Announcing the S.A.L.T. (Southland Alternative Lens Test)

[Phil Jordan]

FYI there is no such a thing as a "4K lens"
There are Academy lenses that cover, well, Academy format(27.16mm circle) and those will clip(100% vignetting) 32mm so they are not usable and 35mm full aperture(motion picture) lenses that cover 31.1mm diagonal-in reality 32mm and these would not have a problem since the coverage is so close.
If you are trying to use academy lenses for full (5k) you'll get poor results- it's pretty obvious...
All the newer lenses cover full aperture because Academy format is pretty much dead since they started using tape recorders for sound- and it's been a while...
Jacek Zakowicz, optitek-dot-org

Of course there is no 4K lens except in the context of 4K coverage on a R1 which is what I was referring to and some of the newer lenses do cover 31-32mm but are completely soft out at the edges which is why we want to know what these lenses will do on a larger image circle. Just the fact that they cover isn't enough information.

[Tim Whitcomb]

Lets not forget there are more lenses coming from RED in the next 8 Quarters or so.

Mini Primes, Electronic FF lenses, the RPP's are just the beginning of their continued commitment to providing the best DIGITAL CINEMA gear on the market.

Easier said than done, but the best part is RED really cares about getting it right.

But you seem knowledgeable, and as such, don't you agree that Lens choice is as much about "aesthetic" as much as its about lens projectors, collimators, edge sharpeness, etc? Perhaps that's why so many on here are growing impatient regarding the SALT tests... but for me, Mark P shot a great little short and the lenses looked killer to me. I liked the Bokeh, tones, color rendition and yes, the sharpness. But wtf do I know.

That said, what IS your point?

[Brook Willard]

With all this 32mm talk, I really hope that the Epic and Scarlet have a "4K" mode where we can shoot standard super 35 instead of RED super 35. The 30mm x 15mm sensor seems so arbitrary. It'll make focal lengths result in abnormal/unexpected field of view, it'll potentially increase the burden on post [though this is less of an issue with the Rocket on the horizon], etc. It's an awesome marketing point - 5K is a great number, bigger sensor is more badass, etc... but the sensor itself isn't even big enough to do proper 4-perf work for standard 2:1 anamorphic lenses.

Don't misinterpret this, I'm very excited and I don't want it to come across as an attack... a larger image circle and larger [but not quite large enough to make the difference I was hoping for] sensor wasn't near the top of my list when I dreamt of the "RED TWO". And if there was a dimension that we wanted bigger, it would be the vertical dimension... to make it big enough to cover 4-perf. And that size, unfortunately, won't be achieved by the S35 sensor.

So it's cool that the RPPs cover the new sensor, but I personally don't think people should concern themselves with it that much.

As Matt said, the difference will come when the 8-perf cameras ship. Until then... stick with [standard] super 35... particularly since the S35 model cameras are the only cameras where the 30mm x 15mm sensor size will ever be important [the future may prove me wrong, but as of today... that's it]. The sensor will be big enough to cover two of RED's cameras, sure... but RED has 9 cameras on the drawing board.

[Kyle Presley]

With all this 32mm talk, I really hope that the Epic and Scarlet have a "4K" mode where we can shoot standard super 35 instead of RED super 35. The 30mm x 15mm sensor seems so arbitrary. It'll make focal lengths result in abnormal/unexpected field of view, it'll potentially increase the burden on post [though this is less of an issue with the Rocket on the horizon], etc. It's an awesome marketing point - 5K is a great number, bigger sensor is more badass, etc... but the sensor itself isn't even big enough to do proper 4-perf work for standard 2:1 anamorphic lenses.

Don't misinterpret this, I'm very excited and I don't want it to come across as an attack... a larger image circle and larger [but not quite large enough to make the difference I was hoping for] sensor wasn't near the top of my list when I dreamt of the "RED TWO". And if there was a dimension that we wanted bigger, it would be the vertical dimension... to make it big enough to cover 4-perf. And that size, unfortunately, won't be achieved by the S35 sensor.

So it's cool that the RPPs cover the new sensor, but I personally don't think people should concern themselves with it that much.

As Matt said, the difference will come when the 8-perf cameras ship. Until then... stick with [standard] super 35... particularly since the S35 model cameras are the only cameras where the 30mm x 15mm sensor size will ever be important [the future may prove me wrong, but as of today... that's it]. The sensor will be big enough to cover two of RED's cameras, sure... but RED has 9 cameras on the drawing board.

Custom aspect ratios come as a listed spec for several of the models. It should be very easy to get 4 perf on the FF models.

[Steve Sherrick]

All I can say is that after shooting with the RPPs today, I am sold on them in a big way. Just spent the past few weeks with S4s, also very nice and capable of producing beautiful images, but the RPPs just seem like the perfect match for Red One. I'm in the process of posting a shot to Vimeo, should be up soon.

Nice work on these lenses RED!

Link to Video: http://vimeo.com/5434694

More can be found here: www.redfilmmaker.com

[Mark Pedersen]

This is an excertp from my SALT article:

My opinion on corner performance…
A lot has been made of lens corner performance recently in light of the newer Red sensor being larger and requiring a larger image circle. Although I think this is an important consideration to be taken into account, I would also state that only very infrequently do objects appear at the very corners of the cinema frame in a fashion that requires high detail representation. In fact many times the “shape” a particular lens adds to the image is desirable to subtly focus the viewers attention in the frame. This is of course a personal preference but it’s important to note that all the lenses tested here out performed the Red sensor in the center easily.

If you consider how our eyes work, they contain WAY more cones in the center and WAY fewer on the periphery. The Fovea is where we focus and direct our eyes to read detail. Much denser cones (higher resolution) in that tinyl area of our retinas. Try keeping your eyes fixed on a word on this page, and then pay attention to how much detail you see in the corners of your web browser without looking directly at it. You can barely resolve it.

Our brains are hard-wired to focus on only part of the scene in front of us. Our brain directs our lenses to what it WANTS to see and focus on. Too much detail would confuse the shit out of you because you wouldn't be able to discern what is important and what is not!

While you can't draw a direct parallel from this to cinematic images, the principles are the same. A good DP and Director direct your eye to the area of the fame they WANT you to focus on and pay attention too. That's what DOF is all about, composition, selective lighting, etc.... Then of course there is motion blurring of the image.

While lens corner resolution and consistent illumination are important if you are doing FX plates and lockoffs on deep DOF high contrast scenes, or to Jarred's point, you are doing DSMC work, for the most part it isn't critical (IMHO). In fact, for most narrative feature work, it can even be considered desirable for the reasons Evin & I mentioned.

How many of you have done "power windows" when grading. Hello??

What I find interesting is that there seems to be little discussion on the other lens properties that really matter: consistency of color matching, contrast, flaring, build quality, serviceability, weight, consistent barrel lengths for quick lens changes with a swing away MB, etc.. I hope the SALT covers that.

A set of lenses should be considered in a holistic way... a system that provides consistent performance across the full range of focal lengths, works well in all production environments (sticks, handheld, dolly, Steadicam, etc..) and holds up to the rigors of production over years of use.

There's a lot to be considered. And of course price/performance is a big one.

That's why we're all anxious to see the results.

Oh, and I have to agree. If MPs aren't part of the test—the one benchmark we all look to—it is missing its foundation.

M

[Rick Darge]

Great post Mark.

If you consider how our eyes work, they contain WAY more cones in the center and WAY fewer on the periphery. The Fovea is where we focus and direct our eyes to read detail. Much denser cones (higher resolution) in that tinyl area of our retinas. Try keeping your eyes fixed on a word on this page, and then pay attention to how much detail you see in the corners of your web browser without looking directly at it. You can barely resolve it.

Our brains are hard-wired to focus on only part of the scene in front of us. Our brain directs our lenses to what it WANTS to see and focus on. Too much detail would confuse the shit out of you because you wouldn't be able to discern what is important and what is not!

While you can't draw a direct parallel from this to cinematic images, the principles are the same. A good DP and Director direct your eye to the area of the fame they WANT you to focus on and pay attention too. That's what DOF is all about, composition, selective lighting, etc.... Then of course there is motion blurring of the image.

While lens corner resolution and consistent illumination are important if you are doing FX plates and lockoffs on deep DOF high contrast scenes, or to Jarred's point, you are doing DSMC work, for the most part it isn't critical (IMHO). In fact, for most narrative feature work, it can even be considered desirable for the reasons Evin & I mentioned.

How many of you have done "power windows" when grading. Hello??

What I find interesting is that there seems to be little discussion on the other lens properties that really matter: consistency of color matching, contrast, flaring, build quality, serviceability, weight, consistent barrel lengths for quick lens changes with a swing away MB, etc.. I hope the SALT covers that.

A set of lenses should be considered in a holistic way... a system that provides consistent performance across the full range of focal lengths, works well in all production environments (sticks, handheld, dolly, Steadicam, etc..) and holds up to the rigors of production over years of use.

There's a lot to be considered. And of course price/performance is a big one.

That's why we're all anxious to see the results.

Oh, and I have to agree. If MPs aren't part of the test—the one benchmark we all look to—it is missing its foundation.

M

[Pawel Achtel]

All good points, Mark.

One thing I wanted to add that I am more critical of corner performance in wide lenses than normal and longer lenses. Wide lenses are often used for establishing shots and other scenery shots where corner sharpness is often desirable.

[Rodney James]

I am new here so I do not mean disrespect but I am so perplexed by how people continue to argue about bad things being good. People arguing that lower resolution on edges is actually better than higher resolution? Come on guys, really?

Mark's post about the eye only focusing on the center of the screen is a great biology lesson, but there is one thing you forgot. Your screen is not blurry on the edges, so if your eyes move up into the corner you damn well expect it to be just as sharp as it is when you were looking at the center. Everyone reading this would return their screens to apple if it wasn't. I expect the same from my images. I want to CHOOSE to have the edges blurry if I want to, just like I want to choose to have the edges vignette if I want to.

We are dealing with high resolution imaging here folks. Resolution and Sharpness matters on the entire image. There is always a subjective factor to everything, but let us not start calling compromises artistic expression.

[Matthew Duclos]

Wow this thread went off on a tangent.. I make a living trying to optimize lenses to gain the absolute best image quality throughout the frame, including the edges. Maybe Im reading all of these posts wrong, but in my line of work, sharpness is everything. Sharpness is a blank canvas. If everyone is looking for crappy image quality in the edge of their frame so that it's more "natural" I may very well be out of a job.