Forum.ArchosFans.com
Archos 80 G9 1.5Ghz 1GB RAM ICS 8GB: Buy now (free shipping)
Archos 80 G9 1.5Ghz 1GB RAM ICS 250GB: Buy now (free shipping)
Archos 101 G9 1.5Ghz 1GB RAM ICS 8GB: Buy now (free shipping)
Archos 101 G9 1.5Ghz 1GB RAM ICS 250GB: Buy now (free shipping)
 * Register    * Login 

It is currently Fri Jul 25, 2014 1:59 pm

All times are UTC + 1 hour



Post new topic  Reply to topic  [ 91 posts ]  Go to page Previous  1, 2, 3, 4, 5  Next
Author Message
PostPosted: Fri Aug 20, 2010 11:18 pm 
Offline
Archos Guru
Archos Guru

Joined: Sat Sep 20, 2008 10:52 am
Posts: 5191
Location: Hanapepe, HI
Since the A32 has a camera I bet the others up the line do as well!?!

Allen

_________________
-An old geek 8)
5IT|A70S|A101


Top
   
 
PostPosted: Sat Aug 21, 2010 1:07 am 
Offline
Archos Guru
Archos Guru

Joined: Thu Dec 10, 2009 12:10 pm
Posts: 6598
One can but hope. I'm hoping the other missing features from the Google requirements are there on the larger tablets (where there is more space for them) as well. I'm hoping that by repeating this sentiment over and over, that it will somehow magically come true... ;)

Michael


Top
   
 
PostPosted: Sat Aug 21, 2010 1:16 am 
Offline
Site Admin
Site Admin

Joined: Sun Nov 27, 2005 2:40 am
Posts: 7052
Location: Copenhagen
AllenPapapetrou wrote:
Since the A32 has a camera I bet the others up the line do as well!?!

Allen


If they got a decent 720p camcorder with nice low encoding bitrate compression at good quality, good enough battery runtime when using it, maybe even removable battery or at least same Mini Dock charging option (I could use a battery pack in the back pack), microphone inputs both analog and bluetooth wireless ones, I might probably use that as my main HD video-blogging camcorder instead of my Sanyo HD1000. As one-click youtube upload over WiFi and 3G, as the eventuality that HD camcorder mode even could support simultaneous live broadcast on WiFi or 3G to ustream or similar site with overlay live chat at the same time as it would be doing the full quality HD recording to internal storage, those features are only possible with Android so those would make it the best HD camcorder for video-blogging in the world.


Top
   
 
PostPosted: Sat Aug 21, 2010 1:55 am 
Offline
Archos Guru
Archos Guru

Joined: Mon Oct 22, 2007 12:59 am
Posts: 2880
Location: Toronto, Ontario
Charbax wrote:
AllenPapapetrou wrote:
Since the A32 has a camera I bet the others up the line do as well!?!

Allen


If they got a decent 720p camcorder with nice low encoding bitrate compression at good quality, good enough battery runtime when using it, maybe even removable battery or at least same Mini Dock charging option (I could use a battery pack in the back pack), microphone inputs both analog and bluetooth wireless ones, I might probably use that as my main HD video-blogging camcorder instead of my Sanyo HD1000. As one-click youtube upload over WiFi and 3G, as the eventuality that HD camcorder mode even could support simultaneous live broadcast on WiFi or 3G to ustream or similar site with overlay live chat at the same time as it would be doing the full quality HD recording to internal storage, those features are only possible with Android so those would make it the best HD camcorder for video-blogging in the world.


DO NOT DITCH YOUR CAMCORDER!!!

No device thats primary function is not camcorder but has a video camera will ever hope to reach the quality of your 720p camcorder, now, or in the distant future. The A32, being a $150 device, will be lucky to have 1/10th of your camcorder's video quality.

I'm not criticizing the 32, I'm just saying the truth. You can't realistically expect that. An analogy would be expecting the 32's internal speakers to replace your stereo (except this would not ever be possible with advancement of technology). Your expectations are hard to achieve by sub $200 *dedicated* camcorders.

_________________
Gr.9 tech lover


Top
   
 
PostPosted: Sat Aug 21, 2010 11:32 am 
Offline
Archos Guru
Archos Guru

Joined: Thu Dec 10, 2009 12:10 pm
Posts: 6598
Even an advancement in technology is not likely to replace a camera's most important piece, the glass (lens). All else being equal, given the same CCD, the cam with the bigger hunk o' glass at the right distance from the CCD will always produce a better image.

Michael


Top
   
 
PostPosted: Sat Aug 21, 2010 1:04 pm 
Offline
Site Admin
Site Admin

Joined: Sun Nov 27, 2005 2:40 am
Posts: 7052
Location: Copenhagen
map002 wrote:
Even an advancement in technology is not likely to replace a camera's most important piece, the glass (lens). All else being equal, given the same CCD, the cam with the bigger hunk o' glass at the right distance from the CCD will always produce a better image.

Michael


Sure, while optics are important to make a good image, the sound quality, low light performance, auto-focus speed, compression quality, all those things may be just as important in making and releasing good videos. And the feature of one-click uploading to Youtube over WiFi and over 3G, the feature of editing Youtube title/description/tags directly from the camera, the possibility that one HD camcorder software also includes live ustream uploading with chat as well, all these features together, and as long as sound quality, low light performance, auto-focus speed and compression quality are good, as long as the optics are best for the phone-camcorder size, I think overall could provide the absolute best HD camcorder for HD video-blogging.

If you are trying to make a hollywood masterpiece, of course, at least get one of those new DSLR that record awesome 1080p videos. But for recording 5 hours per day in consumer electronics shows or for interviewing a lot of people in the streets, I don't need to record much depth of field and don't need the optics of a large camcorder.

The main problem with using an Archos of phone-style device as camcorder, is that people being interviewed might not take one as seriously as holding a larger camcorder. Kind of same problem having to justify to every interviewee why one is using a DSLR to film videos.


Top
   
 
PostPosted: Sat Aug 21, 2010 1:14 pm 
Offline
Archos Guru
Archos Guru

Joined: Thu Dec 10, 2009 12:10 pm
Posts: 6598
Well, depending on lens placement, angles and case material type, maybe you can do something like this to your Archos...

http://www.youtube.com/watch?v=cZN_UYPhwpM

:D

Michael


Top
   
 
PostPosted: Sat Aug 21, 2010 2:39 pm 
Offline
Archos Guru
Archos Guru

Joined: Mon Oct 22, 2007 12:59 am
Posts: 2880
Location: Toronto, Ontario
Charbax, you use your camcorder for almost all indoor shootings. With the 32, you won't even get a usable video quality from the 32. Dedicated $150 video cameras struggle with indoor, low light videos. If a dedicated video camera, which spends a good portion of the money on the lens and the imaging system, struggle with low light videos, what makes you think the 32, that spends maybe $1 for the camera components, can do a good job of shooting? The 32 will crash and burn if you try to take low light videos. Not only that, dedicated camcorders all have some sort of image processing (mainly denoising and minimizing distortion) to improve video quality. They are specific to the lens. Will the 32 (or most cameraphones) have that? Certainly not.

@map002. You may improve quality with that hack, but is that doable on the 32, or even most cameras? I highly doubt it. Even if you can, it will only help the quality so much. The low light videos will still be horrible because of the sesame seed sized sensor. There is a reason no non DSLR cameras can have good image quality above ISO 400, their sensor is too small. DSLRs take fantastic low light images because (not even mentioning the lens) they have a sensor 12-13 times the size of normal point and shoot cameras.

_________________
Gr.9 tech lover


Top
   
 
PostPosted: Sat Aug 21, 2010 3:13 pm 
Offline
Archos Guru
Archos Guru

Joined: Thu Dec 10, 2009 12:10 pm
Posts: 6598
Hmm, seems I wasn't as clear as I meant to be. My bad, I'll try to be clearer...
Chengbin wrote:
@map002. You may improve quality with that hack, but is that doable on the 32, or even most cameras? I highly doubt it.

I don't know what the cases on Gen 8 will be like or whether the cases will be shaped like or what they will be made out of or where the lenses will be placed. That's why I said...
Code:
...depending on lens placement, angles and case material type...

I was careful to qualify my statement.
Chengbin wrote:
Even if you can, it will only help the quality so much. The low light videos will still be horrible because of the sesame seed sized sensor.

Again, I qualified my original statement...
Code:
All else being equal, given the same CCD...

...CCD being the sensor.

I have no idea about what size or quality the CCD on all of the different Gen 8 tablets. Some of them might be big/good enough to produce acceptable quality for YouTube style videos shot indoors, or they may not.

Michael


Top
   
 
PostPosted: Sat Aug 21, 2010 3:22 pm 
Offline
Site Admin
Site Admin

Joined: Sun Nov 27, 2005 2:40 am
Posts: 7052
Location: Copenhagen
New 720p mobile phone cameras such as the ones in Droid X, iPhone 4, HTC EVO and Samsung Galaxy S all have decent low light quality. Low light is not only a sensor size issue, there is aperture and resolution. 720p video which is still arguably optimal resolution for web distribution is less than 1 megapixel so it can do pretty fine in low light, even on mobile phone sized cameras with cheap sensors.

If you try take 12 megapixel pictures with a mobile phone, sure low light will be terrible. But for 720p video and with a really fast and high quality encoder, it can be just great even with a small low cost sensor.

A good modern HD mobile phone camera is more like $12 in component cost. I'd film with a 720p camera at the least, so Archos 43 or 48 at the least.

Image processing is getting amazing with OMAP3630 and will be absolutely incredible with OMAP4430 coming out for next year. And prices are coming down so all cheap cameras will have absolutely fantastic compression quality built-in. Look forward to it. With ARM Cortex A9 and the new DSPs that go with it, all camcorders even cheap ones will encode H264 in real-time at about 98% the same quality that you can get from encoding multiple passes using a quad-core intel processor desktop. Yup you heard it right. Near full high profile quality done by the camera itself in real-time, no re-encoding needed.

These new technologies are worth more than just the image quality, at least for uses where you are not trying to film a hollywood masterpiece. The physical advantage of having your HD camcorder in the pocket at all times and have it in a cheaply affordable device that you can use for many other things can outweigh the image quality of getting one of the $1000 or much more latest DSLRs.

If there was a $1000 DSLR from Nikon, Canon, Sony, Panasonic or other with Android features such as WiFi direct upload, title/description/tags editing, in-camera split/join fast editing, OMAP4 quality compressing, as well as built-in Wifi and 3G streaming to ustream and similar video streaming sites with overlay chats, then for sure I would buy that. But for now the morons making those high end cameras just don't understand the opportunity of providing tools for the citizen media revolution.


Top
   
 
PostPosted: Sat Aug 21, 2010 3:58 pm 
Offline
Archos Guru
Archos Guru

Joined: Mon Oct 22, 2007 12:59 am
Posts: 2880
Location: Toronto, Ontario
map002 wrote:
Hmm, seems I wasn't as clear as I meant to be. My bad, I'll try to be clearer...
Chengbin wrote:
@map002. You may improve quality with that hack, but is that doable on the 32, or even most cameras? I highly doubt it.

I don't know what the cases on Gen 8 will be like or whether the cases will be shaped like or what they will be made out of or where the lenses will be placed. That's why I said...
Code:
...depending on lens placement, angles and case material type...

I was careful to qualify my statement.
Chengbin wrote:
Even if you can, it will only help the quality so much. The low light videos will still be horrible because of the sesame seed sized sensor.

Again, I qualified my original statement...
Code:
All else being equal, given the same CCD...

...CCD being the sensor.

I have no idea about what size or quality the CCD on all of the different Gen 8 tablets. Some of them might be big/good enough to produce acceptable quality for YouTube style videos shot indoors, or they may not.

Michael


My post was not directed at you as you had a valid post with no misinformation.

Charbax wrote:
New 720p mobile phone cameras such as the ones in Droid X, iPhone 4, HTC EVO and Samsung Galaxy S all have decent low light quality.


That's because they're comparing to other phones's cameras, not to dedicated camcorders.

Charbax wrote:
Low light is not only a sensor size issue, there is aperture and resolution. 720p video which is still arguably optimal resolution for web distribution is less than 1 megapixel so it can do pretty fine in low light, even on mobile phone sized cameras with cheap sensors. If you try take 12 megapixel pictures with a mobile phone, sure low light will be terrible. But for 720p video and with a really fast and high quality encoder, it can be just great even with a small low cost sensor.


I won't even bother to argue anymore until you read up and get your facts straight. Read up If we use a car analogy, you're trying to get sedan performance from a kart's engine. It's not possible. A kart's engine does not have anywhere the near the displacement needed for such performance. A engine's displacement is *directly* related to its performance, the same as a camera's sensor size's relation is directly related to its image quality.

Charbax wrote:
A good modern HD mobile phone camera is more like $12 in component cost. I'd film with a 720p camera at the least, so Archos 43 or 48 at the least.


While that might be true, they're certainly not used on phones, or the $150 32. It sounds more like the cost for the Flip HD or other pocket 720p camcorders's imaging system cost.

Charbax wrote:
Image processing is getting amazing with OMAP3630 and will be absolutely incredible with OMAP4430 coming out for next year. And prices are coming down so all cheap cameras will have absolutely fantastic compression quality built-in. Look forward to it. With ARM Cortex A9 and the new DSPs that go with it, all camcorders even cheap ones will encode H264 in real-time at about 98% the same quality that you can get from encoding multiple passes using a quad-core intel processor desktop. Yup you heard it right. Near full high profile quality done by the camera itself in real-time, no re-encoding needed.


Again, get your facts straight.

Dedicated, standalone hardware converters used for TV broadcast that costs tens to hundreds of thousands dollars cannot hope to reach the quality you get from a high quality, slow encode from x264 on a computer. Desktop GPUs that cost $100+ alone produces horrifyingly bad video conversion quality. The portable OMAPs will just be even worse.

Again, I am not criticizing the new OMAP chipsets, because it is fundamentally impossible (due to the way video is compressed) to have a hardware based encoder to reach CPU encoded video quality.

I will have to ask you not to post things that you have *no clue* about. Misinformation is bad.

_________________
Gr.9 tech lover


Top
   
 
PostPosted: Sat Aug 21, 2010 4:19 pm 
Offline
Archos Guru
Archos Guru

Joined: Thu Dec 10, 2009 12:10 pm
Posts: 6598
Chengbin wrote:
I will have to ask you not to post things that you have *no clue* about. Misinformation is bad.

I agree that misinformation is bad, but I'm not so sure how successful you'll be asking people not to write whatever they like on their own forums... ;)

Michael


Top
   
 
PostPosted: Sat Aug 21, 2010 4:27 pm 
Offline
Archos Guru
Archos Guru

Joined: Mon Oct 22, 2007 12:59 am
Posts: 2880
Location: Toronto, Ontario
Here is a simplified explanation for people who want to know why video compression done on hardware is so bad compared to CPU: Video encoding done on hardware is done with the GPU. The GPU is consisted of hundreds of "cores". They can be VERY powerful, much more powerful than CPUs for work that is "embarassingly parallel", which means they can be easily split into thousands of threads for the GPU's myriad of cores to process.

For video compression to be embarassingly parallel, each frame of the video has to be split into hundreds of pieces for the hundreds of cores on the GPU to process. This is NOT good for efficient video compression, because a very important part of efficient video compression needs to take into account the whole frame (and the frames beyond). A simple example is if we pretend this string of numbers is a video frame, and suppose I want to represent

111111111111111111111111111111
111111111111111111111111111111
111111111111111111111111111111
111111111111111111111111111111
111111111111111111111111111111

in a compressed method on a CPU. I could say from (1,1) to (5,30) is all 1s. (in video compression, you would put it in a coordinate system)

But if I'm on a hardware encoder, I need to split it into lots of pieces. I split this into 50 pieces, which becomes

11 11 11 11 11 11 11 11 11 11 11 11 11 11 11
11 11 11 11 11 11 11 11 11 11 11 11 11 11 11
11 11 11 11 11 11 11 11 11 11 11 11 11 11 11
11 11 11 11 11 11 11 11 11 11 11 11 11 11 11
11 11 11 11 11 11 11 11 11 11 11 11 11 11 11

Each core on the GPU is only responsible for 2 of the digits. So each core will say from coordinate (1,1) to (1,2) is all 1s 50 times.

Because of that, the GPU's compression efficiency is 50 times less efficient than the CPU. This is just a extreme example used to explain the theory. For real life content it is less extreme, but still very significantly less efficient. Therefore a GPU encoder can never be as good as a CPU encoder.

_________________
Gr.9 tech lover


Top
   
 
PostPosted: Sat Aug 21, 2010 4:49 pm 
Offline
Site Admin
Site Admin

Joined: Sun Nov 27, 2005 2:40 am
Posts: 7052
Location: Copenhagen
Chengbin wrote:
camera's sensor size's relation is directly related to its image quality.


We are talking specifically about low light performance. Low light performance is as much related to sensor size as it is to recorded resolution, the lower the recorded resolution, the higher low light performance. 720p is great for HD video but it's still only 1megapixel in resolution. Thus even cheap small sensors, if made right, and if backed up by new fast DSP can capture good low light videos.

Chengbin wrote:
Charbax wrote:
Image processing is getting amazing with OMAP3630 and will be absolutely incredible with OMAP4430 coming out for next year. And prices are coming down so all cheap cameras will have absolutely fantastic compression quality built-in. Look forward to it. With ARM Cortex A9 and the new DSPs that go with it, all camcorders even cheap ones will encode H264 in real-time at about 98% the same quality that you can get from encoding multiple passes using a quad-core intel processor desktop. Yup you heard it right. Near full high profile quality done by the camera itself in real-time, no re-encoding needed.


Again, get your facts straight.


I get my facts straight from representatives of the semiconductor companies, you?


Chengbin wrote:
Dedicated, standalone hardware converters used for TV broadcast that costs tens to hundreds of thousands dollars cannot hope to reach the quality you get from a high quality, slow encode from x264 on a computer. Desktop GPUs that cost $100+ alone produces horrifyingly bad video conversion quality. The portable OMAPs will just be even worse.


You have no idea how the OMAP compresses video.

Chengbin wrote:
Again, I am not criticizing the new OMAP chipsets, because it is fundamentally impossible (due to the way video is compressed) to have a hardware based encoder to reach CPU encoded video quality.


Very high quality real-time variable bitrate encoding is absolutely doable, as uncompressed source from camera can be stored in large enough cache and then real-time processed by extremely powerful digital signal processor to achieve nearly as good quality as multi-pass high profile encodings on an Intel desktop.

That may not be true if you have very inconsistent scene changes in the overall video and if you are targetting a very specific bitrate average for the whole video. But that is NOT THE CORRECT WAY to get best video quality.

We don't care anymore about reaching exactly 700mb or exactly 4.5GB per video file. CDs and DVDs are technologies of the past. Here with the HD camcorders of the future, AND WITH SCENE RELEASE ENCODES OF THE FUTURE, what we aim for is a consistant quality level throughout any and all videos, no matter what type of action are in the scenes, no matter what type of video, the quality level needs to be the same.

You will find that Youtube encodes all their HD videos based on quality level, not based on standardized bitrates. You will find that a Youtube video with a lot of action automatically is encoded and streamed at higher bitrate than a low action video which has lower bitrate.

When encoding for quality level and not for file size or bitrate average, then the OMAP4 will encode within 2% of any number quad-core Intel whatever desktop and no matter how many slow multi-passes you throw at it. Anyone who deals with any type of video encoding is going to want to have this and it will literally make Intel obsolete for video encoding.


Top
   
 
PostPosted: Sat Aug 21, 2010 5:43 pm 
Offline
Archos Guru
Archos Guru

Joined: Mon Oct 22, 2007 12:59 am
Posts: 2880
Location: Toronto, Ontario
Charbax wrote:
Chengbin wrote:
camera's sensor size's relation is directly related to its image quality.


We are talking specifically about low light performance. Low light performance is as much related to sensor size as it is to recorded resolution, the lower the recorded resolution, the higher low light performance. 720p is great for HD video but it's still only 1megapixel in resolution. Thus even cheap small sensors, if made right, and if backed up by new fast DSP can capture good low light videos.


Do you have any idea how a camera develops an image??? You clearly don't.

A camera needs sufficient lighting to develop an image. This can be done in two ways. Decrease the shutter speed to allow more light in, or increase light sensitivity (otherwise known as ISO). In video recording you obviously can't change shutter speed, as you are recording at a constant frame rate. A camera can only increase ISO. A tiny sensor will collect little light, which will be an image filled with noise, and a large sensor will collect lots of light, which creates a good image. It is a simple concept you fail to understand.
Charbax wrote:
Chengbin wrote:
Charbax wrote:
Image processing is getting amazing with OMAP3630 and will be absolutely incredible with OMAP4430 coming out for next year. And prices are coming down so all cheap cameras will have absolutely fantastic compression quality built-in. Look forward to it. With ARM Cortex A9 and the new DSPs that go with it, all camcorders even cheap ones will encode H264 in real-time at about 98% the same quality that you can get from encoding multiple passes using a quad-core intel processor desktop. Yup you heard it right. Near full high profile quality done by the camera itself in real-time, no re-encoding needed.


Again, get your facts straight.


I get my facts straight from representatives of the semiconductor companies, you?


You certainly did not. Anyone who says that kind of information would have been fired immediately. I've been on Doom9 enough to know how video compression works.

Charbax wrote:
Chengbin wrote:
Dedicated, standalone hardware converters used for TV broadcast that costs tens to hundreds of thousands dollars cannot hope to reach the quality you get from a high quality, slow encode from x264 on a computer. Desktop GPUs that cost $100+ alone produces horrifyingly bad video conversion quality. The portable OMAPs will just be even worse.


You have no idea how the OMAP compresses video.


I don't need to know how OMAP compresses video to know that it is terrible compared to CPU due to limitations of GPU encoding.

Charbax wrote:
Chengbin wrote:
Again, I am not criticizing the new OMAP chipsets, because it is fundamentally impossible (due to the way video is compressed) to have a hardware based encoder to reach CPU encoded video quality.


Very high quality real-time variable bitrate encoding is absolutely doable, as uncompressed source from camera can be stored in large enough cache and then real-time processed by extremely powerful digital signal processor to achieve nearly as good quality as multi-pass high profile encodings on an Intel desktop.

When encoding for quality level and not for file size or bitrate average, then the OMAP4 will encode within 2% of any number quad-core Intel whatever desktop and no matter how many slow multi-passes you throw at it. Anyone who deals with any type of video encoding is going to want to have this and it will literally make Intel obsolete for video encoding.


Yeah sure, if you throw enough bitrate at it, it will eventually be as good as x264. But you need to pile on a magnitude more bitrate, much more than twice or triple the amount of bitrate.

If you're so confident that you're right, and that portable, hardware encoding is so good it rivals CPU encodes, and all the "hype" you're giving it, then go to Doom9, go to the MPEG-4 AVC - H.264 section, and make a thread about that. You'll get flamed worse than the Russia forest fire.

_________________
Gr.9 tech lover


Top
   
 
PostPosted: Sat Aug 21, 2010 6:28 pm 
Offline
Site Admin
Site Admin

Joined: Sun Nov 27, 2005 2:40 am
Posts: 7052
Location: Copenhagen
Resolution matters. I'm not speaking about 12 megapixel low light photography but about 720p video.

Consider a sensor is half the size, but the resolution is also half. Low light sensitivity will be the same on both.


Top
   
 
PostPosted: Sat Aug 21, 2010 6:32 pm 
Offline
Site Admin
Site Admin

Joined: Sun Nov 27, 2005 2:40 am
Posts: 7052
Location: Copenhagen
Chengbin wrote:
I've been on Doom9 enough to know how video compression works.


You go back to hanging out at the old doom9 forums, and read in your high school photography books. Those are not the real world, those don't understand where technology is heading.


Top
   
 
PostPosted: Sat Aug 21, 2010 6:40 pm 
Offline
Site Admin
Site Admin

Joined: Sun Nov 27, 2005 2:40 am
Posts: 7052
Location: Copenhagen
Chengbin wrote:
Yeah sure, if you throw enough bitrate at it, it will eventually be as good as x264. But you need to pile on a magnitude more bitrate, much more than twice or triple the amount of bitrate.

If you're so confident that you're right, and that portable, hardware encoding is so good it rivals CPU encodes, and all the "hype" you're giving it, then go to Doom9, go to the MPEG-4 AVC - H.264 section, and make a thread about that. You'll get flamed worse than the Russia forest fire.



Do a search on your doom9 forum for:

Filesize based encoding vs quality based encoding.


Top
   
 
PostPosted: Sat Aug 21, 2010 6:41 pm 
Offline
Archos Guru
Archos Guru

Joined: Mon Oct 22, 2007 12:59 am
Posts: 2880
Location: Toronto, Ontario
Charbax wrote:
Resolution matters.

Consider a sensor is half the size, but the resolution is also half. Low light sensitivity will be the same on both.


To oversimplify things, yes, that's true. This is why a cameraphone, with its microscopic sized sensors, can't take good low light videos.

Charbax wrote:
You go back to hanging out at the old doom9 forums, and read in your high school photography books. Those are not the real world, those don't understand where technology is heading.


This is the talk from someone who does not know what they're talking about. Anyone who has knowledge will know that the fundamental limitations that is either very difficult/costly or impossible to overcome.

_________________
Gr.9 tech lover


Top
   
 
PostPosted: Sat Aug 21, 2010 6:48 pm 
Offline
Archos Guru
Archos Guru

Joined: Mon Oct 22, 2007 12:59 am
Posts: 2880
Location: Toronto, Ontario
Charbax wrote:
Chengbin wrote:
Yeah sure, if you throw enough bitrate at it, it will eventually be as good as x264. But you need to pile on a magnitude more bitrate, much more than twice or triple the amount of bitrate.

If you're so confident that you're right, and that portable, hardware encoding is so good it rivals CPU encodes, and all the "hype" you're giving it, then go to Doom9, go to the MPEG-4 AVC - H.264 section, and make a thread about that. You'll get flamed worse than the Russia forest fire.



Do a search on your doom9 forum for:

Filesize based encoding vs quality based encoding.


Don't even try to educate me. I know what I'm talking about. Quality based encoding, otherwise known as constant rate factor (CRF), hits a specific quantizer (note I didn't say specific quality, because it doesn't). Yes, you can achieve the level of quality you want on a OMAP, but at what cost? The file would be so big you'd run out of space almost immediately. I have a camera that does 720p recording, and at the highest level of quality, I can only record slightly less than 10 minutes of video on a 8GB card. What use is that?

_________________
Gr.9 tech lover


Top
   
 
Display posts from previous:  Sort by  
Post new topic  Reply to topic  [ 91 posts ]  Go to page Previous  1, 2, 3, 4, 5  Next

All times are UTC + 1 hour


Who is online

Users browsing this forum: Google Adsense [Bot] and 2 guests


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum

Search for:
Jump to:  
Powered by phpBB® Forum Software © phpBB Group
Hosted by Forumatic™