For those who commented about HR, it is discussed at 16:11. HR is slightly higher for the open efforts, and much higher for the standing efforts.

Awesome video! Great to get your take since this is obviously more of an engineering question which I''l admit I have no background in, my degree is in exercise science. The upstream losses didn't occur to me when I was making the video but multiple people pointed it out in the comments. What this means is that when your suspension is open your power meter (for the purposes of training and pacing, not in actuality) is inaccurate. I've never heard anyone talk about this. This should be common knowledge for cross country mountain bikers at this point! That's my big takeaway from this whole experiment. Keep up the good work, love the channel.

You nailed where all the others fail on their analysis, by measuring output power alone and not energy consumed it is not possible to understand eficiency. Having said that, in an MTB, what matters is your efficency climbing on rough terrain (and not tarmac) where the wheel is stuck to the floor (opened) or jumping around (locked). The only way to measure that is to develop a series of sistematically reproducible experiments with different roughnes (from gravel to big stones/roots) and then see when if it is worth to climb with the switch opened or locked. To my knowledge nobody has done this publically and i believe it is not an easy one to do due to the complexity of a reproducible consistent setup.

Pulled out the thermo textbook, made some assumptions, and made a guess at how much thermal energy you loose to the shock. Over the hour at 300W you produce ~1080 kJ, and heating up the shock takes 16-20 kJ (worst case), so less than a 2% loss. There’s far more at play but some perspective is nice

I had never thought about the 'upstream' energy before you even look at the cranks, that's a new perspective for me.

As someone who has a very distant appreciation of engineering (I understand broad concepts mainly from motorsport videos etc), but have never gone deep into it or looked at the maths, you explain things so well to a relative layman and make it super interesting

The best explanation of anti-squat I have ever heard. Great video!

Very interesting stuff. Having originally started watching the channel for the road bike design discussions, this video really highlighted something to me. There is some actual design engineering and science going into mountain bikes. By comparison, road bike engineering seem to just be [bad] design for manufacture and making some "sales engineer" claims about stiffness, very specific use case aero, weight etc. It's refreshing to see some science in bike design!

Mate that was an excellent video, really well put together. I don't know where you're based but I work in the sport and exercise science department at Staffs uni and we have portable gas analysis equipment that could be used to address your point about metabolic cost independent of power output. Would love to run the experiment again in an ecologically valid environment (as you have here) and it would be even better if we got Dylan onboard. I feel a publication coming on.

Awesome video! Clear explanations of a lot of in depth topics and great conclusions to be made. Makes me wonder if a similar thing isn't occurring in road bikes with lower lateral stiffness. I think a video covering the effects of stiffness on drivetrain efficiency and metabolic efficiency could be great.

Great research and results. I watched another similar programme a few months ago aith the same results. This makes choosing my next bike and fiddling with valves much less stressful.

Never seen any tests so detailed on any other bike channels… huge thumbs up

Spot on! When I saw Dylan Johnson video i knew right away, that it would be against law of physics, if the same amount of energy would be spent on rigid bike vs open suspension bike given same power and speed. There is no way that movement of parts in suspension system would be "for free". And he admits that during his tests he had higher average HR with open suspension, but he attributed that to the order of his test runs protocol: first locked, then open. But i knew that in HR differences lies the answer to that question. And measured power from power meter is only one in many variables in that equation, as it only measures what is happening between your foot and pedal, and does’t take into account the whole system. And BTW if it was true that open vs locked suspension doesn’t matter, the question of road bike stiffness also wouldn’t matter which, as we know, is not the case.

This has been a very interesting watch, thank you

Terrific stuff, man. The answer seems obvious in retrospect, but only because you explained it so well.

Great video, very well explained and we'll "exampled" discussion. I don't do any real engineering at my "engineering" job, so it's refreshing to go over this material, and even better that its regarding my favorite hobby.

@Peak Torque The topic on "human energy-output COST" is a great one! It has "vaguely" been addressed I think, when bike fitters say things like "position needs to be aero as possible, while still maintaining highest possible power output within that position". Heart rate is a telling component, if we assume oxygen consumption is linked to pump cycles, and also, oxygen is the key component in "burning" calories, turning them into energy. There was a video on "veritasium" where he finds out, that we weigh about 300g less in the morning due to exhaling the "heavier" CO-2, which is the residue of our energy transformation system. I am completely going off topic, sorry. THANKS for geeking out on this high level, as you do !

Great bit of analysis.

Great video and great point!

You did a great job on breaking down anti squat and explaining the suspension work. One thing that both you and Dylan did not test, is the effect of locked and unlocked on actual off-road trails of various conditions where mountain bikes were designed to be used. What is the efficiency, power output, and metabolic expenditure on trails? Vibration, bumps, traction, and etc, one would think, would have an effect on the test results as well. Would be interesting to see the results off-road!

As someone making an electronic system to control lockout for my final year project this video was an awesome roller coaster of emotions haha! Awesome video as always

Very good sound reasoning there, nice work! I, like you had a 'but wait a minute?" moment watching two other videos which appeared to demonstrate that there was no difference between the efficiency of active Vs locked out suspension. I had never really thought about how power meters work, and what they don't record, (because I can't afford one, and don't care enough). I do have a medium travel mountain bike with a rear lockout though, and I don't think I can be imagining how much easier smooth fire road climbs are with the suspension locked out are. These tests haven't spent much time discussing sprints either. I would have thought that would be an even more exaggerated example of power loss through suspension movement. As you were saying, the only way you can push against something and have most of that energy returned to you is by eliminating all damping. That is not a good thing for suspension control. There is plenty of discussion about lightweight wheels, with lower rotating mass and how they improve acceleration. I bet those gains are insignificant compared to the inertia of a system that loses energy through suspension compression.

Nice experimental design, and clear results. An interesting analysis, if you have the data would be to look at the efficiency factor, as defined by Friel (NP/avg HR), using that as the response variable might yield a different result than using time.

Thanks for that, that makes complete sense. But i needed you to tell me before i could realise that.

First of all: Big Thanks for the information. Despite the fact, that there is a difference in effort to produce for example 300 W, the losses in the system are to small to be measured with the equipment used. So in terms of difference within the system there is none. Dylan showed it and you confirmed it. However: I still lock since it makes me feeling less effort, what brings us back to your conclusion.

Love to see a youtuber build off of another's video and nice job crediting Dylan!

Another great vid to talk and laugh about during the summer afternoons we might spend over meze in that Greek island...each 10 time zones away from it, in different directions. Btw, I flip the shock switch climbing long stretches, but it is not the end of the world if you don't. I don't bother with the fork.

Excellent video dismissing an old chestnut! Thank you!

Great analysis. Your car engine analogy makes perfect sense. In my mind I'm an efficient GR Yaris engine, but reality is more BMC A series.

excellent video and insight.

Thank you for giving a clear explanation of why lockouts help. I’ve seen some videos basically saying they don’t matter, but by using a power meter as a control and your heart rate and effort as the variable, you haven’t proven the value of lockouts. Now I don’t feel like I wasted my money!🤣

Related to your "power meter only measures output power" comment. To take that statement one step further, the power meter only measures your output power at the crank, not the true output power of your legs. An obvious demonstration of this is if you stand on the bike and hop in place without pedaling. A power meter will read zero, but your legs are clearly putting out (significant) power to lift of your center of mass. This energy is dissipated in your muscles and suspension when your center of mass falls. I think this happens when pedaling too, particularly when standing and pedaling it's hard to keep your COM a constant height, and some of this COM-bob energy gets burned in the suspension rather than being recovered during the pedal stroke. Even sitting and pedaling, if the suspension is cycling, that means your COM must be moving up and down, and some of the energy you put into raising your COM gets dissipated by the suspension every cycle. Another example where the power meter and human effort can be dramatically different - if both feet are loaded downwards on both the up and downstroke (like in the pedaling standing example, if you don't fully unweight or pull upwards on your back foot), the leg on the upstroke is doing negative work while the leg on the downstroke is doing positive work. These numbers are extreme, but say the leg on the downstroke is doing 400 watts and the leg on the upstroke is doing -100, the metabolic cost would be much higher than +300/-0 (~1/3 higher if we assume muscle efficiency is constant, we can't regen negative work sadly) , but the power meter would measure 300W for both. I don't think people pedal like this, just illustrating the power meter point.

Your next venture seems clear. The metabolic costs of the suspension setting and body position must be explored! It seems like one of those large, inclinable treadmills and the necessary biometric equipment should do the trick!

Yes! Scientific sense prevails! Perfect video!

Thanks for posting this, I really liked the explanation about anti-squat. Is there any chance you could extend this and tell us how the Yeti Switch Infinity system works? It's been bugging me for years!

Peak Torque: Riding on enduro bike in lycra. GMBN: Wait, that's illegal!

I watch every video both you and Dylan put out and really appreciate both of you. I have a 2018 Specialized Epic (100mm front/rear) with the Brain (rear only) and had been riding with the Brain in the second stiffest setting for 3 years. I thought it was very efficient and every time I rode it, was very impressed on how well it worked. However, after watching Dylan's original video on the topic, I decided to do a ride with it wide open (basically with the Brain off) and was surprised by the results. I have been a little faster but, more importantly, considerably more comfortable over the course of a 3-4 ride. I finish the rides less fatigued but no more physically tired than I did with the Brain engaged. My caloric burns, based on heart rate, not power, are basically identical to the same rides with the Brain engaged. So if having the suspension fully open is more metabolically taxing, I certainly don't notice it. Conditions play a big role here I believe because my local trails are very technical with nothing but roots and rocks 90% of the time. Perhaps all of this matters more on tarmac, but on even slightly technical trails, I won't be damping my compression in any way again. Thanks to both you and Dylan for such great content.

I'm late to the party but as an XC racer and endurance XC racer, I can unequivocally state that racing several hours on a locked shock on anything but the smoothest of climbs will take its toll in terms of fatigue on the rider which in itself is a metabolic/physical cost. I've done my own testing and I simply don't see the metabolic cost for climbing on trail that is being eluded to here, not seated anyhow. It's common knowledge that standing when climbing is less efficient and will absolutely increase the heart rate, this is why most riders only do it for short periods. I have a portable Vo2 master gas analyser that would answer this question quite easily.

Thanks for the insta360 recommendation, I went with a oneR with 1-inch lens, 2hr battery, and wireless mic. I can't wait to start filming with it, need to do the firmware update first... and get a groupset...

Great content as always. My most pressing question is what was that Le col jersey you were wearing? I’ve never seen it

Excellent test! I'm just as surprised as you are about the results. Thinking about it, I think your statement regarding the metabolic cost being the reason for the sensation of 'inefficiency' in the open position is spot on. If you think about it, the power meter strain gauge is more stable in the locked position, and thus, more accurate. In the open position, you reach a certain threshold where some of that power is lost in moving the suspension. In that instant, the position of the power meter is moving with your foot (which would reduce the strain gauge reading, and thus your displayed power), and in order to compensate for this movement, you must've been pushing more than 300w in order to average 300w. I'm not an engineer, but that explanation makes intuitive sense to me, but it might be completely off.

Great video. Not a mtb’r but a lot to be considered for the road bike

As exhaust gas analysis is difficult, expensive, and not feasible while riding, doing a TT at FTP is a good way to keep a consistent effort. Power output measured at the power meter (along with speed) will be lower for the less efficient setup.

Another great video, thanks to both Dylan and Peak for continuing to surface up evidence based research that validates the bike industry's assumptions and associated marketing.

If the maximum amount of human power output is fixed and is assumed to occur at maximum heart rate, then you should be able to see power lost as velocity delta of the bike over the distance traveled during a maximum sprint (Power = Force x Velocity). Measure bike acceleration with an accelerometer at maximum heart rate and integrate to calculate max velocity along with directly measuing bike velocity and you'll have your answer. Conduct the test over a few days alternating which variable is tested first.

What an impressive research guys. Is there anyone else did further research on your job?

My friend pointed out Dylan’s assumption that the power meter measurement is equivalent to your actual power consumed, and I was hoping someone on here would address it. Thanks for setting the record straight! It would be very interesting to see the results of a lab metabolic efficiency measurements across different suspension designs

Very thorough analysis, PT. Would a variance/change in cadence between the different testing scenarios affect the outcome ?

For me I'll lock out the rear shock on climbs that are not as bumpy and typically I feel way better and less exhausted as opposed to when I'm not locked out. So yeah, this video explains the reason why~

Good test. Next steps: You could buy, hire or get sponsored for a portable metabolic gas analyser! Or build a bicycle dyno with a static analyser as your next project! To get a full answer (because we ride at different intensities) you would also have to include a test with different levels of HR because HR is not linear with O2 consumption. So do tests below first lactate threshold turn point, then tests at lactate threshold, then one above (full anaerobic). So overall, with the HR data, could you conclude here that the metabolic cost will be pretty small? If large surely this would show in increased HR? Many thanks. Great work you're doing. DazO

Wow. I was expecting to get lost again …. But this time I understood everything! Well explained. After watching Dylan’s video a while ago I was thinking “why didn’t he just put heart rate against the times?” But I guess presenting the heart rate data against the run times would show quite a low difference probably …. and it would be subject to build up of fatigue during testing. I have never thought before about the power meter reading Vs physical effort cost ….. Having had power meters for a while I guess I kind of linked physical effort with watts …. On any one bike, especially a road bike with fixed geometry and similar road surfaces all the time, then this could be true …. But actually now I understand that this is not the case, especially for MTB where position, geometry and rolling surface changes all the time!

super interesting, and the conclusion makes sense. also feels right. #teamlockout

Great video and conclusion on metabolic cost. The problem to me though is if you’re climbing on technical terrain vs. flat road then having a locked out shock will likely hinder your climbing ability and probably increase the metabolic load of the bike bouncing up and down and demanding more control and stability from the rider. Also. What about the impact of bike geometry on anti squat? I wound have thought the seat tube angle plays a role here and maybe Canyon Strive adjusting the angles on the go provided better climbing anti squat then simply altering the shock compression which the Strive doesn’t do.

Great video, thank you very much for the explanation. Recently, I have watched a few videos about this subject. They all had the same conclusion and the same experiment set up. Basically, the issue is that an experimental setup is incorrect. I wonder whether we can think about this problem as if there would be a spring between shoes and pedals. In order to put 300 w to the crankset, a rider would have to compress the spring first which would require an additional effort.

Hey @Peak Torque, with all the fuzz trying to save a few watts in cycling, it got me wondering how much more distance is covered when standing up vs sitting? See Contador for example, he greatly exagerates when pedalling out of the saddle, going from side to side and doing an "S shape" on the ground... even if that is 2-3%, that's already a lot of time "saved" if just going in a straight line... just an idea for a future video 👍

@3:15 The heat added to the air spring as it's compressed isn't lost, the air decompresses almost instantly. Heat in the air can comes from the air can surrounding the damper.

I like to discuss a crude way to guesstimate the power cost of the moving damper. First, we can safely assume the damping energy loss is less than the peak stored energy in the spring and use the peak spring stored energy as the upper bound estimation. Let’s start with an ideal coil spring without a damper and assume a 75kg rider sag @ 30mm of travel. The stored energy is W_sag = 0.5 kx^2 = 0.5 (Weight/x_sag) x_sag^2 = 11J. To simplify the math, I assume the deepest travel of the pedaling bob is twice the sag, thus the peak stored energy per pedal stroke is 4 W_sag and the change is about 33J per pedal stroke. Further assuming you were pedaling at 60rpm, there are two pedal strokes per second and 66W of power is cycling through the spring when there is no damper. If we use your heart rate near the end of effort to proxy your metabolic cost using a crude linear P_output - HR relationship. 300W @ 158bpm implies additional 13W @ 165bpm. Overall, I would guess about 20% of the spring stored energy is lost to the damper with all the assumptions here.

Great video ! Surprising results. Heart rate would be a better indicator of efficiency,as you saw with the standing efforts. This is what we feel,the energy expense with the bobbing,sag & other losses,we expect the times to be slower. My full sus,has been a frame set only for a long time. So I only have my Kona Kahuna 29er hardtail to go by,usually just locked out on the road or any smooth section. How much does tyre pressure affect the result ? I use 15-30psi.

The data is right there, but it would nice to see this run, even just a single for each one, with your heart rate as close to normalized as possible. The metabolic efficiency isn't particularly important for this, so long as the relative effect of each setting is greater than error.

Here's a test suggestion: Build an insulated calorimeter ("foam box") around the shock to measure temperature gain of the mass of materials inside the box in order to measure the energy loss. Compare energy loss from open vs lock-out mode. Then publish a white paper. Thanks in advance😊

RE; expired gas analysis: you'd need volume as well as oxygen/CO2 inspiation/expiration. One CO2 molecule generated by respiration of glucose does not release the same amount of energy as respiration of a ketone body etc. Even if you assume a "respiratory quotient" you'd still need tidal volumes too. You can already measore CO2 concentrarion quite easily: used in hospitals for 'conscious sedation' sometimes already.

If you put force sensors in the saddle and stem, you can probably calculate power used without gas analysis.

Great video. When Standing on a road bike you can put more weight on the handle bars more easily. Comparing HR when standing Road to MTB isn't just about the shock. What was the HR difference seated between locked and open? Thanks.

Great video - wouldn’t a better test be a max effort climb vs a constant watts. If a closed shock is more efficient- your time to the top of the hill should be quicker. Thoughts?

Love this type of detail and very much appreciate the time investment to pull this off but, the bigger use case that racers are looking for a major advantage is the full gas sprint for race starts and full gas sprint finishes where the watts are much higher and the new technology in shocks and forks might not filter out like the lower watt stuff of a seated or standing climb. When you are trying to get a great single track drop in position from a mass start with a long open start sprint would you see an advantage with a rear shock lockout?

Very interesting video and I love the MTB content! Pinkbike often does a climbing comparison when testing different dual suspension bikes and come out with significantly different times at the same output power. Does this make any sense?

Best video on KZfaq

It would be very interesting to do the same test (standing is useless for mtb) but with 3 different bikes with different levels of AS, measuring time and HR, with powermeter constant. High AS bikes are always thought as efficient pedaing bikes that don't need the lockout but the force transmitted from the chain to extend the rear suspension has to add to the total energy used. Also considering that high AS bikes use their chain to fight off bob when the shock is open, you may see differences between open and locked while using your powermeter.

Would be interesting to see what the difference is like on an off road climb, as locked out you're going to have a harsh ride and less traction, whereas open damping will use more energy but more efficiently.....

Road bike idustry - proud to present next year model rigid frameset, that get 4.7% stiffer MTB idustry - there is no time loss between locked and unlocked suspension

nice testing, thanks. How about a bike with front suspension only? I have a Canyon Pahtlite 6 and it has a 75mm front suspension fork. I use it for road and light paths, but mostly road (of variable quality). Should I lock the front suspension when on tarmac? I like to keep it open but my brother tells me I should lock it when on tarmac. However, there is the video of Dylan Johnson you mentioning, indicating there are studies where locking/unlocking your front fork makes no difference in your performance at all. I mean FRONT fork only of course, no rear suspension or anything like that

@Peak Torque In the case of diving deeper into the metabolic cost when climbing in a standing position ... What would you think about adding strain gauges to the handlbars? Shouldn't this allow for calculating pulling and pushing forces on the bars, which could then be related to the heart rate? They could then give us some more insight on metabolic efficiency, couldn't it?

Fun & interesting experiment, would be very interested in being able to truly quantify the open v locked question, personally I tune my suspension to work for me (at least I think that) on climbs. On dirt say a 50km ride with 2000m climbing, I feel less drained and put in faster times with my suspension open the whole time, if it’s locked out for climbs I find myself very fatigued from the harsh stiff ride, and can get caught out by a technical feature, in the end do I actually burn more fuel open or closed 🧐🤔 🤷‍♂️

There have been lots of tests in the past proving that open suspension does not lead to power loss. It seems the industry forgot for a while, and then all these XC bikes started coming with lockouts again. The suspension movement does dictate the rhythm of the pedaling. If it's counter to how you would like to be pedaling, increasing compression damping or fully locking it out can be beneficial, not for efficiency, but for the ability to hit the highest power numbers. This is most often felt in a smooth pavement sprint to the finish.

Very interesting. Would using a power meter at chainset and hub to compare mechanical losses be useful?

After watching GPLama's video on the Dura-Ace power meter test, do you think the discrepancy was down to the bike being in the trainer? I saw your comment about pivoting. Also, do you think the material differences between Ultegra and Dura Ace will make a difference? Still considering the Shimano PM cranks, but I'd like to see more testing and reviews. Thank you!

What you'd need to do is measure the non-tangential forces. I'd imaging that with the suspension open the pedaling action is more likely to produce higher non-tangential forces at higher metabolic cost. That's why, typically, climbing out of the saddle is less efficient - at the bottom of the stroke you're more inclined to push straight down because of body weight, rather than have your body weight supported by the saddle.

I could tell that it had something to do with how power is measured between the cranks and rear wheel but I was still surprised because I thought that all this motion in the frame would impede the bike movement somehow. Like a spring that oscillated back and forth while moving in a direction parallel to the oscillation, I thought maybe it would add rolling resistance or something. I figured that in one half of the oscillation the back wheel would be "pushed" back thus impeding its movement. Then again I guess the other half would pull the back wheel forward canceling out the effect?

Great jersey! Where it can be bought?

Wasn't DJ's HR 4-5 beats higher on his run that didn't have his shock locked out? Either way, good job on the video and it points out that properly devised testing is just as important as doing the testing itself.

Quick someone take shots for how many times he says “peak torque”

I suppose if there was a way to replace the human with an engine where you can control/measure the power output you could then measure the loss.

It would be interesting to compare ftp test results on both locked and unlocked suspensions. The difference between the two would give an idea of the decrease in metabolic (upstream) efficiency.

Would love to see a test of climbing efficiency on a “brain” shock like specialized has vs a hard tail

C'mon, mate! Antisquat is not a chain tension thing - chain tension affects the antisquat, yes, but just alters it. antisquat is determined by the pivot placement in the first place (by an angle from rear wheel ground contact point to the pivot). you can get a lot of antisqat with a shaft in motorcycle or a car, or in a mountain bike with a high pivot and a pulley - the pulley is there to isolate suspension from a chain tension, but the bikes still have a lot of antisquat, high pivots usually have over 100%. you can check it yourself in linkage - take a commencal supreme for example - chain has almost zero effect on a suspension with a lot of antisquat. or read a book "Motorcycle Handling and Chassis Design: The Art and Science" by Tony Foale - he explains the suspension kinematics for bikes with shafts and chains

I think you could make a reasonable approximation of metabolic cost by measuring expiratory volume (spirometry) or respiratory rate (heart rate monitor).

Smart! Same power into that chain/crank HAS to result in same speed (aerodynamics held constant)… basically comes down to wrong test! Test used is incapable of getting the answer we want!

Sooooo we need hub based power meters on mountain bikes to compensate for the suspension losses and get accurate power output readings? Like crank vs wheel HP in cars?

My Ibis Ripmo with DW link suspension hardly bobs at all, never use the climb switch on the X2 rear shock.

Can't heart rate be the a rough measurable metric to decide how efficient each suspension setup is? In theory if some activity is harder for your body, your heartrate goes up, if you maintain said load. You can also base your load on certain heartrate.

I am wondering if the results would be different on an uneven surface.

Makes sense to me that it would be the same. What goes up must go down etc

Ooooooh we're just missing hambini's PowerPoint's pen and we're in KZfaq's cycling dream.

What is the purpose of the quick release axle inside the bottle cage on the TT bike in the background?

Finally, someone points out that drivetrain-based power meters do not measure any power lost that never goes into the drivetrain.

That logo really made me think "Ah yes, the new video from Teak Pork"

I´m not a scientist nor engineer but when I think about it. Does it matter if it is locked or not when power is measured on the bike (spindle, pedals or cranks)? Power applied is going to be transfered via chain on the rear wheel anyway. If you could measure power from legs (hip- knee-ankle linkage) and testing terrain is very rough so suspension is actively changing biomechanics of your stroke then this would be the place where power loss/gain could be found. Just an idea.

It could have been interesting if have added your heart rate numbers to the power spreadsheet. It could show some of the differences. The conclusion must be that powermeters doesn’t work for mountainbikes, your have to use a heartrate monitor.

I think the only way to settle this is to hit the hill climb every day for about a year and time how long it takes to see which you can do faster.

Try a test of peddling up a bumpy gravel road or single track. I’d wager the metabolic cost is less with the suspension open because the suspension is absorbing the trail chatter instead of your body.