今天买杂志拿到了一个绿色的免费软件miditest...非常方便,大家一看就知道怎么用.
点击下载
说明
Release notes MidiTest 2.6
==========================
This is free software. It can be used and distributed freely as long as it is accompanied by this text. It is not
allowed to use this software for commercial purposes without written consent of the author.
The author takes no responsibility for damage resulting from using this software. The software is
provided as is, without any warranties about the proper functioning or useability.
What it is
----------
MidiTest is a small utility that will test the speed and functioning of a MIDI hardware device. It's use is
quite simple. You need a MIDI device that has at least one input and one output. Connect the output to the
input using a standard MIDI cable. Start MidiTest. Select the output and the input with the drop-down lists.
Click the button marked 'Test' and the testing begins. The application will start transmitting MIDI messages
through the output. If all is well the data will be received again through the input. The application
measures the time it takes for the MIDI data to travel through the hardware and presents this data in a
nice overview. The data can then be saved to a text file for future reference.
MidiTest has a few options to control it. You can set the number of messages that will be transmitted. A
normal test run consists of a number of simple MIDI messages and one system exclusive block. Simple MIDI
messages consist of 1, 2 or 3 bytes. MidiTest sends a selection of all possible simple messages.
You can also control the size of the system exclusive block. The data within the system exclusive block is
structured such that, in case of data corruption, it's possible to analyze the nature of the data corruption.
For this purpose the data is divided into blocks of 4 bytes. Each block of 4 bytes represents a counter that
is increased by one with every 4 bytes sent. The counter uses hexadecimal numbering.
The preload setting controls a special feature of DirectMusic. This feature makes it possible to schedule
MIDI data to be sent in the future. You can specify the number of milliseconds the data will be pre-scheduled.
In theory this improves timing jitter. By default this value is set to 0, meaning that data is sent
immediately. Setting the preload to high numbers will increase the time it takes to finish the test
considerably.
Clicking the 'Hide Details' check box in the lower left corner of the application dialog will show all the
actual measurements. When you have this unchecked, these data will also be saved when you save the measurements.
Version 2.5 offers greater control over the types of messages that are used in the test. You can select
categories by ticking the corresponding checkbox. When testing begins, messages are randomly picked from
the selected categories. Not only are they randomly picked, they are also randomly generated.
The test report has also changed somewhat. Average times are now calculated per byte where it used to be
calculated per message. New results may therefore differ from results obtained with previous versions of
MidiTest.
Version 2.6 shows again considerable changes in data presentation. A distinction is now made between per message
and per byte statistics.
Why was this done? Because in some cases it makes sense to do so. MIDI is a serial protocol which means that data is
transmitted sequentially, byte after byte. It takes time to transmit one byte. With a standard MIDI connection this
is 320 microseconds to be exact. But MIDI messages are not all the same size. A MIDI messages can consist of 1, 2 or 3
bytes. So to transmit a 3 byte message takes 3 times as much time as a 1 byte message.
To complicate things even more this doesn't affect all devices in the same way. For a MIDI device that connects to the
USB port the difference in size between MIDI messages is not all that relevant. Even though USB is a serial connection
too, data is always transferred in packets. Any sensible driver design for such a device will always send a complete
MIDI message in one packet. Therefor the size of the MIDI message will not affect the measured latency time so much.
In these cases the per byte statistics can safely be ignored because they don't give any relevant information. They
can even be misleading.
This is different for MIDI devices that connect to the parallel port or the serial port. In these cases it is not
unusual that the MIDI messages are transferred serially to the device. In such a case the per byte statistics might be
relevant (all depending on the driver design).
A further addition to the statistics are measurements for latency, jitter and maximum jitter. In my view these figures
are the most important of the MIDI test. They can be seen as representative for the tested device.
Let me explain what they mean. To start with latency, this is in fact nothing other than the average roundtrip time of
a message (or byte). You will see that it is equal to the average time in the SND+RCV column.
Latency in the MIDI data path is not really a problem and in fact inevitable. Because of the serial nature of a MIDI
connection it will always take a certain amount of time to transmit a MIDI message. It becomes a problem when the
latency is not constant, but varying. The human brain is capable of compensating for constant latency to a certain
extent. Sound itself takes a certain amount of time to travel so sounds coming from a distance we hear a little later
than they were produced. The brain has however much more difficulty with varying latency. In the context of MIDI this
can be perceived as unsteady timing.
This variance in latency is expressed with the jitter value. In statistical jargon this calculated number is known as
the standard deviation. In the context of MidiTest it tells us how much the measured latency averagely differs from the
average latency. If this figure is small then all messages have about the same latency. The resulting MIDI timing will
be perceived as tight. If this figure is very large than the resulting MIDI timing will be perceived as sloppy and
not very usefull in a musical context.
The last added informative figure is the maximum jitter. This tells us the maximum difference between a measured latency
and the average latency. Because calculating the average has a tendency to even out the extremes it can sometimes be
usefull to have a look at extreme values. Even if the jitter value of a MIDI device seems reasonably small, one or two
of these extreme latency values can still make the MIDI device impractical for musical use. In music we want all the
notes to be on time. One (random) off note can ruin it.
About me
--------
Although this is free software you are welcome to make a donation. I can be contacted through this email
address: <
miditest@earthvegaconnection.com>. I am also interested to receive test results. I will put them
up on the web-site so people can compare the performance of various devices.
Visit my web-site
http://earthvegaconnection.com
Evert van der Poll
Amsterdam
The Netherlands