Small Differences Can Matter

Small Differences Can Matter

Recently, I was on a familiar road, but riding it did not feel familiar. It seemed like I was straining to stay on top of my gear, whereas usually I just spin along. Was I exceptionally tired? I didn’t feel that way…
Then I remembered that I had replaced my 46-30 chainrings with 48-32 for Paris-Brest-Paris. In PBP, I sometimes ride in big groups and with a strong tailwind… Then my tallest gear of 46-14 – big enough for my riding in the Pacific Northwest – might not be quite tall enough.
So my big ring was 2 teeth larger than usual, hence the difficulty to stay on top of the gear. The difference is about 5% – small, but noticeable. Now that I have returned from PBP, I will re-install the 46-30 rings, since they suit my usual riding better.
It’s too bad that customizing your chainrings isn’t as easy as it used to be. Today, most makers only offer only very few chainring sizes, and none are small enough for non-racers. I have not yet been dropped because I spun out and could not keep up… and yet my biggest gear of 48-14 is 20% smaller than 50-12, the smallest maximum gear you can get from mainstream makers today.
If you have a healthy spin, you’ll rarely, if ever, use the 2 or 3 smallest cogs on a modern drivetrain. And that makes your 10-speed cassette effectively a 7-speed. With smaller chainrings, you could get a closer-ratio cassette, and have smaller steps between your gears, while maintaining the gear range that you currently use. Or you can keep your current cassette, and trade the super-large gears you never use for extra-small gears that will come in handy in the mountains.
With that in mind, we are offering dozens of chainring combinations for our René Herse cranks, from 52 to 42 teeth for the big ring, and down to 24 teeth for the small one. That way, you can equip your bike with gears that you’ll actually use!
Click here for more information about the René Herse cranks.

Share this post

Comments (95)

  • Bob

    But sadly, only one crank length.

    October 27, 2015 at 6:05 am
    • Jan Heine, Editor, Bicycle Quarterly

      What length do you need? We are thinking about adding more. Unlike most small makers, who use one forging for all lengths, we will have to make new forgings, so our cranks can pass the stringent EN “Racing Bike” standards. So it’s a costly decision, but one we are prepared to make at some point.

      October 27, 2015 at 6:46 am
      • shiggy

        I have 180mm on my road bikes. Anything shorter feels like walking with my shoelaces tied together. Using White Industries because of the ring options (running 44/30) and it has proper chainline (square taper). Also have an ’80s Specialized touring crank with 46/34, just a bit too big for most rides

        October 27, 2015 at 7:14 am
      • Greg

        175 mm. If you look at old Stronglight, etc. cranks, they tended to offer 165, 170, and 175 in most cases. Don’t bother with the in-betweens, imo. However, some folks get all excited over 172.5 mm cranks, so YMMV.

        October 27, 2015 at 7:28 am
        • Bob

          Coming from the cold, hard background of mechanical engineering, the femur, shin, ankle, foot, crank is one of the most complex levers you’re likely to encounter. In my case it was suggested longer cranks would alleviate some hip pain. I tried them (200mm) and it did. At first they felt like these huge windmill things but that quickly went away. They’re spinnable but not much above 90rpm, but I’ve ended up climbing better, with a lower HR, slightly higher wattage, and just feeling a bit more powerful.
          Once I found a CF source for these cranks, it was time to switch all bikes over and I haven’t looked back.

          October 27, 2015 at 8:17 am
      • Fred Blasdel

        Given that the existing model is the traditional 6.75″ length, why not 7″ for the other size? You shouldn’t just acquiesce to modern sizing after you’ve started, keep it novel for everyone 🙂

        October 27, 2015 at 11:49 am
      • nickskaggs

        I put Sugino Mighty Tours in 160mm on my bike and it has made my bike much more comfortable. My hip flexors don’t bother me now like they did with longer cranks. I’m on the opposide side of the height spectrum, at 5’3″. (160cm.)
        My only complaint is that with a 110BCD they’re limited to a 34T chainring.

        October 28, 2015 at 12:51 pm
      • Greg

        Fred, that is 178 mm. Pretty long. OK with me, but I ride 175 plus-or-minus 2.5 mm. I sell a lot of cranksets, and many are 170, then 175, then 172.5, in that order, then about three per year are 165. About once every other year I sell a 177.5 or a 180. Super-rare to sell those super-long ones.

        October 28, 2015 at 1:47 pm
      • Greg

        170, 172.5, and 175 combined cover perhaps 93% of all riders, male and female. If 2.5 mm of length can’t be detected (and I might give you that), then if you were only going to make one length, it should have been 172.5 imho. Now that you have 170 in production (yes, it’s really 171, but that’s statistically the same number, really) I would (again) strongly recommend 175 as the next step. 165 would be great, at some point in the distant future, but it will take a long time to recoup the forging costs for that particular length. End of semi-rant. 😉

        October 29, 2015 at 8:56 am
        • Jan Heine, Editor, Bicycle Quarterly

          yes, it’s really 171, but that’s statistically the same number, really

          I agree. So is 175. Four millimeters amounts to just 2.28 %. Imagine if we offered stems in 2 mm increments. You’d call us crazy.
          The requests for 180 or 200 mm cranks make a lot more sense. But as you point out, the cranks we offer already cover 93% of riders. As a small company, we must think long and hard before offering components that work for only 7% of the riding population.

          October 29, 2015 at 5:07 pm
      • Bob

        On a side note- how about getting those seat-tube mounted lever operated rd’s in production? SO KILLER! Also- I’d love to put standard-casing RTP’s on my (working) pedicab. You think they’d last? Honestly I’d like to roll them because I think they look cool and I dig Compass products (local, understated style, not mainstream etc etc).

        October 29, 2015 at 11:43 am
        • Jan Heine, Editor, Bicycle Quarterly

          I don’t know anything about tire use on three-wheelers. If your current tires last well, I’d give the Rat Trap Pass a try. And I’d charge extra for the smooth ride…

          October 29, 2015 at 5:09 pm
    • Bob

      Well I’m an outlier because at 6’6″ I ride 190-200mm, using femur measurement as a guide.
      But even 180’s would be nice, and possibly longer, like TA used to offer. I know these would come at a premium, but with the gearing options you’re offering, it would seem to cover the full range of riders.

      October 27, 2015 at 7:46 am
      • Jan Heine, Editor, Bicycle Quarterly

        Thanks for your thoughts, Bob. It certainly seems, from past comments by others on this blog, that you are not the only tall rider who has found longer cranks to be an improvement. Lennard Zinn has written a lot about this as well and is an advocate for the approach you describe.
        -Theo Roffe, Compass Bicycles

        October 27, 2015 at 11:01 am
    • Greg

      No, 171s do not cover 93% of riders, but 170, plus 172.5, plus 175 do. The rest are outliers. I’d put 180s at maybe 0.5% and 200s at 0.0000002%.
      Jan, go ahead and make a bunch of 200s (or even 180s) and see how many sell per year. Good luck with that, good sir! 😉
      175 is most definitely NOT the same as 170. Stems are often made every five mm, why not crank arms? You are being too dogmatic about this sbject, imho. Make what will sell!

      November 2, 2015 at 3:46 pm
  • Jeff Potter

    The racers, go-fasts, and club paceliners are out there, probably in fair numbers, so the bigger rings do serve a need, I’d think. I accidentally left my 50t CX ring on my CX bike that I wanted to test as a roadbike this summer. I frequently spun it out on club rides and got into trouble until I realized what was happening. I was relieved to install the 52. When I get bold enough to ride with the fastest bunch I appreciate a 53.
    I haven’t finished my comparison of tire widths for fast group road-riding but I’m not sure I can avoid being dropped on 32’s on pavement, even if they’re Stampede EL’s. I *did* get dropped the one time I tried them. I want to try them again, though, since we have so much chipseal. I was thinking their anti-vibration aspect might help. I also did a gravel road race-ride with them — and was able to contend. I’d say they were a benefit then, but some riders were using mtbikes so there wasn’t a “technology gang-up” effect. I certainly prefer them over all other tires for my solo non-race riding!
    Fast club riding, racing — city limits sprints — might have different needs than rando type riding. I’d like to see “BQ” explore that style, too.

    October 27, 2015 at 6:24 am
    • Jan Heine, Editor, Bicycle Quarterly

      I agree on the gearing. When I raced, I needed a 53-12 for some downhill race finishes. As a climber, I had a hard enough time staying on the wheel of racers who were strong and heavy. Of course, when things went according to plan, the strong and heavy guys were dropped on the uphills, making my life a bit easier…
      Regarding the tires, I think you need more data. All our testing indicates that wider tires roll no slower, and often faster, than narrow ones.

      October 27, 2015 at 6:49 am
    • Jan Heine, Editor, Bicycle Quarterly

      Of course, even then, I probably was trying to make up for lack of power with bigger gears… After all, without the strength and leadout that a Cipollini enjoyed, why would I need gears as big as his.
      Now, even when riding with racer-types, I find that I don’t get dropped even though my gears are much smaller than theirs.

      October 27, 2015 at 7:58 am
      • Jeff Lyall

        Exactly. Why would a club racer need the same gearing at Julian Dean or Cippo, I’ve always wondered that. And true, while I often suck at long fast downhills, thats not the gearing, its my inabilty to push at speed, a skill that can be picked up with a bit of training in that area. I am currently looking at ideal gearing options for MTBs, and right now I can say 1x and 2x are not there yet for usable bikes that can be ridden and raced in the same format.

        October 27, 2015 at 2:12 pm
        • Jan Heine, Editor, Bicycle Quarterly

          For most downhills, it is much faster and more efficient to get in the aero tuck. You reduce your wind resistance by 33% – at speeds above 40 mph, it’s impossible to make up for that with more power. It’s fun to coast by riders pedaling furiously on the downhills.
          Of course, for sprinters, that doesn’t work. They need to go close to 50 mph on the flats, and you can’t get in the aero tuck for that! Fortunately, for sprinter-types (and tandems and gear mashers), we offer 52-tooth chainrings.

          October 27, 2015 at 4:58 pm
      • Jeff Lyall

        Yes, I must be talking about less steep hills then, because there is no way anyone is hanging on while coasting in a tuck in this particular race, and I am particularly aero, given my wattage to speed ratio in a typical TT.

        October 28, 2015 at 12:39 pm
  • Ian Whatmough

    I would be interested to know what crank arm length you prefer on with your 46-30 setup and do you think a longer crank arm with the larger front rings would reduce the fatigue your were experiencing?

    October 27, 2015 at 6:49 am
    • Jan Heine, Editor, Bicycle Quarterly

      I have found that crank length doesn’t matter to me. When I first rode Mark’s bike, which is very similar to mine, except his cranks were (then) 5 mm longer, I would have sworn he was on 170 mm cranks, too. I was surprised to learn that his were 175s.
      Last weekend, I rode a bike with 180 mm cranks. This time, I noticed many differences, for example, it immediately was obvious that the tubing was stiffer and the bike didn’t plane like mine. I never noticed the crank length difference to my (now) 171 mm.
      It appears that generally, differences of less than 5% are too small to be important. For example, 20 mm wider handlebars are noticeable, but 10 mm are not. The same with chainrings – 2 teeth is noticeable, 1 tooth not so much. Translated to cranks, it would mean that a difference of 8.5 mm should be noticeable. So I should have noticed the 180 mm cranks, but perhaps too many other things about the bike were different and masked the different crank length.

      October 27, 2015 at 6:26 pm
  • David Morgan

    After buying the superb 6800 groupset (mine having 34×50; 11-32-wish a 13×32 would be offered) and loving it on my high performance road bike; I bought another 6800 group for my commuter, but with the 36×46 Cyclocross crank option with the CX-70 brakes. Wow, it is so wonderful to see that wide, top end rims, wide, top quality tubeless tires, Ultegra quality 36×46 cranks, and “Gran Fondo” /Enduro/Gravel geometry frames are now mainstream and widely available.

    October 27, 2015 at 7:20 am
  • David Morgan

    Oh, re cranklength- it is anti climactic to some- but do the most stringent testing you can do (Masters candidates in Sports Physiology are the best folks to find/afford for us non-pros)- and you may find, despite your deepest feelings, hottest opinions, and most passionate beliefs, crank length (this is from someone who swore I MUST have a 170 for touring/172.5mm for road/175mm MTB and did testing to validate that belief) doesn’t affect performance. I don’t want to argue, just to encourage testing in the area.

    October 27, 2015 at 7:32 am
    • Fred Blasdel

      You’re right that crank length does not effect power once you’re used to it, even at the extremes. Most of my bikes have 180mm cranks, but I also have 172.5, 190, 200mm and even 108mm on a cargo bike!
      Crank length does however have a dramatic impact on bike fit, especially on where your saddle wants to be. That alters your hip angle, which changes the balance of your torso, and ultimately how comfortable it is to get lower on the bike.

      October 27, 2015 at 11:59 am
      • B123

        Indeed, the cranks affect the fit. I felt my knee pains eased when I switched to shorter cranks, spinning feels more natural and even climbing standing up feels more natural and less fatiquing. And in addition to being more comfortable when my legs don’t have to go into such a tight angle on the top of the stroke, it’s easier to get low on the drops when my knees are further from my chest. My inseam is on the short side at about 80cm and now use 165mm cranks.

        October 28, 2015 at 3:22 am
        • Bob

          That’s really cool to hear because I had the exact same outcome switching to longer cranks (200mm). Key word in your statement being “felt”. Can’t be quantified, but it counts.

          October 28, 2015 at 3:41 am
      • Bill Gobie

        As a taller rider I feel more comfortable on 180 mm cranks. I feel I use my muscles’ range of motion more effectively.
        If Compass were to manufacture another crank length, I suggest something significantly shorter like 160 or 155 mm. Tall riders can get along on 170s. Shorter riders like B123 need shorter cranks to avoid injury and improve their bike fit.

        October 30, 2015 at 6:26 pm
    • B. Carfree

      I remember reading something from Lennard Zinn last year in which he insists that longer cranks give him much better performance on climbs. He didn’t really do very rigorous testing. In fact, it was barely above anecdotes with power readings. However, his results were consistently obtained. If Jan and company came to the same conclusion, that would clinch things. Then again, if Jan did the testing it would stand on its own anyway.

      October 27, 2015 at 9:09 pm
      • Jan Heine, Editor, Bicycle Quarterly

        As I recall Zinn’s experiments, testers had the best performance both with the longest (200 mm?) and shortest (150 mm?) cranks, but the “standard” 170 mm cranks were worst. This seemed hard to explain, and perhaps is the reason why the testing wasn’t pursued further…

        October 27, 2015 at 9:21 pm
        • Jan Heine, Editor, Bicycle Quarterly

          One major issue that complicates any testing is that the tester might perform better after they adjust to the different-length crank. However, during the adjustment time, the rider’s fitness might also change… So you’d give the rider 200 mm cranks, have them ride the longer cranks for three months, and then compare their max. power or climbing speed or whatever to the results from 3 months earlier on their previous cranks. Of course, now you don’t know whether any change is due to the cranks, or the hill intervals they’ve been doing (or it being the off-season).
          The solution to this is a large sample size, both with different riders and by going back and forth between cranks multiple times for the same riders. For starters, I’d suggest finding 10 riders, each riding long, medium, short, long, short, medium, short, long (or whatever random order you come up with) for three months at a time. After 2 years, you do the statistical analysis. You might need more cycles, since annual form tends to vary, so ideally, you’d compare only early-season with early-season, mid-season with mid-season, etc. So perhaps 10 years is more realistic.
          Of course, the longer and shorter cranks also might need different bike fit, so you are introducing another dimension to the problem. Test each length with three different fit configurations (more/same/less setback of the saddle). It’s a major investment of time and effort, and at the end, people will complain that it wasn’t a double-blind test…

          October 27, 2015 at 9:31 pm
          • Bob

            Which is why I have trouble with any of the published tests published in BQ, an otherwise above-excellent publication, be it tire testing, or whatever, that involve human beings. Subjective opinions really matter, but when it comes to numbers…… Even the slightest variable due to a human tester-almost unavoidable-negates the constants in tests. Labs don’t use machines to save on labor costs, but to control consistency in testing. Then it’s totally possible to say, “…. lab testing indicates X but despite that, we found this (tire, for example) to be more comfortable/faster accelerating feeling/responsive etc.

            October 28, 2015 at 3:11 am
          • Jan Heine, Editor, Bicycle Quarterly

            The big problem with machines is that they don’t replicate humans well. The rider is a major part of the bike. For example, if you test tires without a rider, you cannot measure the suspension losses that occur in the rider’s body. Of course, as you point out, humans introduce additional noise to the testing.
            Fortunately, science has found a way around this problem. You do a statistical analysis to check whether what you see is “slight variables due to the human tester” or actual differences in tires or whatever. The more repeat runs you do, the more the “slight variables due to the human tester” average out, and the finer differences you can discern. It’s a rigorous process, and it takes all “arm-waving” out of the study.
            Some measured differences are too small to be statistically significant (for example, the differences in wind resistance between 25 and 31 mm tires). This doesn’t mean they aren’t real, just that we don’t know. (And for practical purposes, they are too small to matter.)
            All Bicycle Quarterly tests include a rigorous statistical analysis. We are lucky that one of our editorial team, Mark Vande Kamp, has a Ph.D. with a minor in statistics, and works in the field, so he is uniquely qualified. This means that your concern, valid as it is, has been addressed.
            This statistical analysis has to be done with all testing, whether it involves a human or not. Other things can change, too, like the air temperature in a wind tunnel test. Unfortunately, few bicycle-related studies are rigorous enough to include a statistical analysis of the results. Instead, measurements are reported at “face value” without considering whether they may just represent “noise” in the data.

            October 28, 2015 at 3:23 am
          • Bob

            You’re correct of course, and I was not great at statistics, but every if I ever turned in a paper with the human variable being such a large one, as I did (once was enough), it would never have been even considered. The human variable is considered well after other variables are minimized.
            (Sorry, not trying to throw water on a fun discussion).

            October 28, 2015 at 3:38 am
          • Jan Heine, Editor, Bicycle Quarterly

            Again, all I can say that if your statistics are good, and your results are statistically significant, there is no reason your paper wouldn’t be accepted for publication. Human subjects are standard practice in medical research…
            One key thing is to limit the scope of your study to something that you can hope to achieve. For example, testing different tires for performance is not so difficult. Our wind tunnel tests have shown that our testers can replicate the same position reliably, so the error introduced is small. Then you just have to make sure temperature is constant (and correct for the inevitable small fluctuation you see), that there is no wind, and you can test. Do your statistics to check that you have real results, and you are done. It almost sounds easy until you actually do it!

            October 28, 2015 at 4:05 am
  • Paul Hoffmann

    Jan: I recently reread your account of “Raid Pyrénéen 2.0” ( vol. 12:2) in which you mention lowering the gearing from 48-32 chainrings (your normal?) to 46-30 to better cope with the long climbs. Had your preference for gearing changed after that ride?

    October 27, 2015 at 7:59 am
    • Jan Heine, Editor, Bicycle Quarterly

      Yes, I just kept the 46×30 on my bike… Only for this year’s PBP did I switch back to 48×32, and I actually needed the 48-14 that is the largest gear!

      October 27, 2015 at 6:28 pm
  • teamdarb

    Each year since 2011 I’ve ridden the east coast tip to tip. Technically, for times a year. This final up and down for the year I decided finally to try my 27″ wheeled mixte converted to 26″ with 1.75 tire. The true width was more 47mm, cheap Bontrager tire. I found the gearing lacking all the way around. Then I remembered other bikes I’ve used had much smaller chain rings, but larger tire roll out. So I began first playing with some spare new cassettes I’d brought along to put on a bike at the other end of the voyage. In the end I was comfortable even through Spotsylvania country Virginia using 52/42 14,15,16,18,24,28,32. I could push the 42,14 and keep up with the riders in the modern bikes quite well. I’m not much of a spin guy, but at 5’4″- 134lb a surprising masher loaded or not. People asked numerous times how I was keeping up. I tried explaining gearing in inches. That the taller gear was allowing me to match their overall rollout compared to the initial difference of 26 versus 700c. Before someone states the obvious, I just happened to have a bag of cassette cogs with me- yes, swapping the chain rings would be the greatest change possible
    I’ll be glad when I find a bike that is tough, elegant and fits without a sloping top tube.

    October 27, 2015 at 8:20 am
  • Luis Bernhardt

    I think crank length is highly overrated. When I was racing (up to 2008), I used 175mm cranks and could win or place in masters road sprints. I rode a lot of track and used 165mm exclusively. I could detect no difference (and I’ve been riding well over 40 years). I’ve been riding 165’s on my fixie on the road ever since I stopped racing, and I’ve been up Haleakala, Mont Ventoux (4x in one day), 2 PBP’s, etc. all on a 70″ gear and 165’s. When I raced, my preferred sprint gear for a flat finish was 53×14. I have no idea how guys can sprint in 53×12 unless it’s downhill or they’re pro’s. I regularly limited myself to 53×15 on the Tuesday night training races (it’s supposed to be training). My new geared road bike will use 170mm cranks and a 9-spd cassette (9spd chain lasts longer than 10- or 11-spd, and 9spd has the most compatible spacing next to 11spd). Yeah, crank lengths, highly overrated. Another 2.5 or 5mm will not make up for any deficiencies, despite what people say about “marginal gains.” Just use what you’ve got!

    October 27, 2015 at 8:51 am
    • Bob

      I ride 200’s on a fixie winter bike and have some time to train on pancake flat roads in FL (122 ft gain over 50 miles), previously I ran 180’s. The longer cranks really bring more muscles (as in different leg muscles), but I’m able to ride a 53 x 15 and keep a comfy cadence. HR rarely exceeds 122 bpm. Don’t have wattage on the fixie. On the 180’s my legs were flying, so it does, at least in one example, make a difference. On shorter cranks I too get that exact same “shoelaces tied together” feeling.

      October 27, 2015 at 7:12 pm
      • Jan Heine, Editor, Bicycle Quarterly

        I can see that 20 mm make a difference, but some people are talking about 1.5 or 4 mm (172.5 or 175 mm compared to the René Herse’s 171 mm).
        53×15 seems like a huge gear to me… unless you are very fast. A quick calculation shows that with 700C wheels at 90 rpm, you’ll be going 25 mph (40 km/h). What is a “comfy cadence”?

        October 27, 2015 at 7:22 pm
        • Bob

          Ha ha, fast is something I’m not, at least not when things get vertical. My cadence on the Dixie is comfy enough to move along at 20-21 mph, on avg. Again it’s table top flat, and the major variable is wind. However strong I may be, it’s all relative at 6’6″, 220. Plus, I’ve never let pursuit of speed, or performance get in the way of having fun, or enjoying the scenery.

          October 28, 2015 at 3:24 am
    • Greg

      You need 165 on the track so you don’t suffer crash form a pedal strike… Apples to oranges….

      October 29, 2015 at 9:00 am
  • Eric Gierke

    I am glad to hear Jan runs a 46-30, as I have a TA 46-30 (175 length) on one of my favorite steel bikes, and I run it with a 9-speed 11-34 cassette. I almost never get down to the 11, but I usually use the middle 4 or 5 cogs on the cassette, which keeps my chain fairly straight. I guess I haven’t worried too much about having slightly bigger jumps than I would have with, say, a 13-26 cassette. Having big cogs of 30 and 34 is a nice insurance policy — I don’t worry about getting up any hill at the end of a long day despite my aging (50-something) legs. Like Jan, I am in Seattle, and we can get some very steep 1-3 block hills sometimes.

    October 27, 2015 at 8:51 am
  • phr3dly

    Shimano sells Ultegra 10-speed and 11-speed cassettes with 13-tooth, 14-tooth, and even 16-tooth small cogs.
    See, for example:

    October 27, 2015 at 8:57 am
    • DG

      Thanks for that link. If you have a Shimano 7801 hub, you can only use cassettes w/the deeper splines. As a result, the lowest gearing in stock form is the Ultegra Cassette 10-speed CS-6700 12-30. I am wondering if any of you have found a way to get 13 and 14 tooth cogs as the smallest cogs. This would allow me to add back the 19 cog that I like. The lower priced Shimano cassettes have the larger cogs riveted together. This modification would not have the shift ramps as Shimano designed, but I don’t think the shifting would end up being terrible. My 7800 crank is the standard road one so I don’t have the option for smaller chain rings.

      October 28, 2015 at 7:03 am
      • Bob

        I especially like the 16 x 27. With a little fooling around,a la Harris Cyclery, I bet that could become a 16 x 30 or 34.

        October 28, 2015 at 7:52 am
  • Mike Terrell

    I currently run a 48/34 that I got through Velo Orange. As I have gotten stronger as a rider, I have found myself using the big ring far more than the small ring. But, I also have found myself rarely using the bottom 1/3rd of the cassette when in the 48 tooth ring. My plan, come spring, is to switch to a 46/34 or 44/34 or 44/32. Based on the riding I do, the small ring will really only be used for more sustained or steep climbs. With a smaller big ring, I think I will use more of the cassette giving me more usable gears and fewer front shifts.

    October 27, 2015 at 9:40 am
    • Jan Heine, Editor, Bicycle Quarterly

      I (Theo) use my compact double cranks (44/28) the same way: little ring is for steep climbs and the big ring is used with the full cassette for almost all other conditions. When I picked this gearing it was based on Jan’s suggestions on this blog from November 3, 2012:

      In addition to covering the range from low to high gear, a good gear selection will do the following:

      • Put the base gear in the middle of the rear freewheel/cassette, so that I can adjust to changes in speed and terrain with a simple shift or two in the rear.
      • rovide small enough steps between gears, so that I can continue pedaling seamlessly.

      – Theo Roffe, Compass Bicycles

      October 27, 2015 at 10:34 am
  • zigak

    ratio*cadence*circumference*0.06 = 46/14*110*2.0*60/1000 = 43.4 km/h

    October 27, 2015 at 12:15 pm
  • TimJ

    Every time I look at the small bolt diameter on the Compass or Rene Herse cranks I wonder about the lateral stability of the chainrings. I’m pretty sure this issue was addressed in some issue of (V)BQ but I can’t for the life of me find it. Any ideas anyone?

    October 27, 2015 at 1:48 pm
    • Jan Heine, Editor, Bicycle Quarterly

      Chainrings don’t undergo significant lateral loads. It’s that simple. Even at an extreme cross-chaining angle, the chain is effectively pulling backwards. The Herse chainrings are a millimeter thicker than most chainrings, so they are quite stiff. We addressed this here.

      October 27, 2015 at 4:56 pm
  • Conrad

    Racing on the velodrome really changed the way I think about gearing. Just about nobody will ride a taller gear than 100 inches, even in a Keirin. Even the pros. Those races top out at 40+mph. You learn really quickly that a tall gear is useless if you can’t spin it. Anything over 100 gear inches is basically for overdrive, or trying to sprint downhill in a road race. I am a little baffled that most road bikes (that will never even see a race) go out the door with a 53X12. I guess now people are talking about a new 148mm rear wheel width standard because making room for all those cogs (that we never needed) was creating weak, asymmetric wheels. Could it be that the best way to do it was with fewer than 9 cogs? The horror!

    October 27, 2015 at 9:59 pm
    • Bob

      I’ve always wondered about running a straight block, and using the chainrings to get minimal gear overlap.

      October 28, 2015 at 3:27 am
      • Jan Heine, Editor, Bicycle Quarterly

        I tried that, setting up my Rivendell in 1999 with a 54-44-34 triple and a 13-18 straight block. Great in theory – the smallest possible steps between gears, 18 speeds, of which only two were duplicates. Basically, just one cog overlap between each adjacent chainring pair.
        In practice, it didn’t work well at all. I found myself in the 44-18 most of the time, and every time I needed a smaller gear, I had to shift to the small ring and across five gears on the freewheel at the same time, only to repeat the process in reverse when I needed a bigger gear again. With the small steps between gears, it meant constant shifting. It was hard to get into a rhythm.
        It taught me that overlap is a good thing – you want to be in the middle of your rear cluster during “normal” riding, so you can adjust to changing conditions without front shifts. (The exception is half-step, where front shifts are used to fine-tune your gear.)
        These days, I use a “single ring plus granny” approach, if you like. Most of my riding is on the “big” ring (which isn’t all that big), with the gears I use most in the middle of the cluster. When it gets really steep, I switch to the small ring, again to the middle of the freewheel. So in each case, I can adjust my gearing significantly without requiring a front shift.

        October 28, 2015 at 3:36 am
        • Bob

          I think my id a was even worse than what experienced. Going across (at the time) 7 gears on the small ring, and continuing that by starting over on the big ring. OK until the ratios you need happen to fall around the small/small, big/big area!

          October 28, 2015 at 3:46 am
      • Steve Palincsar

        And what happens when you cross over? Shift to the big ring in front and then shift down 7 or 8 times in back? Same thing in reverse going from the big to the small ring: front shift plus 7 or 8 upshifts in back – either way, the transition will be monstrous.

        October 28, 2015 at 3:48 pm
        • Bob

          Yes, practical considerations never entered into it.

          October 28, 2015 at 5:21 pm
  • cbwesq

    I find it very odd that Jan says that small differences in gear ratios can be significant, while at the same time saying that similar differences in weight are mere “noise” that gets lost in all the other factors. I wonder if it might have something to do with the fact that his company sells chain rings that allow for small adjustments in gear ratios but one of the biggest detriments of the tires and other gear he sells is that they weigh more… My own experience, unimpeded by financial interest, is that small differences are significant under certain circumstances (e.g. fast group rides), whether it is gear ratios or the extra weight of Jan’s beloved “supple” 32-42 mm tires.

    October 28, 2015 at 6:59 am
    • Jan Heine, Editor, Bicycle Quarterly

      I don’t think we ever said that small differences in weight are mere “noise”. When all other factors are equal, the lighter bike will be faster. It’s just that other factors tend to be more important, especially since you need to look at the weight of bike and rider combined. For me, 5% weight difference would amount to 4 kg (9 lb)…
      To take your “financial” argument to its logical conclusion, as a maker of very lightweight bike parts, we should be stressing the importance of bike weight at every opportunity. Whether it’s our handlebars, our tires or our cranks, they are among the lightest you can find anywhere.

      October 28, 2015 at 7:30 am
      • Bob

        Oooooo that’s a cynical take CB. FWIW, I’ve only ridden the supple tires in 26mm and 35mm, and with obvious weight differences etc., they are the real deal.
        I liken it to the first radials on motorcycles. When you switched, it was no big deal. But when to switchec back it was a screaming difference. As always YMMV. But until we have a very clear cut reason not to, give the gentleman the benefit of the doubt.

        October 28, 2015 at 7:48 am
    • Matthew J

      “but one of the biggest detriments of the tires and other gear he sells is that they weigh more”
      Actually the 700×32 Extra Light Stampede Pass tires on my Spectrum are about as light as light a set of 700×32 clinchers out there.

      October 28, 2015 at 11:02 am
    • emem1956

      I want to make a couple of comments, but the word limit stands in the way. So I’ll provide this information first and then try to post my comments:
      1. Miche make 13,14,15, and, I think 16 tooth 1st position cogs, for Shimano & Campagnolo cassettes in 8, 9, 10 & 11 speeds. They also make individual cogs up to 32 teeth.
      2. Miche also make complete cassettes, in ‘odd’ sizes, for instance, 14–29.

      October 28, 2015 at 4:05 pm
  • bob

    Jan, i think its time to produce 7 speed cassette hub and the cassette, 130mm spacing..

    October 28, 2015 at 7:13 am
    • Jon

      I second the 7 speed recommendation. I’m old school and I like friction shifting and mixing and matching parts, rather than the indexing/group mentality. I also think the old Shimano Uniglide design seems good because once you wore out one side of the teeth you could just flip the cog around and use the other side since the splines were all the same width (unlike Hyperglide). How about making Uniglide cogs too? Plus, 7 speed on 130mm would have less wheel dishing. Is the shifting on Hyperglide really that much better than Uniglide? I’ve never used Uniglide since I started riding after it’s time had passed.
      The only issue I have is I am severely limited on what I can spend and based on how nice Compass makes their parts, I won’t be able to afford the Uniglide style hub or cogs. I’d like to afford Compass parts but all my extra money goes to my daughter’s expensive sport and I get more joy through watching her than I would with my own hobby.

      October 28, 2015 at 9:52 am
    • Conrad

      Heck yes! Among the worst of our current mainstem trends are 9+ cogs in the back. You are forced to use a more expensive, more fragile, and less durable drivetrain. And a horribly asymmetric and therefore weaker rear wheel. All to cram in more cogs that you don’t need anyway. Its getting to the point that an 8 speed or less drivetrain is becoming obsolete. I really doubt Shimano or Campy are going back so it is time for Compass to fill the gap. Please?

      October 28, 2015 at 1:19 pm
    • Greg

      Agreed!!!!!! I will stock them for sale at Bicycle Classics!

      October 28, 2015 at 1:57 pm
    • Jeff Lyall

      Maybe not relevant to this discussion directly, but someone local here is doing it on mtb wheels. Unfortunately this would be 135mm, not 130mm. Still, the idea of a strong AND light wheel appeals to me.

      October 28, 2015 at 6:45 pm
    • norman

      I think with 10 speed cassette thickness and spacing, 6-7 speed in 130mm dishless hub is possible. Well even 6 is addequate, really. plus you got more secure chain supply esp. those with missing link for easy cleaning.

      October 30, 2015 at 4:53 am
  • Chris Lowe

    My new Cannondale ‘cross bike came equipped with a 46×36 combo up front and a 12-28 10 speed cassette in back. I find that on my 8-12 mile commutes through Seattle I never need the inner ring. I’ve also only used the 12 tooth cog a couple of times and rarely touch the 13. I am thinking about swapping out the rings for a set of Wick Werks 44×34 rings and a 12-30 10 speed cassette for when I do start riding some of the bigger mountains. A 12-30 10 speed has nice 2 tooth jumps through the middle of the range which is where I spend most of my riding time. I’d have 1-2 gears for downhills or tailwinds and a 34×30 combo should allow me to ride up most anything I’m likely to want to ride up. I may also keep the 46 big ring and simply switch out the 36 for a 34. Either way I like not having a huge 16 tooth jump in chainring sizes and with 10 cogs in back there’s little reason to not spend most of time time in the big ring with the inner ring as a bail out option.

    October 28, 2015 at 12:52 pm
  • emem1956

    If your bike’s designed for many conditions, why do you have to change chainrings for different contexts? Using a Miche 10-speed cassette you could have 46/30 13-14-15-16-17-18-19-21-24-28 or 48/34 14-15-16-17-18-20-22-25-28-32 and no need to change them. Sure, sometimes you’ll need to finesse a 4/5 cog change, but you could adjust the ratios so that happens somewhere inconsequential. Or use 8 or 9 speeds. Anyway, I’ve found using brifters and both friction & indexed barcons those crossovers are easy—not fast enough for racing, but who’s racing? Is it harder, slower, or more awkward than using a rod FD? (That’s a genuine question.)
    Other chainring combinations (god forbid, even triples) would give closer ratios & wider ranges. But you’d have to use more modern cassettes & chains.
    Is there solid, non-anecdotal evidence that 9–11 speed chains are weaker, wear faster, or have more friction than 5–8 speeds? Are the differences large enough to justify limiting the number of cogs on freewheels or cassettes? I’ve never broken any 10-speed chain, even on a tandem. I can see reasons for using 5–6 speed freewheels in remote places, but in the USA, Western Europe, Canada, Australia, Japan … really?

    October 28, 2015 at 4:46 pm
    • Jan Heine, Editor, Bicycle Quarterly

      Once every 4 years, I do a ride where slightly bigger gears are beneficial. I have a choice of carrying those extra gears around all the time, or just putting them on for that one ride.
      The point of the post is less that you should change your gears, but that different gearing is appropriate for different types of riding.

      October 28, 2015 at 5:04 pm
    • Conrad

      No scientific evidence that I am aware of. I have bikes with both 8 and 9 speed cassettes. I replace chains at the 0.075 wear mark. I get about twice as many miles out of the 8speed chain, and the 9 speed chain is on the bike that is hardly ever ridden in the rain. So it seems less durable to me. I have only broken chains when mountain biking (8 speed chain). The gear mashing and sudden gear changes under load that are harder to avoid when riding trails make me think that using a narrower chain on a mountain bike isn’t a good idea. I needed to replace my old mountain bike recently… rather than buy a new bike with a 10 speed cassette I opted to find a used bike with a more sensible drivetrain.

      October 29, 2015 at 4:18 pm
      • Jan Heine, Editor, Bicycle Quarterly

        I recall when 8-speed became popular in racing. Suddenly, in every race, somebody’s hand shot up, and they rolled to the side, not because of a flat tire, but because of a broken chain…

        October 29, 2015 at 5:10 pm
        • Jeff Lyall

          So it wasn’t just me !! I got so good at mending chains in those days. I though it was just poor technique on my part, as I was a recent convert from running. Never heard that before though, and it was always MTB racing, never road, in my case.

          October 29, 2015 at 5:19 pm
  • Owen

    For anyone wishing to see their gearing mapped out visually (including the degree of overlap that Jan talked about) I’ve found the HTML5 Gear Calculator website very helpful. You can see the spread of different cassette/chainring combinations easily and even compare two set ups. Real world experience on the road is understandably different but I’ve found this site useful to make comparisons before shelling out for new parts. I just swapped my 48/34 double for a 46/30 and couldn’t be happier!
    I have no stake in the website, financial or otherwise, and it is copyrighted by its owner.

    October 29, 2015 at 1:38 am
    • DG

      wow! His site got quite an upgrade. I came upon the site several years ago when it was Flash based and the url was Thanks for posting the new link.

      October 29, 2015 at 5:29 am
  • marmotte27

    I need to figure out my gearing, too. On my randonneur I have 48/36/26 chainrings with a 13-27 casette, and riding on flats and slight uphills I find myself doing numerous cumbersome frontshifts between the 48/36 chainrings.

    October 29, 2015 at 2:08 am
    • Jan Heine, Editor, Bicycle Quarterly

      It appears that 48 is too big for you, and 36 too small. A good solution may be to just increase the middle ring size to 40 or 42. You have more overlap, which is exactly what you apparently need. Then you can use the 40 or 42 for most riding, the 48 for tailwinds, slight downhills and days when your legs just feel great, while you keep the 26 for the steeper hills.

      October 29, 2015 at 2:22 am
      • marmotte27

        Many thanks for that reply, sounds good! I’m going to look into that right now!

        October 29, 2015 at 4:30 am
  • ron bell

    Hey! I am just glad that I stumbled on this post! I have nothing new to contribute, but, I do feel validated in my normal riding style. Usually, I stay in the middle of a 9 speed cluster with a 52 tooth driving the front, 12-25 in the rear. I hardly ever use the 12, if I am using the 25, then I am on the 42 tooth stock front on my 70’s racing bike. Still riding it today, 38 years later!! I appreciated everyone’s insight and input!

    October 30, 2015 at 2:28 am
  • Chris V.

    That gear calculator works nice! According to the calculator riding in a paceline at 23-24 mph @ 90rpm on my 46×14 seems right to me from what I remember. I agree that I don’t quite understand why road bikes come with 53 tooth chainrings. I wonder if maybe this is more of a way for guys to show their macho side by chain ring size.

    October 30, 2015 at 6:40 am
  • Chuck Davis

    Re “Small Differences Can Matter” there was “study” by Stuart Burgess (U Bristol) showing that drive trains w/larger rings/cogs were more efficient than smaller rings/cogs (w/ same end ratio). Burgess showed that Boardman might have added an 100m to hour record had he doubled his ring and cog # . With this not necessarily “new” trend > “compacts – wide range doubles” and smaller outer/inner rings and cogs now 12 >11, it is odd that “we” haven’t considered the efficiency/wear considerations with larger rings/cog (maybe wouldn’t be so “compact ” may but increased diff twixt the rings would be the same)

    October 30, 2015 at 12:19 pm
    • Jan Heine, Editor, Bicycle Quarterly

      For the hour record, 100 m is very significant, but for the rest of us, 0.2% is not a discernible difference.
      Also, doubling my ring size? That would be a 92-tooth ring!
      It is well-known that cogs smaller than 14 teeth are significantly less efficient, hence my bike uses a 14-22 freewheel.

      October 30, 2015 at 8:39 pm
      • Chuck Davis

        Yes, 92 would be a RING for sure, the main takeaway (today) probably should be looking at bigger than the 12 and 11 cogs. I ended up with A 50-34 and 15-34 on my old Takara in the early-mid 80’s; I found it diffcult at my old shoppe to get folks thinking about the time they might spend on their 11’s on LHT

        October 31, 2015 at 8:06 am
    • Greg

      The reason that rear cogs on freewheels were limited to 14 for so long was The Fundamental Law of Gearing. 14 was considered the smallest overall cog that should be used, but that was based on now-ancient metallurgy, so it’s pretty outdated now, and moot…..

      November 2, 2015 at 3:58 pm
      • Jan Heine, Editor, Bicycle Quarterly

        It wasn’t based on metallurgy, but on the geometry of chains. And that hasn’t changed. Even today, a cog smaller than 14 teeth has significantly more resistance.

        November 3, 2015 at 5:57 am
  • Steve Pells

    I think that a lot of the problem is the cassettes, not the chainrings. In the days of 13 or 14 tooth small sprockets, a 48 or 50 tooth big ring was fine. Now 12 teeth seems to be becoming rarer in favour of 11, but the standard big chainring for a road machine has dropped only from 53 to 50 teeth! The othe problem is that the gaps are too *even*: 11-13-15-17 or 11-12-14-16 is stupid; massive gaps at the top, fast end, yet placebo shifts at the bottom.
    Sometimes massive gears *are* just what is needed. The Klosters bypass tunnel is several km long, narrow and only one lane of very fast traffic in each direction. Luckily I went through it in the downhill direction! On the edge of spinning out 50 x 12, so over 70 km/hr, I was barely able to hold my spot in the traffic which was safer than going slower. Stupid really, wouldn’t do it again.

    October 30, 2015 at 2:16 pm
    • Jan Heine, Editor, Bicycle Quarterly

      At 70 km/h, whether you pedal or not makes little difference. Of course, the aero tuck puts the brakes out of reach, so it’s also not a good idea when riding in traffic.
      My solution would be going at my normal speed and taking the lane. The first car that queues behind you blocks all others – you are safe from getting rear-ended. Perhaps the drivers wouldn’t like it, but better safe than sorry.

      October 30, 2015 at 8:44 pm
      • Steve Pells

        Indeed, normally pedalling wouldn’t make a difference at that speed, but it did in that situation. Although downhill, the tunnel is a shallow gradient, really just enough to keep you on top of things. Riding “normally” you’d be going less than half the speed of the other vehicles. Going that fast (not aero tuck-as I say, it’s too shallow-just getting down in the drops and hammering) allowed me to take the lane and just about hang on in the draft of the vehicle in front (although by the end of the tunnel I’d probably fallen off about 20 or 30 m).
        I can’t remember if it was the case in that particular tunnel, but often the long Alpine tunnels also have a strong tailwind which also helps in situations like that. There are giant fans, looking rather like jet fighter engine pods, hanging from the ceiling to clear the polluted air from the tunnels as they are so long. If you’re going in the right direction so that they give you a tailwind, you can ride (pedalling) very fast and fly out like a pea out of a peashooter.

        October 31, 2015 at 2:24 pm
  • Ed B

    I have a custom Zinn with 200 mm cranks and 46-34 rings. I used short crank arms (177.5) and a 53-38 set of Rotor QXL rings on my Charly Miller ride at PBP. I pretty much used the big ring for the whole ride with 53×22 or 53×25 being my climbing gears. My cadence likes to be 70-75 rpm but a recent bike fit moved me up over the crank a bit more and my preferred cadence is NOW into the stratosphere……82 rpm. I rarely need the 11 but it is as nice to have as the 32. When you need it, you need it.

    November 2, 2015 at 1:43 pm
  • Matt

    Add another tall guy requesting 200mm cranks.
    But I also understand it’s not super feasible given you’re a small outfit. And don’t forget to test for higher loads. A 6’6″ guy is not only 13% taller than a typical 5’9″ rider, they’re also 13% deeper and 13% wider. This means the total load is about 44% larger (13% more cubed), and that’s on a 13% longer lever arm meaning the torque you’d have to handle is 63% larger (44% more on a 13% longer arm) than a normal person running normal cranks. That’s a lot more material and testing than you probably want to get into for just offering another size.
    …doesn’t mean I don’t want you to make some classy 200mm cranks though.

    November 2, 2015 at 3:06 pm
  • alliwant

    Is there actually a difference between the Herse inner and outer rings? I’m so poky I can get by with nothing bigger than a 38t. For example, could I put a 38t on the outside and a 26t on the inside of a double?

    November 2, 2015 at 8:29 pm
    • Jan Heine, Editor, Bicycle Quarterly

      There is a big difference. The teeth are asymmetric to facilitate shifting – only one side of the tooth actually pulls the chain, so the other is cut away a bit. Turning it around would give you teeth that have the wrong shape. Also, there is a slope on the inside of the big rings that facilitates the chain riding up toward the big ring during upshifts…

      November 3, 2015 at 5:59 am

Comments are closed.

Are you on our list?

Every week, we bring you stories of great rides, new products, and fascinating tech. Sign up and enjoy the ride!

* indicates required