Personal View site logo
SLR Magic 2x ANAMORPHIC lens
  • 804 Replies sorted by
  • 1.35x or 1.33x adapter with 77mm thread is optimum for me.

    Also 1.35x m4/3 lenses are REALLY interesting.

    I am not interested on the other aspect ratios at all. Only 1.35x producing 2.40:1 with 16x9 recording.

    (I would much rather crop top and bottom if i wanted any wider than 2.40:1)

    Also closeup lenses (+0,5, +1, +2) with antireflective coating in 95mm or 105mm diameter would be usefull. They seem to be hard if not impossible to find.

  • @jackdoerner and everyone else.

    We were not interested to make a 2X lens before for a few reasons.

    1) 1.33X lenses are expensive and we want to make it affordable 2) 1.5X is already far from our original consideration. 2.66:1 ration however is still acceptable 3) 3.55:1 ratio is a joke and we DID consider it silly at one point

    However, after all we have read we are convinced 1.35X for cropping convenience is not the point since we want to go for the anamorphic look. 1.35X has much better IQ than 1.5X and 2X but sometimes we go for feel an not the best possible IQ. Anamorphic was meant to be 2X to give that extreme look. The new 16:9 format does not translate well after 2X but regardless the recording format is 1:1 or 4:3 or 3:2, or 16:9 anything less than 2X will be less than the original anamorphic look. As so many effects such as horizontal flare, oval bokeh, 2.35:1 can be done in post the only thing that cannot be done in post is the dramatic anamorphic look. If we need to crop 1.5X then it is not much of an extra job compared to cropping a 2X squeeze footage. We shoot anamorphic for a reason and post production is worth it if we want that certain look so dealing with 3.55:1 is worth the extra work. If we want convenience with 1.33X on 16:9 it is not much different from using a 0.7x wide angle adapter.

    It seems 1.33X should be out of the question for the reasons above and 2X should be the way to go. Anyone on the same page as I am or we still in disagreement?

    Kind rgds., Andrew

  • @LPowell I know 1920x1080 is barely worse than 2K, but I have never seen the point of 2K cinema anyway, nor even of IMAX 4K (which is miles behind true IMAX in image quality). The consensus seems to be that 4K can just barely match the quality of 35mm film, and in my experience 2K doesn't come close. I am disappointed that so many cinemas have decided to project 2K, because to make 2K look decent, the image has to be relatively small.

    I also understand that a wider aspect will mean downscaling or cropping for release - but's a compromise I find acceptable, and I don't think it will be necessary for long. Computer monitors have been larger than 1920x1080 for years, and all the manufacturers keep teasing 4K televisions to match the 4K cameras that have so recently become common. I myself have a 4K monitor (or nearly - it's an IBM T221, last built in 2005. You can get them for $600 or so on ebay right now), and I can say from experience that the difference in quality is enormous. 2880x1080 adapts much better to 4K than does 1920x800 and, while I have no scientific basis for the claim, I would expect it adapts better than does 2554 x 1080 (i.e. 1920x1080 with a 1.33x stretch), because more of the enlargement is achieved optically.

  • @slrmagic Here's my ideal widescreen workflow compatible with both 2K DCI and Blu-ray release formats:

    1. Shoot in 1080p24 mode with a 1.5x anamorphic.

    2. In post, stretch horizontally out to a 2048 frame width (DCI spec).

    3. Squeeze vertically down to an 810 frame height (both DCI and Blu-ray).

    4. Letterbox out to a 2048x1080 frame size (DCI spec).

    5. Crop to a 1920x1080 frame size (Blu-ray spec).

    This workflow preserves the full resolution of the 2K master without losing any image quality or screen height in the Blu-ray release. With attention to framing, the slight horizontal crop of the Blu-ray release is negligible.

  • @slrmagic

    If you crop to 2.4:1 from 3.55:1 on a 1920x1080 camera with a 2x stretch, you are effectively wasting 620 pixels of horizontal resolution on your camera's sensor, or around 33% of its total area.

    If you crop to 2.4:1 from 2.66:1 on a 1920x1080 camera with a 1.5x stretch, you are wasting 200 horizontal pixels, or approximately 10% of the sensor's total area.

    If you shoot 1920x1080 with a 1.33x stretch, you waste nothing, but the artistic aspects of anamorphic lenses are lost.

    For reference, cropping to 2.4:1 from 1920x1080 with no stretch (i.e. just cropping a normal shot) wastes 26% of the sensor's total area. As you can see, you actually lose more quality by cropping from a 2x lens than you do by cropping from no anamorphic at all.

    I find the wastage involved in shooting a 2x stretch to be too much, and, while I want the highest possible quality, I also value the 'anamorphic look'. For that reason, I find 1.5x to be the best compromise. It retains the pleasing artifacts of an anamorphic lens to a reasonable degree, produces a valid native format on a standard S35 sensor, and crops to 2.4:1 with much less loss of quality than would a 2x stretch

  • In some of the emails I received, the 2x and 1.5x are more desirable because the anamorphic stretch mode is an available option on most monitors. 2x is more desirable than 1.5x because it is what anamorphic should be like and it is in-camera native option for 2x in Red Epic and Scarlet and most other cameras supporting anamorphic setting.

    As for wastage vs anamorphic feel it is a subjective matter. It is just like saying the sports car waste gas even when cruising but the point of it is not the wastage part. Or the 0.87 Carat diamond is just the same as 1 Carat but to some 1.03 Carat makes more sense as 1 Carat than 0.87 even when that extra bit cost a whole lot more.

    Not my opinions. Just emails I received.

    Kind rgds., Andrew

  • @slrmagic

    For me resolution, IQ and framing are more important than flares etc.

    So 1.35x is the optimum but I could possibly go with 1.5x at the widest if that will be the way it has to be. But I will not consider anything wider at all. So 2x is a no no for me.

  • @slrmagic, I started reading this thread with strong opposition to X2 squeeze. That was with the consideration to 16x9 being the most common format for HD these days. Without cropping this combination creates 3.55:1 which is unusually too wide for any taste. On the other hand X1.3 squeeze doesn't provide a significant anamorphic bokeh and the rest of characteristics of anamorphic visuals. At this point I think an acceptable compromise would be the X1.5 squeeze. We will enjoy a moderate anamorphic visuals and the 2.66:1 which is beautiful format. Also 2.66:1 require less cropping to get to the academy standards. For the FOV I think we all tend to be attracted to as wide as possible. So on an average s35 imager it would be great if we get a 17mm-28mm adapter. The close focus could be enhanced with diopter doublet achromats.

  • @slrmagic In an earlier post you targeted a $1200+ retail price for a 1.5x or 2x anamorphic adapter. I'd definitely pay that kind of money for an alternative to a 1.5x ISCO anamorphic. But for a 2x adapter, I wouldn't even consider a price over $1000, there are plenty of legacy 2x adapters available at less than half that price!

  • @slrmagic

    In reply to your emailer (I understand that their opinion is not yours), wastage is not a matter of opinion, it's a matter of math. Different amounts of wastage may be acceptable to different people and for different applications, but it can be measured and objectively quantified, and (all other things being equal) less wastage produces objectively higher image quality, which is important in many applications. I will admit, however, that sacrifices in absolute image quality are sometimes desirable for the sake of art. Only heartless pixel peepers and Zeiss engineers would deny that.

  • anamorphics have always been messy always been a niche. always been used by the more supple the less lazy the more adventurous directors and cinematographers. anamorphic burnt into the minds eye via french and hollywood cinema has always been dominated by 2x compression. that unique combination of factors. everything else is niche within a niche. 2x may not be the best squeeze resolution wise for sterile digical but the look trumps everything. the greatest names in anamorphic design worked hardest on 2x compression. the greatest cinematographers did the best work in 2x yes a good reason existed but that is the look burnt into memory. 2x is the look a bad director can use to make himself look better than he actually is.

  • I'll chime in one more time here, but I feel I'm beating a dead horse. 1.33x lenses are glorified aspect ratio changers, not anamorphic lenses created with the intention of simulating cinema anamorphics.

    To those who keep bashing 2x lenses for their "resolution loss", I ask, have you ever used a 2x lens the way it was intended (that is, maintaining 1080p vertical resolution and stretching it out horizontally)? It looks gorgeous, even when cropped to 2560x1080. The retention of 1080 vertical lines brilliantly hides the loss of horizontal resolution. But if you're squeezing your footage down to 1920x810, yes. It will look lower res. (I've even done experiments upscaling anamorphic GH2 footage to 4K and then applying a 4K grain overlay effect. It holds up. If there's one thing the GH2 couldn't have much more of, its resolving power.)

    @slrmagic If you're going to make an anamorphic, it ought to have at minimum a 1.5x squeeze factor, but something closer to 2x would be better.

    If you want to make a 1.33x, that's fine, but I have a suspicion that unless it has some form of oval aperture, people who are after an "anamorphic look" won't be particularly interested. You'd instead be catering to those who want to change aspect ratios without loosing any resolution (which is a tricky thing to try and do. 1.33x Anamorphics aren't exactly known for their ability to resolve detail.)

    P.S. in order of importance from most to least for me personally: 1. Out of focus stretch 2. Usability and sharpness at wide apertures 3. Wider than 35mm 4. Flares 5. Mount other than m4/3 and APS-C image circle. (I love GH2, but I won't be using it forever. EF-s lenses are adaptable to lots of stuff, M4/3 lenses are adaptable to almost nothing else. Be wise and widen your market with a longer FFD lens mount. GH2 users will just have to gasp use an adapter.)

  • assuming BMD smartens up and releases a MFT mount for BMCC sometime in the next two years, I think having a MFT mount is fine. You'll definitely have an audience between gh2/3 users and BMCC users.

    That being said, I really think 1.5x is the best compromise after reading the entirety of this thread. you can add letter-boxing in post without too much wasted space of the sensor, same thing goes if you want to widen it by punching in 5%.

  • From what we have read here it seems we have decided 2x is good for 4:3 and 1.5x is good for 16:9 in terms of look and not loosing too much information/pixels

    A our primary goal is to provide a solution for 16:9 HD shooting we need to choose a squeeze 1.5X or more to achieve the anamorphic look. As 1.5X and 2X are standard for most HDMI LCD screens it makes sense we do not choose a format in-between. 2X with a crop is good on a 60" LED/Plasma screen viewing but many feel it is not good for large screen cinema viewing for those with a purpose of putting their work for cinema display and 1.5X is the best compromise between IQ and Anamorphic feeling.

    We will see what we can do about it =D

    Thanks all for the suggestions!

    Kind rgds., Andrew

  • I think LPowell has a very good point with 2x adapters being available relatively cheaply.

    @slrmagic What are your thoughts on the idea of adapter + dedicated lens lineup?

  • I think 2x will hurt IQ when shotting 16:9 and crop for 2.4:1 Main reason people choose GH2 over other cameras is the IQ, so a large crop will decrease resolution and increase softness so much.

    This is why i vote for 1.33x, but if the anamorphic oval bokeh look is a must so the 1.5x is the best option between crop amount/anamorphic look. Maybe the 1.5x can Keep a good IQ and a good aproach to the oval defocus when shooting 16:9 to get from 2.35:1 to 2.4:1

  • Maybe you can develop a mix design with 1.33x and an oval shape exit pupil or oval shape middle element to force the camera to see the oval bokeh. i do not know if it is possible but i got this idea. I think if we keeping speculating about it we will not find if it is possible or not. I think is time to start some real word experimentations to find the best balanced squeze factor and if possible combine it with an oval shape pupil or diafragm. its time to move from the brainstorm to the real world experimentations. Maybe it will need to design an build some different anamorphic adapters and test them before to decide the one to put in market. It will take some money and time but it is important to the success of the product.

  • the following link shows a filter using the oval shape pupil, i know this is not what anamorph hungry people wants but my idea is to mix the 1.33x with some oval shape pupil in the adapter to get the oval look and the best IQ, take a look:

    http://www.vid-atlantic.com/cinemorphic.html

  • assuming BMD smartens up and releases a MFT mount for BMCC sometime in the next two years, I think having a MFT mount is fine

    It looks like thay had been so afraid that released it today :-)

  • A simple and inexpensive test it to cut some oval pupils in different sizes and put them between the LA7200 and the lens and see what happens. the oval pupil filters can be made in different pupil sizes and can be a removable acessory for the anamorphic adapter.

  • @apefos The oval shape will give you oval bokeh but would not give you jaw dropping wide aspect of an anamorphic.

    I am again getting many emails to do a 77mm 2x Anamorphic. I passed the msgs to the board and they agree with @LPowell and @RRRR. 2X anamorphics can be picked up for $300 on ebay. We wanted to make a 1.33X in the first place because they give the right aspect ratio without cropping and they are not really affordable considering they are not user friendly below f/5.6.

    Can someone convince us we should make a 2X considering used ones are sold at $300 on ebay for about a quarter of the price we plan to make ours? When the board looked on eBay and looked on @LPowell's point on 2X they said this project is not feasible at all =S

    Kind rgds., Andrew

  • @christianhubbard @Vitaliy_Kiselev

    With the 2.3 digital multiplier ratio it seems our 12mm T1.6 lens is a good offering for that system as it would give 28mm (35mm Equiv) field of view.

    If we make some lenses considering the BMC as they mentioned it has a passive mFT mount only should we go wider or longer? Is there anything missing in the range? When we asking about the 12mm T1.6 before people said do not make a 17.5mm because there is already a 20mm f/1.7 lens and no one would buy our lens. We have 12mm covered and it seems there is a 11-16mm and 17.5mm already. Is there still something missing in the range? @Vitaliy_Kiselev can we open a new topic for that?

  • @slrmagic 10.5 would make for a ca20mm equivalent lens for gh2 and ca24mm equivalent for bmc cam. Both are pretty sought after. Distortion control would be very important. But unlike with the 12mm f1.6 I don´t think you should sacrifice that much at infinity focus.. The performance between 1 feet and 2 metres of the 12mm is really nice, don´t get me wrong but it becomes a bit strange when infinity focus is hardly useable (my subjective opinion).

    Also, consider making it a more expensive lens / improve the construction / optics as much as you can. The market you´d cater to that uses the bmd can afford a more expensive lens if the optics and usability of it is better.

    RE: squeeze factor I thought you were into 1.5x since our discussion here? I retirate, please consider anamorphic adapter + dedicated lens solution. Aim at equivalents that are made for s35 sensors; 35, 50, 75 - which would translate to roughly 20, 30, 40mm for gh2 sized sensor and start with the widest.

  • Sorry, what I was talking about when referring to an oval iris is THE aperture . . . the adjustable aperture within the lens, not some ghetto-rigged oval cutout.

    I see no logical reason why it would not be possible to make a functioning oval iris. It should be perfectly possible (if not as a fluid aperture, then certainly the blades could be machined in such a way as to be oval at each stop).

    @slrmagic I don't have CAD software, or i'd try to design one myself as example of feasibility. If you've got CAD software and a smart fellow, set him to the task for an afternoon and see what he comes up with. Better yet, hand it to an intern ;)

  • @slrmagic 25mm focal length 1.5x anamorphic lens would be a good starting point.

    Many people use 40mm or 50mm focal length lens with anamorphic adapter on Gh1/Gh2. 25mm would be a heavenly match.

Start New Topic

Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!

Sign In with Google Sign In with OpenID

Sign In Register as New User

Tags in Topic

Top Posters