BRIFFIDI - SW1 Racquet Swingweight Machine
- Thread starter JEDI MASTER
- Start date
Shroud
Talk Tennis Guru
bfroxen
Semi-Pro
The SW1 calibration procedure, using two objects of known swingweight, subtracts out the swingweight of the phone, cradle, etc. Most other machines can be calibrated with one object of known swingweight, as the swingweight of the cradle is fixed.
I wrote a blog article about discrepancies between machines a while back: https://www.briffidi.com/2026/01/30/accuracy/
I wrote a blog article about discrepancies between machines a while back: https://www.briffidi.com/2026/01/30/accuracy/
Brando
Professional
Recapping, my SW1 consistently read ~5sw higher than RDCs across multiple racquets. So I raised the app’s MoI setting 5 points above the app’s preset of 296.8. This worked OK but it bugged me that I couldn’t figure out why my SW1’s readings seemed off. So I never stopped looking for a cause.
Along the way, I had the occasion to cradle two different phones in my SW1 and noticed that the phone that was 50 grams heavier read 1sw point higher. This prompted me to wonder if phone weight differences might somehow account for my and others’ ~5sw differential.
Using key-words, I searched for any intel @bfroxen may have offered on TT about this and found nothing. But then, yesterday, I happened upon his post # 662:
Moving on from cause to solution, each rod has a natural range of periods when you release it consistently. My rod’s periods average around 1.155450 – 1.15550 seconds when my release form is good. Further, I learned that this 0.00100 period range translates into a span of swingweights around 0.5sw wide; to me, an acceptable margin of error.
This discovery prompted me to flip the script. Instead of calibrating to the MoI printed on my rod (296.8– that consistently produced swingweights 5 points too low), I’d calibrate to my rod’s most natural period, 1.15500, and then reset the MoI accordingly. By accordingly, I mean according to a ‘control frame.’ Mine is named Flower, and she’s a 360 Radical MP modded to a 330.3sw confirmed by Racket Doctor in L.A., whose racquet techs recalibrate their Head 3-in-1 on the daily.
For the same reason I had to suss out my rod’s average period, I sussed Flower’s period mean (when aptly released) to range from ~1.20350 to 1.20450 seconds. Then I followed a procedure:
1) Calibrate the rod to a reading that’s as close as you can get to the mean of your rod’s natural period range; in my case 1.15500 seconds.
2) Reset MoI in 1-to-1 proportion with the amount of swingweight you’re seeking to add or subtract from your readings; knowing that, at any period, the higher the MoI it’s calibrated to, the higher the SW output will be. To execute this, press the MoI # in the “CALIBRATION OBJECT #2” section and the app will take you to a simple reset & save screen.
3) Return to the main calibration screen and press the “Ready to Calibrate” button (once you’ve also done your “CALIBRATION OBJECT #1” with the rod fully collapsed), see the calibration-complete notice, and move to the “Measurement” screen.
4) Load in your control frame and attempt to release 3-count groups with period means that are within its acceptable range (in Flower’s case 1.20350 – 1.20450), deleting the rest.
5) Evaluate the swingweight outcomes in terms of whether their average period means (avg. p/m) are within average range, and use this data to repeat the process with an MoI that gets you closer to your SW target (in Flower’s case 330.3sw).
Here are my initial calibrations and the subsequent swingweight readings on my control frame:
Analyzing the first row from left to right, having hit my calibration target of 1.15500, I reset my first MoI 5 points higher than my rod label of 296.8, at 302. After calibration, I loaded Flower in to get a 3-count of 330.0±0.02, basically my 330.3sw target. But wait. This first 3-count, as well as the second (check-count) at 330.1±0.04, have avg. p/m’s (period means) way over my average of 1.20350 – 1.20450 seconds, at 1.20600+. This means these counts were excessively slow, producing artificially high swingweights because evidence suggested I'd withdrawn my finger too slowly from the cradle.
The second row, as with all the rest, hits my calibration period target of ~1.15500, so I raise MoI by a point (to 303) with the idea that a lower, in-range period will decrease SW, so I need to bump up my MoI # to compensate. My expected outcome was at least 330sw. What happened? I got 329. That SW went down instead of up seemed suspect until I noticed that the periods are now below my averages at, 1.20250, instead of above them before at 1.20600.
Why? I’d overcompensated for the previous round, withdrawing so fast that I inadvertently pressed in on the stop before release, adding energy to the system. But then, I was just learning how sensitive SW1 is to technique changes.
The 3rd calibration hit paydirt, setting Mol at 304 and getting a 1.20354 period (at the low end of my objective 1.20350 – 1.20450) that produced a 330.6±0.03 swingweight. Since that’s 0.3sw over my 303.3sw objective, I tweak MoI down from 304 to 303.7 in the 4th calibration and get 330.2±0.05 via an edge-of-low avg. p/m that’s close enough to tell me this MoI to period combo will read swingweight correctly.
Here it’s worth noting that SW1 has linear consistency; if a calibration correctly reads one racquet’s swingweight, it’ll correctly read all others too. This also means that I could’ve correctly calibrated the rod to, say, 1.15600 vs. 1.15500. The apt MoI to calibrate with a 1.15600 period would be higher than 303.7, but no less accurate. This means you can choose your own best period per your cradle release style. What matters is not the period number, per se, but that your style, whatever it is, consistently pulls the rod within that range.
Given the length of this post, some will scan it and get the impression I’m critical of the SW1. I’m not. I LOVE this device. The twistweight attachment has paid for my entire unit in intel alone. If the SW1 weren’t as accurate as it is, none of my experimentations would be as dependable as I believe they are. Indeed, my belief in the device’s accuracy is precisely what prompted me to look for other causes for the apparent discrepancy.
Along the way, I had the occasion to cradle two different phones in my SW1 and noticed that the phone that was 50 grams heavier read 1sw point higher. This prompted me to wonder if phone weight differences might somehow account for my and others’ ~5sw differential.
Using key-words, I searched for any intel @bfroxen may have offered on TT about this and found nothing. But then, yesterday, I happened upon his post # 662:
So it's obvious @bfroxen has already considered this as a cause and has concluded that it couldn’t add enough mass to account for a 5-7sw difference. That, then, puts my phone weight idea to bed.
Moving on from cause to solution, each rod has a natural range of periods when you release it consistently. My rod’s periods average around 1.155450 – 1.15550 seconds when my release form is good. Further, I learned that this 0.00100 period range translates into a span of swingweights around 0.5sw wide; to me, an acceptable margin of error.
This discovery prompted me to flip the script. Instead of calibrating to the MoI printed on my rod (296.8– that consistently produced swingweights 5 points too low), I’d calibrate to my rod’s most natural period, 1.15500, and then reset the MoI accordingly. By accordingly, I mean according to a ‘control frame.’ Mine is named Flower, and she’s a 360 Radical MP modded to a 330.3sw confirmed by Racket Doctor in L.A., whose racquet techs recalibrate their Head 3-in-1 on the daily.
For the same reason I had to suss out my rod’s average period, I sussed Flower’s period mean (when aptly released) to range from ~1.20350 to 1.20450 seconds. Then I followed a procedure:
1) Calibrate the rod to a reading that’s as close as you can get to the mean of your rod’s natural period range; in my case 1.15500 seconds.
2) Reset MoI in 1-to-1 proportion with the amount of swingweight you’re seeking to add or subtract from your readings; knowing that, at any period, the higher the MoI it’s calibrated to, the higher the SW output will be. To execute this, press the MoI # in the “CALIBRATION OBJECT #2” section and the app will take you to a simple reset & save screen.
3) Return to the main calibration screen and press the “Ready to Calibrate” button (once you’ve also done your “CALIBRATION OBJECT #1” with the rod fully collapsed), see the calibration-complete notice, and move to the “Measurement” screen.
4) Load in your control frame and attempt to release 3-count groups with period means that are within its acceptable range (in Flower’s case 1.20350 – 1.20450), deleting the rest.
5) Evaluate the swingweight outcomes in terms of whether their average period means (avg. p/m) are within average range, and use this data to repeat the process with an MoI that gets you closer to your SW target (in Flower’s case 330.3sw).
Here are my initial calibrations and the subsequent swingweight readings on my control frame:
Calibration Period | MoI | Flower SW | Flower avg. p/m | Flower SW recheck | Flower avg. p/m |
1.15507 | 302 | 330.0±0.02 | 1.20608 | 330.1±0.04 | 1.20624 |
1.15502 | 303 | 329.0±0.04 | 1.20250 | 329.3±0.05 | 1.20300 |
1.15508 | 304 | 330.5±0.08 | 1.20329 | 330.6±0.03 | 1.20354 |
1.15505 | 303.7 | 330.2±0.05 | 1.20341 | 330.0±0.02 | 1.20302 |
Analyzing the first row from left to right, having hit my calibration target of 1.15500, I reset my first MoI 5 points higher than my rod label of 296.8, at 302. After calibration, I loaded Flower in to get a 3-count of 330.0±0.02, basically my 330.3sw target. But wait. This first 3-count, as well as the second (check-count) at 330.1±0.04, have avg. p/m’s (period means) way over my average of 1.20350 – 1.20450 seconds, at 1.20600+. This means these counts were excessively slow, producing artificially high swingweights because evidence suggested I'd withdrawn my finger too slowly from the cradle.
The second row, as with all the rest, hits my calibration period target of ~1.15500, so I raise MoI by a point (to 303) with the idea that a lower, in-range period will decrease SW, so I need to bump up my MoI # to compensate. My expected outcome was at least 330sw. What happened? I got 329. That SW went down instead of up seemed suspect until I noticed that the periods are now below my averages at, 1.20250, instead of above them before at 1.20600.
Why? I’d overcompensated for the previous round, withdrawing so fast that I inadvertently pressed in on the stop before release, adding energy to the system. But then, I was just learning how sensitive SW1 is to technique changes.
The 3rd calibration hit paydirt, setting Mol at 304 and getting a 1.20354 period (at the low end of my objective 1.20350 – 1.20450) that produced a 330.6±0.03 swingweight. Since that’s 0.3sw over my 303.3sw objective, I tweak MoI down from 304 to 303.7 in the 4th calibration and get 330.2±0.05 via an edge-of-low avg. p/m that’s close enough to tell me this MoI to period combo will read swingweight correctly.
Here it’s worth noting that SW1 has linear consistency; if a calibration correctly reads one racquet’s swingweight, it’ll correctly read all others too. This also means that I could’ve correctly calibrated the rod to, say, 1.15600 vs. 1.15500. The apt MoI to calibrate with a 1.15600 period would be higher than 303.7, but no less accurate. This means you can choose your own best period per your cradle release style. What matters is not the period number, per se, but that your style, whatever it is, consistently pulls the rod within that range.
Given the length of this post, some will scan it and get the impression I’m critical of the SW1. I’m not. I LOVE this device. The twistweight attachment has paid for my entire unit in intel alone. If the SW1 weren’t as accurate as it is, none of my experimentations would be as dependable as I believe they are. Indeed, my belief in the device’s accuracy is precisely what prompted me to look for other causes for the apparent discrepancy.
Last edited:
Brando
Professional
My last post was already too long, so I didn’t include a point I've kinda been itching to write since…
The reason my SW1 consistently reads ~5 points lower than the RDC I’ve typically relied on (a Head 3-in-1) may be that it’s not the SW1 that’s misreading swingweight but the RDC. I’m not talking about a miscalibrated RDC. I’m saying that it’s possible the SW1 is inherently more accurate than an RDC because it’s more linearly accurate.
One known problem with RDCs, as with weight scales, is ‘measurement creep.’ This is a phenomenon where the farther a frame’s swingweight is from the rod’s calibration MoI, the less accurate the SW measurement is likely to be. @bfroxen goes into this in his post # 662, and it’s worth digesting the whole thing. But the upshot is that his research indicates that older Dunlop and Babolat RDCs lack the linear accuracy of the SW1. I, for one, believe it.
Still, the reason I hold with the Head 3-in-1 readings I’ve received at Racket Doctor in L.A is that over a dozen of the unstrung frames measured there had already been measured by Tennis Warehouse’s Matching Service, and all the readings conform to within 1sw point of each other (Racket Doc’s vs. TW’s), often less than a point. Could both RDCs be equally off in their swingweight measurements? Absolutely.
At this point, I’m thinking the best way to confirm the accuracy of the MoI labeled on my collapsible rod is to use the Rod Swingweight Calculator built into bfroxen’s blog on Swingweight Machine Accuracy to fashion a verification rod out of PVC piping. The directions are clear and my pipe has to be precise in length and cut (or faced) cleanly & squarely, but this is doable and I’ll share the results here…
The reason my SW1 consistently reads ~5 points lower than the RDC I’ve typically relied on (a Head 3-in-1) may be that it’s not the SW1 that’s misreading swingweight but the RDC. I’m not talking about a miscalibrated RDC. I’m saying that it’s possible the SW1 is inherently more accurate than an RDC because it’s more linearly accurate.
One known problem with RDCs, as with weight scales, is ‘measurement creep.’ This is a phenomenon where the farther a frame’s swingweight is from the rod’s calibration MoI, the less accurate the SW measurement is likely to be. @bfroxen goes into this in his post # 662, and it’s worth digesting the whole thing. But the upshot is that his research indicates that older Dunlop and Babolat RDCs lack the linear accuracy of the SW1. I, for one, believe it.
Still, the reason I hold with the Head 3-in-1 readings I’ve received at Racket Doctor in L.A is that over a dozen of the unstrung frames measured there had already been measured by Tennis Warehouse’s Matching Service, and all the readings conform to within 1sw point of each other (Racket Doc’s vs. TW’s), often less than a point. Could both RDCs be equally off in their swingweight measurements? Absolutely.
At this point, I’m thinking the best way to confirm the accuracy of the MoI labeled on my collapsible rod is to use the Rod Swingweight Calculator built into bfroxen’s blog on Swingweight Machine Accuracy to fashion a verification rod out of PVC piping. The directions are clear and my pipe has to be precise in length and cut (or faced) cleanly & squarely, but this is doable and I’ll share the results here…
Last edited:
ichaseballs
Hall of Fame
The SW1 is great! Hard to complain when there is no other product that competes at a similar price range.
So the Prince/Dunlop machine costs over $1000, but can't do decimals?? maximum variance up to 0.9?
So the Prince/Dunlop machine costs over $1000, but can't do decimals?? maximum variance up to 0.9?
- If racquet 1 is 332.5 it will read as 333?
- If racquet 2 is 333.4 it will read as 333 as well?
ichaseballs
Hall of Fame
even if the SW1 is not calibrated exactly, for matching purposes it will work fine.
if your goal is to simply match racquet B to racquet A.
if your goal is to simply match racquet B to racquet A.
Brando
Professional
Mystery solved, and it turns out that I didn’t follow @bfroxen’s directions exactly enough...
This updated Calibrating the SW1 video displays how the expandable rod should be expanded and locked while held vertically. Well, I can attest to that. I’ve been expanding the rod while loaded in the cradle, and so, it wasn't locking down at all its joints. This horizontal expansion technique was what caused my readings to be consistently 5-7sw light.
For those curious, last weekend I indeed fabricated my own calibration rod out of PVC piping at Home Depot via Briffidi’s handy dandy Rod Swingweight Calculator in bfroxen’s blog on Swingweight Machine Accuracy (located at the bottom). That this calibration pipe read true while a replacement expansion rod @bfroxen had sent me (last week) read just as light as the original was a big hint. I mean, it’s possible @bfroxen could’ve misread or mislabeled a rod. But two of them and both sent to me? No.
Something I was doing, or not doing, had to be the cause. So I revisited the calibration directions on the Briffidi site to discover my mistake. The rest is history and my thanks to @bfroxen for providing such good customer support.
Now this has me wondering if this might be the source of @KBlade’s issue too?
This updated Calibrating the SW1 video displays how the expandable rod should be expanded and locked while held vertically. Well, I can attest to that. I’ve been expanding the rod while loaded in the cradle, and so, it wasn't locking down at all its joints. This horizontal expansion technique was what caused my readings to be consistently 5-7sw light.
For those curious, last weekend I indeed fabricated my own calibration rod out of PVC piping at Home Depot via Briffidi’s handy dandy Rod Swingweight Calculator in bfroxen’s blog on Swingweight Machine Accuracy (located at the bottom). That this calibration pipe read true while a replacement expansion rod @bfroxen had sent me (last week) read just as light as the original was a big hint. I mean, it’s possible @bfroxen could’ve misread or mislabeled a rod. But two of them and both sent to me? No.
Something I was doing, or not doing, had to be the cause. So I revisited the calibration directions on the Briffidi site to discover my mistake. The rest is history and my thanks to @bfroxen for providing such good customer support.
Now this has me wondering if this might be the source of @KBlade’s issue too?
Last edited:
bfroxen
Semi-Pro
Kudos to @Brando for actually creating a verification rod and checking the results.
To clarify, I recommend extending the rod and locking the sections while hanging it vertically. This allows gravity to ensure all sections are fully extended and unwanted friction is minimized while locking.
In at least the original extendable rod calibration video, I showed it being locked while mounted in the SW1, but I described that the end must be supported during locking. If the end is unsupported while the rod sections are locked in the horizontal orientation, it can cause a bit (less than 0.5 kg·cm²) of variability in the MoI of the rod. The bigger problem with not supporting the end is that there is considerably more friction in the joints, and the user might not get all four of the joints locked. If the joints are not locked, the "floppy" rod will swing more slowly and significantly (several kg·cm²) affect the calibration.
To clarify, I recommend extending the rod and locking the sections while hanging it vertically. This allows gravity to ensure all sections are fully extended and unwanted friction is minimized while locking.
In at least the original extendable rod calibration video, I showed it being locked while mounted in the SW1, but I described that the end must be supported during locking. If the end is unsupported while the rod sections are locked in the horizontal orientation, it can cause a bit (less than 0.5 kg·cm²) of variability in the MoI of the rod. The bigger problem with not supporting the end is that there is considerably more friction in the joints, and the user might not get all four of the joints locked. If the joints are not locked, the "floppy" rod will swing more slowly and significantly (several kg·cm²) affect the calibration.
Irvin
Talk Tennis Guru
I have one of the early SW1s and was sent a replacement calibration rod. Then I started having issues with readings being off. Long story short I had to manually change the MOI values for my new rod.
I believe the SW1 app reads the NFC chip in the SW device to initiate the MOI setting. If for any reason those MOIs are changed it will raise or lower your readings.
A way you can verify it’s close is to first verify the MOIs on the calibrate page are correct then make a measurement with nothing in the device. It should read close to 0 kgcm2 after calibration.
EDIT: If you ever want to maladjust the SW1 to get similar readings to another SW machine (not sure why you’d want to) you could use the MOI settings to do that as you have found out. You’re basically changing the expected value of the calibration rod.
Last edited:
Brando
Professional
G'ah! A perfect parable, @esgee48. I shoulda' known better indeed. Funny thing? I did read the manual and I thought I was following ALL its directions; but no, I wasn't. This is why police detectives, when they run out of clues, are taught to go back to whatever evidence they have and run through it again; and again, if necessary. Why? Because we're human and we always miss something.
Last edited:
No, I don’t have this issue. My SW1 is one of the older units that used the single piece PVC calibration rod, not one of the extendable ones.
McLovin
Legend
Yeah, I asked that question years ago. After learning that was true, I’ve never bothered calibrating as all I ever do is match one frame to another.
Brando
Professional
Darn. I was hoping that solving my issue would solve yours. (Am kinda bummed.)
Irvin
Talk Tennis Guru
Sounds like either your calibration rod, or the PTC’s rod (maybe both) is out of spec. Something you can do quickly is take a measurement with nothing in the SW1 but your phone of course. You should read Les than 1 kgcm2. If the value is 6-7 points the value marked on you rod is incorrect.
Last edited:
TennisCarsPlanes
New User
I just got me one SW1 + TW add on. Got it for the TW part rather than the SW as I have 3-in-1 and upgrading to the 4-in-1 was just way too expensive.
What are your thoughts on the TW measurement accuracy? Thanks!
What are your thoughts on the TW measurement accuracy? Thanks!
Brando
Professional
My TW attachment, @TennisCarsPlanes, operates perfectly, yielding tighter groups than the SW operation because the racquet's weight is centered over the cradle's pivot point. I've also compared the readings against an RDC's, and so far, they're identical.
TennisCarsPlanes
New User
Thank you for the feedback. I am looking forward to trying it. I do feel that it is one of the specs generally ignored during customization and I have heard too many times: "it is the same frame, same string setup, identical static weight, balance and SW but they play differently..."
Brando
Professional
My prediction? You'll be surprised how many of your assumptions about twistweights get busted. You could even get a couple "Ah-hah!"moments. I, for instance, had no idea the twistweights of my favorite frames were so low, all 0.5-0.6tw below the 13.6tw Tennis Warehouse published. All to say, this is a journey of discovery, and it'll be fun.
TennisCarsPlanes
New User
Thanks. Historically (over the last 45+ years) I have not paid attention to the TW at all. I use non-retail rackets, so, the posted retail specs are irrelevant, however, I have noticed that my frames in rotation, although set up "identically" (same frame, length, three major specs (static weight, balance and SW) all within 1g/1mm/1kg*cm2, same string setup) do feel slightly off from each other, so, that would be the parameter that I omitted.
Brando
Professional
Cheers, @TennisCarsPlanes. I personally have @Irvin to thank for beating the twistweight drum, and I'm glad I heeded the call. Do let us know what you find out; were the diffs among these 'identical' frames TW after all? I, for one, am curious...
Irvin
Talk Tennis Guru
I think @bfroxen did an excellent job with the Briffidi products. They solved a real problem: bringing measurements that were once largely confined to pro shops and labs into something affordable, portable, and repeatable. The combination of sound mechanical design with electronic timing makes them much less susceptible to user error than many homemade swingweight setups.
Before the SW1, I was calculating SW with a homemade jig, a video camera for timing, a lot of time, a wing, and a prayer. I got pretty close with my primitive setup but it took a lot of time. For TW I never got it down right and was happy to see the TW adapter added to the SW1.
Before the SW1, I was calculating SW with a homemade jig, a video camera for timing, a lot of time, a wing, and a prayer. I got pretty close with my primitive setup but it took a lot of time. For TW I never got it down right and was happy to see the TW adapter added to the SW1.
Last edited:
