I'm aware they are different and don't doubt it's bettere unit that the 110. I'm just arguing the point that the LTBC present on the is not as good as the Datavideo. There's a big difference between syncing a few lines as opposed to the whole frame. Follow that link above.....THoff wrote:I think you're being influenced by your experience with the ADVC-110. The ADVC-300 is a completely different unit with different internals, including a fan that is required to cool the added electronics.
Video Editing Business
Moderator: Ken Berry
-
thecoalman
-
THoff
I did, and it's nothing more than conjecture and speculation. People make statements such as "Not real hardware. Not real TBC.", which is obviously false. Few, if any, of the people participating in that discussion have an first-hand experience with the unit.thecoalman wrote:Follow that link above.....
The ADVC-300 has motion detection capabilities, flashable firmware, and a bunch of other features that strongly suggest onboard RAM and the ability to buffer multiple frames. As I said earlier, I will open mine up tonight after I get home and post a picture.
-
thecoalman
That's LS it's hard to get more than a few words out of him much less an entire sentence unless you post it on his private forum.THoff wrote:"Not real hardware. Not real TBC.", which is obviously false. Few, if any, of the people participating in that discussion have an first-hand experience with the unit.
Anyhow I'm looking forward to your pics... how about posting them at VH. Maybe we can both learn a little something.
-
THoff
Sounds like a good plan.
I think all of this controversy stems from some people's focus on Canopus' use of the term LTBC in their literature and product information pages. They use it when discussing jitter removal, for instance.
But on the same page, they also write "ADVC300's 3D Y/C separation separates by cross-referencing each frame with the frames immediately in front and behind, further refining and improving the overall image quality." (indicating that at least three frames are buffered), and "Digital 3D frame synchronizer – Stores incoming signals into a frame buffer to supply a stable, synchronized signal to the DV encoder chip"
So jitter removal accomplishes what a traditional LTBC does, and that may be why they use the term there, but the capabilities of the ADVC-300 go far beyond that.
I think all of this controversy stems from some people's focus on Canopus' use of the term LTBC in their literature and product information pages. They use it when discussing jitter removal, for instance.
But on the same page, they also write "ADVC300's 3D Y/C separation separates by cross-referencing each frame with the frames immediately in front and behind, further refining and improving the overall image quality." (indicating that at least three frames are buffered), and "Digital 3D frame synchronizer – Stores incoming signals into a frame buffer to supply a stable, synchronized signal to the DV encoder chip"
So jitter removal accomplishes what a traditional LTBC does, and that may be why they use the term there, but the capabilities of the ADVC-300 go far beyond that.
-
THoff
OK, never one to be intimidated by warnings that say "DO NOT OPEN THE CASE!" in all-caps on a delicate device, I opened the bloody thing up:

(Click on image to see a larger version)
What you see in the bottom half of the picture are two EM636165TS-7 RAM chips by Etron Tech that are mounted on a small riser module. According to this spiffy secret decoder ring, each provides 64Mbit of 7.0ns SDRAM. The top half of the picture shows the fan that is lacking in the other standalone ADVC models, as well as the LEDs that help configure the unit's filters when used standalone without the Picture Controller utility.
Edit: I was thinking about what I wrote while I was heading to FedEx to pick up UVS 9, and something didn't make sense.
High-bitrate DVDs are encoded at 8000Mbps, and here we have what I thought was 2 * 64Mbit of SDRAM. That clearly wouldn't be enough to buffer three full frames of video for temporal analysis.
So when I got back I double-checked the specs of the memory modules, and they are in fact 64 * 16 Mbit chips. At 8000Mbps / 29.97fps for high-quality NTSC DVD s, you need ~267Mbits per frame, and the Canopus has 2048Mbits of SDRAM at it's disposal. Furthermore, analog TV doesn't encode nearly as much data as high-quality DVD video, so there is plenty of memory to buffer three or more full frames.

(Click on image to see a larger version)
What you see in the bottom half of the picture are two EM636165TS-7 RAM chips by Etron Tech that are mounted on a small riser module. According to this spiffy secret decoder ring, each provides 64Mbit of 7.0ns SDRAM. The top half of the picture shows the fan that is lacking in the other standalone ADVC models, as well as the LEDs that help configure the unit's filters when used standalone without the Picture Controller utility.
Edit: I was thinking about what I wrote while I was heading to FedEx to pick up UVS 9, and something didn't make sense.
High-bitrate DVDs are encoded at 8000Mbps, and here we have what I thought was 2 * 64Mbit of SDRAM. That clearly wouldn't be enough to buffer three full frames of video for temporal analysis.
So when I got back I double-checked the specs of the memory modules, and they are in fact 64 * 16 Mbit chips. At 8000Mbps / 29.97fps for high-quality NTSC DVD s, you need ~267Mbits per frame, and the Canopus has 2048Mbits of SDRAM at it's disposal. Furthermore, analog TV doesn't encode nearly as much data as high-quality DVD video, so there is plenty of memory to buffer three or more full frames.
-
thecoalman
THoff wrote:OK, never one to be intimidated by warnings that say "DO NOT OPEN THE CASE!" in all-caps on a delicate device, I opened the bloody thing up:
Here's the spec sheet for that RAM, at least I think it is. The only varince I see is the addition of VS after the specific models but the top has the base #. Perhaps you can decipher it better than me. It's a google link it's the first return, the damn auto smiley kept screwing up the direct link.
http://www.google.com/search?sourceid=n ... EM636165ts
From the second page.
From what I'm reading you have 2 16Mbit chips for a total of 32 MbitThe EM636165 blah blahblah containing 16 Mbits
Now onto the Datavideo, if what I'm reading is correct it actually has less memory than the ADVC unless I'm looking at the wrong chips. It contains 4 Averlogic chips with 3Mbits of memory each which according to the data sheet is enough to store a full frame of video per chip. (Ibelieve you have your math wrong above and are mixing Mbit with Kbit for the bandwidth required for a frame)
http://www.averlogic.com/password/AL422B051104.pdf
Memory or no memory I still don't think what the ADVC incorporates can perform the full capabilities of a full frame TBC as far as vertical sync is concerned. The reason I'll say this is pretty simple. If in fact it performed the same functions as a full frame TBC they would state it in their literature, this is something you would obviuosly want to advertise. It may buffer up to 3 or 4 frames but not for the purpose of vertical sync but for the other features it has.
What you get is vague statements in the literature but nothing stating specifically that it can correct and maintain vertical sync which is what a full frame TBC provides. Anyhow I'm not going to speculate anymore as to what it does or doesn't do but wish there was some concrete specs or tests as to what it does or doesn't do. I guess it's really useless to go any further without being able to test the differences between the datavideo and the ADVC300 side by side with a anlog source that has vertical sync issues. Ive searched but haven't been able to find any.
-
thecoalman
-
thecoalman


