Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

Late to the party, but Ektachrome is coming back!

  • Please log in to reply
117 replies to this topic

#76 Michael Covington

Michael Covington

    Author

  • *****
  • Posts: 6,868
  • Joined: 13 May 2014
  • Loc: Athens, Georgia, USA

Posted 02 June 2018 - 05:09 PM

That is what is being bitterly debated here.  The best test I have done was with Kodachrome 64, and obvious Tech Pan or even Ilford Pan F is sharper than that.

As I understand it, film like Techn Pan claims a resolution up to maybe 150 lines/mm.  Ignoring MTF, that is equivalent to 300 pixels/mm, or 3.3-micron pixels, slightly smaller than those in a current 24-megapixel APS-C DSLR.

However, near the limits of their resolution, film and digital sensors behave very differently.  A film that tests out at 150 lines/mm will have only minuscule contrast at that resolution.  Contrast will drop as you go up from maybe 50 lines/mm.  A digital sensor with 3.3-micron pixels, on the other hand, will have essentially full contrast at 150 lines/mm (Nyquist theorem).

My actual experiments show that fine-grained color slide films are comparable to maybe 20-micron pixels, i.e., not nearly as sharp as current digital sensors.

There's a reason why the digital takeover happened when it did.  Six- or eight-megapixel sensors rival most films; 24-megapixel sensors surpass them, IMHO.


  • Alen K likes this

#77 TxStars

TxStars

    Mercury-Atlas

  • *****
  • Posts: 2,618
  • Joined: 01 Oct 2005
  • Loc: Lost In Space

Posted 02 June 2018 - 07:11 PM

I think for a given total exposure time film may still have an advantage.


  • SandyHouTex likes this

#78 Michael Covington

Michael Covington

    Author

  • *****
  • Posts: 6,868
  • Joined: 13 May 2014
  • Loc: Athens, Georgia, USA

Posted 02 June 2018 - 08:24 PM

Quite the contrary. Film has reciprocity failure, digital doesn't.  Also, digital is linear, so dark frame subtraction, flat fielding, etc., are feasiable.  Film is nonlinear and image arithmetic generally doesn't work.

 

A digital sensor picks up as much light in 1 minute as film at the same ISO in 15 to 60 minutes.



#79 Faintandfuzzy

Faintandfuzzy

    Lift Off

  • -----
  • Posts: 7
  • Joined: 15 Feb 2011
  • Loc: Victoria, BC, Canada

Posted 10 June 2018 - 05:40 PM

Film is simply a different medium for look and spec.  Using Tech Pan or Adox CMS 20, and an Imacon 848 at 8000ppi, you can pull about 40mp of useful detail from a high rez chart.  In real life contrast ranges of 1:6 to 1:20, it is closer to about 24mp.  The same scanning setup for the finest grain film made, Fuji Astia 100F, one cannot get more than about 20mp of useful info at normal contrast ranges....and about 35mp mp on a rez chart.  Rockwell is simply wrong here. 

 

That said, I'll take the superior reciprocity characteristics of a digital cam for astro work over film any day.  


  • Alen K likes this

#80 TxStars

TxStars

    Mercury-Atlas

  • *****
  • Posts: 2,618
  • Joined: 01 Oct 2005
  • Loc: Lost In Space

Posted 10 June 2018 - 10:30 PM

"A digital sensor picks up as much light in 1 minute as film at the same ISO in 15 to 60 minutes."

 

So you are saying that my 20min shot shown below can be done with a digital camera in 1 min?

https://www.cloudyni...ge/59158-e100g/

 

Imaged with my Takahashi FCT-65  @ F3.7

 

Ummm I dont buy that till someone can show me..



#81 Michael Covington

Michael Covington

    Author

  • *****
  • Posts: 6,868
  • Joined: 13 May 2014
  • Loc: Athens, Georgia, USA

Posted 10 June 2018 - 11:20 PM

"A digital sensor picks up as much light in 1 minute as film at the same ISO in 15 to 60 minutes."

 

So you are saying that my 20min shot shown below can be done with a digital camera in 1 min?

https://www.cloudyni...ge/59158-e100g/

 

Imaged with my Takahashi FCT-65  @ F3.7

 

Ummm I dont buy that till someone can show me..

YES!

 

Let me rummage around for a comparable example...

This is the thing I could most quickly lay hands on.  It is a single 2-minute exposure at f/4 through a 180-mm telephoto lens; that is, 45 mm of aperture.  It looks a little grungy because the usual processing (darks, flats, etc.) has not been done.  (I quickly reprocessed just one frame for you, to show that stacking is not necessary to pick up faint detail, though it helps.)  And it's at only ISO 200.  It is one of a set of frames to be stacked.  If I had not been planning on stacking, I would have set the ISO higher.

M31example.jpg

 

I know this isn't the same area you photographed; let me see what else I have.

You have no idea what reciprocity failure is costing you until you use a digital sensor, which doesn't have any.


 



#82 Michael Covington

Michael Covington

    Author

  • *****
  • Posts: 6,868
  • Joined: 13 May 2014
  • Loc: Athens, Georgia, USA

Posted 10 June 2018 - 11:23 PM

This is a stack of four 1-minute exposures of the Lagoon and Trifid with an AT65EDQ (f/6.5) and Nikon D5300 digital SLR.  It is a closer comparison to yours.  A single exposure would look similar but about twice as grainy.  A filter-modified DSLR would show the nebulae as red; here you're missing much of the H-alpha and mostly looking at blue H-beta, which film doesn't record well because it falls in a gap in the sensitivity.

160701-M8-4x1m-1000h666w.jpg

Digital really does have serious advantages over film.


Edited by MCovington, 10 June 2018 - 11:27 PM.


#83 Alen K

Alen K

    Surveyor 1

  • *****
  • Posts: 1,705
  • Joined: 25 Nov 2009

Posted 11 June 2018 - 09:22 AM

Michael, I'm not disputing the efficiency comparison (although a factor of sixty does seem a little high) but wouldn't you have to compare photos taken with optical systems yielding similar CEFA's to judge fairly? 

 

http://www.clarkvisi...y.and.exposure/

(Let's ignore the cheek of using one's name in an acronym for a figure of merit. The number is IMO a useful metric nonetheless.)


  • Miguelo likes this

#84 Michael Covington

Michael Covington

    Author

  • *****
  • Posts: 6,868
  • Joined: 13 May 2014
  • Loc: Athens, Georgia, USA

Posted 11 June 2018 - 06:13 PM

Same celestial object (a smooth surface such as a nebula) and same f-ratio should be comparable.   I brought in two very approximate comparisons because that was all I had time to dig up.  I know that film is dramatically worse than digital, because of reciprocity failure.  Are you familiar with reciprocity failure and the Schwarzschild formula?



#85 Alen K

Alen K

    Surveyor 1

  • *****
  • Posts: 1,705
  • Joined: 25 Nov 2009

Posted 11 June 2018 - 08:42 PM

I am familiar and that's why I primarily agree with you. But it would be nice to see a closer comparison, e.g., same deep-sky object, same optics, film vs. DSLR. AFAIK no one has done it. If anyone did do it, it would have happened years ago when film was still easily available.


  • Miguelo likes this

#86 Michael Covington

Michael Covington

    Author

  • *****
  • Posts: 6,868
  • Joined: 13 May 2014
  • Loc: Athens, Georgia, USA

Posted 11 June 2018 - 08:46 PM

Yes; I don't think I have ever done an exact comparison but will look back.



#87 TxStars

TxStars

    Mercury-Atlas

  • *****
  • Posts: 2,618
  • Joined: 01 Oct 2005
  • Loc: Lost In Space

Posted 12 June 2018 - 12:52 AM

Well after seeing these it looks like I am going to be spending my extra money on some digital cameras, software and what not for when I don't want to use film.



#88 Michael Covington

Michael Covington

    Author

  • *****
  • Posts: 6,868
  • Joined: 13 May 2014
  • Loc: Athens, Georgia, USA

Posted 12 June 2018 - 10:33 AM

What line of film cameras do you have?  Nikon lenses (even pre-AI ones) fit the Nikon D5300, which is an outstanding performer.  They also fit on Canons with adapters.  Many other SLR lenses also fit on Canons with Fotodiox adapters because the Canon body is the second-shallowest in the business.  The only lenses you can't adapt are Canon FD lenses, because Canon's pre-EOS film bodies were even shallower.



#89 TxStars

TxStars

    Mercury-Atlas

  • *****
  • Posts: 2,618
  • Joined: 01 Oct 2005
  • Loc: Lost In Space

Posted 12 June 2018 - 09:25 PM

Lol 

I put in an order this morning for a Nikon 5300 to use with my Nikon manual focus lenses..


  • Michael Covington likes this

#90 Michael Covington

Michael Covington

    Author

  • *****
  • Posts: 6,868
  • Joined: 13 May 2014
  • Loc: Athens, Georgia, USA

Posted 12 June 2018 - 09:44 PM

Lol 

I put in an order this morning for a Nikon 5300 to use with my Nikon manual focus lenses..

For astrophotography it will be great.  You will find that its exposure meter works with only the very latest Nikon autofocus lenses.  Nikon is weird...  But it may be the single best astrophotography DSLR under $1000 today.



#91 Michael Covington

Michael Covington

    Author

  • *****
  • Posts: 6,868
  • Joined: 13 May 2014
  • Loc: Athens, Georgia, USA

Posted 12 June 2018 - 10:44 PM

Make sure you learn how to do the processing.  It's not like daytime photography.  Calibration and stacking are important.



#92 SandyHouTex

SandyHouTex

    Fly Me to the Moon

  • *****
  • Posts: 5,965
  • Joined: 02 Jun 2009
  • Loc: Houston, Texas, USA

Posted 13 June 2018 - 01:18 PM

Film is simply a different medium for look and spec.  Using Tech Pan or Adox CMS 20, and an Imacon 848 at 8000ppi, you can pull about 40mp of useful detail from a high rez chart.  In real life contrast ranges of 1:6 to 1:20, it is closer to about 24mp.  The same scanning setup for the finest grain film made, Fuji Astia 100F, one cannot get more than about 20mp of useful info at normal contrast ranges....and about 35mp mp on a rez chart.  Rockwell is simply wrong here. 

 

That said, I'll take the superior reciprocity characteristics of a digital cam for astro work over film any day.  

I have thoroughly reviewed what Ken did, and can find no errors.  How about some objective analysis of your own and how does it conflict with what he did?


  • Miguelo likes this

#93 Michael Covington

Michael Covington

    Author

  • *****
  • Posts: 6,868
  • Joined: 13 May 2014
  • Loc: Athens, Georgia, USA

Posted 13 June 2018 - 07:05 PM

The question is: is Ken Rockwell comparing a digital sensor to the same *area* on the film, or is he comparing it to the same *part of the picture* on a picture taken with 4x5 film, much larger than the sensor?  That's what we weren't sure of. 



#94 Alen K

Alen K

    Surveyor 1

  • *****
  • Posts: 1,705
  • Joined: 25 Nov 2009

Posted 13 June 2018 - 07:50 PM

A more balanced view? http://www.clarkvisi....summary1.html/


  • Miguelo likes this

#95 Michael Covington

Michael Covington

    Author

  • *****
  • Posts: 6,868
  • Joined: 13 May 2014
  • Loc: Athens, Georgia, USA

Posted 13 June 2018 - 09:06 PM

That is a good analysis.  Note that many of his comparisons are to older sensors.  24- to 36-megapixel sensors are now common, and sensor noise levels and dynamic range have continued to improve, in some cases dramatically.



#96 SandyHouTex

SandyHouTex

    Fly Me to the Moon

  • *****
  • Posts: 5,965
  • Joined: 02 Jun 2009
  • Loc: Houston, Texas, USA

Posted 14 June 2018 - 09:25 AM

The question is: is Ken Rockwell comparing a digital sensor to the same *area* on the film, or is he comparing it to the same *part of the picture* on a picture taken with 4x5 film, much larger than the sensor?  That's what we weren't sure of. 

Resolution is not a function of area!


  • Miguelo likes this

#97 SandyHouTex

SandyHouTex

    Fly Me to the Moon

  • *****
  • Posts: 5,965
  • Joined: 02 Jun 2009
  • Loc: Houston, Texas, USA

Posted 14 June 2018 - 09:42 AM

He says Velvia 100 is equivalent to a 13 megapixel sensor.  Not to mention that he never corrects for the "deBayerization" process.  Sometimes called demosaicing.  Here's an excellent article on it:

 

https://en.wikipedia...iki/Demosaicing

 

For a cmos sensor with a Bayer filter, you don't get 16 megapixels of information for a 16 megapixel sensor.  You get a bunch of red and blue pixels for each two green pixels.  To get the actual color for a specific pixel, you must interpolate all of the pixels around it.  This is a very exhaustive and lengthy computing process to get anywhere close to the actual color that's there.  Just by doing this you lose resolution.  Even if you do it well, which many cameras don't.

 

Digital sensors have their place.  They're more sensitive than film, allow "lucky imaging", etc.  


  • Miguelo likes this

#98 Michael Covington

Michael Covington

    Author

  • *****
  • Posts: 6,868
  • Joined: 13 May 2014
  • Loc: Athens, Georgia, USA

Posted 14 June 2018 - 10:33 AM

I can't speak for anyone else, but I've done my own tests by comparing digital camera images (both daytime and astronomical) with color slides (both digitized and examined through microscopes).  I came to the conclusion that digital almost always has the edge.  I've already posted links to some of my tests.

As I said, we're not quite sure how Ken Rockwell's test was done, and that's the origin of the dispute; he had a 4x5 camera involved somehow.

Roger Clark is a noted optical analyst, not just some guy with an opinion.  However, the page that was linked was mostly made around 2002, and digital cameras have advanced somewhat since then.  He has done partial but not complete updates to it.

I see insults beginning to be flung here.  I won't stick around for that.  I invite everyone to compare whatever photographic techniques they want to compare, and draw their own conclusions.

 


Edited by MCovington, 14 June 2018 - 10:35 AM.


#99 Erik Bakker

Erik Bakker

    Cosmos

  • *****
  • Moderators
  • Posts: 9,528
  • Joined: 10 Aug 2006
  • Loc: Netherlands, Europe

Posted 14 June 2018 - 11:01 AM

Let's play nice, be friendly and respect others here.


  • Uwe Pilz likes this

#100 donlism

donlism

    Viking 1

  • *****
  • Posts: 538
  • Joined: 20 Oct 2015
  • Loc: San Jose, CA

Posted 14 June 2018 - 12:14 PM

I suspect one of the common limitations of film resolution is focusing.  It's pretty hard to keep film flat, and pretty hard to focus on it.

 

Medium and large format high-resolution film-based photography interests me.  My ill-fated eclipse plan included medium format because of the large area of sky that would be easy to cover with it, inexpensively and easily, with simple manual exposure bracketing, and then with the smaller, complicated, software-driven, yes-people-have-to-fiddle-with-them-all-them-time-and-reload-drivers-and-stuff-because-software-is-ALWAYS-broken digital cameras carrying the load for the close-ups and such.

 

Alas, none of it flew.  But I'm going to do some high-res moon stuff as soon as I can get past issues of life!  Black and white.  Not Ektachrome.  I'm skeptical EP will really make it back -- the processing market just isn't there, I think.  But... best of luck with the endeavor anyway, and I hope I'm wrong!




CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Recent Topics






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics