 |
 | Home | Contact Us | FAQ | Search & Site Map | Link to Us |
 | Sign In | Join | Other 45 Sites in Network |
| PhotoKB HomeDiscussion Groups Digital Photography Digital PhotoDSLR CamerasZLR CamerasPoint & Shoot CamerasFilm Photography 35 mmLarge FormatMedium formatDarkroomFilm and LabsOther EquipmentPhoto Technique Nature PhotographyPeople PhotographyTechnique GeneralGeneral Photo Topics General TopicsAustralian PhotographyUK PhotographyDirectoryPhoto Clubs |
Photo Forum / General Photo Topics / Australian Photography / June 2006Interesting Thread on aus.bushwalking.....
See : How Much Do Digital Photos Weigh ? | See : How Much Do Digital Photos Weigh ? good for a chuckle, but let's be realistic.. mass of: laptop or XS-drive + batteries >/=/< film ? one wonders.. thinks for a moment: a teeny $100 35mm rollei loaded with 25/50/100/200/400 ISO still produces the same theoretical 'filesize' image as a top of the line 35mm SLR (what does a 5000 ppi scan yield) what sort of DSLR weight must one consider to give the same file? must weigh a bit.. k > thinks for a moment: a teeny $100 35mm rollei loaded with 25/50/100/200/400 > ISO still produces the same theoretical 'filesize' image as a top of the > line 35mm SLR (what does a 5000 ppi scan yield) > > what sort of DSLR weight must one consider to give the same file? must > weigh a bit.. But why on earth would you compare a DSLR to a 35mm rangefinder? There are digital PS camera's just as small as the Rollei with 8M pixels or more. Samsung even have a camera phone with 12M pixels. Of course it's quality is unlikely to match the Rollei, that's for sure. (BTW where do you get a Rollei for $100 anyway???) And a 60GB portable Hard Drive will certainly hold a lot more photo's than the equivalent size/weight of 35mm film, regardless of how many Mpixels your camera is! 8GB CF cards would do even better of course, although at a much higher price. MrT. | > thinks for a moment: a teeny $100 35mm rollei loaded with | 25/50/100/200/400 [quoted text clipped - 3 lines] | > what sort of DSLR weight must one consider to give the same file? must | > weigh a bit.. | But why on earth would you compare a DSLR to a 35mm rangefinder? the image filsize as I stated. 5000pixels x 75000 pixels = 37 1/2 Mp at a 5000 ppi scan. There are | digital PS camera's just as small as the Rollei with 8M pixels or more. | Samsung even have a camera phone with 12M pixels. they've released that now? I could only get the 7Mp! - and only if I spent $1000 USD :( it looked more like a camera than a phone too ;) Of course it's quality is | unlikely to match the Rollei, that's for sure. | (BTW where do you get a Rollei for $100 anyway???) Photo swapmeet. bought 3 at under $100, one as low as $40 in Perth. They still turn up from time to time | And a 60GB portable Hard Drive will certainly hold a lot more photo's than | the equivalent size/weight of 35mm film, regardless of how many Mpixels your | camera is! 8GB CF cards would do even better of course, although at a much | higher price. that's true ;) let's up the stakes and talk 4x5 or 8x10 now ;) ;) > | But why on earth would you compare a DSLR to a 35mm rangefinder? > > the image filsize as I stated. 5000pixels x 75000 pixels = 37 1/2 Mp at a > 5000 ppi scan. So which DSLR takes 37.5 Mpixel photos, that you are comparing to? Surely you need to compare to a medium or large format digital back *IF* you want that many *real* pixels. Of course you are just blinded by numbers, and don't understand whether they are all useful information. A scan from that Rollei is unlikely to provide anything like 37Mpixels of *useful* information. > There are > | digital PS camera's just as small as the Rollei with 8M pixels or more. [quoted text clipped - 3 lines] > $1000 USD :( > it looked more like a camera than a phone too ;) Pity the photo's look more like those taken by a phone than a camera :-) > Of course it's quality is > | unlikely to match the Rollei, that's for sure. > | (BTW where do you get a Rollei for $100 anyway???) > > Photo swapmeet. bought 3 at under $100, one as low as $40 in Perth. They > still turn up from time to time Wait a few years and the current digital camera's will be selling for that price S/H at swap meets too. There are already in fact. Still, a Rollei RF for $40 *is* a better buy, no argument from me. > let's up the stakes and talk 4x5 or 8x10 now ;) ;) You already were (37 *real* Mpixels) you just didn't know it. There haven't been all that many bush walkers use 4*5" or 8*10" plate camera's for a *long* time though. (Nor that many landscape photographers either :-) MrT. "Mr.T" < | > the image filsize as I stated. 5000pixels x 75000 pixels = 37 1/2 Mp at a | > 5000 ppi scan. | So which DSLR takes 37.5 Mpixel photos, that you are comparing to? you pick one, I'll compare. | Surely you need to compare to a medium or large format digital back *IF* you | want that many *real* pixels. | Of course you are just blinded by numbers, and don't understand whether they | are all useful information. hahahaha. Interpolated pixels V film data, that's rich! when you say 'real' pixels, do you mean real? ;) | A scan from that Rollei is unlikely to provide anything like 37Mpixels of | *useful* information. 100-300lpi from tech pan, mate, that's plenty of info even the scanner ain't seeing! | Pity the photo's look more like those taken by a phone than a camera :-) haven't seen any images yet to tell - where have you seen them, are they on the web available for download | Wait a few years and the current digital camera's will be selling for that | price S/H at swap meets too. I sold a rough as guts Nikon DCS 620 2 Mp camera for $355 USD not six months back. What happens to all these digital cameras that are obsolete? | > let's up the stakes and talk 4x5 or 8x10 now ;) ;) | | You already were (37 *real* Mpixels) you just didn't know it. nope, that was 35mm and I did know it. 500 ppi scan a 4x5 - then an 8x10 what, the laest 4x4 digi back is .. $? | There haven't been all that many bush walkers use 4*5" or 8*10" plate | camera's for a *long* time though. | (Nor that many landscape photographers either :-) must just be me then :/ k > | > the image filsize as I stated. 5000pixels x 75000 pixels = 37 1/2 Mp at > | > 5000 ppi scan. > > | So which DSLR takes 37.5 Mpixel photos, that you are comparing to? > > you pick one, I'll compare. *You* were the one who said you had to compare to a DSLR, name one that does 37Mpixels? > | Surely you need to compare to a medium or large format digital back *IF* > you want that many *real* pixels. [quoted text clipped - 3 lines] > hahahaha. Interpolated pixels V film data, that's rich! when you say > 'real' pixels, do you mean real? ;) The jokes on you if you think film has infinite resolution, or that Rollei lens either. > | A scan from that Rollei is unlikely to provide anything like 37Mpixels of > | *useful* information. > > 100-300lpi from tech pan, mate, that's plenty of info even the scanner ain't > seeing! 300 lines per inch? Could do that with a cheap digital. But how often do you use Tech Pan anyway? > I sold a rough as guts Nikon DCS 620 2 Mp camera for $355 USD not six months > back. Lucky you, PT Barnum was right of course. >What happens to all these digital cameras that are obsolete? Dustbin I imagine. > | > let's up the stakes and talk 4x5 or 8x10 now ;) ;) > | > | You already were (37 *real* Mpixels) you just didn't know it. > > nope, that was 35mm and I did know it. Nope, you fail to realise that Rollei will not get anywhere near the equivalent of 37 Mpixels, even with Tech Pan. A Canon 1DS mkII would probably beat it at a third of that, simply upsample if you really want extra useless pixels. Try it, then get back to us with the results. MrT. | 300 lines per inch? Could do that with a cheap digital. | But how often do you use Tech Pan anyway? often enough.. still have 100 feet of 4" and 50 feet of 9" left too :) | > I sold a rough as guts Nikon DCS 620 2 Mp camera for $355 USD not six | months | > back. | | Lucky you, PT Barnum was right of course. funny - the guy was buying it for it's known strength - glass over the mirror and a good colour fidelity.. seems they're highly prized for such things.. COmpared to ongoing CCD cleaning I cans ee why. | > | > let's up the stakes and talk 4x5 or 8x10 now ;) ;) | > | [quoted text clipped - 4 lines] | Nope, you fail to realise that Rollei will not get anywhere near the | equivalent of 37 Mpixels, even with Tech Pan. how do you figure that? a 5000 ppi scan isn't good enough, or the maths are wrong? | A Canon 1DS mkII would probably beat it at a third of that, simply upsample | if you really want extra useless pixels. | Try it, then get back to us with the results. again, comparing pretend-O-matic® pixels to real data is typical digi-nonsense. k > | Nope, you fail to realise that Rollei will not get anywhere near the > | equivalent of 37 Mpixels, even with Tech Pan. > > how do you figure that? > > a 5000 ppi scan isn't good enough, or the maths are wrong? So if you do a 10 million ppi scan, it will be even better right? How thick you are to not understand the limitations of simply adding more pixels to a limited resolution photo. But *IF* you *only* want drum scans of Tech Pan photo's then bloody well use it and let the rest of us get on with taking the photo's we want. No need to pretend it is better than a view camera. > | A Canon 1DS mkII would probably beat it at a third of that, simply > upsample [quoted text clipped - 3 lines] > again, comparing pretend-O-matic® pixels to real data is typical > digi-nonsense. Just as I said. *YOUR* 5000 ppi scan is just as "pretend-O-matic" in most cases, but you just don't understand, no matter how many times someone tries to point it out. Just keep carrying that 8*10" view camera, loaded with Tech Pan, on your bush walks. Then you can make some good, contact prints, and forget about the Rollei :-) (contrast may not suit everything though, but as long as you are happy, I'm sure I don't care) MrT. "Mr.T" me: | > hahahaha. Interpolated pixels V film data, that's rich! when you say | > 'real' pixels, do you mean real? ;) | | The jokes on you if you think film has infinite resolution, or that Rollei | lens either. no one mentioned infinite res except you.. 100-300 lpi is real resolution. explain moire. digital 'resolution' is sharpening algorithms applied to fuzzed images projected onto bayer arrays, the RGGB mono images then being interpolated to make a good guess of the scene.. k > no one mentioned infinite res except you.. So tell us exactly what the resolution limit of the Rollei is, using Tech pan, and how many pixels would be necessary to achieve the same result? (You stated 37 Mpixels was not enough in fact.) Now tell us what the figures are for color neg or tranny? (BTW there are tests available for film, and someone has probably tested the Rollei too I imagine) > 100-300 lpi is real resolution. Exactly, but do you know what it means? It doesn't appear so. > digital 'resolution' is sharpening algorithms applied to fuzzed images > projected onto bayer arrays, the RGGB mono images then being interpolated to > make a good guess of the scene.. What a load of mangled crap. All digital photo's have a "resolution" limit before ANY sharpening algorithms are applied to the captured data. All film cameras and film also have a resolution limit before any scanning or printing is done. MrT. | > no one mentioned infinite res except you.. | | So tell us exactly what the resolution limit of the Rollei is, using Tech | pan, and how many pixels would be necessary to achieve the same result? | (You stated 37 Mpixels was not enough in fact.) thats correct, at a 5000 ppi scan rate that's a scan of the film.. whatever it may have on it. | Now tell us what the figures are for color neg or tranny? let's see, Agfa ultra has a stated res of 140 lpmm, but moving on - I understand the Rollei lens in question has a res of 96lpmm. so that 96 is the limiting factor here.. 96x24.5 = 2350 lpi res limitation on the optical system so you're right - 2300 ppi is the max functional res I can expect to see out of a scan so let's limit the scan to that. 2350x3525 = 8mp bugger! oh more here about film resolving power.. http://www.normankoren.com/Tutorials/MTF1A.html | > 100-300 lpi is real resolution. | | Exactly, but do you know what it means? It doesn't appear so. yup.. | > digital 'resolution' is sharpening algorithms applied to fuzzed images | > projected onto bayer arrays, the RGGB mono images then being interpolated [quoted text clipped - 3 lines] | What a load of mangled crap. All digital photo's have a "resolution" limit | before ANY sharpening algorithms are applied to the captured data. what would that be from deliberately fuzzed light hitting monochromatic photo receptors behind a series of RGGB filters, prey-tell? k > | So tell us exactly what the resolution limit of the Rollei is, using Tech > | pan, and how many pixels would be necessary to achieve the same result? > | (You stated 37 Mpixels was not enough in fact.) > > thats correct, at a 5000 ppi scan rate that's a scan of the film.. whatever > it may have on it. So how many more "pretend-O-matic" pixels are necessary by *your* calculation? In fact you seem to be saying that even if the film is blank you need more than 37 Mpixels to represent the information. Is that correct? If not you may want to address the question I raised as to how much actual, useful information is on the film. MrT. | > | So tell us exactly what the resolution limit of the Rollei is, using | Tech [quoted text clipped - 11 lines] | If not you may want to address the question I raised as to how much actual, | useful information is on the film. did that, read the post above this one.. <snip> > let's see, Agfa ultra has a stated res of 140 lpmm, but moving on - I > understand the Rollei lens in question has a res of 96lpmm. [quoted text clipped - 5 lines] > > bugger! Oops. Bad news for you, K. The combined resolution of any film/lens combination is less than the either film or lens resolution separately. Forgive the maths, but the combined res. is calculated by (film res. times lens res)/(film res plus lens res) so for your figures, it would be 140*96/(140+96), = 13440/236, = 56.9, say 57 lines/mm. That gives 57*25.4 (not 24.5), = 1447 lpi, so a scan at 1200 lpi is close, and 2700 lpi is overkill. However, I wonder if the Rollei lens is actually 96 line pairs/mm, in which case the combined resolution will be 140*192/(140+192), = 80.9, say 81 - which sounds a bit more like it, but is still less than your 96. Colin D.  SignaturePosted via a free Usenet account from http://www.teranews.com | <snip> | [quoted text clipped - 15 lines] | 57*25.4 (not 24.5), = 1447 lpi, so a scan at 1200 lpi is close, and 2700 | lpi is overkill. no, my fault - I had forgotten that.. | However, I wonder if the Rollei lens is actually 96 line pairs/mm, in | which case the combined resolution will be 140*192/(140+192), = 80.9, | say 81 - which sounds a bit more like it, but is still less than your | 96. good points. makes you wonder why anyone drumscans anything, hey? yet myself and other folks get better results from their 4000 (and greater) dpi scanners than the older 2700's. wonder why that is k > makes you wonder why anyone drumscans anything, hey? > yet myself and other folks get better results from their 4000 (and greater) > dpi scanners than the older 2700's. > wonder why that is Probably because scanners have optical systems as well. Just because a scanner claims 2700ppi, doesn't mean all of them are perfect. Obviously a 4000ppi scanner *probably* is more likely designed to achieve 2700 or more ppi of actual information. (if they aren't total BS artists anyway) A 2700ppi scanner is unlikely to achieve perfection. However I have seen tests showing clearly some 2700ppi scanners outperforming some 4000ppi models, especially when you compare all factors, not just resolution. For example Dmax is critical for slide scanning, (or your Tech Pan). In *general* though the newer 4000ppi scanners are better than the older 2700ppi scanners in most areas, and most drum scanners better still. Nothing too surprising really. You do seem to keep getting bogged down on numbers though, without thinking what they actually mean in practice. MrT. | Probably because scanners have optical systems as well. Just because a | scanner claims 2700ppi, doesn't mean all of them are perfect. Obviously a [quoted text clipped - 10 lines] | You do seem to keep getting bogged down on numbers though, without thinking | what they actually mean in practice. you are quite right.. been too much going on here lately and I am quick to jump the gun with half remembered fragmnts of info. I have a 1200 ppi scanner that exceeds the quality of my 4000 ppi scanner by miles - but then it's a 3 colour pulsed xenon job :) the dmax issue with this beat is clearly overcoming the limitations of the cold cathode scanner given the amount of light it belts through! and the quality is startlingly different from that of any other scanner I've owned. (it's an A3 film scanner) A friend in the UK set about testing the Dmax's of a number of scanners by attempting to scan a razorblade - figuring he could also learn something about the errors introduced by diffraction. Not a lightweight, he's had quite a bit of experience in this area but tossed the whole task in the bin concluding it was such a hit and miss affair, the variance between scanners of the same model, hardware, interference, software obfuscation and clipping .. k > | <snip> > | [quoted text clipped - 35 lines] > wonder why that is > k Not necessarily - Scanned one image on two occasions had the same result off the Nikon 2000 (2700dpi) and the Minolta Multi scan (4800dpi) They were for big prints. The results printed off the same printer same size. The film being the limiting factor in this case (100asa slide - Nikon lenses) > | <snip> > | [quoted text clipped - 29 lines] > > makes you wonder why anyone drumscans anything, hey? Linearity and Dmax. > yet myself and other folks get better results from their 4000 (and greater) > dpi scanners than the older 2700's. > > wonder why that is > k One major factor is grain aliasing. I have an older 2700 ppi scanner (Acer Scanwit) that just doesn't handle fast fils well, the aliased grain looks like sand in the scans, but fine-grained images do scan well. I currently (very occasionally!) scan with my current flatbed, a Canon 9950F, at up to 4,800 ppi. The images are smoother, but no sharper than the old Scanwit. Colin D. SignaturePosted via a free Usenet account from http://www.teranews.com back to the digital cameras for a tic .. I was thinking again last night and more fragments were surfacing - some CCD/CMOS 's have as little as 70% of the surface covered in photosensors, but ignore that.. looking at something like the coolpix 995 for a moment we have a sensor that's 7.2x5.3mm and yields a 2048x1536 pixel image. that's around 285 lpmm resolution. let's assume the lens has a res of (optimistically) around 85 lpmm so (AxB) / (A+B) = approximately 65 lpmm total optical resolution. forget demosaicing and the rest of the digital stuff for the moment - I'm just trying to get my head around the actual resolution of this 3Mp camera. If there is an optical resolution of 65 lpmm then (65 lpmm x7.2mm) x (65 lpmm x 5.3mm) = ( 468 x 344 ) = 0.16Mp of real resolution does this sound right? k > back to the digital cameras for a tic .. > [quoted text clipped - 22 lines] > > k The surface %age affects the efficiency of the sensor in terms of light gathered, but of course doesn't affect the sensel pitch, which determines the ultimate theoretical definition the sensor can deliver. As for the lens sharpness, I think 85 lines/mm might be a bit on the lean side for the Coolpix's lens, but even at double that, it's still only going to produce 106 l/mm, which works out to 7.2*5.3*106^2, = about 0.43Mp. It doesn't sound right to me, however, but unless the lens is much sharper than we imagine, I can't see how the images would be sharp at all, considering a 6x4 print represents a >20x enlargement. A 20x lift of a 106 l/mm image will give a print sharpness of only ~5 lines/mm, and clearly the Coolpix can do better than that. Search me. Colin D. SignaturePosted via a free Usenet account from http://www.teranews.com "ColinD" | The surface %age affects the efficiency of the sensor in terms of light | gathered, but of course doesn't affect the sensel pitch, which [quoted text clipped - 10 lines] | of a 106 l/mm image will give a print sharpness of only ~5 lines/mm, and | clearly the Coolpix can do better than that. Search me. the sensor resolution limit ("Nyquist limit") is actually lower than that I calculated too, it's calculated as only 145 lpmm for this camera, with the max res on an 8x10 being thus a mere 3.8 lpmm. lets blame the algorithms then ;) k | The surface %age affects the efficiency of the sensor in terms of light | gathered, but of course doesn't affect the sensel pitch, which | determines the ultimate theoretical definition the sensor can deliver. | It doesn't sound right to me, however, but unless the lens is much | sharper than we imagine, I can't see how the images would be sharp at | all, considering a 6x4 print represents a >20x enlargement. A 20x lift | of a 106 l/mm image will give a print sharpness of only ~5 lines/mm, and | clearly the Coolpix can do better than that. Search me. punching this all in again with your input (thanks ColinD) Mr T said it was pointless scanning film with high ppi scanners as the data produced was 'junk' data - stuff that contributed nothing to the image. It was suggested that a 4000 ppi scanner offered no benefit above a 1200 ppi scanner for high quality 35mm film.. let's assume that is accurate for the moment. Using the calculation that the limit of an optical system is based on the equation (lens res x film res) / (lens res + film res) I'll drop more numbers in. If I pop in the figure for tech pan as 100 lpmm (it can go as high as 300 so I understand) and pick an arbritrary lens res value of 80 lpmm for a 35mm lens. 100x80 / 100+80 = 44.4 lpmm So I have a piece of film 24mm x 36mm so that's 1065 vertical lines x 1598.4 horizontal lines = 1,703,255 total 'blocks' of resolution. lets call it 1.7 Mega-'blocks' of resolution. If I call these line blocks 'pixels' (blocks of data) , that's a resolving power equating to a mere 1.7Mp (!) Sure the image has lots of detail and looks great, but the actual *resolving power* in line blocks, pixels, whatever is pretty small. OK, so I went an looked at some digital sensors to see what they could resolve in real terms.. forget demosaicing and the rest of the digital stuff for the moment - I'm just trying to get my head around the actual resolving power of that 3Mp Coolpix 995 camera again for comparison to the 35mm film example above. looking at the coolpix 995 for a moment we have a sensor that's 7.2x5.3mm and yields a 2048x1536 pixel image. that's around 285 lpmm resolution. To eliminate further complications I'll also forget the sensor resolution limit should also consider the "Nyquist" limit, which reduces the actual resolution substantially (it's calculated as only 145 lpmm for this sensor) - agin let's forget this for a moment. let's assume the lens has a res of around 80 lpmm again so (AxB) / (A+B) = approximately 62 lpmm total optical resolution. (niquist limit calcuations in place, the actual resolution is really limited to 51 lines per mm, but I said we'd forget that..) If there is an optical resolution of 62 lpmm then (62 lpmm x7.2mm) x (62 lpmm x 5.3mm) = ( 446.4 total lines x 3428.6 total lines) = 0.14 Mega'blocks' (Mp again?) of real resolution The D30 Canons 27.7 x 15.1mm sensor yields 2160 x 1440 pixels or 95.4 pixels per mm Assume a lens of 80 lpmm res again, that means an overall res of 43.5 lpmm. total resolution possible then is 43.5x22.7 x 43.5x15.1 = 0.64 Mega'blocks' or megapixels jumping up the quality lines LOTS, let's look at the Canon EOS-1Ds Mark II with a maximum resolution of 4992 x 3328 pixels covering a 24x36mm area. 4992/36 = 138.7 pixels (lines) per mm 3328/24 = 138.7 pixels (lines) per mm again using the same 80lpmm lens, we have a total maximum limit to resolution (again not considering the Nyquist limit) of 51 lines per mm 51 lpmm x 24 = 1224 lines vetical resolution 51 lpmm x 36 = 1836 lines horizontal resolution 1836 x 1224 = 2,247,264 mega'blocks' or 2.2 Mp in real terms, including the Nuiquist limit in the calcs, the resolution is actually half this or 1.1 Mp of real resolving power in the optical system, still less than that of the good ol' tech pan. If one can assume then that a stated 16Mp camera has a little over half the actual resolving power of the tech-pan, could one not be excused from concluding the tech pan shot is equal to a 24Mp image? (1.4x16) Would not one then conclude a 4000 ppi scanner producing a 4000x6000 pixel (24Mp) scan is actually gathering legitmate data? just thinking out loud.. karl > Mr T said it was pointless scanning film with high ppi scanners as the data > produced was 'junk' data - stuff that contributed nothing to the image. *NOT* what I said at all. I did say 37 Mpixels from a Rollei RF was just a joke though. > was suggested that a 4000 ppi scanner offered no benefit above a 1200 ppi > scanner for high quality 35mm film.. Not me who said that either. I use a 4000ppi film scanner. Even those reduce quality to some degree. The idea is to minimise the quality reduction in each step, which requires the imaging device to be better than the original, not just the same. > If one can assume then that a stated 16Mp camera has a little over half the > actual resolving power of the tech-pan, could one not be excused from > concluding the tech pan shot is equal to a 24Mp image? (1.4x16) Possibly not far off, with Tech Pan, under ideal conditions. (although that 16Mp image may not have 16M legitamate pixels either) > Would not one then conclude a 4000 ppi scanner producing a 4000x6000 pixel > (24Mp) scan is actually gathering legitmate data? Yes, under ideal conditions, maybe. But that still leaves the Rollei RF, with any normal film, a long way short of 37 real Mpixels :-) MrT. | > Mr T said it was pointless scanning film with high ppi scanners as the | data | > produced was 'junk' data - stuff that contributed nothing to the image. | | *NOT* what I said at all. I did say 37 Mpixels from a Rollei RF was just a | joke though. | > If one can assume then that a stated 16Mp camera has a little over half | the [quoted text clipped - 3 lines] | Possibly not far off, with Tech Pan, under ideal conditions. | (although that 16Mp image may not have 16M legitamate pixels either) rue. Actually a good friend pointed out that the fault in my calcs (which really are quite abstract) lies in my assumption that resolving power can be reduced to lines vertical x lines horizontal - suggesting that could ONLY work for digital if there were 2x pixels sampling each line vertical or horizontal (thus 2x pixels were needed for each line) ..but in fact if the lines to resolve were at 45 degrees then I would need 2.5xsrt2 or 3.5 pixels per line. thus the resolving power equivalents for the theoretical test shot is reduced to: tech pan = 2400 total lines coolpix = 578 total lines 1Ds MkII = 941 total lines So again, if we can accept that a 16Mp camera produces 16Mp of (non-junk data) image, which a lot of digital enthusiasts seem able to do, then 16Mp /941 lines = 0.017 0.017 x the known real 2400 lines of tech pan = 40 Mp so this means.. ;) | > Would not one then conclude a 4000 ppi scanner producing a 4000x6000 pixel | > (24Mp) scan is actually gathering legitmate data? | | Yes, under ideal conditions, maybe. and a 5000 ppi scan of the same technical pan yields 5000x7500 pixels = 37Mp | But that still leaves the Rollei RF, with any normal film, a long way short | of 37 real Mpixels :-) not really.. k > so this means.. ;) Who the hell knows! > and a 5000 ppi scan of the same technical pan yields 5000x7500 pixels = 37Mp Of course, we still don't know what the quality of that data is though. A 5000ppi scan of a blank negative yeilds the same number of pixels. You still have a number fixation. Lots of bogus calculations with *NO* actual measurements to prove whether any of your assumptions are actually valid. We can see you don't have a scientific background anyway. Not sure that you have proven anything else as yet though. MrT. From: "Mr.T" : > so this means.. ;) : [quoted text clipped - 4 lines] : : Of course, we still don't know what the quality of that data is though. you know the difference between analogue (infinite progressive steps) and digital (discrete steps) ? If a pixel could represent a silver halide crystal in size, then digital has succeeded in matching film quality for resolution. Digital has a LONG way to go yet. A : 5000ppi scan of a blank negative yeilds the same number of pixels. duh! and a flat tyre is no good.. (until you pump it up !) Saying something as silly as this is as odd as your statement. : You still have a number fixation. My original post (can you remember that?) was something to do with the subject line above - remember that? Now who was it who brought numbers into the picture? Pretty sure it was you.. no wait, I make a statement, you reply something along the lines of 'rubbish' or 'no it doesn't' .. so *I* bring numbers in to back up my statements.. then I have a number fixation (!!) : Lots of bogus calculations with *NO* : actual measurements to prove whether any of your assumptions are actually : valid. hang on, you want justifications one minute (generally best represented mathematically) then you want actual measurements. Aren't they numbers? Can I use numbers again? You deflected my comment regarding digital fake pixels.. seeming to defend them as more real than film res and attacked the conclusion that tech pan resolution was around 100 lines per mm - go ahead and deny it, I'm expecting that. Then when I used a resolving power calc and concluded film (tech pan) as having 8Mp* and all but conceding your argument, you ignored that.. of course I was wrong then. I thought things through and did some straightforward calcs based on the actual numbers of sensors spread across the stated areas of various sensors. You conclude I am thick.. those numbers are there for anyone to view - where are your calculations refuting my observations? ": We can see you don't have a scientific background anyway" hahahaha.. you can *tell* that can you? I hope you don't stand by your to your deductions! ;) . Not sure that you : have proven anything else as yet though. 'not sure' ain't too scientific! (hint, see size of silver halide grain comment above) Shall we play numbers again? maybe this *real world* comparison would be better: http://www.users.qwest.net/~rnclark/scandetail.htm Warning : Clarke doesn't toe the party line and actually tests his theories.. and there are numbers involved k ><snip> > you know the difference between analogue (infinite progressive steps) and > digital (discrete steps) ? If a pixel could represent a silver halide > crystal in size, then digital has succeeded in matching film quality for > resolution. Digital has a LONG way to go yet. Not too sure who wrote this, but it's wrong. A film negative or slide cannot resolve detail down to individual grain size. What is commonly - and erroneously - call 'grain' is actually clumps of grains that are many times larger than the individual silver grains. Fuji specify grain size of Velvia at 8 microns, and the pitch of high-end sensors is less than 8 microns. Further, since a particular grain is either black or it isn't, it can be said to be binary - either on or off. But a pixel of about the same size in a sensor is not on or off, it can replicate many shades between black and white - in an 8-bit image, each pixel can represent 256 shades from black to white. Have a look at https://www.luminous-landscape.com/essays/clumps.shtml to get the picture (!) Colin D. SignaturePosted via a free Usenet account from http://www.teranews.com ----- Original Message ----- From: "ColinD" <nospam@127.0.0.1> Newsgroups: aus.photo Sent: Tuesday, June 27, 2006 12:15 Subject: Re: Interesting Thread on aus.bushwalking..... : ><snip> : [quoted text clipped - 5 lines] : Not too sure who wrote this, but it's wrong. A film negative or slide : cannot resolve detail down to individual grain size. obviously, yet neither can a pixel - however they contribute to the successful rendering of an image (image definition, resolution etc) by the sum of the parts making the whole. 2 lines of photosensors *minimum* are required to determine the position of a horizontal line, 2.5xsqrt 2 if you want it 45 degrees of axis If silver halide crystals were the same size as photosensors then you'd need the same number for silver based tech.. however they are much smaller and *capable* of resolving far smaller detail : What is commonly - : and erroneously - call 'grain' is actually clumps of grains that are : many times larger than the individual silver grains. that's why I specifically wrote 'silver halide crystals' not grain. there's a big difference which you and I both know ;) Fuji specify grain : size of Velvia at 8 microns, and the pitch of high-end sensors is less : than 8 microns. Further, since a particular grain is either black or it : isn't, not entirely true - take fine grain developers for example, sulphite partially disolves the silver during development potentially leaving much smaller clumps of semi transparent grain. then too, talking Velvia (colour I presume)- what's left is no silver but rather the dyes. : it can be said to be binary - either on or off. it can be treated that way, yet diffraction and diffusion in practical terms render that otherwise, unless one's talking about using high accutance devs with 'grainy' films.. and assuming we're talking black and white But a pixel of : about the same size in a sensor is not on or off, it can replicate many : shades between black and white - in an 8-bit image, each pixel can [quoted text clipped - 5 lines] : : to get the picture (!) Don't get this wrong Colin, I l know what you mean about clumps and halides - I worked in this area for quite a while. I even designed a few dev processes :) Digital is a fantastic tool, no argument from me - it just isn't capable of what film can do. this is worth a peek ("Future Proofing with Film") http://www.henninger.com/library/hdtvfilm/ and a reply sent to me: I forwarded Karl's article RE "future proofing" to a quite knowledgeable friend in the biz and this was his reply (Again, the conclusion is the important part): There are a number of fundamentally incorrect assumptions made in this guys' article. He talks about resolution and I do give him credit for explaining how it is measured, but he misses a major point with regard to resolution of HD video cameras. A typical 2/3" 3-chip HDTV camera has sensors with 1920 pixels across a 9.6mm wide area. In order to make any measurement of resolution of such a chip, one must divide the pixel count by 2. This, because a single pixel (picture element) is a photo site that can have only one single luminance level at any given time. In order to discern the finest bit of detail that any optical imaging system is capable of, you minimally need two pixels or you will have nothing against which to contrast a single pixel. So, with every other pixel being black and rest being white, the 1920 pixel wide chip is optimally capable of 100 lp/mm or 960 tvl/ph in video terms. This is on a good day -In other words, it can't happen. *(he also made the same mistake I did in assuming 2x - which ONLY works on the same axis as the sensors - not at 45 degrees where 2.5xsqrt2 should be used)* So now that we know what is possible in an optimal situation, let's examine how people are really doing HDTV. They use much cheaper zoom lenses with maximum resolution of around 56 lp/mm at perhaps just one focal length and with the iris closed a bit. They typically don't spend the money for the right charts or learn how to properly back-focus their lenses and so performance at infinity focus is compromised reducing resolution at the very focal lengths where it is most needed (most are still using back focus charts designed for LowDTV or none at all). They don 't use matte boxes or if they do, don't really understand how to properly use them and so compromise dynamic range and apparent sharpness further. They stick el cheapo Tiffen window glass filters in front of their lens, screwing up the lens file (assuming they know what one is and have gone to the trouble of creating one) and undoing the effects of having a multicoated lens (for which they paid a premium). And lastly, they are judging focus with a resolution limited viewfinder as opposed to using a high resolution monitor, so they can't really see when stuff is in focus anyway. So anyone that thinks that the average user of even a $100,000 Sony Cine Alta HD-F900 is getting the most out of their camera is probably smoking funny cigarettes. While the Japanese (and a couple of domestic chip manufacturers) continue to cram more and more pixels onto silicon, film continues to get better and better with each new generation of emulsion that Kodak and Fuji bring out. And all but the most serious cinematographers continue to ignore the role that the lens plays in all of this. It's a numbers game and the specs are the manufacturers bragging rights, nothing more. <clipped cine stuff> Bob Zeichner, MFSO > you know the difference between analogue (infinite progressive steps) and > digital (discrete steps) ? Yes, I know that anyone who believes analog has "infinite" *anything* in this universe, doesn't have the *slightest* clue what they are talking about. > If a pixel could represent a silver halide > crystal in size, then digital has succeeded in matching film quality for > resolution. Digital has a LONG way to go yet. Not at all. The grain size of many films *does* in fact limit the 37Mpixels you are fond of quoting. Even you know that, which is why you always mention Tech Pan as a *possible* chance. But have yet to prove the Rollei is actually capable of that even with Tech Pan. That is what was in dispute after all. > : Lots of bogus calculations with *NO* > : actual measurements to prove whether any of your assumptions are actually [quoted text clipped - 3 lines] > mathematically) then you want actual measurements. Aren't they numbers? > Can I use numbers again? Of course, but what we need are real *proven* ones, not just pulled from your arse. > You deflected my comment regarding digital fake pixels.. seeming to defend > them as more real than film res and attacked the conclusion that tech pan > resolution was around 100 lines per mm - go ahead and deny it, I'm > expecting that. Information can be defined and measured. Whether it starts as analog or digital is irrelavant. > Then when I used a resolving power calc and concluded film (tech pan) as > having 8Mp* and all but conceding your argument, you ignored that.. of > course I was wrong then. I never commented at all, since you were just shooting yourself in the foot. > I thought things through and did some straightforward calcs based on the > actual numbers of sensors spread across the stated areas of various > sensors. You conclude I am thick.. No, (well maybe, given your claim above of *infinite* "steps" for analog data.) > those numbers are there for anyone to view - where are your calculations > refuting my observations? Like you, I haven't done the measurements to back them up. I don't even own the Rollei under discussion. > maybe this *real world* comparison would be better: > http://www.users.qwest.net/~rnclark/scandetail.htm > > Warning : Clarke doesn't toe the party line and actually tests his > theories.. and there are numbers involved Yes, he contradicts himself quite often. What he does show beyond doubt is that there are many variables to image quality, and simply talking pixels like it is the whole story, is delusional. Garbage in, garbage out comes to mind. MrT. "Mr.T" me: | > you know the difference between analogue (infinite progressive steps) and | > digital (discrete steps) ? | Yes, I know that anyone who believes analog has "infinite" *anything* in | this universe, doesn't have the *slightest* clue what they are talking | about. take a frog Mr T, let him jump half way toward his rock, then let him jump half way again. How many times must he jump to get to the rock? to concede a point though, film/digital has limits of course as there are a finite number of grains/pixels to the media. Interestingly, the challenger space shuttle disaster was filmed digitally and I had the pleasure of examining it shortly after the event. Problem - XxY pixels, no matter how many times they were captured revealed no useful information. Film however with the grain location and size constantly changing would have allowed a lot more detail to be onserved by overlaying and adding/removing. A shame. | > If a pixel could represent a silver halide | > crystal in size, then digital has succeeded in matching film quality for | > resolution. Digital has a LONG way to go yet. | | Not at all. The grain size of many films *does* in fact limit the 37Mpixels | you are fond of quoting. It sure does , as does the TYPE of developer used (sulphite etc), the method of agitation, the fix procedure, etc. However 'Film' (generic, all encompasing term) and "many films" are not the same thing. I love this bit where people talk digital Vs 'film', always limiting 'film' to 35mm and some awful fast, thick E6 emulsion - you haven't specified I film I know, but it seems so prevalent around the 'net. Yes there are low res films, low res lenses too - and some people manage to lower their potential resolution further by neglecting to use a lens hood! Focal plane shutters used at high speeds with a narrow slit, effectively increasing diffraction is another way.. yet the old 35mm SLR is still the 'point of refence' for many. How many other ways are there to lower resolution? This has gotten waaay out of hand. my original post went like this: "thinks for a moment: a teeny $100 35mm rollei loaded with 25/50/100/200/400 ISO still produces the same theoretical 'filesize' image as a top of the line 35mm SLR (what does a 5000 ppi scan yield) -what sort of DSLR weight must one consider to give the same file? must weigh a bit.." you: | But why on earth would you compare a DSLR to a 35mm rangefinder? me: the image filsize as I stated. 5000pixels x 75000 pixels = 37 1/2 Mp at a 5000 ppi scan. it goes downhill from there. Generally when you suggest a blank bit of film has no information on it.. | Even you know that, which is why you always mention Tech Pan as a *possible* | chance. But have yet to prove the Rollei is actually capable of that even | with Tech Pan. That is what was in dispute after all. How would you like it proved - other than shooting test charts and whacking the results under a microscope.. I have the graticule slide sitting right here, I have the microscope - but I haven't resorted to a test chart for some time now. It's a bad habit, one I discourage! ;) | Of course, but what we need are real *proven* ones, not just pulled from | your arse. what, you want an article in Polular Photography or something? k {snip irrelevant stuff} > It sure does , as does the TYPE of developer used (sulphite etc), the method > of agitation, the fix procedure, etc. However 'Film' (generic, all > encompasing term) and "many films" are not the same thing. Exactly what I said when you kept on about Tech Pan. > I love this bit where people talk digital Vs 'film', always limiting 'film' > to 35mm and some awful fast, thick E6 emulsion - you haven't specified I > film I know, but it seems so prevalent around the 'net. Yes, you were the one limiting to Tech Pan, and we were discussing the Rollei RF in particular. > This has gotten waaay out of hand. my original post went like this: > [quoted text clipped - 4 lines] > you: > | But why on earth would you compare a DSLR to a 35mm rangefinder? Yep, the Rollei RF is not an SLR. Current DSLR's are the same size and weight as 35mm SLR's, on average. > me: > the image filsize as I stated. 5000pixels x 75000 pixels = 37 1/2 Mp at a > 5000 ppi scan. Yep. figures which bear no relevance to actual data for the range of films quoted above. You later limit the discussion to Tech Pan, but still fail to actually prove anything. > it goes downhill from there. Generally when you suggest a blank bit of film > has no information on it.. No, when you suggested that 37 Mpixels are required to capture data from a blank bit of film, or that a 35mm Rollei RF has 3 times the performance of a Canon 1DSmkII. > | Even you know that, which is why you always mention Tech Pan as a > *possible* [quoted text clipped - 3 lines] > How would you like it proved - other than shooting test charts and whacking > the results under a microscope.. Would be a good start. >I have the graticule slide sitting right > here, I have the microscope - but I haven't resorted to a test chart for > some time now. It's a bad habit, one I discourage! ;) Why? Better just to make unfounded speculation in your opinion? MrT. | > I love this bit where people talk digital Vs 'film', always limiting | 'film' [quoted text clipped - 3 lines] | Yes, you were the one limiting to Tech Pan, and we were discussing the | Rollei RF in particular. True, and where's this 'We' - I was the one talking Rollei + tech pan. There's no point discussing the Rollei on it's own - it has no resolution beyond the lens unless you pop a film in it.. in the example I used, tech pan. were "we" to be discussing filmS (emphisis on the plural), then we'd have to be talking generally but that would be pointless so 'we' must talk of a base limit. Much as one would when for example, this discussion to take place: "all your digital photos on the web look like crap" 'what is your monitor resolution and bit depth set to?' "what does that matter, all your images look like crap!" One sets a base standard. Mine was Tech pan. I concluded that a 5000lpi scan would yeild a 37M file I then extrapolated back from a 16Mp cameras stated sensor count to conclude the same thing. We could have talked about Type 2745 Recording Film, but that would be as idiotic as saying a photographic print is better than a web image at 256 colours. (as an aside 2475 was the only film I found to maintain it's contrast curve across a range of time/temps .. unique!) | > you: | > | But why on earth would you compare a DSLR to a 35mm rangefinder? | | Yep, the Rollei RF is not an SLR. Current DSLR's are the same size and | weight as 35mm SLR's, on average. Ah.. now that makes sense :) Almost. | Yep. figures which bear no relevance to actual data for the range of films | quoted above. You later limit the discussion to Tech Pan, but still fail to | actually prove anything. sigh.. these are the days when web pages = truth, nothing else matters.. | > it goes downhill from there. Generally when you suggest a blank bit of | film has no information on it.. | No, when you suggested that 37 Mpixels are required to capture data from a | blank bit of film, now where did you pull that from? I think if you look carefully across your posts, you'll find you are the one who frequently referred to blank film. or that a 35mm Rollei RF has 3 times the performance of a | Canon 1DSmkII. can you count the pixels? no, however you CAN check the manufacturers data. Can you now tell me how many lines of pixels are required at the resolving point to determine whether a line is 'here' or 'here' then swing that 45 degrees and work out how many are needed | > How would you like it proved - other than shooting test charts and | whacking the results under a microscope.. | | Would be a good start. you wouldn't believe the results. But humouring you for a moment, what would be the comparison? Hark back to this for a moment.. "What a load of mangled crap. All digital photo's have a "resolution" limit before ANY sharpening algorithms are applied to the captured data. All film cameras and film also have a resolution limit before any scanning or printing is done. Care to explain how we 'measure' the pre interpolated digital image ? If we DO want to make this comparison, lets get a few things straight. I know how to measure the resolution (resolving power, pre-digital-talk-interpretation ). Do you want to take a stab at suggesting a way to measure the resolving power of a digital sensor, pre-demosaicing, pre-sharpening ? Can you TELL me the resolution limit of a digital camera? I'd be keen to see your method The best most people can mange is to work the numbers, after all, once the output from the photosites is digitised it is a numbers game. Actually even before it's digitised.. | >I have the graticule slide sitting right | > here, I have the microscope - but I haven't resorted to a test chart for | > some time now. It's a bad habit, one I discourage! ;) | | Why? Better just to make unfounded speculation in your opinion? lets get started.. speculation is based on 'nothing' fact based on - 'tangible details' going over this again: the limit of an optical system res: (lens res x film res) / (lens res + film res) tech pan = 100x80 / 100+80 = 44.4 lpmm 24mm x 36mm x 44.4 =1,703,255 line-blocks Is this 'unfounded speculation'? Or can we call this a theoretical statement based on accepted facts? Canon EOS-1Ds Mark II (16Mp camera) resolution of 4992 x 3328 pixels covering a 24x36mm area. 4992/36 = 138.7 pixels (lines) per mm 1836 x 1224 = 2,247,264 are we still in speculation mode? How can we measure a signal electronically if it approaches the limit of the sampling technique? the Niquist limit is worth considering. He found you need to sample at a rate that exceed twice the bandwidth, to get the real information. So our ACTUAL photosites above can be concluded as having a sensitivity 50% of the actual number of sites to attain accuracy. Accuracy is a little like resolution, no? Accuracy is a LOT like resolution.. OK, so we've been able to show (not PROVE, we're waiting for an idea of how to do that ) the digital sensor above can use all it's photosites, but effectively a linear half of them are used to determine where the detail is. Another way - How many lines of pixels are needed to determine whether a line of an adjacent pair is at point A or whether it is at point B - answer, a minimum of 2. still pretty stable at this point, fact wise.. How many at 45 degrees? almost 3.5 draw it on a graph.. you'll see. So our digital camera example has 2247274 photosites. Half are used to determine what the OTHER half see are right (assuming the lines are perpendicular to the sensor). So that leaves 1123632 photosites remaining. Canon EOS-1Ds Mark II so far =1,123,632 tech pan so far = 1,703,255 If we work toward maximising the chance that the digital camera can resolve as WELL as possible, we reduce it to tech pan so far = 1,703,255 Canon EOS-1Ds Mark II = 642,075 You know something about audio, extrapolate along the same theories and you'll see. now here's the WILDEST speculation. IF we can assume a 16Mp camera capable of capturing 16Mp, then we can deduce (1.7 / 0.64) x 16Mp = ? Mp from the tech pan. BTW, how did you get on with the frog question? k <snip> > So our digital camera example has 2247274 photosites. Half are used to > determine what the OTHER half see are right (assuming the lines are > perpendicular to the sensor). So that leaves 1123632 photosites remaining. > > Canon EOS-1Ds Mark II so far =1,123,632 > tech pan so far = 1,703,255 > If we work toward maximising the chance that the digital camera can resolve > as WELL as possible, we reduce it to > > tech pan so far = 1,703,255 > Canon EOS-1Ds Mark II = 642,075 I beg to differ here. The 'line blocks' you mentioned in the snipped bit are the equivalent of pixels in a sensor. If you are going to halve the number of pixels in a sensor (actually this is the same as 'line pairs/mm', instead of 'lines per mm'), then you need to do this for the film as well, so TP = 536,628 vs 1Ds = 642,075. However, returning to reality instead of rarified sums, there's no question that 6 MP or better can outrun all but the most careful processing of the slowest 35mm film available; and the comparison between faster 35mm films and digital just leaves film in the dust. <more snip> Colin D. SignaturePosted via a free Usenet account from http://www.teranews.com | <snip> | I beg to differ here. The 'line blocks' you mentioned in the snipped | bit are the equivalent of pixels in a sensor. If you are going to halve | the number of pixels in a sensor (actually this is the same as 'line | pairs/mm', instead of 'lines per mm'), then you need to do this for the | film as well, so TP = 536,628 vs 1Ds = 642,075. No, Colin you recall you said how small halide particles were? very small It's OK.. The resolution is already a stated, measurably determined fact for film. Those grains are SO small that rather than needing to add the number of *grains* up and devide them by 2 and multiply them by the film size (!) we already have a stated resolution. Of course if someone wants to add them up and divide by two.. ;) | However, returning to reality instead of rarified sums, there's no | question that 6 MP or better can outrun all but the most careful | processing of the slowest 35mm film available; and the comparison | between faster 35mm films and digital just leaves film in the dust. That's something the guys at NASA are starting to review post-Challenger. We might just see the old film cameras beeing wheeled out again for critically important stuff for general image making where longevity and accuracy aren't important it clearly stands head and shoulders over film. Sales figures prove it k "k" <fellafel@PING.com> | The resolution is already a stated, measurably determined fact for film. | Those grains are SO small that rather than needing to add the number of | *grains* up and devide them by 2 and multiply them by the film size (!) we | already have a stated resolution. | | Of course if someone wants to add them up and divide by two.. ;) I neglected to mention that to complicate things further the film halide grains are laid out in 3D on the substrate - a double edged sword sure, but it complicates calculations further. Digital has the enviable feature of being arranged in a single layer. k > "ColinD" <nospam@127.0.0.1> wrote in message > | However, returning to reality instead of rarified sums, there's no > | question that 6 MP or better can outrun all but the most careful [quoted text clipped - 10 lines] > > k I understand that 'stacking' frames to reduce grain and noise is done to winkle out the most detail available, it doesn't mean that stacked film images will be any better than stacked sensor images. Astronomers have been doing this for years, and stacked digital exposures are way ahead of film. I guess the NASA folks are/were trying to see as much as they could of the Discovery breakup, using whatever records there were of the disaster. It wasn't so much as case of film vs digital as which medium had the advantage of bigger or sharper images to start with, or even any image at all that was available. Colin D. SignaturePosted via a free Usenet account from http://www.teranews.com > > "ColinD" <nospam@127.0.0.1> wrote in message > [quoted text clipped - 24 lines] > had the advantage of bigger or sharper images to start with, or even any > image at all that was available. Dammitall, not Discovery, I meant Columbia {:-( Colin D. SignaturePosted via a free Usenet account from http://www.teranews.com | > I understand that 'stacking' frames to reduce grain and noise is done to | > winkle out the most detail available, it doesn't mean that stacked film | > images will be any better than stacked sensor images. Astronomers have | > been doing this for years, and stacked digital exposures are way ahead | > of film. you're right in that for eliminating noise digital wins, but for gaining cumulative small fragments of detail the stacking of film can work very well . The other upside to digital stacking lies in it's greatest strength - rapid access and ease of manipulation. I've used SGBRN for noise reduction for both film and digital and digital wins, I've also used darkfield stackers (astronomix) and digital CLEARLY wins on ease of noise reduction! :) But the reduction of noise is just one use for stacking. | > I guess the NASA folks are/were trying to see as much as they could of | > the Discovery breakup, using whatever records there were of the [quoted text clipped - 3 lines] | > | Dammitall, not Discovery, I meant Columbia {:-( Here's the image, take a look, it was only captured digitally so this is *all* they had to work with: http://www.rit.edu/~andpph/a-misc/107_bw_2_07.jpg Of course at a 32x24 pixel crop there really isn't a lot to see :( Andy Davidhazy was sent the images as soon as they became available from NASA for his analysis, he bounced the images on for others to interpret - no one could help deduce anything from the little squares :( karl > Here's the image, take a look, it was only captured digitally so this is > *all* they had to work with: > > http://www.rit.edu/~andpph/a-misc/107_bw_2_07.jpg > > Of course at a 32x24 pixel crop there really isn't a lot to see :( And there's the rub. The question is how many mm^2 of film would also be required. > Andy Davidhazy was sent the images as soon as they became available from > NASA for his analysis, he bounced the images on for others to interpret - no > one could help deduce anything from the little squares :( Is this supposed to be a red herring or a straw man? I'm not sure which. A better/more suitable camera & lens, either film or digital, could have provided a better image, if it was lucky enough to be in the right place at the right time. MrT. > now here's the WILDEST speculation. IF we can assume a 16Mp camera capable > of capturing 16Mp, then we can deduce (1.7 / 0.64) x 16Mp = ? Mp from the > tech pan. Easy enough to prove whether tech pan in the Rollei can actually produce better prints than from a Canon 1DSmkII. (use a similar fixed lens for the Canon of course) Now try it with some color print film and compare. Let us know what your results are. Do you actually get 2.6 times more resolution on the print? Do pigs fly? Or do you just like quoting meaningless numbers. MrT. | > now here's the WILDEST speculation. IF we can assume a 16Mp camera | capable [quoted text clipped - 10 lines] | Do pigs fly? | Or do you just like quoting meaningless numbers. do you actually use a camera? but back to the questions.. HOW can one identify actual resolution for digital bearing in mind your comment "What a load of mangled crap. All digital photo's have a "resolution" limit before ANY sharpening algorithms are applied to the captured data. All film cameras and film also have a resolution limit before any scanning or printing is done. One of the main problems with this digital/interface/film world is that certain terms like 'resolution' have been hijacked by digital. k > HOW can one identify actual resolution for digital bearing in mind your > comment > "What a load of mangled crap. All digital photo's have a "resolution" limit > before ANY sharpening algorithms are applied to the captured data. > All film cameras and film also have a resolution limit before any scanning > or printing is done. The same way you do with any imaging system. If *you* want to apply algorithms to the digital files, go right ahead. You can use identical ones for the digital image, and your 37Mpixel drum scannned film image, if you think that helps. > One of the main problems with this digital/interface/film world is that > certain terms like 'resolution' have been hijacked by digital. No, it's that some people don't understand what they mean apparently. MrT. | > HOW can one identify actual resolution for digital bearing in mind your | > comment [quoted text clipped - 3 lines] | > All film cameras and film also have a resolution limit before any scanning | > or printing is done. | The same way you do with any imaging system. If *you* want to apply | algorithms to the digital files, go right ahead. You can use identical ones | for the digital image, and your 37Mpixel drum scannned film image, if you | think that helps. SO again I ask, how does one determine the resolution of a digital file - you said ""What a load of mangled crap" to my "digital 'resolution' is sharpening algorithms applied to fuzzed images projected onto bayer arrays, the RGGB mono images then being interpolated to make a good guess of the scene.. explain then how this is "mangled crap" or better yet, explain how the resolution of a digital camera can be determined before the algorithms are worked - you suggest it exists.. I would like to know more. | > One of the main problems with this digital/interface/film world is that | > certain terms like 'resolution' have been hijacked by digital. | | No, it's that some people don't understand what they mean apparently. what does 'resolution' mean? Pre digital it meant resolving power, POST digital it has come to mean pixel count. can you correlate the two? > SO again I ask, how does one determine the resolution of a digital file - One determines the resolution of an optical *system*, not a file. > to my "digital 'resolution' is sharpening algorithms applied to fuzzed > images projected onto bayer arrays, the RGGB mono images then being > interpolated to make a good guess of the scene.. Well there's your problem, misuse of terminology. > or better yet, explain how the resolution of a digital camera can be > determined before the algorithms are worked - you suggest it exists.. I > would like to know more. With RAW files, no algorithms need be applied to the captured data (other than to re-organise the pixel order etc). In any case all the algorithms in the world wont create resolution where none exists. You have already shown that. In fact your insistence on the use of Tech Pan for film could probably be equated to any contrast fiddling etc. with color digital sensor array data. It all depends on what part of the system one is trying to measure, and how you wish to *consistantly* define it. > what does 'resolution' mean? Pre digital it meant resolving power, POST > digital it has come to mean pixel count. > > can you correlate the two? The second is *not* resolution. MrT. "Mr.T" | One determines the resolution of an optical *system*, not a file. you know that I meant system not file.. | > to my "digital 'resolution' is sharpening algorithms applied to fuzzed | > images projected onto bayer arrays, the RGGB mono images then being | > interpolated to make a good guess of the scene.. | | Well there's your problem, misuse of terminology. You obviously are not interested in facts, learning, sharing or conversing. | > or better yet, explain how the resolution of a digital camera can be | > determined before the algorithms are worked - you suggest it exists.. I [quoted text clipped - 4 lines] | In any case all the algorithms in the world wont create resolution where | none exists. You have already shown that. I take it you feel you have some special secret which you choose to guard closely. I can accept that, it's your perogative. | In fact your insistence on the use of Tech Pan for film could probably be | equated to any contrast fiddling etc. with color digital sensor array data. | It all depends on what part of the system one is trying to measure, and how | you wish to *consistantly* define it. it's one of the thinner emulsioned films, it has the finest grain and it has a nice sensitivity - it's a good film to use as a base standard. I'd also add that it has an axceptionally wide contrast range and responds to various developers in a highly predictable fashion. It's a good film to use for many tests as well as just being a fantasic film for shooting. | > what does 'resolution' mean? Pre digital it meant resolving power, POST | > digital it has come to mean pixel count. | > | > can you correlate the two? | | The second is *not* resolution. Tell that to the manufacturers, magazines, even *lecturers* these days. When I was told I HAD to teach that filesize was resolution and we were not permitted to go any further in discussing it, I decided my days as a lecturer were over. No I'm not telling you who was delivering that course.. I'm in enough trouble there already for bucking the system You don't want to discuss how one can determin the actual resolution of a digital camera system - so we've nothing I can see here that holds me to this nonsensical discussion. you win Whatever your point may have been.. |
Reply to this Message |
 Sign In
 Join
 My Latest Posts My Monitored Threads My Blog My Photo Gallery My Profile My Homepage Start New Thread Enable EMail Alerts Rate this Thread
|
©2008 Advenet LLC Privacy Policy - Terms of Use
This website includes both content owned or controlled by Advenet as well as content owned or controlled by third parties.