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

[GrahamClark]

Hoping for some magical day when the variation copy to copy and tolerances from a plethora of manufacturers is unified and good enough not require vigilance and shimming seems to just be encouraging bad practice. Those problems will never STAY solved over time.

Marketing a product that solves a problem that can’t actually be solved whilst encouraging the teaching of bad practice doesn’t seem like the best path to be on.

the premise is simple, make it exact to spec, that's it

wouldn't a user adjustable backfocus in the mount eliminate a couple variables from the backfocus equation where shims normally come into play, in those instances of thermal expansion or otherwise?

 

[Jacek Zakowicz]

exactly, minus or plus the model specific deviation

Please do tell...

 

[John Brawley]

the premise is simple, make it exact to spec, that's it

wouldn't a user adjustable backfocus in the mount eliminate a couple variables from the backfocus equation where shims normally come into play, in those instances of thermal expansion or otherwise?

It creates more problems for the end user who now has to re-calibrate their lens everytime they change to a new PL Mount body.

And now they need to recalibrate all their focus scales for the Preston or WCU.

No. It doesn’t solve a problem. It creates more. Which is why no one does it in PL s35 plus size sensors.

Because you can’t accurately do this in the field to a standard thats good enough for the big screen. (Assuming that moat don’t carry pocket / field collimators)

JB

 

[John Brawley]

The Supremes would be the more interesting example as their release is tied to being the main competitor to the Arri Signatures and with that they initially were released shimmed for the Alexa LF and Mini LF. Zeiss is now changing this however to put them right on 52mm.

This doesn't add up for me.

Are you saying Alexa LF / Mini LF have different FFD's to other Alexa cameras? Other PL mount cameras? Other LPL mount cameras?

Isn't it tied to the mount that's on the front of the camera?

I mean sure if Alexa LF, Mini LF are shipping with an LPL mount, which is of course is an FFD of 44mm instead of PL's 52mm.

Last I checked Arri Signatures ONLY come in LPL but Supremes are PL / LPL / EF, but again, why wouldn't they ship set to the actual mount they are fitted with? PL mount Zeiss Supreme is 52mm, an LPL is 44mm. Naturally, this is normally QA'd and checked and re-shimmed if needed before delivery by any reputable dealer.

What you say doesn't make sense in this scenario. Are you saying Zeiss set to a standard that isn't the mount they are fitted with?

JB

 

[Robert Hofmeyr]

it's equivalent to a lens not being properly collimated

I still don't understand. I'm not trying to be difficult, I'm genuinely curious about this. I thought that it the FFD was a bit out it would mean the focus marks on the lens would be off, the lens might not focus to infinity, and zooms would not be parfocal. You seem to be saying that it makes the lens less sharp overall, and it becomes impossible to get perfect focus at any distance?

 

[John Brawley]

I still don't understand. I'm not trying to be difficult, I'm genuinely curious about this. I thought that it the FFD was a bit out it would mean the focus marks on the lens would be off, the lens might not focus to infinity, and zooms would not be parfocal. You seem to be saying that it makes the lens less sharp overall, and it becomes impossible to get perfect focus at any distance?

Exactly.

If the FFD is out you may or may not hit infinity.

You may not have accurate focus scales any more.

You might not be able to have a zoom hold focus though it’s range.

If the lens can be focussed, even with the FFD out, then it’s just as sharp as when the FFD is spot on.

The language being used is misleading. Showing images purporting to be side by sides is only showing you a side by side of two unadjusted unshimmed, unverified mount adaptors as they came from out of a box.

This is why I think this thread has caused so much angst. Basically there’s a lot is smoke and mirrors but this should actually be really simple.

I’m all for innovation and hooray for new players. As a DOP I love options. But the messaging here stinks.

JB

 

[Robert Hofmeyr]

Exactly.

If the FFD is out you may or may not hit infinity.

You may not have accurate focus scales any more.

You might not be able to have a zoom hold focus though it’s range.

If the lens can be focussed, even with the FFD out, then it’s just as sharp as when the FFD is spot on.

The language being used is misleading. Showing images purporting to be side by sides is only showing you a side by side of two unadjusted unshimmed, unverified mount adaptors as they came from out of a box.

This is why I think this thread has caused so much angst. Basically there’s a lot is smoke and mirrors but this should actually be really simple.

I’m all for innovation and hooray for new players. As a DOP I love options. But the messaging here stinks.

JB

Thanks John! That's what I've been trying to understand. So the soft image in the side-by-side images posted could have been sharp if it was focussed by eye (unless the FFD was so far out that it was unable to focus at that distance). That makes sense to me, but I was confused by the way it was described.

I just went back and checked and Graham did say right at the beginning: "Objective: shoot hard stop infinity focus between two mounts, compare which is more critically sharp." But the post by Tim Daust with side-by-sides didn't explain this, hence my confusion.

Now I have another question: if you set your focus to infinity, and something close to the camera is soft, as shown in the side-by-sides, does that necessarily mean your FFD is out?

 

[Phil Holland]

What you say doesn't make sense in this scenario.

Arri is changing what transpired as far as I know, but this isn't just an Arri thing at all. Best I can say about this for now. But I'll be pretty clear in saying that the different cameras in the world themselves are a bigger variable than one might think due to a great deal of things I've previously mentioned. Same cameras from the same manufacturer, depends on the manufacturer how constant their spec is. Some are either not aware or aren't acknowledging the impact of the optical chain behind the lens, pretty easy thing to test if you have the time, but it takes a lot of time. But even now between identical mounts there's subtle and annoying variations, and for several of the mirrorless mounts it's hard to know exactly what is "right" and i'm not just talking about the back focus here, but rather the entirety of the mount. I'm unsure why any of this is going on, but perhaps it's because the two major mirrorless mounts are so new that C and S are tinkering more and more.

Whole industry now is working much harder for spot on .0000+\- spec for all native and adapted mounts, 1st party I'm talking. Taking a look closely at official RF>EF mounts yields interesting variations. Subtle things noticeable to discerning image makers.

Out of the box or shimmed a 0.0002+\- variance is visible in image performance on all of these cameras. And even more highly visible on certain lenses depending on their design.

It's a weird thing going on in this thread in terms of some of the discussion, but BTF has uncovered exactly what I and many others have and they are making mounts to a tighter spec on a per camera basis, where the variables exist. There's the paper spec and the reality, and they are designing to the reality. I don't know if they will get to a day where custom Komodo, R5, C70, etc. adapters are made, but it is a nice effort to ship something a bit more on point in my opinion considering the market and times. This is something a bit unfamiliar in this segment to date as generally adapters are "universal" and you go from there. But I can say there is some consistency on a model by model basis at least that allows for this. RED, Sony, Canon, etc. Some of have subtly changed spec unbeknownst to many along the way, but when you hit that speedbump it's a big Scooby Doo style mystery while figuring it out.

You can absolutely shim whatever mount adapters you have too to get there as well. But as this all alludes to, if you have an RF to PL adapter and say a few of those cameras, each will have to be adjusted slightly differently on a per camera basis. People asked why I had so many E-Mount and RF adapters, this is why. I have them dedicated to camera bodies themselves. All 6 of the FX3s working down under right now have mounts shimmed per each body with little dots of paint on them to let you know which is which. That's a big film and they are C camera, but I get hired to figure weird stuff like that out.


Outside of the tit for tat starting point of this thread (which I'll chalk up to manufacturers having competitive banter really), the real question is do the people in this thread see a value to a per camera calibrated adapter or not? Because a lot of people have strong thoughts about the spec itself without discussing many variables that are at play. And I'll echo my previous point of a lot of people just take these cameras out of the box and shoot with them before doing any sort of calibration, man and the online small house rental world is a minefield when it comes to this. This is a constant reality of the sub $16K arena that is pretty common, if fact if I had to guess perhaps 70% + owners never tend to any of this in this arena. Considering that this is the entry level professional, prosumer, to mass market tier that's a lot of cameras out there. Wild west.


Just to illustrate why it's important to calibrate (out of the box or shimming or whatever). Hitting infinity focus, peak optical performance on image plane, scale accuracy, and lessening artifacts are pretty key things here. That said, a great deal of the entry level lenses out there don't even have accurate scales, which is a whole other bag of worms that I've been trying to sick the birds on. That's a topic for another thread and isn't even that important to "every type of shooter", though should be.

FYI, none of this is new. Theses variables have been going on well before I broke into the industry, even when the world was just film. The only thing that is new is we have WAY more cameras out there made by many different companies.

 

[J. Orlandi]

I disagree, just look at the results as more start to through as units are delivered.

Vocas PL is a great example of how a small amount of out of spec can produce a softer result.

what's wrong with getting critically sharp out of the box, thermal expansion and materials aside?

I see those as valid issues of course, but they only push the spec further out, to my mind.

Then you are completely oblivious as to what I’m saying I can see. The vocas mount I’m taking about is the one that uses the ridged 4 point connection for Komodo utilizing side mounts.

Furthermore the whole point is you keep saying the same thing over and over without any acknowledgment to being “in spec” doesn’t matter when the materials, forces, etc is a moving target. Just you simply calling something “sharper” when talking about these problems at hand show how little you understand optics and this subject for cinema use.

Place a 4lb prime lens on the end of your adapter with your tiny little bottom support. Now add a clip on matte box, you’re around 5lbs. A normal RED camera has no problem with this weight because of their design. However your adapter having only a small bottom support will not give properly support for the torque involved. 5lbs 8”+ forward of your mechanical connection is a huge amount of force that is pulling your adapter downward. Pulling the top of your mount out of spec regardless.

Are your adapters made out of titanium? Magnesium? Have you checked your elasticity of your materials? What’s their thermal spec?

My point still stands. You are chasing microns without solving ANY other problems that tosses those numbers out of the equation.

You keep toting you’re more in spec.... that doesn’t matter. What matters is to be able to put things IN spec. A Komodo being used for a year, will go out of spec. So you saying out of the box your mount is perfect. But that’s according to your hypothetical perfect world of a brand new Komodo with hardly any weight, 59 degree F temp, and babied.

This is just not how reality works. It’s why no one else is trying to chase these numbers, because they DO NOT matter when you take into account reality. You’re talking math. I’m talking physics. They are completely different and you aren’t factoring many variables that are important to cinematography.

Your materials, mount, and design have flaws that make it just as some of these other “Chinese” mounts you complain about. Bottom line is, the Komodo isn’t built with rigidity in its mount, rigidity in its sensor mount is TBD considering thermal and shock over time. Since the sensor cannot be adjusted, the adjustment needed will be at the PL mount adapter.

All you go on about is 1 problem you’re trying to “solve” yet you aren’t taking into account the other problems that have just as much if not MORE impact on the proper distances you’re trying to achieving.

You keep spewing the same excuses with very little explanation on your hypothesis. Bottom line is there are at least a few very nerdy professionals on here, all with long backgrounds in exactly this field who have already argued and brought up very valid points opposing your design and philosophy of use of your product, yet your only response to anybody has still been “but you’ll be in spec closer than the others”.

And in response to me talking about your vaporware, I’m speaking on behalf of the rest of the people who gave you money for a product and waited eight months with still no response from you. So yeah it’s caper ware until it’s arriving at peoples doorsteps. My ultimate point with that statement is that you’re flashing around a $10,000 PR stunt in front of people who were still waiting for a product for eight months. That was their money you’re flaunting around.... it all doesn’t look good, and judging by how many people have already come out opposing your designs and spec thought, it seems futile to continue this discussion with everyone without you changing how you respond.

You sound like a broken record while purposely skipping over any responses to your mechanical issues faced.

 

[BrendanLeahy]

either finish and ship RF EF drop-in or make DSMC2 and PL RF drop-in filter mounts and then ship RF EF drop-in mounts.

with RF EF drop-in mounts already existing by Canon, and DSMC2 / Komodo DFM did not exist, all the pressure was on to make something completely new, which I agreed with filmmakers about, in terms of priority.

now that we're shipping the DSMC2 / Komodo drop-in filter mounts we're already back to RF EF drop-in, which is how this all started.

and after having gone through DSMC2 / Komodo we have a filmmaker version, which if we had shipped the "photographer" unit now I think it wouldn't have bene the "right" unit for filmmakers.

Ah ok. I will still hang on to my order, but can I get the filters I ordered shipped to me now rather than waiting on them being bundled with the RF EF drop-in mount?

 

[Jacek Zakowicz]

I definitely see value in a properly calibrated adapter.
IMO here is the order of adjustments if there is an error
1. Camera base mount- if the RF mount is off the nominal 20mm with a tight tolerance than that should be adjusted first. Why? because that error throws out both the adapter height and the lens back focus. All other adapters and other lenses become unknown variables as well if the base mount is off. If the adapter and/or lens are adjusted to incorrect base mount they will most likely have to be adjusted -again- to another camera body.
2. Lens back focus . Why? because if the lens back focus is off the lens may become difficult to use or unusable with other cameras. It makes sense to have the lens set right.
3. Adapter Why? This is the last component to adjust because it should be set to a fixed height mechanically and left there. The number is very precisely defined -32mm (glass-less) in case of RF to PL with a tight tolerance should always suffice with correctly set base mount flange and lens back focus. unfortunately, erroneously- this is the cheapest and easiest component to adjust so, naturally- it is tackled fist while it should be the very last IMO.
Logically in my opinion this order makes most sense and will comply with the lens/camera interchangeability premise- a staple of the rental facilities excellence.

 

[Vance Colvig]

Just like it was done in the good old film days.

 

[Robert Hofmeyr]

AOutside of the tit for tat starting point of this thread (which I'll chalk up to manufacturers having competitive banter really), the real question is do the people in this thread see a value to a per camera calibrated adapter or not?

While I am a bit out of my depth in this thread, I can answer this one: I do think that an adapter shipped at least close to calibrated for my camera is a really good thing for a person like me. I would not be using it on multiple cameras and I'm unlikely to travel somewhere and get it calibrated perfectly. If it works, I'm happy.

 

[Björn Benckert]

I definitely see value in a properly calibrated adapter.
IMO here is the order of adjustments if there is an error
1. Camera base mount- if the RF mount is off the nominal 20mm with a tight tolerance than that should be adjusted first. Why? because that error throws out both the adapter height and the lens back focus. All other adapters and other lenses become unknown variables as well if the base mount is off. If the adapter and/or lens are adjusted to incorrect base mount they will most likely have to be adjusted -again- to another camera body.
2. Lens back focus . Why? because if the lens back focus is off the lens may become difficult to use or unusable with other cameras. It makes sense to have the lens set right.
3. Adapter Why? This is the last component to adjust because it should be set to a fixed height mechanically and left there. The number is very precisely defined -32mm (glass-less) in case of RF to PL with a tight tolerance should always suffice with correctly set base mount flange and lens back focus. unfortunately, erroneously- this is the cheapest and easiest component to adjust so, naturally- it is tackled fist while it should be the very last IMO.
Logically in my opinion this order makes most sense and will comply with the lens/camera interchangeability premise- a staple of the rental facilities excellence.

I don't know if the RF mount on the Komodo has removable chims if it's to far out or what chims to use to bring it further out. But of course yes it should be checked and corrected. But I doubt RF colimation is something every rental house can quickly sort. most of them got instruments to colimate a PL mount correctly, very few will tackle a RF mount with ease. You can use an adapter on your pl colimation tool to check the RF mount, but then you already started to tamper on accuracy.

The wooden camera guys shows how to colimate their PL mount: https://www.youtube.com/watch?v=fsMHMbMzEOA

 

[Jacek Zakowicz]

I'm just saying how it should be handled to keep the lens/camera interchangeability real.

 

[John Brawley]

This is one of the better FFD measuring tools. And yes they have an RF mount.

https://www.denz-precision.com/en/cinetec/fdc-multi/

What's glossed over in the wooden camera video is that with a simple collimation tool like they show, it doesn't really tell you how much adjustment is needed. You're then doing trial and error with shims to try and improve it, having to take the mount apart and swap shims. Even the way you tighten the screws back onto the mount can change the way a lens seats and collimates.

The Denz does tell you the offset and is likely the kind of device RED is using to factory set the RF mount. Note that it's working on a virtual image so it accounts for the sensor stack / sensor cover glass / OLPF.

I got the impression though that Komodo isn't user adjustable. So if it is out or goes out over time, then you have to adjust through the lens adaptor (preferred) or the lens itself (not so preferred)

JB

 

[Jacek Zakowicz]

this way the camera/lens relation is not maintained. Native rf lenses will not work with that Komodo body- at least not correctly. Wouldn't it be a better solution to correct what is actually wrong?
What difference does it technically make if you correct PL, RF, EF, PV, etc..?

 

[GrahamClark]

Graham what does that actually mean ?

JB

if a lens is not properly collimated your not hitting your marks, like the lens, the mount too

 

[GrahamClark]

Please do tell...

Canon R and R5 have a 3 thou (.003") deviation from each other, that's a standard shim

even the locking pins diameters are different

this jumping around may seem erratic to filmmakers, but I don't see how RF play or flange depth has any negative consequence for photography cameras, so I also wouldn't consider it defective given how it seems like conscious design decisions were made

solution: a shim that says "R5" and "R" on it, rather than numbers so you don't have to do math, problem solved!

 

[Jacek Zakowicz]

Canon R and R5 have a 3 thou (.003") deviation from each other, that's a standard shim

even the locking pins diameters are different

this jumping around may seem erratic to filmmakers, but I don't see how RF play or flange depth has any negative consequence for photography cameras, so I also wouldn't consider it defective given how it seems like conscious design decisions were made

solution: a shim that says "R5" and "R" on it, rather than numbers so you don't have to do math, problem solved!

How do you know if they have 1, 2, 3 thousands and which direction? It is completely random. Making a single shim and marking it "R" makes no sense...