It all comes down to the print

[mauro_franic]

<p>As Stuart mentions, grain is also introduced with faster films. The lack of grain in the film older times was an indicative of either low speed iso or medium format prints.</p>

<p>It wasn't until well after the introduction of digital cameras that people realized that they could use sophisticated digital tools like Neat Image or Noise Ninja to remove grain (a matter of personal choice) the same way digital cameras remove noise.<br /> <br /> Especially if a high resolving scanner is used to provide good sampling for the software algorithm, the results are consistent. Nonetheless, digital today can produce less noise at higher iso than the grain that is produced by high iso films. The exception being TMAX 400 that produces no visible grain on prints 16x20 and larger when shot at iso 800 or 1600.</p>

[mauro_franic]

<p>I personally experienced a time when I bougth my first 2MP camera when I thought the results were matching my wet prints of 35mm (up until 8x10 ish). I was bewildered by its smoothness to a point the lack of detail was irrelevant. It was just the honeymoon period that ended quickly though.</p>

<p>Digital technology definitely added a set of tools to film processing that were not available in the past.</p>

[mauro_franic]

<p>Stuart,</p>

<p>I have a model that generates the spectrum of detail resolved (used to produce the carts above).</p>

<p>If you want it, let me know and I'll email it to you. You can input the resolution at different contrast for any film or digital camera and you can obtain the spectrum on the print for comparison.</p>

[StuartMoxham]

<p>Not talking about anything to do with a v500 Les why would you assume that. I would not even bring a v500 into this. Mauro himself found the smoothness in prints from a 2mp digital yes detail would be lacking from a 2mp file but he found the prints were grainless as many other have found.</p>

[Landrum Kelly]

<p>Mauro, I think that it is time to retire the 40D as your point of comparison. It was not even the best example to use when you first used it. Now it is so far behind that your continued inclusion of it in your examples damages your credibility.</p>

<p>--Lannie</p>

[Landrum Kelly]

<p>Even more damaging, Mauro, is the claim that 5.6 megapixels is the "true" resolution of the 40D.</p>

<p>--Lannie</p>

[mauro_franic]

<p>Now the above is comparing to film.</p>

<p>Here is the result of introducing a scanner in the process:</p>

[mauro_franic]

<p>On the V500 (my own test):</p><div></div>

[c._sharon]

<p>Resolving Power vs Contrast numbers - What is the basis for these numbers? Any MTF chart shows lower contrast response at higher frequencies (lp/mm) while the table shows the opposite - higher lp/mm for higher contrast.<br>

Further, it seems the number for the DSLRs are the jpeg absolute resolution from DPR's tests and are being compared with the extinction resolution of Velvia (where do the 3500 and 4000 for 35mm Velvia come from?). The raw extinction resolution of 7D is around 3800 LP/PH according to DPR's review of 7D. The number for FF DSLR would be correspondingly higher.</p>

[mauro_franic]

<p>On the Coolscan 9000:</p><div></div>

[mauro_franic]

<p>C. Sharon,</p>

<p>All tests for high contrast with the exception of the 20MP DSLR (DPR) are my own. This means I personally tested a Canon 40D, TMAX 35mm, Velvia 35mm, TMAX 6x7, Velvia 6x7, Coolscan 9000, Epson V500 and printed on an Epson 24" 7880. All these results were published.</p>

<p>The values for the film's low contrast response are from the 1 to 1.6 contrast resolution from the Kodak charts.</p>

<p>The mid contrast values are interpolated from the curves published by Kodak in the technical documents.</p>

[mauro_franic]

<p>If your tests show you different values of resolution, please let me know what they are and I will enter them in the model and publish the results using your own values. </p>

<p> </p>

[Landrum Kelly]

<p>Forgive me, Mauro, but I share C. Sharon's skepticism. I am not saying that your numbers are wrong, just that I don't think that they are completely objective. In any case, the 7D may be almost twenty MP, but its resolving power is surely below that of the 5D II.</p>

<p>I have no particular axe to grind in all this. I like film. I love beautiful MF film work. I just doubt that you are convincing too many people with these data.</p>

<p>--Lannie</p>

[mauro_franic]

<p>Lannie,</p>

<p>The true resolution of the 40D as tested by me is aprox 6MP. </p>

<p>I am not concerned about my credibility. LOL.</p>

<p>If your tests give you different results, what resolution did you obtain with a 20MP camera and a 20MP have in actual lines per picture height resolved?</p>

<p>If your results are higher than mine, then my test will be wrong and I will replace the values. I must say though, that I have tested that camera so thoroughly that I'd be shock if your results are more than 5% or 10% higher than mine. Please share them though.</p>

[mauro_franic]

<p>Lannie,</p>

<p>No need to apologize. I know everyone has good intentions to share experiences and advance conclusions.</p>

<p>These are facts though. The only people who need convincing are those who don't regularly print at 24".</p>

<p>Please tell me what values I should use in the model. I 'll generate it for you.</p>

[mauro_franic]

<p>I just need this:</p>

[mauro_franic]

<p>This is the format I need the values from your test (or understanding):</p><div></div>

[Landrum Kelly]

<p>Mauro, I'm going out to shoot.</p>

<p>--Lannie</p>

[c._sharon]

<blockquote>

<p>The values for the film's low contrast response are from the 1 to 1.6 contrast resolution from the Kodak charts</p>

 

</blockquote>

<p>I suppose you are aware that the 1.6:1 number is for contact printed film - that is a directly printed sine wave on film without using a lens. So you cannot use it as it is to compare with the resolution test numbers for DSLRs from DPR because the latter are sensor+ lens numbers.</p>

<p>What you need to do is use the Fuji resolving power equation 1/r_film+lens = 1/r_film +1/r_lens. From Fuji's datasheet for Velvia the resolution at 1.6:1 contrast ratio is 80 lp/mm. Pairing it with an excellent lens of 120 lp/mm you get 1/r_film+lens = 1/80 + 1/120, or r_film+lens = 48 lp/mm which works out to 48x2x24 = 2304 LP/PH. This is lower than the raw tested result for any 12 MP DSLR by DPR.</p>

[c._sharon]

<blockquote>

<p>The EOS 450D is listed at 2300 X 2200 - <a rel="nofollow" href="http://www.dpreview.com/reviews/canoneos450d/page33.asp" target="_blank">http://www.dpreview.com/reviews/canoneos450d/page33.asp</a> As I pointed out above, rotated, the numbers fall by 10%.</p>

 

</blockquote>

<p>Again, you are quoting the jpeg absolute resolution. The jpeg extinction resolution is 2500 in the review, and the raw number would be higher. DPR started publishing resolution test number for raw after they reviewed EOS 450D.</p>

<blockquote>

 

</blockquote>

<blockquote>Of course, more importantly, real results are available and theories usually have a "fudge factor" that you must not have accounted for. Either that or your "idealized" digital resolution are not being realized.</blockquote>

<p>I provided some theory in response to numbers being quoted which are questionable. Daniel's "real results" in this thread speak for themselves.</p>

[mauro_franic]

<p>C. Sharon,</p>

<p>To clarify. Please tell me what is the resolving power for:</p>

<p>35mm Velvia high contrast<br>

35mm TMAX high contrast<br>

35mm Velvia low contrast<br>

35mm TMAX low contrast</p>

 

[mauro_franic]

<p>C. Sharon,</p>

<p>I agree results speak for themselves. Theory, especially when you don't validate it with your own tests is less valuable.</p>

<p>Daniel's results from the 7D are the best I have seen.</p><div></div>

[c._sharon]

<blockquote>

<p>Please tell me what is the resolving power for: 35mm Velvia high contrast<br /> 35mm TMAX high contrast<br /> 35mm Velvia low contrast<br /> 35mm TMAX low contrast</p>

 

</blockquote>

<p>You can get these numbers from the MTF curve in datasheets of the respective films. For low contrast look at MTF curve and the lp/mm at which the MTF is 30%. For high contrast, you can look at the lp/mm for MTF50 or higher.</p>

<p>Please note that these numbers are for contact printed film - you will have to use the Fuji resolving power equation to get the film+lens resolution.</p>

<p>Velvia<br>

MTF 30 - 80 lpmm (low contrast)<br>

MTF 50 - 48 lp/mm<br>

MTF 70 - 35 lp/mm</p>

<p>TMX 100<br>

MTF 30 - 180 lp/mm (low contrast)<br>

MTF 50 - 125 lp/mm<br>

MTF 70 - 95 lp/mm</p>

[mauro_franic]

<p>Thank you. I think you are mixing lines per mm and line pairs in you numbers above.</p>

<p>I don't want to make any assumptions, so could you please give your understanding of resolution in lines per picture height in a 35mm strip in this format:</p>

<p>(PLEASE STATE NUMBERS THAT YOU BELIEVE WILL MAKE IT INTO THE FILM - AFTER LENS AND ALL YOUR CONSIDERATIONS).</p>

<p>35mm Velvia high contrast: ___<br /> 35mm TMAX high contrast: ___<br /> 35mm Velvia low contrast: ___<br /> 35mm TMAX low contrast: ___</p>

[mauro_franic]

<p>As a guideline, for high contrast your numbers should be equal or higher than what I get from my Coolscan:</p>

<p>Velvia Scan (CS 9000) 35mm = 3650 lines per picture height<br /> TMAX Scan (CS 9000) 35mm = 3850 lines per picture height</p>