FROM: Eric Calderwood SUBJECT: Dmax DATE: Wed, 10 May 2000 22:54:38 +0000 ORGANIZATION: BT Internet NEWSGROUPS: comp.periphs.scanners Hi, Sorry if this has already been asked , but does anyone know how to calculate or measure the Dmax for a scanner ? IE The UMAX ASTRA 4000U and the new Microtek 1200 dpi scanner as the companies do not quote a Dmax for either of the above scanners. Thanks Eric C FROM: larry_roat@my-deja.com SUBJECT: Re: Dmax DATE: Wed, 10 May 2000 23:56:32 GMT ORGANIZATION: Deja.com - Before you buy. NEWSGROUPS: comp.periphs.scanners In article , Eric Calderwood wrote: > Hi, > Sorry if this has already been asked , but > does anyone know how to calculate or measure > the Dmax for a scanner ? > IE The UMAX ASTRA 4000U > and the new Microtek 1200 dpi scanner > as the companies do not quote a Dmax for either > of the above scanners. > > Thanks > Eric C Eric, This is a good question! I wish I knew the answer. I own a Umax Astra 4000U, and my best guess, and it is just a guess, would be between 3.0 and 3.2. The 4000U uses the same CCD as the PowerLook III, if that is any help. We really should put this question to Umax and MicroTek Support. The question is: What is the maximum optical density, and how is it measured? If you find out please let me know. I will do the same. Regards, Larry from California. FROM: "Ralf Guminski" SUBJECT: Re: Dmax DATE: Thu, 11 May 2000 11:42:05 GMT ORGANIZATION: @Home Network Canada NEWSGROUPS: comp.periphs.scanners Hi Lârry, et âl, The DMax value (X) is an exponent of 10, the number representing the density point where the CCD array registers the amount of light reaching it as black.(IOW the point at which it cannot effectively detect light) AFAIK the value is obtained through measurement. If you know the Dynamic Range (D) of the scanner in question you can guestimate that its DMax is somewhere close to that figure. (for example, some sources have quoted a dynamic range of 3.2 for the Perfection 1200, but its Dmax is 3.0. Neither figure is specified in the docs which come with the scanner.) Density of 0.0 is the end of the scale representing 100% of light being transmitted/reflected. Divide 100 by 10^X to get the percentage of light reaching the CCD that DMax represents. I've seen a range of D-0.0 to 5.0 quoted for a positive transparency (on glass IIRC). Umax figures on 3.7 as about maximum for regular film transparencies. (Win98 calculator has an exponent function if anyone wants to play with numbers) They didn't list your model unfortunately, but UMAX in UK has a chart comparing transparency DMAx for a number of their scanners. The highest DMax rating was for the Powerlook 3000, at 3.6. Some other models listed are: Astra 600/S6E -DMax 2.6 Astra 1200/S12 - 2.8 = 0.5% of light registers as black Gemini 3.0 = 0.1% of light registers as black Mirage 3.0 Powerlook II 3.2 Powerlook III 3.4 Mirage II 3.3 = 0.05% of light registers as black A scanner with Dmax 3.6 (0.025% of light) is 4 times as sensitive to low levels of light as one with Dmax 3.0. The full chart, and many other interesting details, is at: http://support.umax.co.uk/technotes/iq01.htm Cheers, Ralf FROM: "Mark Ligtenberg" SUBJECT: Re: Dmax DATE: Thu, 11 May 2000 21:45:08 GMT ORGANIZATION: Chello Broadband NEWSGROUPS: comp.periphs.scanners I have a flatbedscanner from Agfa, Snapscan 1236s, with a bit depth of 36 and a Dmax of 2,5. How is this possible? From a 36 bit scanner one would expect it to have a Dmax near 3,6 . Mark FROM: me@here.com SUBJECT: Re: Dmax DATE: 13 May 2000 19:48:02 GMT ORGANIZATION: Plug 'n' Pray! NEWSGROUPS: comp.periphs.scanners Mark Ligtenberg wrote: > I have a flatbedscanner from Agfa, Snapscan 1236s, with a bit depth of 36 > and a Dmax of 2,5. How is this possible? From a 36 bit scanner one would > expect it to have a Dmax near 3,6 . If everything (i.e. optics, CCD) *were* perfect. The Powerlook III does about 3.4 or 3.5 for a roughly $1k pricetag. FROM: Timo Autiokari SUBJECT: Re: Dmax DATE: Thu, 11 May 2000 20:19:07 +0300 ORGANIZATION: Clinet Ltd / Tele1 Europe NEWSGROUPS: comp.periphs.scanners Eric Calderwood wrote: >how to calculate or measure the Dmax for a scanner ? Dmax nor the density range can not be calculated, they need to be measured. There are targets available for this but they are somewhat expensive. Manufacturers do not like to publish the density values since the true, effective density range in general is not good with the scanners, usually much less than 8-bit, due to the noise from the CCD. Some manufactures simply specify the bit-depth of the A/D converter in terms of density. The formula for that is Log(2^bits) so the below density values are sometimes seen: 8-bit = 2,4D 10-bit = 3,0D 12-bit = 3,6D 14-bit = 4,2D But due to the CCD noise these values are false. It is the CCD that limits the dynamic range (or density range or bit-depth). It does not help anything if a CCD that has a density range of 2.0 is digitized by an A/D converter that has 12-bit gradation. The equipment still is only a 2.0D device. When you scan a perfectly homogeneous polished surface (with gamma 1.0 in the scanner driver settings) with scanner that is truly of 8-bit quality (2.4D range) then the resulting histogram over that area will be only three levels wide. Timo Autiokari FROM: David Chien SUBJECT: Re: Dmax DATE: Thu, 11 May 2000 15:09:36 -0700 ORGANIZATION: University of California, Irvine NEWSGROUPS: comp.periphs.scanners http://www.silverace.com/dottyspotty/issue7.html for an description of Dmax. As for measuring it, see Quantel's excellent description of scanning, Dmax, dynamic range, scanning film to digital, etc. at: http://www.quantel.com/digfilm/contents.htm specifically: http://www.quantel.com/digfilm/dyres.htm FROM: "Ralf Guminski" SUBJECT: Re: Dmax DATE: Fri, 12 May 2000 19:03:53 GMT ORGANIZATION: @Home Network Canada NEWSGROUPS: comp.periphs.scanners David's reference page at http://www.quantel.com/digfilm/dyres.htm ; paragraph 5, 1st and 2nd sentence: "It is often assumed that bit depth is related to dynamic range. This is not true." Terms relevent to the debate are "Reflective Density" (ZDNET tests?) and "Transmissive Density". Two different measurements, reflective being a considerably narrower range (<3) and transmissive apparently being the figure commonly quoted for scanner dynamic range and Dmax, as was specifically indicated in the case of the UMAX figures. This inference corroborated elsewhere, ie: http://www.idg.net/crd_resolution_83585.html See also articles relating to this debate at: http://www.shortcourses.com/book01/04-01.htm http://www.ccurrents.com/magazine/national/1704/puba21704.html http://www.image-acquire.com/articles/resscan.shtml And some useful info for beginners at: http://209.133.8.24/chapter04.htm#SCANNING BLACK MAGIC Cheers, Ralf FROM: Timo Autiokari SUBJECT: Re: Dmax DATE: Sun, 14 May 2000 12:46:05 +0300 ORGANIZATION: Clinet Ltd / Tele1 Europe NEWSGROUPS: comp.periphs.scanners On Fri, 12 May 2000 19:03:53 GMT, "Ralf Guminski" wrote: >Terms relevent to the debate are "Reflective Density" (ZDNET tests?) and >"Transmissive Density". Two different measurements, reflective being a >considerably narrower range (<3) and transmissive apparently being the >figure commonly quoted for scanner dynamic range and Dmax, The only difference is the method of illumination of the target, nothing else. The target holds some density values and the CCD records what it can from them. E.g: 2.0D in reflective means 1/100 is reflected. 2.0D in transmissive means 1/100 is transmitted. for the CCD the situation is exactly the same. Film does hold larger density range than a reflective print, but as long as the target shows a larger density range than what the equipment is capable to record, the measurement is valid. It does not depend on the type of the target. Now, better scanners can adjust their exposure, e.g. if the scanner has a density range of 2.4D it can e.g. detect a range from 0D to 2.4D or by changing the exposure it can detect a range from 0.4D to 2.8D. So, the Dmax and Drange are not the same for the better scanners that can adjust the exposure time. Density range and bit-depth: The density range of a capture device is the range between the most bright luminance level where detail is still recorded and the most dark luminance level where detail is still is recorded. What ever this range from bright to dark is, it can be digitized by what ever bit-depth, 16-bit, 12-bit, 8-bit or what ever, if there was such an A/D converter we could digitize it with say 1024 bits. This bit-depth is the same as the precision of the digitalization, it has nothing to do with the density range that the CCD can see. It has to do with the gradation, more bits give better gradation, smaller steps (so the precision is better). The density range of an acquire device only depends on the quality of the capture device, the CCD. It is limited by the electrical noise that is generated in the CCD, mainly the dark current noise, the thermal (random) noise and the readout-noise. All these noises are summed to the useful data (image information) so it is there all over the range. Due to the noise there is some smallest level of change of luminance that can be detected. It is generally accepted that when this smallest level of change is equal to the level of noise the signal is still being detected. So for example if the mean average of the noise is 2/256 from the maximum level of the system then it is considered to be a 1/(2/256)=128 level system, in other words it is of 7-bit effective quality (2^7=128). This in turn is the same as log10(128)=2.1D density range. Timo Autiokari FROM: timothy@clinet.fi SUBJECT: Re: Dmax DATE: Fri, 12 May 2000 06:57:29 GMT ORGANIZATION: Deja.com - Before you buy. NEWSGROUPS: comp.periphs.scanners Eric Calderwood wrote: > the Dmax for a scanner ? Here are good test results and comparisons for density for several desktop scanners: http://web-e6.zdnet.com/pcmag/features/scanners98/decaf005b.html Showing that the advertized density values are not anywhere near the true measured values. The other spec issue with the scanners is the resolution. There are also resolution test results at: http://web-e6.zdnet.com/pcmag/features/scanners98/decaf003b.html Showing that neither the the advertized resolution values are not anywhere near the true measured values. To convert the linepair/mm into the advertized (DPI) values: DPI = lp/mm * 2 * 25.4 Timo Autiokari Sent via Deja.com http://www.deja.com/ Before you buy. FROM: "Carsten J. Arnholm(sr)" SUBJECT: Re: Dmax DATE: Fri, 12 May 2000 10:06:06 +0200 ORGANIZATION: Telenor Online Public Access NEWSGROUPS: comp.periphs.scanners timothy@clinet.fi wrote in message <8fga0k$38v$1@nnrp1.deja.com>... >Eric Calderwood wrote: > >> the Dmax for a scanner ? > >Here are good test results and comparisons for density for several >desktop scanners: > >http://web-e6.zdnet.com/pcmag/features/scanners98/decaf005b.html These are all about flatbed scanner, cannot see any film scanner included. >http://web-e6.zdnet.com/pcmag/features/scanners98/decaf003b.html > Same here, no film scanner included. Regards, Carsten J. Arnholm(sr) Oslo, Norway. carnholm@online.no FROM: timothy@clinet.fi SUBJECT: Re: Dmax DATE: Fri, 12 May 2000 12:41:10 GMT ORGANIZATION: Deja.com - Before you buy. NEWSGROUPS: comp.periphs.scanners "Carsten J. Arnholm(sr)" wrote: > These are all about flatbed scanner, cannot see any film > scanner included. Firstly the original poster Mr. Eric Calderwood was referring to flatbed scanners (UMAX ASTRA 4000U and the new Microtek 1200 dpi scanner). Unfortunately these either are not included in that evaluation. Secondly the technology in regards with the density range and noise is the same with flatbeds and film-scanners. I have not found many evaluations about film-scanners but you should take a look at Mr. Tony Sleep's page at: http://www.cix.co.uk/~tsphoto/tech/filmscan/use.htm and follow there to his gray-scale evaluation at: http://www.cix.co.uk/~tsphoto/tech/filmscan/compare.htm That page does compare various film-scanners. E.g. the Minolta Scan Dual Dimâge seems to be notably noisy with -very- limited dynamic range. In order to extract more information from his scans you need to download them and inspect the range and noise e.g. in Photoshop. > >http://web-e6.zdnet.com/pcmag/features/scanners98/decaf003b.html > Same here, no film scanner included. Yes, it is pity that there are not many technical evaluations available. The very small step size and small target area imposes very difficult technical requirements for the mechanical and optical desing of film scanners, compared to the flatbed. I would not be surprised if a 2700dpi film scanner could only resolve something very similar than that a good 1200dpi flatbed does. Timo Autiokari Sent via Deja.com http://www.deja.com/ Before you buy. FROM: larry_roat@my-deja.com SUBJECT: Re: Dmax DATE: Sat, 13 May 2000 16:45:29 GMT ORGANIZATION: Deja.com - Before you buy. NEWSGROUPS: comp.periphs.scanners In article , Eric Calderwood wrote: > Hi, > Sorry if this has already been asked , but > does anyone know how to calculate or measure > the Dmax for a scanner ? > IE The UMAX ASTRA 4000U > and the new Microtek 1200 dpi scanner > as the companies do not quote a Dmax for either > of the above scanners. > > Thanks > Eric C Eric, I checked with Umax Europe http://www.umax.de/eu/Scanner/astra4000u.htm. They show a Reflective maximum optical density of 2.3D. My Astra 1220 has a Transmissive Dmax of 2.8, and Reflective a of 2.1. Astra 1220 Reflective 2.1Dmax, Transmissive 2.8Dmax. Astra 4000U Reflective 2.3Dmax, Transmissive ? Dmax. I do not have the transparency adapter for the 4000U yet, but when it comes I will measure the Transmissive Dmax--I still have my four color transmissive densitometer and step guide (T-27). Best of regards, Larry from California. ------------------------------------------------------ Sent via Deja.com http://www.deja.com/ Before you buy. FROM: "Ian Brown" SUBJECT: Dynamic Range-please explain? DATE: Thu, 18 May 2000 21:53:13 +1000 ORGANIZATION: A customer of Netspace Internet. NEWSGROUPS: comp.periphs.scanners At present when I scan photos (with my Umax 600s) I loose as detail in the darker areas(shadows) of the photos. I was told that I should get a scanner with a higher dynamic range (pref greater than 3.3). The confusing thing is that I have just done a comparison with the Canon 1200S (dynamic range 3.3) and the Epson 1200S (dynamic range of 3.0) The epson showed the darker areas much more clearly than the cannon (where dark clothes appeared to be black). Why is that so? Is dynamic range important? What criteria (appart from testing all the scanners) can I use to determine which ones show up details (especially in the shadows) better? Ian FROM: fulton@scantips.com (Wayne Fulton) SUBJECT: Re: Dynamic Range-please explain? DATE: Thu, 18 May 2000 14:01:07 GMT ORGANIZATION: . NEWSGROUPS: comp.periphs.scanners In article <8g0lke$2b69$1@otis.netspace.net.au>, igbrown@hotmail.com says... > >At present when I scan photos (with my Umax 600s) I loose as detail in the >darker areas(shadows) of the photos. I was told that I should get a scanner >with a higher dynamic range (pref greater than 3.3). > >The confusing thing is that I have just done a comparison with the Canon >1200S (dynamic range 3.3) and the Epson 1200S (dynamic range of 3.0) The >epson showed the darker areas much more clearly than the cannon (where dark >clothes appeared to be black). Why is that so? Is dynamic range important? >What criteria (appart from testing all the scanners) can I use to determine >which ones show up details (especially in the shadows) better? Price is often as good a criteria as exists, but only for large differences. Good measured dynamic range performance costs money, plenty of it, and is the primary difference in a $200 scanner and a $2000 scanner. But price says nothing if the difference is small, like $1300 vs $1500. Advertised dynamic range ratings often have no correspondance to reality, there are no specifications for how it should be advertised, what the number should mean, esp after marketing gets hold of it. Unfortunately, there is no meaning in the advertised numbers for dynamic range. We need some standards we dont have, which should involve actually scanning something. Dynamic range is very important for film, film has a much wider range than reflective prints. Dynamic range is essentially the zero-signal level performance of the scanner (black areas of film), and is an indicator of noise level in the electonics. Better scanners have better electronics and better noise levels, better dynamic range. Dynamic range is very signifcant scanning film, particularly slides, but normally of much less concern scanning reflective prints. If you are scanning film, consider a good top-end film scanner. Dynamic range gives noise free detail in the darkest areas of film. This is inverted to the hightlights of images from negatives, the problem is less visible there, and cheaper equipment often scans negatives better than slides. See Part 14 at site below. -- Wayne fulton@scantips.com http://www.scantips.com "A few scanning tips" FROM: larry_roat@my-deja.com SUBJECT: Re: Dynamic Range-please explain? DATE: Thu, 18 May 2000 15:01:24 GMT ORGANIZATION: Deja.com - Before you buy. NEWSGROUPS: comp.periphs.scanners ------------------------------------------------------- In article <8g0lke$2b69$1@otis.netspace.net.au>, "Ian Brown" wrote: > At present when I scan photos (with my Umax 600s) I loose as detail in the > darker areas(shadows) of the photos. I was told that I should get a scanner > with a higher dynamic range (pref greater than 3.3). > > The confusing thing is that I have just done a comparison with the Canon > 1200S (dynamic range 3.3) and the Epson 1200S (dynamic range of 3.0) The > epson showed the darker areas much more clearly than the cannon (where dark > clothes appeared to be black). Why is that so? Is dynamic range important? > What criteria (appart from testing all the scanners) can I use to determine > which ones show up details (especially in the shadows) better? > Ian Ian, Dmax is the blackest black in an image. There are two types of Dmax: Reflective and Transmissive. The Dmax most quoted by scanner companies is the Transmissive. Transmissive Dmax densitometer readings are higher than Reflective. Thus, Marketing uses the higher reading, which is a bit misleading. A Dmax of 2.1 to 2.5+ would be a more realistic Reflective Dmax value for a flatbed scanner. If you use a transparency adapter, and scan slides and negatives, you can expect higher Dmax values ranging from 2.6 to 3.4+. Why is a higher Dmax so important? Let's say that you have a photograph with a man hiding in the shadows. If you have a scanner that can only see at 1.8 Dmax, the man will not be visible in the scanned image. However, if your scanner has a Reflective Dmax of say 2.3, the man will be visible. According to Ed Hamrick, of VueScan fame, "Dmax can't be any higher than .301 times the number of bits in the A/D converter. For instance, a 10-bit scanner can't have a Dmax higher than 3.01. (.301 is the log base 10 of 2). In practice, the Dmax is often lower than this due to noise in the A/D converter or noise in the light source." All of this is nice to know, but it doesn't really help you choose the right scanner. My suggestion would be to visit these sites, and see what they have to say about the various flatbed scanners. http://www.image-acquire.com/scanner/index.shtml http://www.smartcomputing.com/ http://www.zdnet.com/pcmag/ By the way, I have a Umax Astra 4000U. I hope I have helped. Best of regards, Larry from California. FROM: Timo Autiokari SUBJECT: Re: Dynamic Range-please explain? DATE: Thu, 18 May 2000 21:58:36 +0300 ORGANIZATION: Clinet Ltd / Tele1 Europe NEWSGROUPS: comp.periphs.scanners On Thu, 18 May 2000 21:53:13 +1000, "Ian Brown" wrote: >At present when I scan photos (with my Umax 600s) I loose as detail in the >darker areas(shadows) of the photos. I was told that I should get a scanner >with a higher dynamic range (pref greater than 3.3). For photos the minimum that is needed is about 2.4 since photographs do not go under that (do not have darker areas than that) but the higher the density range is the lesser is then noise. So high dynamic range is always beneficial. >The confusing thing is that I have just done a comparison with the Canon >1200S (dynamic range 3.3) and the Epson 1200S (dynamic range of 3.0) The >epson showed the darker areas much more clearly than the cannon (where dark >clothes appeared to be black). Why is that so? Is dynamic range important? Dynamic range is very important but the specs are not true, they are just something that the manufacturers likes to put in there so that it looks good, the spec that is. It is specmanship. >What criteria (appart from testing all the scanners) can I use to determine >which ones show up details (especially in the shadows) better? Unfortunately there is no good criterias for that. Those equipments that do output higher bit-depth data are usually better than those that only output 24-bit data. But a 12-bit A/D conversion (that gives 36-bit data) does not mean that the analog information from the CCD that is converted by the A/D converter would be anywhere that good, unless the CCD is of very high quality and cooled way below the freezing point. Devices that cost 10k to 20k USD do output about 9-bit/channel true or real bit-depth. There is some trouble in experimenting the dynamic range of the scanners (especially reflective scanning) since there is not many reflective targets available that have a wide range. I just today experimented with the Kodak Q-13 density scale and a piece of density filter. I had part of every patch on the Q-13 covered by the density filter (plastic foil, horizontally) so the upper part shows the original densities from 0D to 1.9D and the portion that was covered by the density filter shows those densities reduced by the density filter. I only had a 0.75 density filter that is not enough for this, but this method seems to work so one can have a very accurate reflective target with quite a large range. If you like you can see a scan of the Kodak Q-13 at: http://www.aim-dtp.net/aim/calibration/index.htm click the link: -What is Middle Gray-. It is a rater inexpensive target, very good to have, helps a lot in evaluations and calibration. Timo Autiokari FROM: Viken Karaguesian SUBJECT: Minolta Scan Dual and Dynamic Range questions... DATE: Fri, 09 Jun 2000 18:21:43 -0400 ORGANIZATION: MindSpring Enterprises NEWSGROUPS: comp.periphs.scanners Hi all. Can someone tell me the Dynamic Range of the Minolta Scan Dual? I've seen ads for it at places like B&H Photo, but the dynamic range never seems to be mentioned. I've seen the dynamic range specs for the other Minolta scanners (Elite, Scan Speed) but not for the Scan Dual. Is this spec not advertised because it is poor? Also, can someone also explain to me how the dynamic range rating system works? What's the real-world difference between a scanner with a 3.2 DR and one with 3.0? Is that small of a DR difference noticeable? I'm guessing that the Dynamic Range is the total range of tone (contrast?) that the scanner is sensitive to. So a scanner with a 3.6 DR can handle much more contrast than one with 3.2, correct? Thanks for any replies. Viken Karaguesian FROM: "Ralf Guminski" SUBJECT: Re: Minolta Scan Dual and Dynamic Range questions... DATE: Sat, 10 Jun 2000 18:22:15 -0400 ORGANIZATION: Posted via Supernews, http://www.supernews.com NEWSGROUPS: comp.periphs.scanners "Viken Karaguesian" wrote in message news:39416DF7.F9D7FC76@mindspring.com... > Also, can someone also explain to me how the dynamic range rating > system works? What's the real-world difference between a scanner with a > 3.2 DR and one with 3.0? Is that small of a DR difference noticeable? It is a logarithmic scale, 10^DR... therefore, comparing DR 3.0 to DR 3.3 we get 10^3= 1000 versus 10^3.3= 1995. IOW a scanner with D 3.3 is almost twice as sensitive to low levels of light as one with D3.0. Dynamic range of 3.2 is about half again as sensitive as a dynamic range of 3.0. Cheers, Ralf FROM: "Carsten J. Arnholm(sr)" SUBJECT: Re: Minolta Scan Dual and Dynamic Range questions... DATE: Sun, 11 Jun 2000 23:25:39 +0200 ORGANIZATION: Telenor Online Public Access NEWSGROUPS: comp.periphs.scanners Viken Karaguesian wrote in message <39416DF7.F9D7FC76@mindspring.com>... > Hi all. Can someone tell me the Dynamic Range of the Minolta Scan >Dual? I've seen ads for it at places like B&H Photo, but the dynamic >range never seems to be mentioned. I've seen the dynamic range specs for >the other Minolta scanners (Elite, Scan Speed) but not for the Scan >Dual. Is this spec not advertised because it is poor? Dynamic Range, or rather, dmax for the Minolta Dimâge Scan Dual is not in the official specification. I have been informed it is 3.2. Bear in mind though that there is no official standard for measuring, let alone describing such a feature in promotional material. The Minolta Dimâge Scan Dual is a reasonably good scanner for slides and an excellent one for negatives, based on my own personal experience. Regards, Carsten J. Arnholm(sr) Oslo, Norway. carnholm@online.no FROM: larry_roat@my-deja.com SUBJECT: Re: Minolta Scan Dual and Dynamic Range questions... DATE: Sun, 11 Jun 2000 23:29:10 GMT ORGANIZATION: Deja.com - Before you buy. NEWSGROUPS: comp.periphs.scanners Hello Viken, Check out http://www.image-acquire.com/scanner/html/minosdual.shtml for the answer to Dmax question. There has been a lot of discussion about Dynamic Range, or Dmax (shadow or blackest black) in this and other groups. The normal target range for color separations from a transparency is .40 highlight, 1.30 midtone, and 2.40 shadow. I have a transmissive densitometer, and this information is from my Kodak T-27 literature. A Minolta Dimâge Scan Dual $400 is a bargain. Combine it with Ed Hamrick's VueScan software, and you will have an excellent scanner, especially for scanning negatives. Regards, Larry from California. FROM: "Mark McGilvray" SUBJECT: Dynamic Range DATE: Sat, 25 Nov 2000 15:56:14 GMT ORGANIZATION: Verio NEWSGROUPS: rec.photo.equipment.large-format I would like to know the technical definition of dynamic range and how it is calculated. I have an intuitive/empirical idea, but would definitely like to refine this concept. Thank you. FROM: Jean-David Beyer SUBJECT: Re: Dynamic Range DATE: Sat, 25 Nov 2000 11:49:30 -0500 ORGANIZATION: Institute for Regimented Whimsey NEWSGROUPS: rec.photo.equipment.large-format Mark McGilvray wrote: > > I would like to know the technical definition of dynamic range and how it is > calculated. I have an intuitive/empirical idea, but would definitely like to > refine this concept. Thank you. It is the difference between the maximum signal the system can transmit (usually without distortion) to the minimum signal that can be transmitted (usually, the residual noise level). For digital signals, this would still be true, but there is normally no distortion if the system is properly set up so the maximum signal that can be transmitted is that represented by the largest number that can be digitally encoded, and the noise level is normally around the amount of the least significant bit that is transmitted (it gets more complex when serious bandwidth compression strategies are employed, such as differential PCM of various types). -- .~. Jean-David Beyer Registered Linux User 85642. /V\ Registered Machine 73926. /( )\ Shrewsbury, New Jersey ^^-^^ 11:45am up 3 days, 11:40, 1 user, load average: 2.05, 2.06, 2.01 FROM: "Mark McGilvray" SUBJECT: Re: Dynamic Range DATE: Sat, 25 Nov 2000 19:19:07 GMT ORGANIZATION: Verio NEWSGROUPS: rec.photo.equipment.large-format "Jean-David Beyer" wrote in message news:3A1FED9A.4026C2B4@exit109.com... > It is the difference between the maximum signal the system can > transmit (usually without distortion) to the minimum signal that can > be transmitted (usually, the residual noise level). > > For digital signals, this would still be true, but there is normally > no distortion if the system is properly set up so the maximum signal > that can be transmitted is that represented by the largest number that > can be digitally encoded, and the noise level is normally around the > amount of the least significant bit that is transmitted (it gets more > complex when serious bandwidth compression strategies are employed, > such as differential PCM of various types). OK. Is this a logarithmic scale? How does this apply to the dynamic range specifications applied to scanners, and photographic applications in general? How is this measured for a scanner? Thanks. FROM: "Sherman Dunnam" SUBJECT: Re: Dynamic Range DATE: Sat, 25 Nov 2000 21:21:36 GMT ORGANIZATION: Giganews.Com - Premium News Outsourcing NEWSGROUPS: rec.photo.equipment.large-format "Mark McGilvray" wrote in message news:LgUT5.3476$A26.174692@sjc-read.news.verio.net... > OK. Is this a logarithmic scale? How does this apply to the dynamic range > specifications applied to scanners, and photographic applications in > general? How is this measured for a scanner? Thanks. > Dynamic range in photography is the difference between the density of the lightest and darkest areas of a negative. This is measured on a logarithmic scale, with each .3 increase in dynamic range being equivalent to about 1 stop. A good b&w negative has a dmin of .3 or .4 (base plus fog) and a dmax of 3.6 to 3.8 yielding a dynamic range of around 3.4. It is possible to exceed these numbers on both ends of the scale but that is rare. In order to capture all the detail that exists in a negative a scanner must have a dynamic range of at least 3.4. The quality of a scan depends not only on the dpi and the levels of gray available but also on the dynamic range. For instance a scanner with a dynamic range of only 3 will lose either the shadows or the highlights depending upon how it is scanned. Your finished scan may still have 256 levels of gray but not nearly enough detail. It is also possible to use a scanner with less dynamic range and do two scans, one for the shadows and one for the highlights. Combining these two scans in Photoshop you can get the full dynamic range of the negative. I haven't tried this so I can't say if other detail is lost. Maybe someone else can comment on that. HTH, -- --- Sherman Dunnam www.flyfishingjournal.com Free fly fishing software, feature articles and more! FROM: "MK70" SUBJECT: Re: New Nikon Scanners - Drawbacks?? DATE: Sun, 14 Jan 2001 00:48:19 +0800 ORGANIZATION: Hong Kong Boardband Network Ltd. NEWSGROUPS: comp.periphs.scanners >Is there any way to corroborate the claim of D-Range of 4.2? >How much difference 4.2 would make over 3.6 in practical terms? >Thanks, >Balu Most of the scanner manufacturer measure Dmax and Dmin on the B/W film output of a gray scale. You may create a gray scale image and output it to a film recorder to B/W film. D-range = D-max - D-min D-min means the film base color that is usually 0.3 Therefore, the D-range of Nikon may be actually only 3.9 However, some highend and drum scanner manufacturers use zero as the starting point for D-min. It is hard to say which one is correct, and is up to individual manufacturer to decide whether zero or 0.3 should be used. The Nikon D-range may be misleading because it may means D-max instead of D-range. Nevertheless, D-max of 4.2 is a very impressive figure and beat those drum scanners that is 20 to 30 times more expensive than the Nikon. The Fuji Crosfield 5250 and 6250 drum scanner has a D-max of "only" 3.8. Does it mean that the Nikon is better than the Crosfield drum scanner that is expensive, industry accepted, and charged $50 to $100 per scan by the service bureau? I can't believe it !!! Different manufacturers use different techniques in measuring D-Range. I believe they are using different sophisticated instruments to measure the light transmitted from the film and compare the data read from the CCD. In theory, you may make you own scanner target using output of gray scale to a B/W film in a film recorder. You can compare the result with the same scan image from a highend flatbed or drum Scanner. If the Nikon claim is true, the image from the Nikon and the image from a drum scanner should be about the same. You should check carefully at the light and dark tone, there should clear separation at these areas. Here is some figures of those expensive scanners. The one with * means they use a zero Dmin instead of 0.3 Dmin. Drum Scanners ------------------ Heidelburg S3900 D-range 4.2 (0.3 to 4.5) Dmax 4.5 Tango D-range 3.9 (0.3 to 4.2) Dmax 4.2 Fuji (Crosfield) 5250/6250 D-range 3.9 (0 to 3.9) Dmax 3.9 * Howtek HiResolve 8000 and Scanmaster 4500 D-range 4.2 (0 - 4.2) Dmax 4.2 * Optronic Color Getter D-range 4.0 (0 to 4.0) Dmax 4.0 * ScanMate 11000 D-range Up to 4.0 (no info about Dmax and Dmin) ScanMate 3000 D-range Up to 3.6 ( I don't recommend you to test this one) Highend Flatbed Scanners -------------------------------- Heidelburg Nexscan D-range 3.7 (0.3 to 4.0) Dmax 4.0 Scitex Eversmart Supreme D-range 4.0 (0.3 to 4.3) Dmax 4.3 Scitex Eversmart Pro D-range 3.7 (0.3 to 4.0) Dmax 4.0 Fuji C550 D-range 3.7 (0 to 3.7) Dmax 3.7 * ScanMate F10 / F8-Plus / F6 D-range 3.7 (0.3 to 4.0) Dmax 4.0 Screen Cezanne FT-S5000 Dmax 3.9 (no D-range or Dmin info) FROM: z0011@horga.de (Andreas Frackowiak) SUBJECT: Re: Scanfehler / Bilderraetsel DATE: 11 Mar 2001 17:23:56 GMT ORGANIZATION: Zok's Kingdom NEWSGROUPS: de.rec.fotografie Hallo Thomas, Thomas 'TomV' Steffen wrote: > z0011@horga.de (Andreas Frackowiak) writes: > Das heitere Rätselraten kann beginnen. Also, wir wissen: > * SW-Negativ Jep. Agfa Ortho 25 > * Color-Scanner Jep. Nikon LS30 > * Probleme mit Helligkeit-Kontrast > * Überstrahlen der dunklen (!) Partien .... die im Negativ die hellen Partien sind. > * Flecken auf dem Negativ Jugendsuenden .... ;-) > * Farbstich Wegen versehentlich vom Scanner weggerechneter Orangemaske. > * Quantisierung > Die Kontrastprobleme rühren vermutlich daher, dass das Negativ sehr > steil ist (über 3 Dichten oder so), was bei Negativen normalerweise > nicht vorkommt (wegen Gamma 0.6). Jep. Der Agfa Ortho 25 hat nen Gamma > 2,5 und eine Dichte weit ueber 3, vermutlich auch jenseits der 4 (logD). Der Film ist zum Scannen ziemlich gemein, weil er kaum Mitteltoene hat, sondern die Towerte sich hauptsaechlich bei der maximalen Dichte oder minimalen Dichte befinden. Deswegen wurde er frueher auch gerne zur Reproduktion von Strichzeichnungen verwendet. > Das erklärt gleich die > Kontrastprobleme und die Quantisierung auf Bild 3 (Grenze der > Farbtiefe). Jep. > Die ganzen Unterschiede zwischen Bild 1 und 3 entstehen > offensichtlich erst per Software, nicht in der Hardware. Doch, direkt in der Hardware. Die drei Bilder unterscheiden sich nur in anderer "Analogverstaerkung" (jeweils 2 LW), und wurden ansonsten nicht bearbeitet. > Folglich hätte ich 2 Tipps ins Blaue: > a) Andere Software. Deine Scansoftware kann nicht mit Negativen > umgehen. Oder ist falsch konfiguriert. > b) Anderen Scanner mit höherem d_max und diffuser Beleuchtung. Klar. Die Hardware kommt mit dem extremen Kontrast des Films nicht klar. Ich habe nun nochmal 2 weitere Scans (diesmal Scans als Positiv, manuelle Umkehrung in PS, und einmal die Analogverstaerkung so weit herunter- geregelt, dass es keine Ueberstrahlungen gibt) auf der Seite abgelegt, und bin nun doch (fast) zum gewuenschten Ergebnis gekommen. Es bleiben aber noch 2 Fragen offen: 1. Warum gibt es bei: http://www.komplix.de/raetsel1/Seiten/elke3.htm so extreme waagerechte Streifen links neben dem Maedel, und warum nicht auf: http://www.komplix.de/raetsel1/Seiten/elke5.htm ? Bei beiden ist die Analogverstaerkung auf maximum eingestellt, und die Ueberstrahlung/Uebersteuerung muesste doch identisch sein, oder ? Nr. 5 Wurde als Diapositiv gescannt und dann in PS von mir invertiert. 2. Wenn ich Dias scanne (in denen helle Stellen durchsichtig sind) bemerke ich keinerlei Ueberstrahlungen, wenn ich obige Negative (ebenfalls als Dias) scanne, ueberstrahlen die hellen Stellen (die auf dem Film ebenfalls durchsichtig sind) auf die dunkleren Bereiche. Beidesmal ist der Scanner identisch eingestellt. Gruesse Andreas -- "Never send a human to do a machines job". FROM: w.j.markerink@a1.nl (Willem-Jan Markerink) SUBJECT: Dmax bei S/W (war: Scanfehler / Bilderraetsel DATE: Sun, 11 Mar 01 18:36:24 GMT NEWSGROUPS: de.rec.fotografie In article <98gcbc$otl$5@news.ins.net>, z0011@horga.de (Andreas Frackowiak) wrote: >> Die Kontrastprobleme rühren vermutlich daher, dass das Negativ sehr >> steil ist (über 3 Dichten oder so), was bei Negativen normalerweise >> nicht vorkommt (wegen Gamma 0.6). > >Jep. > >Der Agfa Ortho 25 hat nen Gamma > 2,5 und eine Dichte weit ueber 3, >vermutlich auch jenseits der 4 (logD). Der Film ist zum Scannen >ziemlich gemein, weil er kaum Mitteltoene hat, sondern die >Towerte sich hauptsaechlich bei der maximalen Dichte oder >minimalen Dichte befinden. Deswegen wurde er frueher auch gerne >zur Reproduktion von Strichzeichnungen verwendet. Hmm....welche S/W-Filme haben eine gleichartige, oder sogar groessere Dmax? Ich hatte neulich eine Diskussion in die Leafscan Mailingliste, wo behauptet wuerde dass Dia's den groessten Dmax haetten, Negative viel weniger, aber S/W ebenfalls viel weniger....wo ich doch sicher war das man gerade mit Silber sehr groesse Dichten erziehen kann, und der Scanner erst richtig an die Arbeit muss.... -- Bye, Willem-Jan Markerink The desire to understand is sometimes far less intelligent than the inability to understand [note: 'a-one' & 'en-el'!] FROM: "Markus Keinath" SUBJECT: Re: [F] an Mr. Agfa (war:Re: Dmax bei S/W (war: Scanfehler / Bilderraetsel)) DATE: Mon, 19 Mar 2001 22:52:41 +0100 NEWSGROUPS: de.rec.fotografie Hm, ich habe das ursprüngliche Posting nicht gelesen, hab´ich was verpasst? Der TMax 100 läßt sich schon auch auf ein D-Max von 4 bringen. Der 400er ebenso. Kodak Technical Pan auf noch mehr. Soviel mal dazu, Markus