Actually that proves my point. Different people have different ideas for what good key shortcuts should be. And I have no problem at all with anyone using keyboard shortcuts to do anything they want. I just have a problem with the concept that they are some how more intuitive.
Please explain to me how memorizing keyboard shortcuts is more intuitive than a pause or fast forward button? I'm not saying keyboard shortcuts are hard to memorize, and some may even be quite logical, p for play/pause, f for fast forward. But what about full screen? F is already taken, maybe s then for screen, but then what is stop? T?
I can sit down at pretty much any GUI media player and do basic controls without knowing anything at all about the program, keyboard shortcuts on the other hand have a way of changing from program to program, sometimes even between products made by the same company.
IIRC human vision (and hearing for that matter) is logarithmic in nature, not linear like say the output on your computer monitor. So I could see how it could take 64 linear bits per channel to match what the human eye sees (note this is ignoring that humans are more sensitive to certain colors than others). But I'm pulling this completely out of my rectum, so it's probably wrong.
Not quite. Each frame has a high resolution 720x480 8bit green image, and two low resolution 360x240 red and blue images at 8bits. The two low res images are one quarter the size of the high res green so you get 12bits per pixel for each image. Not that this really matters as the human eye really sucks at seeing blue and pretty much sucks for red. You can really drop the resolution on blue and barely notice it even when you know what to look for. If you drop the res on green but keep blue and red high you will notice a big drop in picture quality. MPEG2 compression is all about dropping visual information that won't be noticed.
3.5m assuming even distribution throughout the month, works out to 1.35hits per second. Assume everyone hits it between 8am and 5pm M-F, you're looking at 6.08hits per second. That's not bad. And 170 hours a month is assuming he's working on it full time. I strongly doubt that he's working on it full time.
I think there is only one thing that is certain, we are missing some pretty important information in this story.
The original post was talking about component input for HD signals from a cable box. Component HD is uncompressed analog. You'd need a chip to recompress it to do anything useful with it. You probably could display it, but your system would be useless for just about anything else.
In that case why bother mentioning the operating system at all? The oringal post was giving the impression that tvtime did something on linux that was not possible on windows. That is not true. dScaler was doing it on windows first, and a lot of other projects use dScaler to handle deinterlacing. And with good cause, dScaler is an amazing program, if you have the CPU power to throw at it.
No they won't. No consumer pc on the market can handle recording an HDTV stream. Assuming a 4:2:2 image (12 bits per pixel) you're looking at almost 90MB/s of data. No hard drive can handle a datarate of anything near that. And the only hardware MPEG2 encoders that can handle HDTV are still way above what any consumer can afford. Honestly, I doubt you could even send that stream to your video card over the PCI bus. I think you'd either need the inputs to be right on the video card, or use a special, dedicated high bandwith bus from the capture card to your video card. And even then you would have no chance to process the signal at all, so all of your deinterlacing would have to be done on the video card.
I'm sure someday we'll be able to, but just look how long it took before we could digitally record SDTV. We need a lot more than a capture hard with HDTV capable component inputs.
As much as I hate to rain on your Linux zealot parade, you do realize that TVTime uses the *DLLS* from dScaler, a Windows program, to provide the deinterlacing, right? Just because it's for Windows doesn't mean it's awful.
The older AiW cards 8500 and up, have a very good deinterlacer and scaler built in to them already. The card will definately be able to scale to HDTV resolution, as any bt8x8 tv capture card can do that with any video card made in the last 10 years.
Just a guess, but this will look absolutely great, especially if you have a monitor which can display 1920x1080.
My guess is this card will only work with newer AiW cards. IIRC the ATI video cards contain all of the hardware necisary to decode and display HDTV, they just haven't had any way of tuning it. It really wouldn't surprise me if this card is nothing more than a tuner that sends the encoded MPEG2 stream over to an ATI video card for processing. I doubt it's even possible to send decoded HDTV over the PCI bus. A raw 16bit per pixel 1080i stream would take just under 120MB/s (that's megabytes, not bits). IIRC the max bandwith for 33Mhz PCI is 133MB/s. That doesn't leave much bandwith for doing anything else while watching tv.
No there isn't. There might be some delay that can be measured in milliseconds, but it's nothing a human can detect. I've been doing this for years. Up until about 6 months ago I didn't own a TV and used my computer for TV and my PS2.
My experience was exactly the same. Doing really simple canned reports, easy, 10 minutes tops. Doing anything at all that they didn't expect required jumping through absurd hoops and creating reports that on the design side looked like they should produce complete garbage.
As far as leaving your computer on, with an IP phone you don't need to, but they cost anywhere from $70 to $500 per phone. And there are services such as IConnectHere and VoicePulse, which both will give you a real phone number and connect you to non-voip phone numbers. I think with VoicePulse for something like $25 a month you get a real phone number, voicemail, callerID and all of that, and 600 minutes of US calls per month. IMHO that's a pretty good price.
BTW, I think VoicePulse uses a bunch of Linux boxes running Asterisk to handle the calls.
Or you could use X-Lite, Gnophone, SJphone, or Diax. All of which are completely free. Add about $15 / month for IConnectHere or VoicePulse account with a phone number and you're done.
I'm 99% certain that PCI is 133 Megabytes per second. USB 2.0 is 480 Megabits per second or 60 Megabytes per second. Keep in mind that USB has to travel over the PCI bus to get anywhere in your system so the PCI bus will always (well probably not always) set your max speed.
Since some replies didn't seem to get this either...
RGB is not composite. RGB is transmitted over a minimum of 3 cables RGB, and often at least 5 (RGB and horizontal and vertical sync to allow different resolutions).
Composite is transmitted over 1 wire and has a black and white, red, and blue video signal all crammed into it. It sucks big.
The article is talking about using the three wire RGB output of the PS2, which it does support, as long as your display supports sync-on-green. Apparently the RGB output is disabled for DVDs. At least it still works for component output.
In short, read the full article and then post. Just about everything I just posted can be found on the linked site.
It's most definately the disc. On an NTSC disc the movie will either be encoded at 29.97fps or ~23.9fps at 720x480. On a PAL/SECAM disc it's 25fps at (I believe) 720x540.
The reason for the two different frame rates for NTSC is that the player can do some scan line magic to convert the ~24fps to 29.97fps but retain the better compression of dealing with progressive frames.
The short answer is that there is definately a difference in the discs
This is why credit unions are much better than banks. When my account gets over-drawn they send me a letter saying basically, "your account is negative, fix it now," but written more politely. They don't even charge me the $.37 to mail the letter. ATM deposits are available imediately, regardless of whether I have enought money in my account to cover the deposit. The only ATM fee they charge me is $0.25 for any withdrawl under $20. They also refund up to $4 a month in ATM fees charged by other banks.
It's really nice having a bank that exists to make as much money as possible for the members, rather than the corporation that owns it.
Slashdot Mirror
Actually that proves my point. Different people have different ideas for what good key shortcuts should be. And I have no problem at all with anyone using keyboard shortcuts to do anything they want. I just have a problem with the concept that they are some how more intuitive.
Please explain to me how memorizing keyboard shortcuts is more intuitive than a pause or fast forward button? I'm not saying keyboard shortcuts are hard to memorize, and some may even be quite logical, p for play/pause, f for fast forward. But what about full screen? F is already taken, maybe s then for screen, but then what is stop? T?
I can sit down at pretty much any GUI media player and do basic controls without knowing anything at all about the program, keyboard shortcuts on the other hand have a way of changing from program to program, sometimes even between products made by the same company.
IIRC human vision (and hearing for that matter) is logarithmic in nature, not linear like say the output on your computer monitor. So I could see how it could take 64 linear bits per channel to match what the human eye sees (note this is ignoring that humans are more sensitive to certain colors than others). But I'm pulling this completely out of my rectum, so it's probably wrong.
Not quite. Each frame has a high resolution 720x480 8bit green image, and two low resolution 360x240 red and blue images at 8bits. The two low res images are one quarter the size of the high res green so you get 12bits per pixel for each image. Not that this really matters as the human eye really sucks at seeing blue and pretty much sucks for red. You can really drop the resolution on blue and barely notice it even when you know what to look for. If you drop the res on green but keep blue and red high you will notice a big drop in picture quality. MPEG2 compression is all about dropping visual information that won't be noticed.
http://download.com.com/3000-2168-10267687.html?ta g=lst-0-5
Took me 2 minutes.
3.5m assuming even distribution throughout the month, works out to 1.35hits per second. Assume everyone hits it between 8am and 5pm M-F, you're looking at 6.08hits per second. That's not bad. And 170 hours a month is assuming he's working on it full time. I strongly doubt that he's working on it full time.
I think there is only one thing that is certain, we are missing some pretty important information in this story.
What you posted was in reference to compressed HDTV. The original post was talking about uncompressed HDTV, which is a much higher bitrate.
The original post was talking about component input for HD signals from a cable box. Component HD is uncompressed analog. You'd need a chip to recompress it to do anything useful with it. You probably could display it, but your system would be useless for just about anything else.
I agree 100% with that point. I just felt that he was giving the impression that TVTime had abilities which were not found on Windows.
Sorry about that 4:2:2 is 16 bits per pixel I believe, not 12, 4:2:0 (IIRC) is 12. So my earlier figure of just under 120MB/s would be right.
In that case why bother mentioning the operating system at all? The oringal post was giving the impression that tvtime did something on linux that was not possible on windows. That is not true. dScaler was doing it on windows first, and a lot of other projects use dScaler to handle deinterlacing. And with good cause, dScaler is an amazing program, if you have the CPU power to throw at it.
No they won't. No consumer pc on the market can handle recording an HDTV stream. Assuming a 4:2:2 image (12 bits per pixel) you're looking at almost 90MB/s of data. No hard drive can handle a datarate of anything near that. And the only hardware MPEG2 encoders that can handle HDTV are still way above what any consumer can afford. Honestly, I doubt you could even send that stream to your video card over the PCI bus. I think you'd either need the inputs to be right on the video card, or use a special, dedicated high bandwith bus from the capture card to your video card. And even then you would have no chance to process the signal at all, so all of your deinterlacing would have to be done on the video card.
I'm sure someday we'll be able to, but just look how long it took before we could digitally record SDTV. We need a lot more than a capture hard with HDTV capable component inputs.
As much as I hate to rain on your Linux zealot parade, you do realize that TVTime uses the *DLLS* from dScaler, a Windows program, to provide the deinterlacing, right? Just because it's for Windows doesn't mean it's awful.
The older AiW cards 8500 and up, have a very good deinterlacer and scaler built in to them already. The card will definately be able to scale to HDTV resolution, as any bt8x8 tv capture card can do that with any video card made in the last 10 years.
Just a guess, but this will look absolutely great, especially if you have a monitor which can display 1920x1080.
My guess is this card will only work with newer AiW cards. IIRC the ATI video cards contain all of the hardware necisary to decode and display HDTV, they just haven't had any way of tuning it. It really wouldn't surprise me if this card is nothing more than a tuner that sends the encoded MPEG2 stream over to an ATI video card for processing. I doubt it's even possible to send decoded HDTV over the PCI bus. A raw 16bit per pixel 1080i stream would take just under 120MB/s (that's megabytes, not bits). IIRC the max bandwith for 33Mhz PCI is 133MB/s. That doesn't leave much bandwith for doing anything else while watching tv.
No there isn't. There might be some delay that can be measured in milliseconds, but it's nothing a human can detect. I've been doing this for years. Up until about 6 months ago I didn't own a TV and used my computer for TV and my PS2.
My experience was exactly the same. Doing really simple canned reports, easy, 10 minutes tops. Doing anything at all that they didn't expect required jumping through absurd hoops and creating reports that on the design side looked like they should produce complete garbage.
As far as leaving your computer on, with an IP phone you don't need to, but they cost anywhere from $70 to $500 per phone. And there are services such as IConnectHere and VoicePulse, which both will give you a real phone number and connect you to non-voip phone numbers. I think with VoicePulse for something like $25 a month you get a real phone number, voicemail, callerID and all of that, and 600 minutes of US calls per month. IMHO that's a pretty good price.
BTW, I think VoicePulse uses a bunch of Linux boxes running Asterisk to handle the calls.
Or you could use X-Lite, Gnophone, SJphone, or Diax. All of which are completely free. Add about $15 / month for IConnectHere or VoicePulse account with a phone number and you're done.
I'm 99% certain that PCI is 133 Megabytes per second. USB 2.0 is 480 Megabits per second or 60 Megabytes per second. Keep in mind that USB has to travel over the PCI bus to get anywhere in your system so the PCI bus will always (well probably not always) set your max speed.
Since some replies didn't seem to get this either...
RGB is not composite. RGB is transmitted over a minimum of 3 cables RGB, and often at least 5 (RGB and horizontal and vertical sync to allow different resolutions).
Composite is transmitted over 1 wire and has a black and white, red, and blue video signal all crammed into it. It sucks big.
The article is talking about using the three wire RGB output of the PS2, which it does support, as long as your display supports sync-on-green. Apparently the RGB output is disabled for DVDs. At least it still works for component output.
In short, read the full article and then post. Just about everything I just posted can be found on the linked site.
It's most definately the disc. On an NTSC disc the movie will either be encoded at 29.97fps or ~23.9fps at 720x480. On a PAL/SECAM disc it's 25fps at (I believe) 720x540.
The reason for the two different frame rates for NTSC is that the player can do some scan line magic to convert the ~24fps to 29.97fps but retain the better compression of dealing with progressive frames.
The short answer is that there is definately a difference in the discs
Pennsylvania State Employee's Credit Union. Best thing since sliced bread.
This is why credit unions are much better than banks. When my account gets over-drawn they send me a letter saying basically, "your account is negative, fix it now," but written more politely. They don't even charge me the $.37 to mail the letter. ATM deposits are available imediately, regardless of whether I have enought money in my account to cover the deposit. The only ATM fee they charge me is $0.25 for any withdrawl under $20. They also refund up to $4 a month in ATM fees charged by other banks.
It's really nice having a bank that exists to make as much money as possible for the members, rather than the corporation that owns it.
Uhm...why does DVI = MPEG2 encoding? DVI transmits uncompressed digital RGB.