Have my videos have helped you? why not support the channel in some way You can make a one off/ monthy donation here in only one click: www.buymeacoffee.com/MyRetroWatches If you would like to purchase any tools & equipment then please see my Amazon affiliate & eBay (I do earn a small commission from any sales.) Thank you. Link: Link: myretrowatches.co.uk/tool-sale-links/ If you would like to buy any of my T-Shirts then please look here: teespring.com/stores/my-retro-watches
I know this is kinda old video but it was way so useful for me. I have Orient Mako 40 which was running +17 seconds per day. I purchased watch tuner timegrapher from app store and tried to regulate mako 40 with same app. After 1 hours of tinkering, I got -1 seconds per day. I've been monitoring the watch for around 2 days and it really runs -1 seconds per day, which is amazing result for me. Thank you for video, it helped me a lot. ps: I put microfiber cloth on the desk, removed the watch's back cover and put the watch upside down on the cloth. Then placed Apple EarPods microphone on top of the watch mechanism. Then put glass jar upside down on top of the movement and microphone. That created nearly perfect sound isolation.
My tips for getting the app working better: 1. Stuff a folded up paper towel in the bottom of a glass, then put your movement (and the mic) on a towel on the table, then put the glass upside down on top of the movement. The sound isolation with that is very effective. The paper towels dampen any external noise and any tick tock echoes. 2. If you have a beat error, handle that first. It seems to affect the accuracy everything else.
Heya Mike, Sound engineer here. You were right to try a different microphone as I believe that's the primary difference between your time grapher unit and the app. In the end, they are both software using a hardware microphone. However you used the same type microphone as before, and I believe type matters quite a bit here. I understand the Weishi (pronounced way-shee by the way) uses a piezoelectric mic which is a little more sensitive to physical vibrations, less so for vibrating air as with the other mics. Piezos are often used as pickups in musical instruments, acoustic guitars for example. They generally don't pickup sound great, but pickup vibrations beautifully. And that brings us back to the Weishi mic, which is probably depending on physical contact with the watch to work well. The Weishi is probably also filtering the bandwidth so it 'specializes' in hearing ticking, not everything, as a phone mic will. So, without a specialized watch mic, I doubt the app will be properly testable as compared to the Weishi. Apples, oranges, you know. 👍
I bought a piezo micro from Korg and cannot read anything. With the mobile micro I can record the tik tak perfectly but when using the contact micro I can only listen to surounding sound and zero tik tak. What am I doing wrong?
Test equipment and the regular calibration of test equipment is a science all unto itself. It would be fascinating to see a comparison of two identical Weishi 1000 timegrapher machines with their peripheral microphones pitched against the same watch as a test calibration standard unit. There is no doubt that both machines will produce different results with the same target watch. But just how different remains to be seen.
I just downloaded the app recently out of curiosity and use it mainly for pendulum clocks. That might be the one advantage where this app shines as it can also be used in larger clocks that can`t be fastened into the microphone aparatus. Great video!
Thank for your vídeos! I got the Chinese weishi 1000. It helped me today, my watch wasn't running bad, but I put on the TG and I could see slot of dots like snow. Inmediately I demagnetized the watch and now I can see to paralell and nice lines and a constant +3s. Without the TG I couldn't see that.
On further reading it's likely down to the sensor / mic being used. Most people have the best results with the PC application with a Piezo transducer. The 'electret' type microphone catches too much interference via EMF and other noise to be useful. I would be interested to see what's under the hood of the timegrapher mount - those four pozi screws look awfully tempting!
I purchased the app. Watch Tuner Timegrapher for my apple phone, and attempted to "regulate" my seiko ~ save the ocean ~ turtle. After about 15 attempts to get it close to 0 {+/- 5 sec.s} in only one position, I got it around -5-7 sec.s to +1 sec. Not so easy. BUT, I was originally thinking I didn't need anything, I'd just bump it a small amount faster (it was way slow, after tracking it with app., it was lik 45 sec.s slow a day... ). I can only imagine trying to get this watch anywhere close without the app. or a Weishi 1000 or something. Thanks for the comparison, and video. 🙂 Live and learn 🙂
How do we know if it isn't the app that's actually more accurate? I would like to see two or three of the Weishi timegraphers running at the same time and compare their values. Maybe the app's values will fall somewhere between the three dedicated timegraphers.
I was just about to comment the same. How do we know that the app isn't more accurate than the timegrapher. I would like to see both compared to one of the professional high end machines
It all comes down to the quality of the crystal oscillator in the device, electronic devices that measure high frequencies do so by comparing the input frequencies being measured with the frequency of the device's crystal oscillator which in a perfect world is 32,768Hz, but a low quality oscillator will usually not be temperature compensated and will drift as the temperature of the circuitry and oscillator change. A good timegrapher should have a high quality temperature compensated quartz oscillator in it, a phone or computer I'm not so sure, they generally aren't devices used for measuring high frequencies so I'm not sure manufacturers would bother using high quality oscillators. I work with crystal oscillators in geophysical data measuring instruments, there is a large range of precision/accuracy and cost when buying an oscillator.
MeesterDash it’s a Faceted Crystal. I have a few with these and they are very nice. Always a 70s Watch and quite often an Actus. Some would have had them from new but as they are so hard to find if changed it would have had a plain domed crystal fitted.
Thanks for sharing your experiments. I got decent results using a windows app and a piezoelectric mic. I had an old instrument tuner which has a piezoelectric mic with a convenient plug. I had to run it through a preamp on full volume and then I could even hear the ticks (barely). It works but would require more amp and some EQ to filter noise, but it is not very convenient. I see the comment of the sound engineer guy and I think he cracked it, piezoelectric is the solution. You could tweak to get better results but if you have to buy the sensor, a preamp, filters, you might just buy the device. If you are an engineer and like a project this would be an interesting circuit to make.
I appreciate your response to previous feedback. I suspect the Weishi designers analyzed the audio from a watch and filtered the audio and selected the input device for maximum efficiency. That could take a lot of tinkering to duplicate.
Similar apps are available for Windows, and I thought about trying one to save the cost of a Weishi. Here's the thing: you put an app on your phone, Mac, or PC, and you've added to its complexity, impacted its battery life (unless it's using mains power), and tied it up from other uses while you're using it for time graphing. And, as we see here, you're also buying external peripherals and fussing over settings to see if it's in any way accurate. The beauty of a dedicated device for this is that it's doing only what it's designed to do, it's doing it in excellent fashion, and it's doing it with no hassle. It's roughly analogous to using a properly-sized box wrench instead of an adjustable spanner or pliers to loosen a nut. I imagine a professional can learn more details about how a mechanical watch is going by using a very detailed application on a PC with a proper mic and vice, and I imagine such apps exists for testing a quartz watch as well (I know there are dedicated devices for it). Sometimes I miss the days when we used a phone for making contact, a camera to take pictures, a map to navigate, etc. Everytime I hear, "There's an app for that," I think, yeah, and it's probably marginally effective sometimes.
I have an app on my laptop very similar to what you have on your phone. I also have a TGBC cheapy timegrapher device. I found that when I try to check a watch while it's in it's case it does not work as well as if I check just the movement without the case. The unit came with a cable with a type c connector on one end that plugs into the timegrapher and a USB connector on the other end that plugs into my laptop. With just the movement, the parameters and lines show up very well. I don't have a regular timegrapher like yours to compare with but I did use a watch that I know is running well and the rate was +2, beat error was .1, amplitude was 298 and the line graph was straight and consistent. As I said though without having something similar to your Weishi 1000 to compare it with, I don't know how accurate those numbers really are.
The app IS useful. I used to to tune a Miyota 9015 based watch from +12 sek/day to about +2 sek/day. Ordered the Timegrapher 1900 model , but delivery is 2-3 weeks. While waiting for delivery, the app has definitely made mere more happy with my watch performance.
Agreed the app does have its use as a guide but testing against a dedicated machine does show its weak points. Mainly the complete silence you have do be in to use it. I hope you still found the video useful and thank you for your comment.
I think the best microphone solution would be a contact mike. I think I could make one for about $3. You are also assuming that the results from the Weishi are perfect and infallible. I am sure they are in the ballpark, but I am much less convinced that they are perfect, and in fact I'm a bit dubious about it's count resolution due to the age of the technology. There are also Android apps available (and Android Phones that still have a 2.5mm audio jack).
@@Todestelzer there's a special jack on the back for doing calibration. I'm not sure how that works. I have a Weishi on the way so maybe the manual has more information (or maybe not, we will see). I made the contact mike and it worked very well, but the android software left me rather unsatisfied. None of the programmes I looked at did amplitude so I didn't go ahead and build the microphone into a movment holder. The apps are probably OK for the average user, but since I'm stripping and servicing movements I really need to know what the amplitude is... it's the main indication of whether you've improved the movement or serviced it to an acceptable degree. It's possible to analyse the amplitude by other means, for example by making a video of the movement then measuring the full angle of rotation, but having a timegrapher on the workbench would be so much easier.
Been 5 years. Not sure if they improved it since, but vid I watched just out finds they're very close, within 1-2 seconds which is accurate enough for most.
I have been using a free application called Clock Tuner, which is essentially the same thing you have. I simply press the watch face to the spot on the phone where the microphone hole is, and I get good results. But first, I move to a room with no noise, or eliminate noises, like the Telli, radio, talking, etc. Perhaps you could take the foam cover off your microphone and hold it to the watch? Cheers
Kool, thanks for the heads up on these apps Mike, I appreciate it and G'morning 😇 . I don't have a time graph for yet so for now the time graph for apps will give me something to play with just for fun or to possibly get a baseline idea on how things are running 🤓 ...
Thanks for the update. May not be perfect but for a entry level option, it worth the price. Just a FYI, lav mic are unidirectional, meaning it just picks up directly from in front of the mic head
Thanks, unfortunately it was / is the only mic I had at the time. Granted the app is goo enough for a beginner. Thanks for taking the time to comment and watch.
thanks a lot Mike for this new video and for doing this comparison with such an open mind. my guess (but I'm not an expert at all) is that the app is probably wrongly interpreting a parasite noise on the tock - assuming the tick was right, and that affects the beat error calculation. the weishi machine is clearly superior there with dedicated sensor and software. if my guess is correct, what is instructive is that the app shows there is something abnormal with the tock (friction?). I think the original Swiss made witschi timegraphers can analyze the tick and tock sounds spectrum and identify that kind of abnormal noise. but then of course we are not talking about tinkering hobbyist 😀 nor the same budget 😭 thanks again for your great videos. looking forward the next one! l10n3l
thanks for your videos on the weishi 1000 timegrapher, I have the same one. I just wish I could find the lift angles for vintage American pocket watches. These are the watches I mainly repair and collect along with my vintage Seiko wrist watches. If you find a source for the pocket watch lift angles please let me know.
What are those smaller sized 2nd line values in each column? (small white font -2, -4, -1, -2 etc.) The ones below the 1st line, (highlighted, larger font) actual values?
I used a similar app to check my watch (Orient caliber F49). To minimize the external sounds, I made the tests at night, and put the watch inside the wardrobe, covering the watch+microphone with a fabric. After starting the test (5 minutes) I closed the wardrobe door and let the thing happen. Got an avg 1sec/day and .1ms. After some research, saw that the F49 isn't all that precise, so I ended up with some questions LOL. Is the watch really that underrated or something went wrong while testing it Anyway, great video, man
I always do a run-through if I'm doing something for the first time. Practice help avoid the little (awkward moments.) It would have helped you to know which watch would give you trouble, or what to do for the best results, so the viewers know how to get the proper or best result (withstanding unforeseen outside effects) and to point out important things, like the straight graph lines, angled lines, "snow" etc. But overall, it was a good for general knowledge. Cheers,
Thank you. Granted more planning or explanation could have been used, I do a lot in one take and I dont want videos to be too long as statistically when looking in my analytics people usually loose interest after 10 minutes. This video itself since May 2019 has an average view time of only 6 minutes. At some point in the future I would like to do a more comprehensive video on the timegrapher machine and if that happens I will take into consideration some of your points. Thanks
The problem also is the fact that even if you are able to tweak the app to an appreciable degree, you won’t have any reference to compare it to in order to know this, and if you do have another reference you don’t need the app
Jeff Dumon I think the point I am proving is that an app is good for a quick check for those who are curious . If you want to tinker more then a timegrapher is a purchase that you cannot avoid.
Thanks, I was over thinking the issue. The app is quite a neat idea, I just wanted it to be better than it is. So I am pleased you have produced this video.
So it's now Nov 2021 and this app maker now has a couple of cut down versions of this app as well. So I bought one of the 'lite' apps and I actually read the help page. The instructions clearly state "...make sure that the crown of the watch and the microphone from the earbuds are touching." My earbuds are a no name cheap as chips brand but I get anywhere between 98-100% signal quality. I also live in Jakarta where there is a constant stream of noisy motor scooters roaring past. It doesn't seem to worry the app at all. They may have upgraded the app in the time since this video was made, but it appears to me that the quality of the microphone has little bearing on the matter of signal quality. Just follow the app maker's recommendations.
That maybe but how accurate is the results. Problem is you have nothing to compare it to. The most limiting factor is that you would not be able to regulate a watch real time using this method easily.
I am enjoying the video and really was hoping that there would be an app that would give acceptable or at least reasonable results. I have a few watches I'd be curious to check (one Swiss, one Japanese, and a cheapy Chinese), I really am not going to do this enough to justify spending the money on a Timegrapher. Suggestion, when trying different things, make ONE change at a time. If you change the microphone and two different settings, you are just whistling in the dark since you don't really know what changed what. One item may have helped and the other two messed it up.
Thank you for your appraisal! Two questions: What adaptor did you use to connect the 3.5mm jackplug of the lapel microphone to the lightning connector on the iPhone; and secondly would you like to hazard a guess as to whether a little padded box might exclude background noise which might interfere with audio pick-up? Thanks otherwise for a very level-headed assessment!
Thanks Dave. for the adapter Apple made one and supplied it with I think the iPhone 7 when they first did away with the socket. so its a lightning to jack adapter looks like this. amzn.to/3m8LFzV As for sound deadening , yes you could do this but to me its extreme and then impossible to actually work on the watch to correct the regulation.
@@MyRetroWatches Thanks, Mike. I actually downloaded that timegrapher app for the apple watch, but found the microphone of the phone not sensitive enough for some of my Seiko movements (or the background noise too extreme), so also invested in a Levalier lapel microphone, which is definitely more sensitive, and gives me a more realistic set of data using the app. Definitely not good enough for a regular watchmaker, but enough to allow me to assess whether my couple of watches are very bad, or just bad (ha ha). Keep up the good work!
I really enjoy your videos, they teach me and like so many of the other posters, get me to wondering about so many of the variables. Since the Weishi doesn't record an average over the length of the test, you basically can only compare the end results shown. With that being said, they were both extremely close except for the beat error in this revisited video. My question is, if the watch is off by say 8 seconds a day. what would be the beat error? Would it be like .3 or minuscule? What else can cause a watch to lose or gain time besides the beat error? This video has opened a pandora's box of questions for me to think about and I'm sorry but your the only magician with the answers that I know. :)
Hi Danny and thanks for watching and commenting. Interesting question however beat error is separate to rate with error in beat you can still regulate to a satisfactory level but its the long game with these. Beat error is the error between the tick and the tock, in an ideal world you want them to be as equal as possible as a millisecond over a long period of time will add up. Accurate beat error also effect the amplitude. You can have a beat error of say 0.3 and still regulate the rate to within a few seconds per hour. The ultimate goal is to have one nice straight line across the TG and then in all 5 positions that the watch should be checked in (dial up dial down etc) My full understanding of the weshi and how its plots the graph is limited. I assume the longer you leave it on the more "average" the reading. I often leave movements on there all night to see how its performing long term. This is now the question I have..
@@MyRetroWatches Thanks for your prompt reply Mike! I haven't memorized the proper names for most of the watch parts yet, but with your help and some research I see what you mean. Beat error is centering the mainspring and rate is shortening or lengthening the mainspring, two different animals! By the way, I just received my Weishi TG. I don't plan on working on anything yet, still need to learn more and acquire tools, but I may see how some of my watches are doing with this equipment. Thanks again and keep posting these helpful videos.
another quality video Mike. For my money, the app is a huge waste of time. The biggest problem of which is how does one know what the correct reading is, especially without something to compare it to. As you said, you could tweak the thing all night and you stil wouldn't know what to believe. The old adage is apropos here - buy the best tool for the job and you buy it once-buy a cheap tool and you will buy it more than once. In the USA the Timegrapher costs approx. $150. Well worth it and a lifetime investment if you are a watch enthusiast. Plus no need to monkey around with adjusting settings and multiple takes.
Sadly I can't compare my results to a "real" time-grapher, but I will say that "Watch Accuracy Meter" also seems all over the place WRT beat error and amplitude. Overall deviation seems relatively accurate however - at least when compared to my observations recorded via "WatchCheck" over multiple days of wear and/or sitting on the winder. And that's without me using an external microphone ... I just put the crown right next to my Galaxy S20's built-in mic. Of course, I'm a bit weirded out by very low and/or erratic amplitude results and high beat error for brand new or 12 to 24 month old movements - but I guess there's a reason why you can only set lift angle and get amplitude results in that app by activating "experimental" mode. :D
I used a app to regulate a couple watches so it is useful in that regard. I used a app to determine if a watch needed service by the ampltude so that is usefull. In short the app is good enough for government work.
The app itself may be better than a timegrapher. I would set it up with not a lapel mike, but the mike used with a timegrapher. Or with a calibration mike, or a calibration mike used with the timegrapher measurement body used to rest the watch on. You never know. The timegrapher algorithms may be set up for gross averages only. I'm not a watchmaker. I do have experience with professional audio and audio measurement components. Just a passing thought. Great Video. Like the measurement music choice.
Perhaps but the app does not tell you to use anything more sophisticated than the mic with your headphones. The machine will always win if you want to work on a watch and get a reading at the same time. The timegrapher uses a piezoelectric microphone
Now it probably sounds like heresy, but I do agree with Mark, in the sense that I believe the app is sometimes out-performing the timegrapher. My experience is with digital signal processing, building machines not so dissimilar from the timegrapher. It’s easy to see that they are both making mistakes, but they are different ones, leading to different results. Earlier tests from the first video make it pretty obvious that the app wasn’t receiving good information from the microphone - and the results were therefore bordering on uselessness. Garbage in, garbage out. On the flipside, the app is pretty overt (IMO) about how it goes about making its most basic calculations, and the inclusion of the standard deviation is a significant improvement over the timegrapher. Also, it should be noted the timegrapher likely produces a display that is artificially clean, due to its somewhat lower resolution display and summarized reporting. The phone has orders of magnitude greater processing power. So, with the correct DSP applied, along with the option of a longer integration window (essentially unlimited memory vs 12 sec), the app should actually be able to put the purpose-built device to shame. (Though the flexible, rotating stand is also very handy.) But, does the app (or other apps) have a robust algorithm? - I can’t say. What I can say is that I have now been inspired to drag out my pro gear and do a little bit of an analysis of the raw audio against a high-precision/high-stability time standard. I can then compare that to the app and see how they measure up. (Because, apparently, I feel it’s a wise decision to use $20K of test gear and do hours of tedious work in order to save 100£. Hmm. Must be the watchmakers’ gene mutation…)
The app is pretty good for what it is. It's the best one I've seen by far. That said, if it can't read the beat error correctly, then it is useless to me. I can't believe that I was without a timegrapher for 2 years after starting the hobby. I would have to stay absolutely quiet for hours at a time just to get my watches anywhere near close in timing. Then I got my timegrapher and realized just how bad my beat of error was off and how low my amplitude was a lot if my watches. I wish I had mine sooner, because a lot of watches would have been sent back that I thought were ok. For anyone with any mechanical watches a timegrapher is just a must have.
Hello and thanks for commenting. I can completely relate to this. I do struggled with the app for 3-4 months before spending the money on the timegrapher. You then wonder why you didnt do it right at the start.... Same can be said for a microscope both pieces of equipment are complete game changers to your watchmaking efforts.
The app gives a decent ballpark value. If you buy a new 60$ seiko with a 7s26 and the app shows +20s, you can easily make it more accurate. Tweaking/servicing expensive or vintage watches tho would require precise measurement, that's where the timegrapher shines :)
That is a Omnidirectional mike and picks up sound 360 degrees, I would use a unidirectional Shotgun mike it concentrate on one point and ignores sound from other direction.
good point however like I think I said in the video its the only other mic. I had. This video was really because of curiosity after the first video as to if I could make it better. In all seriousness nothing will compare to the Timegrapher machine, its purpose build and uses a piezo mic . It was just nice to compare the two to see just how close the app came to the real thing.
I think an issue with this setup, not the way that you're using it but just inherently the way that the setup works, is that you are having to use the dedicated time grapher as a reference and then tweaking the phone setup to get as close to the time grapher as you can. If somebody is trying to use the phone app on its own and they don't already have the time grapher, they have nothing to use as a point of reference to try to calibrate the phone for maximum accuracy, which you need a time grapher for, and if you already have the time grapher, then why not use that instead of trying to tweak an inferior system to work and then using the inferior system? You might know somebody who will let you borrow their time grapher to calibrate your phone setup, but again, if they're willing to loan you their time grapher than why not just use the time grapher you're borrowing.
In principle, a good mic and amp setup should definitely make it equivalent - but it would basically require using the mic and stand setup of the weishi with a custom interface to the phone. And I don’t see that being much cheaper than the whole package, because the cost of that weishi is not residing in the cheap little box and screen.
Curiouser and curiouser. It wouldn't by any chance be possible to use the Weishi microphone with the phone, would it? That would be a useful control test, you'd think. Please note that I'm not egging you on to do it, just musing aloud. :-)
GenWivern2 there is a small chance but I don’t think that mic is doing anything . I just think the app is not really all that up for listening and interpreting what it hears .
@@MyRetroWatches Hmm. Reading up about microphones to suit the TG programme I mentioned in the comments on the last video made my head hurt. Apparently the transducer type and the mounting are both, if not exactly mission critical, definitely going to have a significant effect on the signal quality. From what I can gather the Weishi uses a hard-mounted piezo, which is a very different animal from the Boya ... it also uses a highly inconvenient connector by the look of things making a comparison impracticable! Anyway, for the very little it's worth I suspect the microphone rather than the software. That's not to absolve the app - after all it's highly unlikely that the average user is going to have an optimal microphone set up, or anything like. Oh, and forgot to say that a collection video would be very welcome.
@@MyRetroWatches I have a bad habit of getting sucked into researching things, rather than getting on and doing them. My take on all this - pretty worthless of course - is that you can save say £90 on the price of the Weishi by going with the app, but you incur a good £90's worth of hassle with no guarantee that it'll be reliable, so you may as well go with the timegrapher and rest easy. :-)
Mike -- Thanks for the videos. I'm a new owner of an 1890 Waltham, trying to understand what's responsible for its imperfections. I'm using a free Android app ("Watch Accuracy Meter") which is similar to the one you're using on your iPhone. I have a question, applicable to the Timegrapher and app: . . . How is the amplitude calculated ? There must be more happening in the audio stream, than a "tick - tock - tick - tock". You can get "beat error" from those times, but not amplitude. Thanks again -- Charles.
After watching your last video I made a comment about using a better microphone. But in researching I discovered you need a contact microphone. It is sensitive to sound that travels through solid objects. Looking on amazon you can find them that plug into phones and used to tune musical instruments. Thy are the insensitive to air transmitted noise.
I bought one from Korg and litteraly cannot read anything else but environment sounds. When I use the mobile phone earphones I can record the tik tak perfectly. When I use the contact micro zero....What am I doing wrong?
Thank you sir for the multiple reviews on the app! While the Timegrapher is the definitive way to go, for casual tinkerers like me, the app may be good enough. :) Do they sell microphone/pickup adapters like the one attached to the Timegrapher that works with or attaches to a phone? Using the phone as the head unit? Appreciate the videos!
GeckoCkCkCk we are not testing quartz. These are mechanical watches . You are never going to get mechanical to a decimal point a day it’s always seconds per day
I was talking about accuracy of the quartz in the timegrapher and the mobile device. The accuracy of my laptop clock is all over the place, frequently deviating by 20 seconds.
GeckoCkCkCk timegrapher has a piezo microphone that’s listens to the tic and toc of the balance wheel. This is how it gets its timing from the software shows this in an understandable graph . Mechanical watches all have beats per second (vibrations per second) most 70s Seikos beat at 21600 vibrations per hour I can’t remember exactly what that is but it’s around 6 beats a second. You set this parameter in the machine or it will do it automatically , you then set the lift angle which is the angle the impulse jewel passes the pallet fork. With that dialled in the software can work it al out. It can tell if it’s running past or slow . Beat error is the accuracy of the tick and tock so if it’s swinging too far one way this is error showed in milliseconds. We can then adjust the watch to as good as we can get it. It’s all math , I’m not good at it but at least the machine is!
@@MyRetroWatches I think I've been misunderstood. Assuming the timegrapher and app count the same number of ticks from the mechanical watch, why would they give different rates? Because their reference clocks (quartz crystals) are at different rates. Is the quartz crystal in the timegrapher running at the same rate as the quartz crystal in your mobile device running the app? No. Which one is more accurate? It seems you have assumed the Chinese timegrapher is more accurate. The Apple mobile device could reach a time server and calibrate itself. So there is potential for the app to be more accurate than the timegrapher. I don't know how the timegrapher or app are calibrated. I don't know if they are temperature corrected.
GeckoCkCkCk put it this way most hobbyists and many professionals use this timegrapher . None use an app. Main issue is the app relies on a conventional mic and one that is too sensitive to background noise which is not practical to use to check a watch. The machine I can happily talk away whilst regulating and it only hears the watch. I don’t know if the machine or my phone has a quartz crystal and why it would be used in the math of all this. So the hardware in the phone could be better but it’s let down by the mic. The machine is purpose built for the task and on my eyes will always be the best option.
I'm no expert, but you are testing the devices differently. The open case back is a bowl. If this phone set up is so sensitive, could this change the acoustics? You should put the watch above the microphone just like the machine. No need to run the tests at the same time. We can see how close they are.
Very nice. My app works better. But can't use within a few feet of touch lamp, which raiates a strong low rf signal. Also try without florescent lamp or chinese led lamps. Both radiate supersonic and rf noise. Karl
its as accurate as the algorithm it uses plus you can test the watch real time against the atomic clock. 1000s people use this machine and I dont hear of complaints to accuracy but I do when sometimes the watch is running so poor that the machine can't really calculate what its hearing.
Pulling my hair out. The wieshi is a piece of software in a box with a mic. The app is a piece of software in a box that you attach a mic to. we know as smart watch people that all measurements are relative. So why in the world is an ugly 120$ box of plastic, running a piece of dated software that you can never update considered the baseline, against which all alternatives have to perform. A better comparison would be the 12$ app + $110 mic which roughly equals the plastic box on aliexpress. ah, you can also take that app to the watch shop or use the mic on your podcast.
Have my videos have helped you? why not support the channel in some way
You can make a one off/ monthy donation here in only one click: www.buymeacoffee.com/MyRetroWatches
If you would like to purchase any tools & equipment then please see my Amazon affiliate & eBay (I do earn a small commission from any sales.) Thank you.
Link: Link: myretrowatches.co.uk/tool-sale-links/
If you would like to buy any of my T-Shirts then please look here: teespring.com/stores/my-retro-watches
I know this is kinda old video but it was way so useful for me. I have Orient Mako 40 which was running +17 seconds per day. I purchased watch tuner timegrapher from app store and tried to regulate mako 40 with same app. After 1 hours of tinkering, I got -1 seconds per day. I've been monitoring the watch for around 2 days and it really runs -1 seconds per day, which is amazing result for me. Thank you for video, it helped me a lot.
ps: I put microfiber cloth on the desk, removed the watch's back cover and put the watch upside down on the cloth. Then placed Apple EarPods microphone on top of the watch mechanism. Then put glass jar upside down on top of the movement and microphone. That created nearly perfect sound isolation.
My tips for getting the app working better:
1. Stuff a folded up paper towel in the bottom of a glass, then put your movement (and the mic) on a towel on the table, then put the glass upside down on top of the movement. The sound isolation with that is very effective. The paper towels dampen any external noise and any tick tock echoes.
2. If you have a beat error, handle that first. It seems to affect the accuracy everything else.
Heya Mike, Sound engineer here. You were right to try a different microphone as I believe that's the primary difference between your time grapher unit and the app. In the end, they are both software using a hardware microphone. However you used the same type microphone as before, and I believe type matters quite a bit here. I understand the Weishi (pronounced way-shee by the way) uses a piezoelectric mic which is a little more sensitive to physical vibrations, less so for vibrating air as with the other mics.
Piezos are often used as pickups in musical instruments, acoustic guitars for example. They generally don't pickup sound great, but pickup vibrations beautifully. And that brings us back to the Weishi mic, which is probably depending on physical contact with the watch to work well. The Weishi is probably also filtering the bandwidth so it 'specializes' in hearing ticking, not everything, as a phone mic will. So, without a specialized watch mic, I doubt the app will be properly testable as compared to the Weishi. Apples, oranges, you know. 👍
I'm a 1yr late but thanks for that👍
I bought a piezo micro from Korg and cannot read anything. With the mobile micro I can record the tik tak perfectly but when using the contact micro I can only listen to surounding sound and zero tik tak. What am I doing wrong?
Test equipment and the regular calibration of test equipment is a science all unto itself.
It would be fascinating to see a comparison of two identical Weishi 1000 timegrapher machines with their peripheral microphones pitched against the same watch as a test calibration standard unit.
There is no doubt that both machines will produce different results with the same target watch.
But just how different remains to be seen.
That would be an interesting head to head test!
I just downloaded the app recently out of curiosity and use it mainly for pendulum clocks. That might be the one advantage where this app shines as it can also be used in larger clocks that can`t be fastened into the microphone aparatus. Great video!
Thank for your vídeos! I got the Chinese weishi 1000. It helped me today, my watch wasn't running bad, but I put on the TG and I could see slot of dots like snow. Inmediately I demagnetized the watch and now I can see to paralell and nice lines and a constant +3s. Without the TG I couldn't see that.
Thank you for watching and commenting. The TG is such a useful piece of equipment. A watch collector should not be without one in my opinion.
On further reading it's likely down to the sensor / mic being used. Most people have the best results with the PC application with a Piezo transducer. The 'electret' type microphone catches too much interference via EMF and other noise to be useful. I would be interested to see what's under the hood of the timegrapher mount - those four pozi screws look awfully tempting!
I purchased the app. Watch Tuner Timegrapher for my apple phone, and attempted to "regulate" my seiko ~ save the ocean ~ turtle. After about 15 attempts to get it close to 0 {+/- 5 sec.s} in only one position, I got it around -5-7 sec.s to +1 sec. Not so easy. BUT, I was originally thinking I didn't need anything, I'd just bump it a small amount faster (it was way slow, after tracking it with app., it was lik 45 sec.s slow a day... ). I can only imagine trying to get this watch anywhere close without the app. or a Weishi 1000 or something. Thanks for the comparison, and video. 🙂 Live and learn 🙂
How do we know if it isn't the app that's actually more accurate? I would like to see two or three of the Weishi timegraphers running at the same time and compare their values. Maybe the app's values will fall somewhere between the three dedicated timegraphers.
I was just about to comment the same. How do we know that the app isn't more accurate than the timegrapher. I would like to see both compared to one of the professional high end machines
It all comes down to the quality of the crystal oscillator in the device, electronic devices that measure high frequencies do so by comparing the input frequencies being measured with the frequency of the device's crystal oscillator which in a perfect world is 32,768Hz, but a low quality oscillator will usually not be temperature compensated and will drift as the temperature of the circuitry and oscillator change. A good timegrapher should have a high quality temperature compensated quartz oscillator in it, a phone or computer I'm not so sure, they generally aren't devices used for measuring high frequencies so I'm not sure manufacturers would bother using high quality oscillators. I work with crystal oscillators in geophysical data measuring instruments, there is a large range of precision/accuracy and cost when buying an oscillator.
Love the crystal on the Seiko, not seen that before...
MeesterDash it’s a Faceted Crystal. I have a few with these and they are very nice. Always a 70s Watch and quite often an Actus. Some would have had them from new but as they are so hard to find if changed it would have had a plain domed crystal fitted.
Thanks for sharing your experiments. I got decent results using a windows app and a piezoelectric mic. I had an old instrument tuner which has a piezoelectric mic with a convenient plug. I had to run it through a preamp on full volume and then I could even hear the ticks (barely). It works but would require more amp and some EQ to filter noise, but it is not very convenient. I see the comment of the sound engineer guy and I think he cracked it, piezoelectric is the solution. You could tweak to get better results but if you have to buy the sensor, a preamp, filters, you might just buy the device. If you are an engineer and like a project this would be an interesting circuit to make.
I appreciate your response to previous feedback. I suspect the Weishi designers analyzed the audio from a watch and filtered the audio and selected the input device for maximum efficiency. That could take a lot of tinkering to duplicate.
Similar apps are available for Windows, and I thought about trying one to save the cost of a Weishi. Here's the thing: you put an app on your phone, Mac, or PC, and you've added to its complexity, impacted its battery life (unless it's using mains power), and tied it up from other uses while you're using it for time graphing. And, as we see here, you're also buying external peripherals and fussing over settings to see if it's in any way accurate.
The beauty of a dedicated device for this is that it's doing only what it's designed to do, it's doing it in excellent fashion, and it's doing it with no hassle. It's roughly analogous to using a properly-sized box wrench instead of an adjustable spanner or pliers to loosen a nut.
I imagine a professional can learn more details about how a mechanical watch is going by using a very detailed application on a PC with a proper mic and vice, and I imagine such apps exists for testing a quartz watch as well (I know there are dedicated devices for it).
Sometimes I miss the days when we used a phone for making contact, a camera to take pictures, a map to navigate, etc. Everytime I hear, "There's an app for that," I think, yeah, and it's probably marginally effective sometimes.
I have an app on my laptop very similar to what you have on your phone. I also have a TGBC cheapy timegrapher device. I found that when I try to check a watch while it's in it's case it does not work as well as if I check just the movement without the case. The unit came with a cable with a type c connector on one end that plugs into the timegrapher and a USB connector on the other end that plugs into my laptop. With just the movement, the parameters and lines show up very well. I don't have a regular timegrapher like yours to compare with but I did use a watch that I know is running well and the rate was +2, beat error was .1, amplitude was 298 and the line graph was straight and consistent. As I said though without having something similar to your Weishi 1000 to compare it with, I don't know how accurate those numbers really are.
The app IS useful. I used to to tune a Miyota 9015 based watch from +12 sek/day to about +2 sek/day. Ordered the Timegrapher 1900 model , but delivery is 2-3 weeks. While waiting for delivery, the app has definitely made mere more happy with my watch performance.
Agreed the app does have its use as a guide but testing against a dedicated machine does show its weak points. Mainly the complete silence you have do be in to use it. I hope you still found the video useful and thank you for your comment.
Found the video very useful - and it helped me decide to purchase the proper tool.
I think the best microphone solution would be a contact mike. I think I could make one for about $3. You are also assuming that the results from the Weishi are perfect and infallible. I am sure they are in the ballpark, but I am much less convinced that they are perfect, and in fact I'm a bit dubious about it's count resolution due to the age of the technology.
There are also Android apps available (and Android Phones that still have a 2.5mm audio jack).
Yeah. How do you know the Weishi is accurate?
I think you would have to use a oscilloscope to check the accuracy of the measurements.
@@Todestelzer there's a special jack on the back for doing calibration. I'm not sure how that works. I have a Weishi on the way so maybe the manual has more information (or maybe not, we will see).
I made the contact mike and it worked very well, but the android software left me rather unsatisfied. None of the programmes I looked at did amplitude so I didn't go ahead and build the microphone into a movment holder. The apps are probably OK for the average user, but since I'm stripping and servicing movements I really need to know what the amplitude is... it's the main indication of whether you've improved the movement or serviced it to an acceptable degree. It's possible to analyse the amplitude by other means, for example by making a video of the movement then measuring the full angle of rotation, but having a timegrapher on the workbench would be so much easier.
Been 5 years. Not sure if they improved it since, but vid I watched just out finds they're very close, within 1-2 seconds which is accurate enough for most.
I have been using a free application called Clock Tuner, which is essentially the same thing you have. I simply press the watch face to the spot on the phone where the microphone hole is, and I get good results. But first, I move to a room with no noise, or eliminate noises, like the Telli, radio, talking, etc. Perhaps you could take the foam cover off your microphone and hold it to the watch?
Cheers
Yes. Would be great to see your collection
By the way, with the system I'm using (TGBC Timegrapher, WatchOScope app. and USB cable hook-up), External sounds do not seem to affect the readings.
Kool, thanks for the heads up on these apps Mike, I appreciate it and G'morning 😇 . I don't have a time graph for yet so for now the time graph for apps will give me something to play with just for fun or to possibly get a baseline idea on how things are running 🤓 ...
Thanks for the update. May not be perfect but for a entry level option, it worth the price. Just a FYI, lav mic are unidirectional, meaning it just picks up directly from in front of the mic head
Thanks, unfortunately it was / is the only mic I had at the time. Granted the app is goo enough for a beginner. Thanks for taking the time to comment and watch.
Not really, some of them are pretty directional, some of them are omnidirectional, you need to look at the specs
Great test and comparison Mike.
thanks a lot Mike for this new video and for doing this comparison with such an open mind.
my guess (but I'm not an expert at all) is that the app is probably wrongly interpreting a parasite noise on the tock - assuming the tick was right, and that affects the beat error calculation. the weishi machine is clearly superior there with dedicated sensor and software. if my guess is correct, what is instructive is that the app shows there is something abnormal with the tock (friction?). I think the original Swiss made witschi timegraphers can analyze the tick and tock sounds spectrum and identify that kind of abnormal noise.
but then of course we are not talking about tinkering hobbyist 😀
nor the same budget 😭
thanks again for your great videos. looking forward the next one!
l10n3l
thanks for your videos on the weishi 1000 timegrapher, I have the same one. I just wish I could find the lift angles for vintage American pocket watches. These are the watches I mainly repair and collect along with my vintage Seiko wrist watches. If you find a source for the pocket watch lift angles please let me know.
The lift angles are easily calculated (videos on youtube show how). Make your own charts for those that have no published angles.
What are those smaller sized 2nd line values in each column? (small white font -2, -4, -1, -2 etc.)
The ones below the 1st line, (highlighted, larger font) actual values?
Would love to see your collection in a video Mike. Although it would have to be a series with the number you have 😆
Try turning the beat detection noise-cancelling OFF.
I used a similar app to check my watch (Orient caliber F49). To minimize the external sounds, I made the tests at night, and put the watch inside the wardrobe, covering the watch+microphone with a fabric. After starting the test (5 minutes) I closed the wardrobe door and let the thing happen. Got an avg 1sec/day and .1ms. After some research, saw that the F49 isn't all that precise, so I ended up with some questions LOL. Is the watch really that underrated or something went wrong while testing it
Anyway, great video, man
I always do a run-through if I'm doing something for the first time. Practice help avoid the little (awkward moments.) It would have helped you to know which watch would give you trouble, or what to do for the best results, so the viewers know how to get the proper or best result (withstanding unforeseen outside effects) and to point out important things, like the straight graph lines, angled lines, "snow" etc. But overall, it was a good for general knowledge.
Cheers,
Thank you. Granted more planning or explanation could have been used, I do a lot in one take and I dont want videos to be too long as statistically when looking in my analytics people usually loose interest after 10 minutes. This video itself since May 2019 has an average view time of only 6 minutes. At some point in the future I would like to do a more comprehensive video on the timegrapher machine and if that happens I will take into consideration some of your points. Thanks
I would very much like to see videos of your vintage watch collection please! Maybe you could do a video of each watch. thank you.
william rizzo I have already done some of the collection videos. Please see my other videos. Still have more to film
@@MyRetroWatches will do and thank you, when this video popped up I did not realize it was not current. Thank you again.
I found the app. It appears to be Watch tuner timegrapher. How does it compare to watch accuracy meter ?
The problem also is the fact that even if you are able to tweak the app to an appreciable degree, you won’t have any reference to compare it to in order to know this, and if you do have another reference you don’t need the app
Jeff Dumon I think the point I am proving is that an app is good for a quick check for those who are curious . If you want to tinker more then a timegrapher is a purchase that you cannot avoid.
Thanks, I was over thinking the issue. The app is quite a neat idea, I just wanted it to be better than it is. So I am pleased you have produced this video.
To be fair the weishi also shows different results each time, so how to be sure it is correct vs the app...
Razor sharp video en always in focus!!!.....A pleasure to watch
Which Camera do You use
Also Very good review😀
Great test, convinced me to get the machine.
+BigDuke6ixx wise decision. Good luck sir
So it's now Nov 2021 and this app maker now has a couple of cut down versions of this app as well. So I bought one of the 'lite' apps and I actually read the help page. The instructions clearly state "...make sure that the crown of the watch and the microphone from the earbuds are touching." My earbuds are a no name cheap as chips brand but I get anywhere between 98-100% signal quality. I also live in Jakarta where there is a constant stream of noisy motor scooters roaring past. It doesn't seem to worry the app at all. They may have upgraded the app in the time since this video was made, but it appears to me that the quality of the microphone has little bearing on the matter of signal quality. Just follow the app maker's recommendations.
That maybe but how accurate is the results. Problem is you have nothing to compare it to.
The most limiting factor is that you would not be able to regulate a watch real time using this method easily.
I am enjoying the video and really was hoping that there would be an app that would give acceptable or at least reasonable results. I have a few watches I'd be curious to check (one Swiss, one Japanese, and a cheapy Chinese), I really am not going to do this enough to justify spending the money on a Timegrapher.
Suggestion, when trying different things, make ONE change at a time. If you change the microphone and two different settings, you are just whistling in the dark since you don't really know what changed what. One item may have helped and the other two messed it up.
Thank you for your appraisal! Two questions: What adaptor did you use to connect the 3.5mm jackplug of the lapel microphone to the lightning connector on the iPhone; and secondly would you like to hazard a guess as to whether a little padded box might exclude background noise which might interfere with audio pick-up? Thanks otherwise for a very level-headed assessment!
Thanks Dave. for the adapter Apple made one and supplied it with I think the iPhone 7 when they first did away with the socket. so its a lightning to jack adapter looks like this. amzn.to/3m8LFzV
As for sound deadening , yes you could do this but to me its extreme and then impossible to actually work on the watch to correct the regulation.
@@MyRetroWatches Thanks, Mike. I actually downloaded that timegrapher app for the apple watch, but found the microphone of the phone not sensitive enough for some of my Seiko movements (or the background noise too extreme), so also invested in a Levalier lapel microphone, which is definitely more sensitive, and gives me a more realistic set of data using the app. Definitely not good enough for a regular watchmaker, but enough to allow me to assess whether my couple of watches are very bad, or just bad (ha ha). Keep up the good work!
I really enjoy your videos, they teach me and like so many of the other posters, get me to wondering about so many of the variables. Since the Weishi doesn't record an average over the length of the test, you basically can only compare the end results shown. With that being said, they were both extremely close except for the beat error in this revisited video. My question is, if the watch is off by say 8 seconds a day. what would be the beat error? Would it be like .3 or minuscule? What else can cause a watch to lose or gain time besides the beat error? This video has opened a pandora's box of questions for me to think about and I'm sorry but your the only magician with the answers that I know. :)
Hi Danny and thanks for watching and commenting. Interesting question however beat error is separate to rate with error in beat you can still regulate to a satisfactory level but its the long game with these. Beat error is the error between the tick and the tock, in an ideal world you want them to be as equal as possible as a millisecond over a long period of time will add up. Accurate beat error also effect the amplitude. You can have a beat error of say 0.3 and still regulate the rate to within a few seconds per hour. The ultimate goal is to have one nice straight line across the TG and then in all 5 positions that the watch should be checked in (dial up dial down etc) My full understanding of the weshi and how its plots the graph is limited. I assume the longer you leave it on the more "average" the reading. I often leave movements on there all night to see how its performing long term. This is now the question I have..
@@MyRetroWatches Thanks for your prompt reply Mike! I haven't memorized the proper names for most of the watch parts yet, but with your help and some research I see what you mean. Beat error is centering the mainspring and rate is shortening or lengthening the mainspring, two different animals! By the way, I just received my Weishi TG. I don't plan on working on anything yet, still need to learn more and acquire tools, but I may see how some of my watches are doing with this equipment. Thanks again and keep posting these helpful videos.
another quality video Mike. For my money, the app is a huge waste of time. The biggest problem of which is how does one know what the correct reading is, especially without something to compare it to.
As you said, you could tweak the thing all night and you stil wouldn't know what to believe.
The old adage is apropos here - buy the best tool for the job and you buy it once-buy a cheap tool and you will buy it more than once.
In the USA the Timegrapher costs approx. $150. Well worth it and a lifetime investment if you are a watch enthusiast. Plus no need to monkey around with adjusting settings and multiple takes.
Thank you sir. Nice comment and exactly right. buy cheap buy twice...
Sadly I can't compare my results to a "real" time-grapher, but I will say that "Watch Accuracy Meter" also seems all over the place WRT beat error and amplitude. Overall deviation seems relatively accurate however - at least when compared to my observations recorded via "WatchCheck" over multiple days of wear and/or sitting on the winder. And that's without me using an external microphone ... I just put the crown right next to my Galaxy S20's built-in mic.
Of course, I'm a bit weirded out by very low and/or erratic amplitude results and high beat error for brand new or 12 to 24 month old movements - but I guess there's a reason why you can only set lift angle and get amplitude results in that app by activating "experimental" mode. :D
I used a app to regulate a couple watches so it is useful in that regard. I used a app to determine if a watch needed service by the ampltude so that is usefull. In short the app is good enough for government work.
The app itself may be better than a timegrapher. I would set it up with not a lapel mike, but the mike used with a timegrapher. Or with a calibration mike, or a calibration mike used with the timegrapher measurement body used to rest the watch on. You never know. The timegrapher algorithms may be set up for gross averages only. I'm not a watchmaker. I do have experience with professional audio and audio measurement components. Just a passing thought. Great Video. Like the measurement music choice.
Perhaps but the app does not tell you to use anything more sophisticated than the mic with your headphones.
The machine will always win if you want to work on a watch and get a reading at the same time.
The timegrapher uses a piezoelectric microphone
Thanks for the response.
Now it probably sounds like heresy, but I do agree with Mark, in the sense that I believe the app is sometimes out-performing the timegrapher. My experience is with digital signal processing, building machines not so dissimilar from the timegrapher. It’s easy to see that they are both making mistakes, but they are different ones, leading to different results.
Earlier tests from the first video make it pretty obvious that the app wasn’t receiving good information from the microphone - and the results were therefore bordering on uselessness. Garbage in, garbage out. On the flipside, the app is pretty overt (IMO) about how it goes about making its most basic calculations, and the inclusion of the standard deviation is a significant improvement over the timegrapher. Also, it should be noted the timegrapher likely produces a display that is artificially clean, due to its somewhat lower resolution display and summarized reporting. The phone has orders of magnitude greater processing power. So, with the correct DSP applied, along with the option of a longer integration window (essentially unlimited memory vs 12 sec), the app should actually be able to put the purpose-built device to shame. (Though the flexible, rotating stand is also very handy.)
But, does the app (or other apps) have a robust algorithm? - I can’t say. What I can say is that I have now been inspired to drag out my pro gear and do a little bit of an analysis of the raw audio against a high-precision/high-stability time standard. I can then compare that to the app and see how they measure up. (Because, apparently, I feel it’s a wise decision to use $20K of test gear and do hours of tedious work in order to save 100£. Hmm. Must be the watchmakers’ gene mutation…)
What app is that? It does not look like the watch accuracy meter.
The app is pretty good for what it is. It's the best one I've seen by far. That said, if it can't read the beat error correctly, then it is useless to me. I can't believe that I was without a timegrapher for 2 years after starting the hobby. I would have to stay absolutely quiet for hours at a time just to get my watches anywhere near close in timing. Then I got my timegrapher and realized just how bad my beat of error was off and how low my amplitude was a lot if my watches. I wish I had mine sooner, because a lot of watches would have been sent back that I thought were ok. For anyone with any mechanical watches a timegrapher is just a must have.
Hello and thanks for commenting. I can completely relate to this. I do struggled with the app for 3-4 months before spending the money on the timegrapher. You then wonder why you didnt do it right at the start.... Same can be said for a microscope both pieces of equipment are complete game changers to your watchmaking efforts.
What do we know about the accuracy of the Weishi? I see these being cranked out by the Chinese but at what standard?
Interesting Michael!
The app gives a decent ballpark value. If you buy a new 60$ seiko with a 7s26 and the app shows +20s, you can easily make it more accurate. Tweaking/servicing expensive or vintage watches tho would require precise measurement, that's where the timegrapher shines :)
That is a Omnidirectional mike and picks up sound 360 degrees, I would use a unidirectional Shotgun mike it concentrate on one point and ignores sound from other direction.
good point however like I think I said in the video its the only other mic. I had. This video was really because of curiosity after the first video as to if I could make it better. In all seriousness nothing will compare to the Timegrapher machine, its purpose build and uses a piezo mic . It was just nice to compare the two to see just how close the app came to the real thing.
Hi Mike Nice video 👍
I think an issue with this setup, not the way that you're using it but just inherently the way that the setup works, is that you are having to use the dedicated time grapher as a reference and then tweaking the phone setup to get as close to the time grapher as you can. If somebody is trying to use the phone app on its own and they don't already have the time grapher, they have nothing to use as a point of reference to try to calibrate the phone for maximum accuracy, which you need a time grapher for, and if you already have the time grapher, then why not use that instead of trying to tweak an inferior system to work and then using the inferior system? You might know somebody who will let you borrow their time grapher to calibrate your phone setup, but again, if they're willing to loan you their time grapher than why not just use the time grapher you're borrowing.
Lovely timepieces
Thanks for revisiting this Mike, I did wonder whether a better mic would improve the app. My mind is made up...Weishi 1000 all the way :-)
My Mate VINCE good stuff Vince. Thanks.
Hope to film the regulation video at the weekend
@@MyRetroWatches Looking forward to it :-)
In principle, a good mic and amp setup should definitely make it equivalent - but it would basically require using the mic and stand setup of the weishi with a custom interface to the phone. And I don’t see that being much cheaper than the whole package, because the cost of that weishi is not residing in the cheap little box and screen.
A trick : watch movies at Flixzone. I've been using them for watching all kinds of movies recently.
@Landry Benjamin Definitely, have been watching on Flixzone for years myself :)
Curiouser and curiouser. It wouldn't by any chance be possible to use the Weishi microphone with the phone, would it? That would be a useful control test, you'd think. Please note that I'm not egging you on to do it, just musing aloud. :-)
GenWivern2 there is a small chance but I don’t think that mic is doing anything .
I just think the app is not really all that up for listening and interpreting what it hears .
@@MyRetroWatches Hmm. Reading up about microphones to suit the TG programme I mentioned in the comments on the last video made my head hurt. Apparently the transducer type and the mounting are both, if not exactly mission critical, definitely going to have a significant effect on the signal quality. From what I can gather the Weishi uses a hard-mounted piezo, which is a very different animal from the Boya ... it also uses a highly inconvenient connector by the look of things making a comparison impracticable! Anyway, for the very little it's worth I suspect the microphone rather than the software. That's not to absolve the app - after all it's highly unlikely that the average user is going to have an optimal microphone set up, or anything like.
Oh, and forgot to say that a collection video would be very welcome.
GenWivern2 it was always comparing apples with pears . Interesting that you have gone more in depth on the microphones .
@@MyRetroWatches I have a bad habit of getting sucked into researching things, rather than getting on and doing them. My take on all this - pretty worthless of course - is that you can save say £90 on the price of the Weishi by going with the app, but you incur a good £90's worth of hassle with no guarantee that it'll be reliable, so you may as well go with the timegrapher and rest easy. :-)
The app instructions say to place the mic on the crown. Would that make a difference?
I cant advise. I cannot see how that would make any real difference if i am honest.
Mike -- Thanks for the videos. I'm a new owner of an 1890 Waltham, trying to understand what's responsible for its imperfections. I'm using a free Android app ("Watch Accuracy Meter") which is similar to the one you're using on your iPhone. I have a question, applicable to the Timegrapher and app:
. . . How is the amplitude calculated ?
There must be more happening in the audio stream, than a "tick - tock - tick - tock". You can get "beat error" from those times, but not amplitude. Thanks again -- Charles.
I don't know how they work but its probably the same algorithm that is in the phone software.
Can I get an app .. thanks
After watching your last video I made a comment about using a better microphone. But in researching I discovered you need a contact microphone. It is sensitive to sound that travels through solid objects. Looking on amazon you can find them that plug into phones and used to tune musical instruments. Thy are the insensitive to air transmitted noise.
Would be interesting to see results with a contact microphone.
I bought one from Korg and litteraly cannot read anything else but environment sounds. When I use the mobile phone earphones I can record the tik tak perfectly. When I use the contact micro zero....What am I doing wrong?
Thank you sir for the multiple reviews on the app! While the Timegrapher is the definitive way to go, for casual tinkerers like me, the app may be good enough. :) Do they sell microphone/pickup adapters like the one attached to the Timegrapher that works with or attaches to a phone? Using the phone as the head unit? Appreciate the videos!
Thanks for watching. the honest answer is I dont know. Mark Lovick from The Watch Repair Channel made his own TG using some software and a microphone.
Mic on amazon and ebay.
Please do another test
Kidding
The scale on the app is larger. How do you know which one has a more accurate quartz? Your timegrapher doesn't give seconds to any decimal points.
GeckoCkCkCk we are not testing quartz. These are mechanical watches . You are never going to get mechanical to a decimal point a day it’s always seconds per day
I was talking about accuracy of the quartz in the timegrapher and the mobile device. The accuracy of my laptop clock is all over the place, frequently deviating by 20 seconds.
GeckoCkCkCk timegrapher has a piezo microphone that’s listens to the tic and toc of the balance wheel. This is how it gets its timing from the software shows this in an understandable graph .
Mechanical watches all have beats per second (vibrations per second) most 70s Seikos beat at 21600 vibrations per hour I can’t remember exactly what that is but it’s around 6 beats a second.
You set this parameter in the machine or it will do it automatically , you then set the lift angle which is the angle the impulse jewel passes the pallet fork. With that dialled in the software can work it al out. It can tell if it’s running past or slow . Beat error is the accuracy of the tick and tock so if it’s swinging too far one way this is error showed in milliseconds. We can then adjust the watch to as good as we can get it.
It’s all math , I’m not good at it but at least the machine is!
@@MyRetroWatches I think I've been misunderstood. Assuming the timegrapher and app count the same number of ticks from the mechanical watch, why would they give different rates? Because their reference clocks (quartz crystals) are at different rates. Is the quartz crystal in the timegrapher running at the same rate as the quartz crystal in your mobile device running the app? No. Which one is more accurate? It seems you have assumed the Chinese timegrapher is more accurate. The Apple mobile device could reach a time server and calibrate itself. So there is potential for the app to be more accurate than the timegrapher.
I don't know how the timegrapher or app are calibrated. I don't know if they are temperature corrected.
GeckoCkCkCk put it this way most hobbyists and many professionals use this timegrapher . None use an app. Main issue is the app relies on a conventional mic and one that is too sensitive to background noise which is not practical to use to check a watch. The machine I can happily talk away whilst regulating and it only hears the watch.
I don’t know if the machine or my phone has a quartz crystal and why it would be used in the math of all this.
So the hardware in the phone could be better but it’s let down by the mic.
The machine is purpose built for the task and on my eyes will always be the best option.
An análisis using a normal mic is a good result.
Great!!!! Thanks!!!!
I'm no expert, but you are testing the devices differently. The open case back is a bowl. If this phone set up is so sensitive, could this change the acoustics? You should put the watch above the microphone just like the machine. No need to run the tests at the same time. We can see how close they are.
Looks like one has to buy the table top device just in order to calibrate the app :-)
Very nice. My app works better. But can't use within a few feet of touch lamp, which raiates a strong low rf signal. Also try without florescent lamp or chinese led lamps. Both radiate supersonic and rf noise. Karl
The new mic seemed worst than the Apple EarPods mic
Based on that the machine is accurate...
its as accurate as the algorithm it uses plus you can test the watch real time against the atomic clock. 1000s people use this machine and I dont hear of complaints to accuracy but I do when sometimes the watch is running so poor that the machine can't really calculate what its hearing.
Pulling my hair out. The wieshi is a piece of software in a box with a mic. The app is a piece of software in a box that you attach a mic to. we know as smart watch people that all measurements are relative. So why in the world is an ugly 120$ box of plastic, running a piece of dated software that you can never update considered the baseline, against which all alternatives have to perform. A better comparison would be the 12$ app + $110 mic which roughly equals the plastic box on aliexpress. ah, you can also take that app to the watch shop or use the mic on your podcast.
Let’s take this in a serious scientific direction... TURN THAT SHIT UP TO 11
You make the assumption that the Weishi is an absolute reference. That is not a safe assumption.
:)
Следующий выпуск, на русском языке!))