 |
 | 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 / Digital Photography / DSLR Cameras / April 2006Nikon D70s image size
Greetings... I was playing with my friend's D70s yesterday, and we noticed something when we shot in RAW format - the image size was 3038x2012, yet JPG size is 3008x2000 pixels. Nikon's specifications state 3008x2000 in every document I've read, including the owners manual, and all online sources say the same. My Canon produces the same size RAW and JPG files. Does anyone know why the Nikon RAW image is reporting slightly larger? Is the camera merely cropping to 3008x2000 when converting to JPG? >Greetings... > [quoted text clipped - 8 lines] >Does anyone know why the Nikon RAW image is reporting slightly larger? >Is the camera merely cropping to 3008x2000 when converting to JPG? Where is the raw image "reporting" this? Nikon Capture, PSE3 and RSE both "report/convert" images from my D70 to 3008 x 2000. However, RSE "reports/converts" to 3036 x 2010. -- Ed Ruf Lifetime AMA# 344007 (Usenet2@EdwardG.Ruf.com) http://EdwardGRuf.com >>I was playing with my friend's D70s yesterday, and we noticed something >>when we shot in RAW format - the image size was 3038x2012, yet JPG size [quoted text clipped - 10 lines] >"report/convert" images from my D70 to 3008 x 2000. However, RSE >"reports/converts" to 3036 x 2010. Picasa reports it that way...that's where we noticed it. Picasa also converts to JPG at the same image size. > I was playing with my friend's D70s yesterday, and we noticed something > when we shot in RAW format - the image size was 3038x2012, yet JPG size > is 3008x2000 pixels. Nikon's specifications state 3008x2000 in every > document I've read, including the owners manual, and all online sources > say the same. JPEGs are best produced with dimensions that are even multiples of the compression block size, which is typically 16x16. If the dimensions are correct in this way, the image can be rotated with no loss. Also, it is normal for the edges of the data to be removed from the image. Due to the way Bayer interpolation works, less data is available to form those edges; those areas aren't meant to be part of the final image, but they can be used anyway if you really want.  SignatureJeremy | jeremy@exit109.com > Greetings... > [quoted text clipped - 8 lines] > Does anyone know why the Nikon RAW image is reporting slightly larger? > Is the camera merely cropping to 3008x2000 when converting to JPG? That is very interesting and may answer a question of mine. Nikon calls the D70 a 6.1 MP camera, but 3008 x 2000 is only 6.018 MP, so I could not understand why they advertised the camera 6.1 MP. However, 3038x2012 is 6.092 MP, which rounds to 6.1 MP. It does indeed look like the JPEG processing crops the image size. Merritt Today Merritt Mullen commented courteously on the subject at hand > That is very interesting and may answer a question of mine. > Nikon calls the D70 a 6.1 MP camera, but 3008 x 2000 is > only 6.018 MP, so I could not understand why they > advertised the camera 6.1 MP. > > However, 3038x2012 is 6.092 MP, which rounds to 6.1 MP. My calculator says that 3038x2012=6,112,456=5.8292 MP. You need to divide by divide by 1,048,,576 (1024x1024) to convert normal number to mega anything. What easy thing am I missing here, other than Nikon obviously wants to advertise 6 MP and not a tweak under, unlike retailers who price things $9.99 rather than $10.00 for the opposite reason. SignatureATM, aka Jerry "Whether You Think You CAN Or CAN'T, You're Right." – Henry Ford > You need to divide by divide by 1,048,,576 (1024x1024) to convert > normal number to mega anything. No you don't. That only applies to megabytes. -Mike Today Mike Warren commented courteously on the subject at hand >> You need to divide by divide by 1,048,,576 (1024x1024) to >> convert normal number to mega anything. > > No you don't. That only applies to megabytes. Mathematicians created terms like mega, giga, terra, and they are all based on powers of two. So why are mega /pixels/ exempt? SignatureATM, aka Jerry "Whether You Think You CAN Or CAN'T, You're Right." – Henry Ford > Today Mike Warren commented courteously on the subject at > hand [quoted text clipped - 6 lines] > Mathematicians created terms like mega, giga, terra, and they > are all based on powers of two. So why are mega /pixels/ exempt? They are based on factors of 10. Kilo=1000 x Mega=1000000 x Giga=1000000000 x Only when talking about binary is the power of 2 system used. And that's just for convenience. I think different words should have been used to avoid the confusion. -Mike > Only when talking about binary is the power of 2 system used. > And that's just for convenience. I think different words should > have been used to avoid the confusion. Yes, the problem started in electical engineering by calling 1024 byte a kilo-byte. Close enough for government and computer work but the rest of science considers kilo to mean a factor of 1000. SignatureThomas T. Veldhouse Key Fingerprint: 2DB9 813F F510 82C2 E1AE 34D0 D69D 1EDC D5EC AED1 > > Only when talking about binary is the power of 2 system used. > > And that's just for convenience. I think different words should [quoted text clipped - 3 lines] > kilo-byte. Close enough for government and computer work but the rest of > science considers kilo to mean a factor of 1000. Which is why kibi, mebi, and gibi were invented (http://whatis.techtarget.com/definition/0,,sid9_gci825099,00.html) Today Thomas T. Veldhouse commented courteously on the subject at hand >> Only when talking about binary is the power of 2 system >> used. And that's just for convenience. I think different [quoted text clipped - 4 lines] > and computer work but the rest of science considers kilo to > mean a factor of 1000. The "problem" started in the very earlies digital age, which includes electro-mechanical switch where it made perfect sense to denote numbers in powers of 2. Nothing in my mind has changed to this day and I actually don't give a damn if a 6,122,6144 pixel image (or whatever it was) is a 5.9293 or a 6.0892 mega pixel image, and humans think in decimal not binary. SignatureATM, aka Jerry "Whether You Think You CAN Or CAN'T, You're Right." – Henry Ford >> Yes, the problem started in electical engineering by >> calling 1024 byte a kilo-byte. Close enough for government [quoted text clipped - 7 lines] > pixel image (or whatever it was) is a 5.9293 or a 6.0892 mega > pixel image, and humans think in decimal not binary. Isn't that what I said? Who do you think created the first processors, tube based or otherwise? Who do you think created the "byte" and eventually coined the kilobyte? Electrical Engineers (and eventually the more specialized computer engineers). As far as referencing electromechanical counting machines of the pre 1950s era, I have a hunch that the term kilobyte never came up ;) SignatureThomas T. Veldhouse Key Fingerprint: 2DB9 813F F510 82C2 E1AE 34D0 D69D 1EDC D5EC AED1 > Today Thomas T. Veldhouse commented courteously on the > subject at hand [quoted text clipped - 14 lines] > pixel image (or whatever it was) is a 5.9293 or a 6.0892 mega > pixel image, and humans think in decimal not binary. It is not a matter of how you think, but how you count. A million pixels is a million pixels. Changing the base does not change the physical reality. Merritt Today Merritt Mullen commented courteously on the subject at hand >> Today Thomas T. Veldhouse commented courteously on the >> subject at hand [quoted text clipped - 23 lines] > million pixels is a million pixels. Changing the base does > not change the physical reality. the base /is/ how you think, e.g., programmers who "think" in hex, octal, or binary. what's interesting about this thread, the part I was stupid enough to join, is that it no longer has any relation to the OPs question but is the normal elitists on a mission from Allah to squash the infidel. If you understood english, you would realize that when I said "humans think" I actually agreed with you. It hardly matters what base pixels are expressed in anymore than what weight measure or base cottage cheese is measured in. Maybe there're people somewhere who count the weight of their cottage cheese in octal because they had both their thumbs cut off and can't count to 10 anymore. SignatureATM, aka Jerry "Whether You Think You CAN Or CAN'T, You're Right." – Henry Ford > what's interesting about this thread, the part I was stupid > enough to join, is that it no longer has any relation to the > OPs question but is the normal elitists on a mission from > Allah to squash the infidel. Actually, all the elitists on a mission from Allah did was point out that "mega" does not actually imply base-2 except in a particular context. Why do you take it so personally? SignatureJeremy | jeremy@exit109.com Today Jeremy Nixon commented courteously on the subject at hand >> what's interesting about this thread, the part I was >> stupid enough to join, is that it no longer has any [quoted text clipped - 4 lines] > point out that "mega" does not actually imply base-2 except > in a particular context. Why do you take it so personally? Ah, yes, pearls of wisdom from Jeremy, who never says anything positive about anything to anybody. He just sits under his favorite troll bridge until somebody he doesn't like or just doesn't agree with says something, then he rises up in righteous indignation to put the transgressor in his place. Don't you ever get tired of busting my balls for no good reason other than to gratify your own overwhelming ego, you know, making pronouncements indended to make you look smart and me look stupid? Tell you what, do this: put me in your killfile and you won't be subjected to my inane babbling or just be polite to the rest of the good people here and let what ever I say just pass by not even reading it so you aren't tempted to get obnoxious again. Or, just STFU. Have it any way you like, just go back to sleep under your troll bridge again, you're of no use to anyone here and a thorn in my side I hardly need. SignatureATM, aka Jerry "Whether You Think You CAN Or CAN'T, You're Right." – Henry Ford > Ah, yes, pearls of wisdom from Jeremy, When someone points out an inaccuracy in something I've said, my typical reaction is something like "oh, okay, thanks." Yours is vicious attacks and claims of persecution. Having seen your method in action, I think I'll stick to mine. But hey, if yours is working for you, go for it, I guess. SignatureJeremy | jeremy@exit109.com > Yes, the problem started in electical engineering by calling 1024 byte a > kilo-byte. Close enough for government and computer work but the rest of > science considers kilo to mean a factor of 1000. But 1024 in decimal is 10000000000, a nice round number, if all you do is binary, then the words you use to describe numbers are based on binary NOT decimal. -SL > But 1024 in decimal is 10000000000, 1024 (base 10 aka decimal) is 10000000000 (base 2 aka binary). Signature-- r.p.e.35mm user resource: http://www.aliasimages.com/rpe35mmur.htm -- r.p.d.slr-systems: http://www.aliasimages.com/rpdslrsysur.htm -- [SI] gallery & rulz: http://www.pbase.com/shootin -- e-meil: Remove FreeLunch. Today Sam Lowry commented courteously on the subject at hand >> Yes, the problem started in electical engineering by >> calling 1024 byte a kilo-byte. Close enough for [quoted text clipped - 4 lines] > all you do is binary, then the words you use to describe > numbers are based on binary NOT decimal. Early computer programmers had to make do with only two fingers, hence they counted in binary. Later, thanks to Darwin, they gained enough to want to count to 16 but ran out of fingers and thumbs and had to take their shoes and socks off. That is, except for people at some companies who used could get by even without their thumbs. These days, there's a controversy about how the eye sees light and the brain does A-to-D to convert to pixels, so it could well be that pixels will eventually be counted in a retro base - 12. SignatureATM, aka Jerry "Whether You Think You CAN Or CAN'T, You're Right." – Henry Ford > Today Mike Warren commented courteously on the subject at > hand [quoted text clipped - 3 lines] > > > > No you don't. That only applies to megabytes. > Mathematicians created terms like mega, giga, terra, and they > are all based on powers of two. So why are mega /pixels/ exempt? Nope, the powers of two definitions come from the computer side of the world. Scientists in general think "kilo" means 10^3, not 2^10. Those prefixes come to us from the metric system and the scientific universe. SignatureDavid Dyer-Bennet, <mailto:dd-b@dd-b.net>, <http://www.dd-b.net/dd-b/> RKBA: <http://www.dd-b.net/carry/> Pics: <http://dd-b.lighthunters.net/> <http://www.dd-b.net/dd-b/SnapshotAlbum/> Dragaera/Steven Brust: <http://dragaera.info/> > Today Mike Warren commented courteously on the subject at > hand [quoted text clipped - 6 lines] > Mathematicians created terms like mega, giga, terra, and they > are all based on powers of two. So why are mega /pixels/ exempt? Actually they are part of the universal (metric) system, which was established long before computers became commonplace, and it is based on powers of 10, not 2. Mathematicians, scientists, and engineers do not normally work in powers of two. The computer scientists modified the system for counting binary digits, but that has nothing to do with the number of pixels in an array, which is a physical measurement, not a binary calculation. Merritt >> Today Mike Warren commented courteously on the subject at >> hand [quoted text clipped - 17 lines] > but that has nothing to do with the number of pixels in an array, which is > a physical measurement, not a binary calculation. And that modification has also been made part of the metric system. A kilobit is a thousand bits. A kibibit is 1024 bits. There is similar nomenclature for other multiples. For more information you might want to look at <http://en.wikipedia.org/wiki/Binary_prefix> Signature--John to email, dial "usenet" and validate (was jclarke at eye bee em dot net) Today J. Clarke commented courteously on the subject at hand >>> Today Mike Warren commented courteously on the subject at >>> hand [quoted text clipped - 25 lines] > For more information you might want to look at > <http://en.wikipedia.org/wiki/Binary_prefix> Silly me here, I have this romantic notion that since pixels are subject to the vagueness of binary counting in that each channel of a 24-bit image is measure in powers of 2 (0-055), and in either RGB or HSL, the pixels must live within the binary world. Then, extend that to the simple fact that pixels are digital thingies which are either there or their aren't, a classical binary choice. So, foolishly, I reached the conclusion that /counting/ pixels with computer-based prefixes like kilo or mega pixels would naturally follow the computer scienties penchant for base 2 arithmetic. I should never have been so quick to take on those far more important than me in these matters, particularly since /no one/ gives a damn exactly how many pixels there are except for very special casees - they just want to have a general, easy-to-remember way of deciding which camera to choose and which MP size to shoot, and let all the elitists go argue until they tire of the chase. I am so sorry to ever have ventured where only fools go. Now, anyone recognize satire and sarcasm when they see it? Why don't all you "experts" go back to worrying about things that concern them and just passively let obvious idiots like me to my own dillusions. SignatureATM, aka Jerry "Whether You Think You CAN Or CAN'T, You're Right." – Henry Ford > And that modification has also been made part of the metric system. A > kilobit is a thousand bits. A kibibit is 1024 bits. There is similar > nomenclature for other multiples. For more information you might want to > look at <http://en.wikipedia.org/wiki/Binary_prefix> The thing about those names is that they have a serious marketing problem: they are among the stupidest-sounding things I've ever heard in my entire life, and, as a computer person who uses those terms a lot and often has to differentiate between base-10 and base-2 usage, there is exactly zero chance that anyone will ever hear those names pass my lips or that I will ever use them in any context, ever, at all, forever, lest I sound like a drooling baby. In addition, I know exactly zero people who have any interest in ever using those names, ever, at all, forever. I'd rather be confused and/or confusing than ever use them. They really should be considered dead-on-arrival. SignatureJeremy | jeremy@exit109.com >> And that modification has also been made part of the metric system. A >> kilobit is a thousand bits. A kibibit is 1024 bits. There is similar [quoted text clipped - 14 lines] > > They really should be considered dead-on-arrival. Confidence, me lad, confidence. You worry too much about what people will think of you. Signature--John to email, dial "usenet" and validate (was jclarke at eye bee em dot net) > Confidence, me lad, confidence. You worry too much about what people will > think of you. I honestly don't think I could use them in conversation and maintain a straight face. SignatureJeremy | jeremy@exit109.com >> Confidence, me lad, confidence. You worry too much about what people >> will think of you. > > I honestly don't think I could use them in conversation and maintain a > straight face. Then don't. What's so great about maintaining a straight face? Signature--John to email, dial "usenet" and validate (was jclarke at eye bee em dot net) >>Confidence, me lad, confidence. You worry too much about what people will >>think of you. > > I honestly don't think I could use them in conversation and maintain a > straight face. Probably won't have to. The disk drive, optical disk and flash RAM vendors have adopted the SI meanings of the prefixes. I think DRAM manufacturers are the lone holdouts for the old way. That's probably because CPUs still want memory module boundaries to end on a power of two. Developers and sysadmins have to care about the real amount memory in a machine and about page boundaries, but virtually no one does. I've noticed that even among these folks when the kilo and mega prefixes are used colloquially the SI meaning is implied. When details matter kilo and mega (and giga) switch in everyone's mind to the other meaning automatically. > My calculator says that 3038x2012=6,112,456=5.8292 MP. You need > to divide by divide by 1,048,,576 (1024x1024) to convert normal > number to mega anything. That's bytes, not pixels. With pixels, "megapixel" means one million. SignatureJeremy | jeremy@exit109.com Not to be confused with megabits. bits, nibbles, bytes, words and doublewords. If a pixel were to be turned on or off, then we could consider it binary. In reality, a pixel or picture element can have 24 bits per pixel as in the True Color system. So, this whole bit about mega pixels can take on an entire new dimension in binary. As for the facts, go here: http://www.jamesshuggins.com/h/tek1/prefixes.htm -------------------------------------------------------------- Jeremy Nixon - Tues, Apr 11 2006 9:11 pm All Things Mopar <nunofy...@beez.wax> wrote: > My calculator says that 3038x2012=6,112,456=5.8292 MP. You need > to divide by divide by 1,048,,576 (1024x1024) to convert normal > number to mega anything. That's bytes, not pixels. With pixels, "megapixel" means one million. SignatureJeremy | jer...@exit109.com > Not to be confused with megabits. bits, nibbles, bytes, words and > doublewords. [quoted text clipped - 3 lines] > pixel as in the True Color system. So, this whole bit about mega > pixels can take on an entire new dimension in binary. A pixel is a physical object, not a piece of information. The data that are created when a pixel is turned on and off is called bits (or a bit stream), and they are binary. Merritt > Today Merritt Mullen commented courteously on the subject at > hand [quoted text clipped - 9 lines] > to divide by divide by 1,048,,576 (1024x1024) to convert normal > number to mega anything. ' You are correct that 3028x2012 = 6,112,456 or 6.1 million pixels. You are not correct about dividing by that binary factor since we are not counting bits here but simply pixels in an array. > What easy thing am I missing here, > other than Nikon obviously wants to advertise 6 MP and not a > tweak under Except the point is that Nikon does NOT advertise 6 MP, they advertise (and spec) 6.1 MP. And indeed, in picture element (pixel) count, the full RAW image does indeed consist of 6.1 MP (million pixels). Merritt Today Merritt Mullen commented courteously on the subject at hand >> Today Merritt Mullen commented courteously on the subject >> at hand [quoted text clipped - 23 lines] > element (pixel) count, the full RAW image does indeed > consist of 6.1 MP (million pixels). Whatever floats you people's boats. The laws of mathematics have no specific excepts for counting pixels mega is mega is mega. A /real/ example of linear mathematics is the small weight used in medicine the mg or milligram, with is really 1/1,000th of a gram and not 1.1024th. SignatureATM, aka Jerry "Whether You Think You CAN Or CAN'T, You're Right." – Henry Ford > Whatever floats you people's boats. The laws of mathematics > have no specific excepts for counting pixels mega is mega is > mega. A /real/ example of linear mathematics is the small > weight used in medicine the mg or milligram, with is really > 1/1,000th of a gram and not 1.1024th. In fact, the real reference is 1ml of H2O weighs 1g at STP. There are many reference masses kept in vaults used to standarize measurements against. SignatureThomas T. Veldhouse Key Fingerprint: 2DB9 813F F510 82C2 E1AE 34D0 D69D 1EDC D5EC AED1 > Today Merritt Mullen commented courteously on the subject at > hand [quoted text clipped - 5 lines] >> >> However, 3038x2012 is 6.092 MP, which rounds to 6.1 MP. Unlike MB = 1024 * 8 bits a MP = 1000000 pixels. SignatureThomas T. Veldhouse Key Fingerprint: 2DB9 813F F510 82C2 E1AE 34D0 D69D 1EDC D5EC AED1 >Greetings... >I was playing with my friend's D70s yesterday, and we noticed something >when we shot in RAW format - the image size was 3038x2012, yet JPG size >is 3008x2000 pixels. Nikon's specifications state 3008x2000 in every >document I've read, including the owners manual, and all online sources >say the same. >My Canon produces the same size RAW and JPG files. No, it does not. The standard Canon/Adobe converter output is the same size as the JPEG, but the RAW image is larger. For example, the "standard" 20D JPEG or conversion is 3504*2336, but the actual RAW image is 3522*2348. The RAW bitmap is even bigger than that, as there is a vertical strip of pixels on the left that is never exposed to light that is 74 pixels wide, and a horizontal strip on the top that is 12 pixels high. The actual RAW bitmap, including the blind pixels is 3596*2360. Signature<>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<> John P Sheehy <JPS@no.komm> ><<> <>>< <>>< ><<> <>>< ><<> ><<> <>>< >>I was playing with my friend's D70s yesterday, and we noticed something >>when we shot in RAW format - the image size was 3038x2012, yet JPG size [quoted text clipped - 5 lines] > >No, it does not. Yes it does...<phhht!> :-) I took my XT over to my friend's place and compared images in Picasa and in that program it does indeed have different file sizes for RAW and JPG. First time I've seen that since my software always shows the same 3456x2304 pixels regardless of format. Wow...two new things in two days...go figure. :-) |
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
|
©2009 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.