*the best midi interface search... can you help?

[space2012]

Hi,

need help to find the lowest latency, low jitter MIDI interface, ever made.
i got curiosity from reading this post: "keeps spining in my head"
http://www.gearslutz.com/board/3624005-post2.html

Methodology:
the midi loop test, without feedback.
to avoid creating a midi feedback: just select none out midi port in the second midi recording track.

found some strange things in USB1.0,
and diferent perfomance between usb2.0 chips.
http://www.gearslutz.com/board/3639800-post17.html

so far the best is RME hdsp9632 pci midi, has 61-68 samples of latency.
61 - 68 means: if i record the same midi pattern in loop, measure the sample latency, erase it, and record again, measure again, erase and record again, sometimes latency its 61 and sometimes its 68 samples. sometimes between both aka. jitter.
with a generic PSU, old pc.

is there something better?
please post your Midi Loop Test results, including your computer, if you have an aditional pci to usb card, what brand and model is the chip, what brand and model are the cables you used, midi and usb cables, ... any information.
if someone post the midi loop test results of midiman8x8 or motu MTP AV USB for example, thats not definitive measurements becouse the variables like USB chip, usb cable, pc, OS, PSU, etc...
if you have same interface please do the test, to see if your system performs better or worse in the midi loop test, and post your results with all the variables.

Midi Tricks:
digi001 Asio and midi drivers v6.4 can be used in WinXP MCE sp2, with 975x boards, but... midi needs to be active all the time. "activate the send midi clock and MMC out" and that keeps the midi drivers from sleep/shutdown/automatic self-disable.

[space2012]

tested a few USB2.0 pci cards,
none can match the speed of included USB2.0 in Asus P5W DH Deluxe motherboard, so far.
the best ive tested was the NEC smaller chip,
and closer the PCI slot to the CPU gives better USB2.0 performance,

Asus P5W DH Deluxe with intel 975X & ich7r has 34MB/s to 35MB/s Read, measured in HDTach3.0.4
NEC D70****F1/GD chip has 22MB/s
ALi M5271 has 18MB/s
Via 5, didnt worked in my old PC

The MB/s read transfer affects the MIDI latency,
the best i could get was 278samples of latency with the NEC chip.

44100 samples - 1000milliseconds.
x samples - y ms.

RME HDSP9632 61-68samples = >1.38322 ms. in Cubase loop test.

also found two nice web sites:
Download MidiTest v4.6.231:
MidiTest
Results:
MidiTest
others:
8 Port SE Drivers
Solving Directmusic problems in SX/Nuendo

old pc, compaq 4700 p4 1.6ghz
--------------------------------------------------------------------------
MidiTest Results: RME HDSP9632 pci
--------------------------------------------------------------------------
DirectMusic with TimeStamp active:
Message latency: 0.86 ms
Message jitter: 0.22 ms
Message max deviation: 0.51 ms
same DirectMusic Test #2:
Message latency: 0.57 ms
Message jitter: 0.27 ms
Message max deviation: 0.65 ms
Emulated:
Message latency: 6.01 ms
Message jitter: 0.26 ms
Message max deviation: 0.53 ms
MME:
Message latency: 0.58 ms
Message jitter: 2.77 ms
Message max deviation: 253.67 ms
Kernel Steaming:
Message latency: 0.90 ms
Message jitter: 0.99 ms
Message max deviation: 40.64 ms

--------------------------------------------------------------------------
MidiTest Results: Novation ReMOTE 61 SL with NEC USB2.0 chip - Standard USB Modem Cable - far/last pci slot from CPU
--------------------------------------------------------------------------
Emulated: test #1
Message latency: 10.37 ms
Message jitter: 0.49 ms
Message max deviation: 1.07 ms
Emulated: test #2 in Play mode.
Message latency: 6.67 ms
Message jitter: 0.42 ms
Message max deviation: 1.32 ms
MME:
Message latency: 4.91 ms
Message jitter: 1.22 ms
Message max deviation: 6.11 ms

[ekwipt]

Wow hat off mate, I've been reading your posts for the last few weeks and you really seem to know your stuff.

I'm only on Mac, osx with a macpro, is there a osx version of the test, or are you only worried about pc

[space2012]

Originally Posted by ekwipt

I'm only on Mac, osx with a macpro, is there a osx version of the test

you should write to: miditest @ earthvegaconnection.com
asking for MAC OSX support for MidiTest.

but you can make the midi loop test, in any computer/software that has midi I/O, and measure or post the .mid. with the looped midi signal.

[space2012]

lpuma
Lives for gear



Join Date: Oct 2005
Posts: 1,062


Re: can you test the Profire 2626 MIDI interface ?
Quote:
Originally Posted by space2012
http://www.gearslutz.com/board/gear-...-you-help.html

I dont know if the results will be that revelant to your thread:My pc last xp install is pretty old and I'm connected to the web with only free antivirus stuff. I beleive mt PC is far from being a good reference. It's a P4 2.4 + my midi cable is stait junk, a 10 years old low cost midi cable from the hifi store. My firwire connection is a pci card with TI chipset, the driver is the latest vista version (see th "error 10" thread in M audio forums), the cable is the one that comes with the 2626

I give you the result:

Message latency: 11.43 ms
Message jitter: 1.32 ms
Message max deviation: 2.46 ms

you can add this to your thread if you really feel the need.

Peace.

[space2012]

REVIEW: Ploytec GM5 USB Midi Interface


Ploytec GM5 USB Midi Interface MME:
Message latency: 0.39 ms
Message jitter: 0.41 ms
Message max deviation: 1.34 ms

Midex 8 with Directmusic Driver:
Message latency: 0.77 ms
Message jitter: 0.88 ms
Message max deviation: 3.49 ms

Midex 8 with MME Driver:
Message latency: 1.02 ms
Message jitter: 1.04 ms
Message max deviation: 3.18 ms

Midi sport 1x1 MME Driver:
Message latency: 2.18 ms
Message jitter: 1.32 ms
Message max deviation: 3.77 ms

[Jay-]

Midi sport 2x2 rocks because it has a thru button for switching from kit a to kit b.
Wish I never sold that one.

midisport 4x4 sucked bag..

[space2012]

reading lots of motherboard reviews, found that:
AMD 790GX / SB750 chipset is better for stock 3D graphics.
BUT....
Nvidia nForce 780a has much better USB2.0 performance measured with HDtach.
overclock is other story...
also that 780i beats some intel chips too: p35 and x35 in USB2.0 performance.
also that MB circuit design change a bit those numbers, diferent MB manufacturers with same chipset.!
nvidia 8300 chip similar USB2.0 performance as nForce 750i, 21.5MB/s
the SB600 gives in one test 21.5MB/s in other 28MB/s = strange.

the best USB2.0 chips have tranfer rate of 35MB/s measured with simplysoftware HDTach3.0.4.

Bjorn3d.com -Asus M3N78 PRO, Satisfying Your Daily Tech Cravings Since 1996
bit-tech.net | Review - Gigabyte GA-MA790GP-DS4H

[gl3ny]

Hi... I'm having what I consider to be a severe midi jitter problem and have conducted lots of tests recently. Before I post my results and plead for help I would like to get on same page as you guys.

I have been running the standard midi loop test and understand what jitter and latency mean but I don't know what is being talked about when it come to 'Message Deviation'. Can someone please explain?

[space2012]

Originally Posted by gl3ny

Can someone please explain?

Message = good , latency = bad
Message = good , jitter = bad
Message = good , deviation = bad

[gl3ny]

Originally Posted by space2012

Message = good , latency = bad
Message = good , jitter = bad
Message = good , deviation = bad

I like it but was looking for more of how max deviation is calculated?

[qtuner]

ADK proaudio laptop 2.2ghz core 2 duo. USB chipset unknown.

miditest results motu MTP AV USB:

DirectMusic
Message latency: 13.14ms
Message Jitter: 0.95ms
Message Max deviation: 4.46ms

MME
Message latency: 7.41ms
Message Jitter: 1.14ms
Message max Deviation: 4.52ms

What is interesting is that i kept getting sysex errors during the test. I could only get things to work if i just used midi messages from the first column. A few other messages worked, but I didn't use them in this test as i found they caused a feedback loop. I tried playing with the clockworks application for filtering, but it didn't do much. I couldn't get the test to work with sysex or realtime messages. From the log file, it looked like the timepiece was dropping messages.

I ran the test on all the usb ports of my laptop and they're all about the same

[qtuner]

Korg Padkontrol on ADK SR 1Q 2.2 ghz core duo. Same System as above

MME Driver
Message latency: 3.09ms
Message Jitter: 1.52 ms
Message Max Deviation: 8.53ms

[space2012]

Originally Posted by qtuner

ADK proaudio laptop 2.2ghz core 2 duo. USB chipset unknown.
I ran the test on all the usb ports of my laptop and they're all about the same

ADK 1Q MV has Intel® PM45 Chipset + ICH9M, when click "?"
ADK Pro Audio PC, Laptops, Digital Audio workstations for Nuendo, Cubase, Sonar, Samplitude
Configure your ADK Gaming Rig
small letters:
Benchmarks for Pro Audio! Nuendo, Sonar, Cubase tested with AMD Opterons, Dual Xeon, Dual Core AMD X2, Dual Core Pentium D, Dual Core Opteron.

[qtuner]

I have the SR version not the MV version. Santa rosa chipset...whatever that is.

[space2012]

TSI is a Windows application that shows detailed version information about USB and IEEE 1394 bus driver stacks used in a Windows system.
The TSI tool is useful to analyze problems caused by third-party bus driver stacks.

Free Download Tools

[qtuner]

Intel(R) ICH8 Family USB Universal Host Controller

I need to test my sl61 with miditest so we can compare systems.

[gl3ny]

Originally Posted by qtuner

ADK proaudio laptop 2.2ghz core 2 duo. USB chipset unknown.

miditest results motu MTP AV USB:

DirectMusic
Message latency: 13.14ms
Message Jitter: 0.95ms
Message Max deviation: 4.46ms

MME
Message latency: 7.41ms
Message Jitter: 1.14ms
Message max Deviation: 4.52ms

What is interesting is that i kept getting sysex errors during the test. I could only get things to work if i just used midi messages from the first column. A few other messages worked, but I didn't use them in this test as i found they caused a feedback loop. I tried playing with the clockworks application for filtering, but it didn't do much. I couldn't get the test to work with sysex or realtime messages. From the log file, it looked like the timepiece was dropping messages.

I ran the test on all the usb ports of my laptop and they're all about the same

I have MTPAV USB also but I couldn't get it to work with MidiTest at all.

Keep getting a message saying 'No Midi data received! Check if out put is connected to input'

I can get it to work with a HDSPe RayDat. Any chance you can post a screen shot with the settings?

[jeamsler]

AudioPCI Card/1370 chip

Message Latency: .08ms
Message Jitter: .10ms
Message Max Deviation: .98ms

Average Time Send: .02ms
Average Time Receive: .06ms
Average Time Send/Rcv: .08ms

One message out of 31000+ > 1ms

This and a Lynx One are best testing midi interfaces I've ever seen.
AudioPCI is cheap to be had on ebay.

[gl3ny]

I tested my HPSPe RayDat on an Intel DP35DP Quad Xeon with midi quest and the jitter and latency results look awsome to me. I did get a sysex error though so if anyone can help me with that it would be good.

I also test the RayDat in Cubase 4 using the midi loop test technique because that is closer to a real world situation.

Out of 100 quantized note events spat out by Cubase to a RayDat midi port and recorded back in on a RayDat Midi port:
Jitter = 0 - 2.0ms

Peak latency = 3.0ms
Best Latency = 1.0ms
Average Latency = 1.52ms
BTW many notes had a jitter of 0ms

MTPAV with same Cubase loop test:
Jitter = 2 - 6ms

Peak latency = 13.0ms
Best Latency = 7.0ms
Average Latency = 10.3ms

Conclusion MTPAV hardware is at fault or drivers are shit!!!!! Anyone know how I can test this?

--------------------------------------------------------------------------
MidiTest Results
--------------------------------------------------------------------------

================ Info ====================================================
Date: 23 Jan 2009
Time: 09:58:00
AppVersion: 4.6.231
OS: Microsoft Windows XP Professional, Service Pack 3 (Build 2600)
Processor(s): Intel(R) Xeon(R) CPU X3360 @ 2.83GHz
Speed: 2833 MHz
Number: 4
API: DirectMusic
Test type: Simple
Use timestamp: yes
Errors: Error in system exclusive message nr. 1 at position 4800. Expecting 00 04 0A 0F, received 00 04 0E xx.

================ Tested Message Types ====================================
Note off: yes
Note on: yes
Key aftertouch: yes
Controller: yes
Program change: yes
Channel aftertouch: yes
Pitchbend: yes
System exclusive: yes
MIDI time code quarter frame: no
Song position pointer: no
Song select: no
Tune request: no
MIDI clock: no
MIDI tick: no
Start: no
Continue: no
Stop: no
Active sensing: no
System reset: no
System exclusive mixed with realtime messages: no

================ Ports ===================================================
MIDI Output: RayDAT Midi Port 2
Description: RME HDSPe RayDAT
Provider: RME
DriverDate: 1-7-2009
DriverVersion: 3.0.7.4

MIDI Input: RayDAT Midi Port 2
Description: RME HDSPe RayDAT
Provider: RME
DriverDate: 1-7-2009
DriverVersion: 3.0.7.4

================ Results Per Message =====================================
MESSAGES Snd Rcv Snd+Rcv
Message TotalTime: 994.79 ms 26122.60 ms 27117.39 ms
Message MaximumTime: 0.10 ms 0.95 ms 0.99 ms
Message MinimumTime: 0.02 ms 0.58 ms 0.62 ms
Message AverageTime: 0.03 ms 0.84 ms 0.87 ms
SysexTime: 4.35 ms 1.13 ms 5.48 ms
SysexAverage: 0.00 ms 0.00 ms 0.00 ms
< 1 ms: 31250 31250 31250
1 - 2 ms: 0 0 0
2 - 3 ms: 0 0 0
3 - 4 ms: 0 0 0
4 - 5 ms: 0 0 0
5 - 10 ms: 0 0 0
10 - 20 ms: 0 0 0
20 - 50 ms: 0 0 0
50 - 100 ms: 0 0 0
> 100 ms: 0 0 0
Message count: 31250 Sysex count: 1
Sysex size: 9999 Sysex passed: 4802
Message latency: 0.87 ms Total time: 310.285 sec
Message jitter: 0.15 ms
Message max deviation: 0.25 ms

================ Results Per Byte ========================================
BYTES
Byte TotalTime: 10000.69 ms
Byte MaximumTime: 0.34 ms
Byte MinimumTime: 0.31 ms
Byte AverageTime: 0.32 ms
< 1 ms: 31250
1 - 2 ms: 0
2 - 3 ms: 0
3 - 4 ms: 0
4 - 5 ms: 0
5 - 10 ms: 0
10 - 20 ms: 0
20 - 50 ms: 0
50 - 100 ms: 0
> 100 ms: 0
Byte count: 84727
Byte latency: 0.32 ms
Byte jitter: 0.00 ms
Byte max deviation: 0.02 ms

[qtuner]

Originally Posted by gl3ny

I have MTPAV USB also but I couldn't get it to work with MidiTest at all.

Keep getting a message saying 'No Midi data received! Check if out put is connected to input'

I can get it to work with a HDSPe RayDat. Any chance you can post a screen shot with the settings?

uncheck all the messages and just select note on/note off. If you check the log file that is produced you'll see sysex messages get corrupted as well as a few other messages. As soon as you get that working, add more messages. I've never been able to get sysex to work with my motu atp av. I see from your test results that you have sysex checked. Uncheck that and miditest should work.

I will try to get a screen shot tomorrow.

I have an Edirol UM-880 on the way. i will post those tests soon.

[gl3ny]

Originally Posted by qtuner

uncheck all the messages and just select note on/note off. If you check the log file that is produced you'll see sysex messages get corrupted as well as a few other messages. As soon as you get that working, add more messages. I've never been able to get sysex to work with my motu atp av. I see from your test results that you have sysex checked. Uncheck that and miditest should work.

I will try to get a screen shot tomorrow.

I have an Edirol UM-880 on the way. i will post those tests soon.

Don't worry about the screen shot I got it working, thank you!

I tried unchecking everything except note on/off in an earlier attempt but that didn't work before for some reason. Motu told me to perform a factory reset by powering up holding the panic button so I wonder if that fixed it. Anyway thanks again!

[qtuner]

does that change the test results?

I got the impression that the motu MtP av sucks too from miditest. I think it the has to do with the physical implementation of the unit.

[gl3ny]

Below are the MidiTest results from a my Motu MTPAV. I have a big problem with these results. To me jitter is the difference between the highest receive time and the lowest receive time. In this case it would be 10.87- 6.68 = 4.19ms. Yet below jitter is stated as 0.65ms. I've been performing manual midi loop tests in cubase for 3 weeks now and the numbers I see doing this manual test is more in line with 4ms aboveand not 0.65ms.

If people are running this test and seeing 0.65ms and think their jitter is good, think again. I promise you my jitter is definately not 0.65ms. I have been running all the numbers manually in an excel sheet taking the values from the list editor. I can even see the notes are not consistent by looking in the Key edit window.

Am I loosing the plot here? Can somebody please explain how the 0.65ms is logically calculated?

================ Results Per Message =====================================
MESSAGES Snd Rcv Snd+Rcv
Message TotalTime: 761.60 ms 263668.89 ms 264430.49 ms
Message MaximumTime: 0.09 ms 10.87 ms 10.90 ms
Message MinimumTime: 0.02 ms 6.68 ms 6.71 ms
Message AverageTime: 0.02 ms 8.44 ms 8.46 ms
SysexTime: 0.00 ms 0.00 ms 0.00 ms
SysexAverage: 0.00 ms 0.00 ms 0.00 ms
< 1 ms: 31250 0 0
1 - 2 ms: 0 0 0
2 - 3 ms: 0 0 0
3 - 4 ms: 0 0 0
4 - 5 ms: 0 0 0
5 - 10 ms: 0 31092 31086
10 - 20 ms: 0 158 164
20 - 50 ms: 0 0 0
50 - 100 ms: 0 0 0
> 100 ms: 0 0 0
Message count: 31250 Sysex count: 0
Sysex size: 0 Sysex passed: 0
Message latency: 8.46 ms Total time: 318.480 sec
Message jitter: 0.65 ms
Message max deviation: 2.44 ms

[jeamsler]

gl3ny

Jitter should not be interpreted as the difference between the max and min deviation from norm. Those would be extreme values and represent the worst case. Jitter is normally interpreted as the average deviation over time from the expected value. Miditest states in the readme file that it calculates a jitter value using the standard deviation formula. So what it gives you in the results is the average time and then the standard deviation for that average. The standard deviation tells you the average variance from the expected average value: in effect the average change from the average. So what your results tell you is although your receiving latency is quite high, the jitter or deviation from this latency is pretty good. Most of your results are between 5-10ms and about .5% are between 10-20ms with an average variance of .65ms. So most of the time things are stable if slow with occasional peaks.

Having said all of that it should be remembered that midi interfaces are only one part of the chain. Midi sequencers can introduce just as much or even more variance as midi interfaces on either playback or recording. And of course if you are playing hardware synths then you have latency and jitter associated with the hardware. So getting a good interface is only the beginning. The app you are using may be messing things up as much.

Jon

[qtuner]

I'm hesitant to introduce a sequencer into testing because I believe sequencer timing is independent of hardware timing. Since PC software is layered, and calling standardized APIs I think we can assume hardware performance as independent from sequencer performance. I do agree that sequencer latency is part of the sum of total latency. Is this assumption correct?

In documenting these, I think it would also be helpful to identify for PC which Timer each interface uses.

I saw a tool last week that would test drift and identify which driver to use. Does anyone know what I'm talking about? I can't find it now.

the app is miditime.exe
found here
Parapoetica: PC MIDI Timing and Nuendo

[gl3ny]

Jon,
I truly appreciate you pointing this out to me and now it all makes sense. I didn’t think of looking at the readme file.

While I appreciate what you are saying I don’t understand why MidiTest would want to calculate using the standard deviation formula. I can only guess that it is a standard used by manufacturers of a whole host of products to calculate jitter for the purpose of providing a specification. I’m sure you are aware of how clever manufacturers like to be when it comes down to hiding numbers or true values to sell their products. MidiTest quote on their website, “MidiTest is a small utility that will test the speed and functioning of a MIDI hardware device.” That being said there should be no reason they can’t show what I am going to call ‘true jitter’ since they don’t manufacture midi interfaces.

Now that I understand MidiTest is using the standard deviation formula to calculate midi jitter, in my mind it has only strengthened the validity of the Midi Loop Test for showing true jitter.

In general terms I don’t want to knock MidiTest, I still think it’s a great app. MidiTest is basically testing your computer system with the midi interface, and it is fast and has some very useful information. However I do want to knock MidiTest as a viable tool to measure jitter.

Originally Posted by qtuner

I'm hesitant to introduce a sequencer into testing because I believe sequencer timing is independent of hardware timing. Since PC software is layered, and calling standardized APIs I think we can assume hardware performance as independent from sequencer performance. I do agree that sequencer latency is part of the sum of total latency. Is this assumption correct?

Whether this assumption is correct or not, I don’t know? But what I do know is that with a Midi Loop Test, the computer system, midi interface and the host sequencer are being tested and that is closer to a real world situation than with MidiTest. The real numbers that show when data is being received is right there in the sequencer software. After all your midi sequencer makes music, MidiTest does not.

Lets get one thing clear, I am not out to say anyone is right or wrong. To put it simply I believe I am experiencing a midi jitter problem and I’m on a mission to correct it, and that is all! Some people might think that I don’t really have a problem and that is fine everyone is entitled to an opinion. I am in contact with Steinberg and Motu regarding this problem and from the data I have provided to both parties, they agree that this should not be happing.

What matters to me is the data! People can make up their own mind if they have a problem or not.

Here are the MidiTest results ran on a Motu Midi Timepiece AV

================ Results Per Message =====================================

MESSAGES Snd Rcv Snd+Rcv

Message TotalTime: 2067.96 ms 208853.41 ms 210921.38 ms
Message MaximumTime: 0.11 ms 9.69 ms 9.72 ms
Message MinimumTime: 0.02 ms 4.77 ms 4.84 ms
Message AverageTime: 0.07 ms 6.68 ms 6.75 ms
SysexTime: 0.00 ms 0.00 ms 0.00 ms
SysexAverage: 0.00 ms 0.00 ms 0.00 ms

< 1 ms: 31250 0 0
1 - 2 ms: 0 0 0
2 - 3 ms: 0 0 0
3 - 4 ms: 0 0 0
4 - 5 ms: 0 14 10
5 - 10 ms: 0 31236 31240
10 - 20 ms: 0 0 0
20 - 50 ms: 0 0 0
50 - 100 ms: 0 0 0
> 100 ms: 0 0 0

Message count: 31250 Sysex count: 0
Sysex size: 0 Sysex passed: 0

Message latency: 6.75 ms Total time: 268.249 sec
Message jitter: 0.75 ms
Message max deviation: 2.97 ms


This is how I would calculate the data presented in MidiTest based on the same calculation used in a Midi Loop Test:

Message MaximumTime: 9.72 ms
Message MinimumTime: 4.84 ms

9.72 - 4.84 = 4.88 Message Jitter or ‘true jitter’. No fancy calculation here.



Midi Loop Test I performed consisted of 10 quantized midi note events with a length of 1/32, one on each beat lasting for 2.5 bars (1.1.1.0 to 3.2.1.0). Midi events output on MTPAV port 8 and recorded in MTPAV port 8. This was repeated 10 times giving a total of 100 note on/off events.

Here are the results of a Midi Loop Test performed immediately after the above MidiTest results.

Test 1
Message MaximumTime: 13.0 ms
Message MinimumTime: 9.0 ms
True Jitter = 4.0 ms
Test 2
Message MaximumTime: 13.0 ms
Message MinimumTime: 9.0 ms
True Jitter = 4.0 ms
Test 3
Message MaximumTime: 12.0 ms
Message MinimumTime: 7.0 ms
True Jitter = 5.0 ms
Test 4
Message MaximumTime: 12.0 ms
Message MinimumTime: 10.0 ms
True Jitter = 2.0 ms
Test 5
Message MaximumTime: 14.0 ms
Message MinimumTime: 9.0 ms
True Jitter = 5.0 ms
Test 6
Message MaximumTime: 11.0 ms
Message MinimumTime: 8.0 ms
True Jitter = 3.0 ms
Test 7
Message MaximumTime: 12.0 ms
Message MinimumTime: 8.0 ms
True Jitter = 4.0 ms
Test 8
Message MaximumTime: 12.0 ms
Message MinimumTime: 10.0 ms
True Jitter = 2.0 ms
Test 9
Message MaximumTime: 11.0 ms
Message MinimumTime: 7.0 ms
True Jitter = 4.0 ms
Test 10
Message MaximumTime: 11.0 ms
Message MinimumTime: 9.0 ms
True Jitter = 2.0 ms




If I run an average of each true jitter value that would be 3.5 ms. I have actually been generous by showing one of the tests with better results. But if make an average of the 19 total test averages just like the test above then it comes out to 3.99 ms. Wait a minute that is better than MidiTest. Could it be MidiTest itself is introducing a slight amount of true jitter, more than my sequencer?

My point is, it doesn’t matter how you want to manipulate the numbers using standard deviation or what ever formula you want to use. The bottom line is my midi notes are being recorded in my sequencer randomly within a 5ms window on top of any latency.

Originally Posted by jeamsler

Jitter should not be interpreted as the difference between the max and min deviation from norm. Those would be extreme values and represent the worst case.

I strongly disagree and believe that jitter should be interpreted by taking the max and min deviation from the norm and leave your representation of jitter should be left as a calculation for the manufactures. Those extreme values I’m afraid are a reality and there is no getting away from that.

This random 5ms window is not the full story either. The 5ms windows is a combination of the input and output jitter so each stage we could say is 2.5ms so lets look at this in a practical situation. Let’s say I want to capture a keyboard performance via midi and at a later stage play back the midi part and record it as audio.

Midi Recording - Keyboard performance > sequencer
2.5ms produced by Keyboard midi out
2.5ms produced by Midi interface input and sequencer.

Midi Playback
2.5ms produced by Sequencer + midi interface output
2.5ms produced by keyboard midi input

Keyboard generates sound and result is recorded as audio with a 10ms jitter. I think it’s widely accepted that the performance of a competent musician can be lost with anything over 5ms therefore the jitter is unacceptable to me.

My RayDat on the other hand would produce 4ms using the same calculations which would be acceptable. However keyboard jitter is not the issue here but could be measured after establishing a base line with your midi interface.

My conclusion is that MidiTest is most definitely not a valid tool for testing midi jitter. I advise anyone who is looking to purchase a midi interface and who is concerned with jitter not buy one on the strength of the results produced by MidiTest.

And for anyone who is looking at latency alone when choosing a midi interface. Just remember this, jitter is random latency!

[qtuner]

woah... we're all on the same team. I'm trying to put together a miditest, and i'm just trying to figure out an easy way to test a lot of interfaces in a short amount of time. We all have the same problem....Crappy midi. We should all bond over this.

So...I just did a refresher on college stats, and standard deviation is not a bad way to describe the variance with a set of data points. What these jitter measurement from miditest means (i think) is that 68% of the measured data points vary by an average of .75 ms from the average of the measured times. 99% of your midi messages vary by 2.25ms from the average of the message time. This is just how standard deviation works.

I am working from the assumption that the measurements are the sum of bus latency and jitter. We cant get an exact measure of latency, so we average the message time. From there miditest uses standard deviation to get a variance measurement of the data points. again 68% of the messages have an average vary from the average latency by .75ms.

You're assumption that maximum jitter equals jitter is probably not accurate. I'm trying very hard not to be offensive.

I want to reduce both latency and jitter. I think the motu atp has a crappy latency and crappy jitter.

I friend of mine actually rates the attractiveness of women in std deviations.

[gl3ny]

All I’m concerned about and I’m sure people most people who use sequencer are concerned about is the reality of how far their midi timing is off in real terms rather than some formula to make it look better than it really is.

Really come on! LoL surely you can see my point? What does 0.75ms tell me? It doesn’t mean jack to me when in reality the timing can be off by up to 5ms. Where is the correlation there for the average user? So if you want to use the standard deviation, that’s cool! If I was in the business of selling midi interfaces then ok I would use the standard deviation method.

All I wanted to point out is that MidiTest is grossly misleading when it comes to its message jitter report just like a manufacturer would be and I think I accomplished that.

I won’t mention midi jitter again because that seems to be a standard and it’s always been measured that way. I’ll don’t know if there is a name for how I have calculated my problem but if there isn’t I’d like to reserve the name ‘midi true jitter’ or ‘midi titter’ for short.

I want you to know I’m not taking offense here. I actually think it is a good discussion. We do need to work together but we need to be on the same page with what is a good number and what is not. Sorry but 0.75 mean nothing to me so. Can anyone come up with a formula to convert jitter to titter?

Seriously now,

Originally Posted by qtuner

In documenting these, I think it would also be helpful to identify for PC which Timer each interface uses.

I saw a tool last week that would test drift and identify which driver to use. Does anyone know what I'm talking about? I can't find it now.

the app is miditime.exe
found here
Parapoetica: PC MIDI Timing and Nuendo

That is a good reference for anyone at the beginning of trouble shooting a midi interface even though it’s a bit old there is much that is still valid.

One thing I got out of it that article I think deserves some attention is DirectMusic drivers. I’d be interested to know if anyone has any kind of jitter, titter or latency problem with an interface that uses these DM drivers?

My RayDat uses DM drivers and it is extremely low titter @0.9ms (0.15ms Midi Jitter).

[gl3ny]

Originally Posted by qtuner

I want to reduce both latency and jitter. I think the motu atp has a crappy latency and crappy jitter.

I friend of mine actually rates the attractiveness of women in std deviations.

Honestly, I think we are both at the end of the line when it comes to getting any better performance out of the motu MTPAV. Also I wouldn't even bother testing any of those interfaces that don't support DirectMusic drivers.

One good thing that has come out of this problem is that I have a DAW which was fast to begin with is now lightening fast and extremely responsive beacuse of all the tweaks I've tried.

Can I ask you what your MidiTest message jitter value is and what you hope to acheive or find acceptable?

I bet your friend make his final judgement based on what he see