Jump to content




Photo

Matching a Zeiss Superspeed in quick pick ups?


  • Please log in to reply
15 replies to this topic

#1 Rod Lewis

Rod Lewis
  • Basic Members
  • PipPip
  • 30 posts
  • Director
  • Bournemouth

Posted 19 May 2015 - 09:50 PM

Hi!

 

Thanks for reading!

 

First of all I am a director and not a D.O.P, so go easy on me please?!

 

I am trying to grab some quick pick ups in Canon 5DMKIII Magic Lantern RAW for a shoot we did using a PL mount C300 and Zeiss Superspeeds.

 

My question:

 

Using the 5DMKIII now, as opposed to the C300, can anyone advise me on the best choice of lens that might match the Zeiss Superspeeds most closely?

 

It seems to me that it's a toss up between the Canon L (50mm or 85mm f1.2 etc), or the Zeiss Distagon EF Mount f1.4. I'm not overly concerned about the .2 of a stop - more important is getting a look consistent with the rest of the footage in terms of Bokeh and overall tonality.

 

Obviously the idea of shooting in 5D RAW is total flexibility in post, in terms of matching the footage. (I am aware that I am opening a whole can of worms in even attempting 5D RAW and Magic lantern, but I am going to face it head on!)

 

It's possible I'm being too anal about this, I don't know. I've already shot sucker mount footage on my 5D in neutral shooting mode with the contrast -4 and the saturation -2 and it looks kind of ok on my editing monitor, (much better than when I tried it with my Canon 24-70 L), but I just don't want to compromise, as I have put so much hard work into this film and I want to enter it into all the festivals I can.

 

My D.O.P is away on another job, so I can't wait around. I just want to get it done, and I own a Canon 5D MKIII.

 

Please, any advice would be very much appreciated. I had another thread here, where people chimed in, and it ended in happiness and joy (for me anyway), so maybe we can achieve the same?

 

http://www.cinematog...showtopic=67371

 

I shot the footage with the Zeiss f1.2 50mm which I hired for the weekend. All I need now (I think) is to add 5D RAW to the equation and shoot the footage again, as well as the quick picks ups that actually have actor's faces in.

 

Or do I?

 

Experts please tell me!

 

Much love

 

Rod

 

 

 

 

 

 


  • 0




#2 Stuart Brereton

Stuart Brereton
  • Basic Members
  • PipPipPipPip
  • 2575 posts
  • Cinematographer
  • Los Angeles

Posted 19 May 2015 - 11:25 PM

What f-stop was your original footage shot at? Super speeds go soft and low contrast when wide open. If you weren't wide open then the differences between Super speeds and L series or EF series are likely more to do with color than sharpness, and are easily correctible in post. Remember though, you may well see some difference between the 5D and c300. They have similar colorimetry, but the 5D achieves it's image by a number of tricks that the C300 does not. This difference may obvious or virtually unnoticeable, depending on your shot


  • 0

#3 Satsuki Murashige

Satsuki Murashige
  • Sustaining Members
  • 3081 posts
  • Cinematographer
  • San Francisco, CA

Posted 20 May 2015 - 12:29 AM

Contax Zeiss primes from the 80's would be the most similar optically to Zeiss Super Speeds. That said I don't think the particular brand of lens will make that big of a difference. Keep similar distance subject to lens, keep similar lighting ratios and color temperature. That will help a lot more.

To match depth of field of full frame 5D footage to the smaller sensor of the C300, you should stop down an extra 2 stops, so if you shot T2.8 on the C300 with Super Speeds, then shoot f/5.6 on the 5D. To the keep the field of view similar, use a 35mm on the 5D if you would have used 25mm on C300, 50mm for a 35mm, 85mm for a 50mm, etc.
  • 0

#4 Rod Lewis

Rod Lewis
  • Basic Members
  • PipPip
  • 30 posts
  • Director
  • Bournemouth

Posted 20 May 2015 - 09:09 AM

What f-stop was your original footage shot at? Super speeds go soft and low contrast when wide open. If you weren't wide open then the differences between Super speeds and L series or EF series are likely more to do with color than sharpness, and are easily correctible in post. Remember though, you may well see some difference between the 5D and c300. They have similar colorimetry, but the 5D achieves it's image by a number of tricks that the C300 does not. This difference may obvious or virtually unnoticeable, depending on your shot

 

hey thanks. A lot of the shots were right open at f1.2, but for the purposes of these pick ups, most will actually be closed down a bit because it's car bonnet stuff and I'm trying to keep everything in focus as much as possible. Obviously the light is still low though. Some tests with a Zeiss 1.4 50mm the other night seemed to show that f2.0 at iso 1600 was ok. There are one or two shots though that I will need to try and match the Superspeeds wide open at f1.2 though..

 

Hoping the differences between the 5D and C300 can be solved with magic lantern RAW.....


Edited by Rod Lewis, 20 May 2015 - 09:09 AM.

  • 0

#5 Rod Lewis

Rod Lewis
  • Basic Members
  • PipPip
  • 30 posts
  • Director
  • Bournemouth

Posted 20 May 2015 - 09:27 AM

Contax Zeiss primes from the 80's would be the most similar optically to Zeiss Super Speeds. That said I don't think the particular brand of lens will make that big of a difference. Keep similar distance subject to lens, keep similar lighting ratios and color temperature. That will help a lot more.

To match depth of field of full frame 5D footage to the smaller sensor of the C300, you should stop down an extra 2 stops, so if you shot T2.8 on the C300 with Super Speeds, then shoot f/5.6 on the 5D. To the keep the field of view similar, use a 35mm on the 5D if you would have used 25mm on C300, 50mm for a 35mm, 85mm for a 50mm, etc.

 

Whoa say wha? Ok, I will need to try and get my head around that one!

 

So, f1.2 on the C300 = f2.4 on the 5D?

 

This is very good info thankyou. I would never have known this.

 

Slight concern that shooting at f2.4 under the same conditions as the previous shoot is going to mean a substantial increase in iso. Not sure how far I can push the 5D.

 

Maybe if I'm using Magic Lantern 5D RAW it'll help? I'm hoping ML 5D RAW will help me match the C300 footage but it's an awful lot to take in by the looks of things. Huge memory cards required and some way of backing it all up quickly...


  • 0

#6 Mark Dunn

Mark Dunn
  • Basic Members
  • PipPipPipPip
  • 2182 posts
  • Other
  • London

Posted 20 May 2015 - 09:55 AM

It's a fundamental law of optics that DoF for a given stop is less on a larger sensor. Have a look at your college notes again.

Conversely as a rule it will be less prone to high ISO noise however.


  • 0

#7 Rod Lewis

Rod Lewis
  • Basic Members
  • PipPip
  • 30 posts
  • Director
  • Bournemouth

Posted 20 May 2015 - 10:09 AM

It's a fundamental law of optics that DoF for a given stop is less on a larger sensor. Have a look at your college notes again.

Conversely as a rule it will be less prone to high ISO noise however.

 

I'm a director, not a cinematographer and I do not have college notes which is one of the reasons I'm seeking answers here, but thanks.


  • 0

#8 Mark Dunn

Mark Dunn
  • Basic Members
  • PipPipPipPip
  • 2182 posts
  • Other
  • London

Posted 20 May 2015 - 10:29 AM

Nonetheless you appear to have a cinematographer's hat on today and understanding the answers to the questions you are asking requires some photographic knowledge.

May I suggest one of Michael Langford's books.


Edited by Mark Dunn, 20 May 2015 - 10:33 AM.

  • 0

#9 Rod Lewis

Rod Lewis
  • Basic Members
  • PipPip
  • 30 posts
  • Director
  • Bournemouth

Posted 20 May 2015 - 10:40 AM

Nonetheless you appear to have a cinematographer's hat on today and understanding the answers to the questions you are asking requires some photographic knowledge.

May I suggest one of Michael Langford's books.

 

www.rodlewisphotography.co.uk

 

I'm not the world's greatest photographer by any means, but I do have some photographic knowledge.


  • 0

#10 Rod Lewis

Rod Lewis
  • Basic Members
  • PipPip
  • 30 posts
  • Director
  • Bournemouth

Posted 20 May 2015 - 10:44 AM

I also understand everything that has been explained to me by the other people in this thread.


Edited by Rod Lewis, 20 May 2015 - 10:46 AM.

  • 0

#11 Satsuki Murashige

Satsuki Murashige
  • Sustaining Members
  • 3081 posts
  • Cinematographer
  • San Francisco, CA

Posted 20 May 2015 - 04:38 PM

 
Whoa say wha? Ok, I will need to try and get my head around that one!
 
So, f1.2 on the C300 = f2.4 on the 5D?
 
This is very good info thankyou. I would never have known this.
 
Slight concern that shooting at f2.4 under the same conditions as the previous shoot is going to mean a substantial increase in iso. Not sure how far I can push the 5D.
 
Maybe if I'm using Magic Lantern 5D RAW it'll help? I'm hoping ML 5D RAW will help me match the C300 footage but it's an awful lot to take in by the looks of things. Huge memory cards required and some way of backing it all up quickly...

Well, just think of the C300 as having a 7D sized sensor and it should make more sense. 'Crop factor' and all that. You'll have shallower depth of field on a full frame sensor if you choose lenses with the same field of view and keep the same lens to subject distance. The way to keep the same DoF under those circumstances is to stop down.

Magic Lantern Raw will certainly give you sharper images, greater color depth, and greater dynamic range. I would say it looks more like Red MX footage than C300, personally. One of the downsides, besides creating more data, is that the camera is much noisier at high ISOs due to the lack of in-camera processing. I don't like to go much above 640ISO if possible. When in doubt, overexpose. You will also see all of your sensor's dead pixels which you will have to paint out in post. If you're trying to do this on the cheap, then you're probably better off just shooting video in normal h264 All-I mode.
  • 0

#12 Rod Lewis

Rod Lewis
  • Basic Members
  • PipPip
  • 30 posts
  • Director
  • Bournemouth

Posted 20 May 2015 - 06:27 PM

Well, just think of the C300 as having a 7D sized sensor and it should make more sense. 'Crop factor' and all that. You'll have shallower depth of field on a full frame sensor if you choose lenses with the same field of view and keep the same lens to subject distance. The way to keep the same DoF under those circumstances is to stop down.

Magic Lantern Raw will certainly give you sharper images, greater color depth, and greater dynamic range. I would say it looks more like Red MX footage than C300, personally. One of the downsides, besides creating more data, is that the camera is much noisier at high ISOs due to the lack of in-camera processing. I don't like to go much above 640ISO if possible. When in doubt, overexpose. You will also see all of your sensor's dead pixels which you will have to paint out in post. If you're trying to do this on the cheap, then you're probably better off just shooting video in normal h264 All-I mode.

 

Thanks, that's interesting. I've always been aware of the difference in focal length with a cropped sensor and full frame, but although I've made the adjustments a thousand times it's never actually occurred to me the difference in depth of field, although I must have been aware of it on some level to be shooting professionaly. (Not a very technical photographer :unsure: ) But it's good to actually be technically aware of that now...

 

I certainly didn't know about the greater noise with Magic Lantern RAW either. Bearing this in mind, plus the fact that my f.2 would now be f5 (ish), I can imagine I'd be around iso 2000 or so at least so I imagine the noise would be pretty bad given what you're saying. The whole idea of this for me was to try and get the footage to match as closely as possible with the C300.

 

Amazed that I might better off in h264 All-I mode after all my obsessing over 5D ML RAW for so long   :blink:

 

With the latest 5DIII firmware upgrades do I even need Magic Lantern at all? I have heard that even not shooting in RAW, Magic Lantern still improves the bitrate. I have no idea if that's fact or not.....

 

The main thing is, that I have already spent £8k of my own savings on this film, and it would be stupid for me to compromise all my hard work by having several shots just jump out as being from another camera. I'd rather spend the £1 extra on hire of a C300 and Superspeeds and car mount for a day and adifferent D.O.P if it means I get the same look. There are others who can stand in for my guy, but I was hoping to brave it on y own.

 

Do you think that with enough work in grading normal 5D footage or h264 All-I in Magic Lantern footage could be matched closely enough that people won't notice?

 

Thanks for your help on this :)


Edited by Rod Lewis, 20 May 2015 - 06:29 PM.

  • 0

#13 Satsuki Murashige

Satsuki Murashige
  • Sustaining Members
  • 3081 posts
  • Cinematographer
  • San Francisco, CA

Posted 20 May 2015 - 08:34 PM

I think you'll be fine with the stock 5D3. The greater noise when shooting Raw was always there, it was just previously hidden by in the in-camera noise reduction processing and h264 video compression. Neat Video and/or other noise reduction software can help a lot, but if you need to shoot at 2000 ISO then you'll require a lot of clean-up work in post to make the footage usable.

If you're really concerned, I would coordinate with your DP who shot all the main unit stuff and let them guide you. It is after all their work that you'll be matching to and their name on the film, so I'm sure they'd be happy that you at least called and consulted them before shooting.
  • 0

#14 Rod Lewis

Rod Lewis
  • Basic Members
  • PipPip
  • 30 posts
  • Director
  • Bournemouth

Posted 20 May 2015 - 08:55 PM

I think you'll be fine with the stock 5D3. The greater noise when shooting Raw was always there, it was just previously hidden by in the in-camera noise reduction processing and h264 video compression. Neat Video and/or other noise reduction software can help a lot, but if you need to shoot at 2000 ISO then you'll require a lot of clean-up work in post to make the footage usable.

If you're really concerned, I would coordinate with your DP who shot all the main unit stuff and let them guide you. It is after all their work that you'll be matching to and their name on the film, so I'm sure they'd be happy that you at least called and consulted them before shooting.

 

Hi

 

Thanks for answering again!

 

This is doing my head in so much!

 

Me and the DOP have been in touch a lot about it. Although he's very busy on other paid shoots now and can't get involved himself, he has given me his thoughts on the matter. He doesn't think it matters too much, unless there are close ups on the actor's faces, (which there are but briefly) and initially he suggested I just hire some Zeiss EF mount lenses and shoot the pick ups on the 5D. For a long time I was insistent that it wouldn't do, and that all the footage must be an exact match, as we had really been through a lot of pitfalls on the shoot (car breaking, roads closing, etc) and I had spent so much money on the short already, (feature next hopefully) therefore the PL mount C300 and Superspeeds were a must in my mind. But months down the line, I am worn down through repeated failed attempts at pick ups through either actors not being available or the weather throwing a spanner in the works (it's all shot in a car at night). I realised that we really only need a few quick shots, and a lot are external shots with just a couple of very brief shots on the faces to match the previous footage.

 

Ultimately I am still prepared to hire a C300 and Superspeeds, but it adds more crew, a proper DOP and also permits to shoot which all comes to about £1k, whereas if there is a way to just wing it on the sly with my own 5D I not only save about a grand, but I can get away without permits hopefully for just some quick pick ups.

 

5D RAW was my hope of achieving this, as I could shoot it all myself with just my sound guy, a couple of production assistants, as well as my own car camera mount system that works with my 5D, but now it seems that 5D RAW in those conditions could be a bit noise heavy from what you're saying.

 

If I knew for sure that standard 5D would definitely match with enough post work, it'd be a no-brainer, but it's just in the back of my my mind that somehow the shots will just jump out as being from another camera and it'l compromise all the work we've put in.

 

I'm a bit OCD I think!

 

R


Edited by Rod Lewis, 20 May 2015 - 08:58 PM.

  • 0

#15 Satsuki Murashige

Satsuki Murashige
  • Sustaining Members
  • 3081 posts
  • Cinematographer
  • San Francisco, CA

Posted 20 May 2015 - 09:24 PM

It's not the equipment, it's the eye behind it that makes all the difference! 'Mad Max: Fury Road' used stock 5D Mark 2's as crash cameras. I don't think the general public has noticed yet, given all the strong reviews. Just use what you've got and concentrate on lighting, camera placement, lens selection, etc.
  • 0

#16 Rod Lewis

Rod Lewis
  • Basic Members
  • PipPip
  • 30 posts
  • Director
  • Bournemouth

Posted 20 May 2015 - 09:39 PM

It's not the equipment, it's the eye behind it that makes all the difference! 'Mad Max: Fury Road' used stock 5D Mark 2's as crash cameras. I don't think the general public has noticed yet, given all the strong reviews. Just use what you've got and concentrate on lighting, camera placement, lens selection, etc.

 

Ok thanks. I've banged my head against the wall too much on this. I'm gonna do some tests in standard 5DMKIII. 

 

Thanks for your help, Satsuki. I'll let you know how the footage compares finally :)


  • 0


CineTape

Technodolly

Tai Audio

Broadcast Solutions Inc

Pro 8mm

Abel Cine

Paralinx LLC

Willys Widgets

Zylight

Visual Products

Glidecam

CineLab

Rig Wheels Passport

The Slider

Ritter Battery

rebotnix Technologies

Aerial Filmworks

Zylight

Willys Widgets

Pro 8mm

Technodolly

Tai Audio

Glidecam

The Slider

Rig Wheels Passport

Paralinx LLC

Abel Cine

rebotnix Technologies

Aerial Filmworks

CineTape

Ritter Battery

Visual Products

CineLab

Broadcast Solutions Inc