Return of Ektachrome!

[James G. Dainis]

Anyone actually consider why pros dropped LF Speed Graphics for 35mm decades ago?

[rob_calhoun]

<p>Naturally, pros who had used transparency films found digital faster and more productive for their intended uses, but not necessarily better quality, at least without manipulation. The intended use of transparency film is projection, and that is where it really shines over digital. Not many pros used it for that. As for the typical consumer, they're happy viewing their photos on a smartphone.</p>

[glen_h]

<p>When my dad first bought a color TV, around 1969, he explained flying spot scanners for viewing slides on TV. I don't know that he ever looked for one, and I suspect that they were out of the price range.</p>

<p>Now, digital projectors are easy to find, and slides are easy to scan. (Negatives, too.) </p>

<p>How does slide projection compare to a digital (maybe DLP) projector? That is, in image quality, dynamic range, resolution (as seen by the viewer)?</p>

<p> </p>

[Karim Ghantous]

<p>Rob, I might slightly disagree with your point. I mean, yes, slides were meant for projection, but IIRC no stock library would accept your photos unless they were taken on slide film (for reasons I don't need to explain). In fact, a lot of press photographers used slide film, despite the preference for negatives (e.g. Ektapress). National Geographic photographers used slides almost exclusively for the same reasons that stock photographers did.</p>

[bebu_lamar]

<p>How does slide projection compare to a digital (maybe DLP) projector? That is, in image quality, dynamic range, resolution (as seen by the viewer)?<br>

I think today the highest resolution digital projector is 4K (a bit more than 8MP) and please correct me if I am wrong about this. The 35mm slide has a little bit more resolution than that but really not much more. I think the dynamic range of a bright slide projector so that one can see the full dynamic range of the slide then it's about the same as the best of the digital projector. There is one problem with the slide projector though that it seems even with curve field lens or flat field lens it's very difficult to get the entire slide in focus.</p>

[rob_calhoun]

<p>Karim, for publication and stock libraries, slides are generally preferred because they are a positive image and content easily evaluated compared to a negative. However, unlike color negatives, color slides are not masked, and before being published, color correction must take place, generally in the form of making masks to correct for dye impurities in the slides, otherwise color will be degraded <strong>when printed</strong>. (Color negatives have such dye-impurity correction built right into them giving them an orange color, and better image color when printed.) Masking must also be done to lower the high contrast of a slide. Publishing outfits will have a department set up to make and use such masks for all slides being published. </p>

<p>Therefore, slides can be and are used for publication, but that is not their intended purpose, and as a result a bit of work must take place before publication to use them.</p>

[rob_calhoun]

<p>Sorry, in a hurry and forgot to say that I was explaining a more tradional way of dealing with slides, and today the problem can be dealt with using software after scanning slides.</p>

[flavio_egoavil]

<p><em>"I think today the highest resolution digital projector is 4K (a bit more than 8MP) and please correct me if I am wrong about this. The 35mm slide has a little bit more resolution than that but really not much more"</em></p>

<p>Much more. You can extract good 24MP or even more from a really good fine-grained positive (think Provia 100F), if you have a really good scanner. This rules out all flat-bed scanners, by the way. We're talking dedicated film scanners, preferrably Fuji Frontier machines or (even better) drum scanners or the Hasselblad thing.</p>

 

[bebu_lamar]

<p>I found beyond 12MP or about 3000dpi just show up the grain and not more details.</p>

[Karim Ghantous]

<p>BeBu, I have not done serious scanning but I have seen 24Mpx scans of 35mm film and there is definitely ~24Mpx of information there. It's not as clean but it's there. Super 16mm can resolve at least 2K/HD. Super 35 can resolve 4K, roughly equivalent to 12Mpx. Vistavision (which we call 135) therefore can and does resolve 24Mpx.</p>

[Robin Smith]

<p>It is good news. I actually much preferred Kodachrome or Fuji Astia/Sensia 100 for everyday shooting, and Velvia 50 for high saturation stuff, but still it is good to see it revived. The announcement is very vague about <em>which</em> Ektachrome they are bringing back though. Why it is taking until Q4 to do it is not clear, by then we may all have forgotten about it.</p>

<p>I see there are now rumblings about bringing Kodachrome back too.</p>

[rob_calhoun]

<p>As I understand it there are technical problems that prevent any of the previous Ektachromes from coming back in their old exact form so the new film will be a reformulation. To reformulate and test will take some time so that's why the delay. </p>

<p>Bringing back the old Kodachrome is far more problematic as no one currently develops it anywhere. It was a very difficult, complex process (different than Ektachrome, which is much easier), and some of the chemistry is not even made anymore. Kodachrome had technical problems that were partly responsible for its unique look that some liked, but most would prefer the technically superior E6 films of today.</p>

[Karim Ghantous]

<p>Quick poll: which film speed should Ektachrome be, if Kodak only releases one version:</p>

<p>50<br>

100<br>

400</p>

[chuck909]

<p>I'd vote for 100</p>

[glen_h]

<p>I vote for 100. I still remember Ektachrome X (ASA 64), and Ektachrome 64, but 100 seems about right. Nice for daylight and flash pictures, not so good for available light at night or indoors.</p>

[glen_h]

<p>From: http://photoweb.ru/exusr/pdf/kodak/e27.pdf </p>

<p>(pages are hard to find on kodakalaris.com. google indexes them, but then they don't exist.)</p>

<p>The MTF curve goes to about 50 lp/mm, or about eight megapixels (2400x3600). </p>

<p>To reduce aliasing, a digital camera or scanner needs to low-pass spatial filter to a spatial frequency below half the pixel resolution. Most likely, you will need a lot more than 8 MP in a digital image to get equivalent resolution. </p>

[Karim Ghantous]

<p>Glen, thanks for that. But other slide films have a much higher resolution than that. What's also interesting to note is that negative films usually capture more detail than slide films. I don't know why, but it's true.</p>

[glen_h]

<p>50 cycles/mm is the 30% point, and the end of the graph. </p>

<p>But as I noted, we tend to assume that anything digital has the full resolution of the pixels available, but that is pretty much never true. Since optical low-pass filters are much more difficult to make than for digital audio, I suspect that you need 200 pixel/mm to get 50 cycles/mm, instead of the ideal 100 px/mm. </p>

<p>I think the data sheet for Kodachrome 25 is still around.</p>

[glen_h]

<p>Hmmm. From:</p>

<p>http://wwwde.kodak.com/global/en/professional/support/techPubs/e55/e55.pdf</p>

<p>which gives curves for Kodachrome 25, 64, and 200, all three look amazingly similar, all go to 80 cycles/mm at 10%. </p>

<p>But yes, slide films are different. The image is formed from the grains that don't form the latent image. (That is, that don't develop in the first developer.) </p>

<p>For negatives films, the large grains develop to form the shadows, and larger grains form the actual dye clouds. </p>

<p>For slides, it is the grians that don't form the latent image, which tend to be the smaller grains. It isn't so obvious what that does to the image structure.</p>

[Karim Ghantous]

<p>It's worth noting that contrast affects resolution of film - by no means is this a bad thing, but it's worth keeping in mind.</p>

<p>I have a few printed documents from Kodak, one of which is for Elite 100. That has a maximum resolving power or 80 cycles/mm at 30% response (which I don't understand, so I'll have to look that up at some point). The Kodachrome brochure says that K25 & K64 both give 100 "lines per mm" at 1000:1 contrast ratio. I wonder if "lines" should be "cycles" or "line pairs"?</p>

<p>What's interesting is that Fuji Superia 200 out-resolves K25:<br /> <br />http://www.fujifilm.com/products/consumer_film/pdf/superia_200_datasheet.pdf</p>

<p>They both are measured in "lines", so even if I'm confused as to why the measurements appear different, at least Superia 200 and K25 use the same terms.</p>

<p>Edit: I also have a data sheet for Eastman EXR 200T 5293, and it says that on average, the resolving power is over 100 cycles/mm at 10% response. But the same document gives identical resolution figures as for K200: 100 lines/mm at 1000:1 contrast. There seem to be different metrics at play. I am not knowledgable enough at this point to make complete sense of them.</p>

[James G. Dainis]

"line pairs per mm" always seemed odd to me. To me it looks like black lines drawn on a white background. You have 100 black lines drawn per mm or 100 lines per mm. I guess because the white spaces between the black lines are the same width as the black lines that they are called white lines, ergo the black and white line pairs per mm.

[glen_h]

<p>As well as I know, before digital came along lines/mm or lines/inch was well enough defined. You can't have black lines without white spaces between them, or vice versa. To clarify, though, it is now often written as line pairs (black lines and white lines), or cycles. </p>

<p>Cycles is probably more accurate. It is the optical (spatial frequency) equivalent of the frequency response for an audio amplifier. As with audio, it is defined in terms of response to sinusoidal signals.</p>

<p>http://spie.org/publications/tt52_131_modulation_transfer_function</p>

<p>Since square wave (sharp edged) targets are easier to make, I suspect that people use those, and then convert the result to the appropriate sinusoidal response.</p>

<p>In any case, as the spatial frequency (cycles/mm) gets higher, the contrast tends to decrease, for a variety of reasons. In the case of film, the finite size of grains, scattering of light in the emulsion, and for dye image systems, the diffusion of the oxidized developer before forming the dye cloud. </p>

[Robin Smith]

<p>Yes, it is similar to lens performance now being routinely presented as mod trans functs - MTF- charts. As resolution decreases sharp edges become less well defined and the contrast between black and white lines declines too. I assume a 30% difference is assumed to be the useful threshold for just making out that there are separate lines i.e. the limit of resolution.</p>

[iosif_astrukov]

<p>Didn't they announced 100 ASA film… I do remember the 64 ASA Tungsten, but never have seen 200 ASA or 400 ASA Ektachrome? Were there once such films? <br>

I think Elitechrome had 100 and 200 ISO versions… <br>

Last on the market were 100 ASA G and VS versions… <br>

My personal tribute to the great VS<br>

<img src="http://ioshertzweb.appspot.com/images/E100VS.jpg" alt="" /></p>

 

[Karim Ghantous]

<p>Iosif, yes, there were 200 and 400 ISO Ektachrome versions. E.g:<br>

<br /><br />http://wwwde.kodak.com/global/en/professional/support/techPubs/e161/e161.pdf</p>

<p>http://www.visionimagelab.com.au/_literature_84478/EKTACHROME_Film_E200</p>