Antares AMM1 Hardware user manual

							31
This page allows you to specify for each mic model downloaded into
FLASH memory whether or not it will appear in the appropriate mic
selection list. It is also the first step in permanently erasing a model
from FLASH memory (see the Compress FLASH Page below for more
details).
A bit of explanation: When you first take your AMM-1 out of
the box, it contains models of approximately 100 mics perma-
nently stored in EPROM (more accurately, 200 models, in that
each mic model includes both a source version and a model
version). In addition, the AMM-1 includes enough FLASH
memory to hold approximately 100 more mic models (again,
actually 200, counting source and model versions).
Initially the FLASH memory is empty. When you download new
models from our web site and load them into your AMM-1 via
MIDI they are stored in the FLASH memory. This page and the
FLASH management pages that follow let you monitor the
FLASH memory and take appropriate action should you ap-
proach its maximum capacity.
Accessing this page when there are no mic models stored in FLASH
will result in the the message 
no models in FLASH.
If new models have previously been stored in FLASH, you will see
the top screen, where 
yyy:zzzzzzzzzzzzzzzz shows the
currently selected FLASH model. If that model is a Source version,
yyy appears as Src. If that model is a Model version, yyy
appears as Mod.
zzzzzzzzzzzzzzzz is the model name, indicating both the
manufacturer and mic name.
The content of the 
xxxxxxx field depends on the current state
of each particular model. When a new model is loaded into FLASH it
is automatically set to be “active.” That is, it appears in both the
Source and Model mic selection lists. When you select a currently
active mic on this page, the top line of the display will read:
 to: delete
Pressing the Enter button at this point will mark the model as
deleted. Once a model is marked deleted, it will no longer show up
in the appropriate mic selection list. However, note that it is not
removed from FLASH. Deleted models are only removed from FLASH
using the Compress FLASH Page discussed below.
System Edit 
						

							32
When you select a model that has previously been marked as de-
leted, the top line of the display will read:
 to: restore
Pressing the Enter button at this point will restore the model to
active status. It will once again appear in the appropriate mic
selection list and will not be subject to erasure during the FLASH
compression process.
Note: The Source and Model versions of each mic may be indi-
vidually set to active or deleted status. For example, if you
download a model of a mic that you don’t physically own, you
might decide to delete the Source version (since you won’t be
choosing it as a Source mic) but leave the Model version active
so that you can choose it as a modeled mic.
FLASH Info Page
Active models: 47%
Deleted models: 12%
This screen displays the percentage of FLASH that is used by active
and deleted models respectively. If these numbers are added, and in
turn subtracted from 100, the result is the percentage of FLASH that
is unused.
For example, the data in the screen above shows that 59% of the
flash is currently used, with 41% available for future downloads.
Compress FLASH Page
Press Enter to erase
deleted FLASH models
Here’s where the models that you marked Deleted back up in the
Delete/Restore FLASH Models Page can be physically removed from
FLASH to make room for other models. Simply navigate to this page
and press the Enter button. All Deleted models will be removed and
the FLASH will be defragmented. Depending on the number of
models to be removed, compressing may take as long as 45 seconds.
During this time, the message 
...working...  will appear on
the display.
System Edit 
						

							33
Chapter 5:
The Microphone Models
Your AMM-1 comes with a collection of approximately 100 mic
models permanently stored in EPROM. (A complete list will be found
in the Appendix to this manual.) In addition, we are constantly
modeling more mics. Whether new mics that have just come to
market, or classics we’ve just managed to get ahold of, you should
find an ever-growing collection to download from our website.
We’d also appreciate any suggestions you have for specific mics to
model. Email your suggestions to us at [email protected] with
the words “Mic Suggestion” in the subject line. We can’t guarantee
that we’ll be able to include every mic suggested, but if we see
certain mics getting a lot of votes we’ll do our best to include them.
Adding New Mic Models
To add new models to your AMM-1 you will need a Mac or PC
computer equipped with the following:
•An internet connection (to download the model files from our
web site)
•An application capable of opening and playing Standard MIDI
Files (pretty much any MIDI sequencer should do the trick)
•A MIDI interface (for transferring the data to the AMM-1)
When you download a new mic model file from our web site, what
you end up with is a Standard MIDI File. The AMM-1 Standard MIDI
files hold System Exclusive messages (SysEx messages), which in turn
hold model data for the AMM–1.
Once you’ve downloaded the file, open it in a MIDI sequencer and
simply play it into your AMM-1 (making sure that you’ve first se-
lected “Ye s” on the Allow Downloads page). 
						

							34
When the AMM-1 detects the new mic file, the following message
will appear on the LCD:
Busy receiving MIDI
  model download
The models are written to the FLASH memory as they are received.
Once all of the models have been received, the names of the new
mics are merged into the mic selection lists. while this is happening,
the following message is displayed:
 Adding new models
     to menus
This all happens fairly quickly. In the worst case, 100 microphones
(200 models) would take about 60 seconds to download and take
about 15 seconds to sort into the menus. In the more likely case of a
file containing 4 new models, the entire process would take less
than 4 seconds (and the “Adding new models “ screen would go by
almost too fast to see).
Note: Many MIDI sequencers provide the capability to edit
Standard MIDI Files. In the case of AMM-1, you should resist any
temptation you may have to use this ability to fool around with
the contents of the mic model files. Corruption of the FLASH
memory by user-mangled MIDI SysEx messages may cause the
AMM-1 to malfunction beyond the ability to operate and totally
bizarre events may occur. Really. We’re not kidding about this.
These problems are usually recoverable, but it will almost
certainly require a call to Customer Support.
Another Note: Pausing the flow of MIDI data in the middle of a
SysEx message can occur with no lost information. Simply
resume playing the file to complete the data transfer. On the
other hand, data missing from the middle of a SysEx message
will cause the remaining models in that SysEx message to be
ignored.
Adding New Mic Models 
						

							35
If FLASH memory becomes full in the middle of a download, the
following message is displayed:
Error, FLASH full.
Press bypass...
Press the Bypass button to return to normal operation. If necessary,
use the Delete/Restore FLASH Models Page and the Compress FLASH
Page to erase enough mics from FLASH to make room for the new
models.
Adding New Mic Models 
						

							36
Chapter 6:
Realistic Expectations (or, AMM-1
Meets the Space-Time Continuum)
Although the AMM-1 seems in many ways to be almost magic, it is,
in fact, simply very clever science. And as such, it remains subject to
those pesky laws of physics.
To get the maximum satisfaction out of the AMM-1, it is important
to have realistic expectations of exactly what it can and can’t do.
(Most of what it can’t do relates to the physical impossibility of
recovering information that wasn’t in the original signal to begin
with.) Here are the main issues to be aware of:
•Choice of Input MicrophoneLuckily for all of us, the general
quality of “affordable” microphones has reached a remarkably
high level. Consequently, if you stick with well-known manufac-
turers, most any reasonable quality mic will provide sufficient
performance to allow the AMM-1 to do its magic (OK, we said it
wasn’t magic, but we’re speaking metaphorically here).
On the other hand, you can’t expect to go into a large (but
unnamed) mass merchandiser of low-cost electronics gear and
pick up a $19.95 mic and expect the AMM-1 to make it sound
like a U87. If a source mic has massive roll-off in a particular
frequency range, there is no way the AMM-1 can produce the
signal that would have been captured had the source mic had
better response.
•Microphone VariationsWhile there are obviously major differ-
ences between various models of microphones, there are also
often more subtle differences between different samples of the
same model of microphone. Whether due to manufacturing
variances, age or condition, there is no guarantee that the mic
we modeled will be identical to your source mic or to a specific
mic you want to model. In the case of some well-known classics,
we have even provided multiple models of the same (but soni-
cally differing) mic from different sources. 
						

							37
•Microphone TechniqueIn getting the best possible recorded
sound, mic technique and placement are at least as important (if
not more so) that mic choice. A good engineer can record a
great track with an SM57 while a poor one can make a U47
sound like doo doo. If your audio is not well-recorded in the first
place, the AMM-1 can to do very little to improve it. If you start
with a poorly recorded track, all the AMM-1 will do is make it
sound like a track that was poorly recorded with a great mic.
•Excessive Frequency BoostAlthough the AMM-1’s processing
does not itself add noise to your signal, any noise in your origi-
nal audio or noise added by intervening processes (e.g., A/D
conversion, pre-AMM-1 dynamics processing, etc.) will be accen-
tuated by any large amount of frequency boost.  This should
only be a problem when your source mic has a substantial bass
or treble roll-off and the modeled mic has a corresponding
boost or, more likely, when your audio was recorded with a low-
cut filter on the source mic and you do not use a low-cut on the
modeled mic. In both of these cases, the models will apply
substantial gain to the affected frequency ranges, raising the
level of added noise along with the desired signal. If the result-
ing noise level is unacceptable, you should choose a different
combination of mics and/or turn on the modeled mic’s low-cut
filter.
•Polar Pattern SelectionThe AMM-1 can’t recover information
that was not recorded as part of the original signal. For ex-
ample, if the original audio was recorded with a highly direc-
tional pattern (hence picking up little room tone), you can’t set
the modeled mic to Omni and expect the room tone that would
have been recorded if the original was set to Omni suddenly to
appear. You can however, simulate that effect with some judi-
cious use of reverb or an environmental simulator.
Conversely, if your source was recorded with an omni mic and it
picked up some unwanted audio from the rear, you can’t realisti-
cally expect to set the modeled mic to hypercardioid and have
the unwanted audio disappear.
•Off-Axis ResponseThe purpose of the Pattern selection is to
model the varying frequency characteristics that result from
each of the available pattern settings, with the assumption that
the audio was recorded on axis (i.e., from the front of the
microphone). Since the AMM-1 has no way of knowing the
actual placement of the signal source, it does not attempt to
model off-axis performance.
Realistic Expectations 
						

							38
•Transient ResponseOne of the key characteristics of various
types of microphones is their transient response (i.e., the way
that their diaphragms respond to extremely rapid amplitude
fluctuations, typically during a sound’s attack phase).
Intuition would suggest that modeling changes in transient
response between mics would be next to impossible - particu-
larly changing a source mic with a slow response to a modeled
mic with a fast response. Amazingly, that turns out not to be the
case. The AMM-1’s models do model variations in transient
response in both directions.
Rather than explain how this is done (which we are disinclined
to do, anyway), we suggest that you demonstrate it to yourself
with the following experiment:
1. In a computer-based waveform editor, use the Pencil Tool (or
equivalent waveform drawing tool) to create a single high-
amplitude spike.
2. Process this file through the AMM-1 with the Source Mic set to
Bypass (so the spike is passed through the Source section un-
changed) and the Modeled Mic set to the ATM31.
3. Examine the processed signal in the waveform editor. You
should see obvious evidence of the smeared transient.
Realistic Expectations     
						

							39
4. Now take that processed file (which is now a model of the spike
as it would have been recorded by the ATM31) and send it
through the AMM-1 again, this time with the Source Mic set to
the ATM31, and the Modeled Mic section set to Bypass. If that
Source Mic model is doing its job, it should actually remove the
characteristics of the ATM31, including that smeared transient.
5. Once again, examine the processed signal in the waveform
editor. The evidence of transient smearing will be gone and the
spike restored to its former spikey self. (Frankly, the first time we
tried it, we could hardly believe it ourselves. Thanks, Dr. Andy.)
Realistic Expectations   
						

							40
Chapter 7: Get Creative
Up to this point, all of the instructions in this manual have focused
on how to use the AMM-1 for its primary purpose: making one mic
sound as accurately as possible like another. But don’t let that limit
you. We’ve purposely given the controls wide ranges to allow you to
move beyond what might be considered useful for strict modeling.
Try some of the following:
•Select a Source Mic that doesn’t match your physical mic. In fact,
try one whose characteristics are as different as possible from
your physical mic.
•Select Bypass for the Source Mic to combine the sound of your
physical mic with the sound of the modeled mic.
•Create a “Hyper” version of one of your source mics. To do this,
select Bypass for the Source Mic (to pass through the sound of
your mic unmodified) and then select your source mic in the
Modeled Mic menu. The result will be to accentuate all of the
characteristics that give your source mic its unique character,
making it sound like itself, only more so.
•Use the AMM-1 as a mastering tool. Process your entire stereo
mix through a mic model (this will, of course, require two
AMM-1s). Set the Source Mic to Bypass and try a variety of
Modeled Mics. This is basically a trial and error process, but a
number of engineers and producers have reported some truly
amazing results.
•Extreme Proximity settings can give strange, but interesting
effects. Try wildly differing settings in the Source and Model
sections.
•Dynamically change Proximity settings during a performance.
•Overdrive the Tube Saturation section for some serious gruzz.
None of the above are likely to give you the sound of any mic that
exists in nature, but they can definitely give your recordings unique
and striking timbres.