I am trying to test my MIDI timing and have read about MIDItime, a MIDI timing test application, but I can't find an .exe for it.

Good question might be is why are you asking this. Do you suspect something.  (Because if you do you are probably right!!)
 
You are not clear here.  What midi timing are you trying to test?  Are you talking about midi jitter during live performances between your midi controller and a VST or hardware synth for example.  Or are you talking about how well a DAW records midi timing and plays it back.  Two very different situations.
 
I have spent some time in both of these areas and found out some interesting things especially with several DAW's.  They all handle this a little differently.  Oh BTW Sonar fairs poorly in this area.  Sorry to say it but it is true.  For all the amazing midi features Sonar has it can have poor timing during playback.  (The most important thing really!!)  Studio One excels in this area and it is one of the main reasons I use it.  But you have to be someone who uses a lot of midi in their productions and external hardware midi devices too.  Not VST's so much.  Not a common situation these days.  It also depends what port you have your midi interface connected to on your computer as well.
 
If its Midi jitter you are concerned about check this article out:
 
http://www.eigenzone.org/2012/12/04/midi-jitter
 
Interesting when I ran this test on Studio One I got a perfect result. Apparently this was fixed in V2.5 I have not tested Sonar as yet so not sure what its Midi jitter performance is like. (I only have Sonar 8.5 anyway) Midi jitter does NOT relate to how well a DAW records and plays back midi timing.  That is a separate issue.
 
If you think there is an app out there I will look for it as well Mike.  Always interested in these things too.
 
You might want to try this test:
 
http://www.innerclocksystems.com/new%20ics%20litmus.html
 
 
 
 

I just ran the litumus test on Sonar and Studio One and they both did very well.  But this does not tell you anything about how midi timing is effected under heavy audio load for example. This is where some DAW's really can differ.

Since you've done the test, would you share the results?

Hi Jeff,
 I was playing a soft synth yesterday and saw some discrepancies between what I was playing and what was written to the recorded track. At first it was because I had a overlooked making an appropriate latency compensation setting, but making the proper choices in that regard did not leave me with 100% confidence in the timing as I recorded the parts again.
 
 Afterwards I was reading about testing MIDI sync with a test app known as MIDItime.
 
 I've always felt comfortable that playback is fairly accurate, but I often wonder about the accuracy of recording, and yesterday it seemed like there was room for improvement.
 
 I was using a Roland PC-200 keyboard controller through a MOTU MIDI Express XT. My audio buffer is 64 samples.
 
 I've been reading about how different DAWs and MIDI drivers may use different clock sources and so I was hoping to run some tests with MIDI time to compare choices of Timestamp preferences.  

ampfixer
Since you've done the test, would you share the results?

Hi John,
 FWIW, Results from these type of tests are inherently circumstantial to your system, the MIDI drivers, and the sample buffers you are running at.
 Knowledge of one person's test results will not tell you much about what you should expect from your system.
 
 I don't want to discourage you from asking, nor Jeff from sharing, but it seems like this should be mentioned as reminder that the most important thing is to know your own system and how it reacts to loads and changes in buffer size.

That test as well also does not factor in any variations in the latency of the sound generating device.  You need a sharp click type sound but there may be some subtle timing variations too in the time it takes for the hardware device to generate the sound.  (I could be wrong on that too, just thinking aloud)
 
Litumus results were that both programs were creating the sounds within 50 samples or so.  This is relating to variations in only 1 to 2ms something I doubt you would hear or feel too much.  Sounds like what Mike was experiencing could be other things.  Timing starts to get felt when the errors start heading up towards 10 mS or higher. (under in some circumstances too)
 
Mike did you have any heavy audio processing going on eg heavy latency plugins being used while you were working with your midi track. That seems to effect Sonar.  Not so in Studio One though especially with my midi interface being on my serial port.  Heavy audio processing has no effect on midi timing in my system.  Other Studio One users have experienced issues with VST's though. (with the midi interface also being on USB as well as audio)  But I think that is more to do with system latency in general.  I am so glad my midi interface is on its own port and away from everything else.  It is obvious to me that really seems to work.
 
If heavy audio processing is effecting midi timing one workaround is to switch it all off while you record midi data into your system. Remember I am talking external midi hardware only and the performance I am experiencing is related to that. When I am addressing VST's I turn all heavy plugins off and set the systenm latency right down low as well. Then when I am happy with the timing and feel I put it the system latency back to normal mix settings and continue on etc.. And then convert the VST to audio pretty quickly as well to relieve the CPU hit.

Mike I came across this the other day. You may know about it but I have posted the link in case you have not come across this article. Interesting.
 
http://www.soundonsound.c...es/cubasetech_1207.htm
 
It does relate to Cubase of course but you did mention you installed it the other day. Try some of the things they talk about in this article and see if you get the same result.
 
They do mention things like if you are running lots of audio through a USB interface and you have got the midi interface on the same USB port midi things can go a bit out under certain conditions.
 
I am very lucky here because I am using a rock solid midi interface (Unitor 8) and it is still on my serial port as mentioned above. I am sure it has got something to do with it.  My midi performance is tight and relentless no matter what the rest of the program happens to be doing which is ideal and to be honest the way it should be.  I am nervous too because once I upgrade to a better seystem I won't be able to do things the same way.
 
Thanks for those links to the programs. I have downloaded them and will do some tests and see what turns up. There is some interesting info in them as well.

Thanks Jeff,
 That is actually the article that made me want to find the MIDItime app.
 
 I am packing up to go over to Houston for some work for a few days and when I get back I'll try the MIDI testing software.
 
 

WOW... I went to edit the post with my reply to Jeff which included the links we are speaking of and the post disappeared completely.
 
Here it is again:
 
click here to see the wayback machine copy of the MIDItime webpage
 
click here to download MIDItime
 
 
click here to see the wayback machine copy of the MIDItest webpage
 
click here to download MIDItest 4.6