By the Numbers: The New Breathability Gap in WPB Jackets and What the Breathability Tests Get Wrong

Richard: I just looked at your latest dump: Wet Out MVTR Reduction. These tables contain numerous quantitative claims without traceable experimental references. For example, the assertion that 65 W/m² of evaporative heat transfer is equivalent to 8,000 g/m²/day is incorrect; the proper conversion is approximately 2,290 g/m²/day. In addition, ASTM F2370 is not a rain-tower test. Before discussing the conclusions, I would like to see the actual publications and experimental data supporting these numerical claims.

Stephen & BPL Forum members

Many years ago I did primary and secondary research on windshirt breathability for about 10 years in parallel to testing conventional insulations and fabrics. My secondary research required  substantial costs and copyrights restrictions during that period.  After that I switched to other areas of investigative research. Secondary research access may change with time and and an individual’s access rights but the laws of physics don’t.  For an expedited and simple solution, compare easily accessible  peer reviewed scientific research using AI

Today each BPL member can easily ignore both mine and Stephen’s contradictory findings regarding the effect of how dynamic moment changes the MVTR of a moderately breathable wind shirt versus any WPB. Determine the truth for yourself in this era of AI. Choose any free and fee-based AI system and post 2 questions, word it the way you feel comfortable. As an example, I just used, “Does a wind shirt in the 35CFM range increase its MVTR through active hiking more than a microporous WPB? As an example, I posted this to Gemini-3.8-Flash because it seems to have most real-time data. Be independent and use different  queries and AI sources so that increases the chance of finding any contradictory opinion.

This is a copy of my initial query and then a follow up query asking for the source of the information.

https://poe.com/s/YWV4BPiovHgZyY4p3zYV?utm_source=link

https://poe.com/s/aYe8o1PfhNaaw2WoACfx?utm_source=link

By using any AI system, you can EASILY determine for yourself what the peer-reviewed scientific research concludes.

Backpacking and internally soaking with sweat a Dooy, Houdini, impermeable rain jacket and my WPB, isn’t it just common sense that the more porous the fabric is to air, the faster it dries when you’re hiking?

This is an updated version on my prior post on wet out drop rates. There is an apparent bug in the BPL software that doesn’t allow an “Edit”. The only option is “Link to Post”.

For me, the column that I focus on is MVTR  after Wet Out. The Wet Out state occurs quickly and as a result this is the most relevant information if you are actually wearing a rain jacket in rain

How wet-out really hurts breathability
All numbers below come from the mid-points of the best-publicly-available laboratory ranges (Hohenstein Institute, SINTEF, WL Gore IPTS, U.S. Army Natick, Toray and Columbia Sportswear).

• Test protocol assumed: ASTM E96 “Desiccant” (Procedure B), 35 °C / 40 % RH chamber, 20 °C / 65 % RH room.

• “Dry” = brand-new laminate with a pristine DWR.

• “Wet-out” = identical swatch after its face fabric is fully saturated for 10 min, free water film visible, then re-tested without drying.

The fabrics are listed from the largest to the smallest average percentage loss.

Reading the table

Porous laminates (ranks 1-4) lose roughly half their advertised moisture-vapour transmission once the face fabric soaks through because the external water film physically blocks the pores.

Monolithic and mildly air-permeable membranes (ranks 5-7) have no pores to clog; the hit comes mainly from a weaker temperature / vapour-pressure gradient, so the loss hovers around one-quarter.

Membrane-outside constructions (rank 9, e.g., OutDry Extreme) basically sidestep wet-out—the slight drop you see (≈2 %) is caused only by evaporative cooling of the membrane surface in driving rain.

These numbers are directional guidance, not immutable facts: MVTR varies with the exact textile, test method, and environmental gradient. Still, the ranking captures how different technologies behave once your jacket’s DWR finally gives up.

Stephen once again, this a banger of a study. I especially appreciate the YouTube summary from Ryan.

These issues are relevant to us all, but of particular interest to me. I’ve been working on these problems. Wet out, bonding between layers, dissimilar materials bonding (polyamide, aramid, silicone substrates and coatings) from a chemical and molecular level for years and hope to bring to market some real solutions in the near future.

I think the roll to roll industry has gotten fat and happy and things have changed rapidly in the last year alone. The race to meet and beat PFAS is on.

Brace yourselves. Some fun stuff is coming.

How a 35 cfm windshirt compares in MVTR to a WPB when static and when in motion is definitely interesting.

Much more interesting, at least to me, is how an air permiable WPB compares in MVTR to a non-air permiable WPB when static and when in motion.

Apparently the CFM of air-permiable WPBs like Futurelight, AscentShell, and NeoShell ranges from 0.4 to around 2, depending on the version.  (Manufacturers like North Face use versions with CFMs at the high end of the range when less waterproofness is needed.  Indeed, one might debate whether those at the high end are waterproof.)

Depends what you call waterproof. If the thread used to make the fabric is waterproof. Water coming down at an angle is going to run off. It’s not being absorbed into the fabric.

Richard: The table above is preposterous.

ASTM E96 Procedure B is a water test at 23C and 50% RH, not a desiccant test. No E96 procedure exists at 35 °C/40% RH. What you have here is the temperature/humidity conditions for a guarded hot plate, the standard fabric test conditions that fabrics are exposed to before a test, combined with a mischaracterized E96 procedure. Room conditions for E96 do not matter because the test is conducted in a test chamber which is maintained at the prescribed 23C and 50% RH. This sure looks like AI hallucination.

There is no standard wet-out protocol.

How could six testing entities possibly agree on the same non-existent test procedure?

Whatever the wet-out level, it could not be maintained over the entire test period.

Moisture on the test fabric would be absorbed into the upright cup, so the impact of vapor transmission from the chamber could not be distinguished from vapor transmission from the wet-out. This test arrangement proves nothing.

Of course, you still provide no link to any of your claimed results.

The E96B test cannot produce the magnitude of MVTRs listed in your table.

This data is completely bogus!

This message is addressed to Richard specifically, but I also want to provide the broader community with some guidelines for discussing topics like this where evidence may be subject to debate.

I’ve reviewed this thread carefully, and have studied the concerns raised about the sources of the evidence that has been presented in this discussion.

In response, I’d like to step in and offer some perspective grounded in this core principle:

Presenting numerical comparisons as research findings carries a responsibility to make that research identifiable and open to scrutiny.

The wet-out and static-versus-dynamic comparisons have been presented as findings from laboratory testing. However, the supporting publications and reports have not been adequately identified, and specific problems raised about the citations remain unresolved. Naming testing organizations does not establish where the numbers came from or whether the underlying tests support the comparisons when this data cannot be independently corroborated.

These issues represent material barriers to having educated, evidence-based discussions about topics where objective data must be cited to support performance, product, or behavioral claims. Readers are using these claims to interpret how rainwear performs and to inform equipment choices. Stephen – or any other members participating in this discussion – cannot reasonably be expected to reconstruct missing research or disprove each number before we address the reliability of the contributions.

My request to Richard (and a burden that I ask from anyone who makes a claim based on evidence or established data): please provide traceable sources for these comparisons, identify which results support the reported values, and explain any calculations or assumptions used to derive the rankings. If you cannot substantiate them, please withdraw the unsupported claims or reframe them as theories or opinions and establish that as part of your debate.

Until that happens, readers should not rely on these tables as established research findings.

In addition, there are many existing source questions that need resolved. Some sources are missing, some sources point to references that have nothing to do with the source, and some sources are used to substantiate claims that are out of the actual scope of the referenced source study.

AI can help locate or explain research, but its answers do not establish that a study exists or supports a claim. AI is not a student with critical thinking skills – it’s an algorithm human language pattern recognition tool – at the end of the day, as powerful as it is, it’s still just “deterministic software” and is limited in what it is able to infer when it comes to complex scientific reasoning. Asking readers to query other AI systems does not discharge the contributor’s responsibility to verify what they publish when the contributor is claiming to come from a scientific or otherwise authoritative background.

I welcome challenges to Stephen’s work (and to anything Backpacking Light publishes). That scrutiny is valuable when the evidence and reasoning are available for examination. At the end of the day, this process of debate advances our understanding of these topics and makes us all more informed. But the unsubstantiated information and irrelevant source citations that have been posted in this thread makes this process much more confusing, and leads readers to believe things that may not be true – and that compromises their ability to make good decisions.

So on the finetrack website they note the everbreath air membranes’  Breathability: at 10,000g/m²/24h (JIS L 1099 A-1) / 20,000g/m²/24h (JIS L 1099 B-1). So let’s say for example I’m going to consider this product for purchase.

Say I wanted to estimate where a product would fall on your chart of jackets (one you did not test). Is it possible to do this?  When we are reading B1 results – did 20,000 b1 tests typically feel like 3000 in your tests? I’m not sure I followed figure 4 well enough to understand it. Should I take a b1 test and multiple by some cofactor (ie 0.15 x b1 MTVR)  to estimate where a jacket would land in your study?

And importantly, at what mtvr in your chart did you deem a jacket to subjectively feel breathable? 3000? 3500? 4000?

thanks!

Hi Hanz:  The Everbreath is what is used in the Photon jacket I reviewed for them a year or so ago.  It is a microporous polycarbonate/PU membrane.  Around 3700 on my test, so very good. I think the Photon is an excellent jacket, but it is in other, which I have not tested.  Unless the face fabric is very different from Photon’s, I would expect excellent performance.  Their sizing is Japanese.  I am 5’10” and 175 pounds. Their XL is an OK fit, but I wouldn’t mind a bit more room. If it is a Photon you will want to apply zipper lube periodically.  If it is one of their other products, that should not be an issue.  Any other questions, you know how to reach me.

OK, I took the bait and did my own AI search, looking at how vapor transmission of different WPB types is affected by wet out.  And what I found is … very little.

Either my AI is being lazy (I used ChatGPT set for high thinking effort (though perhaps with something beyond my entry-level paid subscription, it would think harder) and made a number of follow-up queries, whether about details, meanings, or simply finding more results), or very little research has been done on this topic (and a fair amount of what has been done is published on backpackinglight).

Of the handful of articles/studies found, three seemed to say something relevant to my question and be interesting:

First, Ruckman ’97.  I haven’t been able to access the full article.  From what I have been able to gather, she apparently looked at which membranes stopped transmitting vapor soonest when exposed to heavy rain.  Soonest, and in therefore (in this sense) worst was microporous PU.  Best was hydrophilic, monolithic (not sure whether PU or something else).  In between was PTFE.  No idea how close they were to each other.

In addition, the abstract to the article says only PTFE had no condensation on the inside.

Second, Marxmeier ’86.  Working with bare membranes, not fabrics, he found that an external water coating reduced PTFE vapor transmission by 20%, but hyrophilic monolithic polyester (not PU) by only 2.3%.  Both percentages seem surprisingly small.  If PTFE only loses 20% of vapor transmission when fully wet out, then the still better performance of some alternative can be easily outweighed by some other advantage of PTFE (if there is one).  But perhaps that results from his test procedure producing more favorable results than others?  And/or maybe one should add another 50% reduction due to the fabric layers, so it should be 70% vs 52.3%, etc.  Also, given where Marxmeier worked, there’s an open question whether he was involved in the development of the Sympatex membrane (also hydrophilic monolothic polyester), in which case one might need to be concerned about bias.

Finally, Salz ’88.  Exposed to simulated rain on the outside, two different ePTFE fabrics lost a much smaller % of vapor transmission than 2 microporous PU membranes (65 & 76%), and worst of all was hydrophilic monolithic PU (81%).

What’s it all mean?  Well, there may be differences between the many fabrics of the various types mentioned, and some may perform much better or worse than those tested.  Without knowing more, these results suggest PTFE is better for vapor transmission in heavy rain and/or when wet out than microporous PU.  In addition – though the evidence is extremely limited, hydrophilic monolithic polyester MAY be best of all (in this particular regard); whereas hydrophilic monolithic PU is either very good or very bad.

just field testing:

in light rain or less the generic WPB worked just fine.  In steady heavier rain it wetted out and water accumulated on the inner surface.  This is consistent with MVTR reducing.  As in Stephen’s testing.

with neoshell, is steady heavier rain it didn’t wet out and water didn’t accumulate on the inside, again, consistent with Stephen’s MVTR testing.

This MVTR concept seems really useful.

I have not seen much studying of under what conditions the DWR wets out, and different DWRs.  Except qualitatively, C8 wets out the least, C6 not quite as good, newer ones worse.  Richard posted some test results that talk about this and are consistent with my field experiences.  It’s tricky because if the DWR gets dirty its much worse.  Testing new fabric isn’t very useful.  Having fabric that’s typical of actual fabric in use is difficult to control.  Stephen has tested fabric that’s been used which is good.

There was an episode of freakonomics.  Someone wrote a paper that showed people’s quality of life doesn’t improve if salary goes above some value, $100K or something.  Someone else got a data set and decided to ask the same question and found the opposite.  Rather than arguing about it on twitter/X and degenerating to personal insults, they collaborated, why would one person’s data say one thing but another person’s data say the opposite?  I forget, I think one person had more rich people in their data and their quality of life kept improving with higher income.

I think I brought this up before, but I worked for a number of years in a well funded audio research lab where user perception was rigorously tested to set audio design targets.

i see a strong parallel here with rain gear and wind shirts.  Without carefully testing in controlled environments how mvtr specs translate to user perception, this is all interesting but not of much practical use

Its like audio in the 1970s where companies quoted specs liberally because they were easy to measure but there was no way to tell where “good enough” lied or which specs really translated into better perceived performance.  It took years of research at Canadas National Research Council before we really understood which specs mattered and it royally upset the apple cart and was significantly counter to the assumed “science” at the time which was based on speculation.

I see the whole mvtr topic the same as audio in the 1970s.  It’s assumed more is better but is it really?  If it degrades quickly is it useful?  Where are diminishing returns?  Where is the lower threshold of utility?

Dunno, no one knows.

I’m not making light of these investigations and the truth is testing these human perception thresholds would be a brutally difficult exercise.  Having done this for years in audio, I can only begin to speculate on the complexity here.  Not only the variance between individuals in perspiration and heat output with effort but differences in how the vapour accumulation in the garment  is perceived.

Over the years I’ve spent hours combing through both Stephen’s and Richard’s data but have gotten nearly nowhere.   More mvtr is better but I’ve never found the “good enough” threshold.   Of most use to me is that while the mvtr concept has value it is still not well understood and I think Richard is definitely right that breathability matters as well.  I own a few wind shirts and 35cfm sits about right as a target from my seat if the pant personal experience.

David you hit the nail on the head. Short of FDA style human trials the stats mean “something” just not everything. Same as the R Value argument. Perceived warmth is everything. Perceived dryness (or lack of wetness) is everything. Our measurement standards are providing important, er, measurements. There’s a whole world in between those nodes.

I find incredible utility in the work that Stephen has provided here. We are now asking the right questions.

You need to do field testing. Different mtvr garments.  Different conditions.  Record the conditions and results.  See if higher mvtr garments are actually more comfortable.

Then, you could establish what mvtr (or whatever) values produce better results.

Then, you could lab test many garments to see which have better mvtr.  You wouldn’t have to field test every garment.

If you want individualized results and had access to a number of different wind shirts and had the patience to test them in the field repeatedly and had some way to measure conditions and control variables and control for bias and false positives then you might get somewhere useful for ones own self but it probably wouldn’t have any general applicability.  But it would start the conversation.

I don’t want to derail this thread with a discussion about how to properly conduct testing to set general MVTR targets.  I don’t think any of us have the necessary background and training to come up with such tests.  Even in audio they were all highly trained PHD cognitive scientists running these controlled experiments in lab settings.

Hi Guys: A little housekeeping. Two items: Back to the Outdoor Vitals Tushar.  In the first post on this thread and in the article, I described the residue inside the jacket sleeves after contact with sunscreen. I got no response to my initial inquiry to OV to discuss this issue.  I finally got to discuss the issue just before publication.  They said to wash the jacket in a tech wash.  I did.

You can compare with the original image.  After washing, most of the residue is gone. I don’t know how oleophobic these new materials are, especially without PFAS DWR, which was hydrophobic and oleophobic. I think it is safe to assume C0 coatings aren’t oleophobic.  Perhaps the lesson here is that if you see residue on the jacket, wash it with a tech wash to try to keep it from invading the pores. I have no evidence that the residue impacts the liner, but other than drop tests, I can’t really test this sort of small fabric area.

Next, I tested two more Toray Torrain WPB shells.

Once again, excellent breathability.  The Storm Shell is 80×70 denier, so a fairly heavy fabric.  The Chugach TR is about 5 oz lighter. 50D face fabric and a couple less features.   Good feature set for my use in the TR.  For me, the Storm shell is too heavy and more features than I need. I’d really like to find a 3 layer, 20 denier face. I have now looked at Toray Torain implemented in jackets from three US manufacturers. I think they are the microporous membrane to beat.  I hope more manufacturers turn to this product.

I think it is safe to assume C0 coatings aren’t oleophobic.

Reasonable to assume this currently, but it depends on the chemistry (although the promising ones are still at the lab stage).

Hi Andrew S: I’ve collected the articles you cited as well as some others. However I cannot find what you are calling Marxmeier ’86.  Can you provide the full citation so I can read it.  After I  read it I will write a summary of these articles and what they mean for how we deal with rainy conditions.

At least, unlike we thought in 2015, we know now that C0-DWR can be oleophobic and as the industry for the first time is forced to seek a DWR that is PFAS free and truelly durable, to stay competive, something that matches and even exceeds Shakedry-levels while being PFAS-free, is just a question of time.

I read that frog toggs are polypropylene.  So they will never wet out.  Supposedly even heavy rain won’t get through but that’s not clear – their pore size is fairly large compared to membrane.  Supposedly they don’t have very good MVTR.

I wonder if Stephen has ever measured MVTR?

This may be an edge case where even with lower MVTR they’re perceived to be better because convection can get air to go through.  Or is that part of MVTR?

They’re not very robust so that’s almost disqualifying.  Easily abraded or torn by a passing limb.

Supposedly, it’ll leak rain through where the shoulder pads are.  Or waist band.

Anyone have any real world experience with this?

I made some gaiters with what I think is the same fabric.  They get abraded and torn by a passing stick.  I keep patching them.  I haven’t had a good test of rain though, or walking through wet brush, but I think I’ll get an opportunity soon.  I also have a shirt and pants I made to test.

Ryan Jordan, Stephen Seeber, and those few members interested in complex moisture transport details

Ryan — your assessment of the pitfalls in using AI for citations was both accurate and quite enlightening to me…. thank you for explaining things so clearly. On a very minor note, although you placed “deterministic software” in quotes, in computer science generative AI and LLMs are essentially non-deterministic, meaning stochastic or probabilistic rather than fixed. Still, your core point holds: AI does not “think,” confirm truth, or conduct lab work; it produces statistically likely word sequences.

Stephen – Approximately 10 years ago, I compared the humility inside both a tightly closed 35CFM wind shirt (1st generation Houdini) and a tightly closed eVent WPB Jacket. When static the inside humidity was similar for both garments but when walking in the wind shirt, it was much less. That was the time frame I provided my primary research findings in a BPL free post. I had to rely on secondary research to understand what I was seeing but I couldn’t distribute that information. In that era most timely secondary research was paywalled and pdf copies were watermarked as copyrighted. There was no obvious legal way to give that information to a 3rd party. It has been difficult for me to learn how to do this both legally and accurately in the age of AI.

2026 Explanation Version 2 + Reference able Citations

https://www.icloud.com/iclouddrive/05fVqZPOEuczEU2qBnbnxS_xg

Version 2 corrected wet-out search citations

https://www.icloud.com/iclouddrive/0acmevsVDW70Y1_saCkHgQ21g

I appreciate all of Stephens work

Richard’s wet out data matches my field experience

Avoiding wetting out seems to be the biggest factor in staying dry.  In the conditions I’m in, cool, not that much heavy rain, not huge levels of exertion, if my rain jacket wets out, then water starts collecting on the inside surface, then my shirt and pockets get wet.

My neoshell jacket never wets out.  In extreme conditions it will start to wet out, but will be restored pretty quickly when it stops raining, and water never accumulates on the inside surface.

My generic WPB jacket wets out quite quickly and water accumulates on the inside surface.  When I washed it and treated with revivex it was much better but still not as good as the neoshell.

This is consistent with the data Richard just posted – neoshell must be c6, generic WPB must be c0.  Worse when worn or dirty.

I think it is more complicated than this but this analysis is in a useful direction.  A number of other researchers that Richard gave have similar conclusions.

I know that I’m like a dog that you can’t pry the bone out of it’s mouth, but it seems like there’s potential with polypropylene that supposedly doesn’t wet out.  Stephen posted some good data about how polypropylene can still leak.  Polypropylene is so slippery and has such a low melting temperature that maybe it’s not possible to make durable waterproof  garments.  And other properties.  Rips and abrades easily.