WatchUSeek Watch Forums banner

61 - 80 of 477 Posts

·
Registered
Joined
·
250 Posts
Your audio looks very clean!

May it be that the nominal beat rate is not 18000 bph? If you know the correct bph, you can just write it in the bph field, even if it is not among those in the drop-down menu. Tg should be able to work with anything in the range 12000-72000.

I would love if you could send me a 30sec sample of your co-axial to play with...
Believe it or not, I'm using my never-used iPhone headset mic and just holding it tight against the case. And I'm certain the bph of that pocket watch is 18000. It loses like 5 minutes an hour. It's been non running and I only wound it to get a reading since it's by far my loudest watch. Here is about 45 seconds of my Omega with the calibre 8500 co-axial.
AT8500.jpg
 

·
Registered
Joined
·
250 Posts
I just got to thinking...you probably meant an actual sound recording. I've PM'd you.

By the way, back to my "clean sound." There's actually quite a bit of background noise in my recording. So I have to conclude your program is exceptional at eliminating noise.
 

·
Registered
Joined
·
315 Posts
I too am having a font issue with OS X. 10.11.0 here.

Is there anything I can do to give you a hand in trying to sort it? I'm used to programming but not a C person I'm afraid.

I'll also have a look at popping together a homebrew formula so people can install a little more easily.
 

·
Registered
Joined
·
87 Posts
Discussion Starter #64
@dom_

Thanks for asking the mic question: I am interested too. I still have to experiment with the electronics. My gut feeling is that, if the signal is not exceedingly weak, no amp is better than a noisy amp. But that's just gut feeling.

@dmnc

Thank you, your offer is very much welcome! I am really keen on getting this thing to work cross-platform. I will probably just go back to the default font and align the figures manually: it's not pretty, but it works. The homebrew formula is an excellent idea. I will write to you as soon as I find the time to do that modification, so we can proceed...
 

·
Registered
Joined
·
232 Posts
I'm just a casual watch tinkerer, not a professional. I loaded the program and it worked fine on Win10. I didn't have a microphone plugged in at first so the program didn't launch. No error messages, just nothing on screen. Plug in your mic first. Plugged in an earbud headset and launched again. Everything works as advertised. I think the headset was a Motorola brand. I'm not getting the three distinct waves on the bottom screen but it was very noisy in my room. A lot of fan noise from computer and heating system. My computer allows me to adjust the mic input level. I need to play with that some more to get a good input level.

So far, much better and useful than any program I've tried on my tablet. Keep up the good work. I tried this on 2 Seikos with the 4R36 movement. I now that these two watches are quieter than others. The tablet programs can't hear them and they works fine on a 7S26 watch.

Hopefully I can provide some useful feedback from a novice user with few computer skills but full of enthusiasm. I think your stated goal was a useful program for people in my category. You are well on track. IOS version would be nice too. I go both ways on computers.
 

·
Registered
Joined
·
315 Posts
I cobbled together something to make this a little easier to install for Mac users with homebrew.

If you just run

Code:
brew install dmnc/horology/tg --HEAD
You should get all the necessary dependencies and be good to go by just running "tg".
 

·
Registered
Joined
·
87 Posts
Discussion Starter #67 (Edited)
Update:

The latest version (0.2.3) should complie on OS X. Untested yet.
Edit: tested. And we have a nice install script. Thanks dmnc!

@bklake

Thanks for your commentary. I suspect that you didn't get any error message because you were not using the very latest version (this program has been updated quite frequently).

Yes, my goal is to make a program that can be used by a novice (a category I belong to) without offering a disappointing first experience, or being an unreliable, dumbed down black box.

Unfortunately I won't make a IOS (iPhone, iPad) version, because to develop for that platform one needs an Apple machine, and an Apple developer account (100$/year), and apps need to be approved by Apple before they can be distributed. In short, it does not make sense to develop for IOS unless you plan to make money out of it. In a distant future, I might attempt to experiment with Android. I guess that I should have used Qt from the get go anyway...
 

·
Registered
Joined
·
274 Posts
The idea of using a clip-on guitar pickup suggested by Klaus in this thread sounds very very good to me. Unfortunately I didn't have time to experiment with that, but I most probably will. There is even the possibility that it might work without the amplifier: Klaus, did you try?
Well, I did try but failed getting a poor signal. The reason might be that the piezo is sensitive predominantly for solid-borne sound which is very low.

If anybody else is going to try the pickup the following might be usefull. I removed the foam rubber for better contact and covered the hole with an adhesive tape to reduce sound-pressure of ambient noise. No big effect for the tape, just as a matter of conscience.

Dynamic microphones didn't satisfy my needs completely but the sound processing of timegrapher is such good that the built-in microphone of a laptop will be enough for a rough shot.

...but I can't get the application to run at all. I'm using Windows 7 Pro 64-bit and when I extract the files from the .ZIP, install the font, and attempt to run the program, nothing happens. I don't see any errors in the Windows error logs. I get a second or two of Windows "spinning wheel" but the program never launches. Could I be doing something wrong?
I'm experiencing as well the issue with version 0.2.0 and 0.2.2 not starting anymore and "the spinning wheel" causing the explorer to stick. Maybe caused by some windows 10 or antivirus updates? The old version 1.0.0 (nov 5th) is still running perfectly.Restore Auto-Saved Content
 

·
Registered
Joined
·
87 Posts
Discussion Starter #69
I'm experiencing as well the issue with version 0.2.0 and 0.2.2 not starting anymore and "the spinning wheel" causing the explorer to stick. Maybe caused by some windows 10 or antivirus updates? The old version 1.0.0 (nov 5th) is still running perfectly.Restore Auto-Saved Content
Thanks for reporting this bug!

Adamant's problem, as it turns out, was caused by his microphone not being plugged in when he started the program, so I don't think that yours is the same issue.

May I ask you to download version 0.2.3 (the very latest one), then try again, and tell me the result. This version differs from the previous one because I got rid of an annoying font issue which killed the OS X build. Your problem might be related to that. (Also the latest version has amplitude and is about 50% faster, so maybe you want to use it.)

I tried it now on a virtual machine with Windows 10 Enterprise (evaluation edition), no antivirus, and it did work.

I would greatly appreciate if other Windows users could give me a feedback on the latest testing version (0.2.3). Something like "Windows version, tg version: works / fails (failure mode, error message if any)." is more than enough.
 

·
Registered
Joined
·
274 Posts
May I ask you to download version 0.2.3 (the very latest one), then try again, and tell me the result.
Windows 10 home + tg 0.2.3: spinning wheel, no entry in taskmanager but "tg.exe" in use, no message
Windows 10 home + tg 0.2.0: spinning wheel, no entry in taskmanager but "tg.exe" in use, no message (has been running perfectly a few days ago)
Windows 10 home + tg 0.1.0: running without problems (and very smooth on i7 laptop)
(just realizing that I do not have 0.2.2)

Win XP home SP3 + tg all versions: running without problems (just a bit overstressed :-D from USB-stick on netbook with Intel Atom b-))

Windows 7 64bit + tg all versions: running without problems (Pentium 4 dualcore)

Seems that the bug is on my Win10 (maybe after KB3116908, Java update, Flash update ...). 12-13-2015 --> some conflict with Avast Free Antivirus
 

·
Registered
Joined
·
29 Posts
I tried version 0.2.3 on Windows 10 Pro (version 1511) and it worked fine. The Microsoft LifeCam Cinema webcam picked up a signal after pressing the watch crystal against the microphone. :-!
 

·
Registered
Joined
·
925 Posts
Thank you for that great software.
I've just bought a cheap piezo guitat pickup from ebay, plugged in my soundcard (esi maya44 pci), and it works. (win10 64bit)

However I've a Horometer, that shows very similar results, but with ~25-30 degree lower amplitude.
Can you modify beat error, to 2 digits?

Did anyone try this, with co-axial? Sometimes it jumps to the place, and shows some amplitude (I don't know whether correct, or not), I added green arrow, to show that. But most time it looks out of sync.
It is Omega 9300, the lift angle is 39.

co_axial.png

9300 with amplitude:
co_axial2.png


5.5 yrs old Omega 1128 (2892-A2):
tg0233.png

Longines L633 (2824-2)
lld.png

Steinhart - ETA 2824-2

steinhart.png
 

·
Registered
Joined
·
87 Posts
Discussion Starter #73
Thank you all! I didn't forget this thread, I just have a deadline tomorrow...

Wizard's observations are very interesting, I will find the time to share some info on Saturday, hopefully. May I ask you to check my amplitude also against some other software (so we know that the signal is precisely the same). For instance you could use Watch-O-Scope or biburo, tg should be able to run in parallel with both.
 

·
Registered
Joined
·
925 Posts
Wizard's observations are very interesting, I will find the time to share some info on Saturday, hopefully. May I ask you to check my amplitude also against some other software (so we know that the signal is precisely the same). For instance you could use Watch-O-Scope or biburo, tg should be able to run in parallel with both.
I've tried W-O-S, but didn't give stable readings. Biburo didn't even found the signal/soundcard :(
I'll try them more on weekend.

The signal was the same, when I tried yesterday, as I've attached the piezo mic to the watch, and the Horometer to the crown at the same time.
Maybe I'll also try to make some videos.

(I'm also planning to make a diy mic amp, however I'm not sure it is needed, as I had fine results without it)

If you need any other help, measurement, picture or video, just ask, I'd like to help you in the development in this great software ;)
 

·
Registered
Joined
·
925 Posts
I've tried W-O-S, but didn't give stable readings. Biburo didn't even found the signal/soundcard :(
I've made some new test.
Biburo is "working" fine, and maybe W-O-S will also work. But the signal level is too low for them. If I knock the mic with my hand, they show something.

Your software is working very nice, with the weak signal.

So I'll make an amplifier.
 

·
Premium Member
Joined
·
1,495 Posts
Awesome software, I regret getting my timegrapher now!
Thank you so much for doing this!
 

·
Registered
Joined
·
87 Posts
Discussion Starter #77
I'm back. Thank you all for participating to this thread!

Windows 10 bug (@Klaus) I tried to make all the updates to my virtual machine, but still I can not reproduce the bug. If someone else is experiencing the crash under Windows 10 please let me know, so we can try and track down the cause.

Amplitude (@Wizard)
I would like to understand what causes the error of 25-30 deg in amplitude that you report. Let's look at your Omega 1128 screenshot, which is very clean.

fig1.png

Here tg is detecting a little less than 7ms between the first pulse and the third pulse, indicated by the blue line. The formula to compute the amplitude from this time is well known

fig2.png

and it gives us a little more than 284 degrees, which is consistent with tg's reading (considering that the other pulse was a few tenths of millisecond shorter). In conclusion, the Horometer device must be detecting a time between the first and the third pulse which is at least one half of a millisecond longer, so at least 7.5ms. I see that that device has a very complete waveform display mode, may I ask you to observe the waveform of the Horometer and try to figure out where this difference is comming from?

Beat error digits (@Wizard) I believe that by "two digits" you mean adding the hundreths of milliseconds figure. Of course I could just print out that number, but it is basically a random digit: tg is just not capable of measuring beat error to that precision. I would argue, also, that it is not possible in general to measure beat error to the hundreth of a millisecond. Let me explain why. Beat error is given by the time between a tic and a toc (and a straightforward formula). Tics and tocs are not moments in time, they are indeed complicated sounds that take several milliseconds. So one needs to fix a reference point to say when in time a tic or a toc has happened. Geometrically the correct reference point is when the impulse jewel crosses the center line of the escapement. Unfortunately, there is no sound occurring at this precise moment, so we need to make do with some other, arbitrarily chosen, reference. As you can see from your screenshots, and the many others posted in this thread, the tics and tocs almost always differ by, say, one tenths of a millisecond in the relative position of their pulses. As a consequence any arbitrary reference point can not be expected to indicate the actual beat error to a better accuracy than about a tenth of a millisecond (just because choosing different references would produce indications that differ by about as much).

Which reference is used in practice? Well, my impression is that the most popular is the beginning, namely the first pulse. The reason appears to be of convenience, because with traditional methods, if the sound is good, this choice gives the most stable result. Tg's approach is to use the whole waveform as a reference and not any particular point on it, this produces a result very similar to using the third pulse as a reference, because that's where the lion's share of the energy is.

Coaxial (@Wizard) Tg's algorithm does not really work well with Daniel's coaxial escapement. In short: the rate is fine, the beat error is just an indication, the amplitude is garbage. Here is the problem: tg works under the assumption that the tics and the tocs have a similar waveform, which is of course not the case in coaxial escapements. The jumps that you see are tg trying to locate the third pulse of the waveform. Of course, we know that there is no such thing in a coaxial, but tg doesn't, so it gets confused and every now and then it changes it's mind about where the first and the third pulse are. Sometimes, tg might manage to convince itself that it is picking up a well defined first pulse and a well defined third pulse, so it displays a value of amplitude. The truth is that it is trying to understand a coaxial thinking that it is a lever, so whatever amplitude tg might show in this case has just no meaning at all.
 

·
Registered
Joined
·
925 Posts
contrate_wheel: I've even made a diy amp for the microphone, to improve this thing.
The biburo is not working for me, in windows10. It won't detect the beat, won't show any information, that can be used, although it can "hear" the watch.
The W-O-S is "working" but very unreliable. When it works it also shows 10-30 degrees smaller amplitude, but it frequently looses the signal. I don't know what is its problem. There is also difference in the beat error, between TG and WOS. Also the daily rate is fluctuate a lot with WOS.
Only the TG can lock on the signal reliable. ;)

Now I can't check with the Horometer, as it would be a Christmas gift from my wife, so I'll be able to check things with it next week :D


I've made 2 videos, where TG and WOS is working side-by-side, and managed to capture a timeframe, where WOS didn't loose the signal.
The first video is with the microphone amplifier, the second is without it, only the guitar pickup plugged in to the soundcards MIC-in. You can hear the volume level difference also.
https://drive.google.com/file/d/0BwKbAzpjbaDORmltNXM3RG5aZG8/view?usp=sharing
https://drive.google.com/file/d/0BwKbAzpjbaDOeUFsa0RCRHNRd1E/view?usp=sharing
 

·
Registered
Joined
·
274 Posts
@Wizard7 cool idea making video-screencaptures. |>

In my opinion there is a bit to much noise in your recording when you are using the amp (I made the same experience several times).
TG is solving this issue with it's crazy good filtering algorythm. But in your first video one can see as well the noise in the graph.
Biburo is not able to detect if it's a signal or if it's noise and has to surrender.

Here some points how I get rid of the noise, depending where it's based:
1. power supply: I first tryed a 5V wall power supply which produced a lot of 50Hz noise. So I switched over to a 9V battery (very effective)
2. grounded shield: as shown in the first picture of the thread, about 15 centimetre length taken from an old coax-cable around the pickup-cable and as an extension the red cable which I can easily grab or even more comfortable just step on with my foot :-d (very effective and you see the difference immediately grabbing it or let it hang loose)
3. amplification (computer): give it a try to play with the amplification of the internal soundcard. For me +12dB and about 30% volume for soundrecording is comfortable (if the battery is getting weak I switch to higher amplification), drawback: higher values will amp as well the noise (basic)
4. amplification (external amp): same as before, mine has two potentiometers (Conrad 197688
one for amplification the other for load voltage) trimming those two potentiometers is very important for a good signal to noise ratio (basic)

for both amplification methods I use biburo to adjust to a combination of good soundsignal and low noise level. Noise then is just a few millimetres (about 1/3 of the signalgraph see picture in post #19
or even less when testing loud vintage watches)

If you give it a try and at least modify 3 and 4 your biburo should work fine as well and you will be able to compare ratio, amplitude and repere/beat error of different timing programms.
Very happy for any feedback and discussion.

@contrate-wheel again many thanks for the TG !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! it is the greatest Christmas gift for the low-budget hobbyist
:-!
 
61 - 80 of 477 Posts
Top