I call Hot Rod Cameras Bluff with a $10,000 Bet (for Sharpest PL RF for Komodo)

[Michael Hastings]

I think the problem here is that Graham came out of the box claiming not only that his mount is more precisely machined but is "sharper", which of course doesn't make sense since precise focus is achieved through the focus mechanism. The only difference would be if the lens is not collimated or the mount/sensor is incorrectly shimmed or adjusted and could not achieve infinity focus. But again that isn't a problem of sharpness - the image is simply not focused.

The adapter is just a pass through with no optics, so unless there is something in the adapter creating vignetting or internal reflections it should have no effect on "sharpness".

It doesn't seem like the differences in flange depth variations between the adapters means much as long as they are within a range that can be shimmed.

More important would be

A That the two mount surfaces - camera side and lens side - are perfectly parallel. Modern CNCs are extremely precise so that parallelism is probably more dependent on how the material is fixtured during machining. We are also assuming that the sensor is perfectly parallel to the base camera mount. If all of the surfaces aren't parallel there would be a slight focus variation and distortion across the sensor.

B the rigidity of the base and adapter mounts - if they are not rock solid it can introduce the same issues caused by lack of parallelism.

As far as the drop in filters changing the FFD one would assume you use a clear filter when no other filter is needed.

 

[GrahamClark]

Then Graham comes in here with zero humility, acting like he's the greatest thing since sliced bread

I quit social media in 2015. I quit email, and forums are definitely not my thing. I'm incognito.

I don't see any problem with saying "I don't like it", but to say "the scientific data says it's bad" when Breakthrough is more spectrally neutral than Singh-Ray, which outperforms every cinema ND we've seen, including TrueND...

And TrueND makes an incredible ND filter.

But when I saw this falsified scientific data rom Hot Rod about our products is out there, how are filmmakers to know if it's signal or noise?

And to test filters with an interferometer (used for measuring surface flatness)... the instrument is called a spectrometer

I'll clean it up with scientific data, by the numbers, not words.

Breakthrough Filters has 300 day returns, 0% tax, free shipping, 24/7 telephone support – Breakthrough really does have incredible customer service.

But we also have zero tolerance for unreasonable or belligerent customers.

Impatience for products is reasonable, but being an asshole gets you fired from Breakthrough. I'd rather my competitors keep them.

There's always the 1%, but not in the companies I run.

The PL mounts appear sharp, but the Komodo is sharper than previously thought.

That's really great news.

 

[GrahamClark]

The microns you are mentioning above is such small numbers so all it does is to offset the focus scale for prime lenses just a super small fraction

This is an image of the side-by-side of Vocas / Breakthrough, they make a great mount. They're off the least, less than the thinnest shim, and the difference is visually discernible.

As a result of this discussion mounts for Komodo will get dialed in, and filmmakers will knowingly or unknowingly get sharper results.

You clearly seem to have a problem with critically sharp out of the box, all good man, just buy any random mount and you're all set to go!

You're either not paying attention or you're lost in space.

I look forward to seeing more side-by-side sharpness results from various PL mounts on this thread in the future, I think that's where the real discussion takes place.

 

[GrahamClark]

Soooooo... if we drop filters in behind a lens don't we have to adjust the FFD by 1/3 the thickness of the filters? So aren't these apples and oranges? And wouldn't you need a clear filter if no filter was used on the filter one?

Nick

Nick, you are correct, without the presence of optical glass in the light path of a drop-in the FFD is out of spec.

If you were to compare the Canon EF RF drop-in adapter vs. the Canon Straight EF RF you'd find the drop-in unit is a bit less than 1mm longer.

If you wanted to shoot without a filter and you didn't have a clear inserted, autofocus won't work, can't hit infinity, despite the fact that the mount is longer (normally you'd overshoot infinity in that case).

Which is why a clear is required to complete the optical path if you want to shoot without a filter.

The thickness of the glass changes based on mount length combined with FFD spec, so a 1/3rd wouldn't necessarily be the figure, from my experience.

(unrelated question and answer to original thread, since this thread is about straight mounts, no drop-in filters)

 

[GrahamClark]

... which of course doesn't make sense since precise focus is achieved through the focus mechanism.

If you were shooting a Canon drop-in filter adapter without a clear filter, your image would be out of focus.

Lens mounts which are not made to spec act in the same way on a smaller scale, and the results may look sharp but they're not critically sharp.

If the shim thicknesses are any indication, precision with regards to flange spec is important, if you're into sharpness.

 

[Kyle Wright]

Well, you quote me then go off on an unrelated tangent about ND filter quality. But you do say some interesting things.

But when I saw this falsified scientific data rom Hot Rod about our products is out there, how are filmmakers to know if it's signal or noise?

And to test filters with an interferometer (used for measuring surface flatness)... the instrument is called a spectrometer

I'll clean it up with scientific data, by the numbers, not words.

Interferometers have many uses. Measuring surface flatness (or more commonly, distance) is just one use.

What about Fourier-transform spectroscopy? It's a method of measuring how much an item transmits and absorbs light at each wavelength. And guess what this uses? A Michelson interferometer! Any reason this interferometer and technique couldn't be used to measure the light transmission characteristics of ND filters?

Is the Hot Rod data (which we haven't seen, as far as I'm aware) correct? Who knows. They haven't provided numbers or testing parameters. Neither have you. At this point, it's all just noise and vague claims from both sides until the data can be looked at. Saying Illya "falsified" scientific data is a bold claim and potentially libelous, however. Sure, it's not unthinkable someone would falsify data, but it could also be the result of bad methodology or flawed testing. Or it could be accurate. That is why scientific data is peer reviewed, not because the scientists are all out there falsifying their results, but because the results they get and honestly believe to be accurate might not be what they think it is.

Breakthrough Filters has 300 day returns, 0% tax, free shipping, 24/7 telephone support – Breakthrough really does have incredible customer service.

But we also have zero tolerance for unreasonable or belligerent customers.

Impatience for products is reasonable, but being an asshole gets you fired from Breakthrough. I'd rather my competitors keep them.

Saying you have great customer service and actually showing it are two things. The consistent story I keep seeing is people paid for your product, were promised it was coming soon, and then they have to keep asking when that means and only get silence in return. I'm not saying you don't have a reasonable excuse for the delays (manufacturing is tough and I'm sure Covid introduced problems). It's the not answering when your paying customers inquire that's the issue. Again, I WANTED to get your filters. I thought (and still think) they are a great product. But I've seen way too many people say that they can't get an answer after having given your company money. This isn't just one or two people saying this. It's a story I keep seeing over and over and over again. It's too many independent sources, all saying the same thing - no communication, for me to ignore. If you can't see how that's a problem, I really don't know what to tell you. What I can tell you, however, is that your super alpha male, aggressive manner on this forum is just plain off-putting and lets me know I made the right decision avoiding your company.

 

[Phil Holland]

Fascinating discussion, yet somewhat how I thought things would unfold. I just wrapped filming in the UK and am on a 3 day journey back to the states and will dig through all this.

I'll start with saying assume the obvious answers are likely taken into consideration among the lens professionals here.

One thing I'll do to spice up the conversation on is regarding PL lenses as well as digital cinema cameras. Some manufacturers previously have been shipping glass collimated for Alexa for instance, in which case right when you buy a set you would need to set your back focus in camera or via your mount or via collimating the lenses. There's been a pretty hardcore movement as of late to be "right on" 52mm for lenses and cameras w/ sensor box filters and coverglass taken into account. This is also having an impact shipping cameras, I would need to ask again when the transition occurred, but Arri is going to the 52mm standard as of pretty recently. DSMC2 has adjustable back focus and I believe stock shipped at 52mm + taking the OLPF into account, Ranger is shimmable only. For those that have filmed PV cams in the past, this was one of the reasons there were optical slugs made for a few different cameras, which is also something people want to move away from.

Quick summary of thoughts. Yes, you need to have a mount that is shimmable or adjustable for a few reasons. Yes, real important to have your lenses collimated and/or setting back focus via lens adjustment via shims or mechanism. The industry has been trying to button up some of this stuff to tighter standard across the board as of late. It's been discussed for years, most visibly for me at NAB amongst various manufacturers. And the more interesting is the state of the market itself when we have entry level cameras + adapters + filmmakers who never collimate, shim, check, or adjust their glass, mounts, or cams.

From what I can tell the BTF RF>PL Adapter is about locking into Komodo's spec to achieve good out of the box sharpness for lenses properly calibrated to 52mm PL spec. That said you need to have your lens properly collimated to likely benefit from that, which is another thing lens manufacturers are trying to button up as a whole right now. Most recently I believe Zeiss is now going "dead on" for Supremes as of very recently onto 52mm.

I'll do some further testing at home, but I have been testing each of the adapters thus far released and a couple that are still on the way. Many have pros and cons regarding build quality, weight, features, etc. Such is life. i.e. I started with the Metabones RF>PL (only one out at the time really) for instance and it on the heavier side of things and that difference was highlighted in the timeline when the Wooden Camera mount was released and now the BTF RF>PL. They are decent in a variety of levels of "decent".

I'll leave with a recommendation/tip. I've already mentioned it in the FAQ in this Komodo topic, but I strongly recommend using a support foot or frame to assist with maintaining stability as well as ensuring your RF mount lasts when using PL glass. I still see a lot of folks using slightly too large glass on their RF cams and I know that doesn't end well. Support foots, chins, and body frames/cages work well. More points of contact and security allow you to use bigger glass with more piece of mind, I'm talking real big lenses. Most who shoot with primes and compact zooms will likely just want a support foot. Then in both cases it's also nice to have lens support itself. There was a shot on FB sometime back of an E-Mount destroyed by a PL mount adapter, it wasn't the adapter that was the problem, but lack of lens support using a long and heavy lens.

A somewhat controversial thought, but there will be a time way down the line when a great deal of this just "works as expected out of the box". The industry (camera and lens manufacturers) very badly wants to get to that point and there has been rather active moment/discussion on this for the last 5-6 years. Things will get a bit tighter likely from decent lens and camera manufacturers. It will still be chaos for a while though, and until that time shim up as needed.

 

[J. Orlandi]

I'm going to chime in here because why the hell not... I have a minor background in engineering / machining and tolerances don't = a win.

Sure you can get something to be within microns of accuracy in machining if you're good, however you are relying on RED's tolerances, which will be much loser than yours, and unless you have a large sample of production komodos measured... well you're data set is just too small.

You could have measured a batch of komodos made during the colder months, now that its summer in CA things change.

This is my point, materials expand, warp, shift. It's the simple thing that is always a nightmare for precision, its never the accuracy of machining... rather the materials ability to keep its original dimensions.

The komodo RF mount is NOT some over engineered titanium mount, it can flex, expand, etc... Without shimming, you have no real ability to claim "sharper" because there's no way to know at what back of komodo, in what temps?


Metal moves and lets not even get started on mechanical strain...



The weight of a cinema lens on the end of your adapter will change things without the best mechanical connections. That vocas 4 point mount is really a very good positive mechanical connection that eliminates the variances by most bottom only secured mounts. The screwing down of an adapter applies a ton of shift without shimming as well.

Mechanical tolerances are one thing, material stresses and material instabilities are another.



I'm all for people wanting to build a better mouse trap, but the photo of this adapter does not bring up any confidence in defeating simple things like stress from mounting a lens.

 

[Björn Benckert]

What graham does not understand is that a slight offset in backfocus does not change a thing except that your focal lenght change extremly slightly. And your focus scale get an offset. You really need to be wack wrong to get blury images across the board / at all focus instaces.

The above is the reason why critical backfocus is not even a thing in still photograpy. Canon screws thier still mounts straight into plastic even on their most expansive still cameras. Why, because perfectly calibrated focus scales are not a thing needed in the world of snapping stills where they use auto focus and the lenses does not have a hard stop at infinite. So no need to care about the still cameras bf distance. What is improtant there is to have rapid autofocus and always be able to reach infinity and past that.

Now when people started to use stills cameras to film and pull focus by hand some 10years ago the ef mount and similar mounts and thier lack of precision became an issue. Simply they where difficult to shim and unstable over time.

Now the pictures of trees infront of the sky that Graham and others post here as proof for his mount is sharper is as I see it complete nonsens. You can not make a judgment of backfocus based on such images... as the difference is so slim if the two mounts have different bf thier focusing point moved just slightly thats all, how can you tell such difference is not caused by lens or simply an other focus point due to different backfocus.

Sure then there can be internal flocking and other things bringing more or less contrast to the image. But this thread is about backfocus and the soul reason for adjusting it is tho get your colimated lenses to focus what says on thier focus scale, no more no les.

Sure a shitty mounted lens adapter can also skew the focus plane like a swing shift lens. But graham does not mention once here thst his mount is mounted in a more rigid or accurate way than the other mounts which is a far bigger issue then the out of tge box shimming.

Simply bolting the mount down onto your dovetail will cause a tilt shift focus to your setup no doubt. So be carefull with doing so and learn from the guys at Kippertie and the procidure they recomend for mounting thier chinstrap.

 

[Vance Colvig]

Much of this discussion has referenced prime lenses it seems to me. Lets consider wide angle parfocal zoom lenses such as Tokina 11-20, Optimo 15-40, or Sigma 18-35 when wide open. All three are popular zoom lenses and there are many more. If back focus is not spot on, those lenses are no longer parfocal in effect throughout their zoom range. Improper back focus has a much more pronounced effect with parfocal zoom lenses than just scale markings being slightly off as on primes. Old school film makers know this only too well. Now that we are into 4k,5k,6k,8k and 12k imaging soft focus is now more visable which makes accurate and stable back focus ever more important.

 

[Robert Hofmeyr]

This is an image of the side-by-side of Vocas / Breakthrough, they make a great mount. They're off the least, less than the thinnest shim, and the difference is visually discernible.

As a result of this discussion mounts for Komodo will get dialed in, and filmmakers will knowingly or unknowingly get sharper results.

You clearly seem to have a problem with critically sharp out of the box, all good man, just buy any random mount and you're all set to go!

You're either not paying attention or you're lost in space.

I look forward to seeing more side-by-side sharpness results from various PL mounts on this thread in the future, I think that's where the real discussion takes place.

What's going on here Graham? This has me confused now.

Is this the same lens and camera, just different mounts?
Is this a crop from the edge? If so, was the centre sharper on both?
Did you focus both lenses on the tree, or are they both at infinity?

As you know, you can still get sharp images if the back focus is out. Hell, you can even get sharp images when using an extension tube (as long as the subject is close), effectively putting the back focus waaaay out.

What exactly do you believe is causing the one image to be less sharp?

 

[James Darren]

But we also have zero tolerance for unreasonable or belligerent customers.
Impatience for products is reasonable, but being an asshole gets you fired from Breakthrough. I'd rather my competitors keep them.

Then you have a lot to learn about basic business 101 as you will always at some point come across difficult or unreasonable customers. Your comments on here immediately make me never want to buy anything from you. It's how you handle situations like this that set you apart, and you've done poorly on this public forum.

 

[John Brawley]

@ Phil....

I don't understand.

PL mount is 52mm.

PL mount Lens and cameras are designed to be collimated to this. They may or may not arrive when you buy them being set to 52mm. Who cares. That is irrelevant. Because a PL mount on a camera can be shimmed. As can a PL mount on the lens. It's a regular part of every prep on every show I've ever done to check this in prep and adjust it. You start off assuming the lens or camera is out until you verify it.

Doing so to 52mm ensures that the lenses and cameras are interoperable, zooms hold focus through their range, something that's important across multi bodies and multi unit shows.

Making claims about a mount being "sharper" is disingenuous when at most it's not been shimmed to the given situation.

It's not sharper, it's just more in (FFD) focus. The mount doesn't make the lens any sharper than what it would be if if they are correctly shimmed.

Even if the camera or lens wasn't sharp out of the box, who cares ? They are literally designed to be shimmed many times in their operating life because as we all know, the FFD changes over time and with circumstance.

Making a long post that honestly appears to be creating FUD around this seems to only muddy the waters and is not at all clarifying.

Sure, stills lenses are all over the place with regards to FFD. Yes they routinely do not actually set their often un-shimmable consumer lenses to the actual FFD of the mount they are made for. Canon especially are guilty of this across their various mounts and cameras, including the C series as well as the DSLRs.

You appear to be conflating those long never-been-made-to-spec-for-a-reason stills lenses with PL mount lenses which has ALWAYS been set to 52mm and if they aren't are easily adjusted, even in the field if needed to be.

It's mischievous to create a problem where there isn't one.

PL isn't the issue here.

Being sharp out of the box is also not at issue because PL is designed to be adjusted as conditions change and we all know they do change over time and with configurations.

I'm still honestly truly baffled by this thread.

JB

 

[Satsuki Murashige]

PL mount is 52mm.

Hi John,

Doesn’t the thickness of the OLPF stack of each camera actually affect the FFD? Yes, the PL Mount is nominally 52.00mm and lenses should be collimated that way, but the camera mounts may vary slightly from 52.00mm to account for this, no? It’s not like anyone is using a flange depth gage and steel plate in the gate on digital cameras, the mounts are checked optically with something like a Denz FFD tool. As long as all of the properly collimated lenses continue to tape out and zooms remain parfocal, I think it’s all good? Please correct me if I’m wrong.

Even on my film cameras, my Arriflex 3C is set to 51.98mm and my Moviecam to 51.99mm, so there was a bit of variation from camera to camera without any OLPF, just due to the different design.

 

[John Brawley]

Hi John,

Doesn’t the thickness of the OLPF stack of each camera actually affect the FFD? Yes, the PL Mount is nominally 52.00mm and lenses should be collimated that way, but the camera mounts may vary slightly from 52.00mm to account for this, no? It’s not like anyone is using a flange depth gage and steel plate in the gate on digital cameras, the mounts are checked optically with something like a Denz FFD tool. As long as all of the properly collimated lenses continue to tape out and zooms remain parfocal, I think it’s all good? Please correct me if I’m wrong.

Even on my film cameras, my Arriflex 3C is set to 51.98mm and my Moviecam to 51.99mm, so there was a bit of variation from camera to camera without any OLPF, just due to the different design.

Yes, film cameras vary. When I worked in prep the standard for Arri was 51.98, which allowed for the thickness of the film itself. Highspeed film versions like the SR2/3HS would be 51.975mm.

On the camera side I understand the 52mm depth is set with the OLPF / sensor cover glass accounted for. RED is unusual for some of their cameras having an adjustable sensor. The sensor cover glass / OLPF filter stack would certainly change the FFD, but when you set the cameras FFD on a digital camera you're taking that into account visually (You're not actually measuring the depth to the sensor) but it's "nominally" 52mm.

Just like on a film camera where each model has it's own actual setting. Also why shimming is an important function (or a moveable sensor)

JB

 

[John Brawley]

Here's some cool numbers with some actual real world sample testing.

Notice that most cameras aren't set to the actual FFD, but many of the consumer / producer cameras that have mounts that can't easily be shimmed, which isn't anything like what you can easily do with PL.

This is most likely because those users that are working with stills lenses, and AF type lenses are built to lower cost, lower precision and they want the AF parts to be very lightweight and don't need the correct scales and consistency that cinema lenses tend to have.

Cine EF mount cameras
https://www.lensrentals.com/blog/20...-sensor-distance-article-part-1-cine-cameras/

DSLRS / mirrorless
https://www.lensrentals.com/blog/20...ensor-distance-article-part-ii-photo-cameras/

JB

 

[Björn Benckert]

Hi John,

Doesn’t the thickness of the OLPF stack of each camera actually affect the FFD? Yes, the PL Mount is nominally 52.00mm and lenses should be collimated that way, but the camera mounts may vary slightly from 52.00mm to account for this, no? It’s not like anyone is using a flange depth gage and steel plate in the gate on digital cameras, the mounts are checked optically with something like a Denz FFD tool. As long as all of the properly collimated lenses continue to tape out and zooms remain parfocal, I think it’s all good? Please correct me if I’m wrong.

Even on my film cameras, my Arriflex 3C is set to 51.98mm and my Moviecam to 51.99mm, so there was a bit of variation from camera to camera without any OLPF, just due to the different design.

The spec distance for pl flange distance is 52mm optically speaking. So yes the physical distance may stray from that as there is all kinds of glas elements between the flange and the actual sensor surface that captures the light the amount of flat glas add an offset to the 52mm. But the collimator tool is calibrated to 52mm and then used on the camera questioned and as the colimeter use light, tack sharp can be set to the right distance no matter if there is behind the lens nd’s, olpfs, and or sensor coverglas etc. in play. So, as there is a bunch of different glas in the cavity infront of the sensor that is different from camera to camera model/make. Its not really true to say that the offset between rf and pl is exactly equal to X as that X factor vary depending on the camera make and the amount of gak / glas it has in in front of the sensor... So yes one need to check backfocus also due to these reasons and that makes the physical distance of the pl standard a moving target. But nobody messure the physical distance on digital cameras, backfocus is calibrated using light.

 

[Satsuki Murashige]

Yes, film cameras vary. When I worked in prep the standard for Arri was 51.98, which allowed for the thickness of the film itself. Highspeed film versions like the SR2/3HS would be 51.975mm.

On the camera side I understand the 52mm depth is set with the OLPF / sensor cover glass accounted for. RED is unusual for some of their cameras having an adjustable sensor. The sensor cover glass / OLPF filter stack would certainly change the FFD, but when you set the cameras FFD on a digital camera you're taking that into account visually (You're not actually measuring the depth to the sensor) but it's "nominally" 52mm.

Just like on a film camera where each model has it's own actual setting. Also why shimming is an important function (or a moveable sensor)

JB

Thanks for the confirmation, John and Bjorn! This was my understanding as well.

 

[Kristi]

Good to have you here graham.

 

[GrahamClark]

Clearly you looked at all the competing products out there before building your own, so you must have a reason for building yours the way you did. Would you mind sharing your thought process?

We didn't have any PL RF mounts when we made the Breakthrough PL RF straight, just a simplified version of our PL RF drop-in filter mount.

When I sent PL RF straight mounts out to filmmakers to check out, I asked that in return they send me whichever units they were shooting with so I could borrow them to see how some of the other ones worked, since I've never seen any before.

So the video of all the various mounts, this was well after our mount had been sent to filmmakers.

There was no development project after we reviewed everything...

I bought the Hot Rod PL RF after you said it was an "easy bet".