If this is your first visit, be sure to check out the FAQ by clicking the link above. You may have to register before you can post: click the register link above to proceed. To start viewing messages, select the forum that you want to visit from the selection below. |
|
|
Thread Tools | Rate Thread | Display Modes |
#46
|
|||
|
|||
film vs CMOS
In article , Mr. Man-wai
Chang wrote: I don't know much about photography films. clearly. And you might need to talk about the size (length x width) as well as the resolution of the senors and films! yep. But isn't film molecular level? everything is. Is your claim based on traditional size of film, which is 135? size doesn't change anything. film is very lossy and much less accurate than digital. But why can't we use a bigger film then? we can. there are larger film sizes, namely medium format and large format, but then you also have to use a larger digital sensor to match. Should we always compare 135 film against CMOS sensors of different size? always the same size format. otherwise it's not a valid comparison. |
Ads |
#47
|
|||
|
|||
Is VLC 3.0.3 for Windows 7?
"Mr. Man-wai Chang" wrote in newskn432$pvc$1
@toylet.eternal-september.org: On 8/11/2018 10:47 PM, Stephen wrote: [] The original capture format is "not lossy" - in the sense that you have all the info you are ever going to get. - however that is really a theoretical thing, since sensors are not ideal, and there will be colour and brightness distortins, or effects from adjacent pixels, timing errors. Once you have a source then it can be compressed. But how do you get a 100% TRUE lossless original? Using good, old film-based cameras? BTW, I am thinking about court use.... Even film is not 'lossless' in reference to capturing all the information in an image. In this case the limitation is physical. A camera lense can only resolve a certain amount of the incoming image, due to the fact it is 'compressing' the image to fit the film format (35mm, 120, 70mm, etc). Then the film has a limit due to its formulation due to the size of the individual 'grains' in the emulsion. The finer grained the emulsion, the more detail that can be captured. That is why two photos of the same scene can have very different amounts of detail captured. Then, there is a small amount of loss when a photographic image is copied. Each 'generation' of copying results in some loss of resolution (data). In real life it usually isn't noticeable except at larger magnifications, such as accurs in watching a movie in a theatre, or when parts of the image are 'blown up' into a larger format. So yes, in court use, the closer one can get to the original image captured the better one is. That is why 'chain of possession' is such an important part of evidence. |
#48
|
|||
|
|||
how original is an original image?
On 8/12/2018 1:46 AM, Tim wrote:
So yes, in court use, the closer one can get to the original image captured the better one is. That is why 'chain of possession' is such an important part of evidence. But how do you determine how close a digital image get to the original without a reference? You have to have a control as in experiment! And which pair(s) of human eyes should we use out of billions to make the decision? Oh well... the topic is becoming "just trust someone but not me"! -- @~@ Remain silent! Drink, Blink, Stretch! Live long and prosper!! / v \ Simplicity is Beauty! /( _ )\ May the Force and farces be with you! ^ ^ (x86_64 Ubuntu 9.10) Linux 2.6.39.3 不借貸! 不詐騙! 不*錢! 不援交! 不打交! 不打劫! 不自殺! 不求神! 請考慮綜援 (CSSA): http://www.swd.gov.hk/tc/index/site_...sub_addressesa |
#49
|
|||
|
|||
how original is an original image?
In article , Mr. Man-wai
Chang wrote: But how do you determine how close a digital image get to the original without a reference? You have to have a control as in experiment! the reference is the original |
#50
|
|||
|
|||
how original is an original image?
On 8/12/2018 2:08 AM, nospam wrote:
In article , Mr. Man-wai Chang wrote: But how do you determine how close a digital image get to the original without a reference? You have to have a control as in experiment! the reference is the original In a court trial, how do you do that? You cannot take the physical reality into a court... there is also the time factor. Whatever happened in reality might not repeat itself before the court. -- @~@ Remain silent! Drink, Blink, Stretch! Live long and prosper!! / v \ Simplicity is Beauty! /( _ )\ May the Force and farces be with you! ^ ^ (x86_64 Ubuntu 9.10) Linux 2.6.39.3 不借貸! 不詐騙! 不*錢! 不援交! 不打交! 不打劫! 不自殺! 不求神! 請考慮綜援 (CSSA): http://www.swd.gov.hk/tc/index/site_...sub_addressesa |
#51
|
|||
|
|||
how original is an original image?
On 8/11/2018 2:08 PM, nospam wrote:
In article , Mr. Man-wai Chang wrote: But how do you determine how close a digital image get to the original without a reference? You have to have a control as in experiment! the reference is the original It is my understanding that when an electronic image is modified it can be detected by studying the pixel arrangement in the file. Therefore if you have an electronic image of the original document that was made from the original document then it is close as you can get to the original with out having the original. Unless when the file is examined by an expert they detect fragments in the pixel that indicate the images was modified. There is one problem with the above statement If you are holding the original document, you can see if any pertinent notes were made on the reverse. Unless otherwise note on the front page you would have no way of know the note existed from an image of the front page. For anything but legal evidence presented to a court, If I would not worry about how close an image is to the original. -- 2018: The year we learn to play the great game of Euchre |
#52
|
|||
|
|||
Is VLC 3.0.3 for Windows 7?
On Sat, 11 Aug 2018 00:50:41 -0400, Paul wrote:
J. P. Gilliver (John) wrote: Thanks. I was just wondering if, once the initial compression had been done, any of the various formats can be converted into each other without _further_ loss. Sounds like you don't know - fair enough, nor do I, hence my asking the question! The "formats" part can be broken down into two pieces. The outside part is the "container". .mkv , .mov , .avi are containers Inside the container are video and audio codecs. SNIP Hi Paul, I know what you wanted to say but that last part didn't come out right. There are no codecs inside the container. That would be quite inefficient. ;-) Otherwise, excellent summary. |
#53
|
|||
|
|||
how original is an original image?
On 8/12/2018 2:56 AM, Wolf K wrote:
Considering that many people can't tell the difference between vertical and horizontal phone-videos, passing off fakes is easy anyhow. However, AI created fake videos have a problem with eyes: we blink, the blink rate varies with how we feel about we're saying, and current AI techniques can't fake natural blinks. Yet. Then you need to look at the video frame-by-frame? Need a prolonged trial then, for just viewing the video evidences carefully ... -- @~@ Remain silent! Drink, Blink, Stretch! Live long and prosper!! / v \ Simplicity is Beauty! /( _ )\ May the Force and farces be with you! ^ ^ (x86_64 Ubuntu 9.10) Linux 2.6.39.3 不借貸! 不詐騙! 不*錢! 不援交! 不打交! 不打劫! 不自殺! 不求神! 請考慮綜援 (CSSA): http://www.swd.gov.hk/tc/index/site_...sub_addressesa |
#54
|
|||
|
|||
Is VLC 3.0.3 for Windows 7?
In message , Stephen
writes: On Sat, 11 Aug 2018 04:13:21 +0100, "J. P. Gilliver (John)" wrote: (And thanks Paul as well.) In message , Tim writes: "J. P. Gilliver (John)" wrote in : Are these assorted video conversions lossless, or is there (possibly only theoretical) degradation at each conversion, like there is most of the time when going JPEG to JPEG for images or mp3 to mp3 for audio? [] The original capture format is "not lossy" - in the sense that you have all the info you are ever going to get. - however that is really a theoretical thing, since sensors are not ideal, and there will be colour and brightness distortins, or effects from adjacent pixels, timing errors. Good point, but I was talking about conversion between different video file formats. [] I have not noticed any degredation doing file conversions, but I am basically measuring with Mark I eyeball, so precision is probably low. Oh, still using the MkI? (-: All video file formats can be and usually are compressed in some fashion. Depending on the compression method used it can be lossless or lossy. A data stream using lossy compression has already thrown away some information, and as such is already degraded from the original. Transcoding from one lossy format to another will lose some information, the amount of the loss will depend on how lossy the compressions used are. In this it is I don't think that's _necessarily_ so, if the conversion is just a repackaging, or the second compression uses the same algorithm and settings as the first. But what I was really asking about was that someone (actually I think two someones - they've been snipped now) listed a lot of conversions, and there was no indication whether any of them were lossless (i. e. resulted in no _further_ degradation). similiar to your JPEG to JPEG example, since JPEG is a lossy compression format. Thanks. I was just wondering if, once the initial compression had been done, any of the various formats can be converted into each other without _further_ loss. Sounds like you don't know - fair enough, nor do I, hence my asking the question! There will always be some further distortion in signal since the compression already done will have added compromises in signal from the original, so you are further from an "ideal picture" starting point. Although not quite true (see above), it's probably best to assume that yes, any further conversions do degrade. I'm a little saddened that there seems little interest in establishing where the is _not_ the case, though. [] - last time I was involved the favorite at 1 place was JPEG2000 - the wavelet oriented schemes seem to degrade more gracefully and survive multiple passes with less overall impact. Interesting. Not one you hear of much these days. A heavily compressed stream being recompressed to a different format seems to generate more artifacts - ie the 2 compression systems can interact to give more artefacts and distrotion in the resulting output Yes, I'd have expected that. [] -- J. P. Gilliver. UMRA: 1960/1985 MB++G()AL-IS-Ch++(p)Ar@T+H+Sh0!:`)DNAf "quidquid latine dictum sit, altum viditur". ("Anything is more impressive if you say it in Latin") |
#55
|
|||
|
|||
Is VLC 3.0.3 for Windows 7?
On Sat, 11 Aug 2018 22:15:01 +0100, "J. P. Gilliver (John)"
wrote: In message , Stephen writes: On Sat, 11 Aug 2018 04:13:21 +0100, "J. P. Gilliver (John)" wrote: (And thanks Paul as well.) In message , Tim writes: "J. P. Gilliver (John)" wrote in : Are these assorted video conversions lossless, or is there (possibly only theoretical) degradation at each conversion, like there is most of the time when going JPEG to JPEG for images or mp3 to mp3 for audio? [] The original capture format is "not lossy" - in the sense that you have all the info you are ever going to get. - however that is really a theoretical thing, since sensors are not ideal, and there will be colour and brightness distortins, or effects from adjacent pixels, timing errors. Good point, but I was talking about conversion between different video file formats. [] I have not noticed any degredation doing file conversions, but I am basically measuring with Mark I eyeball, so precision is probably low. Oh, still using the MkI? (-: All video file formats can be and usually are compressed in some fashion. Depending on the compression method used it can be lossless or lossy. A data stream using lossy compression has already thrown away some information, and as such is already degraded from the original. Transcoding from one lossy format to another will lose some information, the amount of the loss will depend on how lossy the compressions used are. In this it is I don't think that's _necessarily_ so, if the conversion is just a repackaging, or the second compression uses the same algorithm and settings as the first. But what I was really asking about was that someone (actually I think two someones - they've been snipped now) listed a lot of conversions, and there was no indication whether any of them were lossless (i. e. resulted in no _further_ degradation). similiar to your JPEG to JPEG example, since JPEG is a lossy compression format. Thanks. I was just wondering if, once the initial compression had been done, any of the various formats can be converted into each other without _further_ loss. Sounds like you don't know - fair enough, nor do I, hence my asking the question! There will always be some further distortion in signal since the compression already done will have added compromises in signal from the original, so you are further from an "ideal picture" starting point. Although not quite true (see above), it's probably best to assume that yes, any further conversions do degrade. I'm a little saddened that there seems little interest in establishing where the is _not_ the case, though. [] - last time I was involved the favorite at 1 place was JPEG2000 - the wavelet oriented schemes seem to degrade more gracefully and survive multiple passes with less overall impact. Interesting. Not one you hear of much these days. Because the compression is done frame by frame and keeps more info - so lower comression ratios. "lossless" JPEG will reduce a 1.5 Gbps HD uncompressed 1080i video stream to maybe 300 Mbps. A lossy JPEG trades off "quality"for compression like any of the other systems - but the systems I worked on had a sweet spot for the types of sources around 20 - 50 Mbps. distribution systems with more limited bandwdith like TV and DVD need higher compression, so JPEG isnt a good choice. A heavily compressed stream being recompressed to a different format seems to generate more artifacts - ie the 2 compression systems can interact to give more artefacts and distrotion in the resulting output Yes, I'd have expected that. [] Agreed. But it seems to hold true even where different versions of the same compression family are used. You probably have more room to improve the results if you are not having to deal with a real time stream (where you only get to do a single pass unless you are adding a lot of delay). -- Stephen |
#56
|
|||
|
|||
film vs CMOS
On Sun, 12 Aug 2018 00:54:04 +0800, "Mr. Man-wai Chang"
wrote: On 8/12/2018 12:50 AM, nospam wrote: In article , Mr. Man-wai Chang wrote: But how do you get a 100% TRUE lossless original? Using good, old film-based cameras? film is more lossy than digital. I don't know much about photography films. And you might need to talk about the size (length x width) as well as the resolution of the senors and films! But isn't film molecular level? Film is quantized to grain size. -- John Larkin Highland Technology, Inc lunatic fringe electronics |
#57
|
|||
|
|||
Is VLC 3.0.3 for Windows 7?
Char Jackson wrote:
On Sat, 11 Aug 2018 00:50:41 -0400, Paul wrote: J. P. Gilliver (John) wrote: Thanks. I was just wondering if, once the initial compression had been done, any of the various formats can be converted into each other without _further_ loss. Sounds like you don't know - fair enough, nor do I, hence my asking the question! The "formats" part can be broken down into two pieces. The outside part is the "container". .mkv , .mov , .avi are containers Inside the container are video and audio codecs. SNIP Hi Paul, I know what you wanted to say but that last part didn't come out right. There are no codecs inside the container. That would be quite inefficient. ;-) Otherwise, excellent summary. Yeah, that should have been "video and audio streams". Paul |
#58
|
|||
|
|||
film vs CMOS
On 8/12/2018 1:19 AM, nospam wrote:
Should we always compare 135 film against CMOS sensors of different size? always the same size format. otherwise it's not a valid comparison. In reality, we just need to do the job right and fair, not about comparison or superiority! What if... a big what if.... all CMOS on Earth were fried by solar storm? Maybe that explained why a man is up there in ISS. -- @~@ Remain silent! Drink, Blink, Stretch! Live long and prosper!! / v \ Simplicity is Beauty! /( _ )\ May the Force and farces be with you! ^ ^ (x86_64 Ubuntu 9.10) Linux 2.6.39.3 不借貸! 不詐騙! 不*錢! 不援交! 不打交! 不打劫! 不自殺! 不求神! 請考慮綜援 (CSSA): http://www.swd.gov.hk/tc/index/site_...sub_addressesa |
#59
|
|||
|
|||
film vs CMOS
In article , Mr. Man-wai
Chang wrote: Should we always compare 135 film against CMOS sensors of different size? always the same size format. otherwise it's not a valid comparison. In reality, we just need to do the job right and fair, not about comparison or superiority! you're the one making comparisons. What if... a big what if.... all CMOS on Earth were fried by solar storm? Maybe that explained why a man is up there in ISS. what if you stopped posting rubbish? |
#60
|
|||
|
|||
film vs CMOS
On 8/12/2018 11:22 AM, nospam wrote:
In reality, we just need to do the job right and fair, not about comparison or superiority! you're the one making comparisons. What if... a big what if.... all CMOS on Earth were fried by solar storm? Maybe that explained why a man is up there in ISS. what if you stopped posting rubbish? Well, calm down... professor!? Let's continue later. -- @~@ Remain silent! Drink, Blink, Stretch! Live long and prosper!! / v \ Simplicity is Beauty! /( _ )\ May the Force and farces be with you! ^ ^ (x86_64 Ubuntu 9.10) Linux 2.6.39.3 不借貸! 不詐騙! 不*錢! 不援交! 不打交! 不打劫! 不自殺! 不求神! 請考慮綜援 (CSSA): http://www.swd.gov.hk/tc/index/site_...sub_addressesa |
Thread Tools | |
Display Modes | Rate This Thread |
|
|