BPG file format trounces JPG for quality at high compression rates

[Aglet]

this is interesting, not sure where I'd make use of it altho it would make storage of my processed, finished files a bit smaller, perhaps.

http://xooyoozoo.github.io/yolo-octo-bugfixes/#zoo-bird-head&jpg=s&bpg=s

some story at

www.imaging-resource.com/news/2014/12/15/new-bpg-image-format-looks-to-beat-out-jpeg-with-same-quality-at-half-the-s

 

[Mt Spokane Photography]

There have been a number of potential replacements for jpeg, most notable flop was the jpeg2000.

Back in the early 1990's, there was a good reason to change from Gif to Jpeg, and that was due to very limited internet bandwidth that most of us had. so we accepted poorer quality for usefulness.

While photographers like myself would like to see a better image format, its a really hard sell for the average person. In 1992, there were few digital cameras, but now, there are trillions of jpeg files, and 99%+ are happy with what they have. That means that there would be 3 or 4 major formats, raw, jpeg, DNG, and BPG. If they started making cameras that had a option to save to BPG, few would use it, just hard core photographers, and they would have to convert to jpg for the next 5-10 years just to share files.

I really wonder if there is enough demand to go thru the pain of converting cameras, scanners, software, internet browsers, and a lot of other stuff to BPG over the next several years. There is also the thorny subject of silent patents which will suddenly appear after it gets adapted by a significant number of users, and require all software and hardware makers to pay royalties. Those companies will see BPG as a big risk.

 

[JonAustin]

There have been a number of potential replacements for jpeg, most notable flop was the jpeg2000.

Back in the early 1990's, there was a good reason to change from Gif to Jpeg, and that was due to very limited internet bandwidth that most of us had. so we accepted poorer quality for usefulness.

While photographers like myself would like to see a better image format, its a really hard sell for the average person. In 1992, there were few digital cameras, but now, there are trillions of jpeg files, and 99%+ are happy with what they have. That means that there would be 3 or 4 major formats, raw, jpeg, DNG, and BPG. If they started making cameras that had a option to save to BPG, few would use it, just hard core photographers, and they would have to convert to jpg for the next 5-10 years just to share files.

I really wonder if there is enough demand to go thru the pain of converting cameras, scanners, software, internet browsers, and a lot of other stuff to BPG over the next several years. There is also the thorny subject of silent patents which will suddenly appear after it gets adapted by a significant number of users, and require all software and hardware makers to pay royalties. Those companies will see BPG as a big risk.

I concur with your JPEG2000 flop comment.

I'm not so sure about the issues transitioning to the file format, however. Many photogs appear to prefer the DNG format (I've never used it personally), but how many cameras support that internally? Those who want to use it convert their RAW files during post.

For RAW shooters, it would be easy enough to convert to a new format, just as soon as the photo editing application developers added it to their Save / Save As dropdown lists. (Or conversion could be accomplished by means of a third party app.) And if it became popular, the various photo viewer apps / applets could add it to their lists of supported formats quickly enough.

I save a lot of stuff in TIFF/ZIP format, and routinely convert some of those to JPG or BMP, if needed for another application or wider distribution. Once I've used the JPG/BMP, I toss it, but save the TIFF/ZIP file.

 

[Mt Spokane Photography]

There have been a number of potential replacements for jpeg, most notable flop was the jpeg2000.

Back in the early 1990's, there was a good reason to change from Gif to Jpeg, and that was due to very limited internet bandwidth that most of us had. so we accepted poorer quality for usefulness.

While photographers like myself would like to see a better image format, its a really hard sell for the average person. In 1992, there were few digital cameras, but now, there are trillions of jpeg files, and 99%+ are happy with what they have. That means that there would be 3 or 4 major formats, raw, jpeg, DNG, and BPG. If they started making cameras that had a option to save to BPG, few would use it, just hard core photographers, and they would have to convert to jpg for the next 5-10 years just to share files.

I really wonder if there is enough demand to go thru the pain of converting cameras, scanners, software, internet browsers, and a lot of other stuff to BPG over the next several years. There is also the thorny subject of silent patents which will suddenly appear after it gets adapted by a significant number of users, and require all software and hardware makers to pay royalties. Those companies will see BPG as a big risk.

I concur with your JPEG2000 flop comment.

I'm not so sure about the issues transitioning to the file format, however. Many photogs appear to prefer the DNG format (I've never used it personally), but how many cameras support that internally? Those who want to use it convert their RAW files during post.

For RAW shooters, it would be easy enough to convert to a new format, just as soon as the photo editing application developers added it to their Save / Save As dropdown lists. (Or conversion could be accomplished by means of a third party app.) And if it became popular, the various photo viewer apps / applets could add it to their lists of supported formats quickly enough.

I save a lot of stuff in TIFF/ZIP format, and routinely convert some of those to JPG or BMP, if needed for another application or wider distribution. Once I've used the JPG/BMP, I toss it, but save the TIFF/ZIP file.

You can start using BPG right now, but sharing it with anyone else might be very limited. Unless I could do something with it, it would not be worth the trouble. That's the issue, a reasonable number of users is needed to get the players to adapt something new. Who is going to step up and do it? Adobe already has PDF and DNG that they are pushing.

Perhaps Mozilla might put it in their browsers. Apple seems to be dropping photo apps, so its doubtful that they would be pushing it.

Commercial enterprises want payback for $$ spent, so a profit angle is needed.

 

[Marsu42]

There have been a number of potential replacements for jpeg, most notable flop was the jpeg2000.

Not quite, it's used a lot in video processing as individual compression of frames. But for stills, you're correct - all these potential sucessors lack the "killer feature" to replace jpeg, the last duds were Google's webp or Microsoft's jpeg-ex. Having transparancy obivously doesn't cut it, the web still uses png as transparancy in *large* images is not common.

this is interesting, not sure where I'd make use of it altho it would make storage of my processed, finished files a bit smaller, perhaps.

The 14bit color depth (like current dslrs) indeed looks appealing - but it won't make a difference for final exported images. And then the jpeg supports 12bit for a long time, and which imaging apps outside niches use it? Right.

Otherwise, bpg (h265) is about the same as webp (vc8): A image format based on a video compression. Nice space saving, but who cases in times of cheap hd space and fast Internet. But bpg doesn't have a worldwide enterprise supporting it. This basically means "forget about it".

 

[Mt Spokane Photography]

I do think that its time for a improvement, for color corrected web browsers, etc. The real difficulty is in getting the other 99.9% to agree.

 

[Marsu42]

I do think that its time for a improvement, for color corrected web browsers, etc.

I don't get that - why is a file format change necessary, my understanding is jpeg/png work with embedded icc profiles just fine (or jpeg with the appropriate tags for srgb/argb)? The only problem is that chrome's color management is broken atm.

 

[ajfotofilmagem]

The possibility of having a universal format with the new 14 bit per color is interesting. But the success of a JPEG substitute depends on its adoption by Lightroom, Photoshop, Windows and Mac OS.

However, for users to change their workflow, saving the processed images in a new format, it would be good that Canon and Nikon had this format option, in addition to JPEG and RAW traditional.

 

[Marsu42]

The possibility of having a universal format with the new 14 bit per color is interesting. But the success of a JPEG substitute depends on its adoption by Lightroom, Photoshop, Windows and Mac OS.

Photoshop had plugins for most aspiring formats, to no avail. Adobe learned from this and Lightroom is very limited, supporting no output plugins unless you code lua to call ImageMagick. Windows or MacOS are no apps, and I guess few photogs care about what their built-in viewers support.

The real test for widespread picture formats are the web browsers - and for example here mng failed big time to replace gif even if there were very good reasons to do so. With more and more devices supporting browsers, it'll get even harder to adopt a new standard.

However, for users to change their workflow, saving the processed images in a new format, it would be good that Canon and Nikon had this format option, in addition to JPEG and RAW traditional.

Following that logic, you'd have a choice of jpeg-lossless/jpeg-12bit, jpeg2k, webp, jpeg-xr by now. Of course all have to be supported ad infinitum to protect customers' libraries...

 

[adhocphotographer]

Better compression whilst maintaining IQ is never bad... it will make the interweb even faster, especially on mobile devices.

However, until this is integrated as a standard in webpages and browsers it is just nice maths. It took over 10 years for mp3's to become moderately mainstream (first developed in early 80s). Only time will tell.

 

[dgatwood]

Better compression whilst maintaining IQ is never bad... it will make the interweb even faster, especially on mobile devices.

However, until this is integrated as a standard in webpages and browsers it is just nice maths. It took over 10 years for mp3's to become moderately mainstream (first developed in early 80s). Only time will tell.

I think you're off by about a decade there. The concept of psychoacoustic masking was conceived in the early 80s, and its predecessors were in early development at that point, but MP3 itself wasn't publicly released as a standard until 1993, and they didn't even start writing the first line of code for the first implementation of MP3 until 1991. That first reference implementation wasn't finished until 1996, and wasn't published publicly until 1998. By early 1999, just about everybody was using it.

PNG was adopted just as quickly, with just a little over a year between when it came out and when it was supported by every major browser (both of them...).

Standards usually either are immediately adopted broadly or wither on the vine. It is pretty rare for a standard to get a slow start and then gain serious traction, with the exception of standards that were simply too complex for the computational power at the time (MPEG-2 video comes to mind). So assuming there's a published reference implementation of this algorithm, it will probably either be broadly deployed within 18 months or will be declared a dead standard.

 

[adhocphotographer]

I think you're off by about a decade there. The concept of psychoacoustic masking was conceived in the early 80s, and its predecessors were in early development at that point, but MP3 itself wasn't publicly released as a standard until 1993, and they didn't even start writing the first line of code for the first implementation of MP3 until 1991. That first reference implementation wasn't finished until 1996, and wasn't published publicly until 1998. By early 1999, just about everybody was using it.

I stand corrected...