I looked at Velvia 50 under a microscope today

[mauro_franic]

<p>Here Scott, I marked them for you:</p><div></div>

[dave_redmann]

<p>Mauro, I'm with Scott--if you placed the ruler at the series of lines that you regard the smallest ones truly resolved, you have about 6.7 pairs in the space between what appears to be the 0 and the 10, representing 10/100 of a mm, so that would be 67 lp/mm. Note also that the contrast is low; if you break them down into hue, saturation, and value, the black line should have a value of 0, and the white line 255. But by my sampling, the average value of the "black" line is more like 42 and the "white" is 65--pretty low contrast.</p>

<p>If you want to say you can see a little contrast on the next closer lines pairs, maybe just a tad; if you've created this little demonstrative correctly, that's at about 85 lp/mm. But the contrast is extremely low, and would almost certainly provide no real pictorial detail.</p>

[mauro_franic]

<p>Not much room for subjectivity.</p>

[mauro_franic]

<p>Dave, I can easily count lines to the left of the 150 lines/mm I marked in green. Velvia clearly resolves over 150 lines/mm in this scan.</p>

<p>Not sure I understand the disagreement.</p>

[dave_redmann]

<p>Okay, <strong>wait a minute</strong>--Mauro says 160 <em>lines</em> per mm, Scott says 67 <em>line pairs</em> per mm--we are only disagreeing by a factor of 19%.</p>

[mauro_franic]

<p>Here is at 100% with some levels (to avoid any distortion on my previous image I posted at 200%).</p><div></div>

[dave_redmann]

<p>By the way, I will admit that the fact the the lines are getting continuously farther apart as the chart proceeds to the right means that counting lines in all 10/100 of the mm lowers the number a bit. On the other hand, the ones at the right end of the ruler are clearly better resolved than the ones at the left end. So if you want to say that the scan demonstrates, oh, 70 lp/mm, that is, 140 lp/mm, or whatever, I won't quibble. So Mauro and Scott, I basically agree with both of you, and think you are misunderstanding each other with the "lines" versus "line pairs" issue, that is, one line pair is two lines.</p>

 

[swilson]

<p>Mauro,<br>

In your scan the best I see is right around 8 pixels/line pair, if that image is really 8000 ppi then it means you are getting 1000 line pair/inch or around 40/mm. But at the same time I am reading on the scale what should be 0.10 mm as 0.20, did you scale the image up by a factor of 2? 40 line pair/mm seems too low, 80 would make more sense. In no way can I get 160 line pairs / mm out of your image.</p>

[mauro_franic]

<p>Scott, I am talking about lines per mm.</p>

<p>160 lines per mm = 80 line pairs per mm.</p>

[c._sharon]

<p><em>I'd like to see some one show me a digital sensor that can match the resolution of 4x5. </em></p>

<p>Some years ago a comparison between 4x5 Velvia and a 39 MP P45 digital back on Luminous Landscape showed only a slight advantage to 4x5 film. Since then new digital backs like P65 have come out that can very well give 4x5 film a run for its money, but I am not aware of any tests with it.</p>

[jeff_livacich]

<p>Saying this as a film user, Curtis, what the hell are you talking about?<br>

Are you going to use 4x5 for sports or other fast action? Superwides or long teles? 4x5 cameras are suited to certain things. They are not "the best". By your definition, why stop there? Why not 8X10, or Ultra Large Format?<br>

4x5 has always "destroyed" 35mm film, so where's the difference? It surpasses medium format as well, by virtue of larger physical size.<br>

What's "fair" about your comparison? Compare like to like. If I compared a digital at ISO 800 to Astia, would it be fair for me to complain about its noise? No more than it would be for me to complain about grain in high speed film compared to Astia.<br>

I don't use digital, except a little P&S for snapshots, but I would sure hate for anyone to get the impression that we film users are all as silly as your statement.</p>

[swilson]

<p>Just how much resolution and detail does someone need? Even after around 6-8 MP how many photos would benefit form more resolution?<br>

With digital there is no real limit as to resolution, it is more a matter of how much is needed. But if resolution is your thing then 4x5 is pretty limited by today’s standards of what a high resolution photo is. <br>

Here is a 158 MPixel image that I took, it is not one of my highest resolution photos buy any means but even to it has way more than is needed.<br>

<a title="lighthouse small by KonaScott, on Flickr" href=" src="http://farm5.static.flickr.com/4079/4802443535_d7b4d38200_o.jpg" alt="lighthouse small" width="700" height="605" /></a><br>

And a 100% crop shown below. <a title="lighthouse crop by KonaScott, on Flickr" href=" src="http://farm5.static.flickr.com/4117/4802443629_e248d22f94_o.jpg" alt="lighthouse crop" width="700" height="700" /></a><br>

But I find that going over around 20 MP to be a waste, in fact an 8 MP can be just fine in most cases, but really if it is raw detail you are after then digital is the way to go, IMO.</p>

[matthew_stott]

<p>I'm just impressed that Scott Wilson had a sunny day in coastal Oregon when he took that photo!</p>

[Dave Luttmann]

<blockquote>

<p><a href="../photodb/user?user_id=4958594">C. Sharon</a> , Jul 17, 2010; 06:01 p.m.</p>

</blockquote>

 

<blockquote>

<p><em>I'd like to see some one show me a digital sensor that can match the resolution of 4x5. </em><br>

Some years ago a comparison between 4x5 Velvia and a 39 MP P45 digital back on Luminous Landscape showed only a slight advantage to 4x5 film. Since then new digital backs like P65 have come out that can very well give 4x5 film a run for its money, but I am not aware of any tests with it.</p>

 

</blockquote>

<p>Nothing like needing to spend $30,000 to equal a $2 sheet of film ;-)</p>

 

[denniswilliams]

<p>Actually in over three decades of work I have never looked at any film through anything stronger than a loupe.</p>

[david_henderson]

<blockquote>

<p>Nothing like needing to spend $30,000 to equal a $2 sheet of film ;-)</p>

</blockquote>

<p>But if you wanted to take 15 000 photographs over the life of the back it might be worthwhile then.</p>

<p>And you get to avoid buying, storing, carrying and processing film, and for many people, the cost and hassle of scanning. And you don't have to worry about ther space to store the negs/transparencies.</p>

<p>Seems like a good deal to me if you're a volume user and you can raise $30 000. </p>

[chris_burgess3]

<p>If only all this concentration and struggle about taking film and digital resolution to ridiculous heights were used for something truly productive, like developing CO2 capture structures near deposition points and deciding which path to take with clean energy and deploying it, I would take humanity more seriously....</p>

[thomas_lozinski]

<p>Scott,<br>

What are you shooting with? </p>

[swilson]

<p>I was using a Canon 350D with the 50mm f/1.8. I was using a panoramic head and stiched with PTGui. It took 3 minutes to take the photos needed for the stitching, which really is not bad, but I have been reviewing what kind of resolution I really need.<br>

<br />I started stitching not so much for higher resolution but rather a widerFOV, which it does great at. But the resolution part is very easy to get at the same time and can get a bit additive. However there are very few uses for an image that is over 20MP, and so now I have been taking more photos with my 28mm lens. With the 28mm lens it takes less then a minute to get my shots, in fact more like 45 seconds. The big advantage with using the 28mm lens is that I get a lot more DOF for the same f/number. <br>

Here is a photo I took, my wife and I run on this road in the morning and so I wanted to capture the feel of the road. I took 26 panoramic shoot along the road, of which this is one of. <br>

<a title="pan7 rect 50mp 06-26-10 by KonaScott, on Flickr" href=" src="http://farm5.static.flickr.com/4100/4805328132_46c52260fd.jpg" alt="pan7 rect 50mp 06-26-10" width="500" height="391" /></a><br>

I can stitch this a number of ways but I like this one. I ended up with a nice image with 50 MP, the full image can be seen here<br>

<a href=" MPixel image</a><br>

The photo was stitch from photos taken with my 28mm lens, I could have used my 50mm lens and gotten a lot more resolution but what would the point be. </p>

<p>For myself it was never the resolution of film that I had a problem with it was a whole bunch of other issues. Trying to convince me that film is better because it can capture more detail is not going to get anywhere because detail is not the problem with most images.</p>

 

[steve m smith]

<blockquote>

<p>For myself it was never the resolution of film that I had a problem with it was a whole bunch of other issues. Trying to convince me that film is better because it can capture more detail is not going to get anywhere because detail is not the problem with most images.</p>

</blockquote>

<p>Film or digital can only be better in context. For you digital is better because you do a lot with it in digital form. For me film is better because there is no computer involved.</p>

[swilson]

<p>For me there was more computer work involved using film since I scanned my film and Ihave found scanned film needing a lot more post processing then digital. And the reason I scanned my film was for me I could get much better print with scanned film. The other reason for me to scan all the film that I care about is that just about all of my film is either fading or molding or both so if I want to be sure to have my photos in the future scanning is pretty much a must.</p>

<p>If I were doing B/W things might be different but I had by fill of B/W in the 60s and 70s.</p>

[kelly_flanigan1]

<p>Dave; any *REAL* honest lens test uses *PAIRS*; that is what one really looks at; ie a black and white PAIR of bars. This goes back to before WW2; when bull*hitting was not really acceptable.</p>

<p>The same goes if one sells Shoes; socks; gloves; skis; it is understood by honest folks it means pairs. Ie Kilroy sold 5 shoes today; it means 5 pairs.</p>

<p>Today spinelless BS rules; thus injecting a 2X factor is done by lay folks. Dumb folks; con artists; and lay folks love to use BS numbers; thus injecting a 2X factor is common today.</p>

<p>It really has no place in research; ie marketing; cons.</p>

<p>lay folks will boast they sold 10 shoes; or got 100 line per mm; when it is often just 5 pairs of shoes; or 50 line pairs per mm. Newbies and hucksters like big numbers; thus injecting a BS 2X factor is todays con.</p>

<p>Few folks really get even 50 to 60 line pairs per mm on film in the real world with a sharp negative.</p>

[swilson]

<p>Kelly,<br>

What you say about few folks getting even 50 or 60 line / mm may be true but it misses what the OP was addressing which was the limits of film resolution due to dye cloud size.</p>

<p>I do agree that counting both the black and white lines is misleading however.<br>

<br />I think the OP is a bit off on the dye cloud and that it is less then 10 microns, it is hard to tell but I would put it at closer to 6-8 microns. The problem with giving a size to something like a dye cloud is that it does not really have hard edges and there is a large variation from cloud to cloud. In the end trying to figure the limit of film resolution by looking at dye clouds is probably not the best way to go.</p>

[mauro_franic]

<p>The dye clouds must be martially smaller than 6, 8 or 10 microns to resolve 160 lines (80 line pairs) per millimeter. </p>

<p>6 micros would be the maximum size the could be if I could magically align the bars on my tests with the dye clouds. Since that is obviously impossible on a random distribution; the dye clouds could only be 3-4 microns at most. (3 times smaller than the OP's assessment).</p>

[mauro_franic]

<p>This post is not a competition with digital, it is a discussion on Velvia dye clouds.</p>

<p>In 2010, most landscape shooters use either MF film (6x7 Velvia or B&W) or DSLRs (10-20+ MP). Hence the vast majority of landscape pictures taken on film far out resolve the landscape pictures taken with digital. </p>

<p>Film landscape shooters who choose 4x5 over 6x7 do it because of the movements and tonality, not because of resolution - since no one scans 4x5 at 8,000 dpi.</p>

<p>On digital systems the most you can get today is 60MP at $40,000 from B&H. A lot less resolution than just MF film and at a joke of a price.</p>

<p>It makes no sense to choose digital if the interest is resolution alone. But it does when convenience or fast shooting is needed.</p>

<p>As an alternative, as Scott mentioned, one may use stitching (digital or film) for still subjects that allow for it.</p>