Excessive contrast in D810 slide copying

[chuck]

<p>I've been using D810, ES-1 slide copy attachment, 60mm D macro, and a SB-800 flash to copy some of my old Velvia and Provia slides. I use the flash, connected to the camera by cord, to backlite the slides, and use manual exposure on the camera and manual output on the flash, experimenting with exposure and output until I get a reasonable overall exposure in the digital copies of the slide. I shoot raw + JPEG fine.<br>

<br /> Unfortunately, the results generally show excessive contrast, clearly higher contrast then the slides. What is more, in Live View, the monitor on the back of the camera show acceptable contrast. But when the image is actually taken and reviewed on the same monitor, excessive contrast is clear. The excessive contrast is also there when the image is seen on any other monitor.<br /> So it seems almost as if the image captured by the DSLR had insufficient dynamic range to cover what the backlite slide can present.<br /> <br /> Is there anyway around this?<br>

Another question, when the scene in the slide is well lite, the D810 captures the color in the slide reasonably accurately. But if the entire scene is dimly lit, the D810 seem to give the image of the slide a disturbing amount of green cast. What is causing this?</p>

[Andrew Garrard]

<blockquote>So it seems almost as if the image captured by the DSLR had insufficient dynamic range to cover what the backlite slide can present.</blockquote>

 

<p>That's possible, but a little unlikely. Or at least, at minimum ISO, the D810 has pretty much as much dynamic range as any DSLR on the market, and close to the most optimistic claims for film. It is extremely likely, however, that your monitor doesn't have the dynamic range of backlit slides.<br />

<br />

If you're taking JPEGs, my first suggestion would be to turn on active D-lighting (in the shooting menu), which - I believe - does some region-aware tone mapping (exposure fusion or similar) to map a higher dynamic range into the limited output that can be represented by a JPEG. If that doesn't work for you, you could also try the in-camera HDR option, or the "flat" picture control mode. Big disclaimer, I don't shoot in JPEG (except as backup), so I don't know much about the behaviour of any of these.<br />

<br />

My preferred option would be to import to an editing package (DxO, Photoshop, etc.) in a high bit depth mode (16-bit normally suffices) and adjust the result to taste. It's pretty common for me to adjust tone curves or to push shadows and pull highlights, often in the raw converter, which acts similarly to the D-lighting but with interactive control. Or you can just adjust curves to make the results image-wide, of course. Make sure your highlights weren't too clipped in the original capture, but you can pull a lot of shadow detail out of an ISO 64 raw file from a D810. If I couldn't do this, I'd need to do a lot more fill flash than I currently do, and that brings the pain of balancing light source temperatures - these days, I usually rely on ambient lighting as much as possible and on post processing to fill shadows.<br />

<br />

I'd hope some combination of the above can get the look you want. The output of Velvia, especially, is <i>very</i> contrasty anyway, of course - my eyes still burn at the memory of the time I used a yellow/blue polariser on top of Velvia. It's part of the look, and very useful when projecting slides because there's usually so much ambient light present when using a slide projector - but it does make Velvia slides look distinctive on a slide table. (I really must get some old film developed.)<br />

<br/ >

Good luck!</p>

[Andrew Garrard]

<blockquote>Another question, when the scene in the slide is well lite, the D810 captures the color in the slide reasonably accurately. But if the entire scene is dimly lit, the D810 seem to give the image of the slide a disturbing amount of green cast. What is causing this?</blockquote>

 

<p>Sorry, missed this. Are the slides processed normally, or are the dim scenes push-processed? It's possible that there is a colour cast there that your eyes are failing to spot as well as the camera does (hopefully you can fix it by fixing the white balance of the image separately for highlights and shadows - the Photoshop white balance tool can do this) or it may just be that the camera's colour filtration doesn't exactly match the spectral response of the cones in the eye, and the colours in the film aren't being recorded the same way by the sensor as they appear in person. Velvia does deliberately introduce some weird colour shifts in order to "look good" (it always makes grass look too blue, to me) but I can't say for sure what exactly you're seeing. I've had to do major digital surgery on some old slides in the past, so I'd be pleasantly surprised to learn that you'd been able to get away with the defaults. The good news is that you <i>can</i> fix things up a surprisingly long way. My scanner was lucky to get 6MP worth of information out of a slide (it had more resolution - I think the slides were the problem) but I hope you do better with your set-up.<br />

<br />

If all else fails, you could chuck it at a drum scanning facility and ask them to fix it for you...</p>

[chuck]

I think there is a difference between the dynamic range slide film can capture, and the dynamic range exposed slide film

can present when it is backlit. I am thinking when exposed slide film is backlit, it may exhibit a much larger dynamic

range in the image it presents then was ever these in the scene it captured. As a result, its presentation could

overwhelm even the D810, even if it could not capture nearly as much dynamic range as the D810 when it is being used

to take pictures.

 

Does that sound plausible?

 

Exposure backing is a good idea. I will try that.

[dan_brown4]

<p>Try softening your light source. I have done this with the D800, the ES-1, 60 Micro, Provia slides, using an AlienBee B400 320WS studio flash (at about 1/4 power) into a small double diffusion softbox, and the result were satisfactory.</p>

[rodeo_joe1]

<p>Slides have a D-max of 3.6 at very most, and a minimum density of around 0.15D. That's a density range of 3.45D or a dynamic range of 11.5 stops. JPEGS are limited to 8 bits and are geared to showing a subject brightness range of around 7 stops. Since the camera can only show a JPEG version of the image on its review screen then it's obvious that 11.5 into 7 won't go - hence the excessive contrast.</p>

<p>Basically you need to be using RAW capture and the only way you can do that (and see the result properly) is by downloading the image file onto a computer and flattening the tone curve or otherwise reducing the contrast in some way. Changing the Picture Control mode of the camera might help though. You definitely don't want "Vibrant" or whatever set as your JPEG option.</p>

<p>Having said that, in order to get anywhere near to the theoretical 11.5 stop range of a slide you need to use fresh film, exposed immaculately and then have it promptly and <em>professionally</em> processed. Anything else and the DR is going to be more like 9 or 10 stops - which is still a bit much for a crummy 8 bit JPEG file to handle.</p>

<p>Another thing to bear in mind is that the density range of a slide bears no resemblance to the original SBR. The capture range of a slide is probably no more than a standard 7 stops, but the gamma (transfer function) is very high, which gives slides their artificial "zing". Unlike colour-negative film which could capture a 10 stop subject range but only show it as a pitiful 5 or 6 stop range on paper.</p>

[ShunCheung]

<p>Slide films such as Velvia and Provia typically have a 5-stop dynamic range. The D810 is going to exceed that many times over.</p>

<p>Chuck, it would be helpful if you can post an image sample or two showing the issue, but like Dan Brown, I would imagine that your problem is the flash being too harsh. Diffusing the flash may help.</p>

[Ed_Ingold]

<p>I get perfectly good slide copies using a continuous light source, like a daylight LED bulb in an ordinary desk lamp. Exposure for a flash unit is a lot more complicated, which may be causing excessive contrast. I've been using AWB or fixed white balance (e.g., 3200K or 5500K, depending on the bulb) with good results. Auto exposure and AWB can help improve a poorly exposed slide. In general, I've been very pleased with the results.</p>

<p>I bounce the desk lamp off a slanted white card for better light distribution. The ES-1 itself is pretty effective on its own in this regard. It doesn't matter if the exposure takes 1/60 or 4 seconds - the ES-1 is locked in place, even if you hold the camera in your hand.</p>

<p>My setup is a Sony A7ii, Novoflex E to Nikon adapter, PX-13 Extension Tube, 55/2.8 AIS Micro-Nikkor, and an ES-1 Slide Holder, and a 5500K LED Desklamp, fixed at ISO 400. This photo was taken on Kodachrome over 50 years ago in Mexico City. There were no extraordinary adjustment to the image, other than Lightroom defaults.</p>

<p><img src="http://d6d2h4gfvy8t8.cloudfront.net/18079133-lg.jpg" alt="" width="467" height="700" /></p>

<p>There's usually not much detail to extract from dark areas in Kodachrome or Velvia. but you could use dynamic range or HDR techniques to get better results. It's easy to do, and costs nothing but time.</p>

[Matthew Currie]

<p>If contrast is the only issue, what about altering the contrast of the capture? On lesser models, at least, one can adjust the contrast and other aspects of each picture control setting, and can add or delete D-lighting to lighten shadows. </p>

[dan_brown4]

<p>The slide film's dynamic range is irrelevant with respect to copying the slide. Just like Ansel Adam's zone system film and paper analysis, the exposed and developed slide's contrast range is from the clarity of the film base (brightest) to the most blocked-up emulsion (darkest). This issue is about the light source used on the copy stand.</p>

[Matthew Currie]

<p><em>This issue is about the light source used on the copy stand.</em><br>

<em> </em><br>

<em><br /></em>So I would have assumed, except that the original post suggests that acceptable contrast is shown in the live view monitor. That suggests that even if it would be easier to get it with better lighting, it's there, but not being recorded. </p>

[ShunCheung]

<p>According to Chuck, he is using an SB-800 flash with the D810 to copy his Velvia and Provia slides. In the live view mode, the flash cannot possibly be on the whole time. Therefore, Chuck must be depending on some other light source that apparently has less contrast to illuminate his slides (during live view).</p>

<p>Once again, I think he needs to modify the light coming from the SB-800 to reduce contrast.</p>

[chuck]

<p>I am not sure how harshness of the light sources can effect the contrast. <br /> It seems to me that the contrast of the processed slide depends solely on the differing degree to which each part of the slide attenuates the light passing though it. As a result, so long as the light impinging on the slide is basically uniform in color and brightness, it would proportionally attenuate the same way as it goes through different parts the slide, resulting in an image of the same contrast on the other side.</p>

<p> </p>

[chip_chipowski]

<p>Chuck, you are a tough customer.</p>

 

[Ed_Ingold]

<p>Chuck,<br>

By your own testimony, the image looks good in live view (room light) but is harsh in the final version using a flash. Does this sound like the light source could be the problem?</p>

<p>In my experience, results using a soft, continuous light source are indistinguishable from the live view, and close to or even superior to the slide itself (i.e., the camera makes minor corrections in exposure and color balance).</p>

[Matthew Currie]

<p>Of course I forgot that the live view must be under a different light from the exposure if one is using flash.</p>

[rodeo_joe1]

<p>There is a phenomenon called the Callier effect*, whereby the transmitted contrast is greater when using a hard (collimated) light source rather than a soft (diffused) one. It's well-known among darkroom printers that a point-source enlarger produces more contrasty prints than one with a diffuse source. However, the difference isn't startling, amounting to at most one paper grade, and I don't think that it accounts for what's happening here. The ES-1 has an opal plastic diffuser in front of the slide holder, which should diffuse any light source - short of a laser - to much the same degree.</p>

<p>The light source should only vary in brightness and not quality when passed through the copier's diffuser screen. Diffuse light is diffuse light, and double or triple diffusing it should make almost no difference. So I still think it has everything to do with the review image being from a gamma-adjusted JPEG, rather than the direct sensor image viewed during Live View. There's also the possibility that ambient light falling directly onto the front of the slide may well reduce contrast if the backlighting isn't very strong.</p>

<p>Shun, the dynamic range that can be captured by slide film has nothing to do with its transmitted contrast range. As I said, the density range of slide film <em>can</em> amount to over 11 stops, and it would need this amount of dynamic range in a scanner or copy-camera to scrape all the shadow detail out of the slide. I've measured the Dmax and Dmin of slide film with a calibrated densitometer, and can verify that a density range of 10 or 11 stops is not unusual. The density range has absolutely nothing to do with the subject brightness range that a slide film can capture.</p>

<p>* For anyone interested; the Callier effect is caused by diffusion of light within the emulsion of a negative or slide as it passes through the dye or silver particles. Hard light that passes through clear gelatine remains undiffused, while the denser parts of the medium scatter some light such that it never reaches the focussing lens on the other side. Thus the denser parts appear denser than their diffuse density would suggest. When illuminated with an already diffuse light source the scattering of light within the emulsion doesn't make any such difference between areas of high and low density.<br /> Here's a link to the Wikidpedia description (not very clear IMHO): https://en.wikipedia.org/wiki/Callier_effect</p>

[pete_s.]

<p>In the camera it's the tone curve that determines how the tones are compressed for viewing and on jpegs. I think most Nikon's have an automatic tone curve by default. So they add contrast to a low contrast scene. I'd take a look at the settings. It called Picture Control.</p>

<p>Flat picture control should be the starting point on the D810. Since the slide already have added all the contrast needed, a flat and accurate capture should be what you want. And of course shot with manual WB, manual exposure, manual ISO and manual power setting on the flash.</p>

<p>For the optimum setting you should look at the raw files with rawdigger. Basically expose to the right (on brightest white possible) for maximum dynamic range to be captured. Then shoot with the same settings for all slides.</p>

<p>That should give you the exact same "scan" and same contrast in relationship to the slide you have in front of you when the raw data is converted into a jpeg preview.</p>

<p>Jpeg or not has no bearing on the captured dynamic range of the raw file, however it has an effect on the quality of the representation. So you should really capture 16 bit tiffs with the ProphotoRGB colorspace. The D810 have 14 bit raw files. I don't think you could dig out as much shadow information from the slides as a dedicated 16 bit film scanner. I only have limited slide scanning experience as I shoot and scan only B&W and color negs.</p>

<p>This is what Nikon says on flat picture control:</p>

<h3>Flat</h3>

<p>Flat provides minimal dramatization while preserving the material characteristics. Compared with Neutral, the finish shows less contrast and does not look as lively as it is. Flat is most often used when shooting video. When you add adjustment to the video after shooting, overblown highlights, blocked up shadows, or excessive color saturations rarely occur, thus enabling rich tonality of both brightness and color tones. With the wealth of information from highlight to shadow areas, this mode is recommended when you are shooting a scene with post-shoot adjusting in mind.<br /> The Flat Picture Control was added with the Nikon <a href="http://www.nikonusa.com/Nikon-Products/Product/dslr-cameras/D810.html">D810</a> DSLR and can be found in cameras introduced after the D810. To adjust the Flat Picture Control on images, use Picture Control Utility 2 which can be found in Nikon <a href="http://www.nikonusa.com/Nikon-Products/Product/Imaging-Software/Capture-NX-D.html">Capture NX-D</a> or Nikon <a href="http://www.nikonusa.com/Nikon-Products/Product/Imaging-Software/ViewNX-2.html">ViewNX2</a> software programs (as of June 26, 2014) or Nikon <a href="http://www.nikonusa.com/Nikon-Products/Product/Imaging-Software/ViewNX-i.html">ViewNX-i</a> software (as of March 2015).</p>

<p> </p>

[rodeo_joe1]

<p>Chuck, can I suggest that you simplify your setup? There's no need to have the flash off the camera. You simply mount the flash in the camera's hotshoe and point it forward over the copier attachment at a sheet of white paper or polystyrene positioned a couple of feet in front of the copier's diffuser. There's more than enough power in the flash to illuminate a slide copier by reflection from a reasonably close white surface.</p>

<p>I've used this simple setup successfully with a dedicated slide-copier attachment on a D800. The amount of flash "sneaking" into the slide holder directly is insignificant if you point the flash clear of the copier stage with a long zoom setting. You can even use a bit of black card to shield the slide-holder from direct flash if it proves to be problematic.</p>

[Ed_Ingold]

<p>There is another effect called "reciprocity law failure" which can occur in very brief (or very long) exposures. The light from a D900 flash is very brief at low power, as one would expect for something close to the subject - about 1/80,000 (maybe less). It's possible there is a dynamic effect in the Nikon sensor response.</p>

<p>The short answer is much simpler. If the image looks good under live view, take the shot without the flash, just using room light. If that works, keep doing it. Using a reflected desk lamp, my exposures at f/5.6 ranged from 1/60" to 4", depending on the slide. Camera shake is not an issue, because everything is screwed together in the setup (unless you deliberately shake it,hard). Most of my slides were taken long before I had a camera with automatic exposure (automatic anything), which resulted in a lot of variability. My copy camera was a Sony A7ii, which has a 13 stop dynamic range.</p>

[john_hinkey]

<p>I use a small light table rotated vertically and placed behind my PS-6 slide holder of my PB-6 belows to copy slides. Yes, the light source can make a difference. Also, the image on the slide has huge dynamic range, which is a much different thing that the dynamic range that the slide film can capture of a scene.<br>

Some time ago I did some comparisons of my CoolScan 5000ED slide scanner and D700 - the slide scanner won out for dynamic range, but lost due to flare at high contrast transitions. When I got my D800 I scanned the same slide (that had a huge dynamic range in the recorded image) and the D800 was slightly better than the CoolScan for dynamic range and had no flare issues. <br>

I don't recall, however, ever seeing more dynamic range in the LCD than I did in the final image - perhaps because of my continuous light source that was the same during both preview and image capture. Never did use a flash for the very reason of the inability to accurately pre-view the image w/o actually recording it.</p>

[mag_miksch]

<blockquote>

<h3 id="postTopic380821"><em>What is more, in Live View, the monitor on the back of the camera show acceptable contrast</em></h3>

</blockquote>

<p>Dont get the point, how to use life view together with flash?</p>

[chuck]

<p>The live view is not being used with flash. But I don't believe that should matter in determining the contrast of the image.<br>

I think the source of the light should not have any serious impact on the contrast present in the transmitted light through a slide. It should only seriously effect the overall exposure value and color balance. So I was puzzled why the sensor seemingly is able to capture the entire dynamic range present in the transmitted light through the slide before the picture is taken, but the picture review was not able to present the same range of information in the picture after it is taken.<br>

<br /> I think the explanation presented above, that in Live view, the monitor shows the entire dynamic range available to the sensor i, but in review mode it only shows the range present in a JPEG version of the image, makes sense. If the sensor can capture EV range of 13, transmitted light through the slide can have a range of 11, while JPEG only has a range of 8, then LV can show how the slide contrast ought to look when seen through with transmitted light, but picture review would lose a great deal of ends and exaggerate the contrast in the middle.<br>

Now I need to experiment with the adjusting the displayed dynamic range of NEF file on my computer monitor to see if I can make the monitor image match the contrast of the slide image as seen on a light table.</p>

<p>.</p>

[Ed_Ingold]

<blockquote>

<p>Dont get the point, how to use life view together with flash?</p>

</blockquote>

<p> You need some means of accurately centering and focusing the slide. Live view is more accurate than the optical finder in that regard. Even normal room light is enough to see the slide with live view. The flash is employed only during the exposure, when live view is momentarily shut down.</p>

<blockquote>

<p>I don't recall, however, ever seeing more dynamic range in the LCD than I did in the final image - perhaps because of my continuous light source that was the same during both preview and image capture.</p>

</blockquote>

<p>If you use the same light source for viewing and taking, there should be no difference between the LCD and the final image. That doesn't mean live view is identical to the slide. I my experience, the camera will adjust the exposure so that a dark slide appears lighter, etc. If you use AWB, it will also adjust the color balance slightly if there is a green (underexposed) or purple (overcast day) tint. It will also improve faded slides, which have an orange cast. I have found the camera to be much "smarter" than the average scanner, not to mention 10x as fast. RAW files leave ample room for further adjustments in Lightroom.</p>

[john_hinkey]

<blockquote>

<p>I have found the camera to be much "smarter" than the average scanner, not to mention 10x as fast. RAW <a id="itxthook8" href="/nikon-camera-forum/00dT9o?start=20" rel="nofollow">files<img id="itxthook8icon" src="http://images.intellitxt.com/ast/adTypes/icon1.png" alt="" /></a> leave ample room for further adjustments in Lightroom.</p>

</blockquote>

<p>+1<br>

I got far better colors, exposure, etc. using my D800 than with the CS5000ED, though the 5000ED excelled at dust and scratch removal!</p>