.69
Prophet 2002     5.60
Rol. S-330       5.38
E-mu Emax        5.09
Akai S900        5.01
K-250            5.00
Korg DSM-1       4.84
Casio FZ-10M     4.36
Ensoniq EPS      3.99
Ensoniq Mirage   1.74

--------------------------------------------------------------------------

There are also lab tests (above just blind listening tests) with scope output,
plots of various things etc.  plus discussion of all tests lab and listening
and what they did.  It is an interesting issue of KEYBOARD.

FWIW

CHad
T.RTitleUserPersonal
Name
DateLines
1911.1and sounds nicer, tooMARVIN::MACHINTue Feb 14 1989 14:086
    
    
    Well, the FZ certainly gets close enough to ruffle the Emax's
    expensive feathers.
    
    	Richard.
1911.2MIZZOU::SHERMANquality first cause quality lastsTue Feb 14 1989 14:3626
    This issue made me feel really good about my S-10.  Although the
    S-10 only has up to about 13.5 kHz bandwidth, one output and four
    samples, it is pretty quiet and doesn't distort much.  It basically is 
    in line with the review of the S-50, but ranks among the cheapest of
    the samplers.  Did you notice how many of the cheaper samplers have
    been discontinued?  Seems like to get a quality sampler you're talking 
    bigger bucks nowadays.  That seems to be that market trend - kill the 
    low-end stuff so that folks have to buy the more expensive items.  Also,
    I found here that my S-10 *does not* support MIDI sample dump standard.
    Neither do a bunch of others.  But, this will be no big deal
    since software vendors are pushing 'universal' sample dump support.
    
    The comments about the Mirage were funny.  'Dog meat, right?  Well,
    it's hard to deny that the Mirage's results look pretty horrendous
    ... The Mirage's distortion levels surpass those of any other sampler
    by such a wide margin that it may as well be considered a synthesizer.'
    
    Gee, seems to me we should put the Mirage into a class of its own!
    ;^)  
    
    Also surprising was that on a spec basis there sure is a wide range
    of bang for the buck.  It was surprising to see that the specs on
    the high end machines were not so impressive as one might hope.
    Again, made me pretty happy to have chosen the S-10 for my setup.
    
    Steve
1911.32 * 330 = 550 ??TROA01::HITCHMOUGHTue Feb 14 1989 15:176
    Interesting stuff, but can anybody tell me why there is such a
    difference between the S-330 and the S-550? I thought a 330 was
    a 550 with less memory.
    
    Ken
    
1911.4Glad to hear that my ears weren't lyingDREGS::BLICKSTEINSo What? BEEG Deal!Tue Feb 14 1989 16:308
    I'm just always glad to have something tangible to point to backup
    what I've always felt about the S-550, which is that I've always
    thought it sounded great.
    
    However, I now have plenty of things to curse it for as well.  So
    don't consider this a unabashed plug for the S-550.
    
    	db
1911.5Ease of use and ability to understandDREGS::BLICKSTEINSo What? BEEG Deal!Wed Feb 15 1989 17:348
    One of the most interesting parts of the issue was the inset about
    how easy each sampler was to use and understand.
    
    They ranted about most of them, but they did have nice things to
    say about two of them.   It was great to see that it WAS possible
    to make a sampler that's easy to use and understand.
    
    	db
1911.6not very objectiveSUBSYS::ORINA waist is a terrible thing to mindWed Feb 15 1989 18:4418
< Note 1911.5 by DREGS::BLICKSTEIN "So What? BEEG Deal!" >
                   -< Ease of use and ability to understand >-

That rant section about User Interface gripes proved that they really did
not know how to take advantage of the features that each machine has, probably
due to lack of time. They didn't like either the EPS or the S550. The S550
suffers from a lousy manual, and the EPS suffers from a lack of manual
(until now). They particularly mentioned that the EPS lacked the capability to
store and recall global parameters. That was completely erroneous. There is a
SAVE GLOBAL PARAMS feature right in the SYSTEM menu. They also complained about
the number of button pushes, which shows that they had problems learning the
hierarchy of the menus and how to use the EDIT button to switch back and forth
to do global parameter edits. I find the EPS very intuitive to use. You can
take a sample, assign the root key, map it to the keyboard, tune it and envelope
it, set the LFO, in about 20 key strokes. Looping takes longer, of course.
Universal Sound Designer is a life saver for cross-fade looping.

dave
1911.7do tell.MARVIN::MACHINThu Feb 16 1989 09:404
    
    O.K., so which samplers' UI did they prefer?
    
    Richard.
1911.8IMHO, ITHO, what's the difference.CTHULU::YERAZUNISSmurf _Terminator_Thu Feb 16 1989 13:3813
    The Mirage, of course!  :-)
    
    Note that their fidelity rankings corresponded highly with the number
    of bits in each machine, and to a lesser extent to the max sample
    rate of the machine.
    
    I certainly don't agree with their rankings.  
    
    And, of course, cost was not a major object.  Was it my imagination,
    or was the price ratio 5:1 across the field?
    
    	-Bill
    
1911.9pricey stuffSUBSYS::ORINA waist is a terrible thing to mindThu Feb 16 1989 18:3917
>    And, of course, cost was not a major object.  Was it my imagination,
>    or was the price ratio 5:1 across the field?


Yes, the winner was Akai S1000 which retail lists for $5995. I was quoted
a price of $4300 by Caruso's, but that is negotiable. This is a 16 bit
stereo sampler. Not much of a library available, and it is not compatible
with the S900 or S950 12 bit machines or libraries. The price I got for an
S950 was $1795 from Caruso's, plus they will send you the library to copy.
I think the Emulator III costs about twice as much as the S1000? It looks
like a Roland S50, S550 or Akai S950 are choice picks. I still think the
EPS will perform much better when treated properly, and from the list of
credits, I'm surprised the Ensoniq reps didn't do better with assistance.
The new O/S should make quite a difference in sampling quality.

dave

1911.10errata?NORGE::CHADThu Feb 16 1989 19:3518
re: .9

>
>Yes, the winner was Akai S1000 which retail lists for $5995. I was quoted
>a price of $4300 by Caruso's, but that is negotiable. This is a 16 bit
>stereo sampler. Not much of a library available, and it is not compatible
>with the S900 or S950 12 bit machines or libraries. The price I got for an

If I remember the 'keyboard' review correctly, the S1000 can use S900
disks, albeit in 12 bit mode.  And it had a lot of REAL nice features.

>S950 was $1795 from Caruso's, plus they will send you the library to copy.
>I think the Emulator III costs about twice as much as the S1000? It looks
>like a Roland S50, S550 or Akai S950 are choice picks. I still think the
>EPS will perform much better when treated properly, and from the list of


CHad
1911.11just ask Sun MicrosystemsSALSA::MOELLERAudio/Video/MIDIophileTue Feb 21 1989 18:3613
    I just bought the issue with the sampler 'reviews' in it.
    
    Very surprised to see the Emax' poor showing.  It did get points
    for a good user interface, and should have gotten lotsa points for
    its huge sample library.  There's a musician here in town with a
    studio.. I've worked with him and his AKAI S900.. to me it sounds
    thin and brittle.. kind of harsh.  The Emax sounds rich in
    comparison.. plus it     has an internal sequencer and real software 
    panning, not just left and right like the S900/950..
    
    Ah well, just goes to show ya how numbers can lie
    
    karl
1911.12only MOCGVAX2::COREY_JTue Feb 21 1989 20:068
    re: .11
    
    Not to mention the Emax is multi-timbral and has one of the best
    arppegiators Ive ever used.
    
    dito
    
    
1911.13SALSA::MOELLERAudio/Video/MIDIophileTue Feb 21 1989 21:3226
    More thoughts on the EMAX and Keyboard's testing methodology..
    The Emax showed poor on overall digital noise level.. which runs
    completely counter to my experience with it.
    
    The Keyboard tests used the Emax' own sample input, and then measured
    its output on a scope... for THD, noise, freq. response, and square
    wave reproduction. (did great on the square wave)..
    
    My thought is this.  I've done very little sampling with the Emax,
    preferring instead to use my large (100+diskettes) sample library.
    Many of the samples in my library were not created using the A-D 
    converter of the Emax.. or even the Emulator II.. most of these 
    samples come from the E-II library, ported with Digidesign software.. 
    and many of the original samples come from specialized sampling
    computers, with virtually no digital noise from the sampling process.  

    And THAT must be why I consider the Emax to be a superb SGU.
    
    So what, you say.. the tests were testing SAMPLING, not 'sample
    players'.. if true, why'd they include the Oberheim DPX sample 
    player ?
    
    somewhat incoherently yours,
    
    karl
    
1911.14the most musical of the lotMARVIN::MACHINWed Feb 22 1989 07:494
    
    Tell you what -- bet you haven't listened to an FZn. 
    
    Richard.
1911.15S1000 InfoYUPPY::GEALWed Feb 22 1989 09:4219
    I recently got an S1000 and thought I would pass on my observations.
    The UI is fine and even the more difficult operations can be picked
    very quickly. It comes with four factory disks (HD) with Piano,
    Strings, Brass and Drums. The sounds are excellent and there will
    shortly be a fairly extensive library available which has been recorded
    at AIR studios in London. The S1000 will read S900 disks, but the
    sampling algorithms are very different and some sounds convert 
    a lot better than others. Also, what used to fit on one (DD) disk
    on an S900 will NOT fit on one disk when its been converted by the
    S1000. This means using HD disks - expensive. The S1000 comes with
    2Mb of memory which, if you are using the true 16bit samples is
    not enough. You can expand to 8mb but this is expensive in the UK
    - about $1400 for each 2mb increase. If you have one sample using
    the full 2mb, you can't write it to floppy - you have to have a
    hard disk. Both the Atari/Supra and SCSI interfaces are supported,
    but you have to buy the relevant card - about $175 here.
    
    One last point - last night mine went gaga and has had to go back
    for repair. Beware new products!